paperID
stringlengths 36
36
| pwc_id
stringlengths 8
47
| arxiv_id
stringlengths 6
16
⌀ | nips_id
float64 | url_abs
stringlengths 18
329
| url_pdf
stringlengths 18
742
| title
stringlengths 8
325
| abstract
stringlengths 1
7.27k
⌀ | authors
stringlengths 2
7.06k
| published
stringlengths 10
10
⌀ | conference
stringlengths 12
47
⌀ | conference_url_abs
stringlengths 16
198
⌀ | conference_url_pdf
stringlengths 27
199
⌀ | proceeding
stringlengths 6
47
⌀ | taskID
stringlengths 7
1.44k
| areaID
stringclasses 688
values | embedding
stringlengths 9.26k
12.5k
| umap_embedding
stringlengths 29
44
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
8513bdb4-ca94-489a-9466-a15e1ab557cb | group-sparse-matrix-factorization-for | 2104.08928 | null | https://arxiv.org/abs/2104.08928v2 | https://arxiv.org/pdf/2104.08928v2.pdf | Group-Sparse Matrix Factorization for Transfer Learning of Word Embeddings | Unstructured text provides decision-makers with a rich data source in many domains, ranging from product reviews in retailing to nursing notes in healthcare. To leverage this information, words are typically translated into word embeddings -- vectors that encode the semantic relationships between words -- through unsupervised learning algorithms such as matrix factorization. However, learning word embeddings from new domains with limited training data can be challenging, because the meaning/usage may be different in the new domain, e.g., the word "positive" typically has positive sentiment, but often has negative sentiment in medical notes since it may imply that a patient is tested positive for a disease. Intuitively, we expect that only a small number of domain-specific words may have new meanings/usages. We propose an intuitive two-stage estimator that exploits this structure via a group-sparse penalty to efficiently transfer learn domain-specific word embeddings by combining large-scale text corpora (such as Wikipedia) with limited domain-specific text data. We bound the generalization error of our estimator, proving that it can achieve the same accuracy (compared to not transfer learning) with substantially less domain-specific data when only a small number of embeddings are altered between domains. Our results provide the first bounds on group-sparse matrix factorization, which may be of independent interest. We empirically evaluate the effectiveness of our approach compared to state-of-the-art fine-tuning heuristics from natural language processing. | ['Osbert Bastani', 'Hamsa Bastani', 'Xuanyi Zhao', 'Kan Xu'] | 2021-04-18 | null | null | null | null | ['learning-word-embeddings'] | ['methodology'] | [ 2.75134385e-01 2.54670829e-01 -5.73451161e-01 -4.70798403e-01
-6.95138454e-01 -6.72892153e-01 2.32663676e-01 7.32923329e-01
-8.32991540e-01 8.16943467e-01 6.55387163e-01 -3.17495793e-01
-1.17403544e-01 -8.79639804e-01 -6.57165647e-01 -5.20816267e-01
-8.08009803e-02 7.08433986e-01 -2.82312125e-01 -2.58441150e-01
-3.86556871e-02 1.34488687e-01 -1.10752010e+00 2.08917812e-01
8.57302368e-01 7.60753453e-01 2.70137072e-01 4.28448379e-01
-4.54196513e-01 2.32880861e-01 -3.82592350e-01 -5.28214216e-01
3.63222539e-01 5.80227673e-02 -7.06064343e-01 2.39805773e-01
3.54129523e-02 -1.78581119e-01 -4.75287884e-02 1.19922709e+00
3.16479236e-01 3.28596532e-01 8.25884819e-01 -9.10119593e-01
-1.02236879e+00 5.96493900e-01 -4.96920049e-01 1.39544040e-01
3.27407241e-01 -1.27512356e-03 1.50620484e+00 -9.62180853e-01
8.41648459e-01 1.14995027e+00 5.89698553e-01 6.08892679e-01
-1.28213465e+00 -5.84828258e-01 4.02199715e-01 -1.48511618e-01
-9.88341630e-01 -7.12742284e-02 6.49625361e-01 -5.65941334e-01
1.08652532e+00 -1.08831629e-01 5.46586752e-01 1.23067427e+00
3.30554217e-01 6.94879532e-01 6.03737414e-01 -2.61197060e-01
4.70178694e-01 5.58499336e-01 2.74522364e-01 5.94642222e-01
7.60925472e-01 -3.66631687e-01 -5.31325579e-01 -4.62138742e-01
5.20230949e-01 6.14010453e-01 -1.46482483e-01 -5.95710814e-01
-1.21764839e+00 1.21870220e+00 2.03736469e-01 2.39955992e-01
-4.91755694e-01 -1.14607751e-01 6.12752855e-01 4.13068086e-01
6.77182853e-01 8.95996332e-01 -1.14449036e+00 -3.42045158e-01
-4.75462586e-01 2.68232584e-01 9.02798414e-01 9.57935691e-01
1.02364719e+00 -3.11581343e-01 2.72666156e-01 1.00864398e+00
2.28755355e-01 3.79196048e-01 9.84039009e-01 -4.20286268e-01
5.57734907e-01 6.36432946e-01 2.31375083e-01 -9.94598269e-01
-3.28288436e-01 -1.13843568e-01 -7.87320256e-01 -4.15244877e-01
2.12712780e-01 -5.48758268e-01 -8.45660985e-01 1.95626438e+00
3.53158355e-01 4.64166552e-02 2.46180281e-01 6.86375082e-01
4.60504830e-01 5.29372156e-01 -4.70415642e-03 -2.32619867e-01
1.81908453e+00 -7.28762984e-01 -8.31728756e-01 -6.40367985e-01
9.49972332e-01 -5.28431535e-01 1.26798511e+00 4.06673938e-01
-6.51386559e-01 -1.52852133e-01 -8.73263955e-01 -2.04812631e-01
-6.03534937e-01 -3.94353718e-01 8.31927955e-01 6.14513755e-01
-4.38865274e-01 4.15330976e-01 -8.57326031e-01 -6.43115699e-01
4.32061583e-01 2.20421240e-01 -6.35471046e-01 -3.22130412e-01
-1.36985087e+00 7.59450376e-01 3.09136629e-01 -5.49910009e-01
-2.84415215e-01 -1.20882583e+00 -1.22152495e+00 1.76256612e-01
4.85739797e-01 -9.03554797e-01 1.17952442e+00 -8.79734874e-01
-1.10014462e+00 7.18966365e-01 -4.03214186e-01 -4.61620063e-01
7.69844204e-02 -4.15759474e-01 -4.61296916e-01 -3.09943617e-03
3.41303498e-01 3.96865904e-01 8.66915584e-01 -8.76185477e-01
-6.40542686e-01 -4.83051807e-01 2.91082054e-01 1.37561068e-01
-1.03039575e+00 -2.62325495e-01 -2.95335203e-01 -8.92581940e-01
-2.76558787e-01 -7.73460627e-01 -6.12528801e-01 3.08583498e-01
-6.89022616e-02 -2.64310598e-01 4.04546946e-01 -3.75466704e-01
1.35487008e+00 -2.27462959e+00 1.00268781e-01 2.58768588e-01
4.47581053e-01 1.14248827e-01 -2.68471867e-01 5.51944613e-01
-8.01033750e-02 2.16995239e-01 -4.97154266e-01 -2.41437823e-01
-1.22845788e-02 6.85332596e-01 -4.01925951e-01 2.10488752e-01
6.01886272e-01 7.45388627e-01 -1.35045576e+00 -2.66594559e-01
-6.69896826e-02 2.11020604e-01 -9.75221217e-01 2.53366623e-02
-1.66760564e-01 -1.85132816e-01 -6.89528525e-01 3.41075242e-01
5.22777915e-01 -4.71766979e-01 5.79731703e-01 -2.13692799e-01
3.95078272e-01 4.16672021e-01 -1.13324153e+00 1.86541009e+00
-8.80832553e-01 4.14419234e-01 -5.62102497e-02 -1.29398167e+00
6.17536366e-01 1.18042007e-01 5.75945377e-01 -2.75032938e-01
7.35236108e-02 2.56212298e-02 1.47732738e-02 -6.24431789e-01
5.63859284e-01 -6.76055312e-01 -3.89631212e-01 5.72152615e-01
2.46277764e-01 -1.42280579e-01 2.25127965e-01 2.50587106e-01
1.37378705e+00 -5.79860270e-01 7.48240054e-01 -3.76585573e-01
5.00517450e-02 3.80844437e-02 7.68393040e-01 4.46187794e-01
-6.04712069e-02 4.12323803e-01 7.04486668e-01 -3.65540624e-01
-9.23669696e-01 -1.06961977e+00 -3.01620156e-01 1.18958414e+00
-3.46618220e-02 -7.03629017e-01 -3.20090532e-01 -9.03928757e-01
5.29702663e-01 6.18290484e-01 -1.00181496e+00 -3.65547091e-01
-1.81101799e-01 -7.67983556e-01 2.05417484e-01 8.25804472e-01
-3.72993082e-01 -7.00354278e-01 -3.68515849e-01 3.92066866e-01
2.94855803e-01 -1.15103078e+00 -6.81423604e-01 6.16403103e-01
-9.62442756e-01 -1.10271800e+00 -5.99951923e-01 -8.35892260e-01
9.97688949e-01 2.05155402e-01 1.20758271e+00 -3.16980541e-01
-2.40333855e-01 5.75271964e-01 -5.27637184e-01 -5.63007772e-01
-2.16076508e-01 1.49004713e-01 5.47257125e-01 3.09581961e-02
8.05063069e-01 -3.48853946e-01 -5.94993651e-01 -6.73370287e-02
-1.41343498e+00 -3.88602525e-01 5.09420812e-01 1.17530704e+00
5.54591060e-01 -1.42241508e-01 5.86483181e-01 -1.61251628e+00
1.07082343e+00 -8.71875703e-01 -1.85073987e-01 6.23043478e-02
-7.01419532e-01 5.06134450e-01 9.27411139e-01 -8.07809234e-01
-5.80295563e-01 -1.30987108e-01 4.46594693e-03 -3.45363915e-01
3.41841318e-02 7.78757453e-01 -1.65970139e-02 5.76213360e-01
7.53049314e-01 2.44818919e-04 7.77424797e-02 -4.21306968e-01
7.18338072e-01 7.63156056e-01 5.78199849e-02 -7.22187102e-01
8.32830489e-01 4.43988979e-01 -5.06688118e-01 -8.08356345e-01
-1.01203132e+00 -8.73289883e-01 -2.93653607e-01 5.72473705e-01
9.26682293e-01 -1.10043538e+00 -2.34451458e-01 -1.99156299e-01
-1.05921340e+00 -1.34582460e-01 -6.81150019e-01 6.68315411e-01
-2.21048906e-01 3.53767186e-01 -4.28713500e-01 -3.34926099e-01
-2.13309690e-01 -9.57095265e-01 9.70704854e-01 -1.15405411e-01
-7.70137370e-01 -1.40141904e+00 3.46319646e-01 1.76302735e-02
1.14638701e-01 -1.56423301e-01 1.27525556e+00 -1.09427023e+00
2.27322459e-01 -1.90397263e-01 -1.24053702e-01 6.35114372e-01
8.36251378e-01 -4.51973200e-01 -4.88390237e-01 -3.21560740e-01
-7.52642527e-02 -2.78127491e-01 8.43591154e-01 1.69960141e-01
1.03477263e+00 -4.72590744e-01 -2.87183553e-01 4.45394903e-01
1.31535125e+00 -1.24847464e-01 1.81053370e-01 1.31901681e-01
6.83418930e-01 4.30547625e-01 7.04589844e-01 8.91967714e-01
2.90752381e-01 2.70367771e-01 -9.64557976e-02 7.92815015e-02
4.04598445e-01 -2.01428160e-01 3.79573464e-01 1.01974642e+00
3.81771624e-01 -1.06871165e-01 -8.64269257e-01 9.43658829e-01
-1.62341440e+00 -4.74066913e-01 4.36121762e-01 2.04741740e+00
1.35003603e+00 1.67041361e-01 -1.05770096e-01 -1.31697774e-01
4.02511030e-01 1.70927167e-01 -8.25142443e-01 -6.76088333e-01
2.78935343e-01 6.61068439e-01 6.72658026e-01 4.24191445e-01
-9.45568204e-01 8.09155107e-01 5.88706684e+00 5.97191811e-01
-9.34376359e-01 1.84593871e-01 2.72464931e-01 -2.55621582e-01
-8.22220981e-01 -3.31242263e-01 -6.31899714e-01 3.94815624e-01
8.81765127e-01 -5.63728213e-01 1.59214452e-01 1.07392335e+00
2.34273057e-02 2.64582604e-01 -1.42170846e+00 1.12928271e+00
2.15183392e-01 -1.31353700e+00 2.79936194e-01 6.96716085e-02
9.06961143e-01 -7.59335607e-03 2.10169345e-01 3.71198952e-01
6.09477460e-01 -7.18539119e-01 1.09540850e-01 4.53351587e-02
8.79587054e-01 -6.80824697e-01 7.53985345e-01 2.52879918e-01
-1.05056810e+00 -1.84623212e-01 -7.18926668e-01 -9.27723870e-02
1.47333890e-01 1.06187332e+00 -1.02110648e+00 2.94829637e-01
4.56785291e-01 9.74210083e-01 -9.91674438e-02 3.39601934e-01
-2.45095074e-01 5.32695830e-01 -2.92486101e-01 -1.54495344e-01
3.31076294e-01 -1.97057977e-01 2.38580361e-01 1.28794050e+00
4.00059462e-01 1.45464644e-01 1.19226687e-01 5.73657632e-01
-6.16506279e-01 3.00742835e-01 -9.02849019e-01 -5.57224691e-01
4.10513729e-01 1.12983203e+00 -3.66950274e-01 -5.55419624e-01
-8.10765982e-01 1.05238402e+00 3.60108972e-01 2.99651980e-01
-5.51347554e-01 -4.68211472e-01 1.54789400e+00 2.47886896e-01
4.80929971e-01 -1.56615719e-01 -3.07745486e-01 -1.44784820e+00
1.27537608e-01 -9.66712952e-01 4.47443426e-01 -3.69616091e-01
-1.95542026e+00 3.34282786e-01 -2.02967897e-01 -1.28080976e+00
-2.79253423e-01 -9.70771372e-01 -4.32454422e-02 5.02902567e-01
-1.56088865e+00 -5.18116891e-01 2.35415101e-01 5.28838813e-01
6.86931849e-01 -2.90425569e-01 1.07011092e+00 2.59450883e-01
-2.40800366e-01 8.06840301e-01 4.43802327e-01 3.76602322e-01
1.20274520e+00 -1.38348043e+00 3.29312295e-01 5.30287743e-01
2.32418284e-01 1.12816632e+00 5.83356142e-01 -5.64660788e-01
-1.64459932e+00 -1.23497152e+00 1.07555914e+00 -5.84772468e-01
1.15550876e+00 -5.28163135e-01 -9.66579556e-01 7.37076819e-01
-7.79205188e-02 1.87983781e-01 1.37654710e+00 6.22557700e-01
-7.34515548e-01 -1.26574069e-01 -1.21485102e+00 7.25211084e-01
9.67574179e-01 -8.09827626e-01 -1.02660942e+00 4.83732492e-01
1.07478738e+00 -6.55410141e-02 -9.79671896e-01 -6.05767779e-03
5.00960231e-01 -1.84972197e-01 8.24397326e-01 -1.28305578e+00
7.26752222e-01 3.67574841e-02 -3.36561978e-01 -1.72830188e+00
-2.31181100e-01 -3.48313719e-01 -1.57190412e-01 8.00657630e-01
6.57533526e-01 -9.44595695e-01 6.67034268e-01 8.96646202e-01
1.24685310e-01 -9.23634708e-01 -7.30601013e-01 -7.66869545e-01
3.03780109e-01 -7.00348854e-01 6.03301048e-01 1.20111430e+00
3.78669262e-01 4.88259494e-01 -1.76459417e-01 8.63599256e-02
1.06574997e-01 -8.66728947e-02 5.51486850e-01 -1.30758226e+00
-3.89619470e-01 -6.66549504e-02 -6.12825036e-01 -1.23634374e+00
3.77228707e-01 -9.76052761e-01 -4.96850833e-02 -1.49802876e+00
2.64777899e-01 -3.21361691e-01 -6.82001114e-01 5.83593369e-01
-4.34461594e-01 -1.43066958e-01 -1.01620935e-01 -3.26037407e-01
-4.46066201e-01 4.80400711e-01 1.03584576e+00 -3.95053208e-01
-2.06083372e-01 -3.16540271e-01 -1.32136154e+00 7.58819163e-01
4.89168078e-01 -7.52357721e-01 -6.83702111e-01 -5.59465766e-01
6.59258544e-01 -3.20306182e-01 -1.74534693e-01 -3.45474601e-01
-1.56644545e-02 -2.62469202e-01 4.08470519e-02 1.80201277e-01
5.86714000e-02 -9.89105105e-01 -5.02507031e-01 3.32239389e-01
-5.74029803e-01 1.21942364e-01 1.66364491e-01 9.89737153e-01
-2.90065020e-01 -8.07773545e-02 4.65009749e-01 -8.43731463e-02
-7.41109550e-01 4.62836444e-01 -4.87766802e-01 5.09489715e-01
7.99331725e-01 -8.13420564e-02 6.42326996e-02 -1.85620621e-01
-6.60944521e-01 2.10566133e-01 4.48176205e-01 6.96887791e-01
6.60795331e-01 -1.31475067e+00 -4.86486346e-01 3.21752936e-01
5.88419974e-01 1.31382406e-01 1.16034701e-01 4.55559075e-01
-5.79312705e-02 9.50363204e-02 7.22155124e-02 -2.64089674e-01
-1.01418197e+00 7.77575850e-01 -3.56281757e-01 -5.88200450e-01
-4.88035411e-01 8.67114186e-01 5.77971160e-01 -7.03190207e-01
-1.12433299e-01 -8.35813820e-01 5.03760464e-02 5.25110662e-01
6.34490430e-01 -1.12438187e-01 6.04069419e-02 -1.20793834e-01
-5.35581529e-01 4.47611481e-01 -4.64631826e-01 -6.81626936e-03
1.64182532e+00 1.44891679e-01 -4.31231968e-03 6.18121505e-01
1.53173959e+00 1.72638953e-01 -8.50118279e-01 -5.82993209e-01
-3.02340779e-02 -5.54134130e-01 -1.52020574e-01 -4.32004362e-01
-8.66954029e-01 8.91717255e-01 2.97174037e-01 -9.12000388e-02
9.55101728e-01 1.29322708e-01 1.13851976e+00 8.14676940e-01
3.97747934e-01 -1.26939940e+00 9.38862711e-02 5.54273844e-01
4.84129786e-01 -1.26005948e+00 -8.62959772e-03 -2.13370606e-01
-9.72613394e-01 9.35205877e-01 1.82137311e-01 -2.47801214e-01
1.03925288e+00 3.84302586e-01 2.71341596e-02 -1.79171041e-02
-9.05939281e-01 -7.16873333e-02 1.69292882e-01 6.63283288e-01
4.70488340e-01 3.42531800e-01 -2.68783838e-01 1.17711878e+00
-1.85170993e-01 -3.80851850e-02 4.37897325e-01 8.74757588e-01
-2.39803508e-01 -1.37211859e+00 -4.19678241e-02 9.17769909e-01
-4.78424937e-01 -4.46433932e-01 -1.40816018e-01 4.54010427e-01
2.94202287e-02 7.10207403e-01 1.56119496e-01 -2.96124101e-01
4.09533769e-01 2.29662850e-01 2.85939068e-01 -1.33226013e+00
-2.79539168e-01 -3.28600734e-01 2.31391713e-02 -4.56245244e-01
-2.07772955e-01 -6.04753733e-01 -1.37387729e+00 -8.25703666e-02
-1.10024936e-03 2.05936015e-01 3.68640691e-01 1.04446781e+00
5.45123160e-01 5.12572646e-01 4.65894759e-01 -3.68754387e-01
-7.87878215e-01 -7.67658949e-01 -8.36196005e-01 8.23401153e-01
3.75320494e-01 -6.37278318e-01 -3.71635944e-01 2.21356481e-01] | [10.386984825134277, 8.658019065856934] |
7ec13636-e40d-4089-9f95-5ab368b408ef | one-step-multi-view-clustering-with-diverse | 2306.05437 | null | https://arxiv.org/abs/2306.05437v2 | https://arxiv.org/pdf/2306.05437v2.pdf | One-step Multi-view Clustering with Diverse Representation | Multi-view clustering has attracted broad attention due to its capacity to utilize consistent and complementary information among views. Although tremendous progress has been made recently, most existing methods undergo high complexity, preventing them from being applied to large-scale tasks. Multi-view clustering via matrix factorization is a representative to address this issue. However, most of them map the data matrices into a fixed dimension, limiting the model's expressiveness. Moreover, a range of methods suffers from a two-step process, i.e., multimodal learning and the subsequent $k$-means, inevitably causing a sub-optimal clustering result. In light of this, we propose a one-step multi-view clustering with diverse representation method, which incorporates multi-view learning and $k$-means into a unified framework. Specifically, we first project original data matrices into various latent spaces to attain comprehensive information and auto-weight them in a self-supervised manner. Then we directly use the information matrices under diverse dimensions to obtain consensus discrete clustering labels. The unified work of representation learning and clustering boosts the quality of the final results. Furthermore, we develop an efficient optimization algorithm with proven convergence to solve the resultant problem. Comprehensive experiments on various datasets demonstrate the promising clustering performance of our proposed method. | ['Li Shen', 'Tianjiao Wan', 'Yi Wen', 'Siwei Wang', 'Xinwang Liu', 'En Zhu', 'Jiyuan Liu', 'Xinhang Wan'] | 2023-06-08 | null | null | null | null | ['multi-view-learning'] | ['computer-vision'] | [-1.36701405e-01 -6.17515326e-01 -3.74977112e-01 -3.60180438e-01
-8.20376039e-01 -6.18011057e-01 3.37013602e-01 -2.44153947e-01
1.61528122e-02 1.62613690e-01 3.82398903e-01 2.08040327e-01
-3.59618813e-01 -5.23087382e-01 -2.14040503e-01 -1.11797857e+00
4.18526411e-01 3.67253006e-01 -2.06145555e-01 1.38576522e-01
1.77462220e-01 4.43317518e-02 -1.40749896e+00 3.00223231e-01
1.03042889e+00 7.25552142e-01 2.91668415e-01 5.98745905e-02
-1.02256536e-01 4.85656798e-01 -1.51386827e-01 -3.41151774e-01
1.45313293e-01 -4.64684844e-01 -6.00655854e-01 7.51439869e-01
6.66717952e-03 5.68759143e-02 -1.20508358e-01 1.18288028e+00
4.87452865e-01 1.59998298e-01 6.43383980e-01 -1.24794006e+00
-7.90471435e-01 4.54214066e-01 -1.16910529e+00 -2.89990216e-01
3.21126699e-01 -3.32434863e-01 1.22452641e+00 -1.15395999e+00
5.12801707e-01 1.26846933e+00 4.14473116e-01 2.60685921e-01
-1.38239181e+00 -7.16420293e-01 4.43938285e-01 2.34651029e-01
-1.65306473e+00 -3.19086015e-01 1.16379642e+00 -6.24029100e-01
3.23712617e-01 1.31003350e-01 5.26736736e-01 7.04986453e-01
-2.97191292e-01 9.28052664e-01 1.16321015e+00 -1.06191367e-01
1.26338512e-01 2.48627365e-01 8.63465220e-02 5.41343749e-01
1.49983451e-01 -6.10305309e-01 -3.41266721e-01 -2.26621523e-01
4.12318945e-01 5.47811151e-01 -2.97393203e-01 -8.88169050e-01
-1.33345675e+00 8.42634320e-01 2.32278168e-01 4.26689863e-01
-3.28749448e-01 -3.37515593e-01 4.83781010e-01 -4.03153896e-03
3.59675944e-01 -7.88637698e-02 -5.26676215e-02 1.52907908e-01
-8.31707537e-01 -1.62907481e-01 3.22747260e-01 9.73588526e-01
9.94844556e-01 -9.28866267e-02 3.22836697e-01 1.18900442e+00
5.23669720e-01 5.61210155e-01 4.71855342e-01 -9.53558445e-01
6.32156849e-01 1.10211992e+00 -1.30555391e-01 -1.60696435e+00
-2.95662761e-01 -3.43995214e-01 -1.42666435e+00 -2.80355752e-01
1.11160129e-01 5.41847572e-02 -4.25523609e-01 1.65509319e+00
4.43643332e-01 -1.90208897e-01 1.27726153e-01 9.62786376e-01
4.65842605e-01 7.81658351e-01 -2.05060795e-01 -6.45760417e-01
1.23852611e+00 -9.34975266e-01 -8.24254751e-01 6.84933513e-02
4.07926410e-01 -7.50337720e-01 7.96309471e-01 6.03639126e-01
-9.27099466e-01 -6.55904531e-01 -9.89518642e-01 3.10123086e-01
-8.73979777e-02 3.25879604e-01 6.30273461e-01 5.21290720e-01
-7.89241672e-01 -5.05250134e-03 -7.71642268e-01 -2.79634595e-01
2.77186275e-01 3.42642725e-01 -5.51703036e-01 -4.68121588e-01
-7.90779114e-01 3.07537198e-01 5.08239269e-01 2.14941710e-01
-4.55266118e-01 -2.34583214e-01 -6.58325195e-01 -1.44388765e-01
5.44724703e-01 -7.20025659e-01 5.50731540e-01 -6.23125315e-01
-1.24898565e+00 5.51962018e-01 -4.39558953e-01 3.82053912e-01
2.44737923e-01 -5.32590076e-02 -6.10937774e-01 3.82443964e-01
3.74894261e-01 2.20804334e-01 9.32010174e-01 -1.79828429e+00
-7.22793221e-01 -6.78523898e-01 -9.25631151e-02 4.82259452e-01
-7.73050308e-01 -2.24954635e-01 -1.04258585e+00 -4.76393133e-01
8.27613115e-01 -1.01248229e+00 -3.07603359e-01 -5.58045149e-01
-4.80336070e-01 -2.58962482e-01 7.86951780e-01 -4.16513264e-01
1.61889243e+00 -2.27765679e+00 7.64875233e-01 3.17762345e-01
4.75475580e-01 -3.45604420e-02 1.54250681e-01 7.20975041e-01
-2.49146610e-01 3.95546965e-02 -2.67164379e-01 -3.45340014e-01
-9.71489772e-02 1.30892709e-01 -1.10668264e-01 6.14729404e-01
-3.25304687e-01 5.07435024e-01 -8.74025106e-01 -7.06162035e-01
3.29438418e-01 3.79512787e-01 -5.80802441e-01 2.21486062e-01
2.68561125e-01 5.07695913e-01 -6.52536869e-01 6.43522382e-01
7.70203948e-01 -6.42455578e-01 4.71750081e-01 -6.13678217e-01
-1.20509557e-01 -5.40427148e-01 -1.56444240e+00 1.91569936e+00
-2.69669354e-01 -7.31826574e-02 2.98947394e-01 -1.38822079e+00
8.12770009e-01 3.90464544e-01 1.04188967e+00 -2.47997686e-01
-1.84118263e-02 2.20630825e-01 -1.61572203e-01 -4.76740003e-01
2.97604352e-01 -4.04557109e-01 -1.06517069e-01 6.64638996e-01
-1.83865473e-01 2.43037194e-01 3.02227080e-01 4.65088844e-01
4.57895637e-01 -2.65768878e-02 3.45187485e-01 -9.27164406e-02
9.24262524e-01 -9.97354239e-02 8.04278493e-01 1.58233345e-01
-3.67115065e-02 6.97756350e-01 2.94784039e-01 -2.37399071e-01
-8.36786509e-01 -1.03027987e+00 1.47727057e-02 9.68825102e-01
3.76464069e-01 -6.50782406e-01 -6.52660728e-01 -6.80826545e-01
-2.08048612e-01 2.39232421e-01 -4.15614128e-01 -1.73181012e-01
-4.03476089e-01 -1.02970791e+00 -2.91688927e-02 3.39843482e-01
4.19859082e-01 -5.13797998e-01 3.68101038e-02 1.82163388e-01
-4.95730728e-01 -9.25245523e-01 -5.37548006e-01 -1.06601171e-01
-8.60619307e-01 -1.17207265e+00 -7.15109348e-01 -8.52538466e-01
9.56237435e-01 1.12699342e+00 6.63741946e-01 -1.07137725e-01
1.67398036e-01 7.05420613e-01 -5.05338550e-01 3.41009796e-01
-1.18924558e-01 4.77201901e-02 4.14138883e-01 5.46819687e-01
3.76632154e-01 -8.30973446e-01 -6.98048711e-01 5.99499464e-01
-1.10574603e+00 1.26863897e-01 7.48345554e-01 9.36390996e-01
8.00699472e-01 4.58063424e-01 6.26372397e-01 -8.66238892e-01
4.62152839e-01 -6.20746255e-01 -3.28700602e-01 4.71493304e-01
-7.11129963e-01 -1.82524562e-01 9.66961026e-01 -1.52116001e-01
-9.97741699e-01 2.34194577e-01 2.75208324e-01 -8.13435674e-01
-3.71495262e-03 6.94606900e-01 -6.79633975e-01 2.83964455e-01
8.14474970e-02 6.08406425e-01 9.44352001e-02 -4.07375664e-01
7.08123982e-01 6.65902019e-01 2.73205698e-01 -6.17820501e-01
1.05951321e+00 7.36717939e-01 -2.48888284e-01 -7.11658239e-01
-9.71222520e-01 -8.11578274e-01 -9.06828701e-01 -3.00284147e-01
9.62563932e-01 -1.17471814e+00 -7.60259986e-01 3.56570214e-01
-7.04628348e-01 6.31298840e-01 2.40904331e-01 6.33828282e-01
-3.86009812e-01 8.63577843e-01 -3.64558995e-01 -6.30741477e-01
-1.47636011e-02 -1.20869792e+00 7.59476960e-01 8.27976167e-02
1.16781376e-01 -9.70310986e-01 1.60797358e-01 7.84589171e-01
-9.40457284e-02 4.25161049e-02 9.41149771e-01 -4.44715500e-01
-4.39721107e-01 -1.31306142e-01 -3.81712258e-01 2.89283037e-01
5.63197613e-01 -1.01362288e-01 -7.24158645e-01 -5.68123639e-01
2.27716178e-01 -2.55324095e-01 6.97292089e-01 3.31197798e-01
1.07522821e+00 4.94919345e-02 -4.92755771e-01 5.60822845e-01
1.49078178e+00 3.28971207e-01 9.60469246e-03 1.78282887e-01
1.16019773e+00 7.25520849e-01 5.70427597e-01 6.67267919e-01
7.57922709e-01 5.09726048e-01 2.71411777e-01 -6.24717493e-03
3.86638045e-01 -1.34799182e-01 1.76787212e-01 1.80103505e+00
-1.52129069e-01 -2.67603435e-02 -8.75465274e-01 4.96813655e-01
-2.08539772e+00 -1.19721484e+00 -4.78313640e-02 1.99537826e+00
3.89983714e-01 -2.18944937e-01 2.07346201e-01 2.12536052e-01
1.04076552e+00 4.45197403e-01 -5.83497047e-01 2.42636189e-01
-8.69206265e-02 -7.00122297e-01 -1.56747531e-02 1.07389897e-01
-1.09979928e+00 6.33585095e-01 5.44977713e+00 8.93440843e-01
-8.18499386e-01 5.54635711e-02 5.16820192e-01 2.81082019e-02
-6.16616607e-01 -1.33408168e-02 -6.00932360e-01 4.43067014e-01
3.69396180e-01 -1.72758937e-01 5.81482828e-01 7.43706465e-01
3.00123632e-01 1.17199793e-01 -7.48064101e-01 1.53704929e+00
3.51747245e-01 -9.24211383e-01 3.51520985e-01 1.62488207e-01
9.50025201e-01 -4.15268838e-01 1.30762622e-01 1.73507527e-01
1.86724588e-01 -5.32932162e-01 3.73387724e-01 3.42489839e-01
6.41543925e-01 -1.09827793e+00 3.49242091e-01 5.08214355e-01
-1.60227573e+00 -9.37299505e-02 -5.22060513e-01 1.64838940e-01
2.54308730e-01 7.49485791e-01 -1.99441850e-01 1.18907094e+00
5.81897914e-01 1.30611944e+00 -4.73244905e-01 5.30272603e-01
1.58015490e-01 2.76532888e-01 -1.11410387e-01 3.93185705e-01
2.94457048e-01 -9.66845334e-01 2.91583866e-01 7.81818748e-01
3.86092156e-01 2.44534135e-01 6.55502617e-01 5.43263435e-01
1.05369180e-01 4.20899868e-01 -6.20159388e-01 -1.10689849e-01
5.65084159e-01 1.45975935e+00 -8.88459921e-01 -3.41285110e-01
-8.85420918e-01 9.15701926e-01 4.61894780e-01 3.69094938e-01
-6.31145000e-01 -5.13605028e-02 4.33713645e-01 -2.15281785e-01
3.57879639e-01 -3.07029516e-01 -1.93057150e-01 -1.68779171e+00
1.42684609e-01 -1.00414586e+00 5.87393165e-01 -3.38961720e-01
-1.54202998e+00 4.69576478e-01 -1.09354101e-01 -1.82372296e+00
-5.96687682e-02 -1.39241993e-01 -1.92269772e-01 5.15203655e-01
-1.25765944e+00 -1.18884289e+00 -2.18947634e-01 9.48597431e-01
5.55492997e-01 -3.05321455e-01 6.83996379e-01 5.67024171e-01
-1.01798022e+00 3.57946873e-01 7.16884494e-01 9.16948915e-03
9.00484562e-01 -1.23697305e+00 -3.29420805e-01 8.62385154e-01
9.43856686e-02 9.57951009e-01 2.20127165e-01 -4.85903740e-01
-1.84150469e+00 -9.83643055e-01 3.94739866e-01 -2.53669560e-01
7.56506681e-01 -2.50923038e-01 -8.60381663e-01 5.67949057e-01
1.39718682e-01 -1.20287254e-01 1.24832785e+00 3.31375867e-01
-4.85730410e-01 -2.97086567e-01 -6.26374125e-01 5.57144105e-01
8.69900167e-01 -6.06242120e-01 -5.19968152e-01 2.47175172e-01
3.83996576e-01 5.35323322e-02 -1.14516938e+00 2.82109737e-01
4.81529415e-01 -1.26574016e+00 1.00165212e+00 -3.24777424e-01
5.18830121e-01 -7.71605790e-01 -5.51749408e-01 -1.36866879e+00
-6.74136043e-01 -3.30584109e-01 2.64931712e-02 1.62127805e+00
1.69609219e-01 -4.68886763e-01 6.70711696e-01 3.86920989e-01
-3.41010801e-02 -7.28055835e-01 -6.59796834e-01 -4.40063715e-01
-1.82985246e-01 -2.48281524e-01 3.46892506e-01 1.25856078e+00
2.11734965e-01 6.12109900e-01 -7.53295064e-01 2.60831207e-01
9.55219030e-01 8.67853999e-01 8.29041958e-01 -1.08979392e+00
-1.03404664e-01 -3.08862686e-01 -1.56999439e-01 -1.08691645e+00
1.46211267e-01 -1.02072155e+00 -1.54992148e-01 -1.59856462e+00
7.57258534e-01 -4.35136855e-01 -3.96923929e-01 1.18991770e-01
-4.71208930e-01 1.11301377e-01 2.28166744e-01 8.69635582e-01
-1.11057639e+00 7.19684243e-01 1.27566445e+00 -1.72471460e-02
-1.95746809e-01 -7.78088421e-02 -1.03744459e+00 8.51002932e-01
4.75934654e-01 -2.62510240e-01 -6.94300413e-01 -3.75593722e-01
1.24807373e-01 2.56331503e-01 -2.35054091e-01 -8.76554191e-01
3.80542248e-01 -2.62714773e-01 3.88218701e-01 -8.40665817e-01
3.37478876e-01 -1.10395706e+00 2.87792146e-01 2.17268169e-01
-1.04272459e-02 1.22882143e-01 -3.90227497e-01 9.99816358e-01
-6.65514350e-01 1.44241020e-01 7.19064474e-01 -2.41142124e-01
-6.63738608e-01 4.40500617e-01 -2.01370493e-01 1.60205476e-02
1.01999056e+00 -2.63998330e-01 2.69825794e-02 -3.59468609e-01
-7.85502076e-01 5.51635921e-01 6.66013479e-01 3.86043847e-01
4.69258904e-01 -1.69789219e+00 -4.99813646e-01 1.72017485e-01
2.75256157e-01 1.02695458e-01 7.40678489e-01 1.01305807e+00
-1.65141858e-02 3.27355951e-01 -2.86632665e-02 -8.88819695e-01
-1.06101084e+00 9.15362358e-01 -1.88662142e-01 -2.44168997e-01
-5.11553168e-01 3.72718275e-01 5.36327720e-01 -4.86636668e-01
-5.32590002e-02 4.13034379e-01 -4.77465630e-01 5.78330994e-01
4.40056026e-01 4.61367399e-01 -3.38371038e-01 -9.65597868e-01
-3.61885995e-01 1.15964556e+00 -1.53075725e-01 6.10922761e-02
1.49073696e+00 -7.28085756e-01 -3.91914189e-01 6.95290446e-01
1.41485012e+00 9.42281187e-02 -1.01671016e+00 -4.38889354e-01
-9.99990329e-02 -4.60034549e-01 -1.45523161e-01 -1.08316079e-01
-1.28033793e+00 8.76912653e-01 1.80213109e-01 2.91815519e-01
1.37217987e+00 -1.06447354e-01 7.34598696e-01 4.06290174e-01
2.86423564e-01 -1.08219671e+00 1.51560009e-01 1.43902898e-01
4.02274013e-01 -1.41082072e+00 3.04170907e-01 -5.56336761e-01
-8.39266896e-01 1.00875318e+00 5.77320993e-01 -9.70968790e-03
6.53795242e-01 -2.61693895e-01 2.07931817e-01 -3.41566980e-01
-4.87160474e-01 5.01898117e-03 2.93469280e-01 3.07668149e-01
3.06597829e-01 3.13941427e-02 -2.11513638e-01 6.79147363e-01
2.07254201e-01 -3.98615301e-01 2.76758432e-01 6.74533725e-01
-3.57534885e-01 -1.18596363e+00 -5.15303135e-01 3.35447013e-01
-2.40015209e-01 2.31723443e-01 -2.10006028e-01 5.50768495e-01
-1.95198432e-02 1.14167619e+00 -4.59998161e-01 -6.43972039e-01
2.40169317e-01 2.20416579e-03 5.46314605e-02 -6.17300987e-01
-2.28934377e-01 7.10541189e-01 -4.53211963e-01 -4.46729064e-01
-1.01078939e+00 -9.42143679e-01 -9.77284551e-01 -2.18981326e-01
-3.60837102e-01 4.96771485e-01 3.26953292e-01 8.68704796e-01
4.18291211e-01 2.89189339e-01 1.19317508e+00 -7.68697500e-01
-4.14888859e-01 -5.62728882e-01 -9.26521599e-01 7.12563217e-01
-1.11975446e-01 -7.43589163e-01 -4.30860311e-01 2.27161705e-01] | [8.298360824584961, 4.626137733459473] |
ddc5fb40-f8bb-41a2-8430-a07a3b48922b | flexible-framework-for-audio-reconstruction | 2004.11162 | null | http://arxiv.org/abs/2004.11162v2 | http://arxiv.org/pdf/2004.11162v2.pdf | Flexible framework for audio reconstruction | The paper presents a unified, flexible framework for the tasks of audio
inpainting, declipping, and dequantization. The concept is further extended to
cover analogous degradation models in a transformed domain, e.g. quantization
of the signal's time-frequency coefficients. The task of reconstructing an
audio signal from degraded observations in two different domains is formulated
as an inverse problem, and several algorithmic solutions are developed. The
viability of the presented concept is demonstrated on an example where audio
reconstruction from partial and quantized observations of both the time-domain
signal and its time-frequency coefficients is carried out. | [] | 2020-07-29 | null | null | null | null | ['audio-inpainting'] | ['audio'] | [ 9.03475463e-01 1.23554930e-01 1.74192682e-01 -1.44173384e-01
-1.20470440e+00 -4.83723193e-01 2.01561809e-01 -2.53849000e-01
-1.82944477e-01 9.16230142e-01 4.83688742e-01 4.31092493e-02
-2.28124559e-01 -8.15418512e-02 -2.63999075e-01 -8.24095845e-01
-2.27529332e-01 -1.59966037e-01 -1.30020067e-01 -1.28217936e-01
1.73115641e-01 4.70599860e-01 -1.68956256e+00 1.72280237e-01
6.24347448e-01 1.03637576e+00 1.02026761e-01 1.10703433e+00
3.45815301e-01 4.29363310e-01 -1.03273058e+00 -1.89411193e-01
1.92394033e-01 -6.80961490e-01 -5.66714942e-01 5.70242107e-01
2.22577512e-01 -4.47179258e-01 -4.66492623e-01 1.17603350e+00
4.62304235e-01 1.48135394e-01 5.92991590e-01 -1.08994329e+00
-2.98211783e-01 2.13372976e-01 -1.92305133e-01 2.66262680e-01
8.15428674e-01 -3.64767939e-01 7.45734692e-01 -7.55468786e-01
5.19451082e-01 1.15769362e+00 8.00987482e-01 1.79273546e-01
-1.52686882e+00 -2.63045132e-01 -3.06902736e-01 3.16948116e-01
-1.36562908e+00 -7.97656596e-01 1.09895480e+00 -3.60523164e-01
6.80357993e-01 4.80161309e-01 4.29524720e-01 8.35305989e-01
4.23383743e-01 5.58084071e-01 9.78190780e-01 -7.18945086e-01
1.93572044e-01 -5.46010286e-02 -1.86348990e-01 1.61824644e-01
-2.96568602e-01 4.03238028e-01 -8.72773528e-01 -3.33851874e-01
6.03779733e-01 -3.52873772e-01 -6.88910425e-01 -5.57357166e-03
-9.42200780e-01 5.13128161e-01 -2.08919242e-01 3.35744083e-01
-5.36762536e-01 2.86233462e-02 5.93256950e-01 1.04105353e+00
8.50755274e-01 2.28367552e-01 -1.14794344e-01 -3.76201570e-01
-1.46270597e+00 4.11982000e-01 7.06311524e-01 9.23964083e-01
3.72193635e-01 6.50768578e-01 4.19744365e-02 7.60900915e-01
6.63166121e-02 3.29274535e-01 6.29508495e-01 -1.28431392e+00
4.81674314e-01 -6.27064884e-01 4.41289932e-01 -7.50704229e-01
-4.07220237e-02 -3.43182981e-01 -7.70503104e-01 3.66771579e-01
2.71560550e-01 -1.16505019e-01 -6.34890318e-01 1.72832906e+00
2.35362500e-01 2.65582800e-01 1.57115996e-01 8.52375865e-01
1.70068234e-01 9.96655643e-01 -3.70582283e-01 -9.19767141e-01
1.13111651e+00 -5.01176417e-01 -1.57296395e+00 4.55673486e-02
-2.29798883e-01 -1.20269847e+00 6.49135232e-01 1.07905579e+00
-1.52769375e+00 -9.19466197e-01 -1.29511428e+00 -4.05220717e-01
-4.01357710e-02 1.49527341e-02 -2.07425848e-01 4.52240795e-01
-1.15484214e+00 9.09717739e-01 -4.81200457e-01 6.81129992e-02
-2.50034988e-01 2.20755532e-01 -3.63912553e-01 1.88407466e-01
-1.40234196e+00 7.31939018e-01 2.55568147e-01 1.63969725e-01
-9.54466760e-01 -5.46933174e-01 -7.77977407e-01 -2.62409318e-02
-7.49432892e-02 -3.89601648e-01 1.64826930e+00 -1.03220129e+00
-1.66450822e+00 5.59009850e-01 -3.69690180e-01 -5.14998496e-01
4.70726401e-01 -3.13801050e-01 -1.02491307e+00 5.30934274e-01
1.76628888e-01 -1.47310924e-02 2.11099863e+00 -8.56752992e-01
-6.10440016e-01 -2.89297998e-01 -4.41218823e-01 1.85889006e-01
-1.54737890e-01 -3.26873548e-02 2.80896500e-02 -1.42377615e+00
4.25335318e-01 -5.09422123e-01 9.99971628e-02 8.36610347e-02
-1.88867822e-01 4.94561881e-01 9.59630609e-01 -1.45134294e+00
1.49168944e+00 -2.82121110e+00 6.31538510e-01 -1.29289571e-02
-5.29684052e-02 -1.58065274e-01 -1.45917565e-01 8.44149649e-01
-5.70716321e-01 -4.30322647e-01 -4.96885896e-01 -6.42604768e-01
2.79291929e-03 2.11827576e-01 -9.51586962e-01 6.39048100e-01
1.88495591e-01 1.43145040e-01 -8.69523346e-01 -3.14727634e-01
4.12145145e-02 6.68714404e-01 -2.52626985e-01 2.68986315e-01
1.85121566e-01 5.86011827e-01 2.35776324e-02 5.96462250e-01
5.85520983e-01 6.32561743e-01 -5.34260720e-02 -2.40765512e-01
-3.49962711e-01 4.08773869e-01 -1.34770238e+00 1.76249468e+00
-5.16083598e-01 9.67902601e-01 9.19814289e-01 -9.17485774e-01
9.12668824e-01 1.11095393e+00 5.23571253e-01 -3.95937175e-01
-1.13266014e-01 3.81259412e-01 -3.68943900e-01 -7.31263697e-01
8.83270144e-01 -7.31476724e-01 -8.54138583e-02 4.36834186e-01
2.52000362e-01 -7.58632243e-01 9.74543691e-02 -7.54622817e-02
8.72568250e-01 7.79678747e-02 5.07476270e-01 1.47507032e-02
4.97439116e-01 -2.78840989e-01 2.22338498e-01 4.06266004e-01
-1.82262421e-01 7.49045372e-01 3.44652563e-01 2.03648299e-01
-1.32585609e+00 -1.16225958e+00 -2.08507687e-01 9.82126474e-01
-3.38743567e-01 -3.98069799e-01 -8.84220243e-01 4.73507904e-02
-2.51160972e-02 7.19948232e-01 -3.10382724e-01 -1.91499174e-01
-4.45053726e-01 -1.93544269e-01 6.63304448e-01 2.82537062e-02
1.48567885e-01 -7.01324344e-01 -5.15161812e-01 8.40845466e-01
-5.54921508e-01 -9.76198554e-01 -7.11264074e-01 3.97805542e-01
-1.35937679e+00 -5.71332097e-01 -8.55108857e-01 -7.37943172e-01
2.26391807e-01 2.39142478e-01 7.43943870e-01 -5.13530850e-01
-1.40644833e-01 6.36763811e-01 -4.25855488e-01 -4.12324488e-01
-6.62892938e-01 -4.62941200e-01 2.75262386e-01 4.48073715e-01
-2.57994413e-01 -1.17733324e+00 -8.02181289e-02 6.63986057e-02
-1.40941584e+00 -3.68417144e-01 3.13274443e-01 1.01262558e+00
6.34147882e-01 4.98387307e-01 7.27891147e-01 -3.62966508e-01
1.12989235e+00 -2.58282453e-01 -2.89677173e-01 -3.44773471e-01
-3.47886503e-01 -1.19983047e-01 7.47550249e-01 -7.29220510e-01
-1.15508878e+00 1.43076986e-01 -1.85665995e-01 -5.48439205e-01
-5.59996068e-02 4.93410140e-01 -2.52769798e-01 1.24002276e-02
7.52627730e-01 5.59814095e-01 2.29675069e-01 -9.50639069e-01
4.20292556e-01 8.18875551e-01 1.28468966e+00 -3.66559714e-01
8.24786186e-01 4.32093382e-01 -9.91111994e-02 -1.25774658e+00
-4.79064912e-01 -5.67173541e-01 -5.62308729e-01 -1.76511303e-01
3.57532769e-01 -9.92153764e-01 -1.90789420e-02 3.71918797e-01
-1.39830267e+00 1.38080791e-01 -8.43634665e-01 5.48633635e-01
-1.10837114e+00 7.31655657e-01 -6.63479269e-01 -9.38067555e-01
-1.70856565e-02 -9.13671017e-01 1.29093850e+00 -3.52420509e-01
-5.37320554e-01 -9.15178895e-01 2.73333132e-01 -1.32752493e-01
7.15122893e-02 2.47432277e-01 7.81992018e-01 -5.69286309e-02
-9.89562422e-02 -4.74647224e-01 4.20652300e-01 8.16434443e-01
3.12321752e-01 -1.17402174e-01 -1.20211947e+00 -5.08300662e-01
9.73041713e-01 -8.31146091e-02 5.25442183e-01 4.12084669e-01
7.07878232e-01 -5.05230188e-01 2.30309412e-01 5.95773160e-01
1.27511322e+00 1.84000790e-01 5.94319403e-01 -9.80155542e-03
-1.10259920e-01 5.89431763e-01 8.86303723e-01 8.06289554e-01
-4.92486149e-01 8.84062350e-01 3.28534812e-01 2.31116012e-01
-2.85174757e-01 -1.47141859e-01 7.46055603e-01 1.15868282e+00
8.51732492e-02 -7.71209374e-02 -2.10256666e-01 7.77817845e-01
-1.37901151e+00 -1.18021989e+00 6.07650056e-02 2.11614919e+00
8.93314779e-01 -1.05348855e-01 3.42269093e-02 1.13250279e+00
8.25218976e-01 3.27295512e-01 -5.34982622e-01 -7.00644314e-01
-7.21664503e-02 5.58048129e-01 1.25512362e-01 8.72357309e-01
-9.65780437e-01 1.65189832e-01 7.68526459e+00 9.25391674e-01
-9.79209006e-01 2.04089269e-01 -2.30620559e-02 5.96064236e-03
8.27614069e-02 -1.70630902e-01 -6.83466867e-02 4.62160528e-01
1.17549932e+00 -5.23310900e-01 7.32773602e-01 3.44078004e-01
6.61673486e-01 -6.61556423e-02 -1.00655568e+00 1.14023197e+00
2.43094526e-02 -6.90004289e-01 -2.38548741e-01 6.31349441e-03
4.36215371e-01 -7.20126152e-01 4.27720726e-01 -1.27236042e-02
-5.25451303e-01 -5.61845899e-01 1.25374925e+00 5.28879344e-01
1.21762407e+00 -6.31053865e-01 1.94136441e-01 4.12538886e-01
-1.06263936e+00 -2.49668971e-01 -1.66334957e-01 -5.22507846e-01
5.12501776e-01 8.95425916e-01 -8.49839866e-01 7.79780567e-01
4.06254947e-01 8.14973712e-01 3.80662642e-02 1.12289548e+00
-2.77549922e-01 6.51797056e-01 -2.19555259e-01 7.38965631e-01
1.26955081e-02 -4.37259048e-01 1.09502602e+00 1.29495013e+00
7.79442191e-01 2.57496864e-01 -2.89519668e-01 4.15601701e-01
3.39154929e-01 -2.11005822e-01 -4.83639002e-01 6.07147142e-02
3.22046310e-01 7.55664647e-01 -1.60251915e-01 -3.69893521e-01
-8.05553123e-02 1.14097488e+00 -4.71775442e-01 6.09715462e-01
-7.37218022e-01 -8.98678541e-01 6.43021107e-01 6.84614852e-02
4.22457874e-01 -1.40094131e-01 -2.52296440e-02 -9.90465641e-01
1.65213659e-01 -1.00616229e+00 2.19530329e-01 -1.01472664e+00
-9.53236759e-01 6.23566806e-01 1.59314439e-01 -1.96032870e+00
-8.62789452e-01 -1.09779894e-01 -3.39786798e-01 1.12564242e+00
-1.36546648e+00 -4.41606671e-01 1.55196682e-01 8.35629463e-01
7.07305908e-01 -6.21539094e-02 8.53650928e-01 4.33965534e-01
9.80301201e-03 7.35071972e-02 6.44677460e-01 -5.81638753e-01
6.85141206e-01 -1.17597032e+00 1.20648310e-01 9.59229767e-01
1.86469387e-02 1.58389002e-01 1.61456251e+00 -3.63568783e-01
-1.11620331e+00 -1.00144601e+00 1.03581524e+00 2.31313244e-01
7.41669059e-01 -4.45739366e-03 -8.69590759e-01 5.87417006e-01
1.72840029e-01 -2.88239598e-01 6.14452839e-01 -5.95288992e-01
-1.68362945e-01 -1.67418733e-01 -1.28071821e+00 1.31381840e-01
5.52919328e-01 -1.17936051e+00 -1.07293522e+00 2.60933876e-01
5.53278446e-01 -5.53571045e-01 -8.91058326e-01 -9.10322145e-02
4.41749275e-01 -1.00952721e+00 1.02625453e+00 -2.72171259e-01
1.79498777e-01 -4.46120292e-01 -3.47326487e-01 -1.37823582e+00
-2.31656313e-01 -1.40994811e+00 -3.69625390e-01 1.04870057e+00
-2.03498289e-01 -3.71041477e-01 3.99360418e-01 -1.29487932e-01
-3.22070032e-01 -3.10062133e-02 -1.55296278e+00 -8.78482759e-01
-5.83452702e-01 -6.19569719e-01 2.37220675e-01 5.53211987e-01
2.70987272e-01 1.48872584e-01 -8.77158284e-01 3.96065116e-01
6.90499961e-01 -1.51951546e-02 3.36707592e-01 -1.05468953e+00
-5.76132834e-01 5.87599538e-02 -4.91123796e-01 -1.22854054e+00
4.35740734e-03 -3.06842625e-01 3.14785689e-01 -8.97845328e-01
-7.13332117e-01 3.83448780e-01 -1.00493945e-01 -8.60683993e-02
4.17574346e-01 3.20448488e-01 3.80515426e-01 2.20519617e-01
7.64403343e-02 6.88659310e-01 1.11351359e+00 -2.98424661e-01
-1.51360020e-01 4.67196643e-01 -2.01540127e-01 5.34430563e-01
5.37354290e-01 -6.70554578e-01 -6.63643301e-01 -1.93912998e-01
-2.82046974e-01 9.21534181e-01 4.49156404e-01 -1.20980561e+00
4.11384972e-03 2.71072567e-01 7.94422999e-02 -5.07763088e-01
1.07681799e+00 -1.06670046e+00 8.02244663e-01 3.85774046e-01
-3.94130588e-01 1.97739333e-01 1.48840919e-01 8.10268462e-01
-1.04948914e+00 -4.34940457e-01 8.39280486e-01 1.72411680e-01
-4.31884378e-01 -1.97733343e-01 -6.56770170e-01 -2.84873784e-01
5.91991484e-01 -5.21676004e-01 3.55743408e-01 -9.02788222e-01
-1.55780780e+00 -6.54918671e-01 2.06301242e-01 -4.22019511e-02
1.01566648e+00 -1.36077702e+00 -7.39168882e-01 4.40525055e-01
-3.47844571e-01 -4.43022668e-01 3.86535525e-01 7.88208544e-01
-3.71508241e-01 4.04255420e-01 -3.62901002e-01 -3.57679278e-01
-1.52622509e+00 6.30301654e-01 1.48574263e-01 -1.76785663e-01
-5.42400599e-01 6.31462574e-01 -1.75146282e-01 5.03590465e-01
4.32872653e-01 -5.64310312e-01 6.16084784e-02 3.15130770e-01
7.80846417e-01 6.07276976e-01 2.59320438e-01 -7.99270034e-01
1.84674338e-01 5.08843839e-01 2.87758172e-01 -9.52203989e-01
1.12368417e+00 -5.58384657e-01 -2.90153295e-01 8.56900930e-01
1.36169910e+00 1.62191018e-01 -1.35197175e+00 -2.77394682e-01
-1.50925905e-01 -5.77138960e-01 1.88123778e-01 -3.93778622e-01
-5.63003778e-01 9.93392885e-01 4.88289326e-01 6.95739865e-01
1.90067303e+00 -5.77234924e-01 6.18027508e-01 8.29284731e-03
3.12494129e-01 -1.08222532e+00 -1.52267702e-02 2.92585790e-01
1.37570310e+00 -2.72135675e-01 2.09081084e-01 -4.18607235e-01
-1.70697510e-01 1.38480878e+00 -4.55271333e-01 -1.74086899e-01
5.97300172e-01 2.20688492e-01 4.86880504e-02 4.50219840e-01
-6.81736171e-01 1.57277286e-01 1.33258417e-01 9.34018373e-01
3.46151292e-01 -1.54194415e-01 -3.00584793e-01 3.32456052e-01
-3.65288287e-01 5.71082160e-03 8.95322442e-01 1.00564086e+00
-4.39245075e-01 -1.15508485e+00 -1.30774462e+00 6.83227777e-02
-5.78881085e-01 1.24398388e-01 -9.09347832e-02 5.15621185e-01
-1.90166645e-02 1.10326612e+00 -2.83422053e-01 -2.45553315e-01
5.05720615e-01 3.89005899e-01 4.75437820e-01 -4.24524069e-01
-3.21974427e-01 7.94573665e-01 -6.70786873e-02 -4.11488831e-01
-4.69457626e-01 -9.07987237e-01 -6.69600666e-01 4.78855893e-03
-4.00681645e-02 4.29096460e-01 6.49207175e-01 4.59048718e-01
1.04054129e-02 7.40021288e-01 8.51443410e-01 -1.29198301e+00
-1.02308834e+00 -9.70035434e-01 -1.27202499e+00 3.58321041e-01
1.25348771e+00 -3.89505178e-01 -5.48156857e-01 6.99940205e-01] | [15.441868782043457, 5.677613735198975] |
f73e43d5-2a56-4f99-a1e2-8faae73f33e9 | boundary-preserved-deep-denoising-of-the | 1904.06329 | null | http://arxiv.org/abs/1904.06329v2 | http://arxiv.org/pdf/1904.06329v2.pdf | Boundary-Preserved Deep Denoising of the Stochastic Resonance Enhanced Multiphoton Images | As the rapid growth of high-speed and deep-tissue imaging in biomedical
research, it is urgent to find a robust and effective denoising method to
retain morphological features for further texture analysis and segmentation.
Conventional denoising filters and models can easily suppress perturbative
noises in high contrast images. However, for low photon budget multi-photon
images, high detector gain will not only boost signals, but also bring huge
background noises. In such stochastic resonance regime of imaging,
sub-threshold signals may be detectable with the help of noises. Therefore, a
denoising filter that can smartly remove noises without sacrificing the
important cellular features such as cell boundaries is highly desired. In this
paper, we propose a convolutional neural network based autoencoder method,
Fully Convolutional Deep Denoising Autoencoder (DDAE), to improve the quality
of Three-Photon Fluorescence (3PF) and Third Harmonic Generation (THG)
microscopy images. The average of the acquired 200 images of a given location
served as the low-noise answer for DDAE training. Compared with other widely
used denoising methods, our DDAE model shows better signal-to-noise ratio (26.6
and 29.9 for 3PF and THG, respectively), structure similarity (0.86 and 0.87
for 3PF and THG, respectively), and preservation of nuclear or cellular
boundaries. | ['Tzu-Ming Liu', 'Tzung-Dau Wang', 'Lun-Zhang Guo', 'Sheng-Yong Niu', 'Yue Li', 'Yu Tsao'] | 2019-04-12 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 1.93102121e-01 -5.84687769e-01 6.23911619e-01 -9.68465358e-02
-6.58486903e-01 -1.72404975e-01 1.10505760e-01 2.98790067e-01
-6.72040582e-01 8.41847539e-01 -1.39804900e-01 1.28109768e-01
-1.14582963e-01 -7.96886921e-01 -5.09944022e-01 -1.55169332e+00
3.60116363e-01 5.70471101e-02 4.70434546e-01 -1.96118072e-01
1.49560064e-01 6.68445349e-01 -1.35624063e+00 1.45130262e-01
9.16146815e-01 1.03663528e+00 4.34463590e-01 5.11830747e-01
-6.82174340e-02 4.76395547e-01 -5.78988731e-01 -1.87219277e-01
-4.72204052e-02 -4.15180206e-01 -4.48286504e-01 -1.51962280e-01
-9.25143659e-02 -4.20945287e-01 -4.41530287e-01 1.34649432e+00
1.07354641e+00 1.50952861e-01 6.20687485e-01 -5.12903631e-01
-7.35081732e-01 1.84014857e-01 -8.56121421e-01 6.23836458e-01
-2.66701698e-01 1.69557035e-01 1.22566886e-01 -6.66810215e-01
6.53863549e-01 9.10175085e-01 5.90157807e-01 7.00482428e-01
-1.33621001e+00 -4.93931621e-01 -3.46361279e-01 2.63637543e-01
-1.14774454e+00 -3.18187922e-01 6.87285721e-01 -3.30170691e-01
6.36679888e-01 1.12881854e-01 6.45457029e-01 8.56535196e-01
8.75629544e-01 3.00217062e-01 1.41695464e+00 -1.98230624e-01
5.32321297e-02 -2.98786759e-01 -2.03706268e-02 5.74629307e-01
4.05091763e-01 -1.19592240e-02 -1.57876849e-01 -2.17709281e-02
9.25919116e-01 2.11421356e-01 -5.93850136e-01 4.78116482e-01
-1.18505192e+00 3.67611885e-01 1.57974049e-01 7.60083020e-01
-5.79864383e-01 1.20940497e-02 3.34663868e-01 7.26246387e-02
2.77877182e-01 2.99682189e-02 -1.74502775e-01 3.57353091e-02
-6.96987927e-01 -1.36655435e-01 2.54916310e-01 1.67295039e-01
7.03103065e-01 1.66082367e-01 -1.83976337e-01 7.00000405e-01
-6.19022064e-02 5.28810501e-01 5.54508686e-01 -1.30499446e+00
-6.23586595e-01 3.70947778e-01 -4.65496369e-02 -1.15842938e+00
-3.15263122e-01 -4.77528244e-01 -1.43920827e+00 1.49409786e-01
5.41640937e-01 1.81654580e-02 -9.93138373e-01 1.47645438e+00
4.91689295e-01 -3.95641476e-03 6.24969155e-02 1.12173724e+00
1.12814784e+00 7.75294781e-01 -5.74902911e-03 -7.03593254e-01
1.43640316e+00 -3.26547652e-01 -1.09169018e+00 1.50412291e-01
1.82689399e-01 -9.19359624e-01 6.92159057e-01 3.69543523e-01
-1.02883792e+00 -3.98751467e-01 -8.27910244e-01 -1.63876116e-01
2.17499286e-02 3.45626920e-02 5.10235429e-01 2.29292437e-01
-9.09322619e-01 6.84620500e-01 -1.06058645e+00 -9.53977033e-02
6.03769541e-01 3.39841038e-01 -5.29514790e-01 -3.03054184e-01
-9.62785780e-01 5.57645977e-01 -7.19266906e-02 2.46088341e-01
-9.27453816e-01 -6.90467238e-01 -2.70361483e-01 2.44848221e-03
-2.09881440e-01 -5.22568226e-01 8.29713821e-01 -6.86008692e-01
-1.56662834e+00 9.50177491e-01 -2.65146852e-01 -7.36251101e-02
6.84245452e-02 2.68497437e-01 -3.89439255e-01 7.13001013e-01
8.49128515e-02 4.55069751e-01 6.80369020e-01 -1.25926626e+00
-3.14863265e-01 -5.34891367e-01 -3.41695964e-01 -1.41838863e-01
-3.16576481e-01 1.71440005e-01 -8.47034156e-02 -4.44163859e-01
4.06947285e-01 -4.08080727e-01 -3.08589756e-01 6.00897111e-02
-5.56942150e-02 -6.62717298e-02 8.75918210e-01 -8.52684319e-01
7.43849635e-01 -2.27455282e+00 -6.40296116e-02 -1.39049008e-01
3.56426954e-01 3.75104636e-01 -4.66050133e-02 6.44647479e-02
-8.36879686e-02 -3.20304967e-02 -2.58355051e-01 1.99036598e-01
-6.32317245e-01 2.99132556e-01 4.81476963e-01 8.20929050e-01
6.09931089e-02 6.80671394e-01 -6.82662785e-01 -5.65661192e-01
1.28719985e-01 9.04040217e-01 -2.18298972e-01 2.42619589e-02
2.51810730e-01 8.59531939e-01 -4.58690643e-01 7.33381867e-01
1.02757490e+00 -2.27214172e-01 -1.21534891e-01 -7.22817481e-01
-2.85691917e-01 -5.43476880e-01 -7.91074276e-01 1.15708458e+00
8.86685625e-02 5.69080114e-01 6.61023438e-01 -1.24749756e+00
9.61520374e-01 2.91537732e-01 7.85080254e-01 -1.08932114e+00
4.86461133e-01 4.60556000e-01 8.17089677e-02 -7.32578754e-01
-1.25928566e-01 -7.18572497e-01 5.00905216e-01 1.05024502e-01
7.98290037e-03 1.09275185e-01 1.98627993e-01 -6.43864423e-02
9.05965388e-01 -5.10933042e-01 -1.63934231e-01 -5.12399495e-01
6.15154326e-01 -2.78482080e-01 8.47254813e-01 4.53653455e-01
-4.91514921e-01 7.86004901e-01 4.43437696e-01 -5.45731902e-01
-9.78427649e-01 -6.09359145e-01 -3.48970711e-01 4.80192453e-01
3.58648151e-01 3.69430125e-01 -9.50270057e-01 -4.11246829e-02
-2.57905155e-01 -1.15479864e-02 -2.68303186e-01 -1.38005808e-01
-6.03862941e-01 -1.28655136e+00 5.09611845e-01 1.43737167e-01
7.65944541e-01 -9.81815219e-01 -5.51887453e-01 4.26088274e-01
-4.70286548e-01 -1.08595908e+00 -1.47524610e-01 3.50214213e-01
-8.58784020e-01 -9.68953729e-01 -1.04058456e+00 -9.81113553e-01
7.45637894e-01 3.99139553e-01 7.48460710e-01 3.87397408e-01
-4.37388301e-01 2.55398266e-02 -1.35072798e-01 -2.55160123e-01
-3.28733087e-01 -5.77567339e-01 -3.18401232e-02 2.27361731e-02
2.51817614e-01 -7.38717854e-01 -1.03939426e+00 1.71638176e-01
-1.12116671e+00 -2.95225352e-01 6.78804100e-01 1.27909553e+00
1.21246314e+00 6.02337837e-01 4.80228484e-01 -4.47793871e-01
3.91212881e-01 7.01216012e-02 -5.17940938e-01 -3.56623009e-02
-3.46835405e-01 -3.44862372e-01 7.69528091e-01 -4.91397321e-01
-1.20944583e+00 -2.06799194e-01 -4.93918717e-01 -2.85154611e-01
-3.51534635e-01 5.11772454e-01 -4.66924384e-02 -5.59271872e-01
6.50137007e-01 5.54963171e-01 4.01918530e-01 -2.72883803e-01
-4.03259665e-01 5.97423792e-01 6.73216045e-01 -3.80212814e-01
2.67296046e-01 9.46970105e-01 2.58193672e-01 -1.04796076e+00
-5.25786638e-01 -4.07600999e-01 -3.93145867e-02 -3.08570445e-01
8.46081913e-01 -8.00278425e-01 -1.10080576e+00 1.04819500e+00
-1.15410352e+00 -2.13098526e-01 -2.36805137e-02 6.83913767e-01
-1.34487346e-01 6.73622489e-01 -1.21855772e+00 -5.99270165e-01
-6.12440586e-01 -1.23858893e+00 7.51353741e-01 6.95683300e-01
4.91869360e-01 -7.87885666e-01 -1.54377773e-01 3.74888480e-01
5.65534651e-01 6.13662183e-01 1.04950058e+00 -1.26210764e-01
-5.23154497e-01 -1.85117498e-01 -3.91863227e-01 5.80494702e-01
1.12697057e-01 3.56205814e-02 -8.65824103e-01 -2.40431949e-01
6.55150890e-01 -1.84917927e-01 9.72515047e-01 1.12793219e+00
1.38190365e+00 2.20519248e-02 -2.28786260e-01 7.37249970e-01
1.57316065e+00 3.93000305e-01 1.10450590e+00 2.71059394e-01
4.12312686e-01 4.53573287e-01 3.14149261e-01 3.88412774e-01
-5.07618003e-02 2.22153831e-02 4.01284039e-01 -5.95383286e-01
-1.63326666e-01 3.94665271e-01 8.06060731e-02 1.01041162e+00
-2.72118241e-01 -3.18488419e-01 -6.12477303e-01 7.26326704e-01
-1.35246682e+00 -8.74476135e-01 -6.56222224e-01 1.73993301e+00
1.01496291e+00 -1.65468678e-01 -2.29842886e-01 2.21740812e-01
9.08689260e-01 -2.04174936e-01 -5.25621712e-01 1.25522614e-02
-6.17464542e-01 4.09059882e-01 3.83473307e-01 2.34063104e-01
-8.17680418e-01 4.29377168e-01 5.59119225e+00 9.97024596e-01
-1.21848583e+00 1.11290596e-01 9.32240129e-01 2.28360891e-01
-3.25919658e-01 -2.59521008e-01 -4.61611420e-01 7.15996325e-01
5.80268800e-01 2.01234668e-01 3.42206895e-01 8.20037723e-02
5.21083236e-01 -4.14281398e-01 -4.08004194e-01 1.19375408e+00
-2.97715604e-01 -1.40169942e+00 -2.60965861e-02 -1.63191184e-03
6.29041970e-01 -1.64800555e-01 -1.58036407e-02 -2.94618100e-01
-1.72541186e-01 -1.02471066e+00 7.58425891e-02 8.82163048e-01
7.26108372e-01 -9.68885899e-01 1.23745847e+00 3.72329295e-01
-5.90839922e-01 1.44602001e-01 -7.11049199e-01 1.35426015e-01
1.38323337e-01 1.45440269e+00 1.86615884e-01 3.42722833e-01
1.04875851e+00 4.83991057e-01 -4.28113416e-02 9.31355596e-01
8.83879736e-02 4.92613494e-01 -3.27977598e-01 1.31797150e-01
3.54818888e-02 -3.80097836e-01 4.72462326e-01 1.05327559e+00
5.67504644e-01 7.56461143e-01 -1.06141217e-01 7.31435776e-01
-1.10407695e-01 -4.31194603e-02 -1.53117865e-01 -1.50886372e-01
3.18005115e-01 1.64759874e+00 -1.06699061e+00 -1.88791022e-01
-3.54220897e-01 7.24559546e-01 -1.82444111e-01 4.96071070e-01
-4.78160501e-01 -5.67379534e-01 4.32856321e-01 2.32122138e-01
1.23577006e-01 -8.54659155e-02 -1.87976792e-01 -1.00705922e+00
-1.90616310e-01 -8.70057821e-01 1.89891562e-01 -5.59234381e-01
-1.35027587e+00 3.23157787e-01 -6.40876353e-01 -6.66984975e-01
7.88780272e-01 -5.81618905e-01 -7.12700188e-01 8.45739663e-01
-1.73343313e+00 -7.79254556e-01 -5.05126894e-01 6.18712366e-01
8.99775922e-02 1.49189010e-01 6.75557137e-01 5.38832366e-01
-6.45326614e-01 1.60353094e-01 6.36791468e-01 -4.63327877e-02
6.75726950e-01 -9.33503330e-01 -4.14953262e-01 6.83925211e-01
-6.32310390e-01 5.68164527e-01 6.57058299e-01 -5.24902225e-01
-1.28333032e+00 -6.91974282e-01 6.40969038e-01 2.51026690e-01
3.26352566e-01 1.77968636e-01 -1.32660520e+00 1.14281699e-01
2.54631162e-01 3.45313221e-01 6.89042449e-01 -5.97657800e-01
4.30814862e-01 -2.37566113e-01 -1.49235642e+00 4.25953209e-01
5.29754400e-01 -2.75739223e-01 -2.60077212e-02 4.15986061e-01
3.63564342e-01 -4.51689363e-01 -1.24759722e+00 4.85644788e-01
5.23627698e-01 -1.16069615e+00 8.38494897e-01 -6.08936185e-03
3.88475031e-01 -5.11242986e-01 8.23371951e-03 -1.05215919e+00
-5.96797049e-01 -4.48688418e-01 1.83644965e-01 1.01912224e+00
-1.84486985e-01 -6.22274041e-01 7.92498052e-01 7.17460513e-02
-4.16207284e-01 -7.23579884e-01 -1.17414272e+00 -4.69479263e-01
1.13540329e-01 2.42388934e-01 1.92035660e-01 7.97572255e-01
-4.78913695e-01 4.08287952e-03 -9.29629430e-02 4.31887358e-02
8.96858931e-01 -7.99086317e-02 1.58549577e-01 -1.35083461e+00
2.34245777e-01 -1.88090801e-01 -4.35844660e-01 -7.84751892e-01
-1.99873073e-04 -5.21685839e-01 2.14124303e-02 -1.55526960e+00
4.83267248e-01 2.58513503e-02 -4.86725479e-01 2.47061357e-01
-1.01530619e-01 4.38221991e-01 -4.92571890e-01 1.63115636e-01
-3.16998303e-01 6.26025200e-01 1.80357695e+00 -1.80885419e-01
1.18004866e-01 -2.97911316e-01 -7.05507100e-01 6.96859241e-01
6.71645820e-01 -5.02447009e-01 1.16140001e-01 -4.61868584e-01
-4.12508957e-02 1.62809670e-01 5.31062543e-01 -1.08446479e+00
4.50534105e-01 -1.32685721e-01 8.04924071e-01 -5.06480992e-01
4.91274055e-03 -6.49185240e-01 3.25632751e-01 6.35869980e-01
1.10445276e-01 -3.98999304e-01 2.21708901e-02 5.16261220e-01
-5.47349870e-01 -2.40376160e-01 1.27235234e+00 -4.65882540e-01
-4.64431345e-01 2.17811331e-01 -8.84591162e-01 -2.85955906e-01
8.19680750e-01 -4.95520681e-01 -6.38104618e-01 -2.04965144e-01
-7.22118437e-01 8.02411959e-02 4.62083489e-01 -4.09890652e-01
9.58696783e-01 -1.16560721e+00 -6.38221323e-01 2.24765167e-01
-3.61498058e-01 2.20713854e-01 1.00552583e+00 1.45563531e+00
-7.27895081e-01 -1.40917659e-01 -4.90169555e-01 -6.67228222e-01
-1.05248058e+00 1.79107860e-01 6.48284793e-01 -1.27558693e-01
-5.84086716e-01 9.16110218e-01 2.64524430e-01 -5.52260950e-02
-2.59445250e-01 -8.40123594e-02 -2.40531951e-01 -1.16414234e-01
6.63565338e-01 6.12573922e-01 2.60596573e-01 -5.70071518e-01
-3.22487205e-01 7.78791487e-01 -1.27832428e-01 4.21313077e-01
1.62667537e+00 -2.64260828e-01 -5.95622838e-01 -4.74697985e-02
1.17467380e+00 -2.45697692e-01 -1.36250544e+00 6.33839816e-02
-3.90660673e-01 -2.16074020e-01 6.52573168e-01 -6.78019345e-01
-1.35558605e+00 1.04790568e+00 9.04943347e-01 2.02231392e-01
1.51082897e+00 -1.12845220e-01 1.37225497e+00 7.72280395e-02
9.23062116e-02 -1.16127324e+00 2.56143719e-01 2.96134412e-01
5.09781420e-01 -1.05073786e+00 -2.14644670e-02 -3.65662992e-01
-2.87472963e-01 1.30352485e+00 4.77115750e-01 -1.09498605e-01
4.35731649e-01 5.50881445e-01 2.47279719e-01 -4.67074335e-01
-4.86556739e-01 -1.03497028e-01 -2.46854916e-01 8.46568227e-01
4.06521618e-01 -3.56380403e-01 -5.27970970e-01 6.97286665e-01
4.88882989e-01 1.57647878e-01 5.83564818e-01 7.20784307e-01
-7.44587362e-01 -6.79636598e-01 -4.09840167e-01 5.46736121e-01
-1.00174594e+00 1.44817933e-01 9.24510658e-02 4.83161956e-01
1.75859362e-01 8.37629676e-01 -1.13364803e-02 -2.05804780e-02
3.09871048e-01 -3.78527611e-01 5.33954024e-01 -1.19308330e-01
-3.40581477e-01 8.21862578e-01 -4.38214570e-01 -4.02754605e-01
-8.66633236e-01 -4.09301192e-01 -1.56487811e+00 -5.31987786e-01
-4.48914975e-01 7.08706230e-02 4.32978719e-01 8.18836868e-01
2.55249977e-01 6.94781601e-01 4.39055949e-01 -5.59383214e-01
-9.11117345e-02 -7.69370258e-01 -1.03956246e+00 3.46598566e-01
4.40924734e-01 -5.24597049e-01 -4.55101013e-01 1.65192708e-01] | [13.07038688659668, -2.593780517578125] |
1c28703c-1da1-401a-8a79-fd06459443d1 | analyzing-multi-task-learning-for-abstractive | 2210.14606 | null | https://arxiv.org/abs/2210.14606v2 | https://arxiv.org/pdf/2210.14606v2.pdf | Analyzing Multi-Task Learning for Abstractive Text Summarization | Despite the recent success of multi-task learning and pre-finetuning for natural language understanding, few works have studied the effects of task families on abstractive text summarization. Task families are a form of task grouping during the pre-finetuning stage to learn common skills, such as reading comprehension. To close this gap, we analyze the influence of multi-task learning strategies using task families for the English abstractive text summarization task. We group tasks into one of three strategies, i.e., sequential, simultaneous, and continual multi-task learning, and evaluate trained models through two downstream tasks. We find that certain combinations of task families (e.g., advanced reading comprehension and natural language inference) positively impact downstream performance. Further, we find that choice and combinations of task families influence downstream performance more than the training scheme, supporting the use of task families for abstractive text summarization. | ['Bela Gipp', 'Terry Ruas', 'Jan Philip Wahle', 'Frederic Kirstein'] | 2022-10-26 | null | null | null | null | ['abstractive-text-summarization'] | ['natural-language-processing'] | [ 6.46406829e-01 1.62835181e-01 -3.81183326e-01 -3.85041952e-01
-7.69106150e-01 -7.31042385e-01 7.98202574e-01 6.87908709e-01
-7.03108370e-01 6.75850749e-01 9.31544483e-01 -5.89202464e-01
-2.41472900e-01 -4.73360687e-01 -6.47429585e-01 -1.59287557e-01
3.57064724e-01 2.17195317e-01 1.25373483e-01 -3.47542703e-01
8.51656497e-01 -3.85015272e-02 -1.36903620e+00 5.56792736e-01
1.58437896e+00 2.02584267e-01 6.34676099e-01 8.57998192e-01
-3.38697821e-01 8.89536321e-01 -1.07400107e+00 -2.66845226e-01
-1.94248706e-01 -6.40950561e-01 -1.09745419e+00 -1.76192060e-01
5.31869411e-01 -2.58327454e-01 1.11535676e-01 6.45812452e-01
7.37309277e-01 4.00260717e-01 1.04974949e+00 -7.39026070e-01
-4.81748313e-01 9.59463954e-01 -4.68722761e-01 6.63903952e-01
6.65805161e-01 5.19164801e-02 1.14592803e+00 -4.59163636e-01
5.03218353e-01 1.27016544e+00 6.23532295e-01 4.97190475e-01
-1.29230726e+00 -5.66078842e-01 3.00449729e-01 -8.31672624e-02
-6.11605704e-01 -6.66780651e-01 5.28119922e-01 -6.56330049e-01
1.19632638e+00 1.12455033e-01 3.22808653e-01 1.06773281e+00
7.18862414e-01 9.14741039e-01 1.09906900e+00 -5.93967557e-01
2.54969415e-03 -9.84741971e-02 6.35920882e-01 4.07572955e-01
5.50372183e-01 -3.85343820e-01 -7.50237584e-01 -1.65299803e-01
4.22064155e-01 -3.02890271e-01 -2.65263885e-01 2.99153090e-01
-1.28160381e+00 8.42732072e-01 -2.05450237e-01 3.04282397e-01
-2.05411673e-01 -6.53885230e-02 7.72201777e-01 6.11977398e-01
7.58265197e-01 1.20022488e+00 -7.23820508e-01 -3.57544273e-01
-9.93063807e-01 2.51297057e-01 9.70252931e-01 7.47476518e-01
7.34564126e-01 9.24008787e-02 -8.91380668e-01 1.13571417e+00
-9.34958309e-02 2.69611597e-01 7.56472528e-01 -8.15565288e-01
9.68843699e-01 5.35072207e-01 -1.21876337e-01 -5.08300960e-01
-6.07710958e-01 -3.47763717e-01 -5.94976485e-01 -4.49738085e-01
5.74634075e-01 -7.35231757e-01 -7.28119850e-01 1.72344792e+00
-2.26037994e-01 -3.65668237e-01 2.11203754e-01 2.82028109e-01
1.01487482e+00 7.65965760e-01 3.36058378e-01 -7.23395824e-01
1.29515219e+00 -1.22790122e+00 -8.58108759e-01 -4.43694979e-01
1.07895041e+00 -7.57590473e-01 1.33203006e+00 3.92242908e-01
-1.24201143e+00 -6.89822912e-01 -7.84875393e-01 -2.79145628e-01
-2.12686419e-01 -2.96623763e-02 4.13429141e-01 4.59606409e-01
-9.03617382e-01 6.43088698e-01 -5.22156775e-01 -2.33679250e-01
1.75746456e-01 1.14722356e-01 1.02942668e-01 8.22772086e-02
-1.18210363e+00 1.21797681e+00 6.58387601e-01 -4.99156862e-01
-7.55995393e-01 -8.65873933e-01 -8.34792852e-01 3.13238382e-01
4.50579464e-01 -1.13853288e+00 1.56069076e+00 -1.02251792e+00
-1.39989972e+00 6.77144289e-01 -2.93643415e-01 -2.82764822e-01
2.84307629e-01 -6.46693766e-01 2.21002460e-01 -5.68297468e-02
2.01709807e-01 5.55034697e-01 9.61405039e-01 -1.00356627e+00
-6.49314404e-01 -3.78653288e-01 3.87124047e-02 6.49833798e-01
-5.59587359e-01 7.50660822e-02 1.97016492e-01 -8.52093875e-01
-3.77410918e-01 -6.34672821e-01 -6.55140355e-02 -9.55820501e-01
-4.89181846e-01 -8.51257741e-01 4.27276313e-01 -8.23304355e-01
1.56829584e+00 -1.96664023e+00 4.44229811e-01 -3.22171956e-01
3.47279906e-01 2.51815200e-01 -2.84757763e-01 5.90489030e-01
-5.47829606e-02 5.79049528e-01 -7.49709904e-02 -5.46042442e-01
-1.07214861e-01 -3.85751612e-02 -2.61969231e-02 -1.51846692e-01
1.61973745e-01 1.01751697e+00 -1.00127959e+00 -5.23334920e-01
-1.09936558e-01 -1.87748745e-01 -5.17230153e-01 1.70363545e-01
-6.06198370e-01 4.64047194e-01 -5.73674679e-01 1.18613616e-01
1.01017609e-01 -1.51450634e-01 -7.51750693e-02 1.48221835e-01
-3.16518515e-01 8.75226498e-01 -5.17453074e-01 1.75596452e+00
-6.40825391e-01 7.46335864e-01 -1.31661817e-01 -9.73564267e-01
7.38778174e-01 3.24052364e-01 5.31664491e-02 -7.68775344e-01
-1.04682758e-01 3.79777513e-02 4.33487266e-01 -5.55006206e-01
9.47667599e-01 -9.41575840e-02 -1.71981588e-01 9.46299851e-01
1.03950940e-01 -4.74890262e-01 5.60455561e-01 4.35130686e-01
9.82840002e-01 -1.60532724e-02 5.75756609e-01 -5.50437629e-01
9.84021574e-02 1.14092946e-01 3.32158834e-01 1.18544888e+00
2.09862962e-02 2.66829044e-01 7.57651269e-01 3.57425176e-02
-9.23607469e-01 -6.02029145e-01 1.73841402e-01 1.99638891e+00
-2.62449741e-01 -7.16694415e-01 -7.39014804e-01 -5.65864861e-01
-3.22337598e-02 1.29119754e+00 -4.43106085e-01 -4.93195891e-01
-6.33282185e-01 -5.97902656e-01 6.18712544e-01 4.84633833e-01
5.40353596e-01 -1.30127013e+00 -9.16292906e-01 1.00431941e-01
-4.70251739e-01 -8.11349332e-01 -7.62656212e-01 3.64783168e-01
-1.22418296e+00 -7.31899500e-01 -6.90680385e-01 -8.49819899e-01
3.85950357e-01 4.27013338e-01 1.41904557e+00 1.90468922e-01
3.65416914e-01 7.23370731e-01 -6.24804914e-01 -7.45634019e-01
-7.64986396e-01 8.58605206e-01 -9.76248756e-02 -5.62387466e-01
-1.62312716e-01 -5.28411925e-01 -2.04742432e-01 7.12651536e-02
-7.99570918e-01 5.11599004e-01 9.00138259e-01 7.56590962e-01
1.32075369e-01 -1.33333609e-01 1.06783938e+00 -1.14906740e+00
1.65352631e+00 -3.42240036e-01 1.07609838e-01 6.68122113e-01
-6.08407319e-01 2.41977379e-01 6.20889783e-01 -6.51404977e-01
-1.39341354e+00 -3.94423783e-01 1.52302384e-01 1.87085807e-01
-1.72803059e-01 9.05924439e-01 2.26913884e-01 3.36254597e-01
7.86991298e-01 3.13672751e-01 9.71219763e-02 -2.37720668e-01
1.34688050e-01 4.75752234e-01 2.03155696e-01 -9.24232006e-01
4.61330235e-01 -3.22305292e-01 -4.70526963e-01 -1.16339087e+00
-1.23579812e+00 -3.32084954e-01 -5.13686299e-01 -1.68092683e-01
1.21056616e+00 -7.64870167e-01 -4.33739007e-01 5.84402502e-01
-1.28664541e+00 -8.87094438e-01 -1.34698763e-01 3.12503666e-01
-5.12976110e-01 4.49682713e-01 -5.07925987e-01 -4.60913837e-01
-5.75585902e-01 -1.11658657e+00 1.02168846e+00 4.12150145e-01
-7.21584857e-01 -1.28016853e+00 -7.98335746e-02 5.56520343e-01
4.88026172e-01 -8.56706351e-02 1.47294509e+00 -1.13210750e+00
-2.00751051e-02 4.14950818e-01 2.13699844e-02 2.35277191e-01
1.96578935e-01 -1.62237138e-01 -5.91242850e-01 -1.28539667e-01
3.16062570e-01 -7.08422244e-01 1.18445122e+00 7.02211618e-01
1.08903074e+00 -4.03668731e-01 -2.61564314e-01 1.53798252e-01
7.86947012e-01 -8.21623020e-03 2.77304053e-01 3.65253985e-01
7.31010914e-01 7.42443860e-01 4.90314186e-01 2.67206520e-01
5.43011785e-01 3.60232294e-01 -2.63905942e-01 2.28493512e-01
-1.18139766e-01 -1.57727242e-01 5.84829211e-01 9.72077966e-01
-3.88574600e-02 -3.93012315e-01 -1.15455604e+00 3.37399721e-01
-1.64697754e+00 -7.46322334e-01 -1.57142788e-01 1.87076557e+00
1.29096711e+00 3.71420234e-01 1.67484596e-01 -5.04152477e-02
4.96332675e-01 4.02639598e-01 -4.99136299e-01 -9.06117320e-01
1.10970117e-01 6.95361495e-02 2.23174781e-01 5.50243497e-01
-8.23339283e-01 1.12533712e+00 6.32403469e+00 8.90367448e-01
-8.82160068e-01 2.20964681e-02 6.28564417e-01 7.79576004e-02
-5.97410321e-01 -9.38070789e-02 -8.26639593e-01 2.59472579e-01
8.74515116e-01 -6.40991569e-01 1.08549863e-01 1.71191335e-01
4.47621226e-01 -5.63001573e-01 -1.28636277e+00 2.01908052e-01
2.43176997e-01 -1.07013476e+00 3.36538225e-01 -1.68263972e-01
9.14674819e-01 -2.54111052e-01 -2.63356894e-01 7.19072938e-01
4.45311755e-01 -9.33905065e-01 5.53348243e-01 3.31663519e-01
5.01578331e-01 -4.91273046e-01 4.57029372e-01 9.07891929e-01
-6.61264122e-01 -2.91118890e-01 -2.92695761e-02 -5.13474822e-01
3.30823034e-01 4.76844311e-01 -7.08455443e-01 5.95696867e-01
2.42375702e-01 6.41307890e-01 -8.79111409e-01 7.18892694e-01
-2.56620139e-01 9.81787860e-01 1.21512569e-01 -9.67523679e-02
4.24218588e-02 -2.04223022e-02 6.86058998e-01 1.33391702e+00
-6.61681890e-02 2.06233427e-01 4.24692094e-01 4.49493974e-01
-3.02300137e-03 2.24330112e-01 -6.24364018e-01 -2.60554522e-01
4.31980222e-01 7.37745345e-01 -6.71727300e-01 -4.27221864e-01
-2.31921598e-01 7.64446080e-01 4.33376610e-01 3.76703709e-01
-4.34276640e-01 -3.83599728e-01 8.13348740e-02 -2.84342561e-02
-2.33076327e-02 -6.07923627e-01 -9.62401867e-01 -1.05893707e+00
-2.25679770e-01 -1.05590093e+00 4.05941486e-01 -6.62017405e-01
-1.03071988e+00 7.54605234e-02 5.00848472e-01 -4.57437485e-01
-2.15210244e-01 -6.70287162e-02 -1.21577680e+00 6.27722204e-01
-1.31570995e+00 -7.93247163e-01 -1.26120806e-01 3.15980703e-01
1.26240253e+00 -1.78141281e-01 7.07180440e-01 -2.36807138e-01
-8.13254356e-01 5.92358828e-01 2.58622095e-02 -1.97281659e-01
1.01684129e+00 -1.32192981e+00 4.00893390e-01 6.67210162e-01
-1.26906678e-01 7.69269586e-01 6.60898328e-01 -9.41448212e-01
-1.19032657e+00 -8.44281197e-01 1.12713134e+00 -4.70682263e-01
4.73980635e-01 -4.53660965e-01 -8.66269171e-01 8.14520955e-01
6.73774540e-01 -1.15666699e+00 7.98291564e-01 6.84188008e-01
-9.78638381e-02 1.42831221e-01 -5.95493555e-01 7.57009745e-01
1.14690709e+00 -2.36305028e-01 -1.23726237e+00 4.62127596e-01
1.26716816e+00 -3.70374143e-01 -8.58564079e-01 3.76031071e-01
3.00344199e-01 -6.66320622e-01 7.04063356e-01 -6.32893026e-01
9.66831803e-01 4.57398862e-01 4.01624173e-01 -1.88339710e+00
-3.84240776e-01 -6.03853464e-01 3.91848013e-02 1.36439598e+00
8.18825841e-01 -7.63757050e-01 2.58302420e-01 2.86578655e-01
-7.41380513e-01 -5.80315530e-01 -5.04052579e-01 -6.04754388e-01
7.19447196e-01 1.26876384e-01 -2.15003472e-02 9.08936381e-01
7.49468952e-02 1.14416575e+00 -1.32994100e-01 -4.91786927e-01
1.56535625e-01 -9.49780867e-02 8.54340255e-01 -1.56776261e+00
-2.12983936e-01 -8.15255880e-01 5.09014785e-01 -1.25284970e+00
6.36570215e-01 -9.05476272e-01 1.32306129e-01 -2.05752158e+00
2.41593570e-01 -1.52882382e-01 1.56649783e-01 5.82960963e-01
-7.84828305e-01 -6.91537559e-01 2.69150317e-01 3.24676424e-01
-8.43465805e-01 4.84260857e-01 1.58804858e+00 -3.75175476e-02
-6.35713041e-01 3.35152596e-01 -1.08004534e+00 5.20565093e-01
9.86438632e-01 -2.82701790e-01 -6.89777076e-01 -8.49136293e-01
2.65895039e-01 3.08016866e-01 -2.07148507e-01 -6.71678901e-01
3.02041948e-01 -3.26414168e-01 2.39163280e-01 -3.67353648e-01
-1.07383765e-01 -4.88685481e-02 -5.87893009e-01 4.17742878e-01
-9.20484185e-01 4.68757182e-01 5.48639953e-01 4.09132928e-01
-1.20341644e-01 -5.24035513e-01 4.19113487e-01 -2.22073928e-01
-3.80231977e-01 -4.36934739e-01 -8.42778325e-01 6.83994532e-01
7.13423967e-01 -4.77611125e-01 -6.49808645e-01 -5.42071104e-01
-5.93229830e-01 6.84329033e-01 1.44981191e-01 5.73595405e-01
3.08707058e-01 -5.99763691e-01 -1.20475233e+00 -2.44740974e-02
-1.42135188e-01 3.93622339e-01 1.51345670e-01 9.32865083e-01
-5.49755134e-02 5.20280361e-01 -1.48315847e-01 -4.89649147e-01
-1.26184642e+00 4.38658957e-04 -4.82296832e-02 -7.45548189e-01
-4.00839508e-01 6.33232832e-01 2.72091001e-01 -3.11740786e-01
2.42079988e-01 -4.87498820e-01 -5.71814775e-01 5.47314227e-01
2.73040682e-01 6.07477486e-01 -4.73698564e-02 -8.28277171e-02
6.87333047e-02 3.57901722e-01 -6.17182195e-01 -1.31156817e-01
1.12519145e+00 -7.01707453e-02 -2.35280186e-01 7.45671988e-01
8.02100062e-01 -8.99707600e-02 -8.99187446e-01 -1.51001737e-01
4.16351557e-01 1.09045230e-01 -1.85880348e-01 -9.53766584e-01
-2.34227568e-01 6.73578978e-01 -2.48992994e-01 3.90945137e-01
1.02093470e+00 6.13932963e-03 5.11422276e-01 7.56941855e-01
-6.16867952e-02 -1.22538209e+00 5.29530466e-01 1.21940017e+00
9.49629664e-01 -1.05687535e+00 2.66276926e-01 -2.05803573e-01
-8.18703473e-01 8.68050456e-01 9.30354416e-01 1.41264647e-01
2.54551947e-01 -3.67770270e-02 -2.84049600e-01 -1.98869750e-01
-1.03644609e+00 6.52082041e-02 3.32785338e-01 3.19114834e-01
8.55175138e-01 -8.70707408e-02 -8.44868720e-01 5.51526964e-01
-6.42610908e-01 -2.52778530e-01 7.34150290e-01 9.40446198e-01
-7.50205636e-01 -8.58407557e-01 -2.78814495e-01 1.16387475e+00
-4.43324298e-01 -3.68705750e-01 -7.11803079e-01 5.68104148e-01
-2.50132024e-01 1.13324547e+00 1.31670043e-01 -9.78372172e-02
4.57924128e-01 4.63342935e-01 5.16308188e-01 -1.27656722e+00
-1.27813733e+00 -1.76845700e-01 4.92316663e-01 1.20988511e-01
-3.22263300e-01 -7.54677415e-01 -1.02264631e+00 -8.81185159e-02
-5.62311113e-01 3.19260925e-01 2.48109460e-01 1.40543389e+00
3.15489709e-01 7.85711706e-01 2.23441288e-01 -8.01481187e-01
-8.88575435e-01 -1.58349657e+00 -5.20555414e-02 2.14264855e-01
3.67277741e-01 -5.32304883e-01 -2.28132978e-01 2.34334506e-02] | [12.259114265441895, 9.297122955322266] |
5a215277-a8cc-4ec9-98b0-3884f241a41a | view-invariant-probabilistic-embedding-for | 1912.01001 | null | https://arxiv.org/abs/1912.01001v4 | https://arxiv.org/pdf/1912.01001v4.pdf | View-Invariant Probabilistic Embedding for Human Pose | Depictions of similar human body configurations can vary with changing viewpoints. Using only 2D information, we would like to enable vision algorithms to recognize similarity in human body poses across multiple views. This ability is useful for analyzing body movements and human behaviors in images and videos. In this paper, we propose an approach for learning a compact view-invariant embedding space from 2D joint keypoints alone, without explicitly predicting 3D poses. Since 2D poses are projected from 3D space, they have an inherent ambiguity, which is difficult to represent through a deterministic mapping. Hence, we use probabilistic embeddings to model this input uncertainty. Experimental results show that our embedding model achieves higher accuracy when retrieving similar poses across different camera views, in comparison with 2D-to-3D pose lifting models. We also demonstrate the effectiveness of applying our embeddings to view-invariant action recognition and video alignment. Our code is available at https://github.com/google-research/google-research/tree/master/poem. | ['Liang-Chieh Chen', 'Jiaping Zhao', 'Jennifer J. Sun', 'Florian Schroff', 'Ting Liu', 'Hartwig Adam'] | 2019-12-02 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2905_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123500052.pdf | eccv-2020-8 | ['pose-retrieval', 'video-alignment'] | ['computer-vision', 'computer-vision'] | [-2.38315433e-01 -2.09371418e-01 -4.11178946e-01 -2.60406852e-01
-5.27776122e-01 -5.86168647e-01 5.45404494e-01 -2.17267200e-01
-1.95133671e-01 3.40907097e-01 6.73432291e-01 2.63798028e-01
5.23332730e-02 -2.57174313e-01 -6.67026281e-01 -3.14376622e-01
-1.45946518e-01 5.72903275e-01 1.07992060e-01 7.68302828e-02
3.18710622e-03 4.14264172e-01 -1.47058749e+00 -9.95708779e-02
2.04719216e-01 6.51247263e-01 -7.11823180e-02 8.09768677e-01
3.86948586e-01 5.50190032e-01 -2.34626800e-01 -4.09688950e-01
4.46268290e-01 -3.03243160e-01 -5.61534047e-01 3.57085377e-01
8.66066813e-01 -7.09865630e-01 -7.74953961e-01 9.52837408e-01
5.11940956e-01 6.18361793e-02 5.78274488e-01 -1.35369241e+00
-5.81069589e-01 -2.14422345e-02 -5.75298905e-01 2.09259212e-01
1.13384128e+00 9.51631516e-02 9.63089287e-01 -8.49831939e-01
6.71878457e-01 1.32189322e+00 5.18896759e-01 7.10742950e-01
-1.04872620e+00 -2.61437476e-01 1.72646418e-01 3.95538211e-01
-1.41515934e+00 -3.81223381e-01 9.22458231e-01 -7.75335908e-01
5.24594069e-01 2.81083882e-01 1.10600770e+00 1.35789037e+00
4.27695602e-01 9.82588768e-01 7.58386493e-01 -2.58206010e-01
-1.59065008e-01 -1.91785559e-01 -1.17382377e-01 8.16551805e-01
4.42692161e-01 -2.00676229e-02 -8.00742388e-01 -1.42424375e-01
1.06567407e+00 5.81407547e-01 -4.75792408e-01 -1.11917853e+00
-1.45992839e+00 6.47698343e-01 2.63719916e-01 -2.36872956e-02
-2.81829238e-01 4.56050813e-01 1.26930013e-01 4.52813171e-02
2.45298073e-01 2.52732009e-01 -1.91896513e-01 -5.45930505e-01
-4.75064099e-01 3.80349308e-01 6.54717982e-01 9.36959028e-01
2.82593250e-01 -1.45573273e-01 1.68123826e-01 5.67715943e-01
5.01108706e-01 7.71686792e-01 5.35571456e-01 -1.17102826e+00
3.47002864e-01 4.65188324e-01 2.28933081e-01 -1.34545815e+00
-3.34874660e-01 1.71988428e-01 -5.20119011e-01 1.43146604e-01
3.82722110e-01 1.17973924e-01 -6.96002066e-01 1.73030972e+00
5.80204844e-01 6.93627745e-02 -2.62663364e-01 1.24136376e+00
4.82829988e-01 2.15535745e-01 -3.81063759e-01 1.89118132e-01
1.37636185e+00 -8.37395012e-01 -6.69806421e-01 -3.41035634e-01
2.64250219e-01 -7.18248248e-01 8.41150284e-01 1.75382286e-01
-9.65496480e-01 -4.55411434e-01 -9.80047107e-01 -2.13149056e-01
5.26061729e-02 2.50707883e-02 5.03502011e-01 5.13663948e-01
-7.36405313e-01 5.08852243e-01 -1.45814979e+00 -5.93693018e-01
8.50193501e-02 2.89182723e-01 -7.89631486e-01 -9.83759165e-02
-9.34693396e-01 1.07869732e+00 4.68410887e-02 -8.05025026e-02
-7.17524946e-01 -2.99277753e-01 -1.32998633e+00 -4.98192579e-01
5.68723142e-01 -9.36008036e-01 1.09084022e+00 -3.77412617e-01
-1.37217474e+00 8.23287725e-01 -2.86778241e-01 -1.68844521e-01
7.07518041e-01 -8.38607907e-01 -8.44946802e-02 3.69245738e-01
-1.11944145e-02 5.95997393e-01 9.83769536e-01 -9.90342379e-01
-1.64618596e-01 -7.69956291e-01 3.28813344e-01 6.76153898e-01
-1.53164253e-01 -2.61786878e-01 -7.35897183e-01 -6.84307754e-01
4.98081803e-01 -1.37403476e+00 -3.62179577e-02 6.83711946e-01
-2.64387190e-01 7.23951757e-02 6.39219940e-01 -7.77773321e-01
9.09263194e-01 -2.06807756e+00 6.82872653e-01 -8.19385573e-02
1.99277803e-01 -2.42001459e-01 2.51581073e-01 3.97610664e-01
6.99400380e-02 -1.00439087e-01 1.01782948e-01 -2.74960458e-01
1.11675113e-01 3.74220550e-01 -4.77280840e-02 9.06849980e-01
3.74062397e-02 8.32846582e-01 -9.86489594e-01 -5.32139182e-01
6.68126166e-01 6.22012198e-01 -5.45425177e-01 2.05662027e-01
1.90728635e-01 4.90003943e-01 -5.82988024e-01 7.73819506e-01
3.02946448e-01 -2.68881828e-01 3.61616910e-01 -5.12905478e-01
3.65865231e-01 1.19854130e-01 -1.39746618e+00 2.02317023e+00
-1.75618738e-01 5.33572257e-01 -3.75364810e-01 -6.52339518e-01
5.44093907e-01 2.96267658e-01 8.08250487e-01 -1.37230739e-01
2.46448010e-01 -2.98660338e-01 -1.77366227e-01 -3.79196018e-01
4.75458205e-01 6.36561438e-02 -2.55891830e-01 4.21337783e-01
9.05563533e-02 -2.02592030e-01 -8.01541507e-02 3.54567580e-02
9.55182016e-01 4.77172911e-01 6.19975150e-01 1.79161683e-01
2.56719410e-01 -3.64120781e-01 5.35809755e-01 2.25837678e-01
-4.96019512e-01 8.67960393e-01 2.65840352e-01 -6.14202261e-01
-9.54368889e-01 -1.55377197e+00 8.08673166e-03 4.56647873e-01
5.01441240e-01 -6.75015450e-01 -3.06837946e-01 -5.61736941e-01
3.57600749e-01 1.76143676e-01 -5.85296750e-01 -1.31700739e-01
-6.43457353e-01 -4.52464223e-02 1.59554183e-01 8.16436768e-01
1.84534863e-01 -3.98163348e-01 -9.84425664e-01 -1.75878197e-01
-3.29663157e-01 -1.04068005e+00 -9.22418594e-01 -3.78953129e-01
-8.87515724e-01 -1.35252178e+00 -8.32453847e-01 -3.33247244e-01
6.30072355e-01 3.98397148e-01 8.67364287e-01 -2.23353207e-01
-4.81972724e-01 1.26503456e+00 -4.15718049e-01 -1.39946073e-01
-3.22160646e-02 -4.90538329e-01 7.51298308e-01 -1.84307545e-01
5.23886502e-01 -6.06814384e-01 -7.18061388e-01 5.51210225e-01
-5.89002669e-01 1.43138105e-02 2.23129898e-01 8.34890783e-01
6.89750552e-01 -4.76256669e-01 -3.53719890e-01 -6.66137487e-02
2.56122649e-01 -1.34317100e-01 -3.96569252e-01 1.95865214e-01
-1.64649278e-01 4.80602644e-02 4.29526828e-02 -6.51085556e-01
-4.11557943e-01 3.46026152e-01 1.81894138e-01 -9.87854540e-01
-1.68233097e-01 2.49456331e-01 -1.53994426e-01 1.27145573e-01
4.49296057e-01 1.92912132e-01 4.75737214e-01 -3.85103881e-01
4.92355049e-01 3.85552764e-01 4.52431500e-01 -5.51106989e-01
8.70087385e-01 7.11439431e-01 -9.42237526e-02 -7.96257734e-01
-5.95221400e-01 -7.18829215e-01 -1.10920227e+00 -4.65854853e-01
9.76155519e-01 -1.14618897e+00 -5.62473834e-01 2.99993575e-01
-9.30660248e-01 9.34457779e-02 -2.30078280e-01 1.05454183e+00
-9.87510562e-01 8.36495757e-01 -3.61172885e-01 -6.33110464e-01
3.24029364e-02 -1.13660228e+00 1.36245787e+00 1.06104478e-01
-7.40770340e-01 -9.23075616e-01 3.25940579e-01 4.44876134e-01
-4.19854671e-01 4.79640156e-01 2.72904366e-01 -2.34695643e-01
-5.11575699e-01 -3.17154735e-01 3.97332996e-01 3.30333203e-01
5.04282832e-01 -1.28960982e-01 -4.86341983e-01 -4.90603417e-01
-8.74913409e-02 -3.59887779e-01 4.44692016e-01 5.66673338e-01
8.79676580e-01 -1.69327036e-01 -2.65413254e-01 6.48763955e-01
1.02901721e+00 -1.94569737e-01 4.59674180e-01 2.78418809e-01
9.16469276e-01 4.28909749e-01 7.34246254e-01 6.90094709e-01
5.92943847e-01 1.17214441e+00 3.74153018e-01 4.66133416e-01
-8.36511329e-02 -4.85340923e-01 6.53567910e-01 7.82615721e-01
-2.69619882e-01 2.15452351e-02 -8.59937549e-01 6.08078718e-01
-1.96378732e+00 -1.11389899e+00 3.31882030e-01 2.43796587e+00
6.59022689e-01 -1.50985822e-01 2.89783478e-01 -1.00799717e-01
6.54008567e-01 4.66362327e-01 -7.30117440e-01 -4.66898866e-02
3.09175640e-01 -1.66700751e-01 4.71708834e-01 4.19635177e-01
-1.23858273e+00 7.42712080e-01 5.71814156e+00 1.60792753e-01
-9.04842973e-01 -1.50253726e-02 -8.49935636e-02 -4.61970001e-01
-1.49124384e-01 -1.61013037e-01 -6.74089789e-01 3.97197068e-01
4.39700574e-01 -2.59267449e-01 1.53522268e-01 8.11700940e-01
-3.89445648e-02 -2.05937261e-03 -1.39264369e+00 1.51309061e+00
5.09096980e-01 -9.56543565e-01 1.26898751e-01 2.77114987e-01
5.47425687e-01 -8.42528641e-02 1.31115168e-01 -1.41155317e-01
1.90276280e-01 -7.10559011e-01 7.65991330e-01 6.04054809e-01
7.77101934e-01 -4.46824610e-01 3.30440074e-01 1.59946814e-01
-1.22212338e+00 1.26910523e-01 -4.27224308e-01 -8.91151428e-02
4.54231530e-01 1.48357525e-01 -6.32035971e-01 5.28565347e-01
8.36871386e-01 1.07283735e+00 -4.34229255e-01 9.76461411e-01
-3.37178051e-01 -4.37087640e-02 -4.96500611e-01 1.88769966e-01
-3.80531639e-01 -1.51772708e-01 8.65625322e-01 5.17957151e-01
4.29828763e-01 5.89366965e-02 3.81204873e-01 4.49321002e-01
1.06210224e-01 -1.68411881e-01 -9.32658434e-01 -1.29267961e-01
3.70986968e-01 8.38348031e-01 -4.23185706e-01 -9.89985466e-02
-4.63520199e-01 1.38098991e+00 1.46819791e-02 2.82230843e-02
-1.08199549e+00 1.35619000e-01 1.19059157e+00 1.33895114e-01
2.99962550e-01 -7.66017079e-01 2.39872083e-01 -1.67884016e+00
5.28962851e-01 -9.17883039e-01 4.11362469e-01 -8.12135577e-01
-1.21283770e+00 1.54316902e-01 3.69312793e-01 -1.71529746e+00
-5.86285472e-01 -7.91837931e-01 -1.19234912e-01 5.07513404e-01
-6.94577754e-01 -1.17209351e+00 -2.46957451e-01 7.06349432e-01
5.88445127e-01 1.40937269e-01 8.44383895e-01 -8.42864141e-02
-1.59816176e-01 5.42849898e-01 -2.05210634e-02 3.06593210e-01
9.20327842e-01 -1.30234969e+00 3.78610820e-01 7.01908231e-01
7.65076339e-01 7.33615279e-01 6.63316369e-01 -5.73158562e-01
-1.82770562e+00 -7.08776355e-01 5.68444848e-01 -1.17337096e+00
5.86069345e-01 -2.69109756e-01 -4.91412103e-01 9.81848359e-01
-2.80426681e-01 1.20387703e-01 9.42843378e-01 2.99851567e-01
-4.44811970e-01 7.72000104e-02 -9.07833517e-01 7.49496579e-01
1.32528830e+00 -7.00999320e-01 -1.02519286e+00 2.24567965e-01
3.91479552e-01 -7.43956983e-01 -1.13442278e+00 3.27053368e-01
1.17512786e+00 -7.36898839e-01 1.29055417e+00 -5.69012225e-01
2.70266563e-01 -4.74214375e-01 -5.39739907e-01 -1.13476765e+00
-2.62776285e-01 -3.39159220e-01 -4.75938439e-01 5.77398539e-01
-1.12662867e-01 -4.42028016e-01 7.74476051e-01 6.64837778e-01
3.30329865e-01 -5.50291479e-01 -1.08463740e+00 -9.97028768e-01
-2.29931429e-01 -3.26359093e-01 3.49080592e-01 7.30929434e-01
3.26436199e-02 -1.14689514e-01 -6.97129786e-01 5.18528342e-01
6.38646483e-01 3.56230140e-01 1.18928516e+00 -1.01159108e+00
-4.89670455e-01 -1.08205602e-01 -1.30627501e+00 -1.42619777e+00
2.73030717e-02 -5.27418852e-01 -1.41952783e-01 -1.34002352e+00
3.70401859e-01 3.11201036e-01 -2.33283445e-01 3.06116700e-01
-1.07923985e-01 3.21189463e-01 4.44134295e-01 4.42965865e-01
-6.63931191e-01 8.45207095e-01 1.09789789e+00 -2.10725859e-01
1.08883686e-01 -1.17582396e-01 -3.89606625e-01 9.42428589e-01
6.91516876e-01 -3.60892236e-01 -4.69861239e-01 -6.92509413e-01
-7.34075606e-02 8.22765529e-02 7.85033286e-01 -1.18838918e+00
-8.19943007e-03 -2.06451043e-01 6.26387954e-01 -6.50835454e-01
8.50153863e-01 -8.38458598e-01 3.90458852e-01 4.98698354e-01
-2.72767115e-02 3.03571373e-01 3.43007259e-02 8.08963537e-01
-1.91148281e-01 1.37060180e-01 3.96102309e-01 -2.13875696e-01
-8.30383420e-01 5.73412418e-01 -1.02823243e-01 5.09152515e-03
1.06355023e+00 -5.70960999e-01 7.09902793e-02 -7.78439462e-01
-9.90806580e-01 2.04621643e-01 1.12712419e+00 9.02964234e-01
9.11388338e-01 -1.67641747e+00 -3.75485361e-01 1.96566045e-01
3.48601639e-01 -2.22179472e-01 3.47078025e-01 8.22781324e-01
-6.27954364e-01 2.73449630e-01 -4.55024213e-01 -1.05853343e+00
-1.67216492e+00 4.20874804e-01 2.39010260e-01 2.30184078e-01
-7.26132512e-01 6.71933591e-01 1.57995477e-01 -5.21022916e-01
7.52569586e-02 -3.00707728e-01 6.91182688e-02 -1.87556431e-01
5.22647619e-01 3.15674305e-01 -4.98811275e-01 -1.02905965e+00
-6.34102404e-01 1.03469157e+00 5.94057851e-02 -2.18781695e-01
1.01201236e+00 -2.63177812e-01 3.10485333e-01 7.43479371e-01
1.27541184e+00 1.29466698e-01 -1.34198260e+00 -2.14659959e-01
-3.21731150e-01 -1.04753411e+00 -3.13387245e-01 -2.03469917e-01
-7.73395538e-01 8.89377475e-01 7.35447943e-01 -3.96955550e-01
7.85174608e-01 2.48956665e-01 5.89998722e-01 3.62374812e-01
7.17384458e-01 -8.45377386e-01 3.16472948e-01 3.09951365e-01
1.12320578e+00 -1.37916017e+00 4.41354692e-01 -2.59336859e-01
-7.64606118e-01 1.11690378e+00 6.09543741e-01 -1.87936753e-01
8.01064491e-01 -1.83203697e-01 6.18479215e-02 -2.22243086e-01
-5.73611856e-01 -1.75436944e-01 5.66477835e-01 7.70764291e-01
3.92802507e-01 3.29798758e-01 -9.51973125e-02 7.57722035e-02
-2.73830712e-01 -1.98910460e-01 4.97137368e-01 1.15553987e+00
-1.99269876e-01 -1.28639841e+00 -3.99950147e-01 1.63867295e-01
-3.33184391e-01 3.44240695e-01 -4.58322078e-01 8.96587074e-01
-2.19761223e-01 6.15743518e-01 1.48115292e-01 -6.43354058e-01
3.49545479e-01 1.30946845e-01 1.05432534e+00 -7.08656907e-01
4.26806301e-01 4.56566783e-03 -7.40083456e-02 -1.04754710e+00
-6.00448728e-01 -1.01651025e+00 -9.32790160e-01 -2.53667563e-01
-5.22164665e-02 -2.59291619e-01 4.53730583e-01 5.02728224e-01
3.32305014e-01 -1.75668336e-02 4.01166022e-01 -9.62478340e-01
-8.65990996e-01 -6.80150986e-01 -6.69553339e-01 6.54361308e-01
2.93631673e-01 -1.24481142e+00 -2.19805896e-01 2.51862645e-01] | [7.184626579284668, -0.6414316892623901] |
806fa490-3554-46b2-9d72-7511e784ed35 | medlatin1-and-medlatin2-two-datasets-for-the | 2006.12289 | null | https://arxiv.org/abs/2006.12289v2 | https://arxiv.org/pdf/2006.12289v2.pdf | MedLatinEpi and MedLatinLit: Two Datasets for the Computational Authorship Analysis of Medieval Latin Texts | We present and make available MedLatinEpi and MedLatinLit, two datasets of medieval Latin texts to be used in research on computational authorship analysis. MedLatinEpi and MedLatinLit consist of 294 and 30 curated texts, respectively, labelled by author; MedLatinEpi texts are of epistolary nature, while MedLatinLit texts consist of literary comments and treatises about various subjects. As such, these two datasets lend themselves to supporting research in authorship analysis tasks, such as authorship attribution, authorship verification, or same-author verification. Along with the datasets we provide experimental results, obtained on these datasets, for the authorship verification task, i.e., the task of predicting whether a text of unknown authorship was written by a candidate author or not. We also make available the source code of the authorship verification system we have used, thus allowing our experiments to be reproduced, and to be used as baselines, by other researchers. We also describe the application of the above authorship verification system, using these datasets as training data, for investigating the authorship of two medieval epistles whose authorship has been disputed by scholars. | ['Mirko Tavoni', 'Alejandro Moreo', 'Silvia Corbara', 'Fabrizio Sebastiani'] | 2020-06-22 | null | null | null | null | ['authorship-verification'] | ['natural-language-processing'] | [-2.34883223e-02 1.19772039e-01 -3.95361632e-01 -2.37300783e-01
-2.44739547e-01 -1.06337678e+00 1.17897284e+00 4.88430947e-01
-5.59874594e-01 9.34237480e-01 3.09073150e-01 -5.09183645e-01
-1.57744866e-02 -4.31603402e-01 -3.59793246e-01 -3.49318177e-01
5.28995097e-01 1.04830706e+00 -1.42957747e-01 -5.18752560e-02
1.14685929e+00 4.33139235e-01 -1.32537532e+00 4.30235453e-02
7.52958536e-01 4.79464322e-01 -4.83675212e-01 5.58623910e-01
1.16280228e-01 9.18570817e-01 -9.39535558e-01 -1.11213553e+00
-1.28561005e-01 -5.34013391e-01 -1.13073087e+00 -2.76267767e-01
7.23788381e-01 1.18368015e-01 -3.80839348e-01 5.29943049e-01
3.38105887e-01 -3.31973821e-01 1.22029448e+00 -1.50591147e+00
-7.35075355e-01 1.24365413e+00 -4.61834133e-01 3.17504466e-01
4.10411239e-01 -2.06043586e-01 1.23446071e+00 -6.87644362e-01
1.04958570e+00 1.25293732e+00 5.53356588e-01 4.64215189e-01
-9.54049408e-01 -9.65501964e-01 -7.26855934e-01 2.55824119e-01
-1.19115412e+00 -7.19524443e-01 8.80015194e-01 -7.27217019e-01
1.17375933e-01 5.60466796e-02 4.72072929e-01 1.54610944e+00
2.00808480e-01 8.01994920e-01 1.48291945e+00 -5.78407645e-01
5.62734948e-03 3.80952269e-01 5.05055726e-01 6.44579649e-01
4.85294312e-01 -9.18631405e-02 -7.92107701e-01 -6.39731228e-01
6.33578420e-01 -4.78431046e-01 -2.39676982e-01 1.55689120e-01
-1.55398834e+00 9.97432053e-01 1.94162913e-02 7.06034362e-01
-5.04958890e-02 -7.36440942e-02 7.13841736e-01 4.39556211e-01
2.51292318e-01 8.31607819e-01 -2.08591387e-01 -4.34334606e-01
-1.36218059e+00 3.62851232e-01 1.25546169e+00 9.44972157e-01
3.20861369e-01 -3.40779632e-01 -1.44707737e-02 5.72301149e-01
1.74741626e-01 5.23758829e-01 5.55850506e-01 -1.12718439e+00
4.16439325e-01 5.66253841e-01 6.62378371e-02 -1.19215250e+00
-2.29267314e-01 -2.49834627e-01 -9.04357791e-01 -3.38081777e-01
7.61640191e-01 2.14890048e-01 -3.03770334e-01 1.16487944e+00
-1.59876928e-01 -3.54075700e-01 5.14911786e-02 7.67231703e-01
1.27893174e+00 5.40283799e-01 -1.92271829e-01 -2.86107033e-01
1.66960382e+00 -8.37610424e-01 -6.94976568e-01 2.92389423e-01
5.40517688e-01 -1.07476318e+00 7.51648664e-01 4.36875433e-01
-9.14397061e-01 -2.51860648e-01 -7.69582391e-01 -2.44431332e-01
-3.73787850e-01 4.76038486e-01 3.16073835e-01 4.17513490e-01
-4.57392693e-01 8.95418227e-01 -3.86263311e-01 -2.48039216e-01
6.68111920e-01 -1.57643363e-01 -2.68178433e-01 5.48021317e-01
-1.12883520e+00 6.81421578e-01 2.43766695e-01 -1.94805503e-01
-4.14596051e-01 -5.72893620e-01 -5.97172797e-01 -2.21180275e-01
2.22601175e-01 -3.28610331e-01 1.07621157e+00 -1.05457938e+00
-1.14113903e+00 1.74513233e+00 -1.29752293e-01 -2.18279541e-01
9.43724811e-01 2.38020033e-01 -5.84151208e-01 6.54667541e-02
5.98667383e-01 1.93368092e-01 7.36108899e-01 -1.01927555e+00
-4.63067383e-01 -5.45998812e-01 -5.04087389e-01 -4.59002137e-01
-2.62589365e-01 5.47757030e-01 4.30911854e-02 -1.14583230e+00
-1.19073749e-01 -8.62172067e-01 5.89580655e-01 -9.38946530e-02
-9.39183831e-01 -8.52244437e-01 6.01262987e-01 -1.00400150e+00
1.41145194e+00 -1.80265021e+00 2.04690218e-01 4.23963279e-01
6.10864758e-01 -7.19804615e-02 9.90946963e-02 6.01007879e-01
1.68684021e-01 6.41976058e-01 -3.78538549e-01 -4.03303444e-01
1.91670150e-01 5.98714240e-02 -5.99587500e-01 7.84504712e-01
-2.10187659e-01 9.75830793e-01 -8.45359862e-01 -9.02592719e-01
-5.08450031e-01 4.62049991e-02 1.83371216e-01 6.42472729e-02
1.46540910e-01 5.65876722e-01 -4.37621415e-01 8.38187814e-01
3.76692832e-01 -4.58092630e-01 2.11603329e-01 4.74236086e-02
-1.77091062e-01 4.61379915e-01 -9.01807964e-01 1.11534655e+00
9.20565352e-02 1.14239192e+00 -7.84996897e-02 -3.15377533e-01
9.30285990e-01 2.12135032e-01 9.20177475e-02 -4.29953158e-01
7.58037865e-01 7.91122258e-01 3.37582789e-02 -5.06932400e-02
8.63354981e-01 -3.29535268e-02 -4.57374454e-01 1.22389364e+00
-2.42162809e-01 1.30554363e-01 7.72803187e-01 6.38188362e-01
6.91610575e-01 -1.12007275e-01 4.82019156e-01 -4.79198873e-01
8.44984353e-01 -2.87104957e-03 3.74792159e-01 8.38105083e-01
-1.57058641e-01 4.52942520e-01 8.84768546e-01 -4.71284628e-01
-1.54032397e+00 -7.83502221e-01 -7.41847634e-01 8.10697377e-01
-1.98756512e-02 -4.86600012e-01 -3.45458180e-01 -8.20787787e-01
3.18862528e-01 6.58011019e-01 -8.27325463e-01 5.65526724e-01
-8.24867725e-01 -2.32602760e-01 1.35912132e+00 2.25860953e-01
1.89845026e-01 -1.36695814e+00 -7.49136150e-01 -4.96825809e-03
-2.62516230e-01 -7.34852731e-01 -6.41047120e-01 1.28149875e-02
-4.68286067e-01 -1.79180229e+00 -1.11180580e+00 -7.90698111e-01
3.53606343e-01 -3.43936831e-01 1.35702932e+00 4.95305330e-01
-2.44783044e-01 6.26019314e-02 -3.79028767e-01 -2.37663746e-01
-1.00233793e+00 4.11748707e-01 -1.06790662e-02 -1.93572253e-01
4.25344147e-02 -3.81695867e-01 -2.27659824e-03 2.77731150e-01
-7.52467752e-01 -1.32791787e-01 1.29448026e-01 1.04307711e+00
-7.78579153e-03 -6.12009406e-01 4.78793830e-01 -1.42178845e+00
7.71749496e-01 -4.76738691e-01 -4.02264953e-01 1.41974807e-01
-9.05571282e-01 -3.12766549e-03 8.39184999e-01 -4.30786639e-01
-4.75693941e-01 -7.60894120e-01 6.65993765e-02 -2.39356905e-01
6.45893812e-02 4.04512763e-01 1.92445517e-02 1.38537437e-01
4.97429579e-01 9.35158804e-02 -2.09087897e-02 -6.61122382e-01
1.04756482e-01 1.19735777e+00 7.98655808e-01 -6.26831114e-01
7.85210490e-01 9.68345925e-02 8.95372704e-02 -5.64403772e-01
-7.62776673e-01 -2.71484256e-01 -9.23920333e-01 1.34583950e-01
2.57141441e-01 -6.15115643e-01 -1.04535532e+00 6.47190034e-01
-1.45878494e+00 1.09325126e-01 -1.74418420e-01 -7.92105645e-02
-4.59410638e-01 7.21302927e-01 -8.31234038e-01 -4.95304853e-01
-3.68503064e-01 -6.12837136e-01 9.64462519e-01 2.14881375e-01
-9.59910989e-01 -1.29887056e+00 1.60082728e-01 5.91286182e-01
1.56668909e-02 3.72166395e-01 1.23796499e+00 -1.18162608e+00
6.91416208e-03 -1.44987836e-01 -1.83131769e-01 -2.13835865e-01
-9.41734388e-02 3.19618255e-01 -6.87116086e-01 -1.59358352e-01
-4.95486170e-01 -5.63819349e-01 7.55621076e-01 -2.09945053e-01
9.58045125e-01 -7.17439294e-01 -3.03664714e-01 3.88916999e-01
8.94882679e-01 -3.70948613e-01 4.23317254e-01 6.79828703e-01
6.17693186e-01 5.59433639e-01 2.23826915e-01 7.59204268e-01
6.37011945e-01 7.41949916e-01 -1.08943149e-01 1.14126325e-01
-2.11887822e-01 -2.61461735e-01 -7.74836391e-02 8.02441061e-01
-2.54073530e-01 -6.80362165e-01 -1.05412960e+00 5.33674240e-01
-1.79115260e+00 -1.09390891e+00 -8.70672822e-01 1.88437665e+00
1.15616965e+00 -2.52134264e-01 3.80493283e-01 4.66178954e-01
6.85267448e-01 2.10506588e-01 -2.49586403e-01 -7.04531252e-01
-6.89335048e-01 2.11840898e-01 2.45227352e-01 3.46066177e-01
-1.16808474e+00 1.01830995e+00 6.33666277e+00 7.24510074e-01
-8.25176835e-01 7.04575852e-02 4.98445928e-01 3.59076470e-01
-2.80012488e-01 1.18238069e-01 -7.07433522e-01 9.53090250e-01
9.26828146e-01 -4.26563948e-01 2.34546542e-01 2.63672203e-01
-1.30212113e-01 -1.34194329e-01 -1.33434880e+00 1.04339111e+00
4.05879229e-01 -1.32322240e+00 1.81892931e-01 2.72284716e-01
5.70586383e-01 -6.40282512e-01 -1.26097277e-01 -7.24732969e-03
1.41618490e-01 -1.11846566e+00 1.15069091e+00 4.38541830e-01
7.03243673e-01 -5.66045940e-01 9.95925844e-01 4.46987271e-01
-4.37157631e-01 1.93076283e-01 -1.53351203e-01 -3.99980918e-02
1.35816321e-01 4.67758417e-01 -5.05217075e-01 4.59885150e-01
3.55444342e-01 1.20872319e+00 -9.53041911e-01 6.92825437e-01
-7.62684286e-01 9.33129787e-01 1.01623781e-01 -3.58414978e-01
-9.71199721e-02 -1.34976298e-01 8.54048312e-01 1.03677332e+00
3.25906798e-02 2.37835757e-03 -1.41294464e-01 1.13226855e+00
-7.83437610e-01 2.91812837e-01 -2.71261811e-01 -6.42242908e-01
7.53801286e-01 1.20819962e+00 -8.50610971e-01 -5.58985114e-01
3.11912417e-01 1.26636505e+00 2.19627440e-01 -4.32873741e-02
-8.40360999e-01 -6.36320710e-01 8.99817199e-02 3.91273916e-01
2.71258503e-02 4.62368550e-03 -7.11072981e-01 -1.40222991e+00
-1.46124989e-01 -1.03163123e+00 6.24362588e-01 -7.84213185e-01
-1.76774871e+00 6.24509037e-01 -3.47669631e-01 -8.37381721e-01
-1.53779715e-01 -5.03913820e-01 -4.99458551e-01 9.84714270e-01
-1.23802125e+00 -1.36099279e+00 -1.97073873e-02 2.31931165e-01
2.33222544e-01 -5.48325062e-01 5.59784234e-01 6.59206137e-02
-5.43397367e-01 9.53184366e-01 3.85871947e-01 7.12696612e-01
1.19906676e+00 -1.07365358e+00 9.25938338e-02 4.17151570e-01
3.11780721e-01 5.82988381e-01 7.45541155e-01 -1.00088048e+00
-1.21452117e+00 -7.23866463e-01 1.89573979e+00 -9.78490412e-01
1.09334874e+00 -1.79135770e-01 -9.30736125e-01 6.00597978e-01
4.89474535e-01 -6.72823012e-01 7.76579022e-01 1.32852361e-01
-6.28834128e-01 5.22273302e-01 -1.08698034e+00 2.61466891e-01
9.61585581e-01 -3.60720336e-01 -9.28936124e-01 3.52191091e-01
-7.22103461e-04 -1.88220352e-01 -1.19681454e+00 -6.26802817e-02
8.28491151e-01 -6.63089633e-01 4.27465767e-01 -8.31613779e-01
1.15626824e+00 -2.73417011e-02 3.44490588e-01 -8.96864057e-01
-3.52929950e-01 -5.50269008e-01 -1.23874150e-01 1.61206329e+00
3.91554892e-01 -7.76927054e-01 5.83027422e-01 1.50816068e-01
2.11842671e-01 -3.34681600e-01 -1.13552272e+00 -6.09581590e-01
5.42651296e-01 2.77364314e-01 3.50940943e-01 1.48362792e+00
-6.55681863e-02 6.44459546e-01 -3.05574507e-01 -5.42762458e-01
8.08958292e-01 7.19121873e-01 7.97864974e-01 -1.83473325e+00
8.87307599e-02 -1.01701069e+00 -2.10527122e-01 -5.90325058e-01
7.55580187e-01 -1.21811461e+00 -2.35125899e-01 -1.25649965e+00
5.66953123e-01 -6.24738872e-01 4.04367894e-01 7.72304118e-01
5.24819493e-02 7.39240706e-01 2.81742841e-01 1.26161945e+00
-3.18388790e-01 4.87692922e-01 1.16396153e+00 -2.60998905e-01
7.73773491e-02 -2.36691326e-01 -7.52870440e-01 7.28059530e-01
5.10153592e-01 -6.99420989e-01 5.79442143e-01 5.29027842e-02
3.76699835e-01 -2.19505116e-01 5.64662874e-01 -4.84442085e-01
2.82864511e-01 -7.87334070e-02 4.21979755e-01 -5.89525938e-01
-1.76390141e-01 -4.02195513e-01 -2.61654630e-02 5.17511725e-01
-4.96161729e-01 1.45937085e-01 -2.79797018e-01 1.93578601e-01
-1.27300635e-01 -4.69821066e-01 7.71556318e-01 3.72774936e-02
-3.10856998e-01 -1.23836799e-03 -6.94824517e-01 4.43930656e-01
6.99386418e-01 -3.52457285e-01 -7.67670691e-01 -1.59548372e-01
-3.86138082e-01 1.36502326e-01 1.09668148e+00 3.69823128e-01
3.49501014e-01 -1.09719801e+00 -1.27540600e+00 -1.36723854e-02
1.87533513e-01 -6.98303461e-01 -4.35460657e-01 9.54223812e-01
-6.77199006e-01 3.75839740e-01 -3.23255777e-01 -1.96081027e-01
-1.36108220e+00 4.98086780e-01 -2.01196223e-01 -2.04413280e-01
-6.38517082e-01 2.71021008e-01 -4.25538719e-01 -4.97076660e-01
1.40876681e-01 2.96227932e-01 -3.88341844e-01 4.66049731e-01
3.20218235e-01 8.41932118e-01 -2.84387738e-01 -1.17535710e+00
-5.17425895e-01 3.65350664e-01 1.28390074e-01 -7.95649588e-02
1.16988838e+00 -8.21782500e-02 -7.48700619e-01 9.42556441e-01
8.64946663e-01 7.33786106e-01 -1.69776365e-01 -1.82389885e-01
2.41230473e-01 -3.87045860e-01 -3.49563509e-01 -1.32100260e+00
-6.27440095e-01 7.99034953e-01 -9.49998852e-03 1.66313231e-01
3.90638471e-01 3.86725724e-01 6.86964214e-01 1.56405523e-01
2.55478412e-01 -8.45423460e-01 -9.67952684e-02 8.58499944e-01
9.91859913e-01 -1.12132454e+00 3.06483388e-01 -2.24722326e-01
-4.56841379e-01 1.55164111e+00 1.71998471e-01 1.05392717e-01
3.69930983e-01 -2.36370750e-02 1.54774547e-01 -3.16667438e-01
-3.57216060e-01 3.07933301e-01 4.51277614e-01 -7.18085989e-02
6.73089802e-01 1.95316300e-01 -9.82053220e-01 1.00708258e+00
-8.07943761e-01 -2.86091864e-01 1.09048581e+00 8.24171424e-01
-1.01025112e-01 -1.33477831e+00 -6.46591365e-01 5.16196311e-01
-3.26866865e-01 -2.23268151e-01 -1.34818470e+00 9.08860981e-01
2.26508036e-01 7.59543419e-01 1.47484139e-01 4.26474661e-02
-1.32967666e-01 -7.21926019e-02 6.14107788e-01 -2.10231915e-01
-1.11606956e+00 -3.11708927e-01 3.25780421e-01 3.71635497e-01
-3.95560294e-01 -8.34591627e-01 -1.06896758e+00 -1.14288437e+00
-1.33297518e-01 6.33311570e-01 4.55721498e-01 9.05921102e-01
1.34183660e-01 1.11111119e-01 5.45166910e-01 -3.77448112e-01
-4.19631273e-01 -1.16346979e+00 -9.87780690e-01 5.61905801e-01
2.28051580e-02 -5.23118854e-01 -5.02937317e-01 2.21718028e-01] | [9.563549995422363, 10.570634841918945] |
6b494ab9-368e-4149-b6fd-84e206525798 | turl-table-understanding-through | 2006.14806 | null | https://arxiv.org/abs/2006.14806v2 | https://arxiv.org/pdf/2006.14806v2.pdf | TURL: Table Understanding through Representation Learning | Relational tables on the Web store a vast amount of knowledge. Owing to the wealth of such tables, there has been tremendous progress on a variety of tasks in the area of table understanding. However, existing work generally relies on heavily-engineered task-specific features and model architectures. In this paper, we present TURL, a novel framework that introduces the pre-training/fine-tuning paradigm to relational Web tables. During pre-training, our framework learns deep contextualized representations on relational tables in an unsupervised manner. Its universal model design with pre-trained representations can be applied to a wide range of tasks with minimal task-specific fine-tuning. Specifically, we propose a structure-aware Transformer encoder to model the row-column structure of relational tables, and present a new Masked Entity Recovery (MER) objective for pre-training to capture the semantics and knowledge in large-scale unlabeled data. We systematically evaluate TURL with a benchmark consisting of 6 different tasks for table understanding (e.g., relation extraction, cell filling). We show that TURL generalizes well to all tasks and substantially outperforms existing methods in almost all instances. | ['Xiang Deng', 'Alyssa Lees', 'You Wu', 'Huan Sun', 'Cong Yu'] | 2020-06-26 | null | null | null | null | ['table-annotation', 'table-annotation', 'column-type-annotation', 'cell-entity-annotation', 'columns-property-annotation'] | ['knowledge-base', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 1.11531444e-01 4.93525803e-01 -5.18340588e-01 -6.49158120e-01
-1.01937532e+00 -6.65919721e-01 4.73988652e-01 6.47385478e-01
1.12258121e-01 7.39734888e-01 4.76867795e-01 -5.35251617e-01
7.00143129e-02 -1.22616231e+00 -1.28608227e+00 -1.76070128e-02
8.00512824e-03 9.83812749e-01 2.62815952e-01 -5.18415451e-01
-1.57825157e-01 2.78013170e-01 -1.51068246e+00 8.43145490e-01
7.36411214e-01 1.21871316e+00 -9.58948508e-02 2.89761633e-01
-5.31039417e-01 1.27347410e+00 -3.59843642e-01 -1.04013872e+00
1.22355424e-01 -3.44646610e-02 -1.20482254e+00 -1.73507421e-03
2.94035047e-01 -1.31024227e-01 -6.24188125e-01 7.02846467e-01
1.67920992e-01 -2.89826654e-02 2.42591530e-01 -9.97993171e-01
-9.29091632e-01 1.35537791e+00 -3.15364182e-01 -1.46171022e-02
1.42101273e-01 -3.26926172e-01 1.42169118e+00 -9.63701427e-01
8.08514237e-01 1.32211041e+00 7.88000226e-01 3.09151202e-01
-1.36029923e+00 -3.40429723e-01 2.47988850e-01 9.84463319e-02
-1.30485785e+00 -5.62953830e-01 3.18742156e-01 -3.08761567e-01
1.30334997e+00 5.27062304e-02 1.34841666e-01 7.70901740e-01
-8.67060665e-03 8.55110824e-01 6.79671586e-01 -2.36630604e-01
6.43128203e-03 3.58348280e-01 1.74153075e-01 6.44316554e-01
6.87344491e-01 -5.76821327e-01 -6.19654179e-01 1.51404157e-01
7.01273978e-01 1.42822340e-02 6.77412972e-02 -8.24735701e-01
-1.24043131e+00 6.72523737e-01 5.75635433e-01 8.55439529e-02
-9.86136720e-02 1.24255806e-01 7.98415303e-01 2.72488683e-01
3.94421518e-01 4.51770663e-01 -1.07891524e+00 -1.24068093e-02
-3.28682303e-01 2.77565032e-01 1.15960598e+00 1.41995597e+00
9.07747209e-01 -3.30960631e-01 -2.19639316e-01 8.52818370e-01
6.38876483e-02 2.62944818e-01 1.13530628e-01 -5.61404645e-01
1.21552563e+00 1.10472727e+00 -1.18020110e-01 -7.90553749e-01
-3.02717298e-01 -3.74685496e-01 -7.71017730e-01 -4.76762563e-01
3.49238038e-01 2.58877516e-01 -1.00335085e+00 1.62532055e+00
3.52721632e-01 -4.47244614e-01 4.22544926e-01 4.08578247e-01
1.08164477e+00 4.60300565e-01 1.15442604e-01 2.00252205e-01
1.60306311e+00 -9.21645939e-01 -9.30773735e-01 -5.52487493e-01
8.61933529e-01 -4.66816366e-01 1.15403318e+00 3.33018485e-03
-1.05837107e+00 -4.85956311e-01 -1.10213006e+00 -8.99491131e-01
-1.03377473e+00 1.61419466e-01 1.12076104e+00 6.26643598e-01
-7.76299655e-01 4.98622864e-01 -8.86756361e-01 -4.95145679e-01
6.35235131e-01 5.52114964e-01 -7.10071385e-01 -2.26007044e-01
-1.33245385e+00 8.87848318e-01 5.97651899e-01 2.47001991e-01
-4.73247498e-01 -8.83247674e-01 -1.31415665e+00 3.45626026e-01
1.02206349e+00 -7.05244482e-01 1.23318183e+00 -6.88468665e-02
-1.03776276e+00 1.11849701e+00 -3.27982664e-01 -7.81957567e-01
2.27766171e-01 -4.16327864e-01 -3.04531544e-01 -3.89574587e-01
3.79250467e-01 3.82203072e-01 5.25584444e-02 -1.38176298e+00
-5.20997107e-01 -4.23581004e-01 2.78247386e-01 1.46355927e-01
-9.89115015e-02 -1.53832540e-01 -9.79512095e-01 -6.61222160e-01
1.57469273e-01 -4.70065355e-01 -1.28181010e-01 -3.77954662e-01
-8.45630586e-01 -1.65079102e-01 2.71667361e-01 -5.76950073e-01
1.49172163e+00 -1.93811882e+00 -2.31337845e-02 5.72839119e-02
5.21844864e-01 -7.54532823e-03 2.60371417e-01 7.27152646e-01
-4.76454943e-02 2.89632440e-01 -4.88606662e-01 -3.82133007e-01
3.34335238e-01 4.00651872e-01 -6.96129501e-01 -2.21733361e-01
5.13239384e-01 1.32627642e+00 -6.28986239e-01 -5.38807929e-01
2.57041696e-02 2.98993528e-01 -7.31603920e-01 2.61195958e-01
-6.22308433e-01 -4.13551778e-02 -3.63402784e-01 9.41480100e-01
7.06200898e-01 -8.00124288e-01 6.18861854e-01 -3.27119887e-01
3.46659750e-01 9.57862377e-01 -1.20492339e+00 1.76374531e+00
-5.37053823e-01 3.28026801e-01 -2.04963103e-01 -9.02446985e-01
9.04930770e-01 9.12421644e-02 3.23617756e-01 -6.31816924e-01
-6.27061799e-02 4.14079666e-01 -2.17987210e-01 -2.14642346e-01
7.60562241e-01 2.09837884e-01 -4.48743016e-01 3.40116352e-01
3.45778763e-01 -4.83164011e-04 4.04547989e-01 4.58707333e-01
1.21543777e+00 1.62811637e-01 6.27354264e-01 -8.08732137e-02
4.90680337e-01 1.15102068e-01 4.61218297e-01 6.92056119e-01
3.58136415e-01 3.92771870e-01 8.57932329e-01 -6.07918680e-01
-9.23740685e-01 -1.01035368e+00 -2.46125489e-01 1.28232253e+00
1.95196837e-01 -9.05707181e-01 -6.69795096e-01 -8.69472563e-01
5.86353660e-01 3.92435193e-01 -9.49512064e-01 -1.43595904e-01
-7.81048179e-01 -7.39821136e-01 3.59103829e-01 1.04254234e+00
4.55553710e-01 -1.16786015e+00 9.71201882e-02 4.39598441e-01
-6.41993508e-02 -1.56281471e+00 -1.41889989e-01 8.03904474e-01
-9.15907443e-01 -1.07852817e+00 2.22012281e-01 -8.21641862e-01
4.60055798e-01 7.34680295e-02 1.91410422e+00 3.80550660e-02
-7.65420794e-02 -7.04760756e-03 -1.92295372e-01 -3.43941599e-01
-2.78584808e-01 7.15750396e-01 -3.81680399e-01 -2.06964210e-01
8.02810490e-01 -3.87342900e-01 -6.01644590e-02 2.90750086e-01
-8.61388624e-01 1.58725858e-01 6.71116292e-01 7.41636992e-01
1.00213158e+00 5.72444908e-02 4.78688687e-01 -1.86186683e+00
3.10927927e-01 -6.15435183e-01 -6.29297853e-01 6.53877437e-01
-6.17588878e-01 6.03683889e-01 6.05360031e-01 9.87131521e-02
-9.88743365e-01 -1.01363197e-01 -1.63215831e-01 -1.61332376e-02
1.49627745e-01 7.61501193e-01 -9.32187557e-01 5.24266601e-01
4.30855185e-01 -8.54076222e-02 -4.08498943e-01 -7.47708201e-01
6.48486853e-01 2.84058660e-01 7.22449780e-01 -9.23476338e-01
9.29636300e-01 4.92055446e-01 -9.25673842e-02 -2.72597875e-02
-1.38519442e+00 -6.02647290e-02 -8.83116066e-01 6.12327635e-01
7.65007257e-01 -1.25899887e+00 -7.91495860e-01 5.98649196e-02
-7.72425652e-01 -5.07539392e-01 -6.10875607e-01 3.83645184e-02
-5.44895649e-01 -2.44526252e-01 -7.80159712e-01 -3.86416942e-01
-2.32589230e-01 -1.18339920e+00 1.21178544e+00 -7.42660612e-02
-8.40034187e-02 -9.30868983e-01 -1.11237496e-01 4.62187380e-01
1.94192365e-01 1.34817466e-01 1.29533601e+00 -9.95003164e-01
-1.26830637e+00 -4.09597084e-02 -5.95063567e-01 -4.34938073e-02
2.51277685e-01 -2.74392784e-01 -9.77698803e-01 1.78591043e-01
-3.99307042e-01 -7.73760259e-01 1.09862626e+00 -7.47731179e-02
1.39583230e+00 -1.81521609e-01 -5.24804413e-01 9.12499905e-01
1.39669895e+00 1.60201684e-01 7.17238843e-01 5.62183082e-01
1.15323973e+00 6.15268648e-01 6.23045564e-01 1.87502906e-01
1.05247056e+00 5.55284619e-01 3.78282040e-01 1.60024054e-02
-4.55904845e-03 -8.55626345e-01 -2.47000560e-01 6.67017162e-01
3.03610027e-01 -3.22737992e-01 -9.52699065e-01 5.50824523e-01
-1.96278048e+00 -6.22175872e-01 1.91419065e-01 1.99624038e+00
1.21924412e+00 4.64182168e-01 -1.20754443e-01 2.35284314e-01
5.52176833e-01 2.19896957e-01 -6.73687696e-01 -2.18422964e-01
-2.14754939e-01 4.36474711e-01 6.71238005e-01 3.44450116e-01
-1.40090847e+00 1.32449377e+00 5.88788366e+00 5.56162775e-01
-5.85912526e-01 -1.56755939e-01 8.60665321e-01 3.37552011e-01
-6.78283870e-01 3.99382487e-02 -1.32979643e+00 -4.18461971e-02
9.88729656e-01 -2.32703641e-01 4.64499325e-01 9.74075735e-01
-6.17909789e-01 1.92859039e-01 -1.64498019e+00 9.05434608e-01
-1.42516479e-01 -1.76075542e+00 2.32775122e-01 -8.37550405e-03
4.98359889e-01 -1.65246710e-01 1.01896577e-01 8.07654798e-01
6.38189673e-01 -1.28277922e+00 6.38360381e-01 1.39807239e-01
1.15229130e+00 -7.40485668e-01 8.76250565e-01 -2.48219281e-01
-1.63976860e+00 -2.90406048e-02 -3.98401678e-01 1.92560747e-01
-1.60575405e-01 5.29327154e-01 -6.75521791e-01 7.40445733e-01
7.49423265e-01 1.03673458e+00 -7.92804122e-01 3.74601156e-01
-2.06124544e-01 2.67991900e-01 -2.76245981e-01 2.76792049e-01
-9.16086733e-02 1.00195795e-01 -3.38440686e-01 1.16400123e+00
-6.82356805e-02 -5.46091124e-02 -2.88418983e-03 8.12355340e-01
-9.10879552e-01 7.59239793e-02 -6.47976696e-01 -2.64480501e-01
7.12939441e-01 1.06455147e+00 -7.78800726e-01 -5.11267304e-01
-7.00877190e-01 4.68425810e-01 8.76497984e-01 1.22665726e-01
-6.20489478e-01 -3.78597021e-01 7.84341514e-01 1.80586517e-01
7.47155905e-01 8.58977214e-02 -8.65342319e-01 -1.39688492e+00
3.77460897e-01 -1.04407632e+00 6.76557481e-01 -5.06932914e-01
-1.22203219e+00 8.93882275e-01 -2.50812043e-02 -7.41389573e-01
-3.96553546e-01 -7.09515214e-01 2.73866318e-02 8.45121324e-01
-1.53810751e+00 -1.10924923e+00 -2.91255832e-01 5.93358278e-01
3.12839657e-01 -1.71559140e-01 9.08946753e-01 4.28927004e-01
-6.21719182e-01 8.38185608e-01 -1.92251038e-02 7.02436984e-01
7.54161596e-01 -1.65636134e+00 1.09873879e+00 7.52954543e-01
2.29556382e-01 1.03496552e+00 3.78556609e-01 -6.61316276e-01
-1.77727187e+00 -1.28436911e+00 1.21774566e+00 -9.40323710e-01
8.88845205e-01 -1.07312036e+00 -1.12027490e+00 1.43437421e+00
-1.05468817e-01 4.39417601e-01 5.96980214e-01 7.61562049e-01
-8.87516558e-01 -5.36564708e-01 -9.07248437e-01 4.52336222e-01
1.23832166e+00 -8.90714765e-01 -5.22741437e-01 2.25991711e-01
1.17715991e+00 -8.24390113e-01 -1.29567802e+00 6.76903486e-01
4.13618863e-01 -7.40593076e-01 1.14935076e+00 -6.70560241e-01
6.77804053e-01 -2.38091871e-01 -4.55122113e-01 -7.87435293e-01
-3.12268846e-02 -4.76132691e-01 -7.28724241e-01 1.43939066e+00
8.36412013e-01 -4.37855393e-01 1.06204450e+00 7.64517605e-01
-1.44829541e-01 -1.05803537e+00 -3.74998271e-01 -2.78361917e-01
2.04956578e-03 -3.99264008e-01 1.11697114e+00 7.78849423e-01
-2.00171606e-03 5.89837790e-01 -1.84820965e-01 1.52614921e-01
3.79930288e-01 5.16699195e-01 1.10971713e+00 -1.23411787e+00
-2.61191040e-01 -8.06159452e-02 -1.34275496e-01 -1.28327060e+00
-1.30211599e-02 -1.06194556e+00 -1.70362100e-01 -1.85561740e+00
2.83704042e-01 -7.07096934e-01 -3.65921527e-01 6.90643489e-01
-4.58616138e-01 -5.37985750e-02 2.56335437e-02 1.56651244e-01
-8.11404288e-01 3.01600367e-01 1.05121934e+00 -1.35102049e-01
-9.83191803e-02 -6.69520795e-02 -1.34887803e+00 2.27792293e-01
5.22198379e-01 -5.38328588e-01 -6.41680539e-01 -6.29845440e-01
7.25025475e-01 7.49623775e-02 -1.23816334e-01 -7.44237065e-01
5.51647507e-02 1.28220350e-01 4.33356136e-01 -8.83446217e-01
2.83029914e-01 -7.99047589e-01 -9.29229259e-02 -5.86595573e-02
-5.83715260e-01 2.39577472e-01 3.40544790e-01 5.71634114e-01
-6.28912330e-01 3.04085791e-01 3.95111948e-01 -2.49738514e-01
-6.18431985e-01 5.55634856e-01 2.71210521e-01 6.31885350e-01
6.12932205e-01 1.10182300e-01 -4.88931239e-01 -1.00689381e-01
-7.31612384e-01 3.75641763e-01 4.16379780e-01 4.59123731e-01
3.00794780e-01 -1.17900968e+00 -3.70507628e-01 4.20605272e-01
5.66706300e-01 5.54820895e-01 -3.15601751e-02 2.99725950e-01
-6.09560192e-01 7.74218380e-01 1.93343349e-02 -2.61971802e-01
-6.50301099e-01 9.77932155e-01 1.89055592e-01 -1.10021102e+00
-7.43209362e-01 8.98097694e-01 3.24485362e-01 -8.42947662e-01
4.20024455e-01 -7.82246232e-01 -1.37387097e-01 1.45520344e-02
4.54050839e-01 -3.20958316e-01 6.12935603e-01 -2.69912422e-01
-4.04197693e-01 1.06074370e-01 -5.69768846e-01 3.69406015e-01
1.39019907e+00 -5.89017011e-02 -1.42271057e-01 6.51304960e-01
1.04873276e+00 6.41822889e-02 -1.09755945e+00 -6.37834370e-01
5.93447685e-01 -2.48894304e-01 -4.64328706e-01 -9.61530268e-01
-1.14349508e+00 8.29746783e-01 -2.10221961e-01 1.57260284e-01
9.55794692e-01 2.14124739e-01 8.23178530e-01 9.11800265e-01
5.53194523e-01 -7.56338716e-01 -2.59345293e-01 7.48177230e-01
5.60745656e-01 -1.46225679e+00 1.04290145e-02 -8.70689690e-01
-8.65226865e-01 8.91657591e-01 7.66235948e-01 1.10432893e-01
7.85505533e-01 7.48923182e-01 -4.81105410e-02 -2.38980204e-01
-1.10692704e+00 -4.71084982e-01 1.38734728e-01 6.16298556e-01
7.64756918e-01 -5.08211413e-03 2.38231316e-01 9.65803862e-01
-4.26672548e-01 -5.51376939e-02 2.74390668e-01 1.09034562e+00
-1.79804847e-01 -1.29332197e+00 -1.56139033e-02 7.00756729e-01
-5.15660942e-01 -4.75592047e-01 -3.89969349e-01 1.11401331e+00
-5.33336625e-02 5.96435666e-01 1.67856917e-01 -3.66495579e-01
5.76001704e-01 2.14785174e-01 3.60719383e-01 -9.30839360e-01
-7.12578952e-01 -3.56523395e-01 3.91148895e-01 -5.98693609e-01
-9.92567390e-02 -4.41004694e-01 -1.30716383e+00 -5.62483907e-01
1.01821288e-01 2.40386635e-01 2.68015087e-01 7.45532095e-01
5.42012632e-01 7.64329076e-01 1.73887536e-01 -1.22479253e-01
-1.98714554e-01 -8.44429374e-01 -6.07286811e-01 5.10035038e-01
9.77603644e-02 -8.65726590e-01 1.54047340e-01 4.78974469e-02] | [9.556639671325684, 7.882285118103027] |
6d4eafca-e6e3-4774-b3e3-03b4dbceedc7 | dynamic-contrastive-distillation-for-image | 2207.01426 | null | https://arxiv.org/abs/2207.01426v1 | https://arxiv.org/pdf/2207.01426v1.pdf | Dynamic Contrastive Distillation for Image-Text Retrieval | Although the vision-and-language pretraining (VLP) equipped cross-modal image-text retrieval (ITR) has achieved remarkable progress in the past two years, it suffers from a major drawback: the ever-increasing size of VLP models restricts its deployment to real-world search scenarios (where the high latency is unacceptable). To alleviate this problem, we present a novel plug-in dynamic contrastive distillation (DCD) framework to compress the large VLP models for the ITR task. Technically, we face the following two challenges: 1) the typical uni-modal metric learning approach is difficult to directly apply to the cross-modal tasks, due to the limited GPU memory to optimize too many negative samples during handling cross-modal fusion features. 2) it is inefficient to static optimize the student network from different hard samples, which have different effects on distillation learning and student network optimization. We try to overcome these challenges from two points. First, to achieve multi-modal contrastive learning, and balance the training costs and effects, we propose to use a teacher network to estimate the difficult samples for students, making the students absorb the powerful knowledge from pre-trained teachers, and master the knowledge from hard samples. Second, to dynamic learn from hard sample pairs, we propose dynamic distillation to dynamically learn samples of different difficulties, from the perspective of better balancing the difficulty of knowledge and students' self-learning ability. We successfully apply our proposed DCD strategy to two state-of-the-art vision-language pretrained models, i.e. ViLT and METER. Extensive experiments on MS-COCO and Flickr30K benchmarks show the effectiveness and efficiency of our DCD framework. Encouragingly, we can speed up the inference at least 129$\times$ compared to the existing ITR models. | ['DaCheng Tao', 'Li Shen', 'Yang Liu', 'Meng Fang', 'Shuhan Qi', 'Liang Ding', 'Jun Rao'] | 2022-07-04 | null | null | null | null | ['self-learning'] | ['natural-language-processing'] | [-0.05796929 -0.48444325 -0.22110352 -0.27489424 -1.1423652 -0.52051276
0.37745512 -0.03944975 -0.8537331 0.24200277 -0.23558281 -0.31611317
-0.29378945 -0.6972364 -0.655511 -0.7423844 0.4302176 0.7894943
0.2988333 -0.24610098 0.16183321 0.39865834 -1.8157021 0.2090088
1.286368 1.0939223 0.4771063 0.61967736 -0.26650375 0.60243434
-0.42910078 -0.65594226 0.10865966 -0.07292684 -0.6744899 -0.34247914
0.7971404 -0.4386918 -0.42142442 1.1964276 0.9159772 0.317314
0.5527018 -1.2303994 -0.8616377 0.54360366 -0.9638827 0.07186306
0.15463804 0.1846711 0.94672364 -1.1696318 0.23590437 1.4373505
0.40283653 0.81678677 -0.9553588 -0.88453823 0.28789 0.5787028
-1.5813842 -0.19127749 0.8420657 -0.16406281 0.6328892 0.32092142
0.693323 1.0268352 -0.37057188 1.4307251 1.1935042 -0.43581828
-0.22293524 0.35855353 0.29979306 0.9749983 -0.16116378 -0.00629845
-0.527768 0.0477905 0.42874843 0.09448268 -0.41941932 -0.42115414
-1.0141746 0.6701831 0.3707786 0.26498166 0.26092353 0.05598029
0.335642 0.40917948 0.34697366 0.1481879 -0.48495898 -0.21688119
-0.99337035 0.07163549 0.5089769 0.7591741 0.891313 -0.09907147
-0.252092 1.2419301 0.4935663 0.7724714 0.8472088 -0.6228548
0.7382934 0.6704481 -0.44121096 -0.89927274 0.05046045 -0.2851126
-0.9357787 -0.04436896 0.35154587 0.10743064 -1.0934244 1.5887656
0.4259038 0.35175455 0.14140955 0.91192186 1.1659348 0.7050324
-0.04642782 -0.09615894 1.311216 -1.3424644 -0.4067226 -0.3217571
0.7280895 -0.9268473 1.4748024 0.55079156 -1.281614 -0.73549604
-0.8846218 -0.42947447 -0.5348286 0.21979244 0.43109792 0.48643324
-0.9236952 0.23039113 -0.5079911 0.03346334 0.31038657 0.37560645
-0.03217777 -0.49184093 -1.1230118 0.6660806 0.4181908 0.21260782
-1.097553 -1.0190647 -0.72327864 0.32960325 0.58095866 -0.4772642
1.0804676 -0.82984483 -1.5541579 0.9355442 0.06846058 -0.04005688
0.71617997 -0.314662 -0.09793454 -0.01020984 -0.26054126 0.7864055
0.8311399 -1.182777 -0.57427156 -0.46500558 0.11802793 0.6085381
-0.8810236 -0.23131427 -1.1649492 -0.5527808 0.06534569 -0.85366756
-0.07329777 -0.10665161 -0.06817574 -0.6829511 0.8385972 -0.3338778
1.371991 -2.055562 0.27382332 0.07403838 0.24372908 0.7106027
-0.4822365 -0.1021665 0.10109974 -0.06428542 0.00959095 -0.58009344
0.05966891 0.28834775 -0.37390974 0.11205696 -0.01884542 1.0083877
-0.93401814 -0.9392225 0.25355697 0.45159075 -0.69921744 0.37284583
-0.21629046 0.11172008 -0.48438603 0.7200642 0.78786063 -0.14250377
-0.0624486 -0.25142106 0.16901247 0.02197608 -1.3433127 1.760638
-0.71412396 0.38484475 0.16960438 -1.3485091 0.6590516 0.01822453
0.17393045 -1.0486685 0.09517905 0.24610806 -0.25452045 -0.53940094
0.5442734 0.15780866 0.00807828 0.27886832 0.15434775 -0.23166241
0.17061785 0.1683922 0.59461075 -0.0179291 -0.28064758 -0.05605749
0.93665445 -0.29079401 0.4862231 0.81144536 -0.23678702 0.377717
0.13517138 -0.28202453 -0.4391064 -1.0434583 0.01144063 1.5225638
0.31051114 -0.303822 -0.56221527 -0.86656123 -0.19064526 0.41428742
-0.21534778 -0.2364253 -0.7375524 -0.8358271 0.5310386 0.4207573
0.57732385 -0.876976 -0.19115837 -0.15252231 -0.28241363 -1.1016642
-0.63651454 0.03303684 -0.57848847 -0.8425249 -0.880425 -1.0877306
0.6615203 0.6892247 1.2604389 0.33519277 -0.4586075 0.64953756
-0.03365713 -0.21306375 -0.26878417 0.15176018 0.00737655 -0.12057487
0.51631427 -0.3157172 -0.631628 0.22912139 -0.93814105 0.15324079
0.633147 1.0657371 0.93581307 0.21163891 0.17035884 -0.4665008
0.70373267 -0.16506553 -0.78900355 0.80561644 -0.86154395 -0.03072117
0.636123 -0.8516485 -0.95265734 -0.15471032 -0.16261174 -0.96355027
0.16564654 0.48231104 -0.05751687 -0.28309116 0.22731215 0.5464391
-0.15621977 -0.4012833 0.46584615 0.59806067 0.4448942 -1.0955441
0.94258326 0.13548419 -0.2323233 -0.64731467 -1.1176528 -0.5654083
-0.19786695 -0.13448699 0.80182296 -1.1940114 -1.1031481 0.70730567
-0.89665383 -0.42977962 -0.10588023 0.458867 -0.21851392 0.535666
-0.5648414 -0.703028 -0.58096534 -1.6851726 0.970772 0.5323975
0.55939686 -1.0085922 0.03205666 0.8853362 0.33173186 -0.49343538
0.966013 -0.53145385 -0.70287544 0.05421817 -0.46826464 0.54572326
-0.32538322 -0.04043623 -0.9413311 -0.6015255 -0.10261834 -1.0847993
0.98817945 -0.04014502 1.719135 -0.14900765 -0.21641411 0.68236095
1.3547542 -0.1421815 0.31125274 0.10883329 0.97637236 0.40668893
0.7876288 0.03668442 0.71013534 0.72471243 0.32969284 -0.19570838
-0.18498157 -0.31869608 0.42486924 1.4160234 0.19975118 -0.17805098
-1.0271366 0.6996312 -1.785403 -0.63250345 0.18670914 2.1628354
1.255996 0.01229241 -0.2543282 -0.14769295 0.33804032 0.18445052
-0.5129864 -0.29320347 0.17235614 0.39814577 0.23723501 0.60630894
-0.8637352 1.0929215 5.0934496 1.5373473 -1.2132574 0.09157961
0.783846 -0.2504318 -0.4143189 -0.2956532 -1.2661369 0.3291755
0.72017604 0.11624403 0.76984906 1.0056001 -0.25253323 0.03904514
-1.0836546 1.4301349 0.36849886 -1.06348 0.37030813 -0.20474114
0.7655799 0.08460911 0.4784333 0.9661369 0.37024397 -0.9108423
0.34443113 0.31847575 0.67462444 -0.89217806 0.49233785 0.63444144
-1.1444991 -0.11249682 -0.43106526 0.42223695 -0.31454095 0.4806504
-0.6881368 0.4971841 0.97460854 0.33953193 -0.89182574 0.6955012
-0.08336572 0.45009807 -0.34030342 -0.11276641 0.36887 -0.39124647
0.23408958 1.0857463 0.3360298 -0.08025652 0.43305132 0.7018116
-0.2566074 0.12196522 -0.24026595 -0.05624524 0.5164786 1.4158263
-0.24472512 -0.4602915 -0.5574852 0.95439017 0.7575528 0.26359725
-0.95777553 -0.23073646 0.5453549 -0.22552578 0.22404239 -0.05368662
-0.00882664 -1.5196884 0.08290502 -1.2544887 0.57005304 -0.629647
-1.4489503 0.3041861 -0.0385102 -1.021283 0.02469996 -0.5398357
-0.36819556 0.80940783 -2.0423532 -1.1449475 -0.39693663 0.90701824
0.6815223 -0.13922438 0.5167939 0.6633607 -0.752405 1.2367166
0.1058284 0.12768072 0.9655843 -1.2617929 -0.09787458 0.64757633
0.35821646 0.593671 0.18650594 -0.0776747 -1.7925589 -0.9802695
0.66154826 -0.3786204 0.80041033 -0.34543455 -0.9735178 0.36526117
0.11882553 0.13559839 0.67872244 0.10508601 -0.5032155 -0.45616645
-0.8614857 0.81056297 0.7960898 -0.6421119 -0.6979255 0.45848867
0.8207604 -0.63054425 -0.8450647 0.5826973 0.47795624 -0.7162372
1.3582134 -0.34804633 0.4452023 -0.13216399 -0.12799627 -1.0920522
-0.02011259 -0.3014051 -0.3430019 1.5547044 0.06845476 -0.45128372
0.71899587 0.43651024 0.04396444 -1.1823442 -0.90625924 -0.52614164
0.30569965 -0.4161531 0.4040637 1.0389904 -0.4467513 0.47239706
-0.34375092 0.10461069 0.5628216 0.33390138 0.799302 -1.1515158
-0.6669302 -0.45424166 0.16045351 -1.5953734 0.24286942 -0.9050288
0.02884161 -1.0793294 0.52925515 -0.8498753 -0.53421813 0.4988731
-0.5748083 0.2756331 0.32802573 0.04301722 -0.85746896 0.67791206
1.5923569 -0.42286333 -0.15954861 -0.19693778 -0.5048197 0.7021325
0.34591964 -0.31603715 -0.7560793 -0.8286249 0.33584163 0.06437736
0.33309865 -0.81805176 0.42831028 -0.22230396 0.23963344 -0.95857996
0.454512 -0.9560792 -0.45941994 0.40860343 -0.27842933 0.01767774
0.1480083 0.51406765 -0.23063199 -0.45344815 0.88567156 -0.21188213
-0.6936764 0.48550126 0.13186826 0.40967682 0.8365443 0.05625394
-0.47218347 -0.08963209 -0.31347835 0.8079416 0.25917068 0.6016448
0.7284663 -1.315432 -0.3979748 0.24258927 0.07378446 0.48759133
0.48609945 0.8765302 -0.29682752 0.21810727 0.20877416 -0.8451401
-1.5379375 0.8528164 0.16190438 -0.84448093 -0.2689263 1.0649812
0.36559856 -0.7750564 0.6781868 -0.06245569 -0.15274851 0.21189556
0.625474 0.33527765 0.01717603 -0.31020898 -0.20768839 0.8386509
-0.5820782 0.18353899 1.1400634 -0.06329132 -0.06496132 0.23029771
1.3944608 -0.18313026 -0.9435532 -0.4510904 -0.33951083 -0.2745476
0.25950536 -0.7359249 -1.3330077 1.2159923 0.9164765 -0.10546842
1.2217337 -0.12015902 1.0329442 0.70564675 0.12804519 -1.28889
0.4552139 0.5156725 0.4794034 -1.4395317 -0.10328013 -0.23337725
-0.39234114 0.94409806 1.0721368 0.09966177 0.50791276 0.0177197
0.02826372 -0.0643512 -0.7762175 -0.11752746 0.53840816 0.17219779
0.2916886 -0.12670505 -0.22425067 0.41805136 -0.04543702 -0.17525741
0.14519031 0.6401363 -0.32861364 -1.0873214 -0.27398393 0.15228826
-0.32135785 -0.30117357 -0.10423599 0.7069013 0.19148299 0.70798117
-0.14870727 -0.28260145 0.277416 0.08172166 0.5220525 -0.3380627
-0.6586488 -0.03184038 -0.40494153 -0.55848414 -0.26917765 -0.1343037
-0.9939018 -0.16568397 -0.46355876 0.1381609 0.49756142 0.90231746
0.210529 0.563801 0.5507286 -0.6415087 -1.037936 -0.6449818
-0.08331463 0.28627133 0.19886371 -0.5773166 -0.45766595 -0.25652552] | [10.654106140136719, 1.4410185813903809] |
56c47426-b2a8-4d1c-8b7b-2b658533f09f | natural-image-stitching-using-depth-maps | 2202.06276 | null | https://arxiv.org/abs/2202.06276v2 | https://arxiv.org/pdf/2202.06276v2.pdf | Natural Image Stitching Using Depth Maps | Natural image stitching (NIS) aims to create one natural-looking mosaic from two overlapping images that capture the same 3D scene from different viewing positions. Challenges inevitably arise when the scene is non-planar and the camera baseline is wide, since parallax becomes not negligible in such cases. In this paper, we propose a novel NIS method using depth maps, which generates natural-looking mosaics against parallax in both overlapping and non-overlapping regions. Firstly, we construct a robust fitting method to filter out the outliers in feature matches and estimate the epipolar geometry between input images. Then, we draw a triangulation of the target image and estimate multiple local homographies, one per triangle, based on the locations of their vertices, the rectified depth values and the epipolar geometry. Finally, the warping image is rendered by the backward mapping of piece-wise homographies. Panorama is then produced via average blending and image inpainting. Experimental results demonstrate that the proposed method not only provides accurate alignment in the overlapping regions but also virtual naturalness in the non-overlapping region. | ['Nan Li', 'Tianli Liao'] | 2022-02-13 | null | null | null | null | ['image-stitching'] | ['computer-vision'] | [ 5.92367887e-01 -6.79602921e-02 2.86659479e-01 -6.57713711e-02
-3.46507937e-01 -5.95362902e-01 4.00463492e-01 -4.58411723e-01
1.77264120e-02 4.09643710e-01 1.03956409e-01 1.49791017e-01
-7.58818984e-02 -6.00312352e-01 -8.55992675e-01 -7.22692966e-01
4.05962467e-01 3.07885349e-01 2.16029942e-01 -1.37374014e-01
3.68268460e-01 5.20976722e-01 -1.40709245e+00 -1.70585811e-01
1.07216692e+00 7.43129313e-01 2.19841972e-01 4.72707093e-01
2.50428766e-01 3.90797853e-01 -3.21043402e-01 -3.85360181e-01
8.58827412e-01 -4.18942839e-01 -2.54008859e-01 7.49246478e-01
8.30618322e-01 -7.02877223e-01 -4.76484805e-01 1.20780456e+00
-8.21226910e-02 7.54947728e-03 2.67756104e-01 -1.17916179e+00
-7.14904740e-02 -2.02691495e-01 -1.40321028e+00 -4.46598619e-01
6.17655218e-01 2.12646961e-01 5.07697701e-01 -7.44116008e-01
9.76667345e-01 1.15575576e+00 5.00360548e-01 -6.83467537e-02
-1.23978961e+00 -5.63968956e-01 -1.39133111e-01 -1.42636448e-01
-1.52501094e+00 -4.62256759e-01 1.11730540e+00 -2.64353603e-01
2.65206516e-01 4.04586375e-01 7.95883060e-01 4.25016224e-01
4.36199635e-01 1.07859924e-01 1.07223105e+00 -7.87764549e-01
-1.44575223e-01 -6.86820075e-02 -4.64743137e-01 4.70279932e-01
2.09573135e-01 3.74705076e-01 -4.74887729e-01 -1.38271675e-01
1.36524403e+00 4.06145215e-01 -7.01661646e-01 -8.03506434e-01
-1.57805049e+00 2.59777755e-01 2.14755058e-01 1.49187669e-01
-4.00431931e-01 -2.20098123e-01 -1.50800675e-01 1.44350097e-01
2.04737619e-01 2.76858598e-01 1.10020779e-01 1.93452209e-01
-9.01496768e-01 2.41523728e-01 4.09518212e-01 1.15165353e+00
1.10798788e+00 1.14680489e-03 4.69523549e-01 6.05114222e-01
1.54303223e-01 5.83855927e-01 1.89130172e-01 -1.35898912e+00
4.90620673e-01 6.02839112e-01 3.28240544e-01 -1.29846120e+00
5.14398469e-03 5.42428195e-02 -1.00886023e+00 4.06166852e-01
3.63802373e-01 -2.86636818e-02 -6.85655475e-01 1.32712734e+00
7.58672595e-01 2.15522632e-01 -3.65987346e-02 9.89343643e-01
3.67848277e-01 6.52658045e-01 -8.76972198e-01 -4.81637299e-01
1.13972044e+00 -8.67148399e-01 -8.53880882e-01 -3.09612244e-01
4.96474691e-02 -1.22902000e+00 7.30479956e-01 3.52077872e-01
-1.16152382e+00 -4.19528097e-01 -1.10645187e+00 -1.50526986e-01
4.27275211e-01 -9.86994579e-02 1.09597109e-02 1.70025662e-01
-6.30154252e-01 3.40576351e-01 -3.75234127e-01 -7.16586038e-02
-2.21628889e-01 -6.27693115e-03 -6.50276423e-01 -3.45049471e-01
-7.04894602e-01 7.76949584e-01 1.86118558e-01 1.53372809e-01
-4.24527049e-01 -5.62036753e-01 -9.60632205e-01 -6.10515960e-02
4.50807422e-01 -6.77038670e-01 7.98179746e-01 -1.16239917e+00
-1.36736751e+00 9.09453452e-01 -2.84300119e-01 5.03586652e-03
6.45416737e-01 -1.12352215e-01 -1.07237592e-01 1.60553455e-01
1.18286505e-01 4.66956913e-01 8.99165213e-01 -1.59962952e+00
-4.89796638e-01 -4.88541126e-01 -2.18983486e-01 7.42702842e-01
2.04245210e-01 -2.24971548e-01 -6.96907103e-01 -5.52470267e-01
6.72954500e-01 -7.42161155e-01 -3.17697257e-01 2.30227664e-01
-4.97467995e-01 6.58056676e-01 1.01068199e+00 -7.59651303e-01
8.65668654e-01 -2.39286876e+00 2.44993851e-01 4.28782254e-01
1.72770232e-01 -2.17925414e-01 -6.64362162e-02 5.90824962e-01
-8.22918862e-02 -5.53676367e-01 -4.06298608e-01 -3.10375124e-01
-5.73349178e-01 1.47924244e-01 -2.94525415e-01 6.88621640e-01
-3.73456925e-01 5.12745082e-01 -7.07512677e-01 -4.94836211e-01
5.99793017e-01 3.96202564e-01 -2.35327080e-01 3.15713584e-01
-1.76423322e-02 7.81373858e-01 -3.29984501e-02 5.64649403e-01
1.29458845e+00 3.68616998e-01 1.51729539e-01 -3.57887566e-01
-5.14785409e-01 -2.86257505e-01 -1.51205909e+00 1.92917740e+00
-2.16911927e-01 7.06546187e-01 1.52428433e-01 -4.75658059e-01
1.08731592e+00 7.06731156e-02 6.85826361e-01 -6.97643816e-01
2.99939662e-01 2.94048876e-01 -2.80603826e-01 -4.61407810e-01
6.51929140e-01 -6.31384552e-02 2.61103243e-01 4.02854741e-01
-4.94392961e-01 -6.87962890e-01 7.89327454e-03 -9.50467307e-03
5.37078381e-01 2.22717792e-01 3.45885664e-01 2.58234292e-02
3.93911272e-01 4.05315943e-02 7.52240539e-01 1.04953229e-01
2.73795605e-01 1.32204473e+00 2.27711990e-01 -5.32696664e-01
-1.45327365e+00 -1.00250971e+00 -7.89054260e-02 -1.63999975e-01
9.15450096e-01 -1.67761922e-01 -8.79442215e-01 -2.15243641e-02
-2.17888281e-01 4.77513850e-01 -3.58995706e-01 -2.48322338e-02
-5.98993123e-01 -1.06212966e-01 -7.06316084e-02 -1.13079034e-01
7.37576544e-01 -5.63210964e-01 -8.41109633e-01 -5.18310294e-02
-5.75709283e-01 -1.20099223e+00 -9.90174890e-01 -5.15615642e-01
-1.01215565e+00 -1.28132820e+00 -8.89602482e-01 -9.72950757e-01
9.83036280e-01 1.18779445e+00 6.85225248e-01 -1.85229648e-02
-1.52648181e-01 6.07875548e-02 -9.53514427e-02 1.38781488e-01
-3.38275969e-01 -6.78393126e-01 -4.49282676e-02 5.27670979e-01
1.72225032e-02 -7.46239066e-01 -8.39006186e-01 8.32172394e-01
-1.09488678e+00 7.66532302e-01 4.58154768e-01 6.56683862e-01
6.43903434e-01 1.89305007e-01 -1.97034687e-01 -4.45157349e-01
1.30401012e-02 -3.54822315e-02 -8.53804946e-01 2.34837607e-01
-2.06166178e-01 -3.69059294e-01 2.82885790e-01 -4.02456343e-01
-1.37429452e+00 3.21269542e-01 4.14240301e-01 -8.63501728e-01
-4.51632738e-02 7.97205567e-02 -3.55687350e-01 -1.22634277e-01
5.44153214e-01 2.95599371e-01 2.91083843e-01 -2.08958328e-01
4.08600301e-01 5.98605871e-01 9.27477837e-01 -2.05069214e-01
1.13369560e+00 1.01752114e+00 4.54714857e-02 -9.05433297e-01
-1.56160295e-01 -4.14540857e-01 -9.55325484e-01 -3.56507510e-01
7.98367262e-01 -9.42203343e-01 -4.49561983e-01 6.92858458e-01
-1.22880495e+00 1.86461192e-02 -3.60653222e-01 6.06775641e-01
-5.11746585e-01 9.39312518e-01 -1.78008229e-01 -5.04891813e-01
-2.70804882e-01 -1.17975521e+00 1.06428897e+00 3.58499676e-01
-1.40933901e-01 -5.42844892e-01 2.47794166e-01 3.46453071e-01
-5.96786626e-02 5.82649112e-01 6.53159201e-01 5.26301563e-01
-1.08958006e+00 -2.29915977e-01 -1.58958539e-01 1.61122262e-01
2.82474458e-01 3.15217167e-01 -7.49222398e-01 -1.31316736e-01
3.34192693e-01 2.64346242e-01 2.53942877e-01 6.01281703e-01
3.22619140e-01 -1.98462039e-01 -2.57350534e-01 9.20132399e-01
1.56462920e+00 4.36144710e-01 9.47351694e-01 4.39765364e-01
8.24655890e-01 9.30113912e-01 9.02836084e-01 3.07283372e-01
2.12088436e-01 9.61733758e-01 5.23948669e-01 -3.39014053e-01
-4.01461385e-02 -3.92003655e-01 2.39130795e-01 6.89876556e-01
-5.95311895e-02 3.35530527e-02 -5.51558316e-01 5.52462816e-01
-1.78100467e+00 -6.76830351e-01 -4.39385861e-01 2.82791638e+00
6.27749801e-01 -2.64853269e-01 -3.71747196e-01 1.86151136e-02
8.81067514e-01 9.08007845e-02 -4.43162620e-01 -1.92732289e-01
-3.51198584e-01 -2.57828057e-01 6.18484080e-01 8.76214802e-01
-4.78987098e-01 6.34161651e-01 5.10234451e+00 5.94341338e-01
-9.97280836e-01 -1.88808426e-01 2.95920789e-01 1.08760782e-02
-5.25305092e-01 6.32390559e-01 -1.25778466e-01 3.28000993e-01
6.27838075e-02 -9.12311897e-02 4.46372122e-01 3.90757024e-01
2.41046995e-01 -6.99161053e-01 -6.99773967e-01 1.23427153e+00
3.03734124e-01 -1.11782479e+00 -1.62617251e-01 1.56135291e-01
1.17314875e+00 -3.25641662e-01 -1.38433129e-01 -7.40843892e-01
-3.08598988e-02 -4.94328290e-01 7.93318570e-01 5.32107234e-01
9.78195906e-01 -7.82876551e-01 3.40108067e-01 5.26980460e-01
-1.04315805e+00 2.23358959e-01 -3.81580740e-01 3.11048348e-02
6.51840031e-01 6.20148540e-01 -4.41798478e-01 7.79803693e-01
5.92926145e-01 6.17927849e-01 -1.92311659e-01 1.11007643e+00
-1.17363565e-01 -4.08997118e-01 -4.25081998e-01 7.85580397e-01
-1.49729893e-01 -1.14781618e+00 8.38299334e-01 3.61099064e-01
6.37542784e-01 2.80799836e-01 -2.32162848e-01 8.44691515e-01
1.41668767e-01 5.81598515e-03 -9.63329315e-01 5.49328983e-01
4.34224933e-01 1.19831264e+00 -5.76494634e-01 -1.82865232e-01
-4.70182925e-01 1.29220355e+00 -2.06885204e-01 5.64938962e-01
-7.31061697e-01 -3.55803341e-01 4.95544761e-01 3.02549243e-01
-7.43930563e-02 -2.81129181e-01 -4.05236185e-01 -1.39792597e+00
3.51019502e-01 -9.30609941e-01 4.66180360e-03 -1.21608591e+00
-5.75463772e-01 4.97921467e-01 1.27037138e-01 -1.78951585e+00
-1.74463227e-01 6.64837435e-02 -7.55577385e-01 8.13890755e-01
-1.24091542e+00 -1.11694050e+00 -6.74131691e-01 5.54718733e-01
5.88328183e-01 3.50953996e-01 4.00754839e-01 1.56728670e-01
-2.61219323e-01 1.87404498e-01 5.09180307e-01 -2.45112032e-01
9.56840456e-01 -6.31684422e-01 2.91328251e-01 1.11432493e+00
4.37774956e-02 4.99380231e-01 7.03363001e-01 -6.10147774e-01
-1.33111501e+00 -6.44942522e-01 7.01087058e-01 -1.29164413e-01
-7.40064606e-02 -2.10527971e-01 -8.91577721e-01 6.88207865e-01
3.97806436e-01 -3.05435479e-01 -6.09847084e-02 -6.65195763e-01
-2.37356752e-01 -2.73489118e-01 -1.26028609e+00 8.29908609e-01
7.80196011e-01 -2.48731673e-01 -2.96672136e-01 1.41765147e-01
5.89206517e-01 -9.02284920e-01 -6.20770156e-01 3.72176319e-01
9.00213301e-01 -1.45207906e+00 1.03388739e+00 3.72508258e-01
5.00313282e-01 -7.54008532e-01 -4.74673435e-02 -1.16322899e+00
7.85054043e-02 -9.77870762e-01 4.73757356e-01 1.24831092e+00
-1.73238352e-01 -6.29243135e-01 6.71550155e-01 5.84990382e-01
-4.59153317e-02 -3.68750811e-01 -8.78084302e-01 -4.70275581e-01
-5.29993176e-01 4.29750271e-02 5.32281995e-01 1.05588531e+00
-1.64992735e-01 1.60831809e-01 -7.27341354e-01 3.02273512e-01
9.62598383e-01 5.01799643e-01 1.33853650e+00 -9.79224265e-01
-1.73740745e-01 -4.95678410e-02 -2.94659525e-01 -1.17722416e+00
-4.04422551e-01 -1.98471710e-01 5.18482141e-02 -1.07814682e+00
8.13820139e-02 -2.75916815e-01 6.88122034e-01 -1.03437990e-01
4.92546521e-02 3.83916378e-01 1.55198351e-01 6.81042492e-01
2.03098714e-01 5.59171617e-01 1.51222849e+00 2.59211153e-01
-5.46236396e-01 -1.24078616e-01 -2.58436173e-01 9.04421508e-01
4.45407659e-01 -2.18862817e-01 -4.49701726e-01 -7.42409945e-01
9.56459492e-02 5.32268524e-01 2.42789388e-01 -9.36482191e-01
1.82914197e-01 -4.52118993e-01 3.41503859e-01 -8.33744526e-01
6.02270603e-01 -1.17942512e+00 8.60284567e-01 1.30145609e-01
1.25168428e-01 2.13326856e-01 -7.25613385e-02 5.04533052e-01
-2.48503134e-01 -2.45439127e-01 8.71015429e-01 -6.28657565e-02
-3.24108034e-01 2.41281748e-01 9.01253968e-02 -3.39470863e-01
1.29761195e+00 -8.34653020e-01 -1.21123850e-01 -6.00391388e-01
-1.81939587e-01 6.66583255e-02 1.31961310e+00 1.64043069e-01
9.79633331e-01 -1.42301667e+00 -5.70539832e-01 8.27037752e-01
9.63294804e-02 5.73571384e-01 5.97821236e-01 9.51851368e-01
-1.10105836e+00 -5.11195064e-02 -3.46647173e-01 -8.51090133e-01
-1.53671873e+00 4.77969646e-01 2.40791589e-01 1.45168677e-01
-7.67742515e-01 3.61665577e-01 6.62442923e-01 -3.05834562e-01
8.76982585e-02 -2.01756880e-01 2.85732448e-01 -3.17711979e-01
4.64381427e-01 3.37327570e-01 -1.94215819e-01 -9.53746378e-01
5.16512729e-02 1.32696545e+00 -3.00900778e-03 -3.97529811e-01
1.04101396e+00 -6.39010668e-01 -3.10132951e-01 1.11706823e-01
1.19176579e+00 4.93284822e-01 -1.61351407e+00 -4.73461390e-01
-6.82207167e-01 -1.37267888e+00 -2.37672403e-01 -6.34624287e-02
-1.11330831e+00 7.90859044e-01 4.67601955e-01 -2.05150902e-01
1.33260989e+00 -3.57814372e-01 9.60288763e-01 -2.24235609e-01
2.05488607e-01 -7.28895545e-01 -7.97221065e-02 1.20319381e-01
9.01349664e-01 -8.40900302e-01 3.14927518e-01 -6.53638422e-01
-6.04760647e-01 1.27504325e+00 6.17341518e-01 -2.21201360e-01
2.04565540e-01 -1.03986800e-01 2.26934984e-01 -1.95239633e-01
-2.26886854e-01 1.22483075e-01 9.84450877e-02 6.80889904e-01
-1.41907230e-01 -3.59327555e-01 -1.31486863e-01 -4.78454262e-01
-1.08294569e-01 -3.21628928e-01 8.68246973e-01 8.39993536e-01
-3.83401901e-01 -1.00253642e+00 -1.05014956e+00 -1.79677367e-01
1.80808216e-01 9.71430838e-02 -2.92142987e-01 7.71957934e-01
1.14950091e-01 7.53893256e-01 1.89160541e-01 -3.21439773e-01
4.87289160e-01 -4.92298603e-01 6.60910070e-01 -2.69219220e-01
-1.18363425e-02 5.91505408e-01 -1.95851699e-01 -4.93596226e-01
-3.67228627e-01 -5.35454154e-01 -1.08198047e+00 -4.30068821e-01
-4.39595014e-01 -1.31169826e-01 5.31523407e-01 5.93961656e-01
2.74729073e-01 -1.19800754e-01 1.02466750e+00 -1.18512416e+00
-9.92185771e-02 -6.57872319e-01 -7.57300138e-01 5.92260420e-01
5.24550617e-01 -5.19478440e-01 -4.48953897e-01 2.74418294e-01] | [9.27246379852295, -2.403306722640991] |
9f16be85-7dac-4a2c-80aa-e635beea5f82 | aerosense-a-self-sustainable-and-long-range | 2205.11902 | null | https://arxiv.org/abs/2205.11902v1 | https://arxiv.org/pdf/2205.11902v1.pdf | Aerosense: A Self-Sustainable And Long-Range Bluetooth Wireless Sensor Node for Aerodynamic and Aeroacoustic Monitoring on Wind Turbines | This paper presents a low-power, self-sustainable, and modular wireless sensor node for aerodynamic and acoustic measurements on wind turbines and other industrial structures. It includes 40 high-accuracy barometers, 10 microphones, 5 differential pressure sensors, and implements a lossy and a lossless on-board data compression algorithm to decrease the transmission energy cost. The wireless transmitter is based on Bluetooth Low Energy 5.1 tuned for long-range and high throughput while maintaining adequate per-bit energy efficiency (80 nJ). Moreover, we field-assessed the node capability to collect precise and accurate aerodynamic data. Outdoor experimental tests revealed that the system can acquire and sustain a data rate of 850 kbps over 438 m. The power consumption while collecting and streaming all measured data is 120 mW, enabling self-sustainability and long-term in-situ monitoring with a 111 cm^2 photovoltaic panel. | ['Michele Magno', 'Luca Benini', 'Raphael Fischer', 'Weikang Kong', 'Hanna Müller', 'Tommaso Polonelli'] | 2022-05-24 | null | null | null | null | ['data-compression'] | ['time-series'] | [ 1.30000085e-01 5.98415025e-02 1.47136509e-01 2.93315858e-01
-1.18409708e-01 -5.23389220e-01 -3.51951867e-01 3.57336402e-01
-2.84668207e-01 8.34387004e-01 -3.08402628e-01 -3.80146593e-01
-2.65170395e-01 -1.30313230e+00 -1.43347338e-01 -1.15168226e+00
-6.58336520e-01 -1.85006648e-01 2.53173172e-01 2.62454808e-01
-2.62198776e-01 3.14553142e-01 -1.91286588e+00 -6.37760758e-01
8.82398307e-01 1.42731535e+00 5.68798900e-01 7.79223859e-01
6.95727348e-01 -1.99644640e-02 -1.06304121e+00 -3.63072716e-02
-6.93736076e-02 -1.57829409e-03 -2.01599061e-01 -6.80097342e-01
-8.26459005e-03 -9.64130282e-01 2.38893643e-01 4.27394986e-01
1.03112662e+00 -4.38098460e-01 4.60674673e-01 -1.16399121e+00
6.94048256e-02 5.24567306e-01 -3.22690189e-01 4.93014641e-02
2.64761448e-01 9.79069099e-02 6.84789717e-01 -8.38280693e-02
1.99943751e-01 5.69901168e-01 9.90901351e-01 3.11508209e-01
-1.00191915e+00 -1.08548748e+00 -1.25932062e+00 -1.66381344e-01
-1.11479652e+00 -7.40502477e-01 6.93970919e-01 -3.16660367e-02
1.25176764e+00 5.76723218e-01 1.45423353e+00 8.95679295e-01
7.50176251e-01 -5.65614998e-01 9.99183774e-01 -2.74323404e-01
8.28972161e-01 -3.24489991e-03 -6.68610215e-01 5.13066530e-01
1.11104167e+00 3.11996996e-01 -6.30678177e-01 -3.64269614e-01
5.95637023e-01 -3.05181384e-01 -1.42354891e-01 2.64655530e-01
-6.14042759e-01 3.01451743e-01 4.24393684e-01 3.97380590e-01
-2.15788737e-01 6.60405815e-01 2.03032285e-01 2.45849676e-02
2.68578917e-01 3.67107004e-01 -6.78898335e-01 -6.69919670e-01
-7.88200080e-01 -6.63205266e-01 1.09899473e+00 9.62906539e-01
1.21117733e-01 4.22171801e-01 4.16340262e-01 5.40355384e-01
5.60398102e-01 1.70458090e+00 5.56373000e-01 -1.26410699e+00
1.56392172e-01 -7.69607201e-02 1.58763826e-01 -9.65328336e-01
-7.41720557e-01 -2.37152964e-01 -1.06887412e+00 1.67250529e-01
-3.89184773e-01 -6.61389351e-01 -2.07725793e-01 1.39062691e+00
5.96782923e-01 -2.12785587e-01 2.48760372e-01 4.84945655e-01
1.01795042e+00 7.11972594e-01 1.27627715e-01 -8.39911878e-01
1.36861265e+00 -3.44882429e-01 -8.25414658e-01 -3.13530900e-02
3.30862314e-01 -6.20255768e-01 8.36144209e-01 1.60647273e-01
-1.27832925e+00 -4.26159531e-01 -1.47175407e+00 2.03623138e-02
-8.60131532e-02 3.17285955e-01 1.13627091e-01 1.33430636e+00
-1.05028558e+00 4.39859807e-01 -1.11838245e+00 -3.89930755e-01
1.11839518e-01 4.15146276e-02 2.53176168e-02 8.45867753e-01
-8.44861627e-01 3.57484013e-01 5.90450950e-02 -2.70136632e-02
-5.20268083e-01 -8.33808422e-01 -4.04929399e-01 3.71615499e-01
-1.84927940e-01 -1.33046579e+00 1.19550955e+00 3.10032874e-01
-2.33511496e+00 -1.91099886e-02 2.89844256e-02 -5.31183839e-01
-2.72860020e-01 -1.50059059e-01 -5.33270538e-01 9.49405551e-01
-2.70565420e-01 2.43701488e-01 6.98628783e-01 -9.93164718e-01
-6.12810016e-01 -3.51271212e-01 -4.43785608e-01 -6.91200048e-02
-1.11494422e+00 -6.03695452e-01 1.05756676e+00 -3.73608917e-01
2.48317588e-02 -5.74456096e-01 1.36276200e-01 2.28269562e-01
-2.73544699e-01 5.28139025e-02 1.36895013e+00 -5.18906713e-01
9.85698223e-01 -1.87952507e+00 -4.88332510e-01 3.21185231e-01
-1.55416988e-02 2.33068988e-01 3.30478907e-01 6.84206367e-01
6.49723649e-01 3.70388657e-01 -2.16737434e-01 -2.39422038e-01
-2.85300255e-01 4.13455725e-01 -1.00256950e-01 3.05330634e-01
-4.64667469e-01 4.89573687e-01 -8.00139904e-01 -3.48159969e-01
6.09220445e-01 5.47782958e-01 -1.56943113e-01 1.45069361e-01
4.38047558e-01 2.63738573e-01 -5.07923126e-01 9.38476980e-01
6.66562736e-01 4.80782270e-01 8.06297138e-02 -4.08815295e-01
-2.68692642e-01 3.01860809e-01 -1.06959927e+00 1.73031902e+00
-1.02210021e+00 2.15200752e-01 1.19627059e+00 -6.90810204e-01
1.15780544e+00 3.80903959e-01 1.08969104e+00 -7.65343070e-01
1.57864168e-01 4.23935115e-01 -8.51713061e-01 -1.06234491e+00
7.17642605e-01 -3.30630749e-01 -6.16026334e-02 1.11605182e-01
-1.02838837e-02 -7.23700404e-01 4.82349806e-02 -5.11692092e-03
1.48641145e+00 -1.87177390e-01 -1.28931238e-03 -3.88788790e-01
2.60596037e-01 -8.34834725e-02 6.29690766e-01 1.67563930e-01
-8.35451856e-02 -4.13445622e-01 -6.54009461e-01 3.75508189e-01
-6.61059260e-01 -1.20839751e+00 -2.43116260e-01 3.53401721e-01
6.22653365e-01 -6.84151053e-01 -5.82009256e-01 2.06934646e-01
2.98270851e-01 6.94254398e-01 5.20109832e-01 9.44410488e-02
-2.19721258e-01 -2.96017468e-01 5.33347011e-01 5.73083878e-01
6.63178921e-01 -7.80229747e-01 -1.54650235e+00 3.79631370e-01
-2.06320345e-01 -9.52347159e-01 2.69079238e-01 3.33304435e-01
-1.18176711e+00 -9.80136275e-01 4.55082417e-01 -6.20125473e-01
3.66047055e-01 2.08887547e-01 9.79086876e-01 6.57136738e-02
-5.08220494e-01 8.10429215e-01 -2.65203774e-01 -8.77312064e-01
-1.44243300e-01 1.01422602e-02 6.39001966e-01 -1.12557364e+00
-1.04788385e-01 -1.13751185e+00 -9.66782212e-01 3.42499405e-01
-3.01674932e-01 -4.86410737e-01 4.48291153e-01 3.75266641e-01
5.46673179e-01 9.42174733e-01 7.35904932e-01 -1.18973348e-02
6.40206575e-01 -4.75128651e-01 -7.05165148e-01 -3.30608487e-01
-9.84436631e-01 -6.88530505e-01 8.90685380e-01 1.10544920e-01
-9.82698917e-01 5.13870195e-02 -3.61599654e-01 4.56449568e-01
-1.77447215e-01 1.94020689e-01 -2.06184626e-01 -1.93688661e-01
4.98396531e-02 -4.73016381e-01 4.74506259e-01 -3.93491328e-01
3.48290890e-01 1.12683916e+00 1.01114333e+00 -2.63555288e-01
1.39671969e+00 5.26247978e-01 4.26385522e-01 -1.80274379e+00
2.26259649e-01 -1.24849364e-01 -6.03874698e-02 -3.57500643e-01
6.27477050e-01 -1.34039593e+00 -1.36402941e+00 4.46690291e-01
-4.27698880e-01 -4.80246097e-01 -9.40047979e-01 5.91459811e-01
-2.11476892e-01 4.48042184e-01 -4.87052113e-01 -1.15832472e+00
-1.54149938e+00 -5.39405882e-01 9.60045755e-01 4.90103185e-01
-3.74393076e-01 -7.63848603e-01 -9.15297717e-02 4.12424237e-01
9.96968985e-01 3.69810671e-01 4.79095668e-01 5.70807517e-01
-1.73808500e-01 -1.27146468e-01 4.39549983e-01 2.51220733e-01
3.26714784e-01 3.15088660e-01 -1.11654186e+00 -7.71841466e-01
4.05568928e-01 -5.00330627e-01 6.16232574e-01 5.63595831e-01
1.17580509e+00 -5.13388157e-01 -6.94267333e-01 6.89668834e-01
1.63425708e+00 -1.48155808e-01 1.96745455e-01 -2.80912250e-01
3.21995765e-02 3.90314646e-02 3.58341396e-01 6.38518035e-01
1.21606007e-01 2.79243737e-02 1.16187286e+00 1.14282943e-01
1.39875770e-01 -2.04299271e-01 5.93905687e-01 1.74773967e+00
2.46875454e-03 -6.30673051e-01 2.55631387e-01 2.55826235e-01
-6.24848425e-01 -8.84126961e-01 -3.92580152e-01 2.29825068e+00
9.27784383e-01 -1.40091320e-02 -3.21952216e-02 6.42243385e-01
3.03527594e-01 9.98162329e-02 -4.40421373e-01 -7.72617400e-01
2.44608775e-01 5.53416252e-01 9.25695956e-01 3.13190103e-01
-3.79637241e-01 -2.70476460e-01 6.18421984e+00 7.88871646e-01
-9.08164442e-01 2.53624797e-01 -1.06616259e-01 -5.04364967e-01
-7.05713987e-01 -3.73874485e-01 -5.47152698e-01 8.06351364e-01
1.50360715e+00 -3.99860710e-01 9.22003984e-02 6.82868838e-01
2.94909775e-01 -8.51971209e-01 -5.13898492e-01 3.44545871e-01
-4.93099362e-01 -1.01714027e+00 -7.35207260e-01 3.55957776e-01
1.19024456e-01 -5.49955331e-02 -4.65284228e-01 -3.17882597e-01
2.43909508e-01 -3.61255407e-01 2.66536444e-01 1.93218291e-01
1.33555627e+00 -6.22878432e-01 5.26519001e-01 6.32021546e-01
-1.71834767e+00 -5.46203017e-01 -6.93139791e-01 -4.23353881e-01
5.13659120e-01 1.57515287e+00 -6.99386954e-01 9.41335440e-01
1.32718086e+00 3.93118203e-01 9.03239176e-02 6.95053816e-01
-1.48691446e-01 1.23509765e+00 -1.54920709e+00 -6.11225843e-01
-5.81039906e-01 -3.53341758e-01 8.22269499e-01 4.83472079e-01
1.09651649e+00 1.16230875e-01 -1.70325130e-01 3.57563496e-01
-2.48811081e-01 -2.72371799e-01 -7.92437196e-01 4.93188292e-01
1.45096552e+00 1.83642137e+00 -3.47893476e-01 2.32420146e-01
1.60421848e-01 1.73901960e-01 -7.06306815e-01 -1.25484228e-01
-6.31923556e-01 -7.43991911e-01 6.62418008e-01 2.89951801e-01
4.38190639e-01 -6.08165503e-01 -4.24675226e-01 -1.42200872e-01
4.66986477e-01 5.14292121e-01 1.48681412e-02 -9.01543736e-01
-1.16413260e+00 -1.90328628e-01 -1.98612109e-01 -1.20393741e+00
9.87716615e-02 -3.06960970e-01 -9.08066750e-01 2.27239028e-01
-1.49683070e+00 -7.24232435e-01 -7.25923300e-01 1.65739641e-01
-1.23799425e-02 6.55954555e-02 1.43799198e+00 2.82771170e-01
-5.18120289e-01 5.52668646e-02 7.86154926e-01 -6.27557337e-01
3.29109877e-01 -1.02146292e+00 -2.34555915e-01 6.10335827e-01
-5.87514281e-01 1.72104999e-01 6.06014967e-01 -7.03732073e-01
-2.36096025e+00 -9.82578397e-01 5.49721479e-01 1.97780207e-01
6.37647033e-01 -2.26609513e-01 -1.85180396e-01 -4.15588200e-01
7.99072444e-01 -2.16136710e-03 7.41301119e-01 -4.23327714e-01
4.40253794e-01 -9.42520082e-01 -1.96256232e+00 1.38964597e-02
1.09117472e+00 -8.18709284e-02 -1.16592780e-01 3.71697158e-01
6.41563952e-01 -3.62710029e-01 -1.68018949e+00 4.62780178e-01
8.10082138e-01 -5.82856655e-01 9.15024400e-01 6.69400632e-01
5.34736924e-02 -2.23134711e-01 2.59336263e-01 -1.30450714e+00
-1.05937585e-01 -9.20016646e-01 -4.08352792e-01 2.03493452e+00
-1.03998622e-02 -8.40329230e-01 5.21400809e-01 9.56705026e-03
-2.27090404e-05 -6.78331673e-01 -1.77883232e+00 -7.16643274e-01
-2.95390964e-01 -1.01060539e-01 5.13673604e-01 2.14231879e-01
6.60175025e-01 3.60996127e-01 2.12625369e-01 5.02326310e-01
7.53337741e-01 -1.84897985e-02 5.94872832e-01 -1.56497312e+00
3.17537606e-01 2.16321386e-02 -3.03195328e-01 -8.45338941e-01
-1.53582349e-01 -5.21083415e-01 1.44131109e-01 -1.56594086e+00
-4.52097982e-01 -4.96754557e-01 2.14449629e-01 2.19171211e-01
3.98095667e-01 3.46146852e-01 -1.68301687e-01 -2.99024075e-01
3.58510837e-02 7.49958694e-01 9.19051051e-01 -1.60394609e-01
1.03756458e-01 1.59496501e-01 -2.12516114e-01 4.89118725e-01
9.81570244e-01 -5.06086409e-01 -5.96649230e-01 -3.79584044e-01
1.09543309e-01 -3.10164206e-02 5.23496330e-01 -1.63983238e+00
4.53430176e-01 1.60200000e-01 2.25804493e-01 -6.29212201e-01
4.18014735e-01 -1.77065420e+00 3.69252145e-01 1.23367608e+00
5.24058163e-01 -3.04857269e-02 9.19896364e-02 6.07476234e-01
2.46176615e-01 1.00036874e-01 4.43305373e-01 2.93389916e-01
-7.29519352e-02 -3.41255158e-01 -7.25154161e-01 -2.06226945e-01
1.35744059e+00 -3.58806878e-01 -9.21327114e-01 -3.05776983e-01
-1.44300088e-01 4.61626679e-01 4.56580490e-01 -1.83395296e-01
5.29197216e-01 -9.67376769e-01 -1.66591227e-01 4.70775485e-01
-5.13506055e-01 3.40576589e-01 1.76738933e-01 6.08854413e-01
-7.11074769e-01 7.48918056e-02 -1.53499112e-01 -7.59470761e-01
-1.53511930e+00 -3.66851211e-01 1.98806927e-01 2.32952714e-01
-7.08347023e-01 7.36889899e-01 -1.35416842e+00 1.51054978e-01
6.65897802e-02 -5.82493246e-01 6.67255223e-02 -1.05631421e-03
-6.97870106e-02 1.25172269e+00 2.61437356e-01 5.78048453e-02
-3.14419627e-01 9.33534563e-01 1.45719004e+00 2.20888421e-01
1.39874387e+00 -6.47318363e-01 -2.11461946e-01 2.58701503e-01
8.90331805e-01 5.77347875e-01 -7.21820474e-01 3.49138081e-01
-8.00252318e-01 -4.66307849e-01 3.64336640e-01 -7.08802700e-01
-1.34056199e+00 3.53059560e-01 6.37754321e-01 1.06267488e+00
1.81173348e+00 -4.87790406e-01 1.38098562e+00 5.31481981e-01
7.88712144e-01 -1.62484264e+00 -1.45827025e-01 -1.08787507e-01
3.55431587e-01 -4.60158288e-01 7.27611423e-01 -5.00145197e-01
3.74306858e-01 1.42262733e+00 5.62002063e-01 3.69572341e-01
1.05347383e+00 1.15650308e+00 -1.65738985e-01 -6.25649542e-02
-5.97709000e-01 2.42252603e-01 -9.44000721e-01 1.05159092e+00
1.77578211e-01 2.39252880e-01 -5.30370057e-01 4.15248811e-01
-7.32290030e-01 1.07013889e-01 2.43362218e-01 1.17126226e+00
-1.15712774e+00 -8.11111629e-01 -2.46154025e-01 7.24323869e-01
-4.50249344e-01 4.32179987e-01 2.30786026e-01 4.04777229e-01
1.61852062e-01 1.71245241e+00 3.70564252e-01 -6.67269826e-01
5.11802912e-01 -3.44537169e-01 5.59538007e-02 7.83307478e-02
-6.15023375e-01 -1.24598024e-02 5.51742971e-01 -9.19364452e-01
-7.37771034e-01 -4.44436640e-01 -1.41734052e+00 -7.23056376e-01
-8.59653354e-01 3.09368879e-01 1.47758353e+00 3.18194151e-01
7.47360289e-01 7.35489309e-01 1.43066049e+00 -4.90043014e-01
-6.45975530e-01 -1.06143320e+00 -1.20246768e+00 -3.15809906e-01
3.24033409e-01 -4.70530629e-01 -8.41918468e-01 -1.13855332e-01] | [6.289802074432373, 1.3925830125808716] |
d459612f-6320-4d49-8a94-00bc6e98dfda | on-the-efficiency-of-integrating-self | 2204.00352 | null | https://arxiv.org/abs/2204.00352v3 | https://arxiv.org/pdf/2204.00352v3.pdf | On the Efficiency of Integrating Self-supervised Learning and Meta-learning for User-defined Few-shot Keyword Spotting | User-defined keyword spotting is a task to detect new spoken terms defined by users. This can be viewed as a few-shot learning problem since it is unreasonable for users to define their desired keywords by providing many examples. To solve this problem, previous works try to incorporate self-supervised learning models or apply meta-learning algorithms. But it is unclear whether self-supervised learning and meta-learning are complementary and which combination of the two types of approaches is most effective for few-shot keyword discovery. In this work, we systematically study these questions by utilizing various self-supervised learning models and combining them with a wide variety of meta-learning algorithms. Our result shows that HuBERT combined with Matching network achieves the best result and is robust to the changes of few-shot examples. | ['Yuan-Kuei Wu', 'Chia-Ping Chen', 'Hung-Yi Lee', 'Yu-Pao Tsai', 'Zhi-Sheng Chen', 'Wei-Tsung Kao'] | 2022-04-01 | null | null | null | null | ['keyword-spotting'] | ['speech'] | [ 9.11102518e-02 -1.03828616e-01 -7.39674509e-01 -3.33495647e-01
-8.07600677e-01 -4.62776244e-01 8.22146773e-01 3.15735281e-01
-5.16406715e-01 7.70079911e-01 2.79917449e-01 -1.16624148e-03
-4.82010067e-01 -8.18511963e-01 -2.31082663e-01 -3.29149574e-01
-1.25095651e-01 6.39197230e-01 5.66554129e-01 -6.28204286e-01
5.43112457e-01 8.92303512e-02 -1.91124272e+00 3.12148243e-01
8.30207407e-01 4.50936407e-01 2.70623535e-01 6.23553634e-01
-9.24927771e-01 6.48419797e-01 -6.64982796e-01 -1.00970402e-01
-1.45859927e-01 -5.65119505e-01 -9.34191942e-01 -7.21541718e-02
2.50951853e-02 -9.25101340e-03 -3.77177328e-01 9.29445326e-01
4.85920399e-01 5.57421744e-01 5.84651053e-01 -1.34194922e+00
-4.54544455e-01 1.16205370e+00 -1.43256024e-01 5.49117029e-01
8.27757955e-01 -2.27043912e-01 1.16512835e+00 -9.89188790e-01
5.51016748e-01 9.57909226e-01 5.73945105e-01 6.07203960e-01
-9.26235497e-01 -7.62284577e-01 -2.58166753e-02 5.28720081e-01
-1.55927837e+00 -6.06733143e-01 8.28408003e-01 -1.96635783e-01
1.00741923e+00 3.58796686e-01 4.43723440e-01 1.00631702e+00
-2.43061900e-01 7.80464828e-01 8.42372894e-01 -1.02683485e+00
6.13039315e-01 5.46503127e-01 5.76091945e-01 3.89410585e-01
-1.18223190e-01 -7.85575062e-02 -8.41229081e-01 -5.34908175e-01
1.21383943e-01 2.21503481e-01 -2.20434442e-01 -8.65895953e-03
-8.19735050e-01 9.94431198e-01 -1.95857838e-01 1.04319084e+00
-8.76107216e-02 -2.77298391e-01 3.35819244e-01 5.20714641e-01
6.97980821e-01 9.74280477e-01 -4.70754743e-01 -4.02153820e-01
-1.23521566e+00 2.17279702e-01 1.20637810e+00 9.39414442e-01
1.09421945e+00 -9.81917977e-03 -2.31395647e-01 1.21153319e+00
-4.77274917e-02 1.41395424e-02 9.35161114e-01 -8.00165951e-01
-2.94551343e-01 5.49421012e-01 1.48389041e-01 -7.53701568e-01
-1.43536016e-01 -8.41932148e-02 -2.36052752e-01 -3.25325668e-01
1.49395034e-01 -2.73665905e-01 -1.10899973e+00 1.42639256e+00
1.11684695e-01 4.91899043e-01 -3.83875407e-02 5.19529998e-01
1.16351509e+00 9.23214197e-01 -2.84406245e-02 -7.69747019e-01
1.00272930e+00 -1.11546767e+00 -1.16759682e+00 -2.47743517e-01
6.36866748e-01 -6.59929872e-01 1.17455697e+00 1.67321905e-01
-7.58596361e-01 -3.53321671e-01 -9.36738968e-01 4.71514076e-01
-1.04717767e+00 -6.76519275e-01 6.08607471e-01 7.41013706e-01
-7.48347938e-01 6.99520707e-01 -2.02438787e-01 -8.95390272e-01
-1.45291969e-01 2.36771524e-01 -2.96527501e-02 6.71970546e-02
-1.78863204e+00 9.77099121e-01 3.99306506e-01 -7.29233325e-01
-6.99226141e-01 -7.52311110e-01 -6.62958860e-01 6.76866844e-02
8.62530649e-01 -1.80290893e-01 1.61843967e+00 -8.59499753e-01
-1.33124459e+00 6.34484589e-01 -4.01086837e-01 -3.78289878e-01
-3.56094129e-02 4.17083688e-02 -7.79309273e-01 1.66850656e-01
7.40285665e-02 2.89919019e-01 9.37104285e-01 -1.24167776e+00
-8.40023816e-01 -2.60859162e-01 -5.07923737e-02 2.23493308e-01
-7.80557394e-01 2.15104803e-01 -4.78147954e-01 -4.24572766e-01
-3.06566134e-02 -5.10143578e-01 -1.18778907e-01 -5.73387980e-01
-1.23054102e-01 -7.08384812e-01 9.41941917e-01 1.17856469e-02
1.61663473e+00 -1.85791731e+00 -2.59208739e-01 7.63964579e-02
1.59390226e-01 4.37910467e-01 -2.19020789e-04 8.85942936e-01
7.96937644e-02 3.03774089e-01 1.54255986e-01 -5.26302047e-02
-1.00572407e-01 3.87297422e-01 -3.34781528e-01 2.53193751e-02
-3.48815501e-01 8.44923794e-01 -1.32283401e+00 -6.03623688e-01
2.53574550e-01 9.68422890e-02 6.24419861e-02 2.58845419e-01
-4.18637574e-01 -1.12629585e-01 -4.23972189e-01 6.86114669e-01
-3.02072782e-02 -6.40409812e-02 -1.93831197e-03 1.11271448e-01
-9.48950276e-02 1.43086582e-01 -1.25019264e+00 1.61904776e+00
-2.33781472e-01 5.18886626e-01 -4.86664206e-01 -1.15134442e+00
8.42757702e-01 5.73392212e-01 5.42269528e-01 -5.40104270e-01
4.68455330e-02 1.52956650e-01 -3.39905053e-01 -8.57847869e-01
3.99103999e-01 -4.54091936e-01 -1.23251006e-01 6.62803173e-01
5.69076478e-01 8.03236514e-02 3.13343912e-01 4.04567689e-01
1.20892358e+00 -4.75606829e-01 5.51283956e-01 1.52554931e-02
1.60972476e-01 3.68137866e-01 4.10934269e-01 1.38414693e+00
-1.10983580e-01 3.49891603e-01 5.71354851e-03 -1.85929507e-01
-7.37556517e-01 -7.75086105e-01 1.73751146e-01 1.63310516e+00
6.06003404e-02 -7.76281953e-01 -5.48498988e-01 -4.21901643e-01
-8.05766508e-02 1.07080746e+00 -5.39018631e-01 -4.33169246e-01
1.70383573e-01 -4.10581112e-01 5.71808815e-01 6.76684901e-02
2.62445867e-01 -1.03844678e+00 -5.68852484e-01 2.88733870e-01
-6.70567378e-02 -7.88534224e-01 -4.05669361e-01 4.65986460e-01
-6.01338506e-01 -8.17051768e-01 -9.72788930e-01 -8.32738936e-01
2.95798451e-01 6.27044320e-01 9.18584883e-01 7.27853477e-02
-5.01633048e-01 8.21319938e-01 -1.05507064e+00 -4.85330224e-01
-3.73310268e-01 3.17715347e-01 3.38102728e-01 6.29067495e-02
9.03889418e-01 -6.08047962e-01 -2.34883204e-01 1.87150408e-02
-7.83553421e-01 -4.13663864e-01 1.22533552e-01 7.53410459e-01
3.78144011e-02 2.51922309e-01 7.09425390e-01 -1.12293196e+00
1.14017367e+00 -6.34722114e-01 -2.58871187e-02 7.67973065e-01
-1.23705900e+00 2.07504462e-02 2.92528749e-01 -7.81040311e-01
-7.99795985e-01 7.01566041e-02 2.00385123e-01 -4.90373045e-01
-3.29761535e-01 6.73855007e-01 1.10346176e-01 -2.05663502e-01
8.58927071e-01 5.37469804e-01 -2.86895692e-01 -4.96031433e-01
5.83409071e-01 1.09279132e+00 2.30341271e-01 -1.51269257e-01
7.12809861e-01 1.75826043e-01 -8.16201746e-01 -1.32984412e+00
-1.10751081e+00 -1.26022649e+00 -5.42800128e-01 -5.53297937e-01
5.69728732e-01 -5.87542057e-01 -9.30218920e-02 2.26084903e-01
-9.92905140e-01 -2.53675170e-02 -3.94060165e-01 3.31826508e-01
-3.68833303e-01 4.09060001e-01 -2.46359691e-01 -1.24671209e+00
-4.21876550e-01 -4.85820949e-01 5.97007155e-01 5.31032145e-01
-6.89971089e-01 -9.68220949e-01 4.52774197e-01 1.39380977e-01
5.12526333e-01 -3.02140743e-01 6.66201293e-01 -1.58844352e+00
1.50941029e-01 -4.70456988e-01 3.60352308e-01 -1.06678389e-01
5.41033685e-01 -1.62115008e-01 -1.06861067e+00 2.14582980e-02
-3.27828340e-02 -4.71139371e-01 9.12678242e-01 1.60347313e-01
7.87737727e-01 -4.75904286e-01 -4.85507280e-01 1.14587910e-01
1.25140095e+00 4.90942240e-01 3.60124677e-01 2.36467227e-01
2.17247963e-01 5.83043993e-01 6.93373024e-01 6.19627774e-01
3.08513135e-01 5.10451972e-01 -2.85791278e-01 2.19090909e-01
4.15869296e-01 -2.87117422e-01 -7.19521493e-02 7.95896947e-01
8.68513063e-02 -6.31917790e-02 -1.23323321e+00 6.71312332e-01
-2.10412526e+00 -1.37909579e+00 4.57684249e-01 1.94014692e+00
1.16994917e+00 2.97279984e-01 2.42843479e-01 3.31160575e-01
8.30062509e-01 3.54219675e-01 -4.46282774e-01 -2.28873968e-01
-2.56459676e-02 2.89020330e-01 5.93201518e-02 5.10794044e-01
-8.05367768e-01 1.19877279e+00 7.25547886e+00 9.79624867e-01
-9.57823396e-01 5.05194247e-01 2.44673878e-01 -1.06959268e-01
-3.13933462e-01 2.10470587e-01 -8.51681530e-01 3.84397894e-01
1.12090611e+00 -8.29966843e-01 2.44512007e-01 8.65240276e-01
2.73545414e-01 -3.28545004e-01 -9.39412355e-01 1.10636544e+00
6.77438498e-01 -1.48933935e+00 1.00884743e-01 -5.69311202e-01
6.17602766e-01 -8.68340954e-02 -4.63243276e-01 5.42837620e-01
2.73619622e-01 -8.94079924e-01 -1.58122480e-02 6.77486777e-01
3.75899464e-01 -6.08190417e-01 3.35788786e-01 7.66395450e-01
-9.63246584e-01 -2.09238887e-01 -1.82679892e-01 -2.09173068e-01
4.65677977e-02 4.09758955e-01 -9.86578882e-01 1.29626036e-01
7.93031991e-01 3.76255721e-01 -3.09546322e-01 1.35858274e+00
1.05238281e-01 7.74007976e-01 -1.62725598e-01 -5.56157649e-01
3.26221436e-01 2.52900034e-01 6.65811658e-01 1.29138124e+00
1.04594886e-01 4.35488671e-01 4.07524168e-01 6.37155116e-01
1.19071864e-01 3.89022470e-01 -7.66924500e-01 -4.77321804e-01
7.31082439e-01 1.31288755e+00 -9.18657184e-01 -5.53457677e-01
-4.56713319e-01 1.10778022e+00 -3.64338607e-02 4.10312057e-01
-2.80762941e-01 -7.75288880e-01 2.64782250e-01 3.53177786e-01
9.06226039e-02 -5.47919050e-02 8.87478068e-02 -1.12433112e+00
-6.65372670e-01 -7.07019031e-01 6.18325293e-01 -9.30538952e-01
-1.22534847e+00 4.17876959e-01 3.04164380e-01 -1.05233371e+00
-4.78828937e-01 -2.51604989e-02 -9.27430630e-01 2.90252864e-01
-1.29954195e+00 -7.91622460e-01 -2.15312183e-01 6.35120213e-01
1.04586720e+00 -4.04168457e-01 1.13052809e+00 1.50355741e-01
-1.45635456e-01 4.97932017e-01 1.32020965e-01 -2.32162885e-02
9.87766802e-01 -1.15751505e+00 -7.33413249e-02 4.70366329e-01
6.94972813e-01 6.64932609e-01 1.09525073e+00 -8.02160025e-01
-1.28509390e+00 -6.44733727e-01 1.25182247e+00 -1.98072881e-01
7.08578765e-01 -2.73050666e-01 -1.19231153e+00 2.40363181e-01
4.20725703e-01 -1.51854768e-01 1.04856157e+00 2.26224512e-01
-2.54651219e-01 -1.98521808e-01 -7.32334018e-01 4.27838355e-01
7.11230338e-01 -6.46696031e-01 -1.18397176e+00 6.98476732e-01
9.49537754e-01 1.17880769e-01 -6.74036264e-01 3.76384974e-01
2.50324309e-01 -6.21510923e-01 6.72999203e-01 -9.92890716e-01
2.80381064e-03 3.98523547e-02 -1.52997196e-01 -1.48507643e+00
-1.80726171e-01 -7.84265816e-01 -3.62394154e-01 1.48859704e+00
6.44881129e-01 -2.57582963e-01 6.67082250e-01 5.48761964e-01
3.24061841e-01 -4.78799462e-01 -8.20706308e-01 -9.81655717e-01
-1.81590319e-01 -5.63019633e-01 3.26047540e-01 1.33694530e+00
8.58311832e-01 5.82913816e-01 -6.04780614e-01 -2.83720016e-01
5.26553154e-01 1.04001202e-01 4.36787665e-01 -1.35966623e+00
-5.58703057e-02 -3.13882262e-01 -1.22070618e-01 -5.99381745e-01
2.89875060e-01 -7.31765866e-01 3.70916128e-01 -1.54745007e+00
3.65343243e-01 -3.37436795e-02 -4.42722291e-01 5.20617366e-01
-9.46976542e-02 -1.93998232e-01 -2.20284671e-01 2.74608403e-01
-9.68452990e-01 3.05600077e-01 5.12494266e-01 -3.12036157e-01
-6.15350783e-01 3.11853886e-01 -5.88692248e-01 6.31647289e-01
7.90577173e-01 -5.45562625e-01 -6.59302235e-01 8.11560154e-02
2.29771495e-01 2.44362489e-03 -2.17835248e-01 -7.35426307e-01
1.08544576e+00 -3.45511675e-01 -3.14788036e-02 -5.05531073e-01
2.65145093e-01 -4.54030186e-01 -2.49673352e-01 1.87397391e-01
-6.13217115e-01 -2.60279208e-01 -1.25841349e-01 7.71969914e-01
-2.42427751e-01 -8.26998115e-01 6.14578605e-01 -8.38754594e-01
-1.41755748e+00 1.83420524e-01 -7.55558193e-01 3.04713726e-01
1.04232466e+00 -3.69051933e-01 2.29161948e-01 -8.46323311e-01
-8.50613415e-01 2.57131547e-01 2.33797263e-03 7.56167412e-01
9.18059170e-01 -1.09178090e+00 -3.07493448e-01 -1.55718639e-01
6.55047178e-01 -6.31452262e-01 -2.10205227e-01 4.07574445e-01
2.25536868e-01 3.18888128e-01 2.35349566e-01 -3.30995888e-01
-1.33045459e+00 4.51456100e-01 3.62693161e-01 2.57126123e-01
-4.60570306e-01 8.19178522e-01 -8.53426337e-01 -4.32621181e-01
8.92105758e-01 4.41870511e-01 -5.75665593e-01 6.12882257e-01
9.19010162e-01 2.69897848e-01 -1.58450782e-01 -3.89134794e-01
-3.58143717e-01 3.02692771e-01 -2.30421007e-01 -2.73217708e-01
1.24841917e+00 -2.90089339e-01 8.22761282e-02 1.28989315e+00
1.00510788e+00 -5.36880612e-01 -3.39989603e-01 -7.36000121e-01
5.75671911e-01 -2.35775828e-01 2.58358866e-01 -6.82430267e-01
-5.75255334e-01 4.71926540e-01 6.10221267e-01 5.15884280e-01
8.19665372e-01 2.66128480e-01 7.35354900e-01 7.86118984e-01
4.23033565e-01 -1.67005002e+00 3.38149935e-01 6.14284456e-01
3.45844775e-01 -1.45350182e+00 -2.81611141e-02 -9.30443332e-02
-5.57437599e-01 1.12491190e+00 7.33000934e-01 1.94707185e-01
1.15520954e+00 3.13963562e-01 5.45902178e-02 -3.13431084e-01
-1.00494850e+00 -6.44627988e-01 3.76067787e-01 3.41282099e-01
2.77564347e-01 -2.87181795e-01 -4.39754784e-01 8.01279843e-01
1.76614299e-01 2.68215477e-01 3.90607834e-01 1.29194820e+00
-1.10153449e+00 -1.10069108e+00 -1.60774931e-01 8.28551888e-01
-2.10191667e-01 -2.18156904e-01 -7.35965490e-01 3.99021596e-01
-4.97827046e-02 1.24526370e+00 2.51434371e-02 -6.59731269e-01
1.87145263e-01 7.91234672e-01 8.37049186e-02 -1.25310385e+00
-4.38935071e-01 1.21748433e-01 2.34036092e-02 -8.87061208e-02
-7.21593559e-01 -3.20464134e-01 -1.12912822e+00 -3.44855860e-02
-8.39377463e-01 6.77137792e-01 3.01856279e-01 1.41424537e+00
8.07594508e-02 1.42441168e-01 8.52906406e-01 -3.63584340e-01
-5.53343415e-01 -1.10111928e+00 -7.96940863e-01 2.14572296e-01
5.28500490e-02 -4.72907066e-01 -7.46713102e-01 -5.06406464e-02] | [10.21362018585205, 3.5487568378448486] |
f7ed796c-cd66-4f51-bcd8-18bd1b109587 | holistic-recognition-of-low-quality-license | null | null | https://ieeexplore.ieee.org/abstract/document/8078501 | https://sci-hub.tw/https://ieeexplore.ieee.org/abstract/document/8078501 | Holistic Recognition of Low Quality License Plates by CNN using Track Annotated Data | This work is focused on recognition of license plates in low resolution and low quality images. We present a methodology for collection of real world (non-synthetic) dataset of low quality license plate images with ground truth transcriptions. Our approach to the license plate recognition is based on a Convolutional Neural Network which holistically processes the whole image, avoiding segmentation of the license plate characters. Evaluation results on multiple datasets show that our method significantly outperforms other free and commercial solutions to license plate recognition on the low quality data. To enable further research of low quality license plate recognition, we make the datasets publicly available. | ['Pavel Zemˇc´ık', 'Luk´aˇs Marˇs´ık', 'Roman Jur´anek', 'Jakub ˇSpaˇnhel', 'Adam Herout', 'Jakub Sochor'] | 2017-10-23 | null | null | null | ieee-international-conference-on-advanced | ['license-plate-recognition'] | ['computer-vision'] | [ 3.25022966e-01 -5.31450927e-01 2.19221905e-01 -4.99381632e-01
-1.43674111e+00 -1.13122988e+00 5.27581036e-01 -1.01227164e+00
-1.77654371e-01 6.61032498e-01 -1.17417186e-01 -1.42882243e-01
2.14817435e-01 -9.07824636e-01 -8.93995166e-01 -4.70306814e-01
5.43187082e-01 5.33380389e-01 4.25558031e-01 -1.60620958e-01
8.25241029e-01 7.13326454e-01 -1.21353889e+00 7.08764374e-01
4.19729233e-01 8.05151105e-01 -1.13352157e-01 8.90133142e-01
-5.00515513e-02 8.22474301e-01 -7.86053717e-01 -8.27425897e-01
1.08013141e+00 -2.09677760e-02 -7.38631368e-01 8.45530450e-01
1.01914513e+00 -3.21846694e-01 -5.52662730e-01 1.21472812e+00
2.78076619e-01 -4.89643067e-02 8.62527847e-01 -1.01741111e+00
-1.13823855e+00 -1.05645739e-01 -3.75530384e-02 5.19305885e-01
2.65130997e-01 1.86160475e-01 4.53366548e-01 -1.36787391e+00
6.55440927e-01 5.82746387e-01 1.06217492e+00 1.53685138e-01
-7.57526398e-01 -6.71035767e-01 -8.13362658e-01 -2.68460691e-01
-1.99424636e+00 -8.99509192e-01 3.95837694e-01 -5.66290021e-01
8.94055963e-01 2.20872000e-01 2.07524210e-01 6.90688848e-01
7.00410306e-02 6.40880048e-01 1.32178545e+00 -4.89206463e-01
4.15250566e-03 2.93550760e-01 2.80100644e-01 6.69732511e-01
3.73396933e-01 -6.44907653e-02 -1.56543732e-01 4.83210862e-01
1.31071055e+00 4.91630025e-02 3.93822528e-02 2.30025992e-01
-9.11454022e-01 3.80844563e-01 -2.29169652e-01 3.67381662e-01
4.84875031e-03 -1.50348157e-01 -4.56253067e-02 2.46186644e-01
2.15898216e-01 4.23837870e-01 4.24600281e-02 -1.98730856e-01
-1.62184393e+00 1.32486597e-01 7.11188376e-01 1.56594646e+00
8.60327601e-01 3.13752353e-01 -9.86343175e-02 9.04634356e-01
4.90282208e-01 7.53074527e-01 3.90495747e-01 -8.11368585e-01
9.06158686e-01 5.12726545e-01 5.29110551e-01 -1.00301397e+00
1.01037286e-01 -6.58731386e-02 -6.81577742e-01 5.97103894e-01
2.54509866e-01 1.54674530e-01 -1.33223569e+00 2.01845989e-01
-8.69901955e-01 2.26770878e-01 3.23229641e-01 8.98073494e-01
9.35091376e-01 9.89716709e-01 -5.04583061e-01 4.86367851e-01
1.21250415e+00 -1.26230133e+00 -6.75748050e-01 -3.75255913e-01
1.13299355e-01 -1.20303106e+00 8.54806125e-01 2.92428523e-01
-9.71275151e-01 -6.69903815e-01 -1.37882113e+00 2.03677174e-02
-7.05521524e-01 9.52232897e-01 3.05503700e-03 1.24340260e+00
-1.28606188e+00 1.72103986e-01 -2.26797208e-01 -8.97749886e-02
6.51638389e-01 5.44215381e-01 -7.92766511e-01 -5.06277025e-01
-7.24249542e-01 8.58471572e-01 1.24101609e-01 2.07851186e-01
-5.85612059e-01 -1.58536971e-01 -6.21092141e-01 -2.42956862e-01
-4.01097117e-03 1.80067625e-02 1.02712977e+00 -1.31851780e+00
-1.65065515e+00 1.32986271e+00 2.53722042e-01 -9.78462026e-02
3.29859853e-01 1.27206132e-01 -9.68834043e-01 1.51946828e-01
5.49992472e-02 5.03895104e-01 8.60209644e-01 -1.25161099e+00
-5.88438034e-01 -1.76766172e-01 -2.76431650e-01 1.50983155e-01
2.98379540e-01 3.28894883e-01 -8.98805022e-01 -5.49607337e-01
1.16819046e-01 -9.70200181e-01 2.16587707e-01 -5.88229477e-01
-5.91987610e-01 3.14789951e-01 6.14525497e-01 -9.70744729e-01
5.00852764e-01 -2.37052989e+00 -9.54047263e-01 3.49643379e-01
-3.05857472e-02 6.78070307e-01 4.74482030e-02 2.67096847e-01
6.41230345e-02 5.38832486e-01 -3.12307030e-01 -4.64472145e-01
2.18297884e-01 2.14404520e-03 -3.89987111e-01 6.52049005e-01
4.13617760e-01 1.05099940e+00 1.90269455e-01 -4.02927130e-01
4.11386669e-01 2.26457536e-01 4.79939096e-02 7.91515186e-02
5.08519590e-01 -2.25468352e-01 -1.98663265e-01 1.13184559e+00
1.44289601e+00 -4.08178940e-02 -6.17908359e-01 -6.66752905e-02
-3.98942351e-01 -2.83402443e-01 -1.27236104e+00 1.20535028e+00
8.95662140e-03 1.22185516e+00 -6.00443892e-02 -5.25477827e-01
1.34975994e+00 3.59338194e-01 -8.44649225e-02 -8.42644751e-01
-1.08210944e-01 4.15158629e-01 -5.67660034e-01 -6.36019409e-01
9.49280798e-01 -2.33677492e-01 -2.77488768e-01 3.62095842e-03
2.08368808e-01 -1.97647974e-01 2.62891978e-01 -2.05862135e-01
6.89791858e-01 1.42447967e-02 -1.09324969e-01 -4.98826280e-02
8.13041687e-01 3.11461538e-01 1.96073368e-01 7.77664185e-01
-5.43442369e-01 1.33452070e+00 1.33926533e-02 -6.42691910e-01
-1.72846854e+00 -1.02151287e+00 5.04446812e-02 4.34996694e-01
9.97992754e-02 2.71538168e-01 -9.73404348e-01 -7.05649853e-02
-4.28191632e-01 1.83430061e-01 -4.59008068e-01 5.65816998e-01
-6.28458679e-01 -7.67043293e-01 1.47030854e+00 5.12110889e-01
1.20440316e+00 -7.71210253e-01 2.96563298e-01 -1.18062459e-02
5.54410927e-02 -1.60538888e+00 -6.47161901e-01 -2.73242146e-01
-6.00337684e-01 -9.67857718e-01 -7.97075808e-01 -1.21134174e+00
6.35322213e-01 4.16743606e-01 1.07087302e+00 -2.04640806e-01
1.20086223e-01 1.46879256e-01 -9.94774848e-02 -2.14376912e-01
-6.19259298e-01 -1.29170850e-01 -3.64913672e-01 2.84561783e-01
7.45303690e-01 -5.84626421e-02 -7.37397000e-02 6.97448492e-01
-1.11859012e+00 -4.40076739e-02 1.08596218e+00 1.27961308e-01
5.57210803e-01 2.58754879e-01 -4.04177047e-03 -5.31251669e-01
7.85673022e-01 -2.12002859e-01 -1.12088609e+00 2.82695860e-01
-3.19492728e-01 -3.05548787e-01 4.60502565e-01 2.69249082e-01
-9.56657827e-01 4.36237007e-01 -2.64470160e-01 -4.67472106e-01
-8.03951025e-01 9.15958488e-04 -2.51762718e-01 -5.00199795e-01
4.54163760e-01 9.72138584e-01 -8.51506814e-02 -3.23213398e-01
-1.14331350e-01 1.07441401e+00 9.24987495e-01 1.86364576e-02
1.05295849e+00 2.95505702e-01 -2.54119217e-01 -9.32486832e-01
-1.38078436e-01 -8.24703753e-01 -9.38201368e-01 -6.46636605e-01
9.51824188e-01 -9.26778555e-01 -4.50566918e-01 1.06876361e+00
-9.48841214e-01 1.75153032e-01 1.16051100e-01 3.12783867e-01
-4.75376517e-01 2.46477634e-01 -6.52920127e-01 -7.50282645e-01
4.32086624e-02 -1.26601899e+00 1.30796075e+00 4.52924728e-01
5.96974790e-01 -6.69730842e-01 3.69042218e-01 1.10061920e+00
6.07097507e-01 4.32982575e-03 1.80110440e-01 -1.05233443e+00
-1.39528239e+00 -7.33918548e-01 -5.80799639e-01 7.50858665e-01
-2.18350127e-01 3.09850872e-01 -1.13849425e+00 3.78906041e-01
-2.82301486e-01 -2.94866294e-01 6.95227444e-01 3.44553143e-01
3.25246602e-01 -3.62693280e-01 1.36952236e-01 7.12675512e-01
1.96316826e+00 2.37606630e-01 1.47960639e+00 7.05080271e-01
6.80098116e-01 2.43308559e-01 2.50718057e-01 3.07038520e-02
7.47785792e-02 6.49872065e-01 6.02161884e-02 -1.22055210e-01
1.89631544e-02 -7.55543858e-02 5.43212175e-01 8.73019636e-01
-5.20398796e-01 -6.14057600e-01 -1.25419939e+00 2.59500325e-01
-1.24228477e+00 -1.34225810e+00 -4.17641431e-01 1.87474930e+00
1.99871406e-01 1.03384480e-01 1.89696297e-01 1.26413167e-01
1.12792516e+00 -4.82624657e-02 1.61590546e-01 -3.93603891e-01
-6.47152483e-01 -1.54088095e-01 1.13540542e+00 3.77571672e-01
-1.47433722e+00 1.11926067e+00 7.82214260e+00 8.91088665e-01
-1.06184053e+00 -1.20267197e-01 7.85860300e-01 4.13571358e-01
1.94707453e-01 -5.27117491e-01 -1.04647315e+00 4.29084271e-01
1.25440872e+00 -2.13235226e-02 3.41023892e-01 7.26201355e-01
-2.74065100e-02 2.03199044e-01 -6.57346725e-01 1.23365629e+00
5.33826470e-01 -1.90625274e+00 9.91552994e-02 3.87152791e-01
8.71663034e-01 3.85992616e-01 2.78442293e-01 2.41097450e-01
1.78523034e-01 -1.54716599e+00 8.49865496e-01 9.72057641e-01
8.56105685e-01 -9.34981942e-01 1.33156633e+00 3.03710759e-01
-8.21995974e-01 2.93220609e-01 -6.82032168e-01 2.94740289e-01
-2.99436003e-01 -1.14896536e-01 -1.14921319e+00 3.24480236e-01
3.58879775e-01 7.07511604e-01 -9.76730704e-01 1.18399930e+00
3.28844011e-01 7.08736479e-01 -7.92463049e-02 1.72331735e-01
6.61080778e-01 -5.40272593e-01 1.63608015e-01 1.69792402e+00
5.34396589e-01 1.03302933e-01 -1.46851227e-01 1.14031327e+00
-3.18487614e-01 4.42362875e-02 -9.96625781e-01 -2.84370124e-01
2.29001150e-01 1.05812097e+00 -1.01023042e+00 -2.33308226e-01
-6.27601445e-01 1.03445601e+00 -3.84661198e-01 3.07047069e-01
-6.73996210e-01 -4.13617492e-01 2.17489079e-01 1.07593231e-01
5.88350058e-01 -3.42275590e-01 -3.86876047e-01 -1.08746004e+00
1.68297961e-01 -7.61963964e-01 -1.37862518e-01 -1.15922821e+00
-1.08412910e+00 6.82590365e-01 -3.46249133e-01 -1.85259116e+00
-1.09435581e-01 -1.23880196e+00 -5.69687605e-01 1.02608979e+00
-1.32862997e+00 -1.34096909e+00 -2.15285361e-01 7.94605613e-01
8.37526619e-01 -9.69735086e-01 7.43342638e-01 5.54033399e-01
-5.08719742e-01 3.53351265e-01 8.12215686e-01 1.03381944e+00
5.76023340e-01 -9.60029602e-01 7.13216364e-01 1.20935822e+00
1.84468329e-01 4.10767615e-01 2.12131605e-01 -4.90032315e-01
-1.29303861e+00 -1.38480103e+00 6.34149790e-01 -7.28919744e-01
2.82639470e-02 -3.61427367e-01 -7.96009064e-01 8.52981031e-01
4.14353251e-01 9.91862640e-02 1.04665446e+00 -8.40248883e-01
-8.18266720e-02 -1.45266861e-01 -1.41927850e+00 6.10906370e-02
2.08512142e-01 -4.69465762e-01 -7.58302867e-01 2.82239288e-01
3.55273634e-02 -1.07709758e-01 -7.63949811e-01 -2.03569993e-01
4.71703947e-01 -1.00821412e+00 1.01960480e+00 5.85803948e-02
6.54229522e-01 -7.77124047e-01 -5.51027119e-01 -6.65302098e-01
-3.65777910e-01 -1.27185971e-01 8.23671341e-01 1.26571929e+00
6.58187866e-01 -1.55495703e-01 1.01272357e+00 8.19836676e-01
-4.65125769e-01 1.73188180e-01 -1.03582406e+00 -9.00966644e-01
-2.75169238e-02 -4.69712466e-01 4.80658889e-01 7.74612665e-01
-7.06641495e-01 -2.71904379e-01 -6.74794316e-01 4.40053940e-01
4.87243444e-01 -6.47307187e-02 7.78556466e-01 -8.78925145e-01
2.53764093e-01 -4.04393822e-01 -1.21457922e+00 -6.60091937e-01
4.56051342e-02 -6.47947252e-01 3.20534736e-01 -1.43865395e+00
7.61351660e-02 9.41331387e-02 -1.24102868e-01 6.03569672e-02
5.41258991e-01 1.26057684e+00 1.51250750e-01 8.42877090e-01
-8.83395791e-01 -1.33306071e-01 1.04163361e+00 -5.35238206e-01
2.73499697e-01 1.29816934e-01 -4.04267341e-01 6.82036698e-01
8.17402780e-01 -5.93952417e-01 1.27165951e-02 -2.25752532e-01
1.16036758e-01 -2.24374738e-02 2.93140322e-01 -1.36315596e+00
5.87070107e-01 9.43112094e-03 7.78875053e-01 -8.59669685e-01
2.70890385e-01 -5.77323794e-01 3.12623709e-01 -2.38809675e-01
-6.37458861e-02 1.38698399e-01 3.31716537e-01 3.23226780e-01
-8.41522217e-01 -5.64539254e-01 8.85163724e-01 -5.46603084e-01
-1.21218371e+00 1.55170754e-01 -1.06166875e+00 -1.83048531e-01
9.81669128e-01 -1.04409182e+00 -2.56349206e-01 -1.81948811e-01
-4.77827191e-01 -6.92752779e-01 9.22515035e-01 4.12755281e-01
7.70857751e-01 -1.46278977e+00 -9.59730446e-01 4.24526662e-01
2.91657627e-01 -4.34519440e-01 4.35786657e-02 1.39604300e-01
-1.54719102e+00 1.05457962e+00 -9.14740801e-01 -3.17687720e-01
-1.17400229e+00 4.85284887e-02 8.68887842e-01 -5.43643199e-02
-2.90719032e-01 7.52560258e-01 -4.50614184e-01 -6.43349588e-01
-3.44604462e-01 -8.20980296e-02 -4.73822713e-01 -3.79012048e-01
6.27319038e-01 4.71190125e-01 5.09987533e-01 -1.33987856e+00
-4.13864523e-01 7.79362857e-01 5.99040054e-02 -4.84062821e-01
1.48450720e+00 7.75253326e-02 -4.94555831e-02 3.17452639e-01
1.23516858e+00 1.17914125e-01 -1.24403453e+00 1.41995758e-01
5.38379438e-02 -9.04304206e-01 -1.12174831e-01 -8.59082937e-01
-1.02041245e+00 5.10125816e-01 7.40121722e-01 2.68658213e-02
8.81728053e-01 -6.05833888e-01 3.75210583e-01 8.01069021e-01
3.19039285e-01 -1.55780482e+00 -2.00467274e-01 6.01473093e-01
9.18363452e-01 -1.34814775e+00 -1.73527136e-01 -3.42931926e-01
-7.79277921e-01 1.74579453e+00 3.64356339e-01 -6.04338646e-01
2.57497102e-01 7.13191986e-01 3.42234522e-01 -1.65516779e-01
-1.54969841e-01 2.24667609e-01 4.47446138e-01 8.17692757e-01
3.42076033e-01 -7.65008032e-02 3.95837605e-01 6.76943243e-01
-3.53549719e-02 2.12271184e-01 1.15363610e+00 8.49508822e-01
-5.10108531e-01 -9.52308536e-01 -1.02054667e+00 3.15612227e-01
-6.65670156e-01 -3.52445781e-01 -8.95030618e-01 1.07583928e+00
3.51152360e-01 8.24110925e-01 2.29949519e-01 -3.45053524e-01
4.43001360e-01 4.71475661e-01 1.44456476e-01 -3.24923873e-01
-4.15279895e-01 -2.23308359e-03 1.74476177e-01 -1.40879884e-01
-7.61948884e-01 -6.13774002e-01 -7.56577909e-01 -4.76514995e-01
-9.39129367e-02 -2.92626739e-01 1.00977218e+00 9.25900042e-01
2.42832035e-01 2.11781949e-01 5.43785155e-01 -8.75486255e-01
-1.21437036e-01 -8.75186920e-01 -1.02447557e+00 2.82261819e-01
2.56653458e-01 -5.21090254e-02 -1.20288566e-01 7.39724100e-01] | [9.835884094238281, -4.942375659942627] |
8c93cc1f-3d93-4f8a-a1aa-9a1823f09a8c | xcon-learning-with-experts-for-fine-grained | 2208.01898 | null | https://arxiv.org/abs/2208.01898v1 | https://arxiv.org/pdf/2208.01898v1.pdf | XCon: Learning with Experts for Fine-grained Category Discovery | We address the problem of generalized category discovery (GCD) in this paper, i.e. clustering the unlabeled images leveraging the information from a set of seen classes, where the unlabeled images could contain both seen classes and unseen classes. The seen classes can be seen as an implicit criterion of classes, which makes this setting different from unsupervised clustering where the cluster criteria may be ambiguous. We mainly concern the problem of discovering categories within a fine-grained dataset since it is one of the most direct applications of category discovery, i.e. helping experts discover novel concepts within an unlabeled dataset using the implicit criterion set forth by the seen classes. State-of-the-art methods for generalized category discovery leverage contrastive learning to learn the representations, but the large inter-class similarity and intra-class variance pose a challenge for the methods because the negative examples may contain irrelevant cues for recognizing a category so the algorithms may converge to a local-minima. We present a novel method called Expert-Contrastive Learning (XCon) to help the model to mine useful information from the images by first partitioning the dataset into sub-datasets using k-means clustering and then performing contrastive learning on each of the sub-datasets to learn fine-grained discriminative features. Experiments on fine-grained datasets show a clear improved performance over the previous best methods, indicating the effectiveness of our method. | ['Bingchen Zhao', 'Siwei Yang', 'Zhongkai Zhao', 'Yixin Fei'] | 2022-08-03 | null | null | null | null | ['novel-concepts'] | ['reasoning'] | [ 3.12013954e-01 8.02830160e-02 -2.50131339e-01 -6.73200130e-01
-6.66438401e-01 -7.85849273e-01 6.03753328e-01 6.15799762e-02
-2.14284182e-01 4.84995693e-01 -6.58671111e-02 6.61833957e-03
-6.11834109e-01 -5.55669188e-01 -4.50195163e-01 -1.11262393e+00
-9.40774530e-02 6.72236443e-01 2.13227242e-01 1.52304903e-01
5.46120286e-01 3.54192764e-01 -2.09216952e+00 5.83448648e-01
8.95685196e-01 1.07435501e+00 3.37855816e-01 2.08314776e-01
-3.55052412e-01 5.76727509e-01 -4.19226825e-01 2.37997785e-01
4.19064671e-01 -3.13083291e-01 -8.93531680e-01 6.06315851e-01
6.12786412e-01 1.30200520e-01 3.43569160e-01 1.23070836e+00
6.18188195e-02 3.83478552e-01 1.04832923e+00 -1.37089670e+00
-6.04889214e-01 3.26474220e-01 -8.98047328e-01 3.28576893e-01
8.14081952e-02 -2.56007850e-01 1.03421974e+00 -1.17750835e+00
3.65533441e-01 1.49798262e+00 3.68218273e-01 7.33981907e-01
-1.10750890e+00 -7.27010667e-01 6.35217667e-01 5.23477018e-01
-1.53489721e+00 -2.16125056e-01 9.08762932e-01 -7.80841768e-01
2.25771800e-01 3.39822054e-01 1.45290360e-01 7.82456458e-01
-2.47148350e-01 7.49616802e-01 1.51498914e+00 -5.10721862e-01
6.13937676e-01 4.96276408e-01 7.51521289e-01 4.40514237e-01
1.54869676e-01 3.12231153e-01 -1.47052795e-01 -3.89541760e-02
3.93760055e-01 3.64385515e-01 -1.76059768e-01 -7.97816932e-01
-9.79387105e-01 9.89330590e-01 6.52931750e-01 4.10103112e-01
-1.72078192e-01 -3.24601352e-01 1.07557528e-01 3.53279591e-01
4.38205808e-01 4.20782298e-01 -6.07983351e-01 4.80264992e-01
-7.81601906e-01 7.19427457e-03 4.75076735e-01 6.42106116e-01
1.12702835e+00 -2.82010168e-01 -2.76866723e-02 9.19845998e-01
5.16294837e-01 1.36499301e-01 7.37602115e-01 -6.88896835e-01
1.26421660e-01 7.49798775e-01 -4.63875867e-02 -8.72680247e-01
-1.62281364e-01 -5.64158559e-01 -7.67853737e-01 3.95447969e-01
4.75444585e-01 1.67857736e-01 -1.47059572e+00 1.59658337e+00
5.31702042e-01 2.92107433e-01 4.21878882e-02 1.06275129e+00
8.06770325e-01 5.87905109e-01 1.36873469e-01 -4.27853167e-01
1.01892555e+00 -9.94168639e-01 -4.57310438e-01 -3.57229024e-01
2.18986288e-01 -5.15559673e-01 1.01307285e+00 5.46841621e-01
-4.03762519e-01 -1.00116336e+00 -1.04194117e+00 4.26418483e-01
-6.93061531e-01 -1.95303589e-01 5.29051304e-01 4.30489182e-01
-7.36782849e-01 5.28183103e-01 -5.42425275e-01 -2.25527793e-01
5.00682950e-01 4.80280370e-01 -2.51679748e-01 -3.83748531e-01
-6.28220081e-01 4.57286894e-01 8.27113211e-01 1.22271581e-02
-1.13947797e+00 -4.63264406e-01 -6.44691229e-01 2.13471018e-02
6.89510286e-01 -1.95085257e-01 8.18363428e-01 -1.66708684e+00
-7.18835890e-01 1.03933918e+00 -1.87546968e-01 -1.51966214e-02
1.71049044e-01 1.89252887e-02 -5.31583905e-01 1.50479332e-01
4.10230964e-01 6.33361697e-01 1.19272852e+00 -1.97485924e+00
-1.03668928e+00 -6.67992413e-01 -1.41028404e-01 2.28486836e-01
-1.79781035e-01 -3.55328202e-01 -1.84939682e-01 -7.38363326e-01
7.95007527e-01 -9.56732810e-01 -2.10325688e-01 -1.95055515e-01
-5.59557378e-01 -6.45061076e-01 1.33093989e+00 -1.54335991e-01
8.97198796e-01 -2.26392913e+00 -1.48571804e-01 5.37503064e-01
4.80849773e-01 2.26744004e-02 -9.13825557e-02 5.04730381e-02
-5.05484760e-01 7.81648308e-02 -2.28441194e-01 7.08540231e-02
-1.48588121e-01 2.78476894e-01 -3.53289902e-01 3.98809075e-01
2.77135164e-01 5.73702633e-01 -1.06430185e+00 -4.79231805e-01
1.26408786e-01 1.47001237e-01 -2.25361586e-01 3.44453365e-01
-1.10758319e-01 6.63607776e-01 -6.39050484e-01 7.27887988e-01
7.95208931e-01 -4.36293691e-01 1.23435631e-01 -1.01170316e-01
6.33522449e-03 -7.00016022e-02 -1.66167068e+00 1.19212580e+00
-1.82955544e-02 2.07928583e-01 -1.81214303e-01 -1.52294004e+00
7.87039816e-01 6.41657412e-02 2.85757065e-01 -4.52349812e-01
-6.13302924e-02 1.33302689e-01 1.26240700e-01 -4.33973163e-01
4.36759479e-02 -5.01739264e-01 2.75455881e-03 5.22527158e-01
1.43802971e-01 1.96506411e-01 1.96960513e-02 1.35360062e-01
6.03539705e-01 -1.39544293e-01 4.01795745e-01 -5.15082121e-01
3.65529776e-01 1.23860344e-01 6.66990995e-01 1.05043864e+00
-2.17458799e-01 5.79126656e-01 -7.72506231e-03 -4.85131562e-01
-6.18119061e-01 -1.29472542e+00 -2.93210387e-01 1.30228746e+00
4.18170869e-01 3.21010947e-02 -4.62444752e-01 -1.16124451e+00
4.54588830e-02 3.30691308e-01 -1.00257254e+00 -1.16593115e-01
-1.48078755e-01 -7.52049685e-01 -2.40447119e-01 5.20787179e-01
4.19779688e-01 -9.97950852e-01 -3.46921474e-01 -6.92955917e-04
1.82753224e-02 -5.97197831e-01 -2.54456460e-01 7.06707001e-01
-8.87778342e-01 -1.41280687e+00 -3.69301438e-01 -1.27202547e+00
1.02816939e+00 6.70225859e-01 1.15191472e+00 1.34560078e-01
-3.64268750e-01 4.47932273e-01 -4.97674018e-01 -3.37773502e-01
-5.18826246e-02 -5.80365181e-01 2.75128007e-01 4.70446020e-01
7.16649830e-01 -3.50408137e-01 -4.47143406e-01 5.21659613e-01
-8.26138973e-01 -2.94760644e-01 6.54949725e-01 1.08709037e+00
9.76138413e-01 5.77308238e-01 7.86569297e-01 -1.36789954e+00
2.47209281e-01 -7.90786326e-01 -3.61672431e-01 2.74296463e-01
-8.55112016e-01 1.05868511e-01 5.70766628e-01 -6.52859032e-01
-1.12929702e+00 2.02341959e-01 5.88532329e-01 -6.03010774e-01
-6.18988395e-01 4.22640622e-01 -4.32234198e-01 7.97839537e-02
7.12541282e-01 1.33797571e-01 -2.60022730e-01 -5.23609579e-01
4.62256938e-01 7.76929498e-01 5.63485503e-01 -5.64555168e-01
1.00420702e+00 6.25591278e-01 -1.78703964e-01 -6.53699160e-01
-1.25350773e+00 -1.23110700e+00 -1.08161187e+00 -6.12165369e-02
8.10939252e-01 -8.12790334e-01 -3.15776378e-01 1.71342045e-02
-4.79358226e-01 1.18451472e-02 -4.30813044e-01 4.07115817e-01
-3.74433577e-01 5.01439810e-01 -6.90857768e-02 -8.19579124e-01
6.81543797e-02 -1.03479898e+00 9.92283940e-01 4.39419717e-01
-7.81066790e-02 -1.09146297e+00 -5.96246161e-02 3.39087635e-01
-4.80778702e-02 1.95097089e-01 1.13353324e+00 -1.14415801e+00
-4.70794797e-01 -4.37332019e-02 -2.93072492e-01 4.04766202e-01
4.83997881e-01 -3.23318630e-01 -1.37088764e+00 -4.87090349e-01
1.75766051e-01 -5.58854878e-01 1.04613733e+00 2.99384117e-01
1.40548193e+00 -3.40128154e-01 -6.22006416e-01 3.94562572e-01
1.53514552e+00 3.88882458e-01 2.64965773e-01 3.05077970e-01
8.46817553e-01 9.48923588e-01 1.05600953e+00 3.76467228e-01
-6.61538839e-02 3.93848240e-01 4.38562602e-01 -1.14566527e-01
-3.37672569e-02 -3.93583477e-02 -1.18313029e-01 6.50024951e-01
1.73118025e-01 1.18445098e-01 -8.04746330e-01 5.58546126e-01
-1.89657116e+00 -1.09344292e+00 -8.91350135e-02 2.22153425e+00
6.98405206e-01 6.91736192e-02 1.32293269e-01 3.61078620e-01
1.01953042e+00 -2.75198519e-01 -8.12636554e-01 1.31579060e-02
2.32846439e-02 2.14099035e-01 -3.48631740e-02 2.42189303e-01
-1.63184309e+00 6.97962821e-01 5.84842682e+00 8.62053871e-01
-1.04011130e+00 -1.67933151e-01 8.96872818e-01 3.56488079e-01
-1.55534759e-01 5.95153943e-02 -7.92834699e-01 4.65095729e-01
5.04086494e-01 4.13405783e-02 9.85611677e-02 1.13642228e+00
9.69573259e-02 -6.94986358e-02 -1.40985560e+00 1.11166334e+00
1.65556192e-01 -8.56028855e-01 2.21342832e-01 1.04855187e-01
9.93523955e-01 -2.82343715e-01 2.74595529e-01 4.01469797e-01
4.59026724e-01 -1.04247653e+00 5.11201560e-01 3.39299619e-01
6.01643503e-01 -7.19513059e-01 6.11011326e-01 5.55221915e-01
-1.18864501e+00 -5.21649122e-01 -5.30500829e-01 1.55440113e-02
-5.72292149e-01 6.11404181e-01 -9.78831708e-01 3.00978094e-01
9.76109862e-01 9.58888233e-01 -8.04051280e-01 1.17472720e+00
-3.75470932e-04 6.50499642e-01 -1.80459425e-01 2.21612766e-01
4.86453682e-01 -2.80175477e-01 2.45532662e-01 9.93934512e-01
-1.16502270e-01 2.27938548e-01 7.39984214e-01 8.02699804e-01
5.55647798e-02 -1.20566085e-01 -5.40016413e-01 2.68181026e-01
3.99517447e-01 1.27468920e+00 -1.08339643e+00 -4.57334369e-01
-3.49486470e-01 7.92167544e-01 3.56985092e-01 4.07343626e-01
-3.49147648e-01 -2.32120588e-01 2.64279217e-01 2.76808701e-02
5.23452759e-01 1.85683027e-01 -2.74113595e-01 -1.04412031e+00
-2.05556899e-02 -8.87759209e-01 9.45131600e-01 -5.11181831e-01
-1.83038783e+00 4.05706555e-01 1.52064040e-01 -1.63397658e+00
-2.20220566e-01 -6.14247382e-01 -5.37235498e-01 5.91589808e-01
-1.68738139e+00 -1.06780481e+00 -5.47685564e-01 8.80964458e-01
8.28715324e-01 -1.92094266e-01 7.29188502e-01 -4.08932492e-02
-1.19734779e-01 4.57724601e-01 4.91722554e-01 1.77881524e-01
7.78364360e-01 -1.69649136e+00 -4.15016204e-01 8.36596787e-01
4.43250239e-01 7.03985691e-01 4.94124502e-01 -5.90521097e-01
-9.97893214e-01 -1.27474582e+00 4.12245750e-01 -6.04028881e-01
4.57204670e-01 -4.02029961e-01 -1.14748383e+00 5.88062048e-01
-2.32185796e-01 2.55177855e-01 1.03733885e+00 2.82697350e-01
-7.13115036e-01 -1.01393633e-01 -1.23387754e+00 2.01995775e-01
8.67303908e-01 -3.89836222e-01 -1.04142678e+00 4.21164393e-01
4.29875702e-01 2.55695760e-01 -4.28597718e-01 4.20798510e-01
2.56902486e-01 -7.88384438e-01 8.73491764e-01 -7.61351824e-01
1.04253218e-01 -6.90357983e-01 -3.10274422e-01 -1.43074954e+00
-7.05257773e-01 9.17807370e-02 1.09511372e-02 1.30751646e+00
1.49512157e-01 -4.02301878e-01 9.20330822e-01 3.79243910e-01
-9.35091302e-02 -4.50676829e-01 -6.36350095e-01 -9.01988983e-01
6.16117083e-02 -1.67959794e-01 1.94844007e-01 1.39618731e+00
-2.48894125e-01 6.03053629e-01 2.08675899e-02 3.54409188e-01
1.08588469e+00 6.70485139e-01 4.88527238e-01 -1.76406074e+00
-3.14931393e-01 -2.41804972e-01 -4.41528022e-01 -8.84479821e-01
2.06420541e-01 -9.16941047e-01 3.92172575e-01 -1.24693882e+00
6.19731009e-01 -9.61532295e-01 -6.08318269e-01 5.96416831e-01
-4.33304340e-01 2.77071565e-01 8.94831270e-02 6.60406172e-01
-9.25315857e-01 3.12802136e-01 1.02203929e+00 -5.44985116e-01
-2.23067656e-01 7.71816894e-02 -1.07247818e+00 7.60979354e-01
4.67124164e-01 -6.47753417e-01 -7.17561722e-01 -4.82080467e-02
-4.12429273e-01 -4.82046902e-01 3.30910772e-01 -8.46109450e-01
1.46093175e-01 -3.57948303e-01 7.15606630e-01 -7.00607061e-01
1.18834823e-01 -1.02957833e+00 -2.20670998e-02 2.81520128e-01
-3.38307679e-01 -5.05190253e-01 -9.71606150e-02 9.82295215e-01
-4.50395644e-01 -2.69957632e-01 9.68374729e-01 -4.17662591e-01
-1.42819405e+00 3.53332847e-01 -6.21821471e-02 1.83379799e-01
1.18902266e+00 -5.96039712e-01 -2.43780896e-01 -5.04835844e-02
-1.10436952e+00 2.55128384e-01 3.74274880e-01 4.21604931e-01
7.85753131e-01 -1.23879743e+00 -6.29470944e-01 2.80542552e-01
5.57865441e-01 2.07068801e-01 2.17270508e-01 3.88934612e-01
1.33957863e-01 2.80883372e-01 -2.11747229e-01 -1.02768445e+00
-1.17512929e+00 1.13246155e+00 2.67988145e-01 -6.49820790e-02
-3.94198507e-01 9.27707970e-01 1.04077172e+00 -4.19936538e-01
3.53448868e-01 1.13365814e-01 -7.29947388e-01 2.70395815e-01
6.80098832e-01 1.66875467e-01 1.05038397e-01 -6.45521104e-01
-4.69596416e-01 8.20577323e-01 -4.94280338e-01 4.00806040e-01
1.39468074e+00 -2.20197245e-01 -3.83379832e-02 7.86813617e-01
1.27794814e+00 -2.97349453e-01 -1.32503510e+00 -3.89158517e-01
4.36249167e-01 -5.40783644e-01 -1.03622042e-01 -1.03914690e+00
-9.86589909e-01 8.95894527e-01 1.22102773e+00 1.56340435e-01
1.27223015e+00 4.22838748e-01 1.01892628e-01 4.05910641e-01
3.70664209e-01 -1.22060120e+00 5.26861608e-01 3.17077190e-01
5.58718681e-01 -1.71344948e+00 -7.99460337e-02 -6.21396422e-01
-5.93819618e-01 1.08856809e+00 6.83130980e-01 -3.10707361e-01
9.98985708e-01 -2.92686939e-01 3.30972999e-01 -4.12953347e-01
-5.49607098e-01 -3.48138005e-01 6.76621377e-01 9.04739320e-01
1.00430705e-01 1.99231267e-01 6.33121878e-02 7.61435449e-01
1.49011090e-01 -4.36765522e-01 3.35260510e-01 9.93970394e-01
-7.38155067e-01 -8.79423738e-01 -6.41569734e-01 7.50196636e-01
-3.20135444e-01 5.28340340e-02 -5.30713558e-01 7.40612328e-01
4.57398325e-01 1.20316708e+00 1.44511461e-01 -3.64468098e-01
4.34992984e-02 1.16647519e-01 2.25893006e-01 -1.11056209e+00
-3.14446598e-01 2.40682840e-01 -3.75359207e-01 -3.90662342e-01
-7.04642296e-01 -5.47202945e-01 -1.32421744e+00 2.29460448e-01
-5.03208876e-01 4.94293809e-01 3.04081351e-01 1.21825600e+00
8.92091170e-02 2.72795916e-01 1.14552796e+00 -7.50397623e-01
-5.31933308e-01 -8.23490739e-01 -9.44746494e-01 1.04433882e+00
5.02543449e-01 -1.12740612e+00 -8.02674294e-01 3.57355267e-01] | [9.608814239501953, 2.931100606918335] |
657f9ba5-c896-4ac7-ba8d-fc8a6180bbbe | dragon-decentralized-fault-tolerance-in-edge | 2208.07658 | null | https://arxiv.org/abs/2208.07658v1 | https://arxiv.org/pdf/2208.07658v1.pdf | DRAGON: Decentralized Fault Tolerance in Edge Federations | Edge Federation is a new computing paradigm that seamlessly interconnects the resources of multiple edge service providers. A key challenge in such systems is the deployment of latency-critical and AI based resource-intensive applications in constrained devices. To address this challenge, we propose a novel memory-efficient deep learning based model, namely generative optimization networks (GON). Unlike GANs, GONs use a single network to both discriminate input and generate samples, significantly reducing their memory footprint. Leveraging the low memory footprint of GONs, we propose a decentralized fault-tolerance method called DRAGON that runs simulations (as per a digital modeling twin) to quickly predict and optimize the performance of the edge federation. Extensive experiments with real-world edge computing benchmarks on multiple Raspberry-Pi based federated edge configurations show that DRAGON can outperform the baseline methods in fault-detection and Quality of Service (QoS) metrics. Specifically, the proposed method gives higher F1 scores for fault-detection than the best deep learning (DL) method, while consuming lower memory than the heuristic methods. This allows for improvement in energy consumption, response time and service level agreement violations by up to 74, 63 and 82 percent, respectively. | ['Nicholas R. Jennings', 'Giuliano Casale', 'Shreshth Tuli'] | 2022-08-16 | null | null | null | null | ['fault-detection'] | ['miscellaneous'] | [-5.54318666e-01 -4.87142771e-01 -3.18341047e-01 -1.43729657e-01
-5.73831916e-01 -2.97683120e-01 1.14763714e-02 -1.75033152e-01
2.60535389e-01 8.27587008e-01 -2.29538128e-01 -5.17528594e-01
-2.52102882e-01 -9.60604429e-01 -7.76899636e-01 -6.52411819e-01
-1.51300535e-01 7.05905497e-01 -1.82032451e-01 1.45346403e-01
-3.08482815e-02 5.27060807e-01 -1.47679484e+00 3.78404856e-01
9.27309453e-01 1.69872808e+00 -1.30771875e-01 5.26661992e-01
1.56091452e-01 7.64317155e-01 -6.55422390e-01 4.29263301e-02
3.19488972e-01 -2.26195544e-01 -4.44295228e-01 -5.24089709e-02
-5.92592778e-03 -3.34753156e-01 -5.23034453e-01 9.19852078e-01
8.19966257e-01 3.47318873e-02 7.55064711e-02 -2.03709722e+00
-6.30598903e-01 6.59189999e-01 -2.85951585e-01 1.65767759e-01
-1.95204288e-01 1.98137254e-01 7.21455336e-01 -4.87983465e-01
3.62843692e-01 7.27879643e-01 7.06208169e-01 4.34726924e-01
-1.10045290e+00 -6.86682224e-01 2.19964609e-01 5.71962953e-01
-1.55268657e+00 -5.98156989e-01 6.71458542e-01 4.78107780e-02
1.59518814e+00 2.84115404e-01 4.58822995e-01 9.24360573e-01
7.13764191e-01 6.63778543e-01 7.82545745e-01 -3.41032222e-02
8.58193338e-01 -4.45986450e-01 -1.13063529e-01 6.81623280e-01
4.70591754e-01 1.32320285e-01 -6.99377418e-01 -3.47941905e-01
5.58523417e-01 3.33862364e-01 -1.21879905e-01 -1.17614716e-01
-6.13795221e-01 3.82720679e-01 5.50618649e-01 9.79435891e-02
-7.74215281e-01 7.26942480e-01 8.25078011e-01 2.37162247e-01
4.98434007e-01 7.93550760e-02 -8.54024351e-01 -2.88322240e-01
-1.05181277e+00 -1.09911852e-01 8.97432685e-01 1.08059192e+00
3.95596176e-01 6.64588690e-01 -5.09299278e-01 2.82770544e-01
2.26546288e-01 3.63666922e-01 3.15645933e-01 -9.41601157e-01
-4.50165719e-02 7.50143528e-01 2.60936860e-02 -7.43258595e-01
-5.11809111e-01 -6.86664462e-01 -1.06423855e+00 5.26989140e-02
-4.13267285e-01 -2.35002905e-01 -1.01972270e+00 1.68617427e+00
2.63597786e-01 8.53383124e-01 -1.34313434e-01 9.70299244e-01
4.33436930e-01 6.39257789e-01 -6.07752763e-02 -2.93621689e-01
1.22218359e+00 -1.31011093e+00 -5.16117036e-01 -1.92236863e-02
3.69251281e-01 -3.23354036e-01 8.07870507e-01 5.57949841e-01
-9.82275426e-01 -3.96847397e-01 -1.13950753e+00 3.57934088e-01
-1.90680668e-01 6.19532242e-02 1.14155042e+00 8.69588912e-01
-1.32213700e+00 7.62118340e-01 -1.07602012e+00 -1.11205421e-01
5.63875735e-01 7.51264036e-01 1.53102264e-01 -1.23529926e-01
-6.73792601e-01 2.57927120e-01 1.68899059e-01 -8.37728009e-02
-1.23095548e+00 -7.81069696e-01 -3.69073838e-01 5.80707550e-01
3.22918564e-01 -1.28632629e+00 1.05844402e+00 -7.79358327e-01
-1.74371362e+00 1.57782242e-01 1.08406588e-01 -5.32193780e-01
2.59417802e-01 -1.43030390e-01 -8.53052080e-01 -1.93882406e-01
-1.52360722e-01 3.21897209e-01 6.60655558e-01 -1.01155353e+00
-6.22927248e-01 -3.50792468e-01 -6.33552223e-02 -2.70910501e-01
-3.00232559e-01 -1.73034042e-01 -4.76962805e-01 -2.75901049e-01
-1.79094762e-01 -1.00712097e+00 -3.18465233e-01 -3.62657607e-01
-4.70789403e-01 -3.18149984e-01 9.47998822e-01 -4.57681090e-01
1.20881474e+00 -1.91591024e+00 -1.59915686e-01 6.13688529e-02
4.07377332e-01 1.79429010e-01 -2.29630247e-02 2.83485442e-01
3.47494185e-01 9.50425342e-02 3.44328582e-01 -3.28547150e-01
1.92421511e-01 4.28744078e-01 3.59818265e-02 2.95737267e-01
-1.20138712e-01 9.50306416e-01 -7.45358944e-01 -5.71147054e-02
2.55182143e-02 4.28551644e-01 -6.93037748e-01 1.87828124e-01
-5.68088710e-01 2.46282116e-01 -3.22618961e-01 1.21738303e+00
7.17046678e-01 -7.45879948e-01 4.47422951e-01 -3.01525474e-01
2.80405432e-01 1.46714702e-01 -9.92313981e-01 1.74628735e+00
-8.42585564e-01 3.47587615e-01 1.66679263e-01 -9.74612176e-01
8.11243773e-01 4.82543409e-01 8.48608196e-01 -1.10367823e+00
4.79323417e-01 1.90424189e-01 -2.97567308e-01 -2.61760235e-01
3.46429855e-01 3.20867032e-01 -1.97052166e-01 5.76175153e-01
-5.42000905e-02 5.64318180e-01 3.45169529e-02 -8.82744119e-02
1.59678829e+00 -4.89505082e-02 -2.69958049e-01 -2.61256486e-01
-4.80268486e-02 -2.34978184e-01 1.20850444e+00 7.21525550e-01
-3.47588092e-01 6.27272576e-02 3.44413370e-01 -7.96655476e-01
-1.01123261e+00 -1.06021380e+00 1.72089130e-01 1.02637434e+00
1.86438367e-01 -5.32555699e-01 -8.41850996e-01 -8.17676067e-01
1.92986891e-01 7.77949452e-01 -2.74697602e-01 -5.56504488e-01
-1.75206035e-01 -1.01100540e+00 2.49492288e-01 8.44010592e-01
4.66318965e-01 -8.84933650e-01 -6.68290913e-01 3.82009387e-01
8.70602801e-02 -1.06764567e+00 -4.22327071e-01 3.75644833e-01
-8.51687312e-01 -8.26969743e-01 3.07153333e-02 -4.84886378e-01
6.57599747e-01 1.12204812e-01 1.57012773e+00 3.04837316e-01
-4.84384984e-01 1.38308108e-01 -3.16402853e-01 3.15140225e-02
-1.56418383e-01 1.78078357e-02 4.18844521e-01 -1.03761017e-01
3.62042993e-01 -8.58204782e-01 -8.14937174e-01 3.36356580e-01
-7.13635087e-01 -6.44057943e-03 6.18954360e-01 1.03894103e+00
9.30670619e-01 5.25699973e-01 1.01612914e+00 -7.55727530e-01
4.52979892e-01 -8.95865917e-01 -1.01227117e+00 2.57767975e-01
-1.38173926e+00 1.13464380e-02 1.05469453e+00 -3.67745370e-01
-5.15848160e-01 2.63101086e-02 1.02430172e-01 -8.45698237e-01
1.57148838e-01 2.05799654e-01 -4.50991809e-01 -2.89018244e-01
3.84091407e-01 6.28645793e-02 -4.28067356e-01 -3.43419075e-01
1.46754608e-01 6.98279202e-01 5.30775130e-01 -8.13865483e-01
1.79355577e-01 3.42110395e-01 1.23624042e-01 -1.59868032e-01
-3.21209431e-01 6.16255729e-03 2.03885213e-01 -2.79906273e-01
2.57451653e-01 -1.02566743e+00 -1.31661749e+00 2.99839467e-01
-8.63212764e-01 -3.31060410e-01 -1.65523112e-01 7.55992308e-02
-5.94443858e-01 -8.27435404e-02 -9.14343894e-01 -6.57902658e-01
-1.06292725e+00 -1.26354015e+00 1.20289719e+00 5.08604169e-01
1.19989932e-01 -6.39163852e-01 -3.17788899e-01 1.98777050e-01
9.33119655e-01 4.31152731e-01 9.93826389e-01 -5.09359598e-01
-1.01101947e+00 -3.11058551e-01 -3.45032930e-01 3.04235369e-01
4.11882363e-02 -2.08432525e-02 -7.71550477e-01 -6.53098941e-01
2.62072403e-03 2.31599491e-02 4.15697634e-01 5.95462143e-01
1.46821320e+00 -4.15363014e-01 -5.08157432e-01 1.00646961e+00
1.79964721e+00 3.01759511e-01 5.49540579e-01 1.70470998e-01
7.64044940e-01 -3.73491198e-01 1.90087408e-01 8.85749459e-01
3.50235581e-01 6.25553966e-01 8.42597187e-01 -5.64565845e-02
1.69796534e-02 -1.07238524e-01 4.10919428e-01 7.88558602e-01
2.96389908e-01 -6.26729846e-01 -6.85058713e-01 4.38147128e-01
-2.15022326e+00 -5.02066255e-01 5.20617105e-02 2.12183380e+00
3.16943884e-01 2.84976900e-01 6.79126605e-02 1.02510437e-01
5.57576001e-01 -3.99877727e-01 -1.28012168e+00 -4.67511177e-01
1.18185654e-01 2.24818677e-01 7.34752655e-01 -1.40516266e-01
-6.29709899e-01 8.05929542e-01 6.27697325e+00 6.12343371e-01
-1.44258654e+00 4.54572886e-01 9.93452132e-01 -4.27687824e-01
-2.56353378e-01 -3.76983471e-02 -5.40282488e-01 9.97493863e-01
1.64756513e+00 -4.66664940e-01 9.22350109e-01 1.44077945e+00
3.12331140e-01 1.27269551e-01 -1.25331426e+00 1.17569935e+00
-1.52400732e-01 -1.75016391e+00 -3.91037673e-01 4.88162115e-02
1.09242165e+00 4.53547537e-01 -1.50843948e-01 4.03276205e-01
6.16157293e-01 -9.33317602e-01 5.94758511e-01 6.01503730e-01
7.40858316e-01 -1.23551679e+00 9.17245746e-01 -3.26252659e-03
-1.03474617e+00 -3.17553669e-01 -3.29723924e-01 -3.73377539e-02
2.41371453e-01 1.21581304e+00 -6.20044112e-01 7.41133988e-01
1.05713689e+00 7.65527189e-02 -1.54310822e-01 1.14759886e+00
1.76356032e-01 6.02216005e-01 -5.01897037e-01 1.19562909e-01
-2.98260927e-01 1.35490671e-01 2.52547115e-01 7.01925516e-01
5.88088036e-01 -3.06838572e-01 6.24570549e-01 9.92599308e-01
-6.26123726e-01 -3.20708036e-01 -1.04798652e-01 -1.99831292e-01
9.40089703e-01 1.57590616e+00 -9.08220947e-01 -4.38501716e-01
-2.48052135e-01 1.18314743e+00 2.12273677e-03 2.70775557e-01
-1.20111060e+00 -1.03280075e-01 1.10476983e+00 -6.31255284e-02
2.67663091e-01 -2.32167989e-01 -5.28211355e-01 -9.39634979e-01
3.16798240e-02 -8.58968675e-01 2.65772939e-01 -5.98700404e-01
-1.39389074e+00 8.74219120e-01 -7.55864263e-01 -8.57351661e-01
-2.11707383e-01 -4.11325186e-01 -6.91352487e-01 6.77239895e-01
-1.35188258e+00 -1.07086837e+00 -6.59662485e-01 3.48185956e-01
3.84119987e-01 -1.59595296e-01 8.51943135e-01 8.01929355e-01
-1.11279511e+00 7.66674340e-01 2.68405974e-01 -4.05275255e-01
2.41684645e-01 -1.01552856e+00 8.31169784e-01 9.19908285e-01
8.86377320e-02 8.25569108e-02 5.31288803e-01 -6.83191895e-01
-2.11468673e+00 -1.34771240e+00 3.15971762e-01 4.03937735e-02
5.21725297e-01 -3.55668455e-01 -5.81396937e-01 4.87993330e-01
1.61446333e-01 5.15046895e-01 7.04886794e-01 -4.87851128e-02
-7.39404485e-02 -4.47227001e-01 -1.36299610e+00 3.56323987e-01
1.10101056e+00 -3.57497126e-01 5.85142553e-01 8.58685136e-01
9.43911314e-01 -4.22288924e-01 -9.56440747e-01 5.16817451e-01
1.62627414e-01 -9.25422549e-01 4.99917597e-01 -7.16273606e-01
-1.26660526e-01 -4.97359186e-01 -3.80011529e-01 -1.33879232e+00
-5.33307374e-01 -8.61561477e-01 -8.54187131e-01 1.19612205e+00
1.20974161e-01 -4.98712361e-01 1.07333541e+00 5.46243310e-01
-6.19412005e-01 -1.03497446e+00 -9.49004471e-01 -1.16097808e+00
-5.23115754e-01 -3.70292634e-01 1.10672367e+00 7.45414138e-01
-4.60426599e-01 1.08046323e-01 -3.19595844e-01 4.55179542e-01
4.98743862e-01 3.27672452e-01 5.11582077e-01 -1.22109747e+00
-6.38828754e-01 -2.71735609e-01 -4.09663707e-01 -5.43388605e-01
1.86744705e-01 -7.85666883e-01 -3.29070054e-02 -1.51150990e+00
-6.52159452e-02 -7.07631707e-01 -8.09050739e-01 7.11087048e-01
1.29367232e-01 1.32468551e-01 -1.33782223e-01 5.48706874e-02
-9.97841120e-01 5.61379135e-01 5.08197069e-01 3.46576907e-02
-1.14485562e-01 -6.62926491e-03 -7.71211684e-01 3.15055817e-01
7.87640095e-01 -5.55599689e-01 -3.46117586e-01 -6.43024981e-01
2.82814234e-01 1.80784643e-01 1.75452009e-01 -1.40406001e+00
5.17159641e-01 -1.55876622e-01 3.21301490e-01 -2.43903309e-01
-1.03528559e-01 -9.49695528e-01 7.49071896e-01 5.58282912e-01
5.34004986e-01 6.26248777e-01 2.34890401e-01 5.45075357e-01
1.87366784e-01 3.98862809e-01 5.76184511e-01 2.31783897e-01
-7.79015839e-01 7.56475866e-01 -1.17561132e-01 -2.35141858e-01
1.20761943e+00 1.86472699e-01 -5.42513490e-01 -1.41421974e-01
-4.95654434e-01 3.91360164e-01 6.82930291e-01 4.58688766e-01
3.43040496e-01 -1.43663299e+00 -4.54648465e-01 3.48770857e-01
-1.32196203e-01 -9.33471173e-02 6.16226017e-01 8.45169127e-01
-6.14176571e-01 2.89108455e-01 -3.75917464e-01 -6.01443291e-01
-6.54179573e-01 7.62654722e-01 5.23265719e-01 -3.31273139e-01
-5.58642745e-01 7.38255620e-01 -3.69298935e-01 -2.19532736e-02
1.89055204e-01 -7.84588698e-03 6.74809754e-01 -5.81578612e-01
3.67940217e-01 7.15589046e-01 6.98338211e-01 -4.93707135e-02
-5.40813267e-01 -1.96610659e-01 1.11457281e-01 4.99855876e-01
1.50989473e+00 7.53372237e-02 -2.39250302e-01 -1.89327653e-02
8.70789468e-01 -3.90695214e-01 -1.34783483e+00 7.11506680e-02
4.85524870e-02 -2.62735069e-01 7.40097106e-01 -1.05611634e+00
-1.78494465e+00 2.36038014e-01 1.00590396e+00 3.57204497e-01
1.74183869e+00 -3.13250393e-01 1.37277162e+00 -5.42973615e-02
7.37666249e-01 -1.21797264e+00 -1.27892151e-01 8.81320462e-02
2.25325599e-01 -9.63247299e-01 -3.42219412e-01 -1.46559760e-01
-1.13444924e-01 7.70553291e-01 9.47112024e-01 -1.76563352e-01
5.52489400e-01 8.10859740e-01 -2.92254478e-01 -2.31440589e-01
-1.28291011e+00 2.47622132e-01 -1.85819238e-01 4.61373419e-01
-8.59105121e-03 5.14948428e-01 -2.76012216e-02 9.44193542e-01
2.65785098e-01 4.03782606e-01 2.00220063e-01 9.78046536e-01
-4.29846346e-02 -1.20237279e+00 -1.71706937e-02 7.31306136e-01
-3.35697740e-01 -1.11055762e-01 5.02593108e-02 1.45442620e-01
1.44944474e-01 1.01046848e+00 3.06714356e-01 -8.05007577e-01
-1.81653220e-02 3.06456052e-02 2.26789400e-01 -2.92601734e-01
-5.83825707e-01 -1.57216817e-01 2.03719940e-02 -9.95797873e-01
3.71821940e-01 -1.30733430e-01 -1.33310854e+00 -8.22394729e-01
-3.26196253e-01 4.48988676e-02 1.22570622e+00 9.20861900e-01
1.39064646e+00 1.23798382e+00 9.57105339e-01 -7.92463183e-01
-3.67749810e-01 -4.64238048e-01 -6.78517401e-01 4.83285673e-02
-1.90733954e-01 -7.45317519e-01 -1.96092799e-01 -3.46283287e-01] | [7.020698547363281, 2.8956105709075928] |
e4e768ac-6435-4772-983f-ff725086e4d9 | deep-learning-for-3d-point-clouds-a-survey | 1912.12033 | null | https://arxiv.org/abs/1912.12033v2 | https://arxiv.org/pdf/1912.12033v2.pdf | Deep Learning for 3D Point Clouds: A Survey | Point cloud learning has lately attracted increasing attention due to its wide applications in many areas, such as computer vision, autonomous driving, and robotics. As a dominating technique in AI, deep learning has been successfully used to solve various 2D vision problems. However, deep learning on point clouds is still in its infancy due to the unique challenges faced by the processing of point clouds with deep neural networks. Recently, deep learning on point clouds has become even thriving, with numerous methods being proposed to address different problems in this area. To stimulate future research, this paper presents a comprehensive review of recent progress in deep learning methods for point clouds. It covers three major tasks, including 3D shape classification, 3D object detection and tracking, and 3D point cloud segmentation. It also presents comparative results on several publicly available datasets, together with insightful observations and inspiring future research directions. | ['Hao liu', 'Mohammed Bennamoun', 'Hanyun Wang', 'Li Liu', 'Yulan Guo', 'Qingyong Hu'] | 2019-12-27 | null | null | null | null | ['3d-shape-retrieval'] | ['computer-vision'] | [-2.14557782e-01 -4.85969901e-01 -1.06425159e-01 -4.05990094e-01
-4.85140681e-01 -5.71472347e-01 6.31551087e-01 1.40944064e-01
-2.17724741e-01 -4.68791090e-03 -6.76016927e-01 -4.42177236e-01
1.19473405e-01 -6.65023208e-01 -7.04802752e-01 -5.56455016e-01
-2.07330808e-01 7.39467323e-01 2.14522108e-01 -5.58085814e-02
4.39244360e-01 1.19699693e+00 -1.68468070e+00 -1.21603079e-01
5.63551664e-01 1.10219300e+00 3.26414943e-01 2.92784303e-01
-5.56158543e-01 -3.43995132e-02 -3.44010562e-01 -2.90438890e-01
5.02703309e-01 3.40284109e-01 -3.76987576e-01 1.69457719e-01
6.49152696e-01 -3.03980678e-01 -2.34768346e-01 1.03550458e+00
4.29791063e-01 2.34268364e-02 5.56996822e-01 -1.43467093e+00
-5.81079781e-01 -1.65913999e-01 -9.26287174e-01 1.78843528e-01
-8.21953788e-02 2.12160453e-01 9.42043006e-01 -1.14173877e+00
1.88983425e-01 1.32073355e+00 8.47445190e-01 4.67695594e-01
-7.82391131e-01 -9.33364332e-01 1.72125638e-01 2.79135108e-01
-1.02223909e+00 1.44111952e-02 1.06341827e+00 -6.39769435e-01
1.21831799e+00 -9.50783640e-02 9.56102848e-01 6.43663228e-01
-2.34458465e-02 1.26204336e+00 7.79422045e-01 -2.29125381e-01
2.71732509e-01 -2.79258102e-01 7.20151067e-02 3.95238042e-01
4.14899468e-01 1.99353829e-01 2.97843255e-02 -1.61589503e-01
1.08499622e+00 3.56563061e-01 2.26794645e-01 -8.42441261e-01
-1.00164938e+00 1.01863015e+00 7.33945847e-01 2.95102149e-01
-4.44089234e-01 3.23581547e-01 3.65545303e-01 3.97893861e-02
5.75834274e-01 2.32532755e-01 -6.04802310e-01 -7.99187496e-02
-7.24254906e-01 6.44705713e-01 3.07966083e-01 1.11848056e+00
6.85442865e-01 2.92609274e-01 6.97012246e-01 9.00274217e-01
5.63100517e-01 8.34353924e-01 -1.09351337e-01 -1.05442798e+00
2.36441627e-01 8.13763142e-01 9.41345692e-02 -1.00040483e+00
-3.70183766e-01 -3.73673737e-01 -8.15206468e-01 7.96864569e-01
1.05123565e-01 1.12493508e-01 -1.08764648e+00 9.61106539e-01
3.89587879e-01 3.64005834e-01 -2.98339963e-01 1.06509507e+00
1.21024394e+00 7.44201124e-01 -1.50298178e-01 4.01452661e-01
1.11066937e+00 -4.53572869e-01 -9.23455805e-02 -4.76001859e-01
2.30033979e-01 -6.71565175e-01 6.30518913e-01 3.54833186e-01
-8.96313429e-01 -6.42691433e-01 -8.67806911e-01 -2.16370314e-01
-2.52433956e-01 -1.77560821e-01 8.93216252e-01 3.79795730e-01
-8.66435885e-01 4.12233412e-01 -1.15693676e+00 -5.83374143e-01
1.08868253e+00 6.78469777e-01 -5.32620735e-02 -7.59099796e-02
-6.03433013e-01 7.92460263e-01 1.18948624e-01 8.88377726e-02
-5.45936227e-01 -6.29475772e-01 -7.69410789e-01 -8.72448087e-03
6.73421100e-02 -8.87651920e-01 1.42739034e+00 -4.07897443e-01
-1.29326582e+00 1.32911134e+00 -6.64467365e-02 -5.34861505e-01
2.73929328e-01 -3.18293929e-01 7.36484304e-03 -1.65759757e-01
1.26006544e-01 9.73355591e-01 8.44816923e-01 -1.37729394e+00
-9.39838886e-01 -8.28386545e-01 -1.21997140e-01 2.35247403e-01
1.07618794e-01 9.92191806e-02 -5.65991104e-01 -1.29183561e-01
5.05001426e-01 -1.10389197e+00 -3.95009577e-01 2.94275552e-01
-1.75951302e-01 -7.91561425e-01 1.09815538e+00 1.67780221e-01
2.09412307e-01 -2.21633315e+00 2.91517824e-02 6.69416389e-04
4.09434795e-01 5.37861705e-01 -9.19216275e-02 1.42123103e-01
-1.63872931e-02 -4.43888828e-02 -3.63583952e-01 -5.17086089e-01
-1.03707686e-01 2.35338271e-01 -5.83597958e-01 6.49807870e-01
3.36804956e-01 1.12264574e+00 -8.43530357e-01 -1.81136996e-01
8.20550323e-01 5.77795923e-01 -2.18748420e-01 -1.47428721e-01
-4.01896715e-01 4.13461149e-01 -6.00988865e-01 8.57575476e-01
1.04286361e+00 -3.86024386e-01 -5.30360818e-01 2.21840158e-01
-4.75547820e-01 1.48379698e-03 -8.40076566e-01 1.59896255e+00
-1.85909271e-01 9.92049634e-01 1.48582697e-01 -1.08045304e+00
1.25294828e+00 2.71492694e-02 8.64002764e-01 -4.26745474e-01
3.56109977e-01 4.04066712e-01 -8.22962672e-02 -2.83961564e-01
4.25831228e-01 -3.64411980e-01 1.42330110e-01 -1.41799590e-03
-3.18208277e-01 -9.51501369e-01 -3.43634069e-01 -2.82979399e-01
7.26854622e-01 5.19855805e-02 4.38869819e-02 2.29436636e-01
3.85574818e-01 5.19815207e-01 2.52563953e-01 4.50804681e-01
-4.07171965e-01 6.05994940e-01 -5.78731112e-02 -8.39898288e-01
-1.09281874e+00 -8.57330501e-01 -3.68943304e-01 4.14108872e-01
4.09605771e-01 9.80705395e-02 -2.89102137e-01 -2.99268901e-01
6.66885912e-01 4.71578777e-01 -3.07010680e-01 1.57952636e-01
-7.46515512e-01 -3.64349782e-01 2.85717845e-01 6.71302736e-01
4.93736953e-01 -1.17194736e+00 -6.79549217e-01 -1.44629797e-03
3.34882885e-01 -1.19310188e+00 3.56052190e-01 1.54036403e-01
-1.51492000e+00 -9.71020103e-01 -8.76015127e-01 -1.02562094e+00
3.92298400e-01 9.75760639e-01 1.18923175e+00 1.13992080e-01
-1.45106882e-01 4.40049410e-01 -1.06813259e-01 -1.15922236e+00
-2.90190503e-02 1.48364544e-01 9.61252078e-02 -5.26595354e-01
1.18957353e+00 -4.18943673e-01 -3.36544901e-01 1.19646609e-01
-6.17059827e-01 -3.57324868e-01 5.48051059e-01 4.37756985e-01
8.50522339e-01 -5.42438813e-02 2.49177113e-01 -4.73604321e-01
4.41424966e-01 -3.85475338e-01 -1.09461808e+00 -4.34808463e-01
-2.31644109e-01 -4.99050349e-01 2.45528653e-01 1.03811331e-01
-5.50297320e-01 1.43883973e-01 -5.02288043e-01 -1.05302119e+00
-6.87883914e-01 2.71852553e-01 1.54827118e-01 -4.32428598e-01
4.04693604e-01 1.25671327e-01 1.16374388e-01 -5.93097210e-01
3.00550282e-01 5.69849789e-01 3.33488435e-01 -3.08019787e-01
1.08482265e+00 8.82688284e-01 1.38162434e-01 -1.26260579e+00
-7.08853722e-01 -7.80875385e-01 -8.83798301e-01 -1.62535504e-01
9.90014851e-01 -8.95988524e-01 -8.98593545e-01 6.60903990e-01
-1.61582506e+00 -7.50805214e-02 -1.38023734e-01 3.42995584e-01
-6.00773871e-01 4.32765245e-01 -3.70908499e-01 -8.01819801e-01
-3.26371074e-01 -1.25108027e+00 1.45725179e+00 3.41521502e-01
1.23500563e-01 -1.09655571e+00 1.84883639e-01 2.08108917e-01
1.94067899e-02 3.52670640e-01 8.50890934e-01 -4.63687181e-01
-9.99681890e-01 -5.23193419e-01 -4.54031110e-01 3.34183455e-01
-9.81000811e-03 1.03744216e-01 -1.00111675e+00 -2.71825910e-01
6.95447400e-02 -1.85977980e-01 8.02409828e-01 8.48323882e-01
1.34439075e+00 5.96289992e-01 -6.08332157e-01 7.78695166e-01
1.35715425e+00 2.87152916e-01 2.98840374e-01 5.60952544e-01
7.86614418e-01 3.46948743e-01 5.21362603e-01 1.67187333e-01
2.98478037e-01 4.97725487e-01 1.03907549e+00 -2.64399946e-01
2.01793626e-01 6.66713789e-02 -2.25036263e-01 6.27063453e-01
-2.87675023e-01 6.52789623e-02 -1.38671005e+00 4.83288974e-01
-1.55780649e+00 -9.16859984e-01 -5.75685561e-01 1.93693542e+00
2.37217005e-02 3.14155221e-01 1.08547457e-01 3.43621075e-02
6.18266344e-01 1.65368035e-01 -1.01623678e+00 -1.85298190e-01
-1.66159794e-02 2.91103333e-01 3.81252676e-01 1.68191176e-02
-1.36118400e+00 1.15459704e+00 6.37345171e+00 4.36490178e-01
-1.41703808e+00 -2.11467609e-01 2.14000493e-01 1.84800088e-01
-9.43160132e-02 -2.59945452e-01 -9.17073905e-01 3.41926694e-01
1.68315738e-01 6.26858026e-02 -2.82809753e-02 1.19834673e+00
1.32070065e-01 1.04175977e-01 -1.03673697e+00 1.45807242e+00
-1.46466434e-01 -1.54736650e+00 -2.01710258e-02 2.66198695e-01
6.58081651e-01 9.45847571e-01 4.52642947e-01 2.99122959e-01
1.88966841e-01 -8.92573595e-01 4.32526797e-01 9.35165286e-02
7.01203823e-01 -8.33774328e-01 7.43697286e-01 6.86518550e-01
-1.12571299e+00 -6.57648733e-03 -8.46216381e-01 -3.93669218e-01
1.41895637e-01 6.44111514e-01 -8.61770093e-01 2.74352014e-01
9.85324144e-01 1.20398057e+00 -3.87447961e-02 1.58712387e+00
-1.48342907e-01 3.13778967e-01 -5.59963703e-01 -8.97397920e-02
6.78959191e-01 -4.43707496e-01 8.03497314e-01 8.42992187e-01
2.33762890e-01 1.48688838e-01 2.75443941e-01 1.08109426e+00
-9.89493281e-02 -2.41758585e-01 -1.10192168e+00 2.46957187e-02
5.29218376e-01 1.19628668e+00 -8.07286561e-01 -2.04350054e-01
-7.20878005e-01 3.26702386e-01 9.09988284e-02 2.52845466e-01
-5.46108902e-01 -3.23522627e-01 1.25468242e+00 -1.66135132e-02
3.91418546e-01 -9.69981611e-01 -7.99311697e-01 -8.83313835e-01
-1.33324847e-01 -2.78143853e-01 -4.50660400e-02 -8.46505046e-01
-1.36541939e+00 3.43597680e-01 8.18978157e-03 -1.58166516e+00
1.03109755e-01 -9.90880191e-01 -6.73604906e-01 8.91307354e-01
-1.88092625e+00 -8.93903077e-01 -4.37927663e-01 2.27532163e-01
8.34203660e-01 -2.78120965e-01 4.19107407e-01 3.17129761e-01
-2.52403140e-01 -4.97804433e-02 2.09793001e-01 1.52376175e-01
2.67112911e-01 -1.07311022e+00 1.26354873e+00 4.81957167e-01
2.73122668e-01 5.21542668e-01 4.08711374e-01 -6.15392268e-01
-1.62731540e+00 -1.08207643e+00 5.10152102e-01 -6.15123391e-01
4.54536080e-01 -3.41488063e-01 -1.18385184e+00 6.41677022e-01
-1.36092737e-01 1.62127502e-02 3.80285889e-01 6.04941882e-02
-2.10002041e-03 -9.45770368e-03 -1.02982259e+00 2.83992469e-01
9.24433351e-01 -2.51020253e-01 -5.34202814e-01 3.87938768e-01
6.02968454e-01 -7.81496882e-01 -4.66253400e-01 5.46352506e-01
1.56780526e-01 -1.00310445e+00 1.18882620e+00 -5.44771791e-01
2.25429282e-01 -2.79218227e-01 4.65431847e-02 -1.15541899e+00
-6.17115855e-01 -1.41061127e-01 -8.31877813e-02 6.20517969e-01
-1.40817121e-01 -4.57052112e-01 1.44176900e+00 3.70085478e-01
-6.16162360e-01 -8.33682299e-01 -9.68795776e-01 -7.60164559e-01
3.39555174e-01 -8.34853947e-01 7.12002516e-01 8.00246477e-01
-7.77783275e-01 3.40242118e-01 1.57020360e-01 3.02043408e-01
8.95088911e-01 6.09487236e-01 1.09850216e+00 -1.87917817e+00
4.49632585e-01 -8.02096248e-01 -7.20654190e-01 -1.52760661e+00
3.57173622e-01 -1.02236855e+00 3.01543623e-02 -1.88472998e+00
-1.94292024e-01 -8.19475830e-01 1.11399896e-01 3.44302148e-01
1.42953664e-01 4.39503074e-01 3.78047049e-01 5.58653951e-01
-3.42013687e-01 6.59460068e-01 1.23191643e+00 -3.64006013e-01
-1.83081254e-01 5.45894384e-01 -3.83894533e-01 1.09744179e+00
9.63895559e-01 -2.80460387e-01 3.49587649e-02 -1.05106390e+00
2.14577422e-01 -1.91335395e-01 4.62313026e-01 -1.14146876e+00
1.78233668e-01 -9.76824164e-02 5.71675420e-01 -1.44251657e+00
6.26656115e-01 -1.10733700e+00 -2.65309095e-01 4.14295226e-01
3.66649419e-01 1.19319916e-01 5.40182471e-01 5.62142551e-01
-3.44193667e-01 -9.48494077e-02 9.42093849e-01 -3.02237958e-01
-1.20650280e+00 8.19918692e-01 -2.37878636e-02 -3.46949428e-01
1.14959872e+00 -5.73047280e-01 6.69716224e-02 -7.20592886e-02
-4.43640023e-01 5.06772101e-01 6.80494368e-01 6.51161611e-01
9.34509814e-01 -1.28497291e+00 -6.76980734e-01 3.10389012e-01
1.08905703e-01 8.26857150e-01 6.99834675e-02 3.65841389e-01
-6.78307116e-01 6.77399814e-01 -2.04507098e-01 -1.52391601e+00
-1.10360610e+00 5.88240147e-01 2.83411324e-01 4.39401180e-01
-9.60607052e-01 1.00044131e+00 2.36813143e-01 -5.79964042e-01
2.70282686e-01 -5.03966689e-01 -1.68254957e-01 -3.81456912e-01
8.05780888e-02 3.76813829e-01 3.45286936e-01 -5.24638474e-01
-5.32180011e-01 1.16309834e+00 -6.94739744e-02 4.32686895e-01
1.53432679e+00 2.12289527e-01 -3.25000435e-02 5.21241665e-01
1.08994520e+00 -4.72057641e-01 -1.26070774e+00 -2.64832050e-01
1.21473975e-01 -4.91977215e-01 1.34575173e-01 -2.71516740e-01
-1.12042975e+00 1.52588880e+00 6.14053428e-01 3.54124725e-01
6.96632564e-01 3.35953325e-01 1.05151784e+00 4.91688162e-01
6.80330396e-01 -5.83968163e-01 -1.20940112e-01 1.14409709e+00
7.26679265e-01 -1.72357416e+00 1.20255880e-01 -2.40730047e-01
-1.99734166e-01 1.18729997e+00 5.56079805e-01 -5.81893265e-01
7.99533784e-01 1.37537017e-01 1.45372033e-01 -5.64935505e-01
-3.33694845e-01 -2.70331264e-01 2.30931953e-01 7.43619442e-01
3.60680908e-01 -2.61835381e-02 3.59416783e-01 -7.95298815e-02
-1.91590741e-01 -7.84522481e-03 5.78334890e-02 8.45210195e-01
-7.52834380e-01 -9.80552435e-01 -6.38057649e-01 5.51993787e-01
-1.35613242e-02 2.59528726e-01 -4.66618240e-01 8.12752366e-01
5.09444736e-02 4.94997352e-01 6.31246388e-01 -2.09369555e-01
5.49705327e-01 -2.39127100e-01 5.10076702e-01 -7.68505991e-01
-2.51880229e-01 -3.31668034e-02 -6.14158273e-01 -3.02139401e-01
-4.58369166e-01 -9.53288972e-01 -1.47927165e+00 -3.67282212e-01
-2.91572362e-01 -1.87306516e-02 1.22224939e+00 7.29950368e-01
4.85357404e-01 2.28776917e-01 4.99441415e-01 -1.57454181e+00
-3.89611751e-01 -5.32831252e-01 -4.11324710e-01 4.66651097e-02
5.19527555e-01 -1.02327216e+00 -2.32458979e-01 -3.40022683e-01] | [8.001296997070312, -3.3458476066589355] |
0d78b945-ee7f-4eb8-8682-529b34b48330 | spin-qudit-tomography | 2012.06464 | null | https://arxiv.org/abs/2012.06464v2 | https://arxiv.org/pdf/2012.06464v2.pdf | Spin qudit tomography and state reconstruction error | We consider the task of performing quantum state tomography on a $d$-level spin qudit, using only measurements of spin projection onto different quantization axes. After introducing a basis of operators closely related to the spherical harmonics, which obey the rotational symmetries of spin qudits, we map our quantum tomography task onto the classical problem of signal recovery on the sphere. We then provide algorithms with $O(rd^3)$ serial runtime, parallelizable down to $O(rd^2)$, for (i) computing a priori upper bounds on the expected error with which spin projection measurements along $r$ given axes can reconstruct an unknown qudit state, and (ii) estimating a posteriori the statistical error in a reconstructed state. Our algorithms motivate a simple randomized tomography protocol, for which we find that using more measurement axes can yield substantial benefits that plateau after $r\approx3d$. | ['Ana Maria Rey', 'Diego Barberena', 'Michael A. Perlin'] | 2020-12-11 | null | null | null | null | ['quantum-state-tomography'] | ['medical'] | [ 3.15827608e-01 3.00080210e-01 1.10599868e-01 -3.13775897e-01
-1.03505969e+00 -5.14324844e-01 4.54976916e-01 -4.89503145e-01
-6.08511567e-01 8.14289749e-01 -1.01737402e-01 -4.54473615e-01
-2.14993164e-01 -9.49647665e-01 -5.35011470e-01 -8.32605481e-01
-2.16216385e-01 8.67226362e-01 5.28150909e-02 -5.01308031e-02
6.92862570e-01 5.52361012e-01 -1.16441274e+00 -3.73438239e-01
4.38058466e-01 1.04448938e+00 -4.69336182e-01 7.04159319e-01
3.03996205e-01 6.79647148e-01 -2.90396184e-01 -2.95311064e-01
3.78936857e-01 -9.51080799e-01 -1.34326017e+00 -7.64660388e-02
1.96155652e-01 -2.57461458e-01 -8.18112910e-01 1.62212074e+00
3.65256071e-01 1.39400929e-01 5.97456217e-01 -5.56810141e-01
-4.29292589e-01 5.90362370e-01 -2.40927786e-01 2.01024070e-01
5.39241731e-01 1.85430288e-01 1.09491992e+00 -5.85378349e-01
1.00329018e+00 9.26312089e-01 2.71222949e-01 4.96713430e-01
-1.72066951e+00 -5.80227375e-01 -9.07434165e-01 3.60561818e-01
-1.87369204e+00 -1.04660439e+00 6.58426344e-01 -1.45226508e-01
1.04488862e+00 3.25323552e-01 5.39351404e-01 4.53281373e-01
1.43589288e-01 1.83859412e-02 1.70570230e+00 -5.90355098e-01
4.38350528e-01 1.12060525e-01 2.00192586e-01 8.55678916e-01
3.51267397e-01 3.94737601e-01 -7.61894107e-01 -1.28903687e-01
5.86318970e-01 -5.52018106e-01 -2.71322459e-01 -4.67026532e-01
-1.45164466e+00 6.57651663e-01 1.25163227e-01 1.91571534e-01
-1.45830065e-01 3.53351682e-01 1.30336806e-01 5.56114674e-01
-8.51215273e-02 6.48723125e-01 -1.83951706e-01 -3.84859979e-01
-9.06589985e-01 -7.38842711e-02 8.08178902e-01 9.41257954e-01
1.00660753e+00 -1.19438231e-01 3.04020852e-01 -1.61036532e-02
1.65552020e-01 1.39900744e+00 -1.85252056e-01 -1.40420616e+00
1.26641974e-01 -1.32560045e-01 5.21613181e-01 -4.91769046e-01
-3.45392764e-01 1.29899934e-01 -7.90769994e-01 -2.85127182e-02
4.31639642e-01 -2.20311716e-01 -4.61778879e-01 1.71340656e+00
2.54897147e-01 -9.72906500e-02 1.92635819e-01 7.70784140e-01
1.48590375e-02 4.69353259e-01 -9.42624748e-01 -6.62236989e-01
1.35488474e+00 -2.52051532e-01 -7.43150234e-01 6.88573718e-02
7.88978219e-01 -8.02262604e-01 6.43105149e-01 5.49286485e-01
-1.45837700e+00 -2.74634250e-02 -1.19835281e+00 -2.14010589e-02
1.95618212e-01 -2.99741268e-01 8.55102539e-01 1.00234282e+00
-1.22447014e+00 1.00183952e+00 -1.23305845e+00 2.90283971e-02
-7.65545517e-02 5.79629958e-01 -4.35965449e-01 -1.32291749e-01
-9.42709744e-01 1.17701566e+00 6.84480295e-02 8.12928379e-02
-8.94101024e-01 -9.37422067e-02 -5.44144392e-01 -2.44637892e-01
3.27148348e-01 -3.47501218e-01 1.00256300e+00 8.35636854e-02
-1.90469778e+00 9.23850417e-01 -5.12560189e-01 -3.28059137e-01
-1.10434473e-01 4.25947189e-01 -5.49390793e-01 5.19509792e-01
1.90606773e-01 -1.19170688e-01 4.45084125e-01 -5.87809205e-01
1.08713292e-01 -7.58716822e-01 -5.81446663e-02 -1.17597682e-02
3.95111054e-01 1.06770933e-01 -2.93570962e-02 5.51644921e-01
1.19650149e+00 -1.31837273e+00 -1.65047899e-01 -4.75871742e-01
-7.46037662e-01 2.53728658e-01 2.48384655e-01 -2.85999745e-01
8.18030775e-01 -1.92636418e+00 4.05376047e-01 5.85024595e-01
1.42431170e-01 -1.99361131e-01 2.27106482e-01 4.96234208e-01
5.87634221e-02 -7.01850653e-02 -2.56751359e-01 -1.65319964e-01
2.58243799e-01 -6.40816614e-03 -2.48647913e-01 1.10610974e+00
-2.87845314e-01 6.21293187e-01 -6.44074619e-01 -1.89988360e-01
-3.40979174e-03 4.83356565e-02 -7.48757184e-01 -1.85639322e-01
1.70220807e-01 6.72059119e-01 -3.64587247e-01 4.00469601e-01
8.80334973e-01 -6.03840530e-01 6.15959525e-01 -3.37759882e-01
-3.11309397e-01 9.13082242e-01 -1.41950095e+00 1.77140498e+00
-1.89865708e-01 4.32088912e-01 4.67051744e-01 -9.66565430e-01
5.64432502e-01 3.42307329e-01 4.65686113e-01 -8.20962608e-01
3.30577493e-01 6.24073625e-01 1.79929495e-01 -2.52541214e-01
5.41447461e-01 -9.32155788e-01 -4.29176509e-01 1.08450651e+00
1.98431522e-01 -9.43630993e-01 -4.78115976e-02 2.54623771e-01
1.44581425e+00 -2.58541793e-01 1.30829751e-01 -6.19808495e-01
2.60739535e-01 -2.08778501e-01 2.62093246e-01 7.89520741e-01
-4.01159078e-01 9.10142511e-02 6.71460629e-01 -4.24991429e-01
-1.18701649e+00 -1.10030007e+00 -7.20662236e-01 1.83549002e-01
4.74664837e-01 -4.60929215e-01 -7.69550741e-01 -1.91229910e-01
-4.91676033e-01 7.79886127e-01 -5.25266588e-01 -1.81573823e-01
-3.83241057e-01 -1.05029953e+00 5.22644103e-01 1.77688655e-02
6.87645316e-01 -8.10616553e-01 -5.20035207e-01 -1.10341139e-01
-1.76698104e-01 -1.01985586e+00 -2.29369417e-01 6.57831252e-01
-8.57075155e-01 -1.10134053e+00 -7.78087750e-02 -8.84524956e-02
7.29546309e-01 2.29172662e-01 8.85563850e-01 -4.65014964e-01
-1.55847296e-01 2.07630381e-01 8.98680091e-02 3.36875468e-01
-2.62344778e-01 -3.35345477e-01 5.53153813e-01 -3.33228469e-01
5.62505305e-01 -7.93542922e-01 -5.17845452e-01 2.20760182e-01
-5.53069890e-01 -3.35638314e-01 2.52077758e-01 7.24510074e-01
6.94222093e-01 8.95840898e-02 -3.55430245e-01 -6.57922864e-01
1.53920323e-01 -1.20849982e-02 -1.01217377e+00 -5.26176877e-02
-6.37446165e-01 7.54978418e-01 5.78477263e-01 4.23481017e-01
-6.42513454e-01 -1.97545007e-01 -4.40078182e-03 1.51174674e-02
1.81719884e-01 3.59942287e-01 1.58563659e-01 -6.04167461e-01
9.08338070e-01 5.22714913e-01 -1.86781697e-02 -2.21085951e-01
4.04887021e-01 5.55179596e-01 3.51166606e-01 -7.98059225e-01
8.41871083e-01 7.91725278e-01 7.64718890e-01 -8.19342196e-01
-1.03572309e+00 -3.25846195e-01 -8.65868807e-01 4.18851942e-01
9.14320171e-01 -6.24357581e-01 -1.24799168e+00 1.77026123e-01
-9.79999661e-01 -1.16913438e-01 -5.75564921e-01 8.87121260e-01
-9.23795462e-01 6.43646955e-01 -5.36516309e-01 -9.90603626e-01
-8.65114555e-02 -1.34129333e+00 1.02896106e+00 -1.84820756e-01
1.47072986e-01 -5.57297349e-01 4.13794935e-01 5.18697321e-01
6.03028417e-01 -2.61236161e-01 7.47997701e-01 -1.28949374e-01
-1.06540477e+00 -4.22785819e-01 -2.83971369e-01 3.15161943e-01
-3.17222267e-01 -5.43552637e-01 -7.98031390e-01 -4.77412134e-01
4.28882778e-01 -5.83651245e-01 5.74043453e-01 1.64357021e-01
7.52715886e-01 -1.70922980e-01 -1.60145864e-01 7.42204309e-01
1.38801730e+00 1.69972524e-01 8.58327091e-01 -1.40009984e-01
2.68910766e-01 -9.38948840e-02 2.11283252e-01 3.86590242e-01
2.32655317e-01 6.65391028e-01 1.54509142e-01 6.00053906e-01
2.95250773e-01 2.11292326e-01 4.59267110e-01 1.29788506e+00
-4.64059293e-01 2.75571644e-01 -6.79408252e-01 2.36940071e-01
-1.03937614e+00 -1.18077850e+00 -1.27080560e-01 2.67862177e+00
8.66900444e-01 1.50840312e-01 -1.39077917e-01 1.66280881e-01
3.62794787e-01 2.37050176e-01 -4.66187507e-01 -3.55044812e-01
-4.60927449e-02 1.08392191e+00 9.53864515e-01 8.85698318e-01
-5.93362868e-01 8.61613989e-01 7.60439014e+00 4.54442412e-01
-1.10958827e+00 3.05234164e-01 3.73149030e-02 -1.86756626e-01
-5.17340183e-01 5.66100359e-01 -7.45362461e-01 3.74893934e-01
1.49064016e+00 -1.57697394e-01 1.19118416e+00 4.22079295e-01
-4.57042418e-02 -3.68980408e-01 -1.16545510e+00 1.10072064e+00
2.83226185e-03 -1.28904819e+00 -3.29274863e-01 4.92760360e-01
7.48966515e-01 3.31858158e-01 -6.00044243e-02 1.36381418e-01
4.54842240e-01 -9.75345731e-01 5.18038213e-01 4.56155300e-01
1.19212580e+00 -5.90707898e-01 4.99926418e-01 3.49897385e-01
-7.79194772e-01 4.50485349e-01 -6.52553201e-01 -1.07650712e-01
4.38633204e-01 5.55243015e-01 -6.36614084e-01 4.42634702e-01
4.06954169e-01 2.30311215e-01 -5.10853752e-02 3.13771755e-01
-2.34466344e-01 4.23305959e-01 -6.57323003e-01 -8.58832034e-04
6.45760819e-02 -9.65697408e-01 4.83422756e-01 4.79127765e-01
7.53927469e-01 7.34562039e-01 -3.48683327e-01 8.41100574e-01
-6.08791709e-02 -2.76216567e-01 -7.18975604e-01 -2.92533547e-01
5.41660070e-01 8.43025804e-01 -6.64863944e-01 -4.15128142e-01
-3.51192020e-02 1.14777553e+00 1.17650470e-02 6.58878386e-02
-5.61904728e-01 -6.12087548e-01 4.59839314e-01 1.63958251e-01
4.55304205e-01 -6.54992759e-01 -1.93583757e-01 -1.82869208e+00
8.09912607e-02 -7.12772131e-01 -1.48844346e-01 -6.18194461e-01
-6.94633782e-01 4.94919598e-01 -2.19044492e-01 -7.73152292e-01
-1.82768866e-01 -6.93039000e-01 -1.74300358e-01 9.29220498e-01
-1.10021079e+00 -3.08416784e-01 2.92845696e-01 5.96152484e-01
-7.80881643e-01 2.47830123e-01 1.36737156e+00 -5.91980703e-02
-3.91259670e-01 2.68518835e-01 5.79549491e-01 -1.47218937e-02
3.15879643e-01 -1.35437322e+00 3.01744252e-01 8.70629072e-01
3.13398033e-01 1.02592134e+00 9.32028294e-01 -2.64374793e-01
-2.09225273e+00 -3.76682132e-01 9.09825861e-01 -5.58471024e-01
8.84037793e-01 -3.68746191e-01 -3.02743584e-01 9.48965847e-01
1.01468414e-01 4.04193938e-01 8.21782053e-01 2.98135668e-01
-5.32592297e-01 -1.25276400e-02 -1.31723440e+00 2.42779046e-01
1.11134219e+00 -1.29511738e+00 -2.60825455e-01 5.14175713e-01
2.06720009e-01 -3.71627450e-01 -1.15435219e+00 5.65501899e-02
4.33571398e-01 -1.07384694e+00 8.19447756e-01 -4.30942178e-01
-6.06630184e-02 -3.92776072e-01 -6.93895817e-01 -9.18297410e-01
-4.82271761e-02 -1.11315024e+00 1.39411077e-01 -9.58352238e-02
3.61208290e-01 -6.61963165e-01 9.09753919e-01 5.41300356e-01
9.97298211e-03 -1.10711992e-01 -1.70629787e+00 -6.45185113e-01
2.20067665e-01 -4.92676944e-01 2.94713318e-01 8.94835353e-01
5.85851729e-01 5.62998652e-01 -4.17826891e-01 2.84729213e-01
9.96565878e-01 3.32602471e-01 5.40221632e-01 -9.36175764e-01
-7.20820189e-01 -1.01116315e-01 -5.98115146e-01 -1.41448843e+00
-4.50423881e-02 -1.13419783e+00 9.67848375e-02 -6.96354330e-01
4.54879880e-01 -3.78318965e-01 -1.51533917e-01 2.48286948e-02
4.24306124e-01 3.76929134e-01 -1.24381006e-01 2.13298112e-01
-1.14104700e+00 6.33770883e-01 1.26946580e+00 1.50623545e-01
3.35585505e-01 -2.58063823e-01 -4.05134261e-01 6.56261981e-01
1.75260514e-01 -6.43207014e-01 7.45808482e-02 -3.63371223e-01
7.52293289e-01 9.55550969e-01 2.61289775e-01 -1.09907782e+00
2.41616905e-01 -1.45623803e-01 -1.73913181e-01 -4.12201971e-01
7.32054651e-01 -2.73339361e-01 2.83599228e-01 5.38358092e-01
-8.97508636e-02 -3.20435464e-01 -2.78008193e-01 2.68532664e-01
-2.30794549e-02 -4.32503074e-01 1.15754497e+00 -3.79505128e-01
7.05114305e-02 3.06055337e-01 -2.47923449e-01 8.57788846e-02
6.14286900e-01 2.23854125e-01 -3.08135122e-01 -3.76182139e-01
-1.01446986e+00 -5.20887852e-01 9.17274296e-01 -7.38409936e-01
2.19370335e-01 -1.19040763e+00 -1.57894835e-01 3.68579149e-01
-1.94517985e-01 -1.61850467e-01 3.24808300e-01 1.28801632e+00
-8.40712905e-01 8.88070941e-01 1.03976175e-01 -5.28958857e-01
-6.66908324e-01 4.23725307e-01 5.52509427e-01 -2.05891639e-01
-3.18103343e-01 1.01476920e+00 -4.13399279e-01 -5.75240731e-01
-5.34980893e-01 -2.34881505e-01 6.93629980e-01 -4.84744936e-01
5.34371614e-01 4.45558995e-01 1.89247355e-01 -8.45431268e-01
-4.79001105e-01 5.32447219e-01 6.53488263e-02 -7.55138636e-01
1.15410781e+00 -1.90290540e-01 -6.44964457e-01 3.25120568e-01
1.36111867e+00 4.52289492e-01 -8.23623180e-01 -4.36207592e-01
-1.13393307e-01 -3.17423880e-01 2.32097313e-01 -4.54806477e-01
-6.81321025e-01 9.17146862e-01 3.71315628e-01 4.77625459e-01
6.55479789e-01 2.14333534e-01 7.01898217e-01 9.66037214e-01
1.28494585e+00 -8.53786886e-01 -2.37743601e-01 6.37423217e-01
2.32319698e-01 -9.26200151e-01 1.32375374e-01 -1.21360160e-01
-9.83963087e-02 1.01547372e+00 -2.19215021e-01 -3.76278609e-01
7.50744462e-01 2.93387361e-02 -3.44217628e-01 -7.78894961e-01
-5.07382929e-01 1.18866533e-01 -6.81381002e-02 9.22274441e-02
2.51740813e-01 4.38561529e-01 -1.14220448e-01 -1.20676905e-01
-7.00194418e-01 -1.84578434e-01 9.49755609e-01 8.19059491e-01
-6.35837257e-01 -1.27476621e+00 -3.33202749e-01 3.20891410e-01
-4.00472432e-01 -2.43526280e-01 9.08699483e-02 3.76112640e-01
-1.89990714e-01 9.11212564e-01 -1.03672400e-01 -2.94442385e-01
1.92773461e-01 1.89502656e-01 1.13253987e+00 -6.69149637e-01
4.34118390e-01 -1.03963286e-01 1.69226915e-01 -1.05606496e+00
-5.00630498e-01 -1.04530025e+00 -1.28283083e+00 -8.63092184e-01
-4.87205744e-01 7.44909823e-01 7.62079179e-01 1.18773079e+00
3.27156425e-01 -1.45976782e-01 8.04295063e-01 -6.88232005e-01
-1.20840776e+00 -1.00394154e+00 -1.21657562e+00 -1.62836686e-02
2.04085916e-01 -5.35585225e-01 -8.62996042e-01 -2.98747092e-01] | [5.739383697509766, 4.840928554534912] |
df1970ed-4e68-42b3-bcdd-1ba4726a5263 | private-training-set-inspection-in-mlaas | 2305.09058 | null | https://arxiv.org/abs/2305.09058v1 | https://arxiv.org/pdf/2305.09058v1.pdf | Private Training Set Inspection in MLaaS | Machine Learning as a Service (MLaaS) is a popular cloud-based solution for customers who aim to use an ML model but lack training data, computation resources, or expertise in ML. In this case, the training datasets are typically a private possession of the ML or data companies and are inaccessible to the customers, but the customers still need an approach to confirm that the training datasets meet their expectations and fulfil regulatory measures like fairness. However, no existing work addresses the above customers' concerns. This work is the first attempt to solve this problem, taking data origin as an entry point. We first define origin membership measurement and based on this, we then define diversity and fairness metrics to address customers' concerns. We then propose a strategy to estimate the values of these two metrics in the inaccessible training dataset, combining shadow training techniques from membership inference and an efficient featurization scheme in multiple instance learning. The evaluation contains an application of text review polarity classification applications based on the language BERT model. Experimental results show that our solution can achieve up to 0.87 accuracy for membership inspection and up to 99.3% confidence in inspecting diversity and fairness distribution. | ['Po-Yu Chen', 'Tongtong Xu', 'Mingxue Xu'] | 2023-05-15 | null | null | null | null | ['multiple-instance-learning'] | ['methodology'] | [-1.78215176e-01 -4.06772010e-02 -7.33704865e-01 -7.49185741e-01
-9.75387931e-01 -4.67531443e-01 3.69372725e-01 4.39191163e-01
-1.78681627e-01 9.61326063e-01 -6.57390594e-01 -5.44917941e-01
-2.50745386e-01 -7.80024230e-01 -4.58027035e-01 -6.44127011e-01
2.64274299e-01 9.02369499e-01 -1.15460463e-01 -5.12339845e-02
7.28240430e-01 4.92625564e-01 -1.90246916e+00 5.86414933e-01
1.49604392e+00 1.40129244e+00 -2.25200132e-01 3.43306869e-01
-5.77431381e-01 8.61708403e-01 -9.15374517e-01 -6.66861951e-01
2.73691028e-01 -9.38843787e-02 -9.17369246e-01 -3.17761712e-02
3.03704619e-01 -2.60496557e-01 7.27161109e-01 1.13328123e+00
3.03872019e-01 -1.80597231e-01 7.49822795e-01 -1.82669497e+00
-5.63082457e-01 6.37247384e-01 -4.33651656e-01 -8.48118663e-02
2.94919163e-01 -1.33047849e-01 1.21319139e+00 -6.00256920e-01
2.20148712e-01 9.30946290e-01 3.61342162e-01 3.58725280e-01
-7.57239521e-01 -6.34832561e-01 2.35509411e-01 3.51112843e-01
-1.33981061e+00 -4.53033119e-01 5.86205482e-01 -3.96080732e-01
7.01932907e-01 6.56831622e-01 2.69304574e-01 4.35060471e-01
-1.91921785e-01 9.08781826e-01 1.23073804e+00 -6.57318890e-01
5.58956325e-01 1.00574529e+00 5.00745595e-01 1.30051613e-01
4.16736692e-01 -5.63567877e-01 -3.21654975e-01 -3.25648159e-01
-5.05759344e-02 -1.20490983e-01 -3.37304361e-02 -2.38234267e-01
-2.27148205e-01 1.11132765e+00 -1.47311566e-02 2.36354113e-01
-3.78349125e-01 -5.74255645e-01 6.23323321e-01 4.37013835e-01
6.38096631e-01 3.33839089e-01 -7.07304716e-01 -1.92649160e-02
-1.09984875e+00 1.16478890e-01 1.22929609e+00 8.15844715e-01
5.71078360e-01 -2.03907996e-01 1.05157986e-01 6.67879522e-01
3.00687253e-01 3.32825273e-01 5.00607967e-01 -6.64673984e-01
4.43096012e-01 1.03381324e+00 1.30197495e-01 -8.32781017e-01
1.38076007e-01 -3.14366192e-01 -5.50527513e-01 3.09129983e-01
4.02609795e-01 3.05925518e-01 -2.86764562e-01 1.22716808e+00
4.12740618e-01 -1.01010136e-01 2.88824826e-01 6.11371219e-01
7.62561858e-01 4.78636801e-01 -1.02884561e-01 -7.06353486e-01
1.16252053e+00 -7.11105645e-01 -7.61006176e-01 4.80585843e-01
7.90863693e-01 -7.00870275e-01 1.13619161e+00 1.04838896e+00
-9.28667963e-01 -2.87128836e-01 -8.52364957e-01 2.28366658e-01
-6.32681489e-01 3.68709207e-01 8.16598237e-01 1.23761213e+00
-5.52624941e-01 5.26345849e-01 -1.38558745e-01 9.85527039e-02
5.41256905e-01 6.18909895e-01 -8.45312625e-02 -1.14095777e-01
-1.29503930e+00 8.22962224e-01 2.95679092e-01 -2.84571834e-02
-2.98450410e-01 -6.05192244e-01 -5.34145117e-01 9.14760008e-02
2.88257927e-01 -1.27300054e-01 1.05889308e+00 -1.45291352e+00
-1.14598680e+00 9.33746457e-01 4.70253602e-02 -4.36190128e-01
6.21772110e-01 1.61105677e-01 -6.10886216e-01 -1.58717036e-01
1.90789938e-01 -7.34316856e-02 5.43930471e-01 -1.31726658e+00
-1.07418430e+00 -4.74479616e-01 -9.36215073e-02 2.12111212e-02
-4.50527340e-01 1.71659857e-01 -9.67057794e-03 5.49725406e-02
-9.13745984e-02 -4.51011866e-01 1.59984663e-01 -3.56587052e-01
-3.22826982e-01 -5.66070378e-01 7.83533335e-01 -7.16307163e-01
1.49105823e+00 -1.80591202e+00 -6.26407921e-01 8.44353378e-01
3.94325471e-03 5.97083867e-01 2.94879973e-01 1.67658627e-01
1.40874788e-01 4.17385012e-01 -1.52005866e-01 -2.06252813e-01
2.71352142e-01 2.51488775e-01 -2.34528318e-01 4.12859440e-01
-8.27796012e-02 3.61812204e-01 -4.75211173e-01 -5.36900103e-01
5.05359769e-02 1.52723953e-01 -5.34154296e-01 2.92930812e-01
-2.37309113e-01 3.60570326e-02 -2.90638804e-01 9.62344944e-01
1.11965621e+00 -1.76182672e-01 3.48640740e-01 -7.82774091e-02
-6.93623200e-02 3.23905796e-01 -1.44478893e+00 8.18851054e-01
-5.80242038e-01 7.60519579e-02 1.05543397e-01 -1.34768093e+00
9.89377201e-01 1.19066671e-01 5.82593381e-01 -5.02662003e-01
2.07140326e-01 7.94461727e-01 -5.38407490e-02 -5.91468573e-01
4.99484211e-01 -7.77970925e-02 -1.32257468e-04 5.27263701e-01
-2.86346823e-01 2.51104012e-02 2.71501303e-01 2.36234069e-01
4.66173887e-01 -3.68347704e-01 2.62574941e-01 -2.18098581e-01
9.45853353e-01 -1.48227379e-01 3.82973254e-01 5.02570331e-01
-2.35401452e-01 1.41058668e-01 7.79510319e-01 -3.16467226e-01
-7.84491360e-01 -5.51630855e-01 -4.05166566e-01 1.02173805e+00
-2.18496755e-01 -2.77995497e-01 -7.49457717e-01 -9.10752058e-01
2.94868946e-01 1.11109269e+00 -2.82969445e-01 7.37453625e-02
7.11878836e-02 -7.32458413e-01 2.53704578e-01 3.42102051e-02
3.66828620e-01 -8.40941727e-01 -2.21929386e-01 -1.68631017e-01
-2.32553944e-01 -9.96026695e-01 -5.22638522e-02 9.29342806e-02
-4.89626527e-01 -1.22015727e+00 -1.01565637e-01 -4.44360644e-01
6.44310594e-01 -1.45699501e-01 1.12835455e+00 2.50578016e-01
-2.64401108e-01 1.51633620e-01 -4.14083838e-01 -6.63079858e-01
-6.35484397e-01 1.50623336e-01 9.80515629e-02 2.61976004e-01
8.59234929e-01 -3.95947427e-01 -2.05258101e-01 2.43156016e-01
-7.90486395e-01 -4.44954813e-01 2.89763570e-01 6.35738373e-01
5.36849082e-01 3.46566707e-01 1.23326123e+00 -1.26445389e+00
8.73040199e-01 -7.60751307e-01 -8.29123259e-01 6.21991694e-01
-1.26232517e+00 -2.55582154e-01 7.67354429e-01 -1.81057230e-01
-7.64122725e-01 -1.17510051e-01 -5.45944422e-02 -3.33291262e-01
-6.89766407e-02 4.82100487e-01 -6.57238066e-01 1.79722719e-02
5.05265772e-01 -5.23029938e-02 2.16861457e-01 -3.16989928e-01
1.28813997e-01 1.35817397e+00 1.10284656e-01 -6.44440234e-01
1.54470786e-01 5.62802926e-02 -2.95766085e-01 -3.37373048e-01
-8.69204879e-01 -6.96905732e-01 -5.43036044e-01 -1.97318718e-01
2.57470459e-01 -6.01311088e-01 -1.38184357e+00 1.82420969e-01
-8.46587956e-01 3.19554806e-01 -2.29194552e-01 3.10788393e-01
-4.06609565e-01 3.78963351e-01 -2.17670381e-01 -1.49512959e+00
-6.60220325e-01 -1.00692570e+00 7.25528419e-01 1.52384892e-01
4.40620147e-02 -8.04736912e-01 -3.48246127e-01 8.28607619e-01
4.77614194e-01 1.96561851e-02 1.03126442e+00 -1.33946419e+00
-3.42250049e-01 -6.60020411e-01 -1.82479337e-01 7.07410216e-01
2.06280082e-01 3.63145828e-01 -1.12470627e+00 -9.67931598e-02
1.58375472e-01 -5.18805504e-01 4.76493746e-01 2.50488073e-01
1.50133741e+00 -6.06476486e-01 1.28942758e-01 1.35656610e-01
1.34844589e+00 6.48739114e-02 6.39502883e-01 3.66220385e-01
3.52505237e-01 9.09407973e-01 8.80115032e-01 8.43542516e-01
4.46780443e-01 4.59290445e-01 6.14624679e-01 1.89838618e-01
6.00964665e-01 7.63486102e-02 2.39452004e-01 7.28081286e-01
4.66959849e-02 -1.19131610e-01 -8.32047999e-01 3.23533624e-01
-1.94165778e+00 -1.09998453e+00 -4.63412166e-01 2.51017833e+00
9.92335200e-01 2.24117890e-01 3.01517814e-01 6.27284706e-01
7.54689634e-01 -5.61407685e-01 -5.34682095e-01 -1.05579269e+00
-6.15027808e-02 2.48766057e-02 3.64212573e-01 5.35956621e-01
-9.25209165e-01 6.36729240e-01 5.59299755e+00 1.23503864e+00
-1.04468536e+00 -6.92756241e-03 1.12807739e+00 -4.64960188e-02
-6.92321539e-01 -1.34648299e-02 -9.82269526e-01 6.36223197e-01
1.07575285e+00 -2.09705889e-01 3.47258180e-01 1.20363581e+00
1.55202225e-01 -3.68357241e-01 -9.61576819e-01 9.84242678e-01
2.59089470e-01 -1.08802271e+00 4.24967473e-03 1.50995940e-01
6.23096228e-01 -4.83119875e-01 1.89343899e-01 5.21174669e-01
-7.24181682e-02 -1.16361582e+00 5.27941525e-01 6.41630113e-01
7.25425959e-01 -1.17733574e+00 1.09711003e+00 8.17695916e-01
-5.48650086e-01 -2.81697869e-01 -3.16131413e-01 4.82882969e-02
-1.69365168e-01 1.11789477e+00 -8.53886485e-01 6.87892258e-01
4.12451029e-01 5.84276021e-02 -3.51340830e-01 8.04525077e-01
2.05168203e-01 5.28195024e-01 -2.16765836e-01 -2.30832785e-01
-1.66827470e-01 -3.67764354e-01 -1.23875767e-01 9.22403336e-01
2.03337118e-01 5.62663935e-02 2.19319433e-01 8.06902885e-01
-1.27902880e-01 7.80345798e-01 -3.73164952e-01 -8.81512836e-03
3.78085226e-01 1.35465181e+00 -3.54682118e-01 -4.03455973e-01
-3.97512257e-01 5.95983267e-01 1.88526109e-01 -1.45910112e-02
-7.93719292e-01 -3.63361269e-01 4.03915972e-01 2.69302756e-01
-8.51335078e-02 4.60474402e-01 -5.14476180e-01 -1.04414499e+00
3.02380741e-01 -1.33708537e+00 5.01508176e-01 -2.94340611e-01
-1.66290629e+00 3.22407126e-01 -2.02573642e-01 -1.13852370e+00
-3.54843080e-01 -7.52251208e-01 -4.77625132e-01 1.03845823e+00
-1.76640058e+00 -1.29491675e+00 -5.31839281e-02 6.02902770e-01
1.53902590e-01 -5.90884864e-01 8.92420113e-01 5.72342098e-01
-4.51213926e-01 8.51116598e-01 2.01690510e-01 -2.44765952e-01
5.76032162e-01 -1.11567080e+00 -4.91103768e-01 4.57559407e-01
-1.37815893e-01 5.70805848e-01 4.98907685e-01 -5.69117427e-01
-1.12000561e+00 -8.38013709e-01 1.26686966e+00 -4.46001321e-01
3.64285648e-01 -3.34794879e-01 -9.91270006e-01 3.14155430e-01
-8.22219402e-02 2.25072689e-02 1.21868360e+00 3.20140958e-01
-2.31445044e-01 -4.71283764e-01 -1.73704922e+00 -1.07325688e-02
2.70802736e-01 -4.95915204e-01 -7.83439204e-02 6.37349129e-01
2.22337008e-01 -1.25635043e-01 -9.59625483e-01 3.69841665e-01
4.41993177e-01 -1.02226543e+00 4.66752112e-01 -9.00761843e-01
2.83970743e-01 -9.03146192e-02 -5.17180443e-01 -6.19068801e-01
3.37905139e-01 -3.15536648e-01 -1.78002998e-01 1.61604905e+00
6.61154866e-01 -6.38463676e-01 8.21113348e-01 1.24566889e+00
1.31872013e-01 -9.53441799e-01 -8.39544713e-01 -7.19484329e-01
7.76660070e-02 -6.30131602e-01 8.34531903e-01 1.34520710e+00
1.48860022e-01 5.55311292e-02 -3.03542018e-01 4.00020964e-02
5.41530252e-01 5.39536715e-01 8.01061392e-01 -1.36267316e+00
-7.42190927e-02 -4.85949665e-01 -1.79879606e-01 -3.77438068e-01
4.76106733e-01 -9.89231765e-01 -3.46049249e-01 -1.43292844e+00
1.75245091e-01 -9.99520600e-01 -3.06552023e-01 4.21986729e-01
1.05414785e-01 -2.52984464e-01 -4.17245999e-02 2.56672442e-01
-7.49076664e-01 2.30235472e-01 6.51042223e-01 7.96224996e-02
-2.22138658e-01 6.78608656e-01 -9.98818636e-01 5.53116262e-01
1.06526506e+00 -3.33364010e-01 -5.05758166e-01 3.86712462e-01
4.56044465e-01 -1.69076070e-01 1.06911473e-01 -5.25059581e-01
2.45836273e-01 -5.30431986e-01 2.38730028e-01 -6.74500644e-01
2.64002513e-02 -1.08284461e+00 -2.64044479e-02 2.72905856e-01
-3.83653253e-01 -2.41977990e-01 -1.20268449e-01 2.07991153e-01
-1.74701273e-01 -6.58195436e-01 6.39108181e-01 -1.54494178e-02
-3.95567060e-01 1.55724153e-01 1.59308016e-02 -6.68642297e-02
1.12485087e+00 -1.45497039e-01 -3.73364151e-01 -4.50435489e-01
-7.03413367e-01 3.26944858e-01 2.40999356e-01 7.87109062e-02
4.16263789e-01 -1.18453610e+00 -7.57046103e-01 3.62421989e-01
1.77168205e-01 -4.40085322e-01 4.85090539e-02 8.48031342e-01
-3.88752788e-01 4.63914931e-01 -1.11437902e-01 -4.92000550e-01
-1.34245574e+00 6.87605321e-01 2.59188414e-01 -2.31107965e-01
2.48399541e-01 6.78913057e-01 -4.15183991e-01 -7.91523755e-01
4.10756856e-01 -1.17684796e-01 -4.36049193e-01 3.73004943e-01
5.17434359e-01 3.22423398e-01 5.37511826e-01 -5.14967322e-01
-4.07874525e-01 5.25334515e-02 -7.28138909e-02 1.47449210e-01
1.05868971e+00 -4.24676128e-02 -4.61265355e-01 5.04734159e-01
1.16409647e+00 2.84620553e-01 -4.18898761e-01 -9.58576391e-04
1.00368083e-01 -7.89639413e-01 -1.57066375e-01 -9.37085271e-01
-1.07131922e+00 8.31324100e-01 6.70334935e-01 6.83763325e-01
1.15391958e+00 -2.81425387e-01 3.79758030e-01 2.67512977e-01
3.00066262e-01 -1.53657103e+00 -2.68130302e-01 1.79876715e-01
6.00581706e-01 -1.64671469e+00 1.61157429e-01 -4.46663201e-01
-9.64701593e-01 1.04906654e+00 7.06847966e-01 3.18108410e-01
7.10178256e-01 1.21228248e-01 -1.91428810e-02 -2.66010291e-03
-8.50655556e-01 -4.49463017e-02 2.99262375e-01 4.77673233e-01
4.42629665e-01 3.19999218e-01 -7.33940899e-01 1.23804975e+00
-1.70767590e-01 2.49516126e-03 3.93798411e-01 5.59639573e-01
-7.44819283e-01 -1.27326596e+00 -3.43095601e-01 8.43580663e-01
-7.62589991e-01 -2.12748200e-01 -4.19103891e-01 6.38820648e-01
4.46870178e-01 1.10198951e+00 -9.39676613e-02 -2.01372921e-01
2.56275028e-01 2.08649993e-01 8.61296207e-02 -3.98818105e-01
-6.84098125e-01 -8.17815736e-02 3.03004473e-01 -7.20280632e-02
-4.54444706e-01 -6.43086970e-01 -1.30639958e+00 -7.78598487e-01
-7.99999475e-01 6.07479155e-01 1.13273180e+00 8.09938908e-01
2.09779426e-01 6.79912195e-02 1.11670768e+00 -1.41894132e-01
-9.08526540e-01 -5.16519964e-01 -8.76594722e-01 4.50387329e-01
-1.07590042e-01 -2.39528283e-01 -6.43534958e-01 -2.61711955e-01] | [9.17146110534668, 4.40355110168457] |
9322baa5-5712-49d7-bba3-74150ef4bb4c | a-closed-form-solution-to-photorealistic | 1802.06474 | null | http://arxiv.org/abs/1802.06474v5 | http://arxiv.org/pdf/1802.06474v5.pdf | A Closed-form Solution to Photorealistic Image Stylization | Photorealistic image stylization concerns transferring style of a reference
photo to a content photo with the constraint that the stylized photo should
remain photorealistic. While several photorealistic image stylization methods
exist, they tend to generate spatially inconsistent stylizations with
noticeable artifacts. In this paper, we propose a method to address these
issues. The proposed method consists of a stylization step and a smoothing
step. While the stylization step transfers the style of the reference photo to
the content photo, the smoothing step ensures spatially consistent
stylizations. Each of the steps has a closed-form solution and can be computed
efficiently. We conduct extensive experimental validations. The results show
that the proposed method generates photorealistic stylization outputs that are
more preferred by human subjects as compared to those by the competing methods
while running much faster. Source code and additional results are available at
https://github.com/NVIDIA/FastPhotoStyle . | ['Ming-Hsuan Yang', 'Xueting Li', 'Ming-Yu Liu', 'Jan Kautz', 'Yijun Li'] | 2018-02-19 | a-closed-form-solution-to-photorealistic-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Yijun_Li_A_Closed-form_Solution_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Yijun_Li_A_Closed-form_Solution_ECCV_2018_paper.pdf | eccv-2018-9 | ['image-stylization'] | ['computer-vision'] | [ 1.12835430e-01 2.21213847e-02 1.45623863e-01 -1.24804892e-01
-4.53657120e-01 -7.28911102e-01 6.90973103e-01 -3.26674521e-01
-5.94801269e-02 7.04633057e-01 2.52671480e-01 -1.81548908e-01
5.54717183e-01 -4.67522591e-01 -7.27047920e-01 -6.08076036e-01
7.35525131e-01 3.38944085e-02 2.97660917e-01 1.89922929e-01
5.68356216e-01 5.83143115e-01 -1.29959142e+00 5.54730557e-02
1.05125022e+00 7.71266520e-01 1.62501618e-01 6.52371883e-01
-9.89309922e-02 4.77876991e-01 -6.40614331e-01 -4.21039313e-01
4.74035978e-01 -5.87096989e-01 -7.34713554e-01 5.22436559e-01
9.31941032e-01 -4.72431272e-01 -9.22945514e-02 1.19938469e+00
5.19633651e-01 1.82977378e-01 5.14293730e-01 -1.30841196e+00
-1.12426019e+00 4.33104634e-02 -1.24234664e+00 -8.99855867e-02
4.31351990e-01 1.71624437e-01 6.44303024e-01 -9.35490429e-01
8.71805370e-01 1.44346189e+00 3.13533872e-01 6.17331028e-01
-1.44531119e+00 -6.93636656e-01 1.36623457e-01 -3.20133239e-01
-1.28231525e+00 -5.36967933e-01 7.30416834e-01 -2.36616209e-01
3.14516276e-01 2.87984699e-01 6.37438118e-01 8.42427433e-01
3.44583541e-01 7.78186142e-01 1.23668706e+00 -4.39260632e-01
3.09568644e-01 1.32014588e-01 -2.09147170e-01 6.54768884e-01
3.26039374e-01 -3.04572582e-02 -3.65023822e-01 -1.33200824e-01
1.24651778e+00 -1.77782476e-02 -2.70372152e-01 -5.21988332e-01
-1.14292490e+00 4.79848772e-01 3.64769131e-01 -7.60473823e-03
-2.99915373e-01 4.02374059e-01 1.18721329e-01 4.30119544e-04
5.98213911e-01 4.27954286e-01 3.86164427e-01 -2.32170317e-02
-9.79558408e-01 1.83562472e-01 5.13601482e-01 1.21663141e+00
8.27632427e-01 3.01404446e-01 -2.15207294e-01 7.84042180e-01
-3.95827293e-02 6.18296325e-01 2.06066355e-01 -1.14134657e+00
3.19905907e-01 4.30915087e-01 4.05702829e-01 -1.03804290e+00
3.17334644e-02 2.02242017e-01 -7.95955360e-01 6.19472921e-01
4.78893191e-01 -4.68576580e-01 -8.20781291e-01 1.32075870e+00
6.12861454e-01 2.13717461e-01 -2.38211632e-01 9.59010303e-01
7.67637193e-01 8.37198317e-01 7.12474110e-03 9.12356302e-02
1.33757079e+00 -1.12398636e+00 -9.16609704e-01 -9.31682661e-02
3.85218024e-01 -1.38283634e+00 1.53129971e+00 2.59660602e-01
-1.44455957e+00 -4.92517620e-01 -1.00126791e+00 -4.06850576e-01
1.48823857e-01 4.75328833e-01 2.25537956e-01 6.65685177e-01
-1.30562735e+00 3.90860885e-01 -5.31533241e-01 -5.54381192e-01
2.36266077e-01 3.99256544e-03 -2.24577233e-01 1.05505548e-01
-5.30739367e-01 7.17393935e-01 9.37506184e-02 -1.24412335e-01
-5.82632780e-01 -7.21109390e-01 -7.30357409e-01 -8.44050944e-02
3.44353169e-01 -7.97782123e-01 1.35996437e+00 -1.36855781e+00
-1.81251967e+00 9.80139971e-01 -5.93168199e-01 2.72972137e-03
8.39753449e-01 -4.97480989e-01 -5.63654341e-02 1.70718953e-01
9.53683034e-02 9.54030991e-01 1.17668414e+00 -1.66767776e+00
-7.85142720e-01 -4.35341932e-02 2.95758750e-02 4.28840965e-01
-7.80200139e-02 -7.65950158e-02 -8.84504318e-01 -1.04668474e+00
-6.35369718e-02 -1.05793607e+00 -5.63477986e-02 2.04785392e-01
-7.77977645e-01 1.05796687e-01 9.83162940e-01 -4.96860892e-01
1.09252799e+00 -2.22350597e+00 5.49001694e-02 4.37354529e-03
7.95112476e-02 2.41640329e-01 -1.30826220e-01 5.22275984e-01
-1.87822223e-01 1.64737165e-01 9.02791787e-03 -6.27237558e-01
-2.54993349e-01 -1.95106983e-01 -5.61189651e-01 5.54372847e-01
-2.82655265e-02 9.53906059e-01 -7.44574368e-01 -5.80173790e-01
4.08051699e-01 5.10612905e-01 -4.70645308e-01 1.73378706e-01
-3.98126952e-02 4.57376808e-01 -2.62787551e-01 5.29968441e-01
1.06786954e+00 -8.41558799e-02 9.64375660e-02 -3.36711168e-01
-1.89504579e-01 -4.84966636e-02 -1.28404474e+00 1.51867902e+00
-4.13375944e-01 7.01848447e-01 -9.74842608e-02 -2.24748492e-01
8.75948429e-01 1.80636957e-01 1.92908451e-01 -6.36396170e-01
2.49846146e-01 -7.87571911e-03 -5.87959588e-01 -2.56343991e-01
9.35908020e-01 -2.09817812e-01 2.20539972e-01 8.91903877e-01
-4.22852010e-01 -5.86600363e-01 1.77466050e-01 2.76361436e-01
4.71905798e-01 3.65100175e-01 3.76597971e-01 -3.40533167e-01
4.64497238e-01 6.20472617e-02 4.29081529e-01 4.47726101e-01
-3.66328023e-02 8.65135372e-01 5.87786734e-01 -4.02621269e-01
-1.31066954e+00 -1.21340013e+00 2.89628446e-01 6.07509136e-01
6.52656019e-01 -3.29507947e-01 -1.18522060e+00 -4.10732269e-01
-1.63914159e-01 7.55882084e-01 -6.12736642e-01 7.06365705e-02
-5.49727976e-01 -5.06560691e-02 3.62818092e-01 4.18264478e-01
6.70685947e-01 -1.08696973e+00 -6.39761090e-01 -2.86459088e-01
-1.29521444e-01 -1.01558244e+00 -1.22962117e+00 -5.69096148e-01
-1.04153490e+00 -8.44190121e-01 -1.06723201e+00 -8.35665762e-01
1.10730875e+00 8.01169693e-01 9.14771259e-01 2.39516035e-01
-1.32442132e-01 2.68634111e-01 -2.18284741e-01 -3.33362252e-01
-3.51638407e-01 -1.71745226e-01 -4.00148407e-02 1.90107465e-01
-2.96430364e-02 -3.74155521e-01 -8.18114042e-01 5.32882869e-01
-1.18844426e+00 5.05515039e-01 4.03228670e-01 5.43948948e-01
8.45802486e-01 -1.41029581e-01 3.66541445e-02 -1.03378344e+00
8.38951409e-01 6.74919710e-02 -8.60492885e-01 1.26358435e-01
-3.93330306e-01 6.04918748e-02 7.16974735e-01 -4.92871553e-01
-1.39419031e+00 2.78898746e-01 3.20162237e-01 -5.29051721e-01
-2.11078450e-01 -2.06054404e-01 -1.46517813e-01 -3.30146402e-03
5.64829886e-01 1.06811918e-01 9.31319967e-02 -4.17887300e-01
6.32885993e-01 4.62875694e-01 6.48612499e-01 -6.92337573e-01
1.17474592e+00 8.76238346e-01 -2.19684005e-01 -9.88603592e-01
-4.65060383e-01 -1.73669502e-01 -6.89388752e-01 -2.83156276e-01
8.53535116e-01 -7.42612183e-01 -5.47092021e-01 4.81556892e-01
-1.20480490e+00 -4.91356462e-01 -4.37191278e-01 1.12621039e-01
-6.04941189e-01 5.96855342e-01 -3.97789001e-01 -6.13948286e-01
-3.83136749e-01 -1.12899494e+00 1.25110757e+00 5.52397668e-01
-5.72565794e-01 -8.79415274e-01 1.59134850e-01 2.54727095e-01
1.66412413e-01 1.38227895e-01 7.02686608e-01 3.96270305e-01
-5.51550925e-01 -1.45645840e-02 -5.52497327e-01 1.60073489e-02
5.84885120e-01 4.45098042e-01 -8.87782395e-01 -2.15354323e-01
-3.03382397e-01 -2.09845304e-02 5.05631924e-01 5.10393918e-01
9.54286754e-01 -1.94022611e-01 -1.56942189e-01 7.34344006e-01
1.37352657e+00 2.96141177e-01 9.03318405e-01 2.33641788e-01
8.84151936e-01 5.48947513e-01 6.70848846e-01 3.25256735e-01
8.07713270e-02 8.03652525e-01 6.64714053e-02 -5.43720305e-01
-4.30142045e-01 -5.00721157e-01 2.93329954e-01 3.48878741e-01
-9.68185365e-02 -2.85831243e-01 -6.42382681e-01 5.37782013e-01
-1.77028501e+00 -9.95133102e-01 -2.10470662e-01 2.22981572e+00
6.71373427e-01 -3.28056127e-01 2.01293543e-01 -3.45352814e-02
8.99784744e-01 2.68285900e-01 -5.96943796e-01 -8.20141494e-01
8.60891342e-02 3.20339687e-02 5.12574255e-01 7.33753502e-01
-8.13456833e-01 1.32505023e+00 6.42698669e+00 7.77604759e-01
-1.35217607e+00 -1.17162250e-01 7.60597169e-01 -2.87928730e-01
-5.87016582e-01 -5.64348809e-02 -5.36708415e-01 4.06654835e-01
2.94575781e-01 -4.37359899e-01 2.87036330e-01 6.06720030e-01
6.97174251e-01 -6.63908899e-01 -9.01251137e-01 1.09762549e+00
3.97717096e-02 -1.42710912e+00 3.18999439e-01 -6.83632120e-02
1.09658158e+00 -5.95116079e-01 3.04706395e-01 -5.14386714e-01
4.60204363e-01 -9.22671795e-01 1.03450990e+00 3.41164351e-01
1.00306737e+00 -7.98770368e-01 2.05996528e-01 -7.18639418e-02
-1.22195399e+00 4.32746738e-01 -2.78789163e-01 -4.33775783e-02
3.01239252e-01 5.09485543e-01 -4.47654009e-01 4.40378398e-01
6.74586594e-01 5.81088781e-01 -6.40199065e-01 9.56591904e-01
-5.53557456e-01 3.01139802e-01 5.09303957e-02 6.39071465e-02
1.73468485e-01 -7.28415668e-01 4.60488170e-01 9.29790020e-01
4.27159935e-01 1.09318078e-01 -1.75745770e-01 1.03733718e+00
-9.27755460e-02 3.02963883e-01 -7.92740047e-01 6.60590604e-02
4.80018377e-01 1.23855674e+00 -9.83967364e-01 -4.01593685e-01
-2.43208811e-01 1.55711222e+00 2.14569926e-01 7.16458023e-01
-8.53123605e-01 -4.39812690e-01 7.41420269e-01 3.05543929e-01
6.84114173e-02 -2.60494322e-01 -8.41714680e-01 -1.17537439e+00
2.07205102e-01 -7.38130748e-01 8.91370177e-02 -1.25954342e+00
-9.20104504e-01 6.72735274e-01 -1.42930567e-01 -1.43721628e+00
-2.12185960e-02 -1.92932591e-01 -9.79913712e-01 9.41424310e-01
-1.25234985e+00 -9.86267567e-01 -6.22307003e-01 4.70133305e-01
6.09353006e-01 1.68057799e-01 5.15933514e-01 -6.19427906e-03
-5.40961504e-01 4.43741709e-01 8.55313465e-02 -1.38839841e-01
1.05703771e+00 -1.19301283e+00 6.94057941e-01 1.15974033e+00
2.01077089e-02 5.80877841e-01 8.80575955e-01 -7.51459897e-01
-1.08761883e+00 -1.06130075e+00 7.45530605e-01 -2.09419981e-01
3.09998006e-01 -1.39987454e-01 -7.17032909e-01 6.76632345e-01
6.69806182e-01 -1.81121230e-01 2.78609246e-01 -3.41766506e-01
-2.38638714e-01 8.00395310e-02 -1.08699024e+00 1.26791549e+00
9.44154739e-01 -3.12664866e-01 -3.25081110e-01 3.08577698e-02
2.83368587e-01 -5.52233160e-01 -4.05500144e-01 -2.85877019e-01
6.41810119e-01 -1.22214878e+00 9.03108656e-01 -1.30001232e-01
7.09871590e-01 -6.82845771e-01 2.73018450e-01 -1.51299858e+00
-2.16596320e-01 -9.99799550e-01 2.94232726e-01 1.22612238e+00
2.79035002e-01 -6.58722103e-01 8.01410139e-01 6.45769119e-01
1.25727430e-01 -5.54826081e-01 -5.16860425e-01 -9.46969509e-01
2.80222669e-02 1.89707473e-01 5.54429948e-01 7.39216328e-01
-9.73280147e-03 2.71341443e-01 -6.33799374e-01 5.50626367e-02
6.44591689e-01 4.35197920e-01 1.24941647e+00 -7.75715590e-01
8.92041326e-02 -3.22509170e-01 -3.90867144e-02 -1.02721238e+00
5.86257242e-02 -4.12099332e-01 -3.92912515e-02 -1.65556633e+00
1.41801432e-01 -2.47860551e-01 4.96063471e-01 3.22779536e-01
-4.28289413e-01 4.75826740e-01 3.46626908e-01 2.40711212e-01
-2.08538696e-01 5.47057033e-01 1.70896530e+00 7.69876614e-02
-4.88687843e-01 -2.27634713e-01 -7.85878479e-01 7.75290787e-01
1.03386629e+00 -2.51900166e-01 -6.33682430e-01 -7.01870441e-01
-1.53601483e-01 -1.05244696e-01 2.08742902e-01 -8.03172648e-01
7.30055524e-03 -3.53329629e-01 4.31609392e-01 -5.32951772e-01
4.15120035e-01 -6.35970533e-01 4.17234898e-01 4.44324762e-01
-1.20092176e-01 4.23792690e-01 3.28471571e-01 3.83943886e-01
-1.27289236e-01 -1.41237378e-01 1.26898229e+00 9.69026014e-02
-4.91512150e-01 1.91754624e-01 -3.97027403e-01 9.82690826e-02
1.26914346e+00 -6.27400935e-01 -2.44100615e-01 -7.35813856e-01
-2.98001856e-01 3.48005742e-02 1.17649186e+00 4.08496529e-01
7.33450353e-01 -1.47369194e+00 -5.08251965e-01 8.00451860e-02
-1.24577612e-01 -1.43797681e-01 2.37871230e-01 6.61908448e-01
-9.76966202e-01 3.78156692e-01 -4.77841288e-01 -4.89094883e-01
-1.62076473e+00 6.15836680e-01 2.50667453e-01 1.98743030e-01
-7.88341463e-01 6.23841941e-01 7.56908238e-01 -8.39965418e-02
1.31905839e-01 -3.50638121e-01 1.59748033e-01 -1.03682071e-01
5.91074884e-01 5.32876253e-01 -2.66618878e-01 -7.20237315e-01
-9.37965978e-03 9.58820879e-01 -3.68348993e-02 -4.73827600e-01
9.47902679e-01 -5.33748567e-01 7.68780112e-02 1.99098185e-01
9.07515705e-01 4.06577110e-01 -1.50410187e+00 -2.05717072e-01
-2.67508179e-01 -9.48684514e-01 -1.32737413e-01 -4.18843538e-01
-1.17713165e+00 7.91374981e-01 3.55502933e-01 1.53039151e-03
1.22997701e+00 -3.06267053e-01 9.03013527e-01 -9.41843987e-02
1.87639832e-01 -1.08934367e+00 2.05874398e-01 2.47609690e-01
1.08337784e+00 -9.80371535e-01 1.72795728e-01 -7.38951683e-01
-1.02972686e+00 9.04250562e-01 6.92930996e-01 -4.63284731e-01
3.20021182e-01 2.48189643e-01 4.70127344e-01 -1.21452808e-01
-4.72778320e-01 2.94389334e-02 3.87680143e-01 5.98439932e-01
5.36097229e-01 -5.03399782e-02 -3.87502313e-01 -1.03337996e-01
-2.03640044e-01 -5.74084334e-02 6.79567814e-01 6.85941875e-01
-2.43860021e-01 -1.08059835e+00 -7.39115715e-01 -2.85884235e-02
-1.72931194e-01 -5.48389368e-02 -7.55799055e-01 7.84739792e-01
-2.95356840e-01 1.01653492e+00 2.10887313e-01 -1.03083014e-01
4.59860235e-01 -2.29646370e-01 5.18459260e-01 -5.45278966e-01
-4.97601360e-01 3.35713714e-01 1.01530300e-02 -8.07322681e-01
-2.23525479e-01 -4.73971456e-01 -1.25225472e+00 -7.18253314e-01
3.94622348e-02 -1.20344907e-02 3.31640750e-01 4.65969473e-01
4.89251614e-01 2.56805509e-01 6.85056508e-01 -1.18060768e+00
-7.56110549e-02 -6.84225559e-01 -6.33542061e-01 6.97773159e-01
8.37465897e-02 -5.08239985e-01 -1.67664960e-01 4.64629322e-01] | [11.371572494506836, -0.7642155289649963] |
16d66928-eae4-4198-b455-7d5cd0498562 | v2x-seq-a-large-scale-sequential-dataset-for | 2305.05938 | null | https://arxiv.org/abs/2305.05938v1 | https://arxiv.org/pdf/2305.05938v1.pdf | V2X-Seq: A Large-Scale Sequential Dataset for Vehicle-Infrastructure Cooperative Perception and Forecasting | Utilizing infrastructure and vehicle-side information to track and forecast the behaviors of surrounding traffic participants can significantly improve decision-making and safety in autonomous driving. However, the lack of real-world sequential datasets limits research in this area. To address this issue, we introduce V2X-Seq, the first large-scale sequential V2X dataset, which includes data frames, trajectories, vector maps, and traffic lights captured from natural scenery. V2X-Seq comprises two parts: the sequential perception dataset, which includes more than 15,000 frames captured from 95 scenarios, and the trajectory forecasting dataset, which contains about 80,000 infrastructure-view scenarios, 80,000 vehicle-view scenarios, and 50,000 cooperative-view scenarios captured from 28 intersections' areas, covering 672 hours of data. Based on V2X-Seq, we introduce three new tasks for vehicle-infrastructure cooperative (VIC) autonomous driving: VIC3D Tracking, Online-VIC Forecasting, and Offline-VIC Forecasting. We also provide benchmarks for the introduced tasks. Find data, code, and more up-to-date information at \href{https://github.com/AIR-THU/DAIR-V2X-Seq}{https://github.com/AIR-THU/DAIR-V2X-Seq}. | ['Zaiqing Nie', 'Ping Luo', 'Jirui Yuan', 'Juan Song', 'Ning Sun', 'Yifeng Pan', 'Yifeng Shi', 'Xin Hao', 'Xu Gao', 'Yingjuan Tang', 'Zhenwei Yang', 'Hongzhi Ruan', 'Wenxian Yang', 'Haibao Yu'] | 2023-05-10 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Yu_V2X-Seq_A_Large-Scale_Sequential_Dataset_for_Vehicle-Infrastructure_Cooperative_Perception_and_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Yu_V2X-Seq_A_Large-Scale_Sequential_Dataset_for_Vehicle-Infrastructure_Cooperative_Perception_and_CVPR_2023_paper.pdf | cvpr-2023-1 | ['trajectory-forecasting'] | ['computer-vision'] | [-4.18909252e-01 -2.38509208e-01 -3.11652660e-01 -6.87529862e-01
-6.65324926e-01 -5.66401064e-01 6.94883823e-01 -8.36526901e-02
-3.41263972e-02 5.10136485e-01 1.24300972e-01 -6.90095186e-01
-5.07387407e-02 -7.47168779e-01 -7.01318264e-01 -3.82501245e-01
-9.14719254e-02 2.59889483e-01 7.02582717e-01 -4.40607518e-01
1.27552375e-01 4.76200461e-01 -2.04339242e+00 1.48783103e-01
5.95723450e-01 1.17623353e+00 5.78992426e-01 7.82736182e-01
1.70620963e-01 8.21947098e-01 -2.68591531e-02 -2.85785645e-01
3.32597017e-01 1.20980747e-01 -2.34993875e-01 -2.26519540e-01
3.90197277e-01 -4.38730866e-01 -8.86682034e-01 7.71738827e-01
2.83041894e-01 3.24671745e-01 2.52758145e-01 -2.25243664e+00
-1.75898835e-01 -7.67176598e-02 -3.24999392e-01 5.27240455e-01
3.52949739e-01 8.75533998e-01 4.48461562e-01 -8.51851046e-01
7.16494858e-01 1.28151381e+00 5.14653444e-01 4.35441583e-01
-6.54157877e-01 -1.04390240e+00 3.16419244e-01 7.06796587e-01
-1.47042346e+00 -7.07671344e-01 5.55784822e-01 -6.54003441e-01
7.03817606e-01 4.21248078e-01 6.76261187e-01 1.11871600e+00
3.02640557e-01 1.01560104e+00 8.54075432e-01 4.86994743e-01
2.52658352e-02 3.73700559e-02 3.50910723e-01 5.44080317e-01
-7.44064376e-02 7.83036828e-01 -4.61535513e-01 1.94110826e-01
4.15974893e-02 4.07360405e-01 3.19399498e-02 -9.70870405e-02
-1.43525529e+00 5.35193801e-01 1.98733985e-01 -3.43279660e-01
-5.46123385e-01 3.53565305e-01 5.63465774e-01 3.10840189e-01
5.33661246e-01 -3.37645739e-01 -2.85806894e-01 -4.93144363e-01
-5.21018744e-01 5.83779871e-01 1.93943724e-01 1.67286456e+00
1.14578533e+00 9.46401581e-02 -2.67189324e-01 4.54267740e-01
1.46786183e-01 1.22248244e+00 -2.85482526e-01 -1.48532903e+00
7.98951209e-01 3.61041188e-01 2.98160762e-01 -1.11492980e+00
-5.12896359e-01 2.87217975e-01 -6.35517240e-01 9.18937474e-02
1.47074647e-03 -5.18600643e-01 -6.93413973e-01 1.57445121e+00
5.31140447e-01 5.89150608e-01 2.71080546e-02 1.06133211e+00
1.08465183e+00 1.01925731e+00 -8.81112665e-02 -8.68880004e-02
1.22589540e+00 -1.03990114e+00 -6.93660975e-01 -1.47072062e-01
6.17854178e-01 -6.20800734e-01 6.83568656e-01 2.92365998e-02
-7.42230415e-01 -9.66715217e-01 -5.66147149e-01 1.86096102e-01
-6.53654456e-01 1.29252551e-02 1.34116530e-01 6.44070357e-02
-9.81528640e-01 -2.22867236e-01 -5.88689804e-01 -4.50130522e-01
5.44178367e-01 -5.92003129e-02 -3.20116758e-01 -5.24487734e-01
-1.29654336e+00 7.27265835e-01 4.58964892e-02 1.81241371e-02
-1.65043223e+00 -1.25246286e+00 -8.04346502e-01 -2.99229681e-01
7.82237411e-01 -2.40133807e-01 1.16724837e+00 -1.39829785e-01
-8.63933384e-01 4.38808173e-01 -5.92744648e-01 -2.72829562e-01
5.98574519e-01 -8.35955441e-02 -1.06819975e+00 -1.70863152e-01
3.77124131e-01 8.84912312e-01 3.61572921e-01 -1.48421252e+00
-1.23685575e+00 -2.91124821e-01 6.00481555e-02 -4.14204337e-02
3.56318921e-01 -1.58877403e-01 -8.13174367e-01 -7.00610131e-02
-5.51521599e-01 -1.26847541e+00 -2.71895170e-01 -1.29001111e-01
-5.31686246e-01 -4.72698808e-01 1.20876837e+00 -4.08897281e-01
1.09996402e+00 -2.56648088e+00 -5.22087753e-01 1.29120663e-01
4.17395473e-01 2.49874949e-01 -3.95376027e-01 7.71107852e-01
3.64580274e-01 -2.59276748e-01 3.51259440e-01 -2.47504234e-01
2.27788717e-01 3.18353534e-01 -4.65401858e-01 4.71383840e-01
-2.36596763e-01 9.75392520e-01 -1.00781608e+00 -2.99037069e-01
8.26766372e-01 1.66999146e-01 -1.83019131e-01 2.31663898e-01
-2.10244745e-01 4.62883681e-01 -5.95293403e-01 5.93644679e-01
9.64787483e-01 2.26139337e-01 -4.10194218e-01 -8.39068145e-02
-6.82238221e-01 -7.91542903e-02 -1.02890217e+00 1.25249076e+00
-3.91963869e-01 1.26743603e+00 3.41858305e-02 -7.14211881e-01
8.43822598e-01 1.94013894e-01 7.31314659e-01 -1.12055659e+00
4.53998931e-02 6.39753342e-02 -4.88802433e-01 -7.91957080e-01
6.22233450e-01 5.24679363e-01 -2.72767901e-01 1.70137912e-01
-5.45544684e-01 1.36095788e-02 3.84584576e-01 3.98257762e-01
1.28997719e+00 -3.08637857e-01 -1.09281763e-01 -7.57630095e-02
3.51204902e-01 6.56287909e-01 7.38665700e-01 5.70282400e-01
-6.67531431e-01 2.11191922e-01 3.67813498e-01 -8.05380881e-01
-9.83965993e-01 -1.15293980e+00 -5.28195594e-03 8.68017435e-01
8.84235978e-01 -4.03575540e-01 -4.49240446e-01 -4.46794182e-01
3.94578606e-01 1.13303792e+00 -6.83689833e-01 -2.92710345e-02
-5.77004552e-01 8.00907165e-02 3.22648227e-01 3.52573693e-01
6.77329838e-01 -8.79498363e-01 -7.61926293e-01 1.43455222e-01
-5.27426600e-01 -1.49656677e+00 -7.49661386e-01 -5.45267403e-01
-1.51139885e-01 -1.25395310e+00 -2.41747543e-01 -1.43770561e-01
3.75382513e-01 1.22740781e+00 1.02650249e+00 8.83730650e-02
-1.08005069e-01 6.80198073e-01 -2.34266773e-01 -7.87857354e-01
-3.75426203e-01 -2.10157886e-01 3.07488799e-01 1.23218007e-01
7.83758521e-01 -2.82260120e-01 -7.75039077e-01 9.99190807e-01
-2.53319621e-01 3.00715417e-01 1.13630563e-01 -3.25630344e-02
6.63327813e-01 -4.01576310e-02 4.02790070e-01 -3.20856482e-01
1.79261908e-01 -9.82162178e-01 -9.25723255e-01 -6.87412247e-02
-4.29601073e-01 -7.59136438e-01 5.18644750e-01 -2.46645704e-01
-8.12628508e-01 -7.00472742e-02 -3.24467085e-02 -8.85624290e-01
-5.86678505e-01 8.89103636e-02 -4.64499854e-02 4.21314597e-01
3.57517868e-01 3.23029935e-01 2.03922316e-01 -2.85906941e-02
4.30051535e-01 7.61282682e-01 6.39860570e-01 -5.64478748e-02
9.31184590e-01 7.23393440e-01 -2.08307728e-01 -9.82680798e-01
-5.33967614e-01 -8.59494150e-01 -2.84593284e-01 -9.84953642e-01
9.94945586e-01 -1.36969578e+00 -1.18502188e+00 3.92041028e-01
-1.05025089e+00 -7.83859611e-01 -1.07247807e-01 6.24413073e-01
-6.16261423e-01 -1.45533845e-01 -7.97481760e-02 -7.84448802e-01
3.05807918e-01 -1.46437037e+00 1.08937442e+00 2.35848594e-02
1.84794590e-01 -6.74431860e-01 2.25595966e-01 6.00916564e-01
3.68595272e-01 2.10464492e-01 3.02899331e-01 -2.19661623e-01
-9.72448826e-01 -5.96805178e-02 -3.59736532e-01 1.76512841e-02
-1.68659180e-01 7.47024938e-02 -6.84685767e-01 -2.87883937e-01
-7.48391807e-01 4.31075916e-02 7.17763245e-01 4.48709935e-01
1.24025738e+00 -1.76341087e-01 -8.80848467e-01 4.30775613e-01
1.22702789e+00 5.63863575e-01 7.52206922e-01 1.12369493e-01
7.31502771e-01 7.80876279e-01 1.30666506e+00 5.15618086e-01
1.24748278e+00 9.67627287e-01 8.79005611e-01 -4.94850203e-02
-2.81373441e-01 -4.17012423e-01 3.96752477e-01 6.18525028e-01
-8.27187523e-02 -4.97650027e-01 -1.24966407e+00 9.91177917e-01
-1.94685125e+00 -1.42620337e+00 -6.62419319e-01 1.84130728e+00
-1.03326559e-01 1.15509659e-01 4.90281522e-01 -2.42736161e-01
6.57064438e-01 4.05749112e-01 -7.74236739e-01 -1.05606429e-01
3.15174162e-01 -9.05693531e-01 8.42127383e-01 7.00337887e-01
-8.75102162e-01 9.62244630e-01 5.05063438e+00 1.02967548e+00
-9.95980799e-01 4.74210471e-01 3.92413825e-01 -2.81552255e-01
-5.24571359e-01 -1.27718434e-01 -1.09683156e+00 9.93347168e-01
1.20320702e+00 -2.97108829e-01 4.39137191e-01 9.16689813e-01
9.60990369e-01 -1.92756921e-01 -8.05368125e-01 1.07390487e+00
-2.53018498e-01 -1.77936375e+00 -3.34219009e-01 2.88773715e-01
5.90650320e-01 9.56065536e-01 1.55085176e-01 5.00022769e-01
4.98714775e-01 -6.74878776e-01 8.16231489e-01 7.16953039e-01
9.80570614e-01 -7.18993902e-01 5.38545549e-01 4.51358497e-01
-1.76154971e+00 -3.24481666e-01 -2.02214390e-01 1.82174221e-01
7.32186496e-01 5.38749516e-01 -4.22451854e-01 4.98968959e-01
1.02305520e+00 1.13355017e+00 -4.13986325e-01 7.75099993e-01
3.15793902e-01 5.81765592e-01 -1.68605119e-01 -5.16707860e-02
5.03150463e-01 -3.11736912e-01 8.41870844e-01 1.06990635e+00
5.10375142e-01 4.34304953e-01 3.60479444e-01 5.09591043e-01
2.61257023e-01 -4.76888806e-01 -1.32885432e+00 6.16593301e-01
9.48588967e-01 1.21276045e+00 -1.54508099e-01 -5.39088428e-01
-6.60254061e-01 1.31889760e-01 -1.86914444e-01 7.03424215e-01
-1.40709686e+00 -1.13235317e-01 1.48618102e+00 5.50810874e-01
2.51411468e-01 -4.59809810e-01 2.11203203e-01 -6.49702728e-01
-7.44760707e-02 -5.59292734e-01 -2.09336147e-01 -1.11599123e+00
-7.69802451e-01 5.90712368e-01 4.87724215e-01 -1.67390203e+00
-1.38882687e-02 -2.65549481e-01 -7.34974384e-01 3.84028107e-01
-1.71582723e+00 -1.07217491e+00 -8.86516094e-01 8.54621351e-01
9.18545008e-01 -4.61081088e-01 -4.29307036e-02 6.26813054e-01
-6.27228737e-01 2.18440607e-01 1.98304385e-01 -3.44322383e-01
4.99934018e-01 -6.91433430e-01 7.21084774e-01 6.76752508e-01
-3.59790117e-01 -2.07373291e-01 7.20580101e-01 -5.51895261e-01
-1.69291520e+00 -1.84478283e+00 7.54367471e-01 -8.50649774e-01
4.72642303e-01 -4.71728444e-01 -5.34287333e-01 8.05817783e-01
6.25981465e-02 3.13194841e-01 1.44904077e-01 -4.91209865e-01
5.53218573e-02 -6.45615458e-01 -8.36584151e-01 6.80908322e-01
1.42778254e+00 -3.40712726e-01 6.28678128e-02 4.05209303e-01
9.28169012e-01 -4.16113138e-01 -7.01889217e-01 3.46061528e-01
5.38198888e-01 -1.00534606e+00 8.32205653e-01 -9.04531702e-02
9.94021520e-02 -5.40748715e-01 -5.52866042e-01 -1.19662309e+00
-2.71462917e-01 -5.52036583e-01 -1.33068366e-02 1.02989221e+00
3.95505816e-01 -1.01495016e+00 5.19662678e-01 3.22492033e-01
-7.59221971e-01 -6.47953868e-01 -1.11994886e+00 -8.77590656e-01
-3.68086785e-01 -1.13177228e+00 1.16062033e+00 6.02421761e-01
-5.72043777e-01 -1.03263613e-02 -4.64112103e-01 4.29985732e-01
6.89121187e-01 1.79772869e-01 1.46949196e+00 -9.45106030e-01
6.05604291e-01 -1.69081748e-01 -5.58319092e-01 -1.37741137e+00
2.42079824e-01 -6.00527108e-01 1.86154380e-01 -1.49911511e+00
2.17996109e-02 -7.06828713e-01 1.65066838e-01 4.38859820e-01
2.15816587e-01 5.62249050e-02 3.68061900e-01 3.47582936e-01
-9.77229536e-01 7.32075632e-01 1.19797075e+00 -2.59240419e-01
5.99743389e-02 2.41488039e-01 -3.26954842e-01 3.74623209e-01
9.01663661e-01 -3.49950910e-01 -7.56906986e-01 -5.59195518e-01
-1.24767192e-01 4.20984894e-01 6.96691751e-01 -1.05485618e+00
4.72579479e-01 -7.13677049e-01 -2.05339372e-01 -1.52882040e+00
5.62368512e-01 -1.01013935e+00 6.62613809e-01 4.31934655e-01
-2.22053342e-02 5.66980183e-01 4.30189461e-01 6.50379002e-01
-2.75989413e-01 5.62660575e-01 2.27510065e-01 1.32345781e-01
-1.51078105e+00 1.03811586e+00 -7.89517641e-01 2.04490364e-01
1.79113221e+00 -3.58724892e-01 -7.47163177e-01 -4.47532207e-01
-2.92550623e-01 1.20629621e+00 4.17123646e-01 1.02588439e+00
8.03239167e-01 -1.65890145e+00 -8.93946111e-01 2.33883291e-01
5.50907314e-01 -3.28167789e-02 9.45076585e-01 1.10114038e+00
-2.88761467e-01 7.96790421e-01 -2.10242167e-01 -1.00209844e+00
-1.27468228e+00 9.13383186e-01 4.78951540e-03 4.21101093e-01
-8.30291748e-01 1.69486925e-01 3.80310535e-01 -6.70063853e-01
-4.64353710e-02 -1.20173894e-01 -7.65701905e-02 -8.83745924e-02
5.90152323e-01 9.22675848e-01 -2.57630318e-01 -1.26202548e+00
-6.18390918e-01 5.54017842e-01 4.11410064e-01 1.26730487e-01
9.99147475e-01 -7.60147154e-01 4.87115949e-01 4.50796038e-01
1.29582310e+00 -3.06844741e-01 -1.69855499e+00 -1.36745080e-01
-4.07899678e-01 -6.26395881e-01 2.40294784e-02 -3.48994821e-01
-1.30917931e+00 8.40119302e-01 7.97395229e-01 2.08241552e-01
8.01904559e-01 2.28315681e-01 1.18227482e+00 4.75155935e-02
7.44742155e-01 -8.38331223e-01 -2.39103600e-01 5.20559251e-01
9.24911857e-01 -1.56566823e+00 -3.87707055e-01 -3.28495353e-01
-1.00603366e+00 5.16782284e-01 9.20210898e-01 2.30815411e-01
9.96347904e-01 1.07006051e-01 3.59185040e-01 -5.95328748e-01
-1.29325700e+00 -4.88721579e-01 2.81735249e-02 8.10496211e-01
-5.54723918e-01 5.00542581e-01 3.83117616e-01 8.84204209e-02
-2.00174510e-01 7.84691703e-03 4.50427830e-01 5.64923823e-01
-2.17616573e-01 -4.26034689e-01 -1.60242811e-01 4.70565200e-01
3.85861039e-01 4.03099686e-01 2.14899182e-01 8.72918963e-01
4.67027277e-01 1.42120790e+00 3.53356510e-01 -8.00300181e-01
7.01315463e-01 -6.04783893e-01 -2.90339500e-01 1.34861805e-02
2.90018637e-02 -5.59297204e-01 3.60906005e-01 -1.16415000e+00
-2.24541008e-01 -8.77270520e-01 -1.06804621e+00 -1.18752944e+00
2.63907641e-01 3.97290960e-02 6.85094953e-01 7.00616002e-01
8.17546666e-01 4.98134196e-01 1.03312206e+00 -1.11521769e+00
8.00024136e-04 -5.78708112e-01 -2.97420561e-01 1.66425243e-01
7.63348818e-01 -9.17632759e-01 -3.43951076e-01 -1.29711390e-01] | [5.942317962646484, 0.9478492736816406] |
5290cf92-4f03-48e8-b2ba-8e4396daf2a5 | toward-compact-parameter-representations-for | 2111.1032 | null | https://arxiv.org/abs/2111.10320v1 | https://arxiv.org/pdf/2111.10320v1.pdf | Toward Compact Parameter Representations for Architecture-Agnostic Neural Network Compression | This paper investigates deep neural network (DNN) compression from the perspective of compactly representing and storing trained parameters. We explore the previously overlooked opportunity of cross-layer architecture-agnostic representation sharing for DNN parameters. To do this, we decouple feedforward parameters from DNN architectures and leverage additive quantization, an extreme lossy compression method invented for image descriptors, to compactly represent the parameters. The representations are then finetuned on task objectives to improve task accuracy. We conduct extensive experiments on MobileNet-v2, VGG-11, ResNet-50, Feature Pyramid Networks, and pruned DNNs trained for classification, detection, and segmentation tasks. The conceptually simple scheme consistently outperforms iterative unstructured pruning. Applied to ResNet-50 with 76.1% top-1 accuracy on the ILSVRC12 classification challenge, it achieves a $7.2\times$ compression ratio with no accuracy loss and a $15.3\times$ compression ratio at 74.79% accuracy. Further analyses suggest that representation sharing can frequently happen across network layers and that learning shared representations for an entire DNN can achieve better accuracy at the same compression ratio than compressing the model as multiple separate parts. We release PyTorch code to facilitate DNN deployment on resource-constrained devices and spur future research on efficient representations and storage of DNN parameters. | ['Matei Zaharia', 'Hui Guan', 'Xiao Liu', 'Lijun Zhang', 'Wenlong Zhao', 'Yuezhou Sun'] | 2021-11-19 | null | null | null | null | ['neural-network-compression', 'neural-network-compression'] | ['methodology', 'miscellaneous'] | [ 3.42054963e-01 1.10592775e-01 -2.57691324e-01 -5.73987186e-01
-5.81027091e-01 -3.28947425e-01 1.97581843e-01 -2.13455036e-01
-1.04155815e+00 6.60005212e-01 -1.18062412e-02 -5.09828627e-01
-2.84060121e-01 -6.68426216e-01 -6.73626959e-01 -4.53177392e-01
-1.12077683e-01 3.39588523e-01 2.10007951e-02 5.78140244e-02
-3.65805179e-02 7.73233056e-01 -1.71792090e+00 3.34539086e-01
3.64076614e-01 1.47418416e+00 4.52829391e-01 8.63074303e-01
7.48044625e-02 9.14249361e-01 -8.32454026e-01 -7.68657088e-01
6.10605061e-01 6.06342070e-02 -7.16384828e-01 -4.04495269e-01
1.00283301e+00 -8.93151522e-01 -8.18896711e-01 1.08649802e+00
6.77165985e-01 1.78087339e-01 4.55883861e-01 -9.54209328e-01
-5.69771230e-01 8.91036630e-01 -1.65951014e-01 5.76540470e-01
-5.93108356e-01 2.68971592e-01 1.07423639e+00 -4.50370312e-01
5.64043283e-01 1.21786141e+00 8.07015121e-01 7.66905487e-01
-1.24731493e+00 -1.00736392e+00 1.76563218e-01 2.36670896e-02
-1.55674422e+00 -9.32337701e-01 1.42999470e-01 -3.78926806e-02
1.62573421e+00 9.86846015e-02 5.77456236e-01 1.04329503e+00
-1.71503536e-02 6.35226190e-01 3.15487444e-01 1.01617835e-02
2.12398082e-01 -2.01477870e-01 1.59743190e-01 7.27187514e-01
6.89758718e-01 -3.17892730e-02 -4.58505720e-01 5.08849211e-02
8.87570381e-01 2.68473208e-01 -1.34621844e-01 3.66278701e-02
-4.79901910e-01 7.86914170e-01 7.46090531e-01 7.06856698e-02
-4.05371308e-01 1.00999498e+00 7.03672647e-01 5.23065507e-01
2.72587746e-01 2.93589056e-01 -8.54026794e-01 -4.55842346e-01
-1.13192880e+00 4.46153194e-01 6.63858116e-01 1.11964893e+00
6.68759048e-01 5.77711582e-01 -6.72865659e-02 1.12382340e+00
1.11358330e-01 2.90360421e-01 8.36114883e-01 -1.25729764e+00
7.77101040e-01 3.43175977e-01 -6.37317836e-01 -6.83498144e-01
-2.08480999e-01 -6.80746615e-01 -9.73879397e-01 3.92373577e-02
1.06971428e-01 -3.65720302e-01 -1.22842515e+00 1.90552771e+00
-4.19860750e-01 -8.75940397e-02 2.41903916e-01 5.05829871e-01
7.52095819e-01 4.56952423e-01 3.11769664e-01 4.01542306e-01
1.29058707e+00 -7.23931730e-01 -3.99697930e-01 -4.66933489e-01
1.02607858e+00 -1.95467502e-01 6.61396205e-01 3.08270961e-01
-1.39781964e+00 -5.17901838e-01 -1.18146622e+00 -5.67591846e-01
-4.07595992e-01 3.71961176e-01 5.98116219e-01 8.16125751e-01
-1.44283414e+00 1.09909356e+00 -9.19044495e-01 5.59236817e-02
1.29867733e+00 8.63982797e-01 -9.76847112e-02 -3.08627397e-01
-8.69861603e-01 8.37207735e-01 6.60329342e-01 -2.87673950e-01
-8.95707369e-01 -1.10414565e+00 -6.01981223e-01 4.74876553e-01
-1.63432613e-01 -7.00489342e-01 1.41498327e+00 -8.13468754e-01
-1.25991857e+00 7.68905222e-01 2.16231748e-01 -1.39986002e+00
2.18683422e-01 -3.50005448e-01 -3.48773688e-01 3.27729255e-01
-3.77795756e-01 1.36480355e+00 7.23826945e-01 -4.86121893e-01
-6.10810220e-01 -2.60799736e-01 -4.54144068e-02 6.75091073e-02
-6.67981982e-01 -2.91076422e-01 -2.96181500e-01 -7.99682617e-01
1.52501583e-01 -6.71495080e-01 -3.21090758e-01 4.78660762e-01
-1.28853202e-01 -5.26806973e-02 8.84305298e-01 -7.09380031e-01
1.03878963e+00 -2.22783828e+00 -2.26231858e-01 1.37208834e-01
5.85939348e-01 7.82910943e-01 -3.21289897e-01 -2.96534777e-01
6.70811608e-02 4.06436563e-01 -1.36148348e-01 -6.86679602e-01
1.38454875e-02 5.19947946e-01 -3.97421390e-01 1.93315685e-01
1.23952679e-01 1.03615820e+00 -4.02327001e-01 -1.94827631e-01
-9.10911039e-02 7.28860378e-01 -9.27653849e-01 -3.88860792e-01
-2.51235753e-01 -6.59923553e-01 -1.23948410e-01 6.16140842e-01
6.39641047e-01 -1.54545933e-01 2.51675606e-01 -4.11401510e-01
3.10958982e-01 5.76391101e-01 -6.63172960e-01 1.68397248e+00
-4.18547720e-01 1.03598094e+00 1.46659955e-01 -9.79166865e-01
1.03408945e+00 2.03898206e-01 1.81850463e-01 -9.28343713e-01
3.66344571e-01 3.47053230e-01 1.54373929e-01 2.23161891e-01
8.16843271e-01 2.25213394e-01 1.23854212e-01 2.92296141e-01
6.66266918e-01 1.72924280e-01 2.45883837e-02 1.32634610e-01
1.44189382e+00 -4.70300764e-01 6.41884804e-02 -1.38705999e-01
-2.25329489e-01 -4.03384656e-01 4.09981132e-01 8.38473320e-01
-3.01713139e-01 4.77915853e-01 3.88497591e-01 -4.70030695e-01
-1.27206779e+00 -9.58191693e-01 -3.48693013e-01 1.39177072e+00
-5.21238089e-01 -5.34053564e-01 -6.46412849e-01 -4.22676951e-01
2.30452403e-01 5.87838948e-01 -3.55567843e-01 -4.35699701e-01
-5.47643423e-01 -5.77383220e-01 1.24093616e+00 8.47325146e-01
9.47854519e-01 -8.22425663e-01 -9.40809667e-01 1.06067844e-01
2.93424219e-01 -1.20790052e+00 -7.71199167e-02 7.00875401e-01
-1.46072185e+00 -6.67373598e-01 -8.65479052e-01 -8.13406289e-01
4.91422802e-01 4.05232497e-02 1.06150055e+00 1.07837684e-01
-3.41928810e-01 2.10768089e-01 -1.46842366e-02 -3.36806029e-01
-1.28013223e-01 4.70774680e-01 -2.68252790e-02 -5.75907469e-01
3.88600320e-01 -8.29082191e-01 -5.75212955e-01 -9.07489434e-02
-7.58285701e-01 -8.18613172e-02 7.45074451e-01 5.04381835e-01
7.76329041e-01 -1.35084212e-01 5.19004762e-01 -8.90325963e-01
6.12429500e-01 -3.86298597e-01 -6.10049188e-01 -9.93840955e-03
-5.97336888e-01 2.75793046e-01 6.65146530e-01 -5.54040790e-01
-5.98610342e-01 -9.59053810e-04 -2.95484513e-01 -1.05754483e+00
2.89960168e-02 1.92839414e-01 2.25235839e-04 -2.16437042e-01
7.74478078e-01 1.47068664e-01 9.54868365e-03 -7.24825323e-01
4.27903682e-01 5.38208365e-01 7.18389094e-01 -3.43492568e-01
3.45802456e-01 1.21231973e-01 -1.21744066e-01 -7.06358969e-01
-4.85138983e-01 1.64627342e-03 -1.63988456e-01 4.93010581e-01
6.18465662e-01 -1.21849906e+00 -5.15438080e-01 2.34861925e-01
-1.15914512e+00 -6.38344109e-01 -6.75228059e-01 3.48034948e-01
-4.29988980e-01 -1.59782931e-01 -9.18673933e-01 -3.31368059e-01
-7.82530606e-01 -1.01753366e+00 5.23103237e-01 1.60751212e-02
-1.05637185e-01 -7.00358808e-01 -3.15643460e-01 6.60392568e-02
8.18384469e-01 -2.39454255e-01 9.33390379e-01 -9.30798590e-01
-6.11045003e-01 -1.53001860e-01 -6.57136500e-01 8.23100328e-01
-2.89745629e-01 -2.67361045e-01 -1.21722043e+00 -3.77487957e-01
-2.60729760e-01 -6.43637300e-01 1.42465222e+00 4.79611993e-01
1.88751245e+00 -6.48227632e-01 -2.11945385e-01 1.30089986e+00
1.46028817e+00 1.85208455e-01 7.44203568e-01 7.20312968e-02
7.01045811e-01 -6.40261248e-02 -3.21898460e-01 5.81688166e-01
7.69006163e-02 3.43030363e-01 6.44269228e-01 3.48832846e-01
-5.88561058e-01 -1.82698339e-01 1.19594477e-01 7.64808297e-01
-3.49957384e-02 -4.02720094e-01 -6.38117194e-01 4.89406884e-01
-1.17936313e+00 -8.18389237e-01 7.16966510e-01 1.82276428e+00
8.31407666e-01 3.66923988e-01 -2.58493960e-01 9.55674797e-02
4.57310319e-01 3.41832429e-01 -9.32597935e-01 -7.71862984e-01
-2.99141735e-01 6.17240071e-01 1.17112601e+00 1.38436183e-01
-1.00135612e+00 9.06191766e-01 6.28594971e+00 1.05937195e+00
-1.00689542e+00 1.32703677e-01 9.54939723e-01 -6.54706597e-01
-1.05335452e-01 -4.32793945e-01 -1.30102015e+00 3.77067745e-01
1.58267355e+00 -6.20642640e-02 6.52407587e-01 1.04072356e+00
-4.97319251e-01 4.15857375e-01 -1.12822914e+00 1.47074568e+00
-2.54578054e-01 -1.80004311e+00 5.58257401e-01 2.87542045e-01
6.66609466e-01 7.44698703e-01 4.24222261e-01 4.80991393e-01
4.37913120e-01 -1.34007919e+00 7.23939061e-01 3.94939661e-01
1.20155215e+00 -9.49518204e-01 6.18731856e-01 7.55686536e-02
-1.10221279e+00 -3.75962198e-01 -8.57165396e-01 -2.46568844e-02
-1.70782730e-01 3.21609050e-01 -7.64392972e-01 -1.34231314e-01
8.81677806e-01 8.31588387e-01 -6.82773054e-01 9.51838076e-01
2.97095239e-01 6.32065177e-01 -5.91525435e-01 1.32754579e-01
3.01614434e-01 2.53684044e-01 9.74518657e-02 1.27308416e+00
3.08447540e-01 9.09827724e-02 -6.00214899e-01 8.89509380e-01
-9.11938012e-01 -3.15881938e-01 -7.19788909e-01 -2.90326089e-01
8.34404111e-01 8.03366423e-01 -5.00754774e-01 -4.79059160e-01
-8.64279866e-02 9.05869603e-01 6.29226208e-01 4.28925067e-01
-6.91813707e-01 -6.34131551e-01 1.14319849e+00 -1.54566050e-01
8.82933259e-01 -1.88867629e-01 -4.71892476e-01 -7.32816935e-01
-1.07084528e-01 -7.07856596e-01 1.58365697e-01 -5.05819917e-01
-8.85834634e-01 7.81454802e-01 -4.20917161e-02 -8.68278742e-01
-2.88829595e-01 -9.25028801e-01 -1.80140451e-01 6.82666004e-01
-1.31694102e+00 -5.24584293e-01 -1.27768114e-01 5.43038547e-01
4.55791861e-01 -5.45653105e-01 8.48802984e-01 6.63844109e-01
-6.70530438e-01 1.29114473e+00 2.90613681e-01 3.22307438e-01
1.90969273e-01 -6.35437012e-01 7.04015315e-01 5.43533802e-01
2.23470926e-01 5.62194228e-01 1.15194522e-01 -3.52114260e-01
-1.13416958e+00 -1.54361820e+00 6.60462976e-01 1.22842200e-01
3.26627105e-01 -4.01570618e-01 -8.74245405e-01 1.01229763e+00
-1.37310445e-01 3.18267018e-01 5.76449215e-01 -1.83248356e-01
-5.97526908e-01 -3.82172912e-01 -1.41868877e+00 6.53405488e-01
1.42381513e+00 -6.76720917e-01 -2.52250105e-01 2.25475863e-01
1.16813183e+00 -4.12697911e-01 -8.79110813e-01 1.54887646e-01
4.46972698e-01 -6.39602482e-01 1.18702936e+00 -6.12316310e-01
3.11774790e-01 4.88112748e-01 -8.67347479e-01 -8.82272303e-01
-1.50191903e-01 -4.49601203e-01 -5.02794206e-01 1.02671397e+00
3.55378091e-01 -7.51907170e-01 1.21510494e+00 7.70667434e-01
-3.28236341e-01 -9.68676209e-01 -1.08900356e+00 -8.23795497e-01
1.22258514e-01 -7.13303685e-01 7.98285425e-01 4.55389500e-01
-8.33642423e-01 -2.06776801e-02 3.89094539e-02 -2.18557402e-01
4.99844223e-01 -6.64772630e-01 2.99895406e-01 -1.16973078e+00
-1.66363418e-01 -9.29079592e-01 -7.59711504e-01 -1.43768334e+00
1.68246359e-01 -1.12218618e+00 -4.60772842e-01 -1.15752089e+00
-1.83580071e-01 -7.52952039e-01 -4.75799114e-01 7.48344302e-01
6.89808846e-01 4.72768396e-01 4.14503396e-01 2.55794108e-01
-6.57537162e-01 6.20589375e-01 7.81766176e-01 -1.34870529e-01
-5.13851792e-02 -2.11161196e-01 -8.98938239e-01 7.20156372e-01
1.07362938e+00 -4.14074838e-01 -7.23476946e-01 -1.10089028e+00
-8.11483264e-02 -3.16060483e-01 6.44203007e-01 -1.49721324e+00
3.89309555e-01 4.39796627e-01 5.79653084e-01 -5.95290720e-01
7.13494360e-01 -6.85885251e-01 -1.27434403e-01 5.74222147e-01
-6.31807804e-01 3.02101254e-01 6.11354768e-01 4.96045619e-01
3.41615230e-02 -3.78397644e-01 8.09366584e-01 -1.44917890e-01
-7.65368640e-01 6.13490641e-01 -2.23594412e-01 2.67043591e-01
5.53836226e-01 -5.61220288e-01 -5.31282485e-01 -1.05031610e-01
-5.80171406e-01 -7.41769746e-02 1.45109147e-01 6.72996566e-02
9.49424207e-01 -1.11360717e+00 -3.87986332e-01 4.02316570e-01
-3.68500203e-01 2.48613492e-01 2.76413828e-01 1.94204971e-01
-9.03444469e-01 6.14904106e-01 -3.97651434e-01 -3.07809114e-01
-1.07213175e+00 -7.72419060e-03 6.88087702e-01 -2.12844759e-01
-7.62619436e-01 1.31477535e+00 -6.35901392e-02 -1.27638012e-01
6.82079196e-01 -7.35472083e-01 1.84587777e-01 -5.43920733e-02
5.83069205e-01 6.12200379e-01 3.59135330e-01 -3.47969621e-01
-2.12267846e-01 2.10684657e-01 -5.25498688e-01 7.23730177e-02
1.36093652e+00 1.08025730e-01 3.24711353e-01 -1.86243579e-02
1.84564888e+00 -8.53250206e-01 -1.62329030e+00 -4.07013625e-01
-2.42511213e-01 -6.03049360e-02 4.85715419e-01 -6.59750640e-01
-1.74307013e+00 9.17145073e-01 9.28667724e-01 -3.40391576e-01
1.19166231e+00 -1.21905163e-01 1.03926897e+00 9.35945034e-01
1.39746487e-01 -1.01704741e+00 -5.93638644e-02 8.11713159e-01
6.94416225e-01 -7.58400202e-01 2.70941794e-01 2.50315905e-01
-4.12699938e-01 9.87009406e-01 6.21897042e-01 -4.38595235e-01
7.88562655e-01 4.14924771e-01 -3.78839940e-01 -1.10885523e-01
-1.12153792e+00 1.86695144e-01 -3.02640069e-02 5.70066750e-01
1.18987970e-01 1.71517767e-02 4.16303992e-01 6.55555129e-01
-6.20878577e-01 2.75697429e-02 4.59958725e-02 8.77278030e-01
-5.53452015e-01 -6.94187343e-01 2.97353834e-01 1.19407129e+00
-6.29185319e-01 -4.53537226e-01 1.11714296e-01 7.73866951e-01
2.14590624e-01 3.33992451e-01 6.65864646e-01 -6.49220169e-01
1.76066726e-01 9.33571607e-02 4.34947819e-01 -4.82016444e-01
-7.24272013e-01 -5.57994604e-01 2.17753779e-02 -7.20800579e-01
1.06784843e-01 -3.18076104e-01 -1.33350313e+00 -6.67922080e-01
3.01126763e-02 -3.86400789e-01 9.17075515e-01 5.63341737e-01
7.99496412e-01 7.33293176e-01 -4.99466658e-02 -9.41083014e-01
-9.84227359e-01 -7.35954106e-01 -6.20038331e-01 5.95846362e-02
4.85871136e-01 -4.14533645e-01 -3.23456466e-01 -2.18777224e-01] | [8.592135429382324, 3.113121271133423] |
d3c2a6ff-7cb2-472e-b684-2848c7fd3182 | a-coarse-to-fine-approach-for-dynamic-to | null | null | https://www.sciencedirect.com/science/article/pii/S0031320321005537 | https://www.sciencedirect.com/science/article/pii/S0031320321005537/pdf | A coarse-to-fine approach for dynamic-to-static image translation | Dynamic-to-static image translation aims to convert the dynamic scene into static so that dynamic elements are eliminated from the image. Recent works typically see the problem as an image-to-image translation task, and perform the learned feature mapping over the whole dynamic image to synthesize the static image, which leads to unnecessary detail loss in original static regions. To that end, we delicately formulate it as an image inpainting-like problem to fill the missing static pixels in dynamic regions while retaining original static regions. We achieve this by proposing a coarse-to-fine framework. At coarse stage, we utilize a simple encoder-decoder network to rough out the static image. Using the coarse predicted image, we explicitly infer a more accurate dynamic mask to identify both dynamic objects and their shadows, so that the task could be effectively converted to an image inpainting problem. At fine stage, we recover the missing static pixels in the estimated dynamic regions on the basis of their coarse predictions. We enhance the coarse predicted contents by proposing a mutual texture-structure attention module, which enables the dynamic regions to borrow textures and structures separately from distant locations based on contextual similarity. Several losses are combined as the training objective function to generate excellent results with global consistency and fine details. Qualitative and quantitative experiments verify the superiority of our method in restoring high-quality static contents over state-of-the-art models. In addition, we evaluate the usefulness of the recovered static images by using them as query images to improve visual place recognition in dynamic scenes. | ['Changyin Sun', 'Lin Wu', 'Teng Wang'] | 2021-10-25 | null | null | null | pattern-recognition-2021-10 | ['image-to-image-translation', 'image-inpainting', 'visual-place-recognition', 'image-to-image-translation'] | ['computer-vision', 'computer-vision', 'computer-vision', 'miscellaneous'] | [ 6.24002695e-01 1.13850988e-01 -1.67054400e-01 -3.97669017e-01
-7.97806561e-01 -4.07218665e-01 5.01500249e-01 -5.74485779e-01
-5.35703972e-02 6.89591765e-01 4.23383683e-01 1.79057837e-01
3.08169693e-01 -8.68742585e-01 -1.19780231e+00 -9.63683367e-01
5.61765313e-01 1.30000249e-01 4.20413315e-01 -3.12004417e-01
1.65503189e-01 5.07570684e-01 -1.50642788e+00 7.02768147e-01
9.57541466e-01 9.41303909e-01 7.81177044e-01 4.32807058e-01
-1.68112516e-01 8.72672856e-01 -4.07481581e-01 -1.81616217e-01
4.65667874e-01 -5.47650814e-01 -4.88191485e-01 4.90550458e-01
5.51714361e-01 -6.68402016e-01 -4.52469766e-01 1.11543941e+00
1.73726156e-01 1.58156186e-01 2.95033723e-01 -6.94407821e-01
-9.63291407e-01 3.00183535e-01 -1.05047154e+00 -7.88113251e-02
3.84557188e-01 2.74496913e-01 7.14787781e-01 -9.57587898e-01
7.37672150e-01 1.35311878e+00 3.46524328e-01 4.43421096e-01
-1.43644559e+00 -6.17350221e-01 3.09624046e-01 1.66661888e-01
-1.21284211e+00 -7.24612951e-01 1.08236921e+00 -1.69221207e-01
5.41536868e-01 3.54593903e-01 4.75587875e-01 9.90266681e-01
3.17547023e-01 8.01453829e-01 1.09708118e+00 -4.46845591e-01
-9.21778604e-02 4.84507978e-02 -4.53989178e-01 7.33520329e-01
-1.19361952e-01 3.38400334e-01 -3.74410659e-01 4.01437193e-01
1.03500605e+00 3.15813690e-01 -6.54299855e-01 -3.29520851e-01
-1.12035012e+00 3.48271877e-01 6.45461917e-01 2.84167588e-01
-4.11866277e-01 1.23655081e-01 6.57981411e-02 8.32925439e-02
6.98525131e-01 2.15551779e-01 -4.01465654e-01 3.34531009e-01
-9.41985667e-01 1.97826922e-01 1.80127636e-01 8.18261623e-01
1.14893329e+00 2.10471928e-01 -2.77745366e-01 8.96306038e-01
4.47458848e-02 5.93746245e-01 2.94352800e-01 -1.23543680e+00
5.88731885e-01 4.02481467e-01 4.03975219e-01 -1.15096939e+00
2.68938631e-01 -5.04031599e-01 -9.50671732e-01 1.87310234e-01
3.58403027e-02 2.58082598e-01 -1.02648163e+00 1.78775275e+00
4.13411200e-01 2.68253893e-01 -2.37045929e-01 1.02041376e+00
3.17783535e-01 1.00595295e+00 -2.20075607e-01 -2.64930874e-01
1.08544660e+00 -1.35292840e+00 -7.89655626e-01 -5.42694271e-01
-8.13807920e-03 -8.11719120e-01 1.24951708e+00 6.91208467e-02
-1.26730967e+00 -8.51389647e-01 -7.85924315e-01 -3.94859672e-01
6.93887472e-02 1.30138755e-01 3.08258802e-01 1.99139342e-01
-1.14043105e+00 4.88164961e-01 -6.40058398e-01 8.57095271e-02
3.73117894e-01 -4.64818180e-02 -3.57500970e-01 -3.72531474e-01
-1.00393748e+00 6.41452014e-01 3.59666705e-01 2.45969012e-01
-1.09469891e+00 -8.18652987e-01 -1.01769125e+00 1.67170286e-01
5.10305464e-01 -8.66558671e-01 9.11360979e-01 -1.51989090e+00
-1.46175897e+00 8.29655707e-01 -6.16544366e-01 -1.60983965e-01
5.60188353e-01 -7.67545551e-02 -2.84149408e-01 2.56375581e-01
5.52005291e-01 7.18933582e-01 1.11933815e+00 -1.78652930e+00
-7.26365209e-01 -1.90754235e-01 5.31399064e-02 4.05936539e-01
-5.20661920e-02 -2.66056091e-01 -9.93290901e-01 -1.08624935e+00
1.90384328e-01 -7.10347176e-01 1.81060769e-02 3.53710532e-01
-3.56116772e-01 5.73381186e-01 9.53369915e-01 -1.06291151e+00
1.06886315e+00 -2.13897729e+00 3.84634376e-01 -1.01723365e-01
2.54698306e-01 -3.36044356e-02 -3.49867254e-01 6.54093549e-02
-9.21165720e-02 -2.86491126e-01 -4.50008661e-01 -5.46741247e-01
-4.29121137e-01 3.10385048e-01 -7.82732368e-01 2.29640484e-01
1.80136397e-01 1.10756564e+00 -8.91315043e-01 -3.09301436e-01
4.71328586e-01 5.76119244e-01 -6.97363496e-01 3.64418358e-01
-3.61074686e-01 8.04111540e-01 -4.98634905e-01 7.62684464e-01
9.23394859e-01 -1.41302586e-01 -5.73867187e-03 -5.01659334e-01
6.20708195e-03 -9.80267823e-02 -8.73082161e-01 1.79735756e+00
-8.04001331e-01 6.00350022e-01 2.91051954e-01 -8.96632314e-01
7.17275560e-01 -1.77769899e-01 5.31224549e-01 -1.08098352e+00
-1.42306522e-01 2.08663195e-01 -3.16682428e-01 -4.79438752e-01
5.25796533e-01 -2.79968441e-01 1.12865768e-01 1.70243666e-01
-3.54090333e-01 -1.04037508e-01 -2.22164705e-01 -2.07735077e-02
6.14283621e-01 5.63952863e-01 -1.42788142e-01 8.57258141e-02
6.02961659e-01 -1.94253892e-01 6.37454629e-01 5.15790105e-01
1.22854389e-01 1.07886457e+00 1.70800716e-01 -4.59469110e-01
-1.24918270e+00 -1.18790352e+00 1.30562156e-01 8.84415388e-01
5.81261158e-01 -7.71384537e-02 -8.50636184e-01 -4.74946886e-01
-2.91542113e-01 5.90291858e-01 -8.06376219e-01 -3.25066537e-01
-7.66327024e-01 -3.51078272e-01 -3.02898251e-02 3.47341508e-01
9.21601951e-01 -1.10850060e+00 -3.18195194e-01 3.50469947e-01
-7.58625984e-01 -1.09875786e+00 -1.00703263e+00 -4.22934145e-02
-6.57011271e-01 -7.43963420e-01 -9.73136306e-01 -9.09680128e-01
7.78273404e-01 6.69945598e-01 9.95073557e-01 4.79131825e-02
-1.57068491e-01 -2.68166989e-01 -2.44077370e-01 2.56943345e-01
-4.26929057e-01 -3.89665931e-01 -2.84905344e-01 4.88462716e-01
-3.77137572e-01 -7.71415830e-01 -9.22122478e-01 4.97102946e-01
-1.24656057e+00 7.24299252e-01 7.62632430e-01 1.05481768e+00
8.77852440e-01 3.55954856e-01 1.55664563e-01 -8.78433466e-01
-6.37650192e-02 -1.58521444e-01 -4.49038565e-01 3.19451392e-01
-4.81087089e-01 1.78582639e-01 8.76103938e-01 -5.12337565e-01
-1.55571651e+00 1.32577762e-01 -6.61399588e-02 -8.88068557e-01
1.27014443e-01 -3.18564735e-02 -5.99456370e-01 -1.13501638e-01
3.09665620e-01 9.50964093e-01 -7.36999512e-02 -5.94896138e-01
4.23578024e-01 3.24023038e-01 7.24135995e-01 -6.32208645e-01
8.59520555e-01 8.59234929e-01 -3.50400299e-01 -3.03489864e-01
-1.25733209e+00 -7.00227916e-02 -3.71359408e-01 -1.40892401e-01
7.64017940e-01 -1.08738852e+00 -2.53003210e-01 5.29386997e-01
-1.05176616e+00 -7.07328439e-01 -4.70893234e-01 -9.23848152e-02
-6.72697961e-01 3.89811486e-01 -6.99976087e-01 -2.90158600e-01
-2.05325782e-01 -1.33602703e+00 1.64800763e+00 1.67629406e-01
3.42150122e-01 -6.63009882e-01 -2.12075099e-01 5.22694528e-01
4.52227443e-01 1.50157034e-01 8.69456053e-01 4.80442375e-01
-1.09493577e+00 1.95018142e-01 -5.39387226e-01 4.41845953e-01
2.89635569e-01 -2.82694489e-01 -9.33015406e-01 -2.49746680e-01
1.67414576e-01 8.79567862e-02 1.17857206e+00 4.73352671e-01
1.46228814e+00 -5.84470212e-01 -3.14560533e-01 9.79233027e-01
1.58491004e+00 2.11880952e-01 1.00556540e+00 2.57822335e-01
1.03177881e+00 6.29193962e-01 7.92799354e-01 1.28733262e-01
3.38270158e-01 8.81942630e-01 2.08200619e-01 -3.03937495e-01
-7.28961647e-01 -6.28953695e-01 5.15386403e-01 4.29251105e-01
1.59127519e-01 -2.60175288e-01 -3.83570343e-01 5.37525773e-01
-1.64124131e+00 -1.07651651e+00 2.63326108e-01 2.07170939e+00
1.02961946e+00 1.02240657e-02 -3.67640644e-01 -2.76596695e-01
7.85738647e-01 5.15779853e-01 -5.44469357e-01 1.09795138e-01
-3.99871975e-01 -6.59162700e-02 3.37725788e-01 8.17035854e-01
-8.02123547e-01 1.19469953e+00 5.35073948e+00 1.33997989e+00
-1.31836963e+00 2.08394289e-01 1.23131061e+00 -1.17091380e-01
-6.71008945e-01 2.45348811e-01 -6.32729948e-01 8.40681553e-01
4.81467336e-01 1.45971492e-01 7.49180019e-01 5.25099397e-01
5.94342351e-01 -3.30442399e-01 -8.31168473e-01 9.41483855e-01
2.07898781e-01 -1.52123702e+00 2.90316284e-01 -1.78066213e-02
1.19077313e+00 -3.34435463e-01 1.65107176e-01 1.52007371e-01
-1.01891443e-01 -7.59575486e-01 1.07350588e+00 8.76308501e-01
1.16732800e+00 -5.69828272e-01 3.06656569e-01 2.44059682e-01
-1.32094848e+00 -1.58581093e-01 -4.41681713e-01 2.70962000e-01
4.86488670e-01 8.32196653e-01 -1.05678178e-01 6.00360572e-01
7.24783242e-01 8.83548200e-01 -4.59411711e-01 6.22129858e-01
-1.61483049e-01 1.77818164e-01 -1.13135017e-01 8.18532348e-01
-3.63726052e-03 -4.64669049e-01 4.34838742e-01 8.04811537e-01
3.34412366e-01 6.17492246e-03 1.09416001e-01 1.24278426e+00
-1.74358010e-01 -2.05381081e-01 -5.15611529e-01 4.08136874e-01
3.13080192e-01 1.06525195e+00 -5.90231121e-01 -5.88418245e-01
-2.49453306e-01 1.70996690e+00 2.39917397e-01 5.18984973e-01
-9.75905120e-01 -1.23045556e-01 4.70382601e-01 4.65707928e-01
4.26295251e-01 1.46510199e-01 -3.35906923e-01 -1.51613283e+00
3.58361542e-01 -8.81121099e-01 -1.38846159e-01 -1.22963297e+00
-9.97549534e-01 6.64691329e-01 -2.03161538e-01 -1.51914334e+00
-1.40920147e-01 2.04521464e-03 -4.81195927e-01 9.11234438e-01
-1.60938644e+00 -1.38979781e+00 -4.97535586e-01 6.18327856e-01
8.68722141e-01 2.04985723e-01 3.36105436e-01 5.84902704e-01
-4.10610676e-01 5.17728090e-01 3.18338245e-01 -1.05038688e-01
8.60221207e-01 -7.55376756e-01 2.49986514e-01 1.06775033e+00
-7.47775733e-02 3.44611198e-01 5.92784464e-01 -8.30087721e-01
-1.16742766e+00 -1.43562627e+00 5.84243357e-01 -3.01814973e-01
2.18571663e-01 -3.72438490e-01 -9.93756831e-01 4.85628694e-01
-1.25240102e-01 2.68644899e-01 -3.22819740e-01 -6.82457387e-01
-3.56727391e-01 -3.68561298e-01 -1.00976467e+00 7.62894571e-01
1.10941100e+00 -6.45207465e-01 -2.16098160e-01 1.53995842e-01
1.06077492e+00 -5.66209137e-01 -4.67800558e-01 3.39754462e-01
5.17520964e-01 -1.08602321e+00 1.24109590e+00 5.90147190e-02
9.29883838e-01 -6.01845920e-01 -2.41609365e-01 -1.12320209e+00
-2.97416806e-01 -5.51251113e-01 -9.62411091e-02 1.27501357e+00
4.27133366e-02 -4.78011012e-01 7.49501228e-01 4.47436690e-01
-3.21602702e-01 -7.48287082e-01 -6.66168213e-01 -4.51343834e-01
-2.90901333e-01 -3.48066464e-02 4.92360145e-01 8.71057212e-01
-7.99907863e-01 1.56032532e-01 -7.49708772e-01 1.16427168e-01
6.40097737e-01 7.03026950e-01 6.50653124e-01 -5.31444371e-01
-6.31432176e-01 -1.81575507e-01 4.53820340e-02 -1.49122548e+00
2.77456343e-01 -5.29905617e-01 2.34601215e-01 -1.26192486e+00
4.73582000e-01 -3.97745281e-01 -1.23552168e-02 4.41312075e-01
-3.96110117e-01 6.92837954e-01 1.28196672e-01 5.09895921e-01
-3.36946934e-01 9.60612178e-01 1.91302168e+00 -2.15986922e-01
-1.30985275e-01 -1.93244338e-01 -8.78597796e-01 4.74700332e-01
4.12764370e-01 -3.30196232e-01 -4.20517713e-01 -6.14190280e-01
-3.01555753e-01 4.27744448e-01 5.33747911e-01 -7.15112269e-01
-1.40366346e-01 -4.27735597e-01 7.61501849e-01 -4.74942684e-01
5.47676563e-01 -8.10045302e-01 4.01393801e-01 2.89720416e-01
-2.59785861e-01 -2.77039856e-01 3.23114470e-02 7.99575388e-01
-4.69059229e-01 1.88790351e-01 1.04123354e+00 -9.50819328e-02
-8.55052769e-01 5.01686454e-01 1.71418250e-01 -1.38598889e-01
1.01297498e+00 -4.92072493e-01 -1.80323482e-01 -4.75417405e-01
-6.67712569e-01 -3.43909524e-02 9.39237237e-01 5.08121908e-01
7.28724778e-01 -1.26583421e+00 -5.66457629e-01 5.93635857e-01
-4.97108288e-02 1.75662830e-01 8.22116017e-01 8.11296165e-01
-6.10256433e-01 1.69882327e-01 -2.82673895e-01 -5.30040860e-01
-9.83295321e-01 8.02532554e-01 5.10268271e-01 -1.48647740e-01
-9.40665126e-01 6.11053944e-01 1.09006774e+00 -8.93943384e-02
5.20855673e-02 -2.21483737e-01 2.43542910e-01 -2.82158107e-01
6.58524752e-01 -1.92768559e-01 -1.25677273e-01 -6.86711609e-01
1.42855663e-02 7.95612633e-01 -1.72133312e-01 -1.42357066e-01
1.39405954e+00 -7.69003570e-01 -1.86086610e-01 3.88937257e-02
1.40581107e+00 2.58550853e-01 -1.90031600e+00 -5.09206891e-01
-5.96525311e-01 -1.03066754e+00 9.81311351e-02 -7.76896656e-01
-1.44296885e+00 7.13096559e-01 6.55934334e-01 -4.49515074e-01
1.55282581e+00 3.38843204e-02 1.06714714e+00 -2.12098792e-01
2.85132349e-01 -8.28593552e-01 2.66679823e-01 3.11387390e-01
1.01718915e+00 -1.21005440e+00 -1.69026986e-01 -5.66708326e-01
-6.37496889e-01 9.05509651e-01 7.42042005e-01 -1.87138021e-01
1.97535202e-01 2.73101836e-01 -2.10253522e-01 1.36825383e-01
-6.10502601e-01 -8.72548763e-03 3.23927373e-01 5.15806258e-01
-9.04048700e-03 -8.43937099e-02 3.81249972e-02 3.26392025e-01
1.13845840e-01 -1.15065038e-01 4.94336873e-01 5.21408796e-01
-3.56069386e-01 -1.04623759e+00 -4.16621923e-01 2.37147585e-01
-2.57140517e-01 -2.86660671e-01 -5.94587736e-02 4.37302142e-01
2.28636906e-01 6.68221116e-01 1.65285580e-02 -2.79922277e-01
1.67364389e-01 -3.27047110e-01 4.62914675e-01 -6.05688810e-01
-1.60394549e-01 5.27173579e-01 -1.83113500e-01 -9.31044996e-01
-2.37564832e-01 -3.21905524e-01 -1.00129318e+00 -3.04666638e-01
-1.07769608e-01 -3.29514444e-01 3.13007355e-01 7.11176574e-01
4.87498671e-01 8.25250626e-01 8.76461804e-01 -1.30908597e+00
-2.26137027e-01 -6.66264832e-01 -6.25907600e-01 5.74262798e-01
6.09747708e-01 -5.66125751e-01 -3.47734690e-01 3.09561402e-01] | [11.271093368530273, -1.2609461545944214] |
23bfbf4b-9293-4a0d-8e75-ccd2eb5b6bc3 | spatial-temporal-transformer-for-dynamic | 2107.12309 | null | https://arxiv.org/abs/2107.12309v2 | https://arxiv.org/pdf/2107.12309v2.pdf | Spatial-Temporal Transformer for Dynamic Scene Graph Generation | Dynamic scene graph generation aims at generating a scene graph of the given video. Compared to the task of scene graph generation from images, it is more challenging because of the dynamic relationships between objects and the temporal dependencies between frames allowing for a richer semantic interpretation. In this paper, we propose Spatial-temporal Transformer (STTran), a neural network that consists of two core modules: (1) a spatial encoder that takes an input frame to extract spatial context and reason about the visual relationships within a frame, and (2) a temporal decoder which takes the output of the spatial encoder as input in order to capture the temporal dependencies between frames and infer the dynamic relationships. Furthermore, STTran is flexible to take varying lengths of videos as input without clipping, which is especially important for long videos. Our method is validated on the benchmark dataset Action Genome (AG). The experimental results demonstrate the superior performance of our method in terms of dynamic scene graphs. Moreover, a set of ablative studies is conducted and the effect of each proposed module is justified. Code available at: https://github.com/yrcong/STTran. | ['Bodo Rosenhahn', 'Michael Ying Yang', 'Hanno Ackermann', 'Wentong Liao', 'Yuren Cong'] | 2021-07-26 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Cong_Spatial-Temporal_Transformer_for_Dynamic_Scene_Graph_Generation_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Cong_Spatial-Temporal_Transformer_for_Dynamic_Scene_Graph_Generation_ICCV_2021_paper.pdf | iccv-2021-1 | ['video-visual-relation-detection', 'visual-relationship-detection'] | ['computer-vision', 'computer-vision'] | [ 3.35325152e-01 1.62379533e-01 1.00375257e-01 -3.03636491e-01
-1.59377784e-01 -3.75808269e-01 6.37561977e-01 -5.89214452e-02
-2.15676799e-01 4.99798566e-01 3.49034876e-01 -1.72705159e-01
9.28730518e-02 -7.80139446e-01 -9.10529494e-01 -6.39213800e-01
-4.61504422e-02 -4.44190167e-02 5.50719798e-01 -2.37219557e-02
9.01098922e-02 3.57556313e-01 -1.46259701e+00 5.23310304e-01
6.61579669e-01 9.27493930e-01 7.28940427e-01 7.08833158e-01
-9.86465625e-03 1.32399774e+00 -3.67769510e-01 -1.98568404e-01
-1.47693185e-02 -8.39839280e-01 -7.38149643e-01 3.56190056e-01
1.78353608e-01 -5.35256684e-01 -6.83782518e-01 9.54433322e-01
2.78199434e-01 3.57380360e-01 2.30148375e-01 -1.36401105e+00
-5.32893062e-01 6.39481962e-01 -4.77085501e-01 3.33849400e-01
5.29391646e-01 3.29034209e-01 8.88528049e-01 -6.27139151e-01
7.62115657e-01 1.36006320e+00 1.55259268e-02 3.80061388e-01
-7.46903539e-01 -5.29840708e-01 4.80166078e-01 5.59650540e-01
-1.20761526e+00 -4.83535975e-01 8.63259554e-01 -5.01346707e-01
7.22107768e-01 1.24536632e-02 9.28340673e-01 1.06901085e+00
2.74576277e-01 7.96316206e-01 7.05355585e-01 -3.03943843e-01
2.18196079e-01 -3.66545111e-01 1.47440257e-02 6.17723346e-01
-1.80449054e-01 -1.61010679e-02 -5.92429996e-01 3.31097722e-01
9.66330409e-01 8.12719315e-02 -4.69010651e-01 -2.44257972e-01
-1.28376353e+00 4.68243271e-01 6.32143140e-01 2.70406067e-01
-4.10915583e-01 3.73538703e-01 3.90561283e-01 -5.61964512e-02
1.66383848e-01 -1.77400131e-02 4.40843366e-02 -4.01040316e-02
-6.46193326e-01 1.76611334e-01 4.21958327e-01 1.08537459e+00
4.78037953e-01 9.24566537e-02 -3.96159738e-01 4.88708884e-01
1.99386403e-01 2.90861756e-01 3.46726000e-01 -9.40064788e-01
5.79994857e-01 6.85389578e-01 2.39236038e-02 -1.21378088e+00
-3.21492106e-01 -1.04249969e-01 -7.89687335e-01 -6.16580946e-03
3.44695210e-01 -4.60844971e-02 -7.88542688e-01 1.72906065e+00
3.82444620e-01 5.75430393e-01 9.02163908e-02 1.14102042e+00
1.00675344e+00 9.54559565e-01 2.11878493e-01 -2.24806279e-01
1.45427918e+00 -1.10315216e+00 -8.75243545e-01 -3.58183295e-01
4.38875079e-01 -6.13812149e-01 1.02559543e+00 8.07161853e-02
-1.16000414e+00 -6.49808824e-01 -8.53295028e-01 -3.56889427e-01
-2.05929771e-01 2.74452657e-01 4.56711590e-01 -1.58878982e-01
-9.85269725e-01 2.92808950e-01 -9.94433463e-01 -3.66817266e-01
3.16535532e-01 1.13748521e-01 -2.62788057e-01 -7.44315982e-02
-1.19456482e+00 4.12348747e-01 8.64158630e-01 4.29638594e-01
-8.88922334e-01 -3.96657526e-01 -1.15562654e+00 1.33934781e-01
5.14981151e-01 -6.95790052e-01 1.11831117e+00 -1.17663205e+00
-1.35254872e+00 5.05568743e-01 -3.01131845e-01 -4.11137789e-01
6.08360052e-01 -2.60424107e-01 -3.17471206e-01 4.04591560e-01
5.10032568e-03 7.57844746e-01 6.37246490e-01 -1.00981510e+00
-7.70057142e-01 -1.80194661e-01 4.57378447e-01 4.43045259e-01
-6.64169937e-02 -2.17353906e-02 -1.11518788e+00 -6.99493289e-01
-1.48308977e-01 -9.86836016e-01 -2.11628750e-02 -6.42225221e-02
-6.23265207e-01 -1.57049403e-03 1.12032223e+00 -8.72485578e-01
1.36950052e+00 -2.31481099e+00 4.88995016e-01 -9.93822590e-02
1.17989391e-01 1.14832863e-01 -8.96394178e-02 5.72584152e-01
-3.26938003e-01 -2.76339233e-01 -1.61836252e-01 -1.07115552e-01
-3.12421560e-01 1.01602294e-01 -3.02724510e-01 2.16736048e-01
2.32371598e-01 9.20474470e-01 -1.01294231e+00 -4.79819030e-01
5.08415759e-01 5.79688251e-01 -5.25014758e-01 3.74897212e-01
-5.86188376e-01 7.06175685e-01 -5.44321239e-01 2.70789802e-01
3.96693647e-01 -5.13491750e-01 3.66812825e-01 -4.43689734e-01
-9.80671197e-02 1.07378565e-01 -1.15743887e+00 1.97733438e+00
-2.21169367e-01 6.75843716e-01 -2.70797372e-01 -7.61337101e-01
7.43017018e-01 2.98454314e-01 3.46394032e-01 -8.73985529e-01
2.09223911e-01 -2.03674346e-01 -6.92380518e-02 -7.64166713e-01
3.88840973e-01 2.14381725e-01 -4.81039546e-02 2.41445735e-01
-1.05927996e-01 1.34905800e-01 5.62260509e-01 4.31142777e-01
1.05715704e+00 4.67820525e-01 2.84631640e-01 6.52820477e-03
7.12699354e-01 -1.15909517e-01 7.09866226e-01 1.53144658e-01
8.31237361e-02 5.61409414e-01 7.59381413e-01 -5.89636683e-01
-9.01109099e-01 -8.99444997e-01 4.59596425e-01 7.18650818e-01
7.08069384e-01 -6.49002492e-01 -7.28316128e-01 -4.76613969e-01
-3.28047425e-01 8.68081152e-01 -6.54859602e-01 -1.22447640e-01
-7.38281846e-01 -2.94094533e-01 1.04377568e-01 6.57024622e-01
7.27257252e-01 -1.33936167e+00 -1.01752293e+00 -1.75715238e-02
-5.54751813e-01 -1.58328474e+00 -7.32016802e-01 -4.10650283e-01
-6.91003859e-01 -1.30996180e+00 -3.82416099e-01 -7.34211504e-01
7.46039212e-01 3.85699630e-01 8.85582447e-01 8.79797935e-02
-1.11278646e-01 1.44982770e-01 -4.91973996e-01 -1.30550086e-01
-3.22237283e-01 -2.30837986e-01 -4.44022268e-01 2.79346168e-01
-3.46469395e-02 -5.45206189e-01 -8.11553538e-01 3.62235844e-01
-1.18687522e+00 9.36581552e-01 4.47482735e-01 5.91529727e-01
6.34276569e-01 2.30229005e-01 1.18819743e-01 -8.78014147e-01
2.18908101e-01 -3.94179910e-01 -7.73386180e-01 3.02559853e-01
1.17452197e-01 -2.01938432e-02 6.95077181e-01 -1.94676593e-01
-1.23668265e+00 2.12719291e-01 1.17051616e-01 -5.48645854e-01
5.52235506e-02 3.75741363e-01 -3.51660550e-01 5.23664713e-01
3.53815764e-01 3.93607557e-01 -2.92627215e-02 -1.96563423e-01
3.96167517e-01 2.59573817e-01 6.21901810e-01 -3.31164181e-01
6.01015449e-01 5.43919086e-01 -1.07029028e-01 -7.06112266e-01
-7.95200765e-01 -3.50516587e-02 -8.08239579e-01 -6.10332072e-01
1.25235641e+00 -9.31766927e-01 -5.62327802e-01 5.02261758e-01
-1.23154998e+00 -6.71846747e-01 -1.69034362e-01 4.24064308e-01
-7.17746496e-01 2.87125349e-01 -5.98522305e-01 -4.90181029e-01
-1.93602338e-01 -1.26544189e+00 1.08124781e+00 4.70020890e-01
-1.41416071e-02 -9.87315595e-01 -3.11113924e-01 3.83253813e-01
-8.80780220e-02 6.15500689e-01 8.30688596e-01 -1.62704866e-02
-1.07632172e+00 1.17012821e-01 -3.01771492e-01 4.06995304e-02
2.18269438e-01 2.23874316e-01 -7.06048310e-01 -2.74286121e-01
-2.27197945e-01 -4.88803424e-02 6.01940811e-01 4.13645208e-01
1.30950844e+00 -3.97771925e-01 -3.33759755e-01 6.45748138e-01
1.37451327e+00 6.73092604e-01 9.00890589e-01 2.44731635e-01
9.23923612e-01 6.23030245e-01 7.56241381e-01 3.28877419e-01
4.75905448e-01 8.73372436e-01 5.82373917e-01 -1.10618271e-01
-3.78656954e-01 -4.66566086e-01 5.18027186e-01 7.21177280e-01
-9.65871811e-02 -5.63002646e-01 -8.66192937e-01 4.84358668e-01
-2.20775270e+00 -1.13123715e+00 -7.80694708e-02 1.96077120e+00
3.91577363e-01 -2.19483371e-03 9.54429880e-02 5.52938879e-02
8.92590344e-01 3.28646272e-01 -6.37466669e-01 -1.21349797e-01
3.84419709e-02 -3.07038069e-01 8.75904411e-02 4.74614233e-01
-9.14312184e-01 1.06929493e+00 4.91115761e+00 6.57809496e-01
-1.16698599e+00 -4.80636321e-02 7.27284908e-01 -1.87802091e-01
-1.91664398e-01 6.06066100e-02 -3.64192307e-01 6.34650767e-01
4.84304965e-01 -2.12448940e-01 4.21078980e-01 4.98414457e-01
5.38928449e-01 -2.55390823e-01 -1.01252890e+00 9.70002294e-01
1.26690948e-02 -1.30377758e+00 2.34386876e-01 -2.39475995e-01
4.47791338e-01 -2.48530641e-01 -2.59289950e-01 -8.55216160e-02
1.48634344e-01 -8.02562118e-01 1.13578045e+00 6.11012757e-01
7.69363582e-01 -6.06456280e-01 4.40234542e-01 3.19326431e-01
-1.50034738e+00 -1.33867949e-01 -1.81307167e-01 -1.04533657e-01
4.54033345e-01 4.98924702e-01 -6.13042057e-01 7.64021814e-01
6.74700141e-01 1.16726887e+00 -5.98830521e-01 8.91482830e-01
-6.39097393e-01 3.06966335e-01 1.30389519e-02 2.14438438e-01
1.45119756e-01 -3.78310412e-01 4.84449387e-01 1.12125003e+00
4.22236353e-01 3.06262523e-01 8.91843140e-02 7.97345638e-01
1.31583631e-01 -3.69024277e-03 -7.83384144e-01 -7.91161656e-02
3.47076505e-01 1.20647836e+00 -9.19712961e-01 -3.42619985e-01
-4.25694674e-01 1.18582308e+00 3.20324898e-01 5.46524465e-01
-1.19098556e+00 -1.20245919e-01 3.02478760e-01 2.22328767e-01
3.16446602e-01 -2.43550181e-01 1.49409041e-01 -1.17058480e+00
1.75559998e-01 -7.09742963e-01 6.65317655e-01 -1.24300385e+00
-5.45354426e-01 6.82346880e-01 1.34301767e-01 -1.36075926e+00
-3.09458315e-01 -3.11569750e-01 -6.56272173e-01 5.89145541e-01
-1.30048871e+00 -1.12953699e+00 -8.09066951e-01 6.46784604e-01
8.40679824e-01 1.55460611e-01 3.76432091e-01 2.72935688e-01
-8.18076611e-01 1.16708443e-01 -4.02367562e-01 3.23767364e-01
2.72558391e-01 -1.00238967e+00 3.89951110e-01 1.21951568e+00
1.25624686e-01 2.38656968e-01 4.66771960e-01 -6.60861909e-01
-1.45714080e+00 -1.31136143e+00 6.14868999e-01 -1.46678671e-01
5.03133714e-01 -5.35776556e-01 -7.25007832e-01 8.88543546e-01
3.14730376e-01 5.31434976e-02 2.55663276e-01 -5.56544065e-01
-6.04267493e-02 -1.72796160e-01 -6.54371798e-01 7.61571109e-01
1.32274294e+00 -3.43298554e-01 -2.14563698e-01 2.86501586e-01
8.47343266e-01 -7.58978367e-01 -6.12169802e-01 2.92731911e-01
5.14896691e-01 -1.17134869e+00 8.10344577e-01 -2.31811479e-01
8.19920301e-01 -7.47027695e-01 4.45178077e-02 -1.09599066e+00
-2.40296096e-01 -4.44563031e-01 -5.40978797e-02 1.23387492e+00
1.28503472e-01 -3.31260353e-01 5.09849429e-01 3.82369161e-01
-1.73805565e-01 -6.45527482e-01 -5.70223212e-01 -5.78412414e-01
-5.91288984e-01 -4.21817124e-01 5.21163225e-01 8.66102636e-01
-9.14442018e-02 4.85105515e-01 -4.86893922e-01 2.06421986e-01
3.31985056e-01 3.59377354e-01 8.38560820e-01 -5.70336521e-01
-4.47090566e-01 -1.42504126e-01 -4.88289237e-01 -1.14829230e+00
1.14226423e-01 -7.13972926e-01 7.51239508e-02 -1.71393323e+00
2.81470537e-01 -7.95949921e-02 -6.94477707e-02 4.65007871e-01
-2.69618213e-01 2.59024724e-02 4.72312063e-01 1.20139517e-01
-7.60717034e-01 7.09949851e-01 1.64093411e+00 -1.03709951e-01
-2.33461022e-01 -3.19959283e-01 -4.58907306e-01 6.67938709e-01
7.61047304e-01 -9.91020352e-02 -9.92651999e-01 -6.88157201e-01
2.69462075e-02 5.56289077e-01 6.06847286e-01 -1.12764394e+00
1.67054132e-01 -2.45337948e-01 3.38190854e-01 -5.78954816e-01
3.57267052e-01 -8.43725383e-01 6.83002055e-01 5.82448781e-01
-2.07942307e-01 3.52464229e-01 2.54565030e-01 5.89163184e-01
-3.82442802e-01 1.33096546e-01 6.80697381e-01 -7.16740415e-02
-1.12571454e+00 3.72371405e-01 -3.46895084e-02 -6.06507482e-03
1.35598338e+00 -2.18347490e-01 -4.61124033e-01 -4.57674891e-01
-7.83946574e-01 3.44477147e-01 5.77785015e-01 6.78697467e-01
7.82902002e-01 -1.38290739e+00 -4.99157161e-01 1.79118469e-01
1.36008084e-01 2.47793943e-01 5.39434493e-01 8.39944124e-01
-7.81342685e-01 3.33464861e-01 -2.91661978e-01 -6.68327987e-01
-1.36241829e+00 6.23409450e-01 3.46074224e-01 -3.87249365e-02
-1.04975486e+00 6.07250154e-01 8.71726871e-01 2.42895752e-01
1.12707205e-01 -5.29906154e-01 -3.76420259e-01 -1.02491856e-01
6.09805942e-01 1.17468573e-01 -2.88989991e-01 -8.09050381e-01
-2.18412429e-01 4.95124221e-01 1.02845110e-01 -7.57837817e-02
1.19763267e+00 -1.88210204e-01 -2.25363508e-01 4.44852173e-01
9.49412405e-01 -3.49616468e-01 -1.65047300e+00 -1.46415979e-01
-1.44275576e-01 -5.68917751e-01 -1.96076095e-01 -4.53355491e-01
-1.33018315e+00 7.08886623e-01 3.45055640e-01 3.44578810e-02
1.50917029e+00 1.59519576e-02 8.37310612e-01 -8.13192353e-02
2.94048488e-01 -8.06390822e-01 3.37793678e-01 2.95431674e-01
1.11593938e+00 -7.75392890e-01 -1.13136061e-01 -7.15377986e-01
-7.54141450e-01 1.11003697e+00 6.33543491e-01 3.68695334e-02
3.96764100e-01 1.28666818e-01 -1.32784650e-01 -3.71153474e-01
-8.28026474e-01 -1.66799068e-01 1.90656602e-01 3.23561370e-01
3.79893392e-01 -1.55252293e-01 -2.48419106e-01 2.15250656e-01
-1.26563922e-01 2.23759830e-01 5.32668948e-01 8.23701382e-01
-5.18194884e-02 -9.13190603e-01 -5.33965454e-02 8.63938928e-02
-1.62852064e-01 9.61276218e-02 -2.53288269e-01 7.85389304e-01
2.35944182e-01 9.07713413e-01 2.10004672e-01 -3.39711010e-01
3.53965938e-01 -1.22336619e-01 3.60161960e-01 -6.05096877e-01
-1.37160257e-01 1.04095787e-01 1.23878032e-01 -9.58121538e-01
-5.99281788e-01 -7.20079243e-01 -1.58317578e+00 -5.85746057e-02
-1.28945177e-02 5.49400821e-02 3.26755285e-01 7.77078986e-01
3.49721283e-01 9.50632274e-01 4.86894816e-01 -7.20611811e-01
3.74156237e-01 -7.22500503e-01 -4.29495215e-01 8.06699812e-01
2.17225209e-01 -6.39671743e-01 1.50174228e-02 6.83456540e-01] | [9.181679725646973, 0.7255865931510925] |
973d83c4-bfb4-4fbd-9387-fc2cfe45899b | recognizing-abnormal-heart-sounds-using-deep | 1707.04642 | null | http://arxiv.org/abs/1707.04642v2 | http://arxiv.org/pdf/1707.04642v2.pdf | Recognizing Abnormal Heart Sounds Using Deep Learning | The work presented here applies deep learning to the task of automated
cardiac auscultation, i.e. recognizing abnormalities in heart sounds. We
describe an automated heart sound classification algorithm that combines the
use of time-frequency heat map representations with a deep convolutional neural
network (CNN). Given the cost-sensitive nature of misclassification, our CNN
architecture is trained using a modified loss function that directly optimizes
the trade-off between sensitivity and specificity. We evaluated our algorithm
at the 2016 PhysioNet Computing in Cardiology challenge where the objective was
to accurately classify normal and abnormal heart sounds from single, short,
potentially noisy recordings. Our entry to the challenge achieved a final
specificity of 0.95, sensitivity of 0.73 and overall score of 0.84. We achieved
the greatest specificity score out of all challenge entries and, using just a
single CNN, our algorithm differed in overall score by only 0.02 compared to
the top place finisher, which used an ensemble approach. | ['Kumar Sricharan', 'Anurag Ganguli', 'Ion Matei', 'Saigopal Nelaturi', 'Jonathan Rubin', 'Rui Abreu'] | 2017-07-14 | null | null | null | null | ['sound-classification'] | ['audio'] | [ 2.17823595e-01 2.14926973e-01 5.10219455e-01 -3.18844765e-01
-1.21256256e+00 -5.12659669e-01 -4.68933098e-02 3.47165257e-01
-4.22474325e-01 5.09914279e-01 1.22764789e-01 -6.54726982e-01
-2.62937754e-01 -3.76957238e-01 -2.92752713e-01 -5.01369596e-01
-4.80444670e-01 3.88274282e-01 -1.45366201e-02 1.11255124e-01
4.00513299e-02 2.96255887e-01 -1.10398674e+00 4.61437076e-01
5.12071490e-01 1.21010172e+00 -5.84694266e-01 1.45643437e+00
3.92468601e-01 9.11072731e-01 -7.65091956e-01 -1.59354210e-01
1.86368644e-01 -8.17496121e-01 -8.76350522e-01 -5.84683478e-01
6.74305022e-01 -4.53628823e-02 -1.01602562e-01 3.76497060e-01
1.22040248e+00 -5.21475561e-02 2.39894167e-01 -7.96961188e-01
-1.35813132e-01 4.26733911e-01 1.73793226e-01 9.42454278e-01
1.14851624e-01 3.03269178e-01 1.01850390e+00 -5.46711802e-01
1.47324830e-01 3.92757833e-01 1.57120538e+00 3.53368640e-01
-1.27213788e+00 -5.22065997e-01 -6.28026843e-01 -2.18408301e-01
-1.23203075e+00 -3.51173252e-01 4.01400805e-01 -6.56980038e-01
1.11138332e+00 4.91872132e-01 9.18052852e-01 5.76236248e-01
2.90699959e-01 1.27846703e-01 1.03295290e+00 -2.42505878e-01
1.01974785e-01 -1.23734951e-01 2.85575539e-02 7.30916858e-01
1.59836099e-01 1.06843337e-01 -1.94792524e-01 -5.70560098e-01
4.94299322e-01 3.84107046e-02 -2.06702292e-01 1.81767166e-01
-1.33721292e+00 6.59094572e-01 2.39118472e-01 4.80702788e-01
-2.85496563e-01 2.79290020e-01 6.09549284e-01 3.06266099e-01
1.79969132e-01 1.24307787e+00 -7.60097086e-01 -6.07819617e-01
-1.13815653e+00 2.76744485e-01 1.03328180e+00 -8.72249305e-02
-5.95902791e-03 -2.18774155e-02 -4.18439716e-01 9.27959859e-01
-2.46473730e-01 2.83412308e-01 4.26259011e-01 -1.03057241e+00
1.36618197e-01 2.26135492e-01 -1.14011236e-01 -1.00476110e+00
-9.70138609e-01 -1.04415607e+00 -8.12870324e-01 1.11512005e-01
5.02778232e-01 -5.87991297e-01 -9.20281112e-01 1.55568731e+00
-2.95938947e-03 3.78992081e-01 -1.28724128e-01 1.02704978e+00
9.96849239e-01 1.25978515e-01 1.72279343e-01 2.89045926e-02
1.24949169e+00 -5.45136034e-01 -4.55940634e-01 2.31903344e-01
7.18697011e-01 -6.56969428e-01 9.55334961e-01 5.72318852e-01
-1.27021551e+00 -5.29462695e-01 -1.21785998e+00 5.49096204e-02
-2.34090656e-01 4.33940403e-02 3.08639497e-01 7.56585896e-01
-1.18560600e+00 1.06131291e+00 -8.86297703e-01 -1.09761290e-01
6.86799228e-01 3.82154346e-01 -9.49535593e-02 3.65715891e-01
-1.26914918e+00 8.23530257e-01 -3.39863524e-02 -2.14276724e-02
-6.06366873e-01 -1.28238225e+00 -4.45404142e-01 2.63902843e-01
-1.96283653e-01 -7.95524180e-01 1.34998500e+00 -6.94433451e-01
-1.21639705e+00 9.48793054e-01 4.64570373e-01 -6.68351948e-01
7.36526668e-01 -1.65709332e-01 -4.97976363e-01 3.08183610e-01
-9.83864665e-02 3.33035260e-01 3.83043736e-01 -5.62726557e-01
-6.67673588e-01 -1.08855002e-01 -2.29048193e-01 -1.80360436e-01
-7.58257136e-02 -1.54673249e-01 1.38208404e-01 -6.59691513e-01
1.07396685e-01 -9.15328205e-01 -2.18171403e-01 -1.08647592e-01
-2.81139910e-01 -8.21127370e-03 3.73554379e-01 -9.13216829e-01
1.32034421e+00 -2.12395644e+00 -2.37306312e-01 3.25859338e-01
8.88831139e-01 4.29493666e-01 2.52126306e-01 2.62160480e-01
-2.96637952e-01 4.89134014e-01 -4.82446194e-01 -1.71800718e-01
-2.89186597e-01 -1.93162307e-01 5.62219396e-02 4.08585995e-01
4.68503654e-01 8.54355454e-01 -9.83333349e-01 -3.25285405e-01
1.53947994e-01 6.34372592e-01 -5.18626869e-01 3.29606801e-01
3.66715342e-01 4.00753766e-01 1.16921768e-01 4.24921870e-01
2.31532812e-01 -4.06473190e-01 9.20758396e-02 -2.08905950e-01
-8.06868076e-02 5.65589547e-01 -8.07432473e-01 1.67274928e+00
-5.53464234e-01 9.94626343e-01 -3.37450415e-01 -1.01881385e+00
9.41489577e-01 6.34949803e-01 9.45020974e-01 -3.65124315e-01
-4.01957482e-02 5.51196337e-01 6.69662535e-01 -6.55180871e-01
-2.81545967e-01 -3.95036697e-01 1.34914309e-01 4.79632586e-01
6.24619657e-03 -1.79780334e-01 -2.77506083e-01 -1.78127497e-01
1.72652936e+00 -2.67533392e-01 1.40654951e-01 -4.64424938e-01
6.52789623e-02 -6.00340627e-02 4.00460094e-01 1.17310572e+00
-5.71658313e-01 1.39770663e+00 6.21729970e-01 -1.12739313e+00
-9.78598952e-01 -1.00725722e+00 -4.76029903e-01 7.08586037e-01
-6.55337274e-01 -4.42559868e-01 -5.76608121e-01 -6.90654635e-01
2.35125255e-02 8.20287019e-02 -7.79601455e-01 -2.46344313e-01
-8.39461446e-01 -8.82303298e-01 1.19871557e+00 7.55666792e-01
2.33214796e-01 -1.15347433e+00 -1.35058951e+00 4.39542651e-01
-1.09441392e-01 -7.31167376e-01 -2.47360155e-01 3.48669291e-01
-7.38096297e-01 -1.16727722e+00 -7.55640328e-01 -5.83638191e-01
-5.59300035e-02 -5.74573338e-01 1.52582288e+00 3.81251723e-01
-1.17551517e+00 1.18024662e-01 -2.43989274e-01 -8.36508036e-01
-2.39785001e-01 3.54498237e-01 -3.00070375e-01 -1.83431342e-01
2.14318439e-01 -6.88987434e-01 -1.15655506e+00 -3.50717574e-01
-5.96566260e-01 -3.42925489e-01 3.23482573e-01 1.06037092e+00
2.89604247e-01 -4.61047888e-01 9.15285528e-01 -7.64408410e-01
8.01550567e-01 -4.91394341e-01 5.48902825e-02 -1.70250446e-01
-8.16997409e-01 -3.77502292e-01 6.21569037e-01 -1.23558939e-01
-1.42411798e-01 8.30468442e-03 -4.08512026e-01 -3.93850982e-01
-7.55662844e-02 2.92708188e-01 8.31376195e-01 7.67245218e-02
1.01982498e+00 5.73652890e-03 2.96759605e-01 -3.69505376e-01
-2.83582240e-01 5.79515696e-01 7.88532555e-01 -6.90510422e-02
2.02531606e-01 2.18599156e-01 3.03207219e-01 -3.62548918e-01
-8.91963184e-01 -4.07652736e-01 -3.76261294e-01 -9.32300985e-02
1.03725266e+00 -6.24620140e-01 -8.01223218e-01 1.79991364e-01
-8.31053317e-01 -3.15839589e-01 -6.46101594e-01 6.30827785e-01
-3.82227153e-01 8.64222795e-02 -5.89025795e-01 -7.63885319e-01
-7.86389530e-01 -7.52412736e-01 9.60078657e-01 -1.59784049e-01
-7.48822987e-01 -1.06939638e+00 5.37103117e-01 1.03724234e-01
9.96600151e-01 9.60706711e-01 8.14730644e-01 -1.07274091e+00
1.58443764e-01 -3.44874084e-01 -1.74688429e-01 6.56913340e-01
3.03719342e-01 1.05125541e-02 -1.30123651e+00 -3.10613960e-01
-1.31684259e-01 -4.02996689e-01 9.85777080e-01 5.52780151e-01
1.61153948e+00 -1.29797682e-01 1.35746062e-01 8.74641061e-01
1.20051122e+00 3.66261601e-01 5.54080307e-01 2.25393116e-01
5.56382060e-01 3.05844873e-01 -7.22322091e-02 4.91048783e-01
1.94063753e-01 3.14247787e-01 1.25620738e-01 -7.12024212e-01
-1.95601910e-01 1.71936378e-01 -4.30490643e-01 5.98244667e-01
-2.62629718e-01 1.69031829e-01 -1.51381516e+00 6.02080703e-01
-1.58211410e+00 -8.78107369e-01 -8.24802592e-02 2.13709545e+00
7.83882558e-01 2.22775608e-01 3.08014184e-01 4.58680511e-01
4.39503789e-01 -6.75654411e-02 -4.36281860e-01 -6.59041822e-01
9.22999009e-02 1.02457142e+00 1.11131191e-01 3.62165570e-01
-1.52573729e+00 1.52688503e-01 7.25423527e+00 1.21936373e-01
-1.51144111e+00 1.90560352e-02 1.01959467e+00 -2.96802759e-01
1.98217362e-01 -4.01749521e-01 4.44049239e-02 5.86768985e-01
1.35997093e+00 1.67707264e-01 1.46120191e-01 4.03499842e-01
2.33089505e-03 4.20638740e-01 -9.75443423e-01 1.11737728e+00
6.57285899e-02 -1.54367959e+00 -5.67690790e-01 -1.81300253e-01
5.87664545e-01 4.22983438e-01 -2.36143414e-02 2.01207906e-01
-2.30139643e-01 -1.58131063e+00 1.65342093e-01 8.24633837e-01
1.06782961e+00 -5.23322582e-01 1.12501383e+00 -1.54597536e-01
-8.04901421e-01 -2.63099134e-01 2.20947802e-01 -2.79234380e-01
-2.06062078e-01 7.13972211e-01 -1.10550320e+00 1.14697151e-01
9.92369056e-01 6.15343571e-01 -3.40546817e-01 1.29539573e+00
4.05203104e-01 9.82908010e-01 -4.63443071e-01 8.12051520e-02
3.49737071e-02 5.05384982e-01 5.46447694e-01 1.74172914e+00
3.56450886e-01 3.33739460e-01 2.33095065e-01 7.37316310e-01
6.94588898e-03 2.00300768e-01 -4.64983255e-01 1.95738792e-01
4.06607628e-01 1.16395259e+00 -4.73618478e-01 -5.95067978e-01
-2.08325535e-02 3.46144378e-01 5.68179898e-02 -3.61670665e-02
-6.20054066e-01 -9.41981733e-01 3.90881598e-01 2.59958088e-01
1.07161514e-01 4.04739022e-01 -9.77212727e-01 -7.39292383e-01
7.64756724e-02 -9.13762450e-01 6.11389518e-01 -5.44081151e-01
-1.20797825e+00 8.60372543e-01 -5.03169596e-01 -1.15800917e+00
-3.23580295e-01 -3.41787398e-01 -8.84124756e-01 1.13709188e+00
-1.23883224e+00 -4.27348763e-01 -4.15058494e-01 2.21511628e-02
2.60063797e-01 -1.05947562e-01 1.25604725e+00 4.46783692e-01
-3.95335704e-01 7.85378218e-01 -3.51936877e-01 3.48564267e-01
6.65695727e-01 -1.67654622e+00 3.73579919e-01 4.19504970e-01
-2.21429154e-01 5.51283062e-01 5.08280933e-01 -1.86802641e-01
-7.15911090e-01 -1.09435153e+00 1.19932485e+00 -7.14671075e-01
3.79478216e-01 7.86049291e-02 -8.44847679e-01 1.99288696e-01
-2.46524997e-02 5.34913123e-01 1.00850701e+00 1.72595382e-01
-3.43135774e-01 -3.11168641e-01 -1.17472935e+00 -4.54555266e-02
7.19348550e-01 -5.64239323e-01 -4.72848952e-01 2.82860219e-01
3.58611435e-01 -6.68793440e-01 -1.38802469e+00 7.53219306e-01
1.09765697e+00 -1.04216182e+00 9.23705757e-01 -9.20827329e-01
6.55479550e-01 2.00767480e-02 1.62768990e-01 -1.31487572e+00
-5.64567149e-01 -6.90506220e-01 -4.01011296e-02 4.12624717e-01
5.99651873e-01 -5.99645853e-01 6.12253666e-01 4.07528162e-01
-4.62284148e-01 -1.30306494e+00 -8.74358058e-01 -3.82849634e-01
2.65105844e-01 -3.55653703e-01 4.23334718e-01 1.01279640e+00
8.76501277e-02 1.34481981e-01 -1.72100514e-01 -2.16891259e-01
3.38767320e-01 1.80164594e-02 2.73555338e-01 -1.46697938e+00
-5.16081750e-01 -6.32905841e-01 -6.48910224e-01 -1.01107359e-01
-5.51954925e-01 -9.62268114e-01 1.15991972e-01 -1.30204237e+00
2.10496873e-01 -5.95144033e-01 -1.03241193e+00 6.97693348e-01
-3.93639654e-01 8.14769924e-01 1.96936220e-01 1.77673101e-01
-3.31723124e-01 -4.14073795e-01 9.86362576e-01 2.39825509e-02
-4.08087164e-01 9.53703150e-02 -7.50206470e-01 4.48392749e-01
1.26814556e+00 -4.76821899e-01 -1.62789121e-01 -2.90749669e-01
2.36683652e-01 1.56804264e-01 6.45047009e-01 -1.36343765e+00
-8.53588209e-02 4.22896415e-01 6.09730899e-01 -1.35539383e-01
1.73724219e-01 -3.74524891e-01 1.42010882e-01 9.09665167e-01
-8.19741428e-01 2.58033723e-01 3.13816875e-01 3.29277366e-02
-1.94367111e-01 3.48687500e-01 8.80495310e-01 -2.97133863e-01
1.54026449e-01 1.67651936e-01 -3.04876298e-01 5.12598276e-01
4.57711726e-01 -9.15409997e-02 -2.56409883e-01 -2.08341807e-01
-1.16593421e+00 -1.82614088e-01 -2.95962930e-01 2.49970704e-01
4.88199770e-01 -1.03714311e+00 -1.17319727e+00 2.28110865e-01
9.35119390e-02 -1.68176442e-01 2.29922652e-01 1.24851501e+00
-8.72670531e-01 3.64213049e-01 -1.40044957e-01 -8.80742013e-01
-1.25440395e+00 -2.17824012e-01 1.09761286e+00 -2.93592632e-01
-1.00134575e+00 8.72095108e-01 -3.54718059e-01 -3.31244081e-01
2.82929093e-01 -5.39182186e-01 -6.55542389e-02 -2.83760309e-01
4.83238339e-01 5.56219339e-01 5.79408586e-01 8.00553709e-02
-5.25201857e-01 4.24847901e-01 2.87476122e-01 -7.58843422e-02
1.29206002e+00 5.19018948e-01 5.49563095e-02 5.04862309e-01
1.50058854e+00 -3.55407864e-01 -8.23643804e-01 2.39137828e-01
-1.57940894e-01 -2.83098012e-01 1.88440010e-01 -1.42550802e+00
-1.10526669e+00 1.14547229e+00 1.15977573e+00 6.46170139e-01
1.11167979e+00 -2.13943020e-01 9.62709069e-01 3.83257687e-01
-4.38308030e-01 -9.04042602e-01 1.47417411e-02 3.63514900e-01
6.53691113e-01 -1.24118245e+00 -2.02914461e-01 2.54324406e-01
-3.74286473e-01 1.30097568e+00 3.29369247e-01 -2.49698877e-01
9.20943499e-01 4.62439865e-01 5.09396255e-01 -5.63416660e-01
-8.41971874e-01 1.20375842e-01 3.32405210e-01 5.60763955e-01
9.34971929e-01 2.49272168e-01 -3.87591600e-01 5.85089743e-01
-4.87420559e-01 2.70793468e-01 2.99958915e-01 9.83935535e-01
-3.73891890e-01 -5.08945048e-01 -2.52565052e-02 8.86916816e-01
-1.28955972e+00 -1.91635281e-01 -3.49312067e-01 4.67215449e-01
3.46152455e-01 9.27669406e-01 2.46442840e-01 -5.41626811e-01
5.02186120e-01 3.58352631e-01 1.18289307e-01 -4.95389640e-01
-1.22614551e+00 -1.25490457e-01 2.42139071e-01 -5.75305820e-01
-1.57596350e-01 -5.63286722e-01 -1.03971648e+00 1.19928561e-01
1.49258271e-01 2.25553721e-01 5.88930309e-01 6.11702025e-01
5.20599961e-01 9.97562170e-01 7.93665230e-01 -4.05433893e-01
-6.99497104e-01 -8.93637002e-01 -4.04189914e-01 5.46575963e-01
1.07660747e+00 -1.09145373e-01 -5.95290780e-01 1.95247692e-03] | [14.339444160461426, 3.3163557052612305] |
c3205798-563e-49db-8a95-8bee955bb108 | deep-temporal-graph-clustering | 2305.10738 | null | https://arxiv.org/abs/2305.10738v1 | https://arxiv.org/pdf/2305.10738v1.pdf | Deep Temporal Graph Clustering | Deep graph clustering has recently received significant attention due to its ability to enhance the representation learning capabilities of models in unsupervised scenarios. Nevertheless, deep clustering for temporal graphs, which could capture crucial dynamic interaction information, has not been fully explored. It means that in many clustering-oriented real-world scenarios, temporal graphs can only be processed as static graphs. This not only causes the loss of dynamic information but also triggers huge computational consumption. To solve the problem, we propose a general framework for deep Temporal Graph Clustering called TGC, which adjusts deep clustering techniques (clustering assignment distribution and adjacency matrix reconstruction) to suit the interaction sequence-based batch-processing pattern of temporal graphs. In addition, we discuss differences between temporal graph clustering and existing static graph clustering from several levels. To verify the superiority of the proposed framework TGC, we conduct extensive experiments. The experimental results show that temporal graph clustering enables more flexibility in finding a balance between time and space requirements, and our framework can effectively improve the performance of existing temporal graph learning methods. Our code and supplementary material will be released after publication. | ['Xinwang Liu', 'Sihang Zhou', 'Siwei Wang', 'Ke Liang', 'Yue Liu', 'Meng Liu'] | 2023-05-18 | null | null | null | null | ['graph-clustering', 'deep-clustering', 'deep-clustering'] | ['graphs', 'miscellaneous', 'natural-language-processing'] | [-2.00577110e-01 -2.64091015e-01 -2.01905996e-01 -1.35724872e-01
-8.85394141e-02 -6.08488739e-01 4.37370926e-01 3.12575728e-01
-1.35784522e-01 1.17201813e-01 3.25024687e-02 -4.10974950e-01
-5.83563268e-01 -7.62014389e-01 -3.28761190e-01 -8.79151881e-01
-6.94716573e-01 4.92446423e-01 4.93110418e-01 -3.92018110e-02
-5.54099828e-02 4.00416106e-01 -1.32642639e+00 2.31940761e-01
8.46359909e-01 5.99616706e-01 2.55386084e-01 3.60324621e-01
-2.49763623e-01 6.80801749e-01 -4.40235347e-01 -2.31518254e-01
2.19943613e-01 -4.01555270e-01 -6.23299956e-01 1.51711002e-01
-3.52145284e-01 -2.44544400e-03 -8.05737555e-01 9.31042075e-01
5.02450585e-01 3.86871964e-01 3.63360316e-01 -1.52405572e+00
-3.20346892e-01 1.06323218e+00 -7.16308355e-01 3.23482245e-01
1.06204547e-01 3.22326630e-01 8.53650808e-01 -1.68462813e-01
5.37897944e-01 1.28857756e+00 5.18989444e-01 3.61217648e-01
-1.04174459e+00 -6.46063447e-01 7.16945410e-01 4.52026576e-01
-1.43367696e+00 -1.26030907e-01 1.12347615e+00 -4.34149772e-01
8.94164979e-01 1.35237604e-01 9.25197363e-01 1.04370701e+00
8.67382213e-02 8.38837802e-01 5.08255899e-01 -1.44465983e-01
2.97996283e-01 -5.52179277e-01 2.26846784e-01 6.05501354e-01
9.65238288e-02 -1.02701172e-01 -2.38511354e-01 1.17865093e-01
5.71611822e-01 2.38060668e-01 -1.95624784e-01 -6.66890025e-01
-1.18837321e+00 6.03855848e-01 5.93705893e-01 5.64045548e-01
-1.68457344e-01 4.23530787e-01 7.58560061e-01 3.34373474e-01
3.47165376e-01 4.44431640e-02 -4.41517353e-01 -3.06631863e-01
-6.48648381e-01 -3.93657684e-02 4.70922709e-01 1.00280893e+00
5.65704882e-01 7.13014696e-03 -3.65688615e-02 6.48585916e-01
2.03695536e-01 5.24430573e-02 4.70224679e-01 -6.36706710e-01
4.40847278e-01 1.05609345e+00 -3.69832307e-01 -1.45992935e+00
-6.43592298e-01 -4.95375603e-01 -1.26690197e+00 -5.74512005e-01
2.73153037e-01 -9.08393711e-02 -8.04761946e-01 1.74454665e+00
4.59178746e-01 4.54252481e-01 -2.07407832e-01 7.42112041e-01
6.86096191e-01 6.55586600e-01 3.62047134e-03 -5.84944487e-01
9.35506582e-01 -8.08483362e-01 -8.95957530e-01 -5.06413132e-02
9.01917338e-01 -4.12074566e-01 8.99861515e-01 1.09700069e-01
-6.32717848e-01 -4.71788287e-01 -8.50246072e-01 2.77444690e-01
-3.79285902e-01 -2.33958960e-01 1.19796193e+00 4.27665740e-01
-1.20231783e+00 5.84428966e-01 -1.37233031e+00 -6.91741407e-01
8.50638822e-02 4.55166042e-01 -1.56476557e-01 -5.73821291e-02
-1.12813020e+00 1.06916443e-01 7.46075153e-01 4.30493653e-01
-7.15422034e-01 -3.55561286e-01 -9.26132739e-01 2.39052027e-01
7.33676314e-01 -3.70197892e-01 8.66245508e-01 -6.66758835e-01
-1.24539900e+00 4.01377589e-01 -1.02155596e-01 -3.88688624e-01
4.85573053e-01 1.92155495e-01 -5.91647983e-01 1.20429046e-01
1.22372340e-02 3.32739592e-01 5.07850587e-01 -1.04901958e+00
-3.48039329e-01 -4.73009527e-01 1.07425742e-01 1.24801464e-01
-5.87972701e-01 -1.72170758e-01 -1.33891785e+00 -6.98321819e-01
2.56428421e-01 -1.11411345e+00 -6.36810541e-01 -4.48909402e-01
-3.42783809e-01 -3.81822705e-01 9.03407097e-01 -2.74537235e-01
1.84886992e+00 -2.29463792e+00 2.82581121e-01 3.20993274e-01
4.14250553e-01 8.65275711e-02 -1.67026252e-01 8.17678213e-01
-1.99922249e-01 2.18980759e-01 -1.52716964e-01 -1.34684891e-01
1.77829172e-02 2.12403148e-01 3.76206674e-02 4.44195598e-01
-2.89948940e-01 1.01825035e+00 -9.80872869e-01 -5.95154643e-01
3.52975011e-01 3.16379666e-01 -4.55620795e-01 2.85650529e-02
-2.56493568e-01 4.15873885e-01 -5.01176238e-01 4.06175971e-01
6.92849159e-01 -7.20392048e-01 8.60108256e-01 -1.80658773e-01
1.17788963e-01 -9.70478728e-02 -1.21668911e+00 1.69417608e+00
-1.16508625e-01 4.06077892e-01 -4.39336486e-02 -1.35031831e+00
6.99221969e-01 1.68566078e-01 1.01619923e+00 -7.42834687e-01
1.92280442e-01 -2.26594806e-01 3.91017556e-01 -3.38491708e-01
2.39326850e-01 2.96087891e-01 -1.45447299e-01 5.56867838e-01
-2.08440870e-01 4.92988408e-01 5.68821907e-01 6.19362295e-01
1.40404439e+00 -1.39833510e-01 4.83266935e-02 -1.72349766e-01
3.67721200e-01 -7.42386952e-02 6.99563146e-01 4.36939269e-01
-3.50917280e-01 1.57433867e-01 7.02875972e-01 -5.17971933e-01
-4.85667706e-01 -8.67427707e-01 4.56720918e-01 9.29748476e-01
3.98808897e-01 -9.24533904e-01 -6.20130360e-01 -8.02035570e-01
-7.33258426e-02 1.43160000e-01 -6.38006389e-01 -4.61302966e-01
-6.98736072e-01 -9.97324944e-01 6.08822964e-02 5.23720741e-01
3.13419104e-01 -9.79582787e-01 -1.73369050e-01 2.97498554e-01
-2.77635425e-01 -1.14575171e+00 -5.76009512e-01 1.77584440e-01
-1.04383039e+00 -1.38055027e+00 -1.58882961e-01 -9.55347955e-01
6.22926295e-01 7.21121550e-01 9.29969549e-01 5.19255102e-01
-6.18339665e-02 4.20118153e-01 -6.42468214e-01 1.94645032e-01
-1.88880190e-01 2.85438150e-01 -1.72147807e-02 1.09015182e-01
4.04827297e-01 -9.62209940e-01 -6.61522210e-01 3.58443469e-01
-1.03881180e+00 1.63141266e-01 3.70446444e-01 4.28826302e-01
4.31939512e-01 9.19793248e-01 4.23549265e-01 -9.97915685e-01
5.36956549e-01 -4.45986539e-01 -7.73339033e-01 1.75376818e-01
-5.98693609e-01 4.14462015e-02 7.99748003e-01 -5.17206371e-01
-6.62517190e-01 1.53890073e-01 2.54588872e-01 -7.23084629e-01
1.44852176e-01 9.61678863e-01 -3.66115183e-01 2.70934880e-01
1.42212123e-01 3.27160150e-01 -6.52895272e-02 -3.65826130e-01
4.70845222e-01 1.71295166e-01 2.97932357e-01 -5.59521258e-01
8.69754612e-01 5.06144345e-01 -5.53874895e-02 -5.79193175e-01
-3.90768826e-01 -7.54876375e-01 -8.74945045e-01 -4.08642352e-01
6.76328063e-01 -7.33291090e-01 -1.05504894e+00 5.37493348e-01
-9.31885898e-01 -7.01374352e-01 1.76819652e-01 3.36975455e-01
-2.58073688e-01 7.70865619e-01 -7.09176838e-01 -6.66124463e-01
-8.81090853e-03 -1.09416199e+00 9.02416050e-01 -8.25818181e-02
3.12362635e-03 -1.27106261e+00 -2.69110221e-02 4.26977389e-02
1.04104087e-01 4.45725530e-01 1.04418004e+00 -5.00925481e-01
-7.60492504e-01 -1.06092922e-01 -8.83525014e-02 -3.18659663e-01
2.44295448e-01 2.01496154e-01 -4.03813243e-01 -6.19442284e-01
-2.81386703e-01 1.20987959e-01 8.82974625e-01 4.46169019e-01
1.39328408e+00 -1.89829692e-01 -7.70776749e-01 6.42453551e-01
1.35484743e+00 5.62735140e-01 5.60652792e-01 1.65364921e-01
1.18541944e+00 5.92161238e-01 5.51742613e-01 5.39308548e-01
4.88175809e-01 7.49055326e-01 5.46913266e-01 -5.38191013e-02
4.23420370e-02 -1.61848366e-01 2.86901027e-01 1.38636410e+00
1.03526056e-01 -6.84620142e-01 -1.11458147e+00 7.02460289e-01
-2.40532207e+00 -9.99391913e-01 -2.54559189e-01 1.93139637e+00
4.42606866e-01 2.03170255e-01 2.63768107e-01 3.29896480e-01
9.99193370e-01 3.59390199e-01 -6.75290585e-01 8.37392658e-02
1.95579067e-01 -1.86886817e-01 2.02014923e-01 3.18322390e-01
-1.11211133e+00 1.10682404e+00 5.77818489e+00 8.47176373e-01
-1.05840611e+00 -9.10042748e-02 5.83793104e-01 2.10593753e-02
-2.14777529e-01 1.28189862e-01 -2.06740335e-01 6.98814034e-01
9.95306075e-01 -3.88285190e-01 6.40917361e-01 6.62632346e-01
5.02560914e-01 2.95911103e-01 -1.25729358e+00 1.22863531e+00
-3.55932742e-01 -1.11709654e+00 1.02258317e-01 2.28977025e-01
5.43369830e-01 -1.33246854e-01 -9.49808285e-02 6.39738739e-01
6.63418531e-01 -7.35854149e-01 2.57620931e-01 3.06839436e-01
4.73547280e-01 -9.07467902e-01 5.84781706e-01 3.92600536e-01
-1.94242835e+00 -2.33252749e-01 -1.96757808e-01 -9.30278823e-02
1.10536315e-01 7.42755711e-01 -5.97622812e-01 8.92839372e-01
8.51298630e-01 1.17614853e+00 -7.58752942e-01 8.32937002e-01
6.14092126e-02 6.47164702e-01 -2.68594593e-01 6.51266947e-02
3.02879930e-01 -4.07497525e-01 1.73715189e-01 1.07470655e+00
2.39808872e-01 7.68353194e-02 6.94989145e-01 4.87005055e-01
-1.99888386e-02 1.55685529e-01 -6.62169874e-01 -4.76989388e-01
6.53607309e-01 1.22701800e+00 -1.35159433e+00 -2.80142009e-01
-3.25727552e-01 8.88605773e-01 4.57058966e-01 4.93140221e-01
-1.06780076e+00 -3.04943442e-01 4.22567397e-01 9.20423269e-02
3.79733801e-01 -6.99236572e-01 7.44694471e-02 -1.22637272e+00
4.99400385e-02 -8.28135729e-01 8.02423060e-01 -3.59684795e-01
-1.16094184e+00 4.20806408e-01 5.11487052e-02 -1.18954062e+00
-3.99298258e-02 -3.14461589e-01 -7.09722102e-01 1.57936037e-01
-7.29198515e-01 -1.11867142e+00 -3.91385078e-01 8.31132531e-01
3.75512332e-01 -5.31043820e-02 2.74518341e-01 5.32090306e-01
-9.41958070e-01 5.68980217e-01 1.07930966e-01 2.48539701e-01
4.50917155e-01 -1.23633134e+00 4.66994375e-01 1.22179639e+00
3.62698063e-02 8.32740426e-01 5.50897300e-01 -7.22664416e-01
-1.78300893e+00 -1.44078469e+00 3.75159025e-01 -2.70218074e-01
8.82268906e-01 -6.16185665e-01 -1.01991367e+00 6.43716156e-01
4.53038514e-02 -4.36571315e-02 5.41525066e-01 3.53451520e-01
-1.77171022e-01 -4.11702633e-01 -5.49319506e-01 8.32122922e-01
1.40554702e+00 -3.74309540e-01 1.39227331e-01 4.76317674e-01
1.21596694e+00 -1.63200244e-01 -9.19795334e-01 4.99269605e-01
1.59764498e-01 -8.47266853e-01 7.53954768e-01 -3.47616524e-01
-4.10668962e-02 -6.40045583e-01 1.44218609e-01 -1.06228542e+00
-6.08779967e-01 -8.74873161e-01 -3.73693496e-01 1.39691436e+00
6.88472316e-02 -4.17226732e-01 8.61846507e-01 2.74836093e-01
-1.90599233e-01 -6.31458819e-01 -6.41369462e-01 -9.57817137e-01
-2.83161968e-01 -6.16435111e-01 8.61979485e-01 1.21159124e+00
1.06167234e-01 5.10802567e-01 -2.49896482e-01 3.34576070e-01
5.04005194e-01 4.32135403e-01 9.27635729e-01 -1.17856002e+00
-3.68947476e-01 -6.09548628e-01 -5.66506565e-01 -9.85255897e-01
2.24784121e-01 -1.06502879e+00 -1.08353727e-01 -1.74364960e+00
3.12865376e-01 -3.49018365e-01 -4.26259011e-01 5.22192240e-01
-2.57252604e-01 -2.15826765e-01 -6.40807748e-02 1.91011235e-01
-1.17621171e+00 7.80523181e-01 1.19519615e+00 -3.35693508e-01
-4.73348856e-01 -1.49480104e-01 -6.37417018e-01 3.54618818e-01
7.48876750e-01 -1.80112466e-01 -1.04549384e+00 -3.50547731e-01
2.29894713e-01 2.55813360e-01 -2.38339659e-02 -1.01154029e+00
4.98664200e-01 -2.63567477e-01 -9.69621316e-02 -7.86225319e-01
-6.22728765e-02 -1.00773203e+00 4.75948453e-01 5.59475601e-01
-1.00516751e-01 3.28781754e-01 1.99618071e-01 1.09177434e+00
-2.83135474e-01 5.09663701e-01 5.87937593e-01 1.69772357e-02
-8.15863669e-01 8.02046895e-01 -4.84068066e-01 -1.16231024e-01
1.09116411e+00 -1.74574077e-01 -2.04811677e-01 -5.00034153e-01
-8.31792355e-01 8.18178713e-01 5.11656523e-01 5.27377129e-01
4.24693406e-01 -1.40814626e+00 -1.70478761e-01 -1.50583848e-01
1.64377272e-01 1.94292814e-01 5.43548644e-01 9.86063421e-01
-3.95368725e-01 3.13371569e-01 2.11967647e-01 -7.57144213e-01
-1.17919171e+00 1.28120017e+00 1.33794412e-01 -4.79937494e-01
-8.03131104e-01 3.99049133e-01 4.31334108e-01 -4.49377358e-01
4.09973979e-01 -3.31179202e-01 -2.28302330e-01 -6.75460994e-02
2.11048827e-01 2.48560011e-01 2.71961056e-02 -3.59967649e-01
-5.09334862e-01 3.91918421e-01 -1.90699652e-01 3.28965276e-01
1.25168967e+00 -2.75252759e-01 -3.04952353e-01 4.45350111e-01
1.08712006e+00 -2.90127844e-01 -1.06576169e+00 -1.10177279e-01
2.78060645e-01 -1.81321487e-01 -8.54798332e-02 -3.05020660e-01
-1.49008882e+00 7.94166565e-01 4.39369202e-01 5.21623850e-01
1.36817741e+00 2.31413320e-02 8.79515111e-01 4.19428051e-01
3.45404148e-01 -9.49349523e-01 2.93426365e-01 5.31220198e-01
4.58395720e-01 -1.13238609e+00 -2.10692119e-02 -5.29494464e-01
-4.00488347e-01 8.94354105e-01 8.83158743e-01 1.73556015e-01
8.86815369e-01 8.66068453e-02 -1.71208769e-01 -6.00185275e-01
-9.86098409e-01 -3.28167737e-01 1.09970858e-02 6.27394915e-01
4.83678132e-01 1.24028422e-01 -3.70209008e-01 5.35799742e-01
5.77277839e-02 -4.98455584e-01 3.35710645e-01 6.70728326e-01
-1.34261742e-01 -1.25073659e+00 1.25047013e-01 2.89229393e-01
-8.46794173e-02 1.45480901e-01 -5.32053351e-01 8.88751507e-01
-7.84523189e-02 1.09296870e+00 5.37031293e-02 -6.41058803e-01
9.84332934e-02 -3.54063064e-01 2.07282498e-01 -5.56988657e-01
-2.26933822e-01 4.29263234e-01 -1.26661226e-01 -7.85884261e-01
-4.85678196e-01 -4.89514768e-01 -1.38212240e+00 -5.84294498e-01
-2.27204353e-01 4.72386450e-01 9.24847648e-02 7.83487320e-01
5.56907177e-01 8.94214153e-01 6.28419578e-01 -6.69238985e-01
1.82962745e-01 -8.08395028e-01 -5.99221170e-01 4.56073880e-01
-2.18782723e-02 -7.56129503e-01 -3.03803295e-01 -4.32768092e-02] | [7.258851528167725, 5.996743202209473] |
c8221e35-ef3c-4caa-8338-7544f5503cf4 | semantic-parsing-of-interpage-relations | 2205.1353 | null | https://arxiv.org/abs/2205.13530v1 | https://arxiv.org/pdf/2205.13530v1.pdf | Semantic Parsing of Interpage Relations | Page-level analysis of documents has been a topic of interest in digitization efforts, and multimodal approaches have been applied to both classification and page stream segmentation. In this work, we focus on capturing finer semantic relations between pages of a multi-page document. To this end, we formalize the task as semantic parsing of interpage relations and we propose an end-to-end approach for interpage dependency extraction, inspired by the dependency parsing literature. We further design a multi-task training approach to jointly optimize for page embeddings to be used in segmentation, classification, and parsing of the page dependencies using textual and visual features extracted from the pages. Moreover, we also combine the features from two modalities to obtain multimodal page embeddings. To the best of our knowledge, this is the first study to extract rich semantic interpage relations from multi-page documents. Our experimental results show that the proposed method increased LAS by 41 percentage points for semantic parsing, increased accuracy by 33 percentage points for page stream segmentation, and 45 percentage points for page classification over a naive baseline. | ['Onur Deniz', 'Mehmet Yasin Akpınar', 'Berke Oral', 'Mehmet Arif Demirtaş'] | 2022-05-26 | null | null | null | null | ['page-stream-segmentation'] | ['natural-language-processing'] | [ 3.73090655e-01 2.23718345e-01 -3.70469987e-01 -4.29324150e-01
-1.23820794e+00 -7.33986318e-01 6.83373094e-01 6.37921810e-01
-3.64107311e-01 3.20643693e-01 3.88308138e-01 -2.57143438e-01
-6.81563979e-03 -5.14746428e-01 -6.80253148e-01 -3.57022464e-01
-7.59077221e-02 5.44864178e-01 3.27487320e-01 2.11123437e-01
3.01000386e-01 2.29475141e-01 -1.35187399e+00 6.93163991e-01
7.97948122e-01 1.07675552e+00 8.13176036e-02 1.03009045e+00
-7.40770102e-01 4.62197542e-01 -5.98456502e-01 -6.82334363e-01
-3.86725426e-01 -2.67235041e-01 -1.28494287e+00 4.95145380e-01
4.24492210e-01 -6.46306053e-02 1.23602495e-01 8.85591686e-01
2.39077717e-01 -4.97541428e-02 7.97046483e-01 -1.14182794e+00
-3.84938627e-01 6.20832860e-01 -8.07165921e-01 -3.96032855e-02
6.52311087e-01 -7.49898553e-01 1.62427306e+00 -9.41285908e-01
6.20928705e-01 1.38430834e+00 2.72650421e-01 1.52833790e-01
-1.02842987e+00 -1.09780394e-01 3.32146347e-01 1.48192286e-01
-8.51004362e-01 -1.51093140e-01 7.00793505e-01 -3.00850958e-01
1.04761398e+00 4.60808694e-01 3.87098581e-01 1.02583861e+00
-8.65930691e-02 1.38003850e+00 1.00324631e+00 -7.43053734e-01
9.32488292e-02 2.27337614e-01 7.73273766e-01 6.35126531e-01
2.01503992e-01 -9.99885082e-01 -5.98993421e-01 4.56768693e-03
4.01322812e-01 -3.54592264e-01 -1.89393520e-01 -2.90017575e-01
-9.62933958e-01 8.17463458e-01 5.54814152e-02 2.64841765e-01
-2.23798290e-01 -5.95806465e-02 4.85006511e-01 -2.40889460e-01
5.37387609e-01 2.40813091e-01 -5.29713333e-01 -3.86745483e-01
-6.87001050e-01 1.26814060e-02 1.21950519e+00 9.67813253e-01
5.58343649e-01 -7.44117856e-01 1.79480147e-02 1.02739823e+00
6.40942931e-01 3.83224994e-01 2.02018037e-01 -7.93493867e-01
1.09740806e+00 7.18549967e-01 -1.91498473e-02 -9.63612020e-01
-5.41487873e-01 -2.38168500e-02 -3.20376128e-01 -4.10539001e-01
3.00810486e-01 4.74628136e-02 -8.34592283e-01 1.16662014e+00
4.10039395e-01 -2.36859754e-01 2.20529407e-01 7.11673200e-01
5.70520282e-01 7.43223429e-01 4.10901040e-01 -8.42417702e-02
2.06183147e+00 -9.98111129e-01 -8.98904622e-01 -4.12147194e-01
9.21774447e-01 -1.01073909e+00 1.25382400e+00 3.16599876e-01
-8.79759192e-01 -3.02802324e-01 -1.09804559e+00 -1.42647013e-01
-5.11579692e-01 3.00829381e-01 6.59465015e-01 6.59101903e-01
-5.56381106e-01 1.50514558e-01 -8.73077631e-01 -6.91573143e-01
4.80279565e-01 2.23139644e-01 -3.68963242e-01 -1.75824657e-01
-1.05814350e+00 5.47080696e-01 3.83445442e-01 -1.25946447e-01
-9.78916213e-02 -3.90970230e-01 -8.92814517e-01 2.38466308e-01
6.36278689e-01 -4.43403512e-01 1.34190893e+00 -6.60424590e-01
-1.11384881e+00 7.39088774e-01 -3.68205398e-01 -1.93292946e-01
-3.36675160e-02 -8.29473376e-01 -4.49219316e-01 6.09118104e-01
2.61110812e-01 4.54751879e-01 5.73482871e-01 -1.62224448e+00
-7.81597614e-01 -6.34752512e-01 9.72958431e-02 3.96294653e-01
-7.22512484e-01 2.43776530e-01 -1.17530370e+00 -3.99480313e-01
5.82395978e-02 -9.35138106e-01 7.81967193e-02 -4.91621554e-01
-8.36077332e-01 -2.67409533e-01 1.01003301e+00 -1.07358897e+00
1.21143126e+00 -2.04669428e+00 1.51250616e-01 1.88499779e-01
1.44549891e-01 6.49769325e-03 -1.19103663e-01 4.89633709e-01
8.73513371e-02 4.90082800e-01 -3.47500771e-01 -6.82301998e-01
3.18518817e-01 2.81474292e-01 -2.46487960e-01 8.36401450e-05
2.88189858e-01 8.07381630e-01 -6.36353374e-01 -9.88309622e-01
-1.16552629e-01 5.38129866e-01 -3.21824133e-01 3.20571035e-01
-3.39042783e-01 -1.53294921e-01 -7.46433973e-01 7.06689835e-01
6.26542866e-01 -4.14495260e-01 5.15731096e-01 -3.81672233e-01
1.73788935e-01 2.14782074e-01 -9.62380052e-01 1.60689008e+00
-5.35748959e-01 6.20222092e-01 7.55185587e-03 -7.62685061e-01
6.68989182e-01 3.19533944e-01 4.87066001e-01 -7.40728974e-01
1.44425735e-01 -1.10123090e-01 -5.24502993e-01 -7.31683791e-01
8.03107679e-01 2.35545814e-01 -6.01475954e-01 4.28068459e-01
-4.08483781e-02 1.53222293e-01 5.60291946e-01 4.94203806e-01
9.92668629e-01 1.43777549e-01 -2.29514926e-03 2.77318448e-01
5.93497753e-01 1.01551071e-01 1.08991943e-01 4.11155283e-01
1.14804819e-01 4.71886694e-01 1.06085181e+00 1.50268123e-01
-8.01357627e-01 -1.09135807e+00 6.40753955e-02 1.24397731e+00
1.69750601e-01 -7.22880840e-01 -1.11491573e+00 -1.11523879e+00
-1.39557913e-01 5.45836926e-01 -3.52183342e-01 3.39892030e-01
-7.75499523e-01 -9.65299606e-01 3.94945681e-01 8.56399894e-01
3.85697156e-01 -6.58810735e-01 -3.33904237e-01 1.27179503e-01
-4.56648588e-01 -1.88721645e+00 -2.89433926e-01 2.26837635e-01
-6.95946932e-01 -1.31406546e+00 -5.40977478e-01 -1.04770136e+00
6.02053642e-01 -2.09330171e-02 9.06659782e-01 -3.09742987e-01
-2.43101850e-01 8.51242840e-01 -6.86834931e-01 -9.26959291e-02
-7.48838410e-02 4.00608689e-01 -4.45728242e-01 8.13816935e-02
4.03316379e-01 -1.08790137e-01 -3.86417687e-01 2.46838510e-01
-1.08561218e+00 9.93918628e-02 8.66842210e-01 5.19423485e-01
6.42069221e-01 7.88634717e-02 1.66969761e-01 -1.29924190e+00
8.95845532e-01 -4.43967372e-01 -2.58609295e-01 4.83204365e-01
-5.60612798e-01 1.91361904e-01 4.35310185e-01 -1.06089883e-01
-1.43186140e+00 1.14362866e-01 -5.12780063e-02 1.75692350e-01
-3.70534569e-01 6.84983373e-01 -5.30413032e-01 3.21436882e-01
1.07923364e-02 -6.82497099e-02 -4.38310385e-01 -6.86902642e-01
7.03725040e-01 1.01567411e+00 4.02029574e-01 -7.12618351e-01
5.39900899e-01 4.61600214e-01 -3.95164713e-02 -9.48963940e-01
-9.83772755e-01 -7.58027554e-01 -8.47911119e-01 -1.21661112e-01
1.43492222e+00 -6.72941983e-01 -6.29948318e-01 2.69640684e-01
-1.20897889e+00 -9.88436937e-02 2.44950026e-01 3.52068901e-01
-4.15456027e-01 9.07978475e-01 -7.37157106e-01 -6.88523650e-01
-2.90510178e-01 -9.70685780e-01 1.34309685e+00 2.85706997e-01
-4.42134172e-01 -1.01234651e+00 9.48033109e-02 9.40849364e-01
-2.32720762e-01 -1.52894137e-02 1.15567172e+00 -9.00072694e-01
-4.99619275e-01 -3.38143706e-01 -6.82272911e-01 6.61343038e-02
-8.61657085e-04 -6.14334121e-02 -9.98573005e-01 1.23207197e-01
-2.83671319e-01 -1.90107912e-01 8.50039244e-01 4.68983091e-02
9.31296885e-01 -1.69045463e-01 -5.73776126e-01 3.16119283e-01
1.41595912e+00 1.43493727e-01 5.33473790e-01 6.65276110e-01
1.06406677e+00 9.50998068e-01 9.68645930e-01 5.42662621e-01
6.95044994e-01 5.27928948e-01 2.92373478e-01 5.15615828e-02
-1.35339424e-01 -1.44216955e-01 2.41857201e-01 8.92750084e-01
3.23022336e-01 -7.48794138e-01 -1.06081641e+00 5.74375451e-01
-1.87119806e+00 -4.06484455e-01 -4.48910385e-01 1.62119722e+00
6.66582584e-01 2.25858733e-01 -2.31526047e-02 1.48932070e-01
7.65174389e-01 2.73409367e-01 -1.12616224e-02 -5.22250533e-01
1.52835190e-01 -2.10491661e-02 5.71350694e-01 3.69148105e-01
-1.43821549e+00 9.69562173e-01 5.28346443e+00 7.05050826e-01
-5.17610490e-01 -7.51420707e-02 5.94006300e-01 2.72356600e-01
-2.50945300e-01 8.86079296e-02 -1.11648619e+00 2.76119798e-01
1.06091106e+00 3.46158355e-01 -4.34853844e-02 6.97585762e-01
-8.13023373e-02 -3.69054288e-01 -1.00840163e+00 7.84467518e-01
4.07418847e-01 -8.62019420e-01 9.17186290e-02 1.01033650e-01
4.20996606e-01 -5.18667221e-01 -1.18528880e-01 3.81273739e-02
4.21890169e-02 -7.79795349e-01 4.17463720e-01 2.86925673e-01
3.07187438e-01 -8.70598853e-01 8.05468321e-01 -9.66619328e-02
-1.39122140e+00 1.64699107e-01 1.12731762e-01 4.08789366e-01
5.29673398e-01 6.42518640e-01 -8.73831928e-01 7.10025549e-01
6.38673723e-01 4.69491154e-01 -8.68977368e-01 7.73528934e-01
-3.71668220e-01 5.61912298e-01 -7.47292489e-02 -2.45501056e-01
1.63458034e-01 -4.71564606e-02 2.47270063e-01 1.60506237e+00
4.14645188e-02 -1.81042686e-01 2.66573548e-01 2.34868601e-01
-1.81276366e-01 2.89119035e-01 -1.40910596e-01 -6.04640186e-01
3.19445789e-01 1.56153691e+00 -1.10298550e+00 -3.68101984e-01
-6.80837512e-01 1.31780171e+00 2.38058880e-01 4.08274412e-01
-8.36426914e-01 -8.04356396e-01 4.16196674e-01 -2.33675554e-01
5.58336496e-01 -4.53886509e-01 -5.53801477e-01 -1.00694120e+00
3.55019301e-01 -4.85626906e-01 6.01748586e-01 -7.39299893e-01
-1.06824911e+00 5.67461610e-01 7.69079402e-02 -7.88095117e-01
-7.03328550e-02 -8.25264096e-01 -3.30625325e-01 5.31103551e-01
-1.56820011e+00 -1.43787777e+00 -3.10751021e-01 2.30981231e-01
6.23836458e-01 2.29582071e-01 8.45576763e-01 1.99332133e-01
-7.71675885e-01 3.82558733e-01 -7.45459720e-02 4.44135636e-01
6.55188918e-01 -1.54754925e+00 3.02459955e-01 8.42144012e-01
2.31145307e-01 3.71964425e-01 5.01508534e-01 -7.41840899e-01
-1.66371036e+00 -9.76424813e-01 1.22775280e+00 -2.98796684e-01
8.18616927e-01 -5.99278688e-01 -9.91411507e-01 6.43080115e-01
4.33825970e-01 -4.67330426e-01 1.02842438e+00 4.08546120e-01
-4.97899234e-01 8.98944214e-02 -7.50353873e-01 6.52267933e-01
6.15027487e-01 -6.46018147e-01 -5.51157653e-01 2.95953065e-01
7.52079487e-01 -2.54445642e-01 -1.19202673e+00 5.03249327e-03
5.11902094e-01 -4.67258930e-01 1.04634631e+00 -2.18382135e-01
6.05638087e-01 -2.16481034e-02 -2.26383105e-01 -9.54373300e-01
2.28814012e-03 -1.95621792e-02 -2.08454326e-01 1.86605072e+00
6.88979030e-01 -3.91276836e-01 8.40908289e-01 6.47705674e-01
-9.92580503e-02 -6.38862431e-01 -5.51877260e-01 -4.01966423e-01
-2.14692339e-01 -6.26896322e-01 2.63450861e-01 5.62936246e-01
2.98726171e-01 6.56261981e-01 -6.16718922e-03 5.03602743e-01
6.54017329e-01 3.63107324e-01 6.48563802e-01 -9.81196702e-01
-3.60793054e-01 -3.49549264e-01 -5.38328774e-02 -1.16030049e+00
2.53161371e-01 -7.03096688e-01 1.05810523e-01 -1.99770987e+00
3.64566714e-01 -1.15827695e-01 -2.35361867e-02 3.00748587e-01
-2.63997644e-01 4.21192534e-02 3.96033853e-01 2.73658454e-01
-9.75517154e-01 3.13969672e-01 9.87777054e-01 -2.16078490e-01
-1.26914427e-01 -3.19591194e-01 -7.62155473e-01 5.82194269e-01
9.55222070e-01 -3.93275976e-01 -3.02246749e-01 -4.02923226e-01
2.39478916e-01 1.62764251e-01 8.32477510e-02 -6.65215075e-01
1.48885444e-01 5.43082915e-02 4.53311682e-01 -7.61672735e-01
4.57946360e-01 -7.41017997e-01 -5.98781645e-01 -8.93192366e-02
-3.47607404e-01 -4.90916707e-02 2.57493526e-01 7.67780602e-01
-3.72003883e-01 -3.34851295e-01 2.03875571e-01 1.08764201e-01
-8.81424963e-01 -1.24424241e-01 -5.74896932e-01 1.10289291e-01
9.96434748e-01 2.11358443e-02 -4.23276752e-01 -8.95496085e-02
-7.78192997e-01 3.54270548e-01 2.12763056e-01 5.36500156e-01
5.23363590e-01 -1.02368999e+00 -2.00185895e-01 -7.52735734e-02
1.58329278e-01 -2.28317063e-02 2.76052225e-02 7.86999285e-01
-5.15942872e-01 7.49182999e-01 2.38475472e-01 -5.13309598e-01
-1.69929349e+00 2.66685724e-01 -4.97944564e-01 -5.47981024e-01
-3.74356627e-01 6.82629824e-01 1.25839621e-01 -2.40386531e-01
3.06915581e-01 -4.05766755e-01 -5.75685680e-01 6.96717024e-01
2.98851967e-01 2.70093352e-01 9.81868878e-02 -5.84592283e-01
-4.44794983e-01 8.14744949e-01 -3.78489673e-01 -4.26079482e-01
1.17615891e+00 -3.58251065e-01 -7.07332417e-02 4.57602888e-01
1.60954118e+00 1.69906616e-02 -9.53331888e-01 2.35418957e-02
6.45511925e-01 -3.27642888e-01 -5.04923388e-02 -8.90299201e-01
-7.04486668e-01 9.41592813e-01 2.76013225e-01 5.22271335e-01
1.08447683e+00 5.04173398e-01 1.12460732e+00 1.45776764e-01
-7.00824261e-02 -1.33910489e+00 1.65694743e-01 3.90307635e-01
4.98036176e-01 -1.19069910e+00 -8.68340284e-02 -8.94763350e-01
-8.60920489e-01 1.29134703e+00 6.16866589e-01 2.77234167e-01
4.13525969e-01 2.77825624e-01 -7.01532364e-02 -2.59618789e-01
-5.20740032e-01 -3.83649200e-01 4.33277965e-01 4.59983706e-01
6.22109711e-01 2.92558484e-02 -3.06870013e-01 7.79660225e-01
1.52454913e-01 -3.44294548e-01 2.90063173e-01 1.24127853e+00
-5.46800315e-01 -1.44934821e+00 -4.88729060e-01 3.17567408e-01
-5.87091744e-01 -1.03838019e-01 -6.09444916e-01 8.74734104e-01
-3.71227115e-01 1.00512826e+00 1.85524985e-01 -1.90999776e-01
3.62184197e-01 3.46903652e-01 4.10611868e-01 -5.72956026e-01
-2.77323306e-01 3.72977734e-01 6.83359444e-01 -3.81326675e-01
-4.45154190e-01 -7.08504379e-01 -1.70608044e+00 1.93738684e-01
-9.70595255e-02 2.65740454e-01 1.00558698e+00 9.67808843e-01
3.42842847e-01 6.73028052e-01 4.93149340e-01 -6.93368673e-01
-6.89658746e-02 -6.53924644e-01 -4.20709878e-01 4.44311947e-01
-7.60068744e-02 -3.47681612e-01 -3.00765276e-01 2.74526447e-01] | [11.610614776611328, 2.769895315170288] |
f0f47ca0-d7b3-420b-89e2-76860b718371 | sleepposenet-multi-view-learning-for-sleep | 2005.02176 | null | https://arxiv.org/abs/2005.02176v2 | https://arxiv.org/pdf/2005.02176v2.pdf | SleepPoseNet: Multi-View Learning for Sleep Postural Transition Recognition Using UWB | Recognizing movements during sleep is crucial for the monitoring of patients with sleep disorders, and the utilization of ultra-wideband (UWB) radar for the classification of human sleep postures has not been explored widely. This study investigates the performance of an off-the-shelf single antenna UWB in a novel application of sleep postural transition (SPT) recognition. The proposed Multi-View Learning, entitled SleepPoseNet or SPN, with time series data augmentation aims to classify four standard SPTs. SPN exhibits an ability to capture both time and frequency features, including the movement and direction of sleeping positions. The data recorded from 38 volunteers displayed that SPN with a mean accuracy of $73.7 \pm 0.8 \%$ significantly outperformed the mean accuracy of $59.9 \pm 0.7 \%$ obtained from deep convolution neural network (DCNN) in recent state-of-the-art work on human activity recognition using UWB. Apart from UWB system, SPN with the data augmentation can ultimately be adopted to learn and classify time series data in various applications. | ['Theerawit Wilaiprasitporn', 'Subhas Chandra Mukhopadhyay', 'Nattee Niparnan', 'Supasorn Suwajanakorn', 'Nakorn Kumchaiseemak', 'Theerasarn Pianpanit', 'Pitsharponrn Leelaarporn', 'Payongkit Lakhan', 'Patchanon Warin', 'Maytus Piriyajitakonkij'] | 2020-05-02 | null | null | null | null | ['multi-view-learning'] | ['computer-vision'] | [ 3.4552169e-01 -1.9534017e-01 8.0644295e-02 -4.8591748e-01
-4.0964890e-01 1.9607385e-01 -2.3044689e-02 -3.7952083e-01
-5.5471987e-01 7.5662869e-01 1.8635295e-02 -1.3238871e-01
-4.2391101e-01 -5.7883376e-01 -1.1966287e-01 -1.0587217e+00
-4.0157327e-01 -6.9864161e-02 -1.9777445e-01 -3.8086978e-01
-1.5373376e-02 1.4410606e-01 -1.6426909e+00 1.4729521e-01
6.4195579e-01 1.7171214e+00 -6.3307501e-02 4.9324256e-01
1.8479867e-01 2.2825828e-01 -1.0276489e+00 2.9799500e-01
-2.0477591e-02 -3.9016148e-01 -2.0426778e-01 -3.9743692e-01
2.8923529e-01 1.4426593e-01 -2.5445831e-01 5.9909880e-01
1.0599735e+00 4.0972346e-01 2.5216952e-01 -1.0689719e+00
-5.7847029e-01 1.7968026e-01 -4.4438282e-01 1.0264581e+00
4.2006686e-01 -9.7892597e-02 3.6121318e-01 -4.6890265e-01
-1.9239593e-01 3.0879551e-01 1.1067846e+00 5.7180500e-01
-6.5151536e-01 -8.6055410e-01 -5.1943254e-01 6.4312023e-01
-1.2750065e+00 -4.1742903e-01 9.3178207e-01 4.8524749e-02
1.1446804e+00 4.5406419e-01 1.2656275e+00 1.6315441e+00
9.6166009e-01 3.2554224e-01 1.1198635e+00 -2.2459075e-01
2.8291774e-01 -3.8447258e-01 4.3531057e-01 6.5891761e-01
2.0563358e-01 -3.9691284e-02 -8.5160220e-01 7.0124604e-02
2.5949052e-01 2.9232830e-01 -2.0165898e-01 2.0790333e-01
-8.7480664e-01 4.6442053e-01 4.7939321e-01 6.0475367e-01
-4.3458211e-01 2.3446005e-02 3.9950943e-01 3.9047066e-02
7.2822946e-01 3.3024901e-01 -4.3920743e-01 -7.1962792e-01
-1.1063648e+00 -2.8307599e-01 5.5675328e-01 3.9796004e-01
3.2695796e-02 6.2937152e-01 -2.6011223e-01 8.7298191e-01
9.6426427e-02 8.7069952e-01 1.0361980e+00 -7.7523160e-01
9.1823518e-02 5.1061213e-01 -5.0791789e-02 -7.8975689e-01
-1.3288907e+00 -9.0235341e-01 -9.9827474e-01 -2.9106432e-01
-7.3584286e-03 -3.8819030e-01 -7.6423138e-01 1.5607342e+00
2.8695336e-02 2.6464784e-01 2.4826321e-01 6.8828237e-01
1.1812500e+00 5.0000685e-01 -2.0841670e-01 -5.5900717e-01
1.7549263e+00 -6.7933583e-01 -1.0201091e+00 -5.9112304e-01
2.5582543e-01 -3.6025473e-01 6.4554244e-01 6.7828006e-01
-9.3668360e-01 -9.1317427e-01 -1.4244095e+00 4.4753093e-01
-1.4732717e-01 2.3133212e-01 3.7159452e-01 1.0649554e+00
-8.1163037e-01 4.1068327e-01 -1.3576945e+00 -6.4364707e-01
4.8673770e-01 6.3914782e-01 -1.1604996e-01 3.6884311e-01
-1.3076978e+00 1.0127262e+00 -1.8230812e-01 5.8895439e-01
-4.1018698e-01 -7.5978476e-01 -7.7501649e-01 8.8507801e-02
2.4725893e-01 -7.2420794e-01 1.2514952e+00 -4.3893087e-01
-1.4477282e+00 4.5928177e-01 -5.9562173e-02 -8.9116663e-01
-1.3734024e-02 -5.2839148e-01 -1.3502977e+00 2.1434149e-01
7.5856730e-02 -2.0862368e-01 8.5562164e-01 -2.7230704e-01
-3.3736816e-01 -6.9341487e-01 -5.9349686e-01 -2.7726495e-01
-4.2167157e-01 -2.6710236e-01 3.0485618e-01 -4.2872438e-01
2.1638225e-01 -1.0049288e+00 2.6519603e-01 -6.8452954e-02
-6.3656032e-02 -3.3660680e-01 9.2470217e-01 -6.0253936e-01
1.2525959e+00 -1.9782686e+00 -2.6553047e-01 4.1108835e-02
2.9015502e-01 4.4087538e-01 3.7279302e-01 2.7656546e-01
-1.4875962e-02 -6.0752541e-01 -2.0015900e-01 -2.0110735e-01
-2.2618255e-01 5.1726413e-01 2.4176919e-01 8.4200215e-01
-3.4326470e-01 7.6972079e-01 -5.3134996e-01 -2.7550256e-02
2.1063755e-01 4.7972390e-01 -2.7282834e-03 8.3217509e-03
3.4628579e-01 3.4437439e-01 -3.3190098e-01 6.6296500e-01
4.0969300e-01 -1.8414201e-01 -2.1989094e-01 -3.6756918e-01
-3.0160329e-01 -1.2271835e-01 -7.0814985e-01 1.8543438e+00
-7.2194386e-01 7.6707280e-01 -1.6228169e-01 -1.2207283e+00
1.1064512e+00 2.9347277e-01 9.1974115e-01 -1.2298511e+00
4.1825172e-01 -6.8199724e-02 3.9364871e-01 -1.0123340e+00
1.1860516e-01 -2.1653648e-01 -2.7018601e-01 5.2348208e-01
2.0346193e-01 5.9748513e-01 -1.6108470e-01 -5.3589159e-01
1.3559340e+00 -2.0109721e-01 4.9033245e-01 -4.3326989e-01
4.7723007e-01 -4.9858853e-01 4.8769590e-01 5.6940907e-01
-4.5143619e-01 2.7148378e-01 -5.6579612e-02 -7.1597683e-01
-3.0208167e-01 -1.1966789e+00 -2.3544998e-01 9.7489858e-01
4.8248298e-02 -1.9526631e-01 -6.3198531e-01 -3.3458695e-01
-2.0887610e-01 7.1526581e-01 -8.9869094e-01 -6.9374335e-01
-5.0992566e-01 -1.0130099e+00 1.0141275e+00 9.5642275e-01
9.6163291e-01 -1.1621587e+00 -1.6323665e+00 2.2890791e-01
-2.5241303e-01 -1.0833006e+00 -3.8059273e-01 5.9835505e-01
-6.8964249e-01 -1.0792286e+00 -4.9536785e-01 -4.1709638e-01
5.2179709e-02 3.1381613e-01 5.7784665e-01 -3.5017458e-01
-6.0343325e-01 5.9896821e-01 -4.3150246e-01 -8.7906700e-01
2.9931536e-01 3.1178929e-02 7.6418477e-01 2.6391998e-01
6.7500347e-01 -1.2517346e+00 -1.0996913e+00 3.3721250e-01
-3.3134317e-01 -4.3265104e-01 7.9619211e-01 5.1306468e-01
3.5354584e-01 -2.2877462e-01 8.1585264e-01 -3.0990651e-01
7.3925221e-01 -4.4796732e-01 1.1653639e-02 1.7560983e-01
-6.6773981e-01 -2.3652187e-01 5.7370639e-01 -3.1750304e-01
-7.5702989e-01 -4.4261044e-01 -4.5924065e-01 -2.9997310e-01
-2.6324487e-01 3.7459496e-01 1.6055818e-01 -5.4205529e-02
8.4718424e-01 8.3579308e-01 3.3645672e-01 -3.5308567e-01
-2.4567109e-01 7.9285222e-01 8.1147778e-01 -1.8442132e-02
7.2660230e-02 5.7911336e-01 1.1338674e-01 -1.5550357e+00
-1.1610576e+00 -6.7868072e-01 -4.0943322e-01 -3.5692111e-01
1.2867329e+00 -7.4053484e-01 -9.1811341e-01 4.1404253e-01
-5.9485233e-01 -1.4263786e-01 -1.6006649e-01 6.6959578e-01
-6.7008585e-01 2.5677538e-01 5.3794645e-02 -9.0470022e-01
-1.1134593e+00 -5.2596039e-01 7.8169852e-01 4.5915347e-01
-7.3460317e-01 -6.6611987e-01 1.2604149e-01 6.1351007e-01
7.7212185e-01 3.4202653e-01 5.6003481e-01 -8.3945680e-01
3.6284867e-01 -3.3175471e-01 4.2097163e-01 3.8096941e-01
4.0450585e-01 -5.8840430e-01 -1.1314914e+00 -2.4677850e-01
7.2276175e-01 3.9526168e-02 5.3023070e-01 8.3555257e-01
1.1705871e+00 -4.9702622e-02 -2.6278093e-01 7.8807366e-01
9.1720760e-01 6.3853288e-01 5.7398027e-01 2.8931117e-01
2.3523885e-01 -4.1761018e-02 3.2018605e-01 7.2951090e-01
1.4637388e-01 6.1485374e-01 4.9061507e-01 2.8016296e-01
2.3174417e-01 5.4812056e-01 2.9831067e-01 7.1930009e-01
-5.3291452e-01 6.9396924e-03 -8.8583136e-01 1.0968515e-01
-1.5160767e+00 -1.1751068e+00 4.4176973e-02 1.7098600e+00
1.2811039e-01 3.7467763e-01 3.0319116e-01 4.1396940e-01
2.4075341e-01 4.4963717e-01 -6.2107754e-01 -6.5572774e-01
4.2941660e-01 6.9865960e-01 2.9923925e-01 -3.6566970e-01
-1.1229097e+00 1.1124986e-02 5.4235606e+00 2.8903791e-01
-1.3553808e+00 3.5821271e-01 -7.1638942e-02 -6.0776383e-01
6.1765277e-01 -1.0857638e+00 -4.7334006e-01 8.8754088e-01
1.7269022e+00 1.4502449e-01 2.5151652e-01 9.4670659e-01
5.0314867e-01 -2.8788921e-01 -7.9933155e-01 1.4601202e+00
5.5580431e-01 -1.0540247e+00 -8.1586909e-01 9.4003901e-02
2.6560679e-01 1.9575854e-01 2.2141699e-01 4.4972107e-01
-8.3704549e-01 -8.2853317e-01 3.6509383e-01 9.1056907e-01
6.6052008e-01 -7.9591566e-01 1.0371485e+00 4.7951919e-01
-1.3019048e+00 -5.0246370e-01 -7.5974055e-02 -2.8499725e-01
2.7009717e-01 3.1757128e-01 -4.9032974e-01 6.5840733e-01
1.1920607e+00 7.2236788e-01 -5.7339180e-01 9.2944181e-01
2.9575055e-02 6.3032329e-01 -4.0043646e-01 -2.6224357e-01
1.2427657e-01 4.1343048e-03 3.5618091e-01 8.5720688e-01
8.1413019e-01 2.4147741e-01 -5.7448413e-02 4.2965534e-01
3.1689829e-01 -4.9895257e-01 -5.8577222e-01 1.7922069e-01
1.1691463e-01 1.2095515e+00 -7.2115862e-01 6.2642105e-02
-4.1963336e-01 7.6788276e-01 -3.6352879e-01 -3.4712683e-02
-1.0743732e+00 -3.7123114e-01 7.3295420e-01 1.6770117e-01
3.7451449e-01 -3.3842793e-01 -3.3081502e-01 -5.7325590e-01
1.7318937e-01 -2.9394311e-01 5.2017850e-01 -9.2334098e-01
-1.1035515e+00 8.7157440e-01 8.4945641e-02 -1.7124372e+00
2.0784488e-02 -5.1341259e-01 -9.6557021e-01 6.0136193e-01
-8.5227937e-01 -9.9995059e-01 -8.6354244e-01 5.3529924e-01
8.8358784e-01 -2.1381961e-01 1.0386883e+00 2.7367830e-01
-7.2024506e-01 6.4186150e-01 -2.6878871e-02 -2.6457384e-01
4.7200018e-01 -9.9916929e-01 1.6501202e-01 5.2386260e-01
6.7111254e-02 7.0847058e-01 7.2609228e-01 -2.7930820e-01
-1.3577977e+00 -9.8257238e-01 5.0063670e-01 -3.7841311e-01
2.6515332e-01 -9.9226378e-02 -5.9036642e-01 3.8521445e-01
3.7042612e-01 2.6487368e-01 1.5163615e+00 9.3334369e-02
2.7994636e-01 -7.2895175e-01 -1.2315753e+00 1.5822355e-01
1.0356604e+00 -2.0572844e-01 -9.4257176e-01 -2.1518556e-02
9.4939597e-02 -4.7904080e-01 -9.5054281e-01 4.0285361e-01
1.1339750e+00 -9.5571679e-01 1.0206757e+00 -3.9507619e-01
6.4642741e-03 6.5159872e-02 -2.0011082e-01 -1.3698442e+00
-2.4418411e-01 -4.0044919e-01 -6.7975938e-01 6.1196870e-01
-3.6541816e-02 -7.1076059e-01 8.2440430e-01 1.4043489e-01
-8.5753298e-01 -1.1511853e+00 -1.5777385e+00 -7.5226992e-01
-7.0329541e-01 -6.7765528e-01 2.4451078e-01 2.6606894e-01
2.2752697e-02 4.4911784e-01 -5.9465498e-01 2.0513891e-01
4.3569359e-01 -3.1459417e-02 2.9780099e-01 -1.4854450e+00
5.6306962e-02 -4.5501713e-02 -6.6716564e-01 -4.5212346e-01
-1.8238303e-01 -5.7001793e-01 -1.1021183e-01 -1.7765701e+00
-3.7138394e-01 2.4800554e-01 -8.4496641e-01 3.7823367e-01
4.7617921e-01 4.0449604e-01 -2.8895974e-01 -2.1983834e-01
-6.2914956e-01 7.4558359e-01 1.0066645e+00 -1.6186681e-01
-3.8659906e-01 9.4737297e-01 -6.8570995e-01 8.5207206e-01
1.0201198e+00 -3.5578641e-01 -7.6621842e-01 -7.9558410e-02
-6.5147556e-02 2.4488352e-01 4.2703366e-01 -1.8145806e+00
3.8291377e-01 2.8489366e-01 8.3412170e-01 -7.6302099e-01
1.0299138e+00 -8.3121663e-01 8.1075169e-02 8.0246514e-01
1.8598577e-01 3.7236255e-01 3.0916250e-01 7.5531948e-01
2.5895160e-01 2.0309445e-01 7.4652100e-01 -1.4063366e-01
-7.8866899e-01 7.3271967e-02 -8.6609262e-01 -9.9023163e-02
1.1363330e+00 -8.0120188e-01 -4.3585521e-01 -2.3362090e-01
-1.1314207e+00 5.8433395e-02 -5.4369360e-01 4.7105810e-01
8.1007218e-01 -1.3579421e+00 3.6686361e-02 6.3586003e-01
2.0691244e-01 -4.2248696e-01 8.7157160e-01 1.4883456e+00
-2.6704144e-01 5.6385797e-01 -7.3588741e-01 -7.3228496e-01
-1.3513614e+00 1.4899395e-01 5.2840024e-01 1.8374255e-01
-1.0299375e+00 7.2098261e-01 -6.6392463e-01 1.8439078e-01
2.3376133e-01 -7.1268284e-01 -5.3270310e-01 3.3357581e-01
5.0584924e-01 8.6724341e-01 5.4687208e-01 -4.4975138e-01
-7.9669589e-01 5.6550926e-01 2.4945998e-01 2.3262104e-01
1.7393491e+00 4.8799835e-02 1.2671399e-01 6.9161916e-01
9.6176827e-01 -5.1380402e-01 -8.3539593e-01 2.1708404e-01
-1.3290554e-01 1.6341661e-01 -7.6972581e-02 -8.9782465e-01
-9.1973257e-01 9.4291705e-01 1.5422423e+00 4.3285945e-01
1.2580849e+00 -1.8228754e-01 1.3390555e+00 6.6029567e-01
5.3809106e-01 -1.1582441e+00 2.7064639e-01 2.1718569e-01
6.8552428e-01 -8.8234425e-01 3.8901180e-02 2.4673885e-01
-5.1888448e-01 1.3516043e+00 6.7274553e-01 -8.9548171e-02
9.7727674e-01 1.9233268e-03 1.1194959e-01 -5.3827482e-01
-3.3645928e-01 1.1698722e-03 3.8316733e-01 8.2433915e-01
1.5109041e-01 -3.9565157e-02 -3.8555974e-01 1.3166723e+00
-5.8467603e-01 9.4364583e-02 4.0246886e-01 1.1976012e+00
-6.1189783e-01 -5.3117168e-01 -2.6744953e-01 9.7595477e-01
-6.5755451e-01 3.4195137e-01 2.9503667e-01 7.6031625e-01
5.9230244e-01 9.6938890e-01 2.8494918e-01 -5.9117544e-01
7.3053104e-01 4.7943565e-01 5.6152517e-01 -6.5931702e-01
-5.8029634e-01 -1.5158096e-01 6.2735096e-02 -6.7660785e-01
-7.5538164e-01 -6.0056210e-01 -1.1647823e+00 1.2514709e-01
1.7362243e-01 2.1945177e-01 5.8258295e-01 1.1908774e+00
3.2660434e-01 1.2129277e+00 3.4372959e-01 -8.3173037e-01
-3.9990360e-01 -1.2895341e+00 -8.7954217e-01 -1.9413719e-01
6.1024767e-01 -1.0264050e+00 -6.7998126e-02 -2.9646543e-01] | [13.53908634185791, 3.483686923980713] |
589ad364-a184-4ff7-a4fb-958b09da3929 | when-crowd-meets-persona-creating-a-large | 2304.0035 | null | https://arxiv.org/abs/2304.00350v1 | https://arxiv.org/pdf/2304.00350v1.pdf | When Crowd Meets Persona: Creating a Large-Scale Open-Domain Persona Dialogue Corpus | Building a natural language dataset requires caution since word semantics is vulnerable to subtle text change or the definition of the annotated concept. Such a tendency can be seen in generative tasks like question-answering and dialogue generation and also in tasks that create a categorization-based corpus, like topic classification or sentiment analysis. Open-domain conversations involve two or more crowdworkers freely conversing about any topic, and collecting such data is particularly difficult for two reasons: 1) the dataset should be ``crafted" rather than ``obtained" due to privacy concerns, and 2) paid creation of such dialogues may differ from how crowdworkers behave in real-world settings. In this study, we tackle these issues when creating a large-scale open-domain persona dialogue corpus, where persona implies that the conversation is performed by several actors with a fixed persona and user-side workers from an unspecified crowd. | ['Nam Soo Kim', 'Sowon Hahn', 'Moosung Kim', 'Sangah Park', 'JiHwan Kim', 'Seoyeon Bae', 'Yoon Kyung Lee', 'Won Ik Cho'] | 2023-04-01 | null | null | null | null | ['dialogue-generation', 'dialogue-generation'] | ['natural-language-processing', 'speech'] | [ 2.66593188e-01 8.40401709e-01 2.76872665e-01 -6.26524150e-01
-6.65601730e-01 -9.35567081e-01 1.10249615e+00 4.14960980e-01
-6.49033785e-01 1.18702137e+00 8.25240254e-01 -7.52643645e-02
3.03825915e-01 -7.81626582e-01 -4.54840302e-01 -5.77421010e-01
4.47058678e-01 9.16575611e-01 1.44789904e-01 -3.60121846e-01
-2.55748164e-03 -1.64934203e-01 -1.25261736e+00 3.82634014e-01
6.50489092e-01 6.48914099e-01 8.52932706e-02 6.30314350e-01
-5.19729257e-01 7.07202494e-01 -1.24471414e+00 -1.13895476e+00
3.28430623e-01 -6.89148545e-01 -1.40562189e+00 5.23720920e-01
1.70348212e-01 -2.09436473e-02 3.25064540e-01 9.16481614e-01
4.48124498e-01 4.96736020e-01 6.11185730e-01 -1.27535248e+00
-3.14965397e-01 9.29099917e-01 -1.96763679e-01 -2.03244582e-01
6.94509327e-01 1.89360246e-01 1.02510560e+00 -3.55361432e-01
9.54462588e-01 1.47268510e+00 6.04999423e-01 9.19546783e-01
-1.28581786e+00 -3.26172441e-01 1.93607926e-01 -5.09665787e-01
-1.02825999e+00 -5.14877915e-01 6.26010776e-01 -8.21084619e-01
4.10588354e-01 4.23954308e-01 5.20111442e-01 1.82540882e+00
-1.85348064e-01 3.78486663e-01 1.07238567e+00 -3.82559896e-01
5.65217793e-01 7.54905522e-01 1.18087068e-01 7.05761388e-02
2.97285795e-01 -6.28488839e-01 -6.46656334e-01 -8.01536620e-01
2.69884855e-01 -1.95568904e-01 -2.06634834e-01 -1.19764917e-02
-1.38840663e+00 1.21105540e+00 1.36981616e-02 4.87142533e-01
-4.43374813e-01 -1.60949424e-01 6.82999790e-01 5.06866992e-01
6.02744699e-01 9.92558002e-01 -4.40600902e-01 -5.34476876e-01
-6.23637915e-01 6.97994232e-01 1.49761736e+00 9.61096346e-01
8.27287674e-01 -4.99117762e-01 -2.63045102e-01 9.94779170e-01
9.04288217e-02 1.96179658e-01 9.35499668e-01 -8.31405401e-01
5.33447742e-01 6.26641572e-01 5.74648380e-01 -9.83056664e-01
-1.81816727e-01 3.62936616e-01 -5.91395080e-01 -5.55085242e-02
1.14785647e+00 -8.20335984e-01 -3.82367939e-01 1.72341132e+00
5.65808237e-01 -5.72213113e-01 2.03313842e-01 8.44832301e-01
9.35364485e-01 4.75497842e-01 1.64458871e-01 -1.94279775e-01
1.76387906e+00 -5.45682251e-01 -9.27748382e-01 -4.22432154e-01
4.29344952e-01 -6.67698443e-01 1.13743281e+00 2.07020447e-01
-8.50747764e-01 -2.44882196e-01 -6.59969985e-01 -3.09886307e-01
-5.61444640e-01 -5.10588229e-01 4.79042590e-01 8.30135345e-01
-7.01997399e-01 2.35525012e-01 -2.49310598e-01 -8.33027601e-01
2.71022171e-01 -2.59473640e-02 -4.19784456e-01 2.63087839e-01
-1.43849647e+00 6.66564584e-01 2.22985834e-01 -1.85009271e-01
-5.09545088e-01 -4.44472373e-01 -8.24598551e-01 -2.90927291e-01
7.72199333e-01 -7.27206647e-01 1.43724370e+00 -1.18796420e+00
-1.47801673e+00 1.37238991e+00 -1.54343247e-01 -4.79033142e-01
1.09150207e+00 -2.27074355e-01 -1.10478826e-01 -3.31852049e-01
4.28185463e-01 4.42627132e-01 9.16502595e-01 -1.26251066e+00
-4.96579766e-01 -5.07240474e-01 5.32051921e-01 3.13798159e-01
-4.71764244e-03 9.04216692e-02 1.73861206e-01 -6.96936429e-01
-4.08184856e-01 -9.13067639e-01 -4.02827829e-01 -1.79191679e-01
-6.52495861e-01 -3.62278610e-01 7.33301759e-01 -4.09345806e-01
7.37411976e-01 -1.96942914e+00 -1.87159076e-01 -4.50782552e-02
3.50027949e-01 1.83113307e-01 2.74004817e-01 7.69247830e-01
2.35079795e-01 4.36531693e-01 -7.79680014e-02 -5.22225499e-01
1.46851942e-01 1.18692137e-01 -4.91315693e-01 2.85184294e-01
-1.43327072e-01 5.75167537e-01 -1.12515211e+00 -4.47607666e-01
-3.01547021e-01 -2.15808023e-02 -3.08650881e-01 3.86179507e-01
-6.12861454e-01 5.57804644e-01 -7.10690200e-01 1.00364208e-01
1.79409578e-01 -1.45400181e-01 2.83263266e-01 3.70506287e-01
-1.48852577e-03 5.23198128e-01 -1.13259149e+00 1.72844994e+00
-6.04063511e-01 7.55507529e-01 2.64873475e-01 -4.20433581e-01
9.93184865e-01 7.21824467e-01 1.47697970e-01 1.30250221e-02
3.14449698e-01 -4.68903705e-02 -7.88489655e-02 -6.85780585e-01
9.58408415e-01 -4.20794219e-01 -4.62684661e-01 1.20060527e+00
-1.26118392e-01 -4.36170220e-01 2.19540998e-01 1.82965174e-01
9.77723479e-01 -3.83338362e-01 4.28710431e-01 -5.03505170e-01
3.90662611e-01 1.92227751e-01 5.77896178e-01 9.28304911e-01
-3.77378941e-01 6.13309503e-01 8.19713175e-01 -4.57812816e-01
-8.30919087e-01 -5.90064704e-01 -1.55776933e-01 1.42211092e+00
-7.97093511e-02 -3.45012188e-01 -1.00632131e+00 -8.04921329e-01
-2.15551332e-01 7.52875268e-01 -6.16166294e-01 2.11433619e-01
-2.41461620e-01 -4.08814520e-01 6.36051893e-01 -3.59383449e-02
5.08718610e-01 -1.14776659e+00 -7.25495636e-01 3.90010327e-01
-6.53987765e-01 -1.12425733e+00 -4.71900612e-01 -1.88636124e-01
-1.78191319e-01 -9.38197136e-01 -6.99604750e-01 -5.59657276e-01
6.98263645e-01 2.71529332e-02 1.19453096e+00 -1.05221272e-01
3.85590941e-02 2.59814441e-01 -6.11279309e-01 -8.41461539e-01
-8.24117482e-01 3.14221680e-01 -2.22039949e-02 3.84975940e-01
7.11303115e-01 -4.01579559e-01 -2.96572357e-01 3.55970711e-01
-9.44548547e-01 -5.53059876e-02 -2.11586535e-01 7.80655503e-01
-2.13333920e-01 4.43733819e-02 7.41336465e-01 -1.61468947e+00
1.41087019e+00 -5.30240178e-01 -1.85252711e-01 1.86448768e-01
-1.64586127e-01 -1.13835722e-01 4.42736477e-01 -5.33917487e-01
-1.33166075e+00 -2.03156918e-01 2.13236690e-01 3.10934573e-01
-6.23808265e-01 1.08442917e-01 -4.20092016e-01 6.47401989e-01
1.12494981e+00 -2.21379519e-01 7.32618570e-02 -2.74318397e-01
4.38833535e-01 1.13590026e+00 2.50623792e-01 -8.67949843e-01
8.76847565e-01 5.44876993e-01 -6.75919294e-01 -1.01135397e+00
-8.18915784e-01 -3.39456409e-01 -5.67762554e-01 -2.13612407e-01
1.11074817e+00 -8.62497389e-01 -7.39360750e-01 4.55270052e-01
-1.60552847e+00 -5.70715487e-01 -6.81797087e-01 1.98575165e-02
-3.00355583e-01 -6.06191643e-02 -2.75583476e-01 -8.04953516e-01
-2.09504396e-01 -9.97977257e-01 9.81661439e-01 3.00852835e-01
-1.19197273e+00 -1.28566992e+00 1.16656341e-01 7.57939279e-01
3.84252310e-01 4.03106868e-01 6.12180948e-01 -1.29284132e+00
-4.58144471e-02 -2.43160069e-01 1.43349275e-01 7.06747472e-02
3.94732207e-01 -2.02360332e-01 -1.32978499e+00 -9.46003869e-02
2.92242110e-01 -7.76888609e-01 4.01500612e-02 -1.94326118e-01
6.71167612e-01 -8.14021409e-01 -1.73416138e-01 -2.62813002e-01
6.60044670e-01 3.63161713e-02 3.15407634e-01 1.08640753e-01
4.15515304e-01 1.24624538e+00 3.88488978e-01 6.75874889e-01
6.18790388e-01 4.16911155e-01 -2.41248801e-01 3.08553688e-02
3.32197994e-01 -4.06629980e-01 2.06559896e-01 2.79807389e-01
-5.35043376e-03 -4.71482933e-01 -9.03186679e-01 6.49066091e-01
-1.86913860e+00 -9.22182381e-01 -6.62201270e-02 2.03697252e+00
1.26313877e+00 -1.36695966e-01 3.45471770e-01 -1.27850221e-02
9.90086734e-01 4.68827069e-01 -2.46684924e-01 -6.60592496e-01
6.37555346e-02 -3.13284874e-01 9.27908495e-02 5.17470479e-01
-7.38377154e-01 7.68164694e-01 5.24660778e+00 3.99353117e-01
-1.02736020e+00 4.02465165e-01 8.32197845e-01 -1.17747247e-01
-3.15369308e-01 7.86092281e-02 -5.68822026e-01 5.49200952e-01
7.27080047e-01 -4.97429609e-01 2.07706735e-01 8.20226073e-01
1.50778666e-01 -2.72739857e-01 -1.43220532e+00 8.85615826e-01
1.68131679e-01 -9.08242702e-01 -1.50362402e-01 3.56705815e-01
6.43502474e-01 -3.68128359e-01 -3.48456353e-01 2.03887448e-01
7.46745825e-01 -9.00754452e-01 7.93404222e-01 1.51994184e-01
5.96186280e-01 -3.90738666e-01 7.66545832e-01 6.42806292e-01
-5.82113504e-01 2.16959849e-01 -2.40581736e-01 -2.85426825e-01
4.02502358e-01 6.81626439e-01 -1.25001717e+00 -5.94980195e-02
4.52157259e-01 9.76876821e-03 -3.67112041e-01 3.98490489e-01
-4.54016805e-01 4.61477667e-01 -1.24150656e-01 -3.87435406e-01
1.59158558e-01 -2.77465075e-01 6.82310045e-01 1.19360101e+00
-7.44935647e-02 1.33405492e-01 1.80219084e-01 9.13025916e-01
-3.79121035e-01 1.37962028e-01 -8.37682009e-01 -2.09920317e-01
5.92697561e-01 1.33943009e+00 -7.83021808e-01 -2.20582500e-01
-3.71963710e-01 9.73079026e-01 1.45035818e-01 4.26687032e-01
-4.28523213e-01 -3.68307412e-01 7.96854377e-01 5.57278693e-01
-1.70438841e-01 1.68068901e-01 -2.73494363e-01 -1.02879131e+00
4.91771176e-02 -1.15556228e+00 4.00711954e-01 -6.05039001e-01
-1.58756626e+00 7.23009288e-01 8.53956044e-02 -6.77973926e-01
-6.83188736e-01 -2.17854843e-01 -7.38262296e-01 6.87411308e-01
-6.48977637e-01 -8.25698912e-01 -3.98526818e-01 5.56626022e-01
6.13371372e-01 -1.54897943e-01 8.36989045e-01 -5.47075234e-02
-3.67256075e-01 4.49777871e-01 -4.51575696e-01 4.83910590e-01
1.00838172e+00 -1.27432048e+00 4.67334241e-01 7.28780866e-01
6.96985200e-02 7.75259078e-01 9.87705350e-01 -6.31020606e-01
-9.41814184e-01 -8.72281373e-01 1.46924055e+00 -8.14629138e-01
8.35237741e-01 -1.03455865e+00 -9.39439595e-01 7.05089808e-01
4.49063450e-01 -3.12238157e-01 1.03769898e+00 3.92651230e-01
-6.52670339e-02 2.73844868e-01 -1.37426782e+00 7.49786854e-01
9.34939504e-01 -7.14004755e-01 -1.06137979e+00 6.89936042e-01
7.41267860e-01 -4.20663238e-01 -6.35184646e-01 -3.87570590e-01
1.95536643e-01 -8.82017076e-01 3.34463298e-01 -7.34384179e-01
3.13320786e-01 -5.12416735e-02 1.38269559e-01 -1.47084594e+00
2.93959826e-01 -1.22084904e+00 3.63086730e-01 1.80775213e+00
6.50120258e-01 -9.34800684e-01 4.57364261e-01 1.73915327e+00
3.88443023e-01 -1.19113229e-01 -9.01806474e-01 -4.40803975e-01
1.09065063e-01 -3.44395190e-01 9.45803285e-01 1.28460896e+00
3.41509998e-01 8.11502159e-01 -3.65725964e-01 -3.19930047e-01
-9.75952018e-03 -7.00931847e-02 1.32715797e+00 -1.39477921e+00
-1.79233700e-01 -1.23510681e-01 -1.16041176e-01 -9.07044351e-01
2.39962146e-01 -4.03099418e-01 4.31533933e-01 -1.13459289e+00
2.85552256e-03 -6.46780670e-01 6.44526482e-01 2.86593139e-01
-3.07701290e-01 -1.94039419e-01 1.03316560e-01 1.86363272e-02
-5.11423767e-01 3.22937697e-01 1.18221796e+00 1.65129006e-01
-3.31131279e-01 3.30911756e-01 -1.13458323e+00 8.85894835e-01
5.72620153e-01 -6.83075666e-01 -6.08848214e-01 -1.79746076e-01
5.28729737e-01 -3.71642634e-02 3.37433755e-01 -5.83553851e-01
1.88652933e-01 -3.09051543e-01 -1.68423757e-01 1.96224332e-01
3.33017319e-01 -7.45100617e-01 7.23684803e-02 4.89035957e-02
-7.11026430e-01 -1.77335337e-01 -3.25245917e-01 7.14377344e-01
-1.86818779e-01 -6.49931788e-01 4.50068384e-01 -5.28470039e-01
-2.28240058e-01 -1.57411192e-02 -6.59891725e-01 6.97107911e-01
1.10320044e+00 -3.33528131e-01 -5.53732395e-01 -9.20403659e-01
-4.99013215e-01 1.37802288e-01 7.74744332e-01 4.74603862e-01
-7.78385177e-02 -8.92149866e-01 -8.06228399e-01 -1.91842988e-01
1.55883431e-01 5.38010240e-01 6.39993930e-03 2.49962389e-01
-1.85238868e-01 6.22227862e-02 2.53462315e-01 -1.83604777e-01
-1.05310094e+00 1.83208048e-01 7.58477151e-02 -2.66829044e-01
-5.21199107e-01 9.33184505e-01 4.05059636e-01 -4.72000867e-01
2.71172464e-01 -2.10839719e-01 -1.64985985e-01 6.43373966e-01
6.92423046e-01 2.85902619e-01 -1.93126947e-01 -8.06482613e-01
-9.35574099e-02 -2.63410658e-01 -1.87268898e-01 -4.78788137e-01
1.01410460e+00 -3.82954687e-01 -1.62916005e-01 7.61819184e-01
8.63195717e-01 2.67171413e-01 -1.06075382e+00 -4.01373655e-01
-5.08176461e-02 -4.76159722e-01 -5.29923499e-01 -5.89319289e-01
-4.49343085e-01 4.94445056e-01 -1.01295225e-01 8.95069242e-01
3.20407391e-01 2.72764593e-01 9.23302591e-01 5.01029551e-01
5.77127099e-01 -1.33373606e+00 2.84123302e-01 4.86082911e-01
1.07713592e+00 -1.56307173e+00 -9.76485163e-02 -3.16685945e-01
-1.20604873e+00 8.17620456e-01 6.15748405e-01 4.25391704e-01
4.78477776e-01 3.70476842e-02 5.68767309e-01 -3.22084397e-01
-7.36286581e-01 1.12275546e-02 -1.03203483e-01 7.78648198e-01
4.57069963e-01 1.47057161e-01 -4.95268196e-01 9.46290731e-01
-9.25411582e-01 -2.90676683e-01 8.87577116e-01 9.23731625e-01
-2.72779781e-02 -1.07793891e+00 -4.28378791e-01 3.56513441e-01
-4.90916610e-01 1.03762485e-01 -1.10828960e+00 5.57523429e-01
2.35266954e-01 1.41838777e+00 2.80806422e-01 -6.11570515e-02
2.48278111e-01 2.93477923e-01 -2.68559754e-01 -1.19017947e+00
-1.01579189e+00 -5.36082029e-01 7.54692376e-01 -1.12229504e-01
-5.02298534e-01 -9.27154362e-01 -1.03015697e+00 -3.66415411e-01
-2.46569663e-01 5.01138568e-01 6.47621751e-01 1.26411080e+00
2.47074008e-01 -1.10003896e-01 4.95865703e-01 -4.91407126e-01
-2.99891055e-01 -1.18853581e+00 -4.37324286e-01 9.92920518e-01
2.16347277e-01 -3.25494140e-01 -4.10792500e-01 3.86499852e-01] | [12.757132530212402, 8.024104118347168] |
39d4c8f0-5dd3-45c1-85bd-48ff75d63f57 | flexer-flexible-entity-resolution-for | 2209.07569 | null | https://arxiv.org/abs/2209.07569v2 | https://arxiv.org/pdf/2209.07569v2.pdf | FlexER: Flexible Entity Resolution for Multiple Intents | Entity resolution, a longstanding problem of data cleaning and integration, aims at identifying data records that represent the same real-world entity. Existing approaches treat entity resolution as a universal task, assuming the existence of a single interpretation of a real-world entity and focusing only on finding matched records, separating corresponding from non-corresponding ones, with respect to this single interpretation. However, in real-world scenarios, where entity resolution is part of a more general data project, downstream applications may have varying interpretations of real-world entities relating, for example, to various user needs. In what follows, we introduce the problem of multiple intents entity resolution (MIER), an extension to the universal (single intent) entity resolution task. As a solution, we propose FlexER, utilizing contemporary solutions to universal entity resolution tasks to solve multiple intents entity resolution. FlexER addresses the problem as a multi-label classification problem. It combines intent-based representations of tuple pairs using a multiplex graph representation that serves as an input to a graph neural network (GNN). FlexER learns intent representations and improves the outcome to multiple resolution problems. A large-scale empirical evaluation introduces a new benchmark and, using also two well-known benchmarks, shows that FlexER effectively solves the MIER problem and outperforms the state-of-the-art for a universal entity resolution. | ['Avigdor Gal', 'Roee Shraga', 'Bar Genossar'] | 2022-08-23 | null | null | null | null | ['entity-resolution'] | ['natural-language-processing'] | [ 4.11519945e-01 5.22375643e-01 -4.50873703e-01 -3.93990755e-01
-1.04411840e+00 -5.01180887e-01 3.68739069e-01 8.44679594e-01
-1.99501947e-01 9.39252913e-01 2.31683373e-01 -1.10202909e-01
-5.32527506e-01 -1.20970619e+00 -1.08895886e+00 -1.85778618e-01
-2.06784174e-01 1.11156464e+00 -1.82617083e-02 -4.66786057e-01
-2.35780865e-01 3.28737080e-01 -1.63538826e+00 2.83926278e-01
7.74705052e-01 8.06922853e-01 -1.98445484e-01 3.74670029e-01
-3.67881477e-01 6.65375412e-01 -8.42309475e-01 -6.23357177e-01
3.29719633e-01 7.15098297e-03 -1.29609501e+00 -3.62491995e-01
7.08939016e-01 3.63213390e-01 -1.27648897e-02 1.18193901e+00
3.17186892e-01 2.64398545e-01 5.68124771e-01 -1.53607726e+00
-1.16090775e+00 9.91350114e-01 -5.50578713e-01 1.21898413e-01
6.01625443e-01 -4.41651076e-01 1.48516405e+00 -4.93204743e-01
1.13493848e+00 1.21215713e+00 7.29852855e-01 4.38170761e-01
-1.42513978e+00 -7.00340271e-01 1.78335920e-01 1.88971296e-01
-1.54976892e+00 -3.23715329e-01 2.00692013e-01 -1.80599913e-01
1.07075715e+00 5.62931418e-01 -1.77253738e-01 6.57573164e-01
-4.06427197e-02 3.58108968e-01 3.80146593e-01 -2.56907403e-01
1.76252812e-01 -4.13104296e-02 7.57236838e-01 4.45447713e-01
9.79333282e-01 -3.05789351e-01 -1.18423335e-01 -4.17600960e-01
2.55108416e-01 9.74897072e-02 -4.16905999e-01 -4.49821353e-01
-1.03353763e+00 7.30449021e-01 8.00804913e-01 4.30454314e-01
-5.02057672e-01 5.95470797e-03 3.98135275e-01 3.85928780e-01
2.97374099e-01 8.69927526e-01 -6.79489374e-01 5.58296859e-01
-4.89313066e-01 4.31878924e-01 1.19793189e+00 1.29212272e+00
1.19701934e+00 -4.57011491e-01 -2.27653936e-01 6.19751513e-01
1.52353182e-01 9.92668420e-02 2.67057240e-01 -6.46925986e-01
7.30193257e-01 1.20838773e+00 3.21499646e-01 -8.91782582e-01
-6.29775286e-01 -3.47685158e-01 -8.83257806e-01 -1.14765793e-01
7.51213133e-02 1.82629880e-02 -9.92147744e-01 1.96035063e+00
5.54005265e-01 3.89573485e-01 5.26515603e-01 8.96771491e-01
1.26022136e+00 4.80721653e-01 4.76684511e-01 -3.44053954e-01
1.63311636e+00 -6.94975555e-01 -8.81507874e-01 -2.62138069e-01
6.38216078e-01 -2.30025575e-01 2.89279133e-01 2.43718755e-02
-7.00608790e-01 -2.91195452e-01 -1.12853944e+00 -3.58862132e-01
-1.10123575e+00 -3.01657051e-01 6.93728805e-01 3.29719901e-01
-9.02071655e-01 5.97110450e-01 -6.31830320e-02 -6.14556968e-01
6.51477128e-02 5.66166878e-01 -9.02988970e-01 -1.70675859e-01
-1.74489915e+00 9.07119751e-01 8.33213091e-01 -1.71206251e-01
-1.33965984e-01 -9.13156927e-01 -1.19170547e+00 3.33180308e-01
8.51530612e-01 -1.04732788e+00 1.06882572e+00 -5.37694633e-01
-3.34080547e-01 9.51634526e-01 -1.78164646e-01 -6.32606149e-01
-3.36715952e-03 -1.87430739e-01 -1.05383253e+00 -5.02336621e-01
5.31946659e-01 3.45784575e-01 3.29566926e-01 -1.57732713e+00
-1.03787708e+00 -4.21806872e-01 3.83126020e-01 5.83547130e-02
1.20518908e-01 -1.18318260e-01 -4.20988977e-01 -3.40272546e-01
8.58395696e-02 -6.48582757e-01 -2.72176296e-01 -6.91017449e-01
-5.83353698e-01 -5.47892153e-01 4.77169633e-01 -6.28451765e-01
1.53077567e+00 -1.85050046e+00 3.07283640e-01 1.91955730e-01
5.64226985e-01 2.27250010e-01 -1.48033977e-01 4.68218267e-01
-6.77254260e-01 5.93092382e-01 -4.54979032e-01 -4.42529470e-01
1.33856416e-01 3.95155013e-01 -3.84979695e-01 1.54763177e-01
2.32864603e-01 1.01429486e+00 -9.00138617e-01 -2.48689801e-01
-2.90747911e-01 1.81451082e-01 -5.41178808e-02 2.42407337e-01
-2.62625843e-01 -1.51314273e-01 -9.08308998e-02 6.87075615e-01
8.19123089e-01 -4.96615201e-01 5.44459820e-01 -3.97793829e-01
7.13976175e-02 1.57025918e-01 -1.71991086e+00 1.58508110e+00
-2.45467275e-01 1.63778052e-01 -2.60589086e-02 -8.43684435e-01
8.79640281e-01 2.92201996e-01 5.57383955e-01 -7.71250248e-01
-2.43102476e-01 3.04451734e-01 -4.74773169e-01 -3.88743192e-01
1.04639363e+00 1.07345641e-01 -5.62809527e-01 3.32359016e-01
1.08002104e-01 6.51202857e-01 5.27529299e-01 3.69700253e-01
1.40689301e+00 -2.09310934e-01 9.19104695e-01 6.59747273e-02
4.22548175e-01 9.27028731e-02 9.52819943e-01 7.74217665e-01
1.40531272e-01 6.01278424e-01 4.62980717e-01 -6.82618678e-01
-6.16126955e-01 -9.59907651e-01 4.69685011e-02 1.11895061e+00
5.73911905e-01 -6.38168514e-01 -4.19633955e-01 -9.40919101e-01
5.30002773e-01 6.86269164e-01 -8.37602913e-01 2.77737025e-02
-8.25245380e-01 -8.79870296e-01 4.72779155e-01 3.57993335e-01
5.40950820e-02 -1.11683083e+00 -2.62756377e-01 4.17465061e-01
-3.20183903e-01 -1.10252059e+00 -1.60071954e-01 3.96140099e-01
-2.92545885e-01 -1.50924432e+00 -2.72864606e-02 -6.74670398e-01
4.38191265e-01 1.76701382e-01 1.62882328e+00 3.30840856e-01
-3.77707124e-01 2.07872808e-01 -1.90123811e-01 -2.72075027e-01
-4.03904140e-01 4.30625290e-01 6.66520968e-02 -8.86998475e-02
6.23960733e-01 -3.40422720e-01 -1.26494288e-01 8.75563323e-02
-1.16022456e+00 -3.43219608e-01 4.67734814e-01 5.16999662e-01
9.02895272e-01 5.94620518e-02 9.58971083e-01 -1.67809117e+00
6.97517097e-01 -1.21451569e+00 -4.38491434e-01 8.84785175e-01
-1.04879892e+00 2.35599086e-01 5.32911777e-01 -1.05208993e-01
-7.97098398e-01 -1.75502062e-01 1.48803264e-01 -3.66286188e-01
-2.70493776e-01 7.46412337e-01 -4.33257848e-01 3.98672432e-01
7.40690112e-01 -4.31741416e-01 -6.79556310e-01 -4.82683599e-01
5.99003315e-01 4.20468360e-01 9.67780530e-01 -5.97362399e-01
9.36515689e-01 3.57620083e-02 6.68701530e-02 -5.29011600e-02
-8.40137303e-01 -8.10579419e-01 -4.80745286e-01 2.88135558e-01
7.62951076e-01 -9.89864826e-01 -6.90476596e-01 -1.17215283e-01
-1.28015161e+00 2.58082211e-01 -4.58468527e-01 -2.06458077e-01
-2.47930110e-01 1.18630372e-01 -3.53953242e-01 -4.77713853e-01
-6.42892778e-01 -7.86466599e-01 1.09690475e+00 4.69822288e-01
-4.25943673e-01 -8.11845601e-01 3.39341581e-01 2.86518276e-01
3.26126903e-01 6.07393384e-01 1.20331979e+00 -1.33360672e+00
-7.52055883e-01 -1.09812096e-01 -4.60927635e-01 -4.32619423e-01
3.69511455e-01 -2.83362716e-01 -8.02006185e-01 -2.04623252e-01
-4.92088944e-01 3.59259807e-02 7.28050530e-01 -1.07738957e-01
7.58942306e-01 -5.26125669e-01 -6.60433352e-01 6.86410010e-01
1.76182270e+00 6.05693199e-02 6.23998702e-01 6.02344334e-01
7.76975095e-01 5.19530714e-01 7.84911752e-01 1.16012961e-01
8.96055818e-01 6.38216078e-01 6.54464066e-01 -3.10787410e-01
3.43034044e-03 -1.68122351e-01 -4.82166916e-01 1.69994146e-01
1.12388775e-01 -6.41804695e-01 -8.38119447e-01 7.18939841e-01
-2.32360625e+00 -9.45473850e-01 -3.03991139e-01 2.13997746e+00
7.22963214e-01 -2.38859057e-01 -1.21993683e-01 -1.49614781e-01
1.10365772e+00 1.31450463e-02 -7.04053104e-01 -3.64741266e-01
-1.55737892e-01 6.02392517e-02 7.85152614e-01 2.82840669e-01
-1.41738248e+00 7.03921497e-01 5.66950369e+00 3.20406646e-01
-7.85512745e-01 4.17754846e-03 2.75953084e-01 3.37322652e-01
-5.60506999e-01 -6.01723194e-02 -9.91386414e-01 4.90724370e-02
9.54107761e-01 -5.53514183e-01 5.72393537e-01 7.64280081e-01
-6.27798378e-01 2.02447146e-01 -1.54966760e+00 8.13679218e-01
-1.18511096e-02 -1.54599118e+00 1.61477312e-01 3.82536165e-02
6.03460252e-01 6.02398021e-03 -3.04901630e-01 7.82533824e-01
5.92647672e-01 -1.21705544e+00 3.45258266e-01 5.08723199e-01
8.59689116e-01 -6.76796973e-01 9.28846419e-01 2.02951908e-01
-1.54048765e+00 -1.44740358e-01 -2.44269744e-01 3.42094719e-01
9.47016776e-02 4.61392134e-01 -9.63439941e-01 1.42835271e+00
7.60885537e-01 6.30940139e-01 -6.11744344e-01 8.70138824e-01
8.30739811e-02 -2.25930065e-02 -1.53506055e-01 6.22733772e-01
-2.31497943e-01 -1.09794782e-02 4.93964553e-01 1.45757115e+00
1.30282253e-01 1.79885924e-01 1.18705094e-01 8.00778747e-01
-8.74293566e-01 6.67646527e-02 -8.44517291e-01 1.29633546e-01
8.95702541e-01 1.45231783e+00 -3.73933107e-01 -2.71937758e-01
-3.44973981e-01 6.69523060e-01 8.56689751e-01 3.51327300e-01
-7.23235309e-01 -4.38774079e-01 1.03024852e+00 -1.14088073e-01
2.11315602e-01 4.04888183e-01 -2.66606569e-01 -1.07618320e+00
8.57485905e-02 -8.99703324e-01 1.15019894e+00 -4.83058959e-01
-1.29516721e+00 5.03623068e-01 -2.96636727e-02 -9.85520661e-01
-3.59720916e-01 -1.78975895e-01 -4.85516250e-01 8.73890817e-01
-1.80823755e+00 -1.16315699e+00 -4.14278209e-01 3.20190758e-01
3.12069058e-01 1.44037664e-01 1.17534912e+00 6.75740302e-01
-7.95532346e-01 6.95444405e-01 -2.16238007e-01 2.77355790e-01
1.00692201e+00 -1.60702109e+00 7.82754421e-01 1.03737497e+00
1.30996093e-01 1.05970001e+00 6.99799776e-01 -8.74975920e-01
-1.57670391e+00 -1.66270554e+00 1.27874672e+00 -4.88862365e-01
5.55277467e-01 -2.11112529e-01 -1.45190299e+00 1.17030180e+00
1.11259431e-01 2.25689769e-01 7.12064862e-01 5.41205943e-01
-5.83606482e-01 -6.37322664e-02 -1.49910712e+00 2.47602895e-01
1.41614544e+00 -3.61638427e-01 -8.96240115e-01 1.68615222e-01
1.10422456e+00 -5.53703666e-01 -1.44895339e+00 5.68036199e-01
1.67021662e-01 -4.11168575e-01 1.15295768e+00 -1.03737128e+00
8.08415338e-02 -6.63903356e-01 -3.09430778e-01 -1.23488367e+00
-5.76746821e-01 -4.10070032e-01 -5.05267560e-01 1.69211888e+00
6.79917872e-01 -6.91751182e-01 3.25793266e-01 1.05214953e+00
5.48872985e-02 -4.84864950e-01 -9.99767482e-01 -4.89938200e-01
-2.99543738e-01 1.38790503e-01 1.43997669e+00 1.43009186e+00
-1.25301600e-01 5.96028984e-01 -1.84679180e-01 9.05042827e-01
5.33719599e-01 5.92163503e-01 7.76330590e-01 -1.87665415e+00
8.48795399e-02 -1.75723985e-01 -3.64709735e-01 -3.54287773e-01
4.24924225e-01 -1.11904991e+00 -4.61073443e-02 -2.00529289e+00
4.59124148e-02 -7.38419950e-01 -6.27707243e-01 7.53147960e-01
-4.58532840e-01 -3.94798741e-02 6.70587420e-02 1.01043269e-01
-1.03977871e+00 -7.10830465e-02 3.76691073e-01 -5.02993405e-01
-1.59842521e-01 -2.09584519e-01 -1.30848050e+00 2.73245692e-01
4.22671080e-01 -7.07345843e-01 -2.18559667e-01 -3.72101724e-01
7.45889544e-01 2.53012985e-01 -3.58447284e-02 -7.92216420e-01
6.16338193e-01 -3.21517736e-01 2.02227291e-02 -3.86672527e-01
-8.27939659e-02 -8.63179624e-01 8.46547902e-01 -9.37302336e-02
-4.24418569e-01 1.35624528e-01 1.36935404e-02 7.70078838e-01
-2.89216101e-01 -2.02296704e-01 3.01638752e-01 -7.11726919e-02
-1.32562172e+00 4.13967907e-01 4.23181683e-01 2.86075652e-01
1.01252103e+00 3.46095413e-02 -1.05238676e+00 1.85880095e-01
-7.30981648e-01 7.01528728e-01 6.04379594e-01 8.77983630e-01
3.21383774e-01 -1.21307218e+00 -8.19031894e-01 -2.14549564e-02
5.71433485e-01 4.56323743e-01 -7.94390589e-03 2.73032606e-01
-5.42977378e-02 1.50190473e-01 2.41736509e-02 -2.16808468e-01
-1.32504749e+00 9.68449712e-01 4.40386027e-01 -6.42717719e-01
-6.40047193e-01 4.30271417e-01 1.22973770e-01 -7.34524190e-01
2.17785671e-01 -2.28016108e-01 -7.37875879e-01 3.10708642e-01
5.32845080e-01 3.80428970e-01 5.93061745e-01 -6.73710167e-01
-5.42534947e-01 2.75054693e-01 -1.33636460e-01 6.32131755e-01
1.36835814e+00 -1.16658106e-01 -3.88891757e-01 2.97664732e-01
9.76976216e-01 -3.08177192e-02 -2.54087955e-01 -4.66317683e-01
6.15055084e-01 -1.25207379e-01 -5.14718235e-01 -9.42956328e-01
-1.02568066e+00 -1.10319242e-01 3.69723886e-01 5.70821881e-01
1.04787517e+00 1.09598830e-01 6.73747420e-01 7.10020304e-01
5.70920110e-01 -7.05408990e-01 -8.50127935e-01 4.70837951e-01
8.40694904e-01 -1.34994805e+00 1.09247439e-01 -5.13398111e-01
-4.72796798e-01 8.91636848e-01 7.61615455e-01 1.30741581e-01
2.77376592e-01 1.33116677e-01 -2.13689283e-02 -4.22090560e-01
-8.45383048e-01 -4.48628157e-01 3.73052597e-01 5.58398962e-01
3.15281928e-01 4.77447242e-01 -2.68051863e-01 8.53577018e-01
5.13329990e-02 -1.26011357e-01 4.95457441e-01 7.69431531e-01
-2.21379980e-01 -1.15486193e+00 -2.93203384e-01 6.04057968e-01
-4.19757515e-01 5.50097339e-02 -4.16508168e-01 8.43211532e-01
4.79138464e-01 9.00508046e-01 4.64490652e-02 -2.39937961e-01
7.74154842e-01 2.15303615e-01 -1.94994703e-01 -7.91611195e-01
-9.68451679e-01 -7.56472826e-01 4.63888377e-01 -5.68705499e-01
-3.18866044e-01 -4.55701083e-01 -1.63326502e+00 -1.97443053e-01
-2.65005827e-01 3.41897994e-01 5.22497475e-01 8.39855433e-01
6.70174479e-01 7.52925992e-01 2.86187947e-01 -2.19870716e-01
-1.65456623e-01 -5.82614064e-01 -4.77282226e-01 9.09266233e-01
6.44876242e-01 -6.15799606e-01 -6.78937435e-02 -2.06983566e-01] | [9.348283767700195, 8.40921401977539] |
b896a300-e1cf-42a8-aab3-a0a7f1b1b758 | a-survey-of-risk-aware-multi-armed-bandits | 2205.05843 | null | https://arxiv.org/abs/2205.05843v1 | https://arxiv.org/pdf/2205.05843v1.pdf | A Survey of Risk-Aware Multi-Armed Bandits | In several applications such as clinical trials and financial portfolio optimization, the expected value (or the average reward) does not satisfactorily capture the merits of a drug or a portfolio. In such applications, risk plays a crucial role, and a risk-aware performance measure is preferable, so as to capture losses in the case of adverse events. This survey aims to consolidate and summarise the existing research on risk measures, specifically in the context of multi-armed bandits. We review various risk measures of interest, and comment on their properties. Next, we review existing concentration inequalities for various risk measures. Then, we proceed to defining risk-aware bandit problems, We consider algorithms for the regret minimization setting, where the exploration-exploitation trade-off manifests, as well as the best-arm identification setting, which is a pure exploration problem -- both in the context of risk-sensitive measures. We conclude by commenting on persisting challenges and fertile areas for future research. | ['Krishna Jagannathan', 'Prashanth L. A.', 'Vincent Y. F. Tan'] | 2022-05-12 | null | null | null | null | ['portfolio-optimization'] | ['time-series'] | [ 2.91309118e-01 2.96596229e-01 -9.85805333e-01 -1.44612446e-01
-8.96196127e-01 -6.99833393e-01 3.03897243e-02 4.41381454e-01
-5.25518119e-01 1.03017223e+00 2.55345464e-01 -7.21244216e-01
-9.67787802e-01 -6.79012299e-01 -4.52255636e-01 -8.35485399e-01
-3.17123443e-01 4.37386602e-01 -5.18259883e-01 9.58226398e-02
3.70930851e-01 4.36305404e-01 -1.01561046e+00 -1.97660431e-01
9.81349349e-01 1.45245695e+00 -3.63041431e-01 1.36318117e-01
5.46426587e-02 5.09814382e-01 -4.28594708e-01 -8.83323073e-01
4.57190871e-01 -4.91507560e-01 -5.49149394e-01 -2.23094136e-01
-2.86009401e-01 -2.40309164e-01 1.95026532e-01 1.00743222e+00
3.58836889e-01 6.54702187e-02 7.39048481e-01 -1.10815811e+00
-1.33440882e-01 6.87835395e-01 -7.14738786e-01 3.78019571e-01
-5.69087379e-02 -1.15792505e-01 1.37895787e+00 -1.12053663e-01
1.41939372e-01 1.05178833e+00 2.95089602e-01 4.81130064e-01
-1.02506876e+00 -5.54871738e-01 6.02617323e-01 -1.53848872e-01
-7.97966361e-01 -3.03338766e-01 5.83615541e-01 -5.45522630e-01
5.11331201e-01 6.27776206e-01 7.75756776e-01 7.08503783e-01
3.32099140e-01 9.48669910e-01 1.06464171e+00 -3.13044816e-01
4.23601836e-01 8.21501538e-02 1.12875246e-01 1.98672831e-01
6.89675450e-01 5.34233093e-01 -4.12559599e-01 -5.46896875e-01
5.31813920e-01 1.36752322e-01 -3.39354396e-01 -7.27753878e-01
-8.61742079e-01 1.17652309e+00 -3.46185043e-02 -2.08818894e-02
-6.82071388e-01 1.92050878e-02 3.01396519e-01 2.98333675e-01
7.80742109e-01 6.07057512e-01 -4.17400062e-01 -4.59589995e-02
-1.12990427e+00 3.41804415e-01 7.10200369e-01 3.71972442e-01
7.48786901e-04 -1.06753238e-01 -6.06225193e-01 5.20140409e-01
1.77074015e-01 2.45002419e-01 -7.42225870e-02 -8.04619610e-01
7.53315151e-01 2.13263035e-01 5.96851349e-01 -4.25725102e-01
-4.46407318e-01 -7.51739323e-01 -6.48858249e-01 2.10436434e-01
4.31651831e-01 -5.43089807e-01 -2.70359874e-01 1.80832684e+00
2.03598708e-01 -1.70077607e-01 -2.12747723e-01 6.70253754e-01
-5.80765046e-02 3.20042700e-01 1.02586709e-01 -1.36832225e+00
1.17613125e+00 -7.03694284e-01 -8.27476501e-01 -2.26656154e-01
3.85393113e-01 -3.38215441e-01 5.23342073e-01 5.91397524e-01
-1.54278028e+00 4.92469251e-01 -8.21040392e-01 6.76632524e-01
6.87761679e-02 -4.84413445e-01 7.32288301e-01 1.16360402e+00
-4.63190943e-01 6.41219795e-01 -4.26500171e-01 1.63251221e-01
7.32161105e-01 2.81900376e-01 3.09571475e-01 1.87652886e-01
-1.03904247e+00 8.04624200e-01 1.26870587e-01 1.20415553e-01
-6.92089677e-01 -1.08440840e+00 -5.34805775e-01 2.37634256e-01
7.75885761e-01 -7.52980113e-01 1.33637476e+00 -8.80783975e-01
-1.40794861e+00 6.07649028e-01 3.41164768e-01 -7.38794804e-01
1.08729398e+00 -2.14966163e-01 -1.93228386e-02 -1.15772918e-01
-8.46082866e-02 -1.44776329e-01 3.43302548e-01 -4.94890064e-01
-7.11299598e-01 -4.52969432e-01 1.96704850e-01 3.30316961e-01
-2.02985704e-01 4.87345964e-01 2.06242606e-01 -8.72286022e-01
-4.29688305e-01 -8.21336687e-01 -6.62657320e-01 -1.94510952e-01
-4.17845994e-01 -2.86231693e-02 -2.55303383e-01 -3.30314040e-01
1.71279907e+00 -1.84557939e+00 -5.31086046e-03 4.26018536e-01
-2.35315591e-01 -2.82884222e-02 2.30036870e-01 4.39460456e-01
-2.52436548e-01 3.21841955e-01 -4.53826904e-01 -4.43366766e-02
9.74962935e-02 -2.55504251e-01 -4.32496101e-01 7.39165962e-01
9.69460234e-02 8.33909154e-01 -6.69925511e-01 -1.50352076e-01
-5.83647713e-02 -1.34839028e-01 -5.71627498e-01 -9.10345092e-02
-3.51928294e-01 2.52069116e-01 -9.38170910e-01 5.55023015e-01
5.59576988e-01 -2.20626742e-01 1.44014180e-01 3.53196919e-01
-1.83532909e-01 1.37249291e-01 -1.04721200e+00 1.06673491e+00
-4.86042023e-01 1.42056659e-01 3.99646275e-02 -1.31482005e+00
4.19384241e-01 2.14303836e-01 7.10922241e-01 -5.02781630e-01
1.05851434e-01 3.10370684e-01 -1.21621266e-02 -1.72287732e-01
1.72299508e-03 -7.05891430e-01 -1.58422470e-01 7.69768119e-01
-6.86562777e-01 3.43778193e-01 -8.67446326e-03 -8.75500143e-02
8.19110632e-01 -3.63249362e-01 7.57939279e-01 -4.28109050e-01
1.92616552e-01 -2.77730256e-01 6.16537988e-01 7.47346640e-01
-1.86112970e-01 2.74252176e-01 1.12511396e+00 -2.21321911e-01
-4.76305693e-01 -8.70236635e-01 -3.87404948e-01 9.09347177e-01
-8.34623799e-02 1.70376003e-01 -4.78610128e-01 -7.32401967e-01
4.30226505e-01 8.09275389e-01 -9.89270806e-01 -1.07957594e-01
-1.39421195e-01 -1.51927769e+00 -6.11746311e-03 3.33258718e-01
-1.22088334e-02 -7.01124430e-01 -9.47483361e-01 2.41181865e-01
1.42840832e-01 -2.99337298e-01 -4.35891122e-01 2.76115596e-01
-1.09216797e+00 -1.11644924e+00 -1.23435593e+00 -7.90024698e-02
2.06328735e-01 -2.44692594e-01 1.03812754e+00 -3.81724745e-01
-1.28402323e-01 2.93425262e-01 -7.44180381e-02 -8.65039766e-01
1.54930949e-01 -4.04383630e-01 -7.48435780e-02 1.86607555e-01
1.17168732e-01 -2.87375659e-01 -1.05863976e+00 3.43206048e-01
-6.49295628e-01 -4.44900423e-01 4.07233924e-01 7.79633045e-01
6.68243587e-01 -1.69247061e-01 1.02893758e+00 -1.00072062e+00
7.66918004e-01 -7.02728868e-01 -9.51876819e-01 6.32704020e-01
-8.97303641e-01 2.03326121e-02 -8.28068182e-02 -3.76453698e-01
-1.09946132e+00 -3.63293260e-01 1.99260637e-01 2.51423963e-03
6.27940297e-01 7.68078446e-01 -1.62382528e-01 2.70164423e-02
3.63222361e-01 -2.40970805e-01 -4.31962647e-02 -6.02949679e-01
1.21648788e-01 5.14109552e-01 -2.86297232e-01 -6.49640560e-01
-2.59755133e-03 3.19317818e-01 2.93469518e-01 -2.69875646e-01
-1.12299871e+00 -2.46961936e-01 3.43831569e-01 4.59381863e-02
5.34741521e-01 -4.76677626e-01 -1.01291919e+00 1.30520090e-01
-8.43405902e-01 -2.44681060e-01 -6.94703341e-01 7.37426937e-01
-1.09941316e+00 9.13433433e-02 -4.30732340e-01 -1.53842533e+00
-3.46814752e-01 -1.11848402e+00 4.57409680e-01 2.24918649e-01
6.31382018e-02 -1.01657176e+00 2.72700816e-01 4.88888592e-01
3.75035971e-01 5.29119611e-01 9.21238422e-01 -7.23327041e-01
-4.99653786e-01 -2.57651180e-01 3.43804918e-02 5.62840924e-02
-1.10123634e-01 -4.08251405e-01 -6.30281270e-01 -1.78602546e-01
4.10958260e-01 -1.82553247e-01 9.61180568e-01 1.19218588e+00
1.19537854e+00 -6.71647489e-01 -3.21578830e-01 4.80382055e-01
1.26562321e+00 5.96163809e-01 3.10827285e-01 4.51142013e-01
-1.68538660e-01 1.00918758e+00 9.71694350e-01 9.22182322e-01
-5.08238524e-02 6.79463506e-01 6.51158929e-01 2.40902022e-01
7.45907009e-01 1.19911127e-01 -2.61544399e-02 -8.40165690e-02
-5.37998695e-03 -3.84356380e-01 -6.49642050e-01 4.22551811e-01
-1.94696593e+00 -7.62060225e-01 3.94360930e-01 2.67757559e+00
8.79044294e-01 9.88144502e-02 8.00219715e-01 5.39956763e-02
7.76553035e-01 9.31390747e-02 -9.75469887e-01 -7.12548137e-01
-1.18034936e-01 8.68452638e-02 8.68506312e-01 4.15587991e-01
-9.39419270e-01 1.98257431e-01 7.31129646e+00 9.65568721e-01
-7.50223696e-01 1.02727473e-01 1.39083517e+00 -7.27976561e-01
-3.91258836e-01 -1.92975610e-01 -5.81960201e-01 5.57168961e-01
9.01102126e-01 -5.96266866e-01 1.19892769e-01 7.42442310e-01
2.83363372e-01 -1.95663854e-01 -1.07467997e+00 7.20806420e-01
-3.49785656e-01 -1.35980678e+00 -4.35445100e-01 5.01380026e-01
6.78640127e-01 -2.53352880e-01 5.35741985e-01 -1.55429661e-01
1.77286223e-01 -1.12440538e+00 7.88916111e-01 5.59002697e-01
7.00521111e-01 -1.19690311e+00 6.65044725e-01 3.41427028e-01
-3.27083439e-01 -3.33186954e-01 -1.14829257e-01 -8.06429982e-02
4.22368854e-01 9.86550570e-01 -2.34833956e-02 5.72239399e-01
5.20115912e-01 4.30159450e-01 3.21487516e-01 1.50490415e+00
8.72702003e-02 3.75123829e-01 -3.24943006e-01 -1.96053863e-01
2.75739104e-01 -3.62150788e-01 5.27010500e-01 8.24926138e-01
3.52889627e-01 3.20823967e-01 -2.02423781e-01 6.98194265e-01
7.91483596e-02 4.55393672e-01 -2.81933278e-01 -1.67360604e-01
2.60359466e-01 7.07453847e-01 -5.62567472e-01 9.14935768e-02
-4.91289437e-01 3.06336641e-01 -7.27141872e-02 2.14764252e-01
-6.40340209e-01 -2.20866814e-01 8.38846922e-01 2.80767772e-02
1.42007142e-01 4.10839438e-01 -5.25214612e-01 -8.56349587e-01
2.04204142e-01 -8.10973763e-01 9.75291073e-01 -8.58303607e-02
-1.19585669e+00 7.69165829e-02 3.14763397e-01 -9.38365579e-01
-2.24056438e-01 -5.08896828e-01 -6.31114662e-01 9.29749548e-01
-1.67367125e+00 -3.52841258e-01 6.86397374e-01 1.32858232e-01
2.01402321e-01 1.69089288e-02 2.72889704e-01 4.96370085e-02
-9.27813888e-01 6.49344623e-01 5.87681532e-01 -4.87507701e-01
1.98752955e-01 -9.69323337e-01 -9.69760865e-02 4.94029224e-01
-1.58523709e-01 5.22475183e-01 7.72286355e-01 -5.85856915e-01
-1.08857656e+00 -7.32098341e-01 5.64060807e-01 -2.03560904e-01
7.72821248e-01 1.01145893e-01 -3.82148981e-01 4.58931386e-01
-1.21882133e-01 -4.13772166e-02 9.46422338e-01 3.48317266e-01
-3.23200375e-02 -2.57298499e-01 -1.45401037e+00 5.09922385e-01
9.97302949e-01 -2.11550146e-02 -1.40738502e-01 5.00451148e-01
5.05737782e-01 -1.34982392e-01 -8.52435589e-01 4.23623383e-01
8.45536470e-01 -1.20186865e+00 1.06138134e+00 -8.92734826e-01
1.44082636e-01 5.81610143e-01 -3.21190655e-02 -1.11884785e+00
-1.07433818e-01 -1.19965148e+00 -2.21338153e-01 9.06622469e-01
5.84047556e-01 -8.69486570e-01 1.11314440e+00 9.48147297e-01
3.48638505e-01 -1.37805653e+00 -1.27680302e+00 -1.07273138e+00
5.28773248e-01 -3.27807456e-01 7.30970979e-01 7.03677952e-01
3.43549460e-01 -2.31199145e-01 -5.09974897e-01 -4.55588371e-01
8.50743413e-01 4.03753430e-01 -3.79713550e-02 -9.75514233e-01
-6.16250932e-01 -9.29549992e-01 4.65689227e-02 -6.44160688e-01
-9.89798903e-02 -5.46407104e-01 -3.63279581e-01 -1.18965840e+00
4.67311412e-01 -4.37208772e-01 -7.30854332e-01 1.26069933e-01
-1.40938640e-01 -1.69019163e-01 2.05680326e-01 -1.08368568e-01
-4.24403548e-01 4.10875052e-01 1.06516647e+00 -1.22587353e-01
-2.77035177e-01 9.14964974e-01 -1.28392208e+00 6.35486126e-01
9.66190696e-01 -4.81444359e-01 -5.36279321e-01 4.98276800e-02
9.53735650e-01 7.00072944e-01 4.32884917e-02 -1.80187345e-01
-1.69282705e-01 -9.65404212e-01 -6.78887516e-02 -4.03433919e-01
1.43403724e-01 -6.91409171e-01 2.06417412e-01 7.11919546e-01
-6.15236044e-01 -2.23481990e-02 -6.31271377e-02 8.42364669e-01
-1.42580103e-02 -5.56625128e-01 8.14158320e-01 3.02616111e-03
2.35293463e-01 4.77739751e-01 -9.26602781e-02 3.79165083e-01
1.35842776e+00 -1.78149521e-01 -2.88120061e-01 -4.96254563e-01
-6.69170678e-01 4.90262538e-01 4.90965098e-02 6.84785470e-02
2.68463582e-01 -1.01078117e+00 -7.41305768e-01 -2.44310454e-01
1.40792280e-01 -3.44516844e-01 3.99305373e-01 9.28397536e-01
-1.17689587e-01 6.01271808e-01 1.29082263e-01 5.46069331e-02
-9.09186244e-01 8.28534365e-01 5.38953245e-01 -7.07683563e-01
-4.63925488e-02 8.34581196e-01 6.22640371e-01 4.68294173e-01
5.45806706e-01 -1.69005439e-01 -2.58627295e-01 4.54376489e-01
6.73745573e-01 6.82689726e-01 4.96981442e-02 3.01489350e-03
-5.55861652e-01 4.77214366e-01 8.07965472e-02 -2.37066135e-01
1.46560550e+00 -1.82990909e-01 5.97065054e-02 2.29547158e-01
8.06393743e-01 -1.55819878e-01 -1.14918578e+00 -1.34036586e-01
2.59668618e-01 -3.46981406e-01 2.46796563e-01 -9.64148879e-01
-1.08173239e+00 7.70265520e-01 3.99536580e-01 4.09498900e-01
1.27266109e+00 -6.36846721e-02 3.82624984e-01 2.35109236e-02
3.82354081e-01 -1.12637639e+00 -1.72780558e-01 -2.88953427e-02
9.77793276e-01 -9.44979012e-01 2.51653284e-01 -1.46909028e-01
-6.85178816e-01 6.05313778e-01 -1.58768237e-01 -4.56455126e-02
1.06699491e+00 3.06936223e-02 -5.15577137e-01 1.74684227e-02
-7.31022358e-01 -9.21480358e-02 4.30250555e-01 3.08577180e-01
2.15942636e-01 3.33679080e-01 -7.69322038e-01 1.02432096e+00
-8.92595295e-03 -9.77063179e-02 3.00392121e-01 9.10148144e-01
-3.16696733e-01 -1.29900730e+00 -4.07746196e-01 1.02833116e+00
-1.10162771e+00 -1.45805359e-01 -2.74275124e-01 3.62022996e-01
-3.24054897e-01 9.45209742e-01 1.27359137e-01 2.70716101e-01
3.76573175e-01 -1.86842158e-01 5.36262035e-01 -3.66086274e-01
-6.75556004e-01 1.56718597e-01 2.14248881e-01 -5.04956424e-01
-4.55458522e-01 -9.55737591e-01 -2.43407667e-01 -3.52711707e-01
-6.23632431e-01 3.82121444e-01 5.38004816e-01 7.66913176e-01
1.21625904e-02 2.63968170e-01 7.84646273e-01 -3.08964521e-01
-1.03159785e+00 -2.88656116e-01 -1.08483279e+00 -3.26814145e-01
6.52887523e-01 -6.44093454e-01 -3.92035753e-01 -7.81006932e-01] | [4.556595802307129, 3.305640459060669] |
f2e41e6a-0710-4ce4-b277-802ac7286bde | spoken-pass-phrase-verification-in-the-i | 1809.11068 | null | http://arxiv.org/abs/1809.11068v1 | http://arxiv.org/pdf/1809.11068v1.pdf | Spoken Pass-Phrase Verification in the i-vector Space | The task of spoken pass-phrase verification is to decide whether a test
utterance contains the same phrase as given enrollment utterances. Beside other
applications, pass-phrase verification can complement an independent speaker
verification subsystem in text-dependent speaker verification. It can also be
used for liveness detection by verifying that the user is able to correctly
respond to a randomly prompted phrase. In this paper, we build on our previous
work on i-vector based text-dependent speaker verification, where we have shown
that i-vectors extracted using phrase specific Hidden Markov Models (HMMs) or
using Deep Neural Network (DNN) based bottle-neck (BN) features help to reject
utterances with wrong pass-phrases. We apply the same i-vector extraction
techniques to the stand-alone task of speaker-independent spoken pass-phrase
classification and verification. The experiments on RSR2015 and RedDots
databases show that very simple scoring techniques (e.g. cosine distance
scoring) applied to such i-vectors can provide results superior to those
previously published on the same data. | ['Jan Cernocky', 'Hossein Zeinali', 'Lukas Burget', 'Hossein Sameti'] | 2018-09-28 | null | null | null | null | ['text-dependent-speaker-verification'] | ['speech'] | [ 2.46959493e-01 -1.69217125e-01 -1.67324334e-01 -8.85880291e-01
-1.30562949e+00 -7.08824039e-01 4.76323336e-01 2.67529190e-01
-4.51246232e-01 4.53161925e-01 4.34360594e-01 -7.22475529e-01
6.63101450e-02 -1.86586305e-01 -2.37214103e-01 -7.41502643e-01
8.49264637e-02 5.09992957e-01 6.39806837e-02 -3.96552563e-01
3.18565935e-01 5.15672863e-01 -1.52798057e+00 1.82622686e-01
1.78568527e-01 8.08618128e-01 -3.43298942e-01 1.21516716e+00
-5.09141721e-02 3.87304902e-01 -9.67176616e-01 -2.47557983e-01
-9.88003537e-02 -5.12454629e-01 -8.70780766e-01 -1.87234506e-01
4.57988113e-01 -1.99653834e-01 -1.19371928e-01 9.50201631e-01
8.83038402e-01 1.98448114e-02 7.32701063e-01 -1.18014073e+00
-4.56917197e-01 7.96372890e-01 -4.00615372e-02 2.96870679e-01
9.51950014e-01 -2.39304021e-01 1.02734518e+00 -1.15937781e+00
1.77473083e-01 1.17669773e+00 7.84581125e-01 1.07874334e+00
-1.21223652e+00 -9.31522131e-01 -2.35276386e-01 4.93055016e-01
-1.76482332e+00 -1.04830337e+00 5.44193625e-01 -1.65504068e-01
1.15617502e+00 6.37148261e-01 1.80139855e-01 1.25024211e+00
-2.05259159e-01 8.62491429e-01 6.75292552e-01 -8.49941730e-01
2.66317546e-01 4.52332824e-01 6.82436824e-01 4.61636752e-01
-5.11902153e-01 2.65520096e-01 -1.10985708e+00 -2.48864010e-01
-8.73612612e-02 -5.39591670e-01 -6.84431911e-01 3.30563635e-01
-1.17914748e+00 1.03109443e+00 -2.99638659e-01 6.73697650e-01
-5.86637445e-02 -3.99807453e-01 6.39127970e-01 5.65644622e-01
-8.18080211e-04 -3.78260463e-02 -6.59401178e-01 -3.76536041e-01
-1.37342882e+00 1.36442855e-01 1.18696642e+00 9.07787204e-01
3.21517169e-01 3.10435414e-01 -5.32972991e-01 7.33523071e-01
7.03107238e-01 7.36698627e-01 8.43027472e-01 -2.53319621e-01
4.23966348e-01 9.06467512e-02 5.32706194e-02 -4.55434978e-01
-3.07661831e-01 -1.21669248e-01 -7.28648603e-01 -3.74899246e-02
3.85566175e-01 -1.81099802e-01 -1.07751143e+00 1.59539676e+00
1.22947738e-01 4.23044572e-03 3.73343378e-01 6.45405948e-01
1.32120097e+00 6.23090029e-01 -1.54516727e-01 -8.26462060e-02
1.51280737e+00 -4.71678615e-01 -1.05465150e+00 2.46051550e-01
6.69851959e-01 -1.00231171e+00 8.71942103e-01 3.92063648e-01
-6.62620604e-01 -5.68191290e-01 -1.18832147e+00 2.63647228e-01
-4.30265963e-01 7.84379914e-02 -1.44289553e-01 1.51884794e+00
-1.44103944e+00 7.39728287e-02 -4.17649090e-01 -3.09344262e-01
-1.35104150e-01 5.98108172e-01 -5.89445889e-01 2.04989150e-01
-1.64618576e+00 7.96543896e-01 1.29296556e-01 1.37467563e-01
-8.92902195e-01 -5.10265470e-01 -1.12401557e+00 1.55663043e-01
-1.43996954e-01 -5.56503795e-02 1.64194000e+00 -5.12508631e-01
-1.82810986e+00 9.55102623e-01 -8.47434878e-01 -5.33826709e-01
1.18223689e-01 2.14669496e-01 -9.11875606e-01 2.13346407e-02
-3.07591796e-01 4.53724325e-01 1.29040039e+00 -6.60214067e-01
-5.48464358e-01 -4.89880472e-01 -7.39175081e-01 -8.91693756e-02
-2.82223612e-01 6.80733860e-01 7.64955878e-02 -2.58870155e-01
3.17382783e-01 -9.24105346e-01 3.15374315e-01 -5.75810671e-01
-9.03539658e-01 -5.81659973e-01 1.11826563e+00 -9.39691901e-01
1.22805572e+00 -2.18642974e+00 -1.26891490e-02 4.44692343e-01
-1.70970008e-01 5.47441602e-01 -4.21731211e-02 4.03152615e-01
-4.11888301e-01 3.20654780e-01 -1.09083846e-01 -7.38683581e-01
3.65910709e-01 7.18771741e-02 -2.73549348e-01 5.93662381e-01
1.49538413e-01 6.74292028e-01 -1.24256238e-01 -4.43482697e-01
1.43426269e-01 7.04865813e-01 4.87452187e-02 3.68538946e-02
3.80531996e-01 3.98723632e-01 1.88603789e-01 5.87407529e-01
6.68750226e-01 3.36402148e-01 -2.67482311e-01 1.95406809e-01
-4.64589931e-02 9.08228993e-01 -1.37104177e+00 1.29923880e+00
-2.49491259e-01 1.02560759e+00 2.53332853e-01 -6.73807263e-01
1.07999182e+00 1.15205252e+00 -2.83410788e-01 -3.93557519e-01
2.63053685e-01 2.55543977e-01 -1.58488788e-02 -3.54259163e-01
6.48611844e-01 -4.05402422e-01 -1.96809635e-01 5.11783481e-01
3.62387508e-01 -1.24575272e-01 -1.81401908e-01 1.29317194e-01
7.86195338e-01 -6.70291722e-01 2.95179814e-01 -4.15024720e-02
1.26547372e+00 -4.45726633e-01 2.26722717e-01 8.17289829e-01
-6.49763942e-01 8.87998760e-01 1.25040919e-01 2.14927107e-01
-7.34873891e-01 -8.29192519e-01 -3.33817214e-01 1.21983528e+00
-5.44170737e-01 -3.95640701e-01 -7.30869412e-01 -6.12547398e-01
-1.79902554e-01 8.00635815e-01 -3.81995320e-01 7.94594735e-02
-3.79507512e-01 -2.81612307e-01 1.29172170e+00 4.89094615e-01
3.13504189e-02 -9.07749474e-01 1.58373311e-01 2.66250134e-01
-2.85529435e-01 -9.49811220e-01 -6.79727614e-01 5.92166543e-01
-3.76104653e-01 -5.03322840e-01 -1.04130113e+00 -1.16898930e+00
1.54710755e-01 -2.25001900e-03 7.92728186e-01 3.28635871e-02
-3.76306772e-02 5.15004694e-01 -5.09912670e-01 -5.47280133e-01
-9.96032774e-01 1.47512764e-01 4.74612862e-01 1.42542511e-01
1.06828427e+00 -8.58476385e-02 -3.23491581e-02 5.88250995e-01
-6.19715810e-01 -7.73443639e-01 5.61170317e-02 7.32696235e-01
7.54693598e-02 -1.22338556e-01 8.09853852e-01 -1.85934991e-01
8.86476338e-01 9.25792605e-02 -3.83855313e-01 3.08339626e-01
-3.05906802e-01 5.27593270e-02 1.35652676e-01 -4.90056723e-01
-5.07848263e-01 3.05441409e-01 -9.27502990e-01 -2.15628281e-01
-6.66496694e-01 5.03596723e-01 -3.28956723e-01 -1.26223266e-01
6.14759147e-01 1.01027215e+00 -1.23499604e-02 -3.85924041e-01
-1.98386069e-02 1.41141176e+00 4.48171616e-01 1.96995020e-01
9.09484982e-01 -1.00480556e-01 -6.09170258e-01 -1.22908890e+00
-3.75990897e-01 -1.13055718e+00 -8.39730144e-01 -1.03091918e-01
9.75805938e-01 -8.28434348e-01 -1.22197223e+00 6.23013616e-01
-1.23200893e+00 3.60446334e-01 6.72540516e-02 6.92177653e-01
-3.16543072e-01 5.96536338e-01 -5.61677754e-01 -1.30314827e+00
-6.31900191e-01 -1.05351698e+00 1.29124916e+00 1.94135249e-01
-7.23118544e-01 -7.74909914e-01 2.24182129e-01 4.77237076e-01
5.54173172e-01 -8.13718557e-01 6.23089731e-01 -1.55305183e+00
1.79604784e-01 -6.37340248e-01 3.43074024e-01 6.66944206e-01
6.84930161e-02 3.92437913e-02 -1.59041536e+00 -2.22736180e-01
4.21382844e-01 -2.41112903e-01 6.96371734e-01 1.75673321e-01
5.59520066e-01 -3.50822747e-01 4.00512330e-02 8.46190825e-02
9.28095281e-01 1.86263323e-01 5.72702527e-01 -2.75008287e-02
2.92809635e-01 5.92904091e-01 2.92527378e-01 2.64447033e-01
1.31464526e-01 7.41767108e-01 -1.69321433e-01 2.02685833e-01
8.09895545e-02 -7.34568164e-02 7.21459627e-01 9.96227980e-01
4.69388366e-01 -4.42075640e-01 -1.18282163e+00 6.34702504e-01
-1.31933010e+00 -1.24253285e+00 -4.76290464e-01 2.25946331e+00
8.89404297e-01 1.29995406e-01 2.07569808e-01 1.15530908e+00
9.32205260e-01 -8.52546692e-02 -1.17624730e-01 -7.47198999e-01
-1.76633418e-01 4.12402362e-01 4.32722926e-01 8.72593999e-01
-1.18116069e+00 6.18632019e-01 6.65266228e+00 7.89575398e-01
-1.38458920e+00 2.19629869e-01 2.32219055e-01 2.21586540e-01
-1.77858118e-02 -4.47551280e-01 -1.57537127e+00 1.23403966e-01
1.59228861e+00 -4.01474237e-02 -2.52560675e-01 6.81364775e-01
3.31919611e-01 1.00041851e-01 -1.44436431e+00 1.28085375e+00
5.58209360e-01 -1.06398332e+00 -2.89445043e-01 -1.85251474e-01
1.51616022e-01 1.31141394e-01 2.06032321e-01 5.59871197e-01
-1.65691093e-01 -1.11896372e+00 8.11240733e-01 -3.35178338e-02
6.81382895e-01 -7.78017759e-01 1.06318498e+00 4.41060662e-01
-1.02217650e+00 2.83639491e-01 -1.56544130e-02 2.54698485e-01
2.28020400e-01 2.14873269e-01 -1.62637568e+00 3.23956102e-01
4.40560639e-01 2.92448342e-01 -3.30002695e-01 9.64640379e-01
-3.93313944e-01 1.13981450e+00 -3.44813734e-01 -3.20787817e-01
9.92908403e-02 5.16189694e-01 8.24344993e-01 1.60463321e+00
1.49007559e-01 -2.71858990e-01 -3.18609864e-01 5.40210485e-01
2.09850758e-01 2.19118163e-01 -4.93556231e-01 5.80557098e-04
2.80630499e-01 9.03398871e-01 -2.26620078e-01 -3.67573351e-01
-1.72992051e-01 1.14727497e+00 -3.49080652e-01 3.14662695e-01
-6.33375764e-01 -7.13390887e-01 5.56027532e-01 -1.09468363e-01
7.36689448e-01 -4.12467092e-01 -6.26508668e-02 -9.13171053e-01
-8.75988044e-03 -9.83500183e-01 2.14260161e-01 -4.18466568e-01
-1.29617620e+00 8.92177403e-01 -4.34637040e-01 -1.31807899e+00
-6.68669224e-01 -4.35342014e-01 -6.89324439e-01 1.29107666e+00
-1.66491890e+00 -7.57305026e-01 2.97526896e-01 9.48823690e-01
5.65953314e-01 -3.97753149e-01 1.50338590e+00 4.61866885e-01
-4.52142447e-01 1.17370152e+00 4.19196896e-02 7.08119452e-01
8.01789165e-01 -1.12228227e+00 4.14809108e-01 7.60110319e-01
4.99370873e-01 6.33316278e-01 8.86006296e-01 -2.21383587e-01
-1.23395646e+00 -7.12424338e-01 1.79978311e+00 -5.20859599e-01
4.59016830e-01 -7.74500608e-01 -1.10868073e+00 4.31747645e-01
1.56676739e-01 -4.28905785e-01 1.27981341e+00 2.28115469e-01
-4.45066899e-01 1.23888582e-01 -1.23269808e+00 9.23741087e-02
2.24151582e-01 -1.11328900e+00 -8.55804861e-01 1.79688543e-01
7.44157076e-01 -1.70909956e-01 -5.91521263e-01 1.33847639e-01
4.36441422e-01 -8.23219001e-01 7.75766373e-01 -7.19844460e-01
-3.54058206e-01 -3.36796939e-01 -4.14356858e-01 -9.44263458e-01
3.92834395e-02 -7.59532869e-01 1.83310851e-01 1.49187779e+00
1.06367111e+00 -4.38264430e-01 8.23304355e-01 7.46756673e-01
1.06534459e-01 2.27283105e-01 -1.67259538e+00 -9.49754119e-01
-2.76650220e-01 -8.07729661e-01 5.56295455e-01 8.32680881e-01
3.45771670e-01 5.88113546e-01 -2.76186585e-01 4.39016998e-01
3.70955884e-01 -4.60448295e-01 5.70734859e-01 -1.11682296e+00
-5.48100993e-02 -3.45044196e-01 -8.44712675e-01 -9.99651670e-01
4.72595721e-01 -7.95635343e-01 5.83854437e-01 -9.85179484e-01
-1.45641968e-01 -1.70250624e-01 -3.26052070e-01 5.96398711e-01
-5.61456531e-02 1.70133531e-01 7.55804181e-02 -7.39702135e-02
-3.22366953e-01 5.71535587e-01 4.24741924e-01 -5.12700737e-01
-3.67138565e-01 6.96302950e-01 -3.30297649e-01 3.32383752e-01
9.28804338e-01 -8.06251347e-01 4.71843779e-03 1.49410188e-01
-2.90754437e-01 3.59875768e-01 8.51966813e-02 -1.04492605e+00
4.72553879e-01 4.70015615e-01 8.10810179e-02 -9.13204908e-01
5.69201052e-01 -8.43761742e-01 -4.92887229e-01 5.43498099e-01
-5.78410625e-01 -8.72663334e-02 3.42040479e-01 4.59417701e-01
-6.47096992e-01 -6.17738903e-01 6.04394376e-01 2.70311415e-01
-4.60145324e-01 -7.04072788e-02 -1.23038316e+00 -4.66020912e-01
5.98062575e-01 -2.75191665e-01 1.24007434e-01 -8.42447221e-01
-9.50373113e-01 -1.33909509e-01 -1.34260163e-01 6.32973552e-01
1.00473118e+00 -1.07671928e+00 -1.14394772e+00 6.46153867e-01
3.17465484e-01 -7.03550220e-01 2.82437857e-02 8.39396417e-01
-2.67211534e-02 8.72920156e-01 2.42131859e-01 -1.11480701e+00
-2.39191175e+00 2.21623436e-01 3.31802189e-01 1.69143438e-01
-1.62456125e-01 1.23780167e+00 -5.25252819e-01 -8.06453347e-01
9.11677241e-01 -4.40403670e-01 -3.84510249e-01 2.38525927e-01
1.08235145e+00 -1.41684666e-01 6.59233510e-01 -1.11751640e+00
-8.75184417e-01 1.85101166e-01 -2.28341997e-01 -7.58467436e-01
1.06962109e+00 -1.94311708e-01 2.28421435e-01 6.22713745e-01
1.50376153e+00 1.83456525e-01 -1.41832545e-01 -2.55187392e-01
2.63565004e-01 -4.42812219e-02 1.86667189e-01 -7.82429814e-01
-5.86527526e-01 1.03454733e+00 8.80336404e-01 3.41870993e-01
7.31159210e-01 1.33874072e-02 6.90629244e-01 5.10449409e-01
9.87401232e-02 -8.29351604e-01 -4.49969560e-01 7.38112271e-01
8.16090882e-01 -1.49616885e+00 -4.89165515e-01 -2.01056208e-02
-6.07258320e-01 1.14855194e+00 -1.68971587e-02 6.10322356e-01
9.88094866e-01 2.16885448e-01 4.62156564e-01 1.29574746e-01
-4.87288892e-01 -3.66895124e-02 4.84956682e-01 8.44521761e-01
5.69224477e-01 2.17735007e-01 1.99298859e-02 4.05515820e-01
-5.33521891e-01 -2.95562416e-01 5.99297166e-01 9.14306045e-01
-5.08560598e-01 -1.35842478e+00 -8.28845501e-01 1.25615776e-01
-5.73961139e-01 -3.35564941e-01 -7.71006107e-01 3.47999722e-01
-4.99408305e-01 1.54463518e+00 -1.66369185e-01 -7.17303514e-01
1.78935349e-01 7.91902006e-01 1.25403494e-01 -7.35448539e-01
-1.01903987e+00 -1.64720312e-01 1.97829932e-01 -1.14593677e-01
-4.25990909e-01 -7.90819645e-01 -1.18222034e+00 -3.20467234e-01
-8.07097673e-01 4.09021437e-01 1.43723500e+00 9.30289924e-01
1.02735817e-01 5.70345558e-02 6.60379767e-01 -5.98724067e-01
-7.37248898e-01 -1.29771900e+00 -6.61457360e-01 4.85486127e-02
9.55214381e-01 -2.24031080e-02 -7.33773232e-01 1.05988957e-01] | [14.341604232788086, 6.0935211181640625] |
94bba269-a19e-4423-95eb-261d0124fc30 | rocnet-recursive-octree-network-for-efficient | 2008.03875 | null | https://arxiv.org/abs/2008.03875v1 | https://arxiv.org/pdf/2008.03875v1.pdf | RocNet: Recursive Octree Network for Efficient 3D Deep Representation | We introduce a deep recursive octree network for the compression of 3D voxel data. Our network compresses a voxel grid of any size down to a very small latent space in an autoencoder-like network. We show results for compressing 32, 64 and 128 grids down to just 80 floats in the latent space. We demonstrate the effectiveness and efficiency of our proposed method on several publicly available datasets with three experiments: 3D shape classification, 3D shape reconstruction, and shape generation. Experimental results show that our algorithm maintains accuracy while consuming less memory with shorter training times compared to existing methods, especially in 3D reconstruction tasks. | ['Juncheng Liu', 'Steven Mills', 'Brendan McCane'] | 2020-08-10 | null | null | null | null | ['3d-shape-retrieval'] | ['computer-vision'] | [-6.20391183e-02 1.52746692e-01 2.13240594e-01 -4.78075355e-01
-7.23382056e-01 -1.08419945e-02 4.04166400e-01 1.47555888e-01
-3.43501449e-01 3.96392226e-01 5.96327424e-01 -2.47942179e-01
-1.86192706e-01 -1.44323647e+00 -8.59932601e-01 -6.05674505e-01
-5.28945923e-01 1.12181079e+00 9.66333225e-02 2.66176134e-01
1.72882915e-01 9.99011874e-01 -1.54080749e+00 4.04723465e-01
2.62502074e-01 1.14705980e+00 1.91507339e-01 7.19699740e-01
3.75261195e-02 6.52904868e-01 -3.04437757e-01 7.30797350e-02
4.73721892e-01 3.54736060e-01 -1.08194339e+00 2.16020681e-02
5.19647300e-01 -7.74474025e-01 -6.27396882e-01 6.79571688e-01
7.25771666e-01 1.94939181e-01 8.02741468e-01 -7.27727354e-01
-5.18068433e-01 6.76233590e-01 -1.67952538e-01 -4.20992486e-02
8.85792747e-02 -3.88522930e-02 6.60566807e-01 -9.63698149e-01
8.00723970e-01 1.28070784e+00 9.36294198e-01 2.75195569e-01
-1.17357814e+00 -4.38601285e-01 -4.56630677e-01 -6.40414432e-02
-1.44807804e+00 -1.03846259e-01 7.37234056e-01 -3.93132299e-01
1.56665468e+00 8.17330405e-02 1.33404303e+00 8.62245560e-01
4.01679009e-01 7.09163249e-01 9.06825602e-01 8.91661197e-02
3.75529677e-01 -6.40024245e-01 -1.61018029e-01 9.43303347e-01
1.27986148e-01 4.35522735e-01 -3.54735047e-01 -4.66699064e-01
1.49460673e+00 -1.17649771e-01 1.15278549e-01 -2.78140515e-01
-1.14028311e+00 1.05986774e+00 7.01183975e-01 9.16496664e-02
-6.17165685e-01 3.64679664e-01 3.93633485e-01 1.31717861e-01
7.89977670e-01 1.82213753e-01 -3.46219987e-01 -1.51935637e-01
-1.08213913e+00 4.14989978e-01 8.69701982e-01 1.09016335e+00
4.49988633e-01 4.23889577e-01 5.97714819e-02 7.40457356e-01
-2.97306702e-02 3.74861956e-01 5.46309114e-01 -1.09672189e+00
2.74064481e-01 3.08455080e-01 -3.87001544e-01 -9.79130685e-01
-5.50118148e-01 -5.01786947e-01 -1.42552209e+00 2.54668742e-01
-2.30658531e-01 2.33530983e-01 -1.10086977e+00 1.15992749e+00
5.88858068e-01 3.46092045e-01 -3.04200966e-02 7.99227834e-01
1.10850561e+00 7.91202068e-01 -1.07057452e-01 1.56281993e-01
9.32432532e-01 -8.93526971e-01 -2.73097247e-01 3.09343815e-01
3.82395059e-01 -7.31558859e-01 7.55255342e-01 4.13213938e-01
-1.47594714e+00 -5.62484920e-01 -8.73354137e-01 -4.48207676e-01
1.84716992e-02 -1.17755920e-01 1.17526650e+00 1.69771597e-01
-1.16779160e+00 1.31773174e+00 -1.13259435e+00 1.65575743e-01
7.76543796e-01 3.17411035e-01 -4.07516062e-01 6.31127274e-04
-6.69491291e-01 4.12876636e-01 4.48697716e-01 -1.37610078e-01
-1.06662035e+00 -1.00958228e+00 -1.02304840e+00 1.21759742e-01
-2.82618403e-01 -1.22172832e+00 1.17324233e+00 2.98853129e-01
-1.48388076e+00 7.71170199e-01 1.16694616e-02 -8.33233356e-01
4.23252910e-01 -6.31856546e-02 2.59692878e-01 1.99055806e-01
-1.29351780e-01 9.05883789e-01 6.76345348e-01 -9.11243618e-01
-2.73209304e-01 -3.74528438e-01 -2.88350463e-01 9.15667787e-02
-5.62935062e-02 -5.73613763e-01 -1.34177387e-01 -7.73001373e-01
8.81360292e-01 -6.31933033e-01 -6.47664666e-01 3.26069146e-02
-4.32321012e-01 -5.08112490e-01 6.54236197e-01 -4.64712977e-01
4.89553124e-01 -1.88265276e+00 1.56831458e-01 2.24327296e-01
5.69502831e-01 -4.25514698e-01 -5.31835705e-02 1.47449911e-01
-4.17713262e-02 -7.60715164e-04 -4.05737668e-01 -6.13901675e-01
-6.76191747e-02 6.28198564e-01 -2.39566565e-01 5.06757498e-01
-3.97145540e-01 1.06548262e+00 -6.74113572e-01 -5.51022649e-01
3.01275849e-01 7.85743833e-01 -9.06122446e-01 2.84639776e-01
-3.26717366e-03 1.97600290e-01 -4.38044310e-01 5.75960994e-01
8.10641527e-01 -6.46830261e-01 -2.94887394e-01 -3.24573159e-01
7.15860352e-02 6.08122408e-01 -1.09020329e+00 1.94998527e+00
-5.12221277e-01 3.81020635e-01 -2.34384939e-01 -9.31783259e-01
9.33534682e-01 3.02456558e-01 8.63393128e-01 -4.37326074e-01
1.08294196e-01 4.01359275e-02 -4.14565682e-01 -1.23614743e-01
5.49176335e-01 -2.65428662e-01 1.22856542e-01 7.70671070e-01
2.67568715e-02 -8.65796328e-01 -1.92984015e-01 2.96138842e-02
1.11960602e+00 -6.98040575e-02 1.13574013e-01 -1.44096613e-01
-1.50717452e-01 -8.43295082e-02 1.75557420e-01 8.31127167e-01
1.81725532e-01 5.99194765e-01 3.45856324e-02 -1.10750067e+00
-1.60906446e+00 -1.17574477e+00 -4.08110857e-01 4.43079084e-01
-2.92370707e-01 -4.44827169e-01 -6.01069927e-01 -1.79551527e-01
3.57592106e-01 3.97167861e-01 -6.13206744e-01 1.90738216e-01
-8.41223419e-01 -3.43222499e-01 5.29349744e-01 7.63940394e-01
5.90283632e-01 -1.21853733e+00 -8.15401912e-01 2.55098671e-01
1.09127678e-01 -1.20158112e+00 -1.59686744e-01 1.91902563e-01
-1.79736233e+00 -8.30064178e-01 -5.06128132e-01 -9.73840892e-01
8.53301883e-01 -5.64853251e-02 1.25765622e+00 1.86256126e-01
-4.21191961e-01 1.29347622e-01 -7.36998916e-02 -1.25565052e-01
-2.80174136e-01 1.35207385e-01 3.61816995e-02 -9.20650661e-01
-1.20876372e-01 -1.28801215e+00 -6.83047295e-01 -1.32893443e-01
-8.51256609e-01 5.35662889e-01 4.57248628e-01 9.54619527e-01
1.20750427e+00 3.90831143e-01 -2.94007361e-01 -8.09977949e-01
4.80879635e-01 -2.77816653e-01 -6.38989031e-01 -4.89597321e-01
-4.43155617e-01 3.88780415e-01 4.93044823e-01 -3.07864696e-01
-6.98177934e-01 2.56418049e-01 -6.82025194e-01 -6.27080977e-01
-3.56193990e-01 3.91133279e-01 4.92133975e-01 -1.00823186e-01
4.67379361e-01 5.34890473e-01 -5.52856326e-02 -7.88555384e-01
2.15435296e-01 2.39203408e-01 6.27786040e-01 -7.70906806e-01
8.80611777e-01 7.57338881e-01 4.01483685e-01 -7.68038094e-01
-6.37942076e-01 2.53745485e-02 -7.82290101e-01 1.86076507e-01
6.06900215e-01 -7.75942743e-01 -8.05327535e-01 3.15560311e-01
-1.21794713e+00 -4.84348148e-01 -8.27378809e-01 6.14774048e-01
-1.06484210e+00 3.63279074e-01 -1.21239352e+00 -1.87812507e-01
-8.59340668e-01 -1.17351699e+00 1.37350059e+00 -3.65898728e-01
3.40734385e-02 -8.13341498e-01 1.74097136e-01 2.70133279e-02
4.76252377e-01 4.43210989e-01 1.28923261e+00 -9.13674831e-02
-9.09190178e-01 -9.53987464e-02 -2.25344017e-01 7.48624699e-03
-1.09596997e-01 -5.31464219e-01 -7.57744253e-01 -3.32555652e-01
1.16585724e-01 -4.87507612e-01 9.51389968e-01 7.25164652e-01
2.02837491e+00 -5.37583530e-01 -1.36380553e-01 1.31192684e+00
1.52895737e+00 -2.08315179e-01 5.07687211e-01 5.61024211e-02
4.60844994e-01 1.29714400e-01 1.32899201e-02 7.70132959e-01
2.56862313e-01 2.60055244e-01 6.11133456e-01 -1.50658295e-01
-3.15230668e-01 -4.77055579e-01 -3.32854867e-01 1.53159726e+00
-3.97807479e-01 1.35785667e-02 -9.09718692e-01 6.24755085e-01
-1.31231308e+00 -6.16276860e-01 9.54638645e-02 1.91879225e+00
9.39544857e-01 1.58801809e-01 -2.67279863e-01 1.80715308e-01
9.89333019e-02 2.58124650e-01 -4.74186957e-01 -3.72618735e-01
4.95279133e-02 7.61253893e-01 4.56735402e-01 5.98858237e-01
-9.42112803e-01 1.07648063e+00 7.40877724e+00 7.76839972e-01
-9.54899311e-01 8.85302797e-02 6.94473982e-01 -1.95071518e-01
-4.98379380e-01 -3.62114012e-01 -5.11346281e-01 -1.21083707e-02
7.10930526e-01 -8.37199911e-02 4.00365800e-01 1.36003339e+00
-6.24082386e-02 1.78320795e-01 -1.14876401e+00 1.29764736e+00
-1.84151530e-01 -1.84047592e+00 3.98027718e-01 2.74309337e-01
7.46332824e-01 3.78812999e-01 5.43724708e-02 6.04505837e-02
4.91360426e-01 -1.34344018e+00 6.11484945e-01 2.77565062e-01
9.76815164e-01 -9.86896932e-01 5.56244016e-01 4.50512856e-01
-1.25293827e+00 4.15298522e-01 -1.09400427e+00 -1.10064127e-01
2.90833652e-01 9.80976105e-01 -1.00574493e+00 1.83754027e-01
8.14416230e-01 5.50332427e-01 -1.42354563e-01 9.53032136e-01
-3.59140895e-02 3.86540294e-01 -8.14964592e-01 9.13784653e-02
3.11623275e-01 1.67227611e-01 5.83024621e-01 9.75667000e-01
5.40310860e-01 5.58816135e-01 2.81951070e-01 9.67539310e-01
-1.60951823e-01 -6.35237023e-02 -9.30667818e-01 2.27178022e-01
4.90221351e-01 8.33623052e-01 -7.94358611e-01 -5.64198375e-01
1.76577017e-01 9.75299656e-01 4.28105801e-01 1.32822404e-02
-5.04960716e-01 -1.35549247e-01 4.03506905e-01 2.64547348e-01
5.50163925e-01 -6.94945812e-01 -4.99240786e-01 -7.52678871e-01
-5.65547571e-02 -3.04602265e-01 2.87355989e-01 -8.34728420e-01
-1.23250735e+00 9.83607590e-01 1.18643403e-01 -1.02226615e+00
-3.56748611e-01 -5.57149708e-01 -1.99212506e-01 4.24984306e-01
-1.43081760e+00 -1.02932131e+00 -3.65358979e-01 5.65101802e-01
5.25917351e-01 -2.21995249e-01 1.16705012e+00 7.17448890e-02
2.11600333e-01 4.20327574e-01 -9.11433771e-02 1.77115977e-01
-1.13990061e-01 -1.22216988e+00 1.01548135e+00 2.01340988e-01
3.53016585e-01 4.52548414e-01 3.07127625e-01 -9.07311738e-01
-1.61385512e+00 -9.70213771e-01 9.03787553e-01 1.61321506e-01
1.90672994e-01 -1.64192349e-01 -7.64699996e-01 7.61978030e-01
-2.51224115e-02 3.08585852e-01 5.43443441e-01 1.54580116e-01
-2.41129488e-01 3.23489040e-01 -1.37253582e+00 1.87894329e-01
1.47254086e+00 -4.36666101e-01 -5.95159352e-01 6.41315520e-01
9.70814705e-01 -1.20144367e+00 -1.41096270e+00 5.74036479e-01
6.88927710e-01 -7.97860742e-01 1.46915686e+00 -6.44672692e-01
9.40577686e-01 1.91998392e-01 -4.05569553e-01 -1.11463141e+00
-6.92098081e-01 -2.00578392e-01 -3.98275465e-01 8.32303241e-02
-8.15655738e-02 -2.20661551e-01 1.42844439e+00 3.44928682e-01
-2.81684816e-01 -1.36128151e+00 -1.32011080e+00 -5.30491471e-01
2.16061026e-01 -8.86520743e-01 1.15160215e+00 7.25735188e-01
-6.73055947e-01 -1.16946079e-01 -2.42634356e-01 1.39951095e-01
1.07355011e+00 6.93218708e-01 6.77163899e-01 -1.24758649e+00
-3.06887716e-01 -2.28749305e-01 -5.69555819e-01 -1.42058527e+00
5.67451455e-02 -1.32093060e+00 -2.01347619e-01 -1.66681767e+00
3.35285187e-01 -6.78743184e-01 -7.25083938e-03 5.57846546e-01
4.86043036e-01 4.77268606e-01 -1.18505701e-01 1.29461035e-01
-1.88195705e-01 8.90368104e-01 1.61539268e+00 -2.69065559e-01
-1.08666939e-03 -1.90897003e-01 -1.41207259e-02 9.34692204e-01
5.84963799e-01 -6.50558591e-01 -2.11615592e-01 -9.93279517e-01
9.23993886e-02 3.80172700e-01 4.63509291e-01 -1.13586199e+00
2.84388989e-01 1.90353487e-02 9.19642031e-01 -1.46977961e+00
4.94163990e-01 -7.11639762e-01 2.16873825e-01 6.41113698e-01
-2.23914608e-01 1.55750513e-01 2.54551351e-01 2.70799339e-01
-5.93441278e-02 -1.63428932e-01 6.79970682e-01 -5.02584875e-01
-3.95660549e-01 8.83303940e-01 -6.51890486e-02 -3.09410810e-01
4.45694447e-01 -3.78676862e-01 2.76083589e-01 -4.39440340e-01
-1.15765595e+00 -1.32522196e-01 1.71360508e-01 1.43937662e-01
1.37379873e+00 -1.79656649e+00 -8.11534882e-01 7.16608644e-01
-4.11171675e-01 6.57431901e-01 4.62603837e-01 1.23209767e-01
-9.03885603e-01 5.93900561e-01 -3.44769984e-01 -9.99842584e-01
-1.09687698e+00 1.12061203e-01 2.44445771e-01 -5.45896173e-01
-1.51422989e+00 1.06881976e+00 -8.63018110e-02 -6.61004663e-01
4.55921382e-01 -6.14044905e-01 1.16393723e-01 -4.81044203e-01
3.09212416e-01 3.39889288e-01 1.57214850e-01 -2.83078343e-01
7.14494288e-02 7.28634119e-01 1.57716617e-01 8.37350786e-02
2.10292315e+00 5.07638752e-01 -2.60549217e-01 1.62169293e-01
1.52275491e+00 -3.14245284e-01 -1.04072404e+00 -2.45216966e-01
-6.04795516e-01 -5.50657868e-01 4.88790512e-01 -2.32678384e-01
-1.30730891e+00 7.49055505e-01 6.86478794e-01 -1.34600297e-01
8.84039938e-01 1.77102700e-01 1.38056791e+00 6.92124605e-01
7.51851201e-01 -7.93526113e-01 3.19956839e-02 8.32117558e-01
1.14912045e+00 -8.71724188e-01 4.87437546e-01 -3.34925711e-01
1.31698912e-02 1.23204505e+00 2.22832143e-01 -5.00440717e-01
1.10021448e+00 7.11623251e-01 -5.49660146e-01 -6.10134840e-01
-8.42142880e-01 3.14134657e-01 2.63085783e-01 4.32269692e-01
1.76750392e-01 2.06666321e-01 1.47228345e-01 2.87036538e-01
-1.13075030e+00 3.16926204e-02 1.61643192e-01 8.90420437e-01
-3.41825962e-01 -1.05233300e+00 -3.55112776e-02 7.04276145e-01
-9.67938974e-02 -7.27726072e-02 1.88631043e-01 6.45107150e-01
-9.85540301e-02 -1.34724781e-01 5.10020852e-01 -3.83022040e-01
6.76750243e-02 -3.44970040e-02 8.71197820e-01 -5.57521939e-01
-2.98556238e-01 1.33857906e-01 -3.15836757e-01 -8.82499754e-01
-1.89560577e-01 -4.58709449e-01 -1.33429933e+00 -5.53437233e-01
3.71978506e-02 6.64135348e-03 7.51966417e-01 4.92219210e-01
3.86585861e-01 5.27931869e-01 5.51819742e-01 -1.13380921e+00
-6.23725414e-01 -1.01905036e+00 -5.40036440e-01 3.34803522e-01
2.19328001e-01 -5.99027514e-01 -6.75738677e-02 -3.95112298e-02] | [8.512042045593262, -3.646287202835083] |
fab4107f-934c-4a2e-9f88-8a7869a908ae | anti-exploration-by-random-network | 2301.13616 | null | https://arxiv.org/abs/2301.13616v2 | https://arxiv.org/pdf/2301.13616v2.pdf | Anti-Exploration by Random Network Distillation | Despite the success of Random Network Distillation (RND) in various domains, it was shown as not discriminative enough to be used as an uncertainty estimator for penalizing out-of-distribution actions in offline reinforcement learning. In this paper, we revisit these results and show that, with a naive choice of conditioning for the RND prior, it becomes infeasible for the actor to effectively minimize the anti-exploration bonus and discriminativity is not an issue. We show that this limitation can be avoided with conditioning based on Feature-wise Linear Modulation (FiLM), resulting in a simple and efficient ensemble-free algorithm based on Soft Actor-Critic. We evaluate it on the D4RL benchmark, showing that it is capable of achieving performance comparable to ensemble-based methods and outperforming ensemble-free approaches by a wide margin. | ['Sergey Kolesnikov', 'Denis Tarasov', 'Vladislav Kurenkov', 'Alexander Nikulin'] | 2023-01-31 | null | null | null | null | ['d4rl'] | ['robots'] | [ 2.48108253e-01 3.31229478e-01 -3.68324444e-02 -2.51322627e-01
-7.93793201e-01 -6.29510164e-01 9.28550422e-01 -2.61764750e-02
-8.52615476e-01 1.37417817e+00 -1.90239865e-02 -5.05321145e-01
-4.44264442e-01 -3.87471944e-01 -6.53591156e-01 -1.03017867e+00
-2.45736033e-01 4.87780601e-01 -1.13134846e-01 -2.13508070e-01
1.46917224e-01 4.62469608e-01 -1.54471421e+00 -2.62035340e-01
8.89521241e-01 8.29348028e-01 6.14612438e-02 4.86005306e-01
4.08515692e-01 8.62285018e-01 -9.52127039e-01 -1.02102891e-01
4.25031632e-01 -8.20972264e-01 -4.56037879e-01 -2.39102364e-01
1.36544809e-01 -4.66002882e-01 -3.45915854e-01 9.73910451e-01
9.01296198e-01 5.37801325e-01 7.92669535e-01 -9.72476244e-01
-7.94637129e-02 8.70308042e-01 -2.72979051e-01 9.33437496e-02
1.10794261e-01 1.91757128e-01 1.09315252e+00 -4.25900966e-01
4.86972243e-01 1.19314408e+00 3.76086861e-01 7.53477156e-01
-1.63215840e+00 -7.27860034e-01 1.84922367e-01 -1.35903060e-01
-1.22702384e+00 -4.99966949e-01 4.91517454e-01 -1.30984768e-01
1.02551651e+00 1.41455904e-01 6.65776968e-01 1.30748951e+00
1.94602147e-01 7.36508071e-01 1.56651700e+00 -6.49564862e-01
8.06562781e-01 3.03380843e-02 -3.33092183e-01 5.80870628e-01
2.37983406e-01 7.38719404e-01 -3.02468419e-01 -3.81526724e-03
6.69941187e-01 -6.73287868e-01 -9.31099653e-02 -5.67170262e-01
-8.55917513e-01 9.98586059e-01 2.59796947e-01 2.39553243e-01
-3.99242073e-01 5.76881707e-01 3.49043787e-01 4.99603868e-01
4.33311522e-01 7.85515964e-01 -2.44536579e-01 -4.63279486e-01
-9.80521023e-01 5.64713180e-01 7.45691597e-01 3.67622882e-01
2.49484435e-01 6.40586972e-01 -2.96827734e-01 6.00402892e-01
1.97669983e-01 4.87060279e-01 3.76743108e-01 -1.46257842e+00
2.10939467e-01 1.23644382e-01 2.67931879e-01 -4.61427033e-01
-4.75864470e-01 -6.94659173e-01 -6.92311704e-01 9.60833848e-01
6.80535734e-01 -7.11220562e-01 -9.23735321e-01 2.21360397e+00
1.36912137e-01 1.96664125e-01 1.22815780e-01 7.33573079e-01
1.21107452e-01 3.13811004e-01 5.41765951e-02 -4.60378885e-01
7.09592819e-01 -4.91051227e-01 -7.44285703e-01 -2.52014250e-01
6.47181153e-01 -1.53336346e-01 6.78660154e-01 8.38555694e-01
-9.41469073e-01 -3.36053520e-01 -1.44961226e+00 6.66448832e-01
-3.98629718e-02 -1.90387383e-01 6.26219690e-01 9.16942060e-01
-8.20075214e-01 1.19374919e+00 -9.04572785e-01 8.60913098e-02
4.75378215e-01 5.65362394e-01 -3.15069944e-01 2.98445553e-01
-1.33174193e+00 1.38012004e+00 8.22456300e-01 -5.72738945e-02
-1.26843727e+00 -4.04316515e-01 -6.02279365e-01 -1.55019641e-01
5.27669430e-01 -4.66376841e-01 1.29308844e+00 -1.17102194e+00
-2.16471982e+00 2.22096637e-01 3.16668183e-01 -1.09232199e+00
6.63410544e-01 -2.01293275e-01 -2.15535879e-01 1.63800758e-03
-4.77174193e-01 6.79752529e-01 1.00176442e+00 -1.14458096e+00
-5.28259933e-01 -1.99730709e-01 2.71401525e-01 4.35537577e-01
-7.17277229e-02 -2.93512821e-01 1.53714433e-01 -6.35648608e-01
-3.63409638e-01 -1.18981206e+00 -5.82892001e-01 -4.74525779e-01
-2.91178107e-01 -1.73419029e-01 5.93146086e-01 -5.21797657e-01
1.33568943e+00 -1.78253746e+00 2.25634754e-01 3.40017527e-01
-8.52834061e-02 4.40106988e-01 -1.81216180e-01 2.71098763e-01
-2.26561844e-01 1.05973169e-01 -4.47494507e-01 -3.54209691e-01
3.42651606e-01 4.88869399e-01 -2.62561768e-01 6.12701416e-01
2.63240755e-01 6.60629332e-01 -1.00232637e+00 -2.40408495e-01
4.42104220e-01 4.34565574e-01 -7.04621613e-01 -3.19835002e-04
-4.38464016e-01 6.21020138e-01 -3.66253048e-01 1.32118478e-01
3.56055319e-01 1.69409305e-01 5.30825317e-01 3.19356203e-01
1.44871566e-02 2.47738928e-01 -1.40817869e+00 1.68868887e+00
-5.48326194e-01 5.68253756e-01 -1.10033154e-02 -1.28762615e+00
7.06217587e-01 3.60148668e-01 4.17421281e-01 -7.02136338e-01
3.11688632e-01 3.97075862e-01 3.76468450e-01 8.98061991e-02
2.28344083e-01 -2.61513054e-01 -1.10177010e-01 4.66235459e-01
4.31663185e-01 -2.47553691e-01 2.80881137e-01 -7.81802759e-02
1.26358759e+00 5.03345549e-01 2.60694385e-01 -5.02604485e-01
4.77271497e-01 -4.10437852e-01 3.94801199e-01 9.64169025e-01
-9.39354897e-02 3.47551376e-01 6.72155261e-01 2.01246604e-01
-1.08118618e+00 -1.01014447e+00 -2.39805147e-01 7.60441542e-01
-1.83595955e-01 -3.22738230e-01 -8.89872730e-01 -9.63150203e-01
1.78572983e-01 1.35279906e+00 -6.80947065e-01 -2.54657507e-01
-4.83038634e-01 -9.14380789e-01 6.18967533e-01 4.27075714e-01
2.96799332e-01 -8.79554510e-01 -7.45130122e-01 3.77002925e-01
2.86502182e-01 -7.34273672e-01 7.39461705e-02 9.59617138e-01
-8.65243316e-01 -8.13669801e-01 -7.64511585e-01 1.15680404e-01
3.76651615e-01 -4.18249995e-01 9.73334849e-01 -4.32788193e-01
1.15485534e-01 4.47255790e-01 -6.73275143e-02 -5.21573067e-01
-6.46872103e-01 8.22657347e-03 2.76467323e-01 -3.35192889e-01
5.33186235e-02 -8.30112278e-01 -6.13922119e-01 1.54677480e-01
-7.06171811e-01 -4.30906028e-01 4.82266754e-01 1.21363926e+00
2.39783540e-01 9.54742581e-02 8.56017172e-01 -7.01336026e-01
8.92717004e-01 -3.15617859e-01 -8.02225530e-01 -1.40040861e-02
-9.39840436e-01 5.91114163e-01 7.13428199e-01 -3.78354847e-01
-1.06212127e+00 -3.55060473e-02 -2.31719017e-01 -5.28548479e-01
5.40249376e-03 2.17682183e-01 8.70878249e-02 -1.49042696e-01
7.80280828e-01 4.27899286e-02 2.71680862e-01 -2.73864418e-01
3.42224360e-01 4.58903342e-01 9.72995386e-02 -6.59043908e-01
6.21788800e-01 1.56700209e-01 3.03767145e-01 -6.03356361e-01
-7.53328741e-01 2.20048532e-01 -2.03032449e-01 -3.64902109e-01
8.03603768e-01 -8.87489021e-01 -8.09366822e-01 3.97357270e-02
-6.33494616e-01 -5.89308500e-01 -7.97566891e-01 7.05080807e-01
-9.97084558e-01 1.67683676e-01 -2.49024481e-01 -1.28663814e+00
7.39952326e-02 -1.08815753e+00 5.00637472e-01 2.04581857e-01
-2.59576797e-01 -1.03548491e+00 3.35473299e-01 -1.00359403e-01
5.46511114e-01 4.79522198e-01 5.62350750e-01 -1.03285241e+00
-1.41683623e-01 -4.55988832e-02 3.68192196e-01 8.42441559e-01
-2.49526665e-01 -7.96747431e-02 -1.08075118e+00 -4.67751950e-01
-9.77698192e-02 -6.36138797e-01 9.73344088e-01 4.00975704e-01
1.11853230e+00 -6.13730513e-02 3.81542705e-02 3.28373849e-01
1.41361928e+00 3.36510986e-01 5.99555314e-01 3.53247702e-01
1.52504966e-01 3.66030306e-01 4.63128358e-01 6.31637216e-01
-2.14721456e-01 7.42984295e-01 6.39058769e-01 3.14386070e-01
1.87057376e-01 -3.09946984e-01 5.72527349e-01 2.70954639e-01
-3.58164579e-01 -5.17786741e-01 -5.48756957e-01 2.31672704e-01
-1.91960764e+00 -1.04260397e+00 3.03575516e-01 2.36908746e+00
7.60923862e-01 4.38772947e-01 2.44016573e-01 1.19296677e-01
4.67267156e-01 2.25549549e-01 -6.92330301e-01 -6.26474023e-01
-2.31475741e-01 6.16474152e-01 7.40387022e-01 5.83795547e-01
-1.10850739e+00 5.67215204e-01 7.40855360e+00 1.22253728e+00
-6.08744621e-01 1.65241119e-02 5.78866243e-01 -2.68540919e-01
-2.19192058e-01 1.99250236e-01 -6.93751812e-01 5.00597894e-01
1.29585195e+00 1.12740286e-01 7.41719365e-01 6.79309845e-01
1.14203811e-01 -4.41608399e-01 -1.05890667e+00 6.49439514e-01
-1.74087659e-01 -9.69608665e-01 -5.46036780e-01 1.50277555e-01
9.00948584e-01 3.13694179e-02 2.07586586e-02 7.58199751e-01
6.98451161e-01 -1.37000597e+00 6.68130994e-01 3.71457815e-01
4.90367502e-01 -1.09632707e+00 5.88194549e-01 5.34509122e-01
-4.01611149e-01 -2.74516165e-01 -2.28907123e-01 -6.14939854e-02
1.18766725e-01 4.29056466e-01 -5.25445282e-01 5.36929369e-01
1.87790036e-01 3.29925805e-01 -8.89948681e-02 7.90576637e-01
-3.42513770e-01 7.14731991e-01 -5.94783008e-01 -2.25494966e-01
5.55332720e-01 -2.12348282e-01 7.71204770e-01 1.13098359e+00
3.83500576e-01 -6.53993040e-02 -7.29449615e-02 6.90959573e-01
2.96984855e-02 -2.67006099e-01 -6.54199541e-01 -1.67392313e-01
2.61509538e-01 1.00898170e+00 -3.89812052e-01 -5.15028164e-02
7.88361207e-02 6.86844170e-01 3.87888700e-01 2.72379309e-01
-9.76130486e-01 -2.88222313e-01 4.37777996e-01 -3.00943315e-01
5.97668350e-01 -2.91622221e-01 -3.75488587e-02 -8.67563128e-01
-2.83403963e-01 -1.13909745e+00 2.47467220e-01 -3.80618215e-01
-1.00633717e+00 4.61483300e-01 2.75564820e-01 -1.10934699e+00
-8.46105695e-01 -7.19878912e-01 -3.35367560e-01 8.20292473e-01
-1.23197997e+00 -4.17436332e-01 3.34641367e-01 4.33243930e-01
6.68633580e-02 -3.87004793e-01 9.71819103e-01 9.38858837e-02
-6.41246259e-01 5.16938090e-01 5.90419948e-01 -3.80763501e-01
5.17888963e-01 -1.60299480e+00 -1.84233785e-01 6.59886897e-01
1.72416180e-01 3.35694075e-01 1.25270486e+00 -3.53272229e-01
-1.15549207e+00 -7.00324357e-01 1.51648819e-01 -4.62416619e-01
6.15599573e-01 -2.52210051e-01 -4.56128716e-01 5.06317675e-01
5.07221162e-01 -1.10001154e-01 3.64958137e-01 5.40748388e-02
5.61876129e-03 -4.88821007e-02 -1.11996174e+00 6.59757555e-01
9.34893906e-01 -1.77753031e-01 -5.66561460e-01 1.87688485e-01
2.57096410e-01 -3.75281215e-01 -8.34479988e-01 5.82348228e-01
4.67052311e-01 -1.17202401e+00 8.17202210e-01 -6.48187220e-01
2.30078295e-01 -8.83498415e-02 -2.78730541e-01 -1.69449663e+00
-6.64598942e-02 -9.14512157e-01 -5.12740254e-01 1.04883766e+00
3.30119103e-01 -7.26559401e-01 7.36249566e-01 4.18002069e-01
-3.75557728e-02 -9.07071114e-01 -1.44493413e+00 -1.04816568e+00
4.88079101e-01 -5.36327899e-01 2.05623463e-01 5.06925702e-01
5.47449626e-02 3.80832553e-01 -5.83339572e-01 -3.18140596e-01
7.25808978e-01 -2.88511366e-01 6.18966341e-01 -1.03184247e+00
-7.63456106e-01 -5.96135259e-01 -2.16107979e-01 -1.04844439e+00
5.05267143e-01 -7.79331207e-01 9.30064395e-02 -1.03329945e+00
-2.65559942e-01 -5.40051043e-01 -7.01585174e-01 2.46080890e-01
1.19011827e-01 -2.67381221e-03 2.45187476e-01 -2.57827222e-01
-5.32300770e-01 7.91618288e-01 1.16103983e+00 4.19472530e-02
-8.72550383e-02 4.70676012e-02 -4.69051927e-01 7.20903456e-01
8.54380786e-01 -6.24231100e-01 -6.17682576e-01 1.88298687e-01
3.95928234e-01 2.63866276e-01 4.03121322e-01 -1.17853975e+00
-2.05829874e-01 -5.01286425e-02 4.47602868e-01 -3.67273420e-01
4.72666889e-01 -7.81764269e-01 9.72419828e-02 6.01723850e-01
-6.30506516e-01 -2.15861261e-01 3.95002455e-01 8.13606143e-01
-1.59808751e-02 -6.85084939e-01 8.79491329e-01 -3.09703425e-02
-4.45753753e-01 -1.11713544e-01 -5.26414096e-01 3.41066986e-01
1.10823488e+00 4.04943563e-02 -7.64567312e-03 -4.57540840e-01
-6.89246595e-01 1.98342770e-01 3.31209630e-01 6.32185861e-02
1.83087453e-01 -1.25513387e+00 -7.49961436e-01 -1.24114277e-02
-3.37360442e-01 -3.89157891e-01 2.31595546e-01 7.43383944e-01
-2.78629303e-01 4.62225825e-01 -1.83877215e-01 -3.82757157e-01
-9.21667695e-01 3.44347268e-01 7.32078433e-01 -7.10734129e-01
-5.07443786e-01 6.98921621e-01 -2.66201854e-01 -3.96614254e-01
4.35516983e-01 9.63076428e-02 -7.83859044e-02 -8.89661014e-02
2.31650814e-01 3.31851482e-01 2.67482782e-03 -9.68367904e-02
-3.16441476e-01 1.45811275e-01 1.53162643e-01 -6.40710533e-01
1.39559937e+00 1.97045684e-01 4.77948785e-01 5.57491124e-01
8.97322357e-01 -8.78487155e-02 -1.67632437e+00 2.14869171e-01
-1.38115529e-02 -2.42474124e-01 3.53126198e-01 -1.03081465e+00
-7.47535288e-01 8.12878668e-01 8.29799652e-01 2.89692134e-01
1.02670014e+00 -5.24315476e-01 1.96483433e-01 5.56674480e-01
4.74507630e-01 -1.54345703e+00 -8.64548683e-02 5.97703755e-01
7.88202226e-01 -1.11362743e+00 2.93911368e-01 4.62795198e-01
-7.07327306e-01 1.04733801e+00 4.60024863e-01 -6.10756814e-01
5.83987713e-01 3.69931996e-01 -4.32334840e-01 2.76255488e-01
-1.02419829e+00 -2.23021761e-01 2.37660199e-01 5.78684330e-01
1.70475125e-01 -3.14702094e-02 -4.32856709e-01 1.40868589e-01
7.06206635e-02 -1.75698832e-01 3.41208071e-01 9.43340421e-01
-4.27481294e-01 -1.35399580e+00 -8.85554627e-02 4.35470313e-01
-6.62675023e-01 4.90601435e-02 -2.98470668e-02 1.14480269e+00
2.51638424e-02 7.81224608e-01 4.52534370e-02 -3.76027256e-01
1.23062566e-01 3.11717659e-01 9.79741216e-01 -1.36311769e-01
-6.96734190e-01 -1.72265455e-01 5.45919955e-01 -5.63761890e-01
-5.43888688e-01 -7.09222198e-01 -9.82195795e-01 -1.59298763e-01
-5.82298398e-01 1.99018642e-01 6.68083131e-01 9.94742155e-01
2.41906166e-01 7.07983911e-01 6.27190471e-01 -8.76235962e-01
-1.35937583e+00 -9.33449209e-01 -7.30057955e-01 3.11779361e-02
3.46439749e-01 -9.96549547e-01 -8.26497793e-01 -7.05036938e-01] | [4.161418437957764, 2.3536648750305176] |
500566d7-9e93-4786-b54d-26b96b9b1268 | face-presentation-attack-detection | 2212.0368 | null | https://arxiv.org/abs/2212.03680v1 | https://arxiv.org/pdf/2212.03680v1.pdf | Face Presentation Attack Detection | Face recognition technology has been widely used in daily interactive applications such as checking-in and mobile payment due to its convenience and high accuracy. However, its vulnerability to presentation attacks (PAs) limits its reliable use in ultra-secure applicational scenarios. A presentation attack is first defined in ISO standard as: a presentation to the biometric data capture subsystem with the goal of interfering with the operation of the biometric system. Specifically, PAs range from simple 2D print, replay and more sophisticated 3D masks and partial masks. To defend the face recognition systems against PAs, both academia and industry have paid extensive attention to developing face presentation attack detection (PAD) technology (or namely `face anti-spoofing (FAS)'). | ['Zhen Lei', 'Chenxu Zhao', 'Zitong Yu'] | 2022-12-07 | null | null | null | null | ['face-presentation-attack-detection', 'face-anti-spoofing'] | ['computer-vision', 'computer-vision'] | [ 0.699676 -0.12883623 0.07982825 -0.14368641 -0.07816652 -0.95875365
0.7166943 -0.2400868 -0.02447005 0.3801353 -0.21925238 -0.64875853
-0.04899801 -0.5525583 -0.2129689 -0.83132017 -0.04061539 -0.32384518
0.12842086 -0.06652182 0.3227649 1.1154596 -1.620072 0.3623388
0.18522315 1.1156361 -0.59531575 0.5583341 -0.17156549 -0.06573397
-0.83484 -0.89208853 0.3220579 -0.28713733 -0.42310482 -0.25009713
0.24145013 -0.45643285 -0.09752899 1.0947475 0.47700146 -0.42951727
0.45292953 -1.4701815 -0.526098 0.06288316 -0.87040657 0.13479911
0.84770346 0.13853863 0.02728435 -0.9800652 0.24741173 1.2880385
0.48321563 0.94314194 -1.0776918 -1.0974745 -0.54732704 -0.10535251
-1.5956383 -0.7658057 0.7042883 -0.26548538 0.7723745 0.55613595
0.09888923 1.01647 0.5207849 0.05746 1.2903805 -0.46440175
-0.06668947 0.5433294 0.127634 0.42404696 0.7016698 0.14778645
-0.40966693 -0.56582654 0.7959852 -0.08835357 -0.15003589 0.04935388
-0.40902233 0.35929817 -0.19305894 0.4340279 -0.41617215 -0.30388823
0.31068632 0.31102425 -0.08000708 -0.19849168 -0.07213549 0.03489554
-0.38156572 0.03899468 0.842641 0.5103801 0.36910963 0.05750161
0.14666378 0.56970835 0.7475926 0.86171246 0.1912829 -0.32321745
0.26750645 0.5057028 0.14580141 -1.0066184 0.06891818 0.25084218
-0.6953682 0.40794176 0.31706783 -0.16698556 -0.66938865 1.2637281
0.3055181 0.10594165 -0.0100625 0.4121298 0.7874106 0.25819612
0.21677727 -0.21778545 1.5379255 0.15651116 -0.97000647 0.11011219
0.18571024 -1.2943013 0.80404204 0.37060767 -0.79724646 -0.4235408
-1.2618698 0.5230895 -0.5119229 -0.2817523 0.3726098 2.0042057
-0.9433478 0.10020532 -0.3797607 -0.3727603 0.606873 1.0914314
-0.88739777 0.0690815 -0.9242995 0.72081655 -0.28436187 0.23004769
-0.20014791 -0.32599083 -0.72935325 -0.11163648 -0.13296865 -0.05365765
0.8593795 -0.8642664 -1.7483735 1.1100546 -0.09823848 -0.25566778
0.176742 -0.25920278 -1.1506785 0.15907998 -0.48892573 0.03899177
1.175848 -0.8571771 0.01334977 -0.7303819 -0.18237968 -0.3765066
-0.39928412 0.8512601 0.1431019 -0.461509 0.09327219 -0.82692885
0.61387324 -0.06785963 -0.2770759 -0.01359372 1.7328454 -0.6679117
1.215243 -2.294151 -0.71089303 0.6218842 -0.44040135 1.1570596
0.13121152 0.4632993 -0.28341556 0.44694752 0.089895 0.12364338
-0.19183896 0.0698606 -0.30335563 0.7735323 0.25775394 0.52909863
-0.21584411 -0.1758966 0.3690145 0.8883033 -0.06263504 0.0310033
0.56817925 0.20949328 -0.44098157 0.9922646 1.2273786 0.37047434
0.2737385 -0.06744132 0.0629862 0.11835486 -1.2227259 0.5414359
-0.02840482 0.4350249 0.521129 -0.3657661 0.9878342 0.7347105
0.31766406 -0.57473654 0.55299574 0.26194686 0.12075406 -0.6838575
0.08101127 0.01966434 0.28027898 0.57750946 -0.19507459 0.38962808
-0.5074196 -0.16015562 0.6440471 -0.11379648 0.275019 -0.41876107
1.2873371 -0.8990697 0.03583058 0.19659384 -0.56276137 0.09548145
0.08199285 -0.08070281 -0.47349614 -0.9706531 -0.27407408 0.43800473
0.09315681 -0.17162924 -1.0223553 -0.8929103 0.19859593 0.11382214
-0.2532281 -0.10829737 -0.67829084 -0.23353206 1.0672649 0.28960028
0.71430475 -0.91203225 -0.5470486 -0.15291412 0.3496461 -0.9840845
-0.3545077 -0.4433468 -0.51308596 -1.348507 -0.5670611 -0.5743401
0.7306006 0.35070607 0.25296584 0.59464264 -0.53422844 0.575069
-0.10077182 -0.661046 -0.66100496 -0.31610927 0.36442086 0.44728923
0.8700843 -0.30540973 -0.65770245 0.6529257 -0.95486677 -0.78077966
0.25092763 0.39028457 -0.26314104 0.03525013 0.4717172 -0.87054217
0.75917023 -0.13494527 -0.25157475 0.3076293 -0.3324737 -0.41011178
0.17100988 -0.41182032 -1.0266308 -0.08969261 -0.37981692 0.01892159
-0.4859779 -0.11316538 -0.84898543 -0.9807964 0.44443676 0.13765328
0.5221777 -0.269802 -0.30494457 1.1920254 0.4338586 -0.33330444
1.0360644 0.34766707 0.18690404 -1.1668721 0.10228191 -0.5250902
-0.3650341 -0.50360656 0.43489888 -0.29550362 -1.320312 1.1154705
-0.9396327 0.46611986 0.731151 0.21497445 0.04820411 0.76311433
-0.53020513 -1.351612 -0.6332983 -1.0030845 0.65252095 0.54049766
-0.29523158 -0.6410518 -0.27048168 0.2523533 0.7892963 0.35224694
0.7490692 -0.2988604 -0.12538284 -0.7396982 -0.23169592 0.4107702
0.7409681 0.331055 -1.37076 -0.3865514 0.3725826 0.11323113
0.08501184 0.08572102 0.8082846 -0.466949 -0.2884555 0.3898416
1.4762845 0.66598296 1.0460958 0.01317541 0.1724786 0.87688315
0.37733403 0.2478271 -0.38666293 0.5709049 0.32083553 0.28574964
0.21677175 -0.09415042 0.5574584 -0.18491438 -0.09574351 -0.24042271
-0.6550258 -0.16088794 -1.1504858 -0.92600167 -0.11880016 2.5273352
0.18303235 0.06615536 0.174131 0.6518736 1.0481725 -0.11184406
-0.13240016 -0.9090669 -0.01167384 0.6958049 0.5489027 0.26782975
-1.1044102 0.6459646 6.4708195 0.43627548 -1.4644306 -0.01986565
0.5143553 0.3217042 -0.06175082 -0.35505438 -0.90724 0.5457666
0.8372252 0.03916708 0.2762785 0.5384666 -0.05761368 -0.16477236
-0.68219835 1.3129843 0.20409764 -0.91404444 0.26884562 0.56124216
0.08832414 -0.79633224 0.38627264 -0.22221799 -0.50949556 -1.0237023
0.18997231 0.02322169 0.9889184 -0.9114709 0.7570943 -0.13512349
-1.2724937 -0.1014582 0.05038434 -0.05245441 0.19702914 -0.10556679
-0.81396043 0.23210368 0.39557803 -0.41622207 -0.33704272 0.58575654
0.11143711 0.48802742 -0.51214486 0.04898079 -0.16962722 -0.02567913
0.4110154 0.9418332 0.32754263 0.35186073 -0.5497468 0.44416356
0.20842113 0.03040472 -0.8453492 -0.2954711 0.50470877 1.0644952
-0.6454031 0.23027311 -0.4130417 0.9652475 -0.669832 0.08887085
-0.4822091 -0.3772505 0.8745985 0.5041073 0.134918 -0.30287886
-0.30928728 -0.43156487 0.11045615 -0.92077595 0.39727104 -0.17668363
-1.0106807 0.4866365 -0.334589 -0.8899307 -0.10277279 -0.82694715
-0.6283091 1.3087299 -0.96971285 -1.4161327 -0.02296617 0.8203536
0.04271613 -0.35384798 1.0458272 0.48627636 -0.47948027 1.0302147
-0.57854337 0.19043478 0.53674084 -0.44400927 0.41229907 0.92756706
-0.2885371 1.2033325 0.5091704 -0.58642215 -1.9085232 -0.25534725
0.95157725 -0.36737323 0.1238563 -0.1869493 -0.7944082 0.22878385
0.07174107 -0.2601359 1.0707881 -0.3828933 -0.49149638 -0.1857021
-1.8836851 0.3657597 0.52248085 -0.9538999 -0.12699813 -0.28753805
-0.12962872 0.05997961 -0.6125788 0.44448903 1.2793167 -0.9395188
1.0508345 -0.2517966 -0.3971591 -0.18165642 -0.10804083 -0.06221887
-0.01579288 -1.1305548 -0.16444097 1.671604 -0.07398704 -0.94267416
0.8787683 1.03022 0.7024591 -0.18705562 -0.8802514 -0.6273297
-0.5902494 -0.10187472 0.8982973 0.7800333 0.09779572 -0.21527915
-0.32111266 0.7542753 0.6837795 -0.5893738 0.7235602 -1.2162557
0.012215 -0.36326537 -0.59357244 -0.40035054 -0.2276848 -0.22220601
-0.4842187 -0.41068584 -0.30770215 -0.20312525 -0.41948473 0.0824536
0.28956646 0.6890543 0.09633578 0.05635197 0.43317935 -0.14673215
0.6846408 0.1533364 -0.08687024 0.46897858 -0.4300569 0.4921815
0.7448284 -0.29503292 -0.35118893 0.07511456 -0.11324503 0.04533535
0.36184785 -0.85783094 -0.09465174 -0.03253586 0.36883223 -0.46290815
0.14874572 -1.1246259 0.41784585 0.7043569 0.33635622 0.18547721
0.33515328 0.03750035 0.1095139 -0.28520533 0.7448219 0.11819129
-0.32211652 0.23338704 -0.5345405 -0.9035889 1.1687175 -0.9460856
-0.46899152 -0.1130376 -0.52202976 -0.5318704 0.6707235 0.48265994
0.76517785 -0.9278433 -0.23484312 1.1475201 -0.12416757 -0.7858885
0.2492285 0.3834911 -0.7961931 0.4737971 -0.4216958 -0.3037631
-2.2103455 0.42294535 0.17708531 0.39185506 -0.20739245 0.6565286
-0.07956219 0.23088685 0.26101452 0.45338905 -0.35887805 -0.35107303
1.2628651 0.51200324 0.2708207 -0.8481457 -0.7248705 0.4797359
-0.04951588 -0.13252082 0.71071285 -0.09556387 -0.26570144 -0.2690449
1.0430931 0.1910345 -0.67632234 0.21675248 0.1486916 -0.89643633
-0.36909062 -0.65131503 -0.7041472 0.7674437 1.0584561 0.50374174
0.8734354 -0.4914924 0.716862 0.17450878 0.43674645 -0.6402555
-0.19102634 0.00659476 0.6355653 -0.8433191 -0.23367207 -0.7798632
-0.26719704 1.1044691 0.38027495 0.11337324 1.1757241 0.47989932
0.22195566 -0.24039109 -0.13428621 0.53620994 0.29273665 1.0165095
0.51936597 -0.12936515 -0.5301651 0.43768078 0.08438571 0.0942601
0.20342 1.113327 -0.24830468 -1.5236056 -0.70687634 0.26994163
-1.0053917 0.3629736 -0.78078556 0.6041113 0.20813656 1.0732365
-0.16944985 -0.40822688 0.12688585 0.34814355 0.6057104 -0.24156699
-1.0037073 -0.02056429 0.01851368 -0.34335968 -0.6226972 -0.5702009
-0.7139265 -0.8210343 -0.37283367 -0.0511608 1.0706806 0.984489
0.36872786 -0.2608375 0.76705366 -0.3593596 -0.40821066 -0.7941551
-0.72173434 0.30422395 0.29378995 -0.53466046 -0.12641132 0.06301395] | [13.12984848022461, 1.1315404176712036] |
5edda055-0c3c-48d4-ac3e-f637a45fb814 | nonlinear-hyperspectral-unmixing-with-robust | 1401.5649 | null | http://arxiv.org/abs/1401.5649v2 | http://arxiv.org/pdf/1401.5649v2.pdf | Nonlinear hyperspectral unmixing with robust nonnegative matrix factorization | This paper introduces a robust mixing model to describe hyperspectral data
resulting from the mixture of several pure spectral signatures. This new model
not only generalizes the commonly used linear mixing model, but also allows for
possible nonlinear effects to be easily handled, relying on mild assumptions
regarding these nonlinearities. The standard nonnegativity and sum-to-one
constraints inherent to spectral unmixing are coupled with a group-sparse
constraint imposed on the nonlinearity component. This results in a new form of
robust nonnegative matrix factorization. The data fidelity term is expressed as
a beta-divergence, a continuous family of dissimilarity measures that takes the
squared Euclidean distance and the generalized Kullback-Leibler divergence as
special cases. The penalized objective is minimized with a block-coordinate
descent that involves majorization-minimization updates. Simulation results
obtained on synthetic and real data show that the proposed strategy competes
with state-of-the-art linear and nonlinear unmixing methods. | ['Nicolas Dobigeon', 'Cédric Févotte'] | 2014-01-22 | null | null | null | null | ['hyperspectral-unmixing'] | ['computer-vision'] | [ 4.84287411e-01 -5.26402593e-01 -9.03677046e-02 -3.47340941e-01
-4.34054226e-01 -7.17730284e-01 6.30539894e-01 -1.92192912e-01
-3.54434192e-01 8.34831357e-01 -1.23628870e-01 -1.55362427e-01
-7.18947709e-01 -4.40258563e-01 -2.83235282e-01 -1.25806451e+00
-2.09713921e-01 1.93811640e-01 -3.24525803e-01 -1.48282468e-01
6.89527616e-02 8.04943800e-01 -1.46586287e+00 -1.62495121e-01
1.41355371e+00 9.08409536e-01 2.13155940e-01 4.84864324e-01
2.26628169e-01 5.81524074e-01 -2.96534657e-01 -1.33140162e-01
7.31965721e-01 -4.84249413e-01 -2.91009784e-01 5.39975047e-01
4.67112154e-01 -7.03377873e-02 -2.29232356e-01 1.55670619e+00
4.01389629e-01 4.85995054e-01 7.00570285e-01 -1.38428891e+00
-5.89649796e-01 3.14117581e-01 -8.46875906e-01 -6.58184513e-02
1.74127780e-02 -1.82785660e-01 9.55819011e-01 -9.25482094e-01
3.53018165e-01 1.06066930e+00 7.51555622e-01 -3.02293114e-02
-1.52568877e+00 -4.76537168e-01 -6.98755085e-02 2.48064864e-02
-1.63266456e+00 -3.55213910e-01 8.27917159e-01 -6.68398023e-01
5.77746212e-01 7.26357341e-01 5.00952184e-01 2.80921191e-01
-2.57968511e-02 3.50767910e-01 1.22180665e+00 -5.71898997e-01
1.94086790e-01 1.22102290e-01 1.63224056e-01 5.62755167e-01
5.80271065e-01 1.12151131e-01 -8.31040293e-02 -5.90244472e-01
4.33608621e-01 9.29992422e-02 -7.79514492e-01 -8.62690389e-01
-1.12330127e+00 8.42611492e-01 3.44269991e-01 3.13950479e-01
-5.38548827e-01 -3.09487522e-01 -8.34231898e-02 3.02634060e-01
4.78013933e-01 3.74167442e-01 -4.37948480e-02 3.92321765e-01
-1.15191519e+00 6.50533810e-02 8.40368330e-01 7.98335433e-01
1.01019466e+00 5.03358722e-01 8.23017135e-02 1.01345742e+00
5.79247177e-01 8.79808247e-01 3.51307362e-01 -8.74068260e-01
4.11446989e-01 5.03191471e-01 3.11201543e-01 -1.01283884e+00
-5.09785652e-01 -6.20289981e-01 -1.20873427e+00 2.53690422e-01
2.05188096e-01 -1.43638894e-01 -8.15146029e-01 1.83921552e+00
1.94414347e-01 3.73504728e-01 1.93750948e-01 9.77071404e-01
3.80046040e-01 8.49965513e-01 -3.48686904e-01 -9.94318485e-01
7.54211605e-01 -7.88035929e-01 -1.05276835e+00 1.29039735e-01
1.70110706e-02 -9.42081153e-01 6.86152697e-01 5.04366815e-01
-9.50503051e-01 -3.02968234e-01 -1.32382429e+00 2.85009772e-01
-2.39983797e-01 1.48568273e-01 5.16571701e-01 9.73573804e-01
-9.34215546e-01 5.81905723e-01 -8.24060261e-01 -1.92417964e-01
-2.70272732e-01 4.44832861e-01 -4.84190226e-01 -4.59386557e-02
-8.11218917e-01 7.97569454e-01 3.95805568e-01 7.25689232e-01
-5.67231417e-01 -7.46681571e-01 -6.95411026e-01 -6.87741861e-02
1.82817832e-01 -6.20225132e-01 6.60556555e-01 -1.03426147e+00
-1.73331618e+00 5.80987334e-01 -1.28009215e-01 -7.86418617e-02
5.68702996e-01 -3.40022892e-02 -7.93382883e-01 1.51207194e-01
-1.44420207e-01 1.48148453e-02 1.00138581e+00 -1.53211915e+00
-8.61293748e-02 -3.14549565e-01 -2.08807915e-01 4.02620494e-01
-5.26861966e-01 -7.84253106e-02 -7.39824250e-02 -8.25191140e-01
5.99733293e-01 -1.18974280e+00 -4.25058603e-01 -1.81897253e-01
-4.61358517e-01 5.54548979e-01 6.90498054e-01 -6.77132010e-01
1.12739718e+00 -2.19170237e+00 6.41190231e-01 7.40472734e-01
-2.78872997e-02 2.45681405e-01 -2.89839655e-01 6.34919405e-01
-5.29367864e-01 -1.74355611e-01 -9.63764310e-01 -2.71820784e-01
-5.58098182e-02 4.31110337e-02 -2.29000911e-01 9.68023002e-01
1.27225148e-03 2.10838556e-01 -8.97145212e-01 1.61752298e-01
3.34599704e-01 5.68390429e-01 -3.34981799e-01 1.55861467e-01
2.34839857e-01 3.36073965e-01 7.43862689e-02 4.93038625e-01
1.20395875e+00 -7.23538995e-02 2.54128009e-01 -5.51268399e-01
-4.62998152e-01 -4.15768117e-01 -1.77942753e+00 1.67019057e+00
-7.65619650e-02 1.87347606e-01 6.38735890e-01 -1.21292758e+00
7.10198045e-01 3.20345283e-01 6.33035600e-01 1.62552118e-01
3.20502743e-02 4.99766886e-01 1.52976289e-01 -3.15548241e-01
4.85521078e-01 -5.76687574e-01 4.37162906e-01 2.13571161e-01
1.31739467e-01 -4.45111185e-01 3.76575530e-01 1.09185735e-02
3.94284308e-01 -1.12534866e-01 5.00981987e-01 -8.52926195e-01
8.46419096e-01 -1.29161313e-01 5.95909476e-01 5.08258522e-01
2.06949249e-01 4.99201477e-01 7.42594600e-02 2.12991580e-01
-6.93198740e-01 -1.00889897e+00 -3.79170954e-01 5.37616670e-01
1.53129742e-01 -5.39013818e-02 -4.69272077e-01 -1.76312625e-01
1.11503683e-01 5.45422971e-01 -4.19049084e-01 -8.27794522e-02
-8.01421106e-02 -1.47265780e+00 2.23119646e-01 -8.31598490e-02
3.99721950e-01 -3.33109885e-01 1.40386358e-01 6.86886609e-02
-6.68832883e-02 -7.51288295e-01 -3.32414389e-01 3.73417556e-01
-1.07680094e+00 -9.21157777e-01 -7.84746289e-01 -5.69272041e-01
8.67151141e-01 7.19074428e-01 5.42731464e-01 -4.42851275e-01
-3.19422543e-01 2.44047031e-01 -1.93754539e-01 -5.85234538e-02
-4.75170910e-01 -3.57637584e-01 5.00888765e-01 6.04809642e-01
1.41567498e-01 -6.75367355e-01 -3.53717297e-01 2.84066647e-01
-1.30540192e+00 -1.94817424e-01 4.71183628e-01 8.93481851e-01
6.06373131e-01 4.82787639e-01 2.99318016e-01 -5.98465621e-01
6.04066312e-01 -3.76342207e-01 -8.36541831e-01 4.91439193e-01
-7.82275856e-01 -3.93638061e-03 5.52253425e-01 -6.05843782e-01
-1.04417920e+00 2.01532587e-01 4.14078325e-01 -5.86541235e-01
1.65768787e-01 8.29160869e-01 -4.74806190e-01 -5.68436086e-01
6.76020026e-01 2.94484138e-01 4.69130054e-02 -5.33711135e-01
6.73094928e-01 7.50499010e-01 5.74004412e-01 -4.43514645e-01
1.25787592e+00 6.84190333e-01 2.80732274e-01 -1.35207939e+00
-4.46066737e-01 -7.53199816e-01 -6.45320952e-01 -7.27012157e-02
5.49003899e-01 -9.76861298e-01 -4.55744386e-01 7.72963047e-01
-6.83409214e-01 1.18008740e-01 -3.78118038e-01 9.29423928e-01
-4.03044403e-01 7.64191329e-01 -4.68848377e-01 -9.46157873e-01
-7.74727017e-02 -1.01617610e+00 4.42075044e-01 1.04973957e-01
2.50541866e-01 -9.83148813e-01 1.81614101e-01 1.80596963e-01
5.03193259e-01 2.81704068e-01 8.00444841e-01 -3.77489328e-01
-3.12274754e-01 -2.43250147e-01 -1.83355957e-01 9.66569364e-01
3.72656494e-01 2.13635206e-01 -9.32520390e-01 -5.85784614e-01
3.83798182e-01 9.59900692e-02 9.73454475e-01 6.42503262e-01
5.90868711e-01 -2.87764043e-01 5.94289554e-03 9.46141481e-01
1.67176974e+00 1.49743348e-01 3.53269100e-01 -4.60172817e-02
8.42286110e-01 4.40783590e-01 2.76909202e-01 5.84017634e-01
-1.79751888e-01 4.36780751e-01 5.48766792e-01 -2.81078726e-01
3.89429778e-01 2.56843954e-01 2.55513787e-01 1.24079168e+00
-3.97156566e-01 -1.73301324e-01 -5.67375481e-01 2.70314872e-01
-1.98765159e+00 -1.12808895e+00 -5.00503719e-01 2.62895298e+00
6.40918136e-01 -4.62279797e-01 -1.42066583e-01 3.87155265e-01
8.68497193e-01 3.46303105e-01 -4.47655320e-01 -4.08521853e-03
-6.30520940e-01 4.85350825e-02 8.59577119e-01 1.01509774e+00
-1.29436350e+00 5.25937796e-01 6.51126623e+00 6.12243474e-01
-1.21996307e+00 2.58518085e-02 -1.94282770e-01 4.85609658e-02
-4.03448850e-01 4.54937108e-03 -2.92848945e-01 2.34844223e-01
5.43276191e-01 -2.35769987e-01 8.00024211e-01 3.64539593e-01
3.10902029e-01 1.30934706e-02 -6.39056683e-01 1.06199777e+00
3.40328604e-01 -7.60599434e-01 3.38788107e-02 2.07936555e-01
1.05485833e+00 -1.69484206e-02 1.90581962e-01 -3.10696393e-01
7.33881518e-02 -6.81782484e-01 6.88687146e-01 6.74755871e-01
6.58411503e-01 -7.69004941e-01 5.38366795e-01 2.07397982e-01
-1.23769331e+00 -2.91208804e-01 -4.90133405e-01 1.87944397e-02
3.06256831e-01 1.04625940e+00 -1.82359189e-01 9.71654296e-01
5.80931120e-02 7.59632349e-01 -1.58101857e-01 1.32969952e+00
-1.41831279e-01 2.82049954e-01 -4.73362774e-01 4.43067372e-01
2.02466071e-01 -1.21360970e+00 8.84407580e-01 1.27152884e+00
8.85277212e-01 2.59719223e-01 2.18551531e-01 6.70821667e-01
1.33426860e-01 3.89944851e-01 -4.97230917e-01 -3.53480905e-01
2.17716560e-01 1.48357129e+00 -5.49859703e-01 -2.03625292e-01
-3.85744900e-01 1.13958049e+00 -2.74208784e-01 6.66830122e-01
-5.47505975e-01 -2.87842244e-01 7.12335825e-01 -1.39256537e-01
1.72704294e-01 -3.71602565e-01 3.33461873e-02 -1.63704348e+00
7.57055059e-02 -9.38265085e-01 3.15599412e-01 -5.33094049e-01
-1.42965519e+00 5.49124718e-01 7.07658231e-02 -1.59074378e+00
-4.15212959e-02 -6.85961187e-01 -3.47159773e-01 1.17977118e+00
-1.65107191e+00 -1.03666580e+00 -3.74496132e-01 7.02526093e-01
-1.80430502e-01 -2.86984503e-01 1.00555551e+00 4.27898794e-01
-7.41821587e-01 9.23004374e-02 7.90066719e-01 -5.37158191e-01
7.22277761e-01 -1.27218735e+00 -4.59570557e-01 1.20765138e+00
-5.71942069e-02 8.65469813e-01 9.62626517e-01 -4.71302420e-01
-1.27479041e+00 -1.02378070e+00 6.08409882e-01 3.14776540e-01
8.44229281e-01 -6.68056831e-02 -8.73989403e-01 4.65721488e-01
2.79045194e-01 1.40467465e-01 1.14016771e+00 -1.61716834e-01
-3.61846924e-01 -2.71447480e-01 -1.23319387e+00 3.06751788e-01
7.50554562e-01 -5.73011458e-01 -2.52584726e-01 5.94323456e-01
2.33713761e-01 -4.11930382e-02 -9.46879864e-01 6.25787377e-01
4.49046761e-01 -8.15186322e-01 1.02082038e+00 -6.26354337e-01
-2.42788181e-01 -8.92088950e-01 -5.99384964e-01 -1.53161478e+00
-6.98176205e-01 -1.01285982e+00 -8.32098871e-02 9.98540759e-01
5.17216206e-01 -6.68587923e-01 3.74320477e-01 4.91425991e-01
-7.77530298e-02 -1.42595768e-01 -7.10630894e-01 -1.11084640e+00
-1.69538587e-01 -1.35072947e-01 3.55363131e-01 1.45512509e+00
9.61776674e-02 3.58548015e-02 -7.10501611e-01 6.86452806e-01
9.52882469e-01 1.31836101e-01 3.43278885e-01 -1.54759049e+00
-4.35780704e-01 -4.62907314e-01 -4.29155529e-01 -7.68001854e-01
2.77184784e-01 -9.91712391e-01 -3.45366411e-02 -1.23850322e+00
3.52152772e-02 -1.92600355e-01 -4.42047834e-01 1.99455202e-01
-1.54209659e-01 3.66314948e-01 2.09390253e-01 2.88376063e-01
1.45796657e-01 6.03644848e-01 9.01025712e-01 -3.12656224e-01
-5.74772358e-01 1.42754763e-01 -5.89963257e-01 7.27683961e-01
5.36275327e-01 -2.55185723e-01 -6.42613769e-01 -2.13560611e-01
2.72574693e-01 9.37252864e-02 2.17884947e-02 -1.11812174e+00
2.21773118e-01 -4.64725435e-01 7.50464201e-02 -3.47768903e-01
5.91962099e-01 -1.17294371e+00 7.55935133e-01 3.30973595e-01
-7.93765187e-02 -1.23995878e-01 1.64747424e-02 5.29835105e-01
-4.28604931e-01 -4.56839889e-01 1.08322477e+00 1.90413490e-01
-3.06611270e-01 2.60144889e-01 -3.04334790e-01 -4.23330367e-01
1.03906202e+00 -1.96493641e-01 -1.70272872e-01 -3.25300932e-01
-9.08761561e-01 -2.23565832e-01 4.67430562e-01 1.56729475e-01
3.35698247e-01 -1.47254181e+00 -9.82772112e-01 4.02489841e-01
5.53123318e-02 -4.53496814e-01 4.09852028e-01 1.22966433e+00
-5.69213867e-01 1.53168619e-01 -1.58038959e-01 -4.78061259e-01
-1.21780968e+00 5.48876524e-01 4.87991542e-01 -8.90657678e-02
-6.75365925e-02 7.95615017e-01 4.45234030e-02 -5.68128884e-01
-1.25002889e-02 -2.26689786e-01 -7.65120424e-03 3.06797773e-01
5.71140468e-01 6.73660338e-01 1.66611820e-01 -1.14579523e+00
-3.48192871e-01 5.80668926e-01 4.41541672e-01 -3.34460676e-01
1.32068074e+00 -8.78447518e-02 -7.39219606e-01 4.71116781e-01
1.24607182e+00 2.47284129e-01 -9.44543660e-01 -3.06806326e-01
-2.73572981e-01 -5.26652694e-01 1.83165491e-01 -6.50373161e-01
-9.70798969e-01 6.98877156e-01 7.94102848e-01 4.78871137e-01
1.33030403e+00 -7.90692389e-01 5.01478240e-02 4.13554072e-01
6.82609305e-02 -8.62601459e-01 -5.56317568e-01 4.47970808e-01
1.07751918e+00 -1.05107367e+00 3.43050838e-01 -6.51247263e-01
-3.52248281e-01 1.01907432e+00 6.33969828e-02 3.44023518e-02
9.84692335e-01 1.82713419e-01 1.52910769e-01 2.92109132e-01
3.84837985e-02 -2.44629860e-01 5.97969174e-01 4.73651588e-01
4.54908073e-01 3.69576991e-01 -6.27771616e-01 2.31449634e-01
4.42537628e-02 -2.43438527e-01 4.94790256e-01 6.83890939e-01
-4.79971439e-01 -1.12271774e+00 -7.07677245e-01 3.02980483e-01
-2.22965658e-01 -1.88555688e-01 -1.80776149e-01 4.52871323e-01
9.84859392e-02 1.17842996e+00 -3.11827183e-01 -1.77028582e-01
2.18560308e-01 1.16114870e-01 4.32133973e-01 -4.88150388e-01
-2.30524257e-01 4.68638569e-01 -2.19597906e-01 -2.51696348e-01
-1.00896657e+00 -6.52660251e-01 -8.06675315e-01 -1.81888282e-01
-8.57414663e-01 4.87076372e-01 9.45885956e-01 6.34804070e-01
1.16519727e-01 1.31967977e-01 8.26464891e-01 -1.12494290e+00
-8.67633581e-01 -1.14897621e+00 -1.30014265e+00 3.79404426e-01
4.86647367e-01 -5.62908828e-01 -7.71056712e-01 2.18763098e-01] | [10.03104305267334, -2.0573439598083496] |
2f0b3d0c-08b3-4698-89ce-21e4120704c8 | semantic-aligned-fusion-transformer-for-one | 2203.09093 | null | https://arxiv.org/abs/2203.09093v2 | https://arxiv.org/pdf/2203.09093v2.pdf | Semantic-aligned Fusion Transformer for One-shot Object Detection | One-shot object detection aims at detecting novel objects according to merely one given instance. With extreme data scarcity, current approaches explore various feature fusions to obtain directly transferable meta-knowledge. Yet, their performances are often unsatisfactory. In this paper, we attribute this to inappropriate correlation methods that misalign query-support semantics by overlooking spatial structures and scale variances. Upon analysis, we leverage the attention mechanism and propose a simple but effective architecture named Semantic-aligned Fusion Transformer (SaFT) to resolve these issues. Specifically, we equip SaFT with a vertical fusion module (VFM) for cross-scale semantic enhancement and a horizontal fusion module (HFM) for cross-sample feature fusion. Together, they broaden the vision for each feature point from the support to a whole augmented feature pyramid from the query, facilitating semantic-aligned associations. Extensive experiments on multiple benchmarks demonstrate the superiority of our framework. Without fine-tuning on novel classes, it brings significant performance gains to one-stage baselines, lifting state-of-the-art results to a higher level. | ['Yan Lu', 'Xun Guo', 'Yizhou Zhao'] | 2022-03-17 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Zhao_Semantic-Aligned_Fusion_Transformer_for_One-Shot_Object_Detection_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Zhao_Semantic-Aligned_Fusion_Transformer_for_One-Shot_Object_Detection_CVPR_2022_paper.pdf | cvpr-2022-1 | ['one-shot-object-detection'] | ['computer-vision'] | [ 3.40251476e-01 -2.69559711e-01 -1.52681470e-01 -5.42372108e-01
-1.09778392e+00 -2.93170899e-01 6.22959077e-01 4.55525443e-02
-1.37373403e-01 3.40961814e-01 1.80267945e-01 1.62334159e-01
-1.80406556e-01 -6.30421042e-01 -7.41474092e-01 -6.09523833e-01
3.44206601e-01 -1.06567398e-01 7.59115517e-01 -2.17059717e-01
2.50852406e-01 2.66503125e-01 -1.84115767e+00 5.61388671e-01
1.00724494e+00 1.20849836e+00 3.64388913e-01 1.13052435e-01
-2.74213105e-01 4.89194810e-01 -3.33092570e-01 -5.12374043e-01
2.76374727e-01 -1.99810825e-02 -5.96301734e-01 1.40255004e-01
6.34107411e-01 -3.70652407e-01 -2.22221196e-01 1.33048022e+00
4.36872303e-01 1.18876770e-02 2.75644243e-01 -1.39525270e+00
-7.71208167e-01 4.30515826e-01 -1.00601530e+00 3.13118100e-01
2.60911167e-01 3.53372127e-01 1.19357121e+00 -1.48226905e+00
4.16185051e-01 1.34799445e+00 6.42255723e-01 6.77631944e-02
-1.17559099e+00 -7.14887977e-01 6.08508468e-01 4.12590176e-01
-1.48779809e+00 -5.58156610e-01 8.39795709e-01 -1.09028064e-01
8.30460846e-01 2.63770193e-01 4.83092010e-01 9.59980130e-01
-1.51364207e-01 1.04357958e+00 1.16006517e+00 -1.13845460e-01
1.07681558e-01 8.54551718e-02 1.99843451e-01 5.56450009e-01
3.53438556e-01 -1.49127319e-01 -8.56301844e-01 6.71384111e-03
5.20018399e-01 3.77971530e-01 -1.67286038e-01 -5.87375104e-01
-1.35627449e+00 6.32682502e-01 8.74540627e-01 2.82765239e-01
-4.90792096e-01 -7.87054673e-02 2.57897913e-01 5.89627363e-02
3.56556594e-01 3.12367946e-01 -5.91447592e-01 2.24876925e-01
-8.90746713e-01 2.67140895e-01 1.88273758e-01 1.12679517e+00
1.03216088e+00 -2.60848135e-01 -5.50080538e-01 8.01315606e-01
9.89434421e-02 3.88542533e-01 3.82677406e-01 -7.56557524e-01
4.14573878e-01 1.00056195e+00 -9.37194154e-02 -1.14372337e+00
-4.10587549e-01 -9.45844293e-01 -5.50107956e-01 -1.00148790e-01
1.02422245e-01 2.88682163e-01 -9.47032213e-01 1.68619025e+00
6.49741292e-01 4.77367938e-01 -6.12701327e-02 1.05216742e+00
8.09906542e-01 2.68652797e-01 1.18950680e-01 -4.94193472e-02
1.65656912e+00 -1.27928388e+00 -4.54189181e-01 -3.54984909e-01
5.88483214e-01 -8.06606770e-01 1.22253525e+00 2.07957700e-01
-9.71890807e-01 -7.65185416e-01 -1.10530794e+00 -1.00001879e-01
-5.45583785e-01 4.32893857e-02 7.77908385e-01 3.52307230e-01
-7.37123311e-01 4.08751607e-01 -7.83396721e-01 -4.26224232e-01
6.17376924e-01 1.34938002e-01 -3.40641826e-01 -2.25460440e-01
-1.03141356e+00 6.59996092e-01 5.17887294e-01 9.43754334e-03
-4.61827606e-01 -8.92149210e-01 -7.31313884e-01 1.06156565e-01
8.63529325e-01 -9.30007398e-01 1.08697057e+00 -7.62873232e-01
-9.05817389e-01 6.43287599e-01 -2.64935285e-01 -2.97724247e-01
3.62127036e-01 -4.17569578e-01 -5.04601657e-01 2.49819025e-01
5.57434738e-01 8.42176974e-01 1.02054739e+00 -1.42892992e+00
-1.16245413e+00 -5.51660955e-01 1.17658511e-01 2.65509278e-01
-5.80088735e-01 -6.16287179e-02 -7.65017331e-01 -8.84448528e-01
5.24836779e-01 -5.83789527e-01 -1.15347914e-01 7.06339329e-02
-4.31991518e-01 -1.33709833e-01 9.31007326e-01 -3.91858965e-01
1.29114997e+00 -2.37955856e+00 -1.84146017e-02 -4.60202806e-02
3.41577709e-01 2.77702719e-01 -2.78762937e-01 7.96644986e-02
2.77962983e-02 -2.06954926e-01 -2.85350829e-01 -2.65944004e-01
-2.61257053e-03 5.53405732e-02 -5.02273500e-01 2.25627065e-01
6.09469235e-01 1.42190719e+00 -9.20570731e-01 -5.91267884e-01
2.55388528e-01 4.15260732e-01 -7.14544475e-01 -1.88444927e-02
-6.59071133e-02 2.03718722e-01 -6.99470162e-01 1.08174157e+00
8.79198015e-01 -5.60708702e-01 -1.88022986e-01 -7.66283512e-01
-6.48635477e-02 4.24753763e-02 -1.19827211e+00 2.16342473e+00
-2.19246656e-01 5.23157716e-02 -1.59333274e-01 -9.25055683e-01
9.16820824e-01 -2.58650154e-01 4.46637541e-01 -8.85728478e-01
-1.96457207e-02 2.85687178e-01 -1.42692178e-01 -3.71711403e-01
6.04437590e-01 3.35117951e-02 -2.59600859e-03 4.33456972e-02
2.86290288e-01 3.07752371e-01 -4.25489619e-02 3.09315920e-01
1.00913548e+00 1.97459698e-01 2.71547794e-01 -2.22386375e-01
5.34991920e-01 -2.03801721e-01 6.72359943e-01 8.07191133e-01
-3.28758359e-01 6.20253444e-01 4.29767035e-02 -2.59488493e-01
-6.61575437e-01 -1.33841431e+00 -4.22080792e-02 1.40144444e+00
6.88766003e-01 -5.94365835e-01 -5.37578821e-01 -8.32743466e-01
2.05963105e-01 5.15818238e-01 -7.27219045e-01 -4.02173817e-01
-4.08577442e-01 -7.68873155e-01 2.55457401e-01 8.11954319e-01
7.80931652e-01 -6.73368335e-01 -7.20819890e-01 2.34083578e-01
-2.21456513e-01 -1.22943866e+00 -5.10677516e-01 6.44997433e-02
-7.05831289e-01 -8.69768322e-01 -6.40547216e-01 -6.54831827e-01
3.92434925e-01 1.02428401e+00 8.71571839e-01 3.44959535e-02
-4.24657464e-01 2.39672720e-01 -5.86495161e-01 -4.17792648e-01
2.96428323e-01 1.00664698e-01 -9.01752338e-03 3.49991411e-01
4.98635888e-01 -6.06105983e-01 -8.51809442e-01 4.19642538e-01
-8.89577210e-01 1.53348938e-01 8.05151701e-01 8.07381451e-01
5.75683653e-01 -6.35166615e-02 7.88344204e-01 -4.79008168e-01
1.41806021e-01 -4.04153079e-01 -2.96782106e-01 4.32788253e-01
-6.08545721e-01 -7.34419078e-02 2.75121868e-01 -2.84073919e-01
-1.12249780e+00 5.18268719e-02 2.39310026e-01 -6.13422751e-01
-2.56642196e-02 2.04882130e-01 -4.26211447e-01 -5.32115549e-02
4.54634190e-01 3.22636038e-01 -1.57023191e-01 -6.61733031e-01
6.31119192e-01 4.39047724e-01 8.03059280e-01 -4.88065451e-01
9.85635459e-01 7.91606247e-01 -2.59128153e-01 -4.10790831e-01
-1.38671398e+00 -6.63329601e-01 -6.76802456e-01 3.68068367e-03
6.49369895e-01 -1.25537169e+00 -4.30332720e-01 3.70020449e-01
-8.85239780e-01 4.00084198e-01 -4.82282460e-01 3.06875050e-01
-3.18644911e-01 1.68220863e-01 -1.97046787e-01 -5.75693130e-01
-2.47692510e-01 -1.13183749e+00 1.53566933e+00 4.32563782e-01
1.55297861e-01 -4.81333464e-01 -3.41582119e-01 4.96169209e-01
5.71759820e-01 -3.12770344e-02 6.12750351e-01 -6.93610907e-01
-7.94594049e-01 -3.65483109e-03 -7.69023538e-01 1.13652170e-01
2.89340913e-01 -1.76562995e-01 -1.22441888e+00 -3.39590371e-01
-4.53430135e-03 -2.38148838e-01 1.31422555e+00 1.43766537e-01
1.22588468e+00 4.25012894e-02 -5.58808923e-01 7.94487059e-01
1.31704068e+00 -1.73872426e-01 2.29516253e-01 5.24960041e-01
8.35171044e-01 5.33492863e-01 9.21766281e-01 4.99540001e-01
6.84889674e-01 7.84864962e-01 5.64957201e-01 -2.48960868e-01
-3.20587218e-01 -2.28640378e-01 2.45830953e-01 4.58211243e-01
2.07137048e-01 1.96121007e-01 -6.00349665e-01 6.04078352e-01
-1.96762228e+00 -8.32266033e-01 1.04503982e-01 1.88978958e+00
6.11424029e-01 3.02425504e-01 1.74827680e-01 -2.20721774e-02
7.71801472e-01 1.74455822e-01 -7.51863539e-01 4.15000439e-01
-3.70155483e-01 -3.12315803e-02 3.12357128e-01 -2.22696681e-02
-1.23089743e+00 1.04264081e+00 5.15812349e+00 1.17216599e+00
-1.09021711e+00 3.11356395e-01 3.49178135e-01 -1.57642156e-01
-3.77344698e-01 1.33500591e-01 -9.31458473e-01 3.18849832e-01
2.55672485e-01 -1.44496590e-01 1.66515142e-01 7.78743505e-01
-2.30446845e-01 1.58060193e-02 -9.86364424e-01 1.07412827e+00
-1.35249002e-02 -1.14481461e+00 1.92020029e-01 -1.47150785e-01
7.60565877e-01 1.50407264e-02 1.38700038e-01 4.96775717e-01
-1.76561996e-01 -6.19607508e-01 8.41200888e-01 5.05597830e-01
5.81524193e-01 -6.63996398e-01 5.40604174e-01 1.67083800e-01
-1.63303483e+00 -3.81801158e-01 -3.67744237e-01 1.35946736e-01
7.97527805e-02 6.74170673e-01 -4.44319904e-01 1.05120313e+00
9.08317626e-01 9.67513561e-01 -1.07703221e+00 1.04849362e+00
4.48170491e-02 2.46025652e-01 -3.71872991e-01 4.60623085e-01
2.81289637e-01 2.07145512e-01 5.48947155e-01 1.17337656e+00
3.87013286e-01 -3.17986943e-02 4.07398373e-01 8.64519298e-01
9.44755320e-03 9.59961042e-02 -2.96457410e-01 1.96354255e-01
7.18969941e-01 1.50244284e+00 -8.54550302e-01 -5.64222753e-01
-7.39851713e-01 1.01786089e+00 3.69531006e-01 3.06719452e-01
-8.79697442e-01 -2.58935541e-01 7.84093380e-01 4.20076922e-02
7.18650520e-01 1.92778498e-01 -4.57346320e-01 -1.37252545e+00
3.59155387e-01 -7.20466137e-01 5.80286086e-01 -7.11147368e-01
-1.46503985e+00 4.08774137e-01 2.90284157e-02 -1.45424855e+00
2.69196838e-01 -3.66516471e-01 -4.24238354e-01 6.29415512e-01
-1.77285874e+00 -1.62715614e+00 -5.46360493e-01 6.96299076e-01
7.50156224e-01 -1.66182995e-01 4.61482555e-01 3.88655841e-01
-5.05193233e-01 7.50742137e-01 -2.47478381e-01 -1.89637214e-01
8.63796830e-01 -9.38614190e-01 4.56498861e-01 9.51280355e-01
2.21011415e-01 8.06659937e-01 5.05160749e-01 -6.33223534e-01
-1.48777473e+00 -1.36429739e+00 4.97166693e-01 -5.45056701e-01
7.20918715e-01 -2.54320562e-01 -1.28275359e+00 4.05775309e-01
-1.45633757e-01 4.48400855e-01 4.21633631e-01 7.60982260e-02
-8.13087046e-01 -3.25267762e-01 -8.18724930e-01 4.76420730e-01
1.27253342e+00 -5.37613809e-01 -7.34316468e-01 6.98223263e-02
1.15495753e+00 -2.13020921e-01 -7.36469507e-01 8.90799165e-01
6.23506606e-01 -1.23908579e+00 1.24769270e+00 -5.21988630e-01
2.71138191e-01 -7.18041539e-01 -5.44685662e-01 -8.91877770e-01
-5.55998445e-01 -3.18691432e-01 -2.71858573e-01 1.38915622e+00
1.23392120e-01 -6.38811588e-01 4.62881148e-01 2.54069537e-01
-2.85090268e-01 -8.76655579e-01 -8.66166413e-01 -9.91841614e-01
-3.76932323e-01 -5.36851287e-01 9.37519610e-01 9.98051345e-01
-2.83719391e-01 4.14929628e-01 -2.57473707e-01 4.20745999e-01
7.01916635e-01 5.52936852e-01 7.82589674e-01 -1.14185596e+00
-2.96336800e-01 -5.73234320e-01 -4.93058234e-01 -1.01025844e+00
-2.05297783e-01 -8.66037548e-01 7.38060996e-02 -1.34282219e+00
6.54179156e-01 -3.16926837e-01 -8.35370362e-01 5.95999002e-01
-7.64791906e-01 4.97408003e-01 4.24333841e-01 2.83675343e-01
-1.02560055e+00 8.08348358e-01 1.17601585e+00 -1.14117965e-01
-6.00544214e-02 -3.59488606e-01 -1.22380435e+00 8.40101838e-01
4.86352772e-01 -3.86960626e-01 -4.31557477e-01 -3.93581718e-01
1.41125191e-02 -4.03433561e-01 7.54197717e-01 -1.09373128e+00
2.74289042e-01 -6.27293214e-02 4.76678073e-01 -7.59766519e-01
2.94643253e-01 -7.78733552e-01 -7.18111023e-02 2.44187996e-01
-1.95048168e-01 -1.71482548e-01 1.48298591e-01 8.27225268e-01
-3.32541764e-01 2.84488469e-01 6.35119379e-01 1.24450073e-01
-1.08167124e+00 4.58228052e-01 5.15474677e-01 -6.77123219e-02
1.05823123e+00 -3.37284684e-01 -4.06747669e-01 1.27260610e-01
-4.46257472e-01 4.10231531e-01 5.88248432e-01 7.77919888e-01
6.29806936e-01 -1.42115605e+00 -5.57888329e-01 3.76489729e-01
6.66416287e-01 1.26270279e-01 5.39188862e-01 1.11753225e+00
2.77954251e-01 3.18048239e-01 -5.35305999e-02 -8.77831280e-01
-1.10427880e+00 8.45662236e-01 -4.52865027e-02 -7.59753883e-02
-7.65113056e-01 1.07916558e+00 8.99386704e-01 -2.15712339e-01
1.43735200e-01 -3.09117526e-01 -5.41133359e-02 2.46841997e-01
7.90331662e-01 2.45065808e-01 1.59638464e-01 -6.59678042e-01
-6.83689594e-01 6.83312118e-01 -3.73472571e-01 2.43500054e-01
1.21724796e+00 -3.99251640e-01 1.19443297e-01 3.52465779e-01
1.14449084e+00 -1.09511934e-01 -1.48623741e+00 -8.10711205e-01
1.13930590e-01 -7.74480820e-01 7.29907304e-02 -7.09973633e-01
-1.08090508e+00 6.78452373e-01 6.86653435e-01 1.96618447e-03
1.46995306e+00 2.27015510e-01 8.54986608e-01 3.58761281e-01
3.76346201e-01 -1.02506983e+00 3.12236667e-01 2.63580769e-01
8.00618052e-01 -1.53704131e+00 -2.35514548e-02 -7.15143383e-01
-6.45871639e-01 8.32683086e-01 7.66089499e-01 -2.48478130e-02
5.36504269e-01 1.37578443e-01 -2.20309377e-01 -2.84647346e-01
-8.49955261e-01 -6.11774206e-01 5.52536845e-01 4.28132772e-01
1.24026969e-01 -6.29242212e-02 -2.75993217e-02 9.53655005e-01
3.99211794e-02 -3.23053114e-02 -1.04772851e-01 1.08056104e+00
-7.56767929e-01 -6.94287777e-01 -3.30939710e-01 5.80349028e-01
-2.17708722e-01 -2.14153871e-01 -2.07782954e-01 6.97639227e-01
3.07505071e-01 8.11666667e-01 2.33831704e-02 -5.25515497e-01
5.66200972e-01 -5.01951762e-02 4.03992146e-01 -4.97786433e-01
-4.33994085e-01 2.44975150e-01 -3.26478094e-01 -8.84356856e-01
-5.34209728e-01 -6.76890790e-01 -1.26246226e+00 1.28525048e-01
-5.09029090e-01 -2.74087518e-01 3.84386510e-01 1.08671343e+00
6.83434367e-01 6.95028067e-01 5.87641537e-01 -7.36483634e-01
-6.80966318e-01 -8.48338783e-01 -4.46424007e-01 4.40031976e-01
3.36247176e-01 -1.03184438e+00 -5.31543866e-02 -8.10279138e-03] | [9.692594528198242, 0.19946053624153137] |
19ea3a7a-0583-44c8-a3f4-7c846bc93157 | meshrir-a-dataset-of-room-impulse-responses | 2106.10801 | null | https://arxiv.org/abs/2106.10801v2 | https://arxiv.org/pdf/2106.10801v2.pdf | MeshRIR: A Dataset of Room Impulse Responses on Meshed Grid Points For Evaluating Sound Field Analysis and Synthesis Methods | A new impulse response (IR) dataset called "MeshRIR" is introduced. Currently available datasets usually include IRs at an array of microphones from several source positions under various room conditions, which are basically designed for evaluating speech enhancement and distant speech recognition methods. On the other hand, methods of estimating or controlling spatial sound fields have been extensively investigated in recent years; however, the current IR datasets are not applicable to validating and comparing these methods because of the low spatial resolution of measurement points. MeshRIR consists of IRs measured at positions obtained by finely discretizing a spatial region. Two subdatasets are currently available: one consists of IRs in a three-dimensional cuboidal region from a single source, and the other consists of IRs in a two-dimensional square region from an array of 32 sources. Therefore, MeshRIR is suitable for evaluating sound field analysis and synthesis methods. This dataset is freely available at https://sh01k.github.io/MeshRIR/ with some codes of sample applications. | ['Jesper Brunnström', 'Natsuki Ueno', 'Takumi Abe', 'Keisuke Kimura', 'Tomoya Nishida', 'Shoichi Koyama'] | 2021-06-21 | null | null | null | null | ['room-impulse-response', 'distant-speech-recognition'] | ['audio', 'speech'] | [-6.19929889e-03 -4.79422480e-01 3.76948714e-01 -7.11227655e-02
-1.41722691e+00 -5.20067453e-01 4.30068016e-01 -5.36089838e-02
-2.56780952e-01 6.15314603e-01 4.23422098e-01 -1.86930448e-01
-3.45281750e-01 -7.67596364e-01 -4.52001214e-01 -1.01318359e+00
-1.27630904e-01 1.11172989e-01 1.18461370e-01 -4.96021397e-02
1.23486884e-01 5.46642482e-01 -1.96734536e+00 2.54691362e-01
5.38151026e-01 1.26195955e+00 4.52191174e-01 6.67402565e-01
1.68242961e-01 5.17763913e-01 -8.01879466e-01 4.04454261e-01
9.10887215e-03 -3.62696797e-01 -2.97700554e-01 -3.04913491e-01
1.32840991e-01 -1.19618058e-01 -2.79690146e-01 9.04868364e-01
1.10658109e+00 5.47308564e-01 4.99725789e-01 -7.39040136e-01
-2.35342667e-01 6.17809296e-01 -2.48066425e-01 3.68403286e-01
6.19823277e-01 -2.87791669e-01 7.54748762e-01 -1.14701688e+00
6.70568645e-02 8.65366697e-01 5.12298822e-01 3.07495445e-01
-9.06819582e-01 -7.07081139e-01 -3.54828238e-01 1.43194497e-01
-1.88658857e+00 -8.19171071e-01 1.05512071e+00 -4.39127833e-01
6.80388510e-01 7.71217644e-01 2.88244426e-01 1.11017764e+00
-4.10184175e-01 2.17214763e-01 1.16670346e+00 -5.12560427e-01
5.96406639e-01 1.50627702e-01 -9.30228382e-02 -2.64427923e-02
-3.00386906e-01 2.71507740e-01 -7.04060614e-01 -4.36145008e-01
9.19845819e-01 -4.77706075e-01 -6.80102468e-01 1.33941993e-01
-1.28775656e+00 5.08346319e-01 1.50772080e-01 7.41400719e-01
-2.64748931e-01 -3.19213301e-01 -4.46063392e-02 1.92277089e-01
7.62252748e-01 3.37528795e-01 -1.91842407e-01 -2.87829876e-01
-8.46136272e-01 1.73643291e-01 5.91216087e-01 7.54316747e-01
6.24873281e-01 3.36657345e-01 1.10294461e-01 1.61393344e+00
3.45356852e-01 1.01455128e+00 3.71970147e-01 -9.99457359e-01
5.52711546e-01 -1.58279493e-01 1.97357342e-01 -8.97229970e-01
-3.45343322e-01 -3.82963717e-01 -9.69820201e-01 2.23084772e-03
4.62124318e-01 -2.74601728e-01 -4.05891478e-01 1.17439628e+00
3.76396805e-01 4.63232458e-01 -9.58868023e-03 1.05181479e+00
1.25732803e+00 9.65081692e-01 -5.64416707e-01 -2.49146670e-01
1.25495470e+00 -4.70080614e-01 -6.50617123e-01 6.41900748e-02
3.08110446e-01 -8.73116434e-01 8.73920321e-01 7.68922627e-01
-1.20514870e+00 -6.35422707e-01 -6.17130458e-01 2.58169830e-01
-3.49345565e-01 2.03384772e-01 7.91942980e-03 6.23495102e-01
-1.12853944e+00 7.04873130e-02 -3.83584112e-01 2.32221037e-01
-1.55352920e-01 -8.87594149e-02 -2.95216173e-01 4.36408594e-02
-1.41745996e+00 3.23662430e-01 -5.67957997e-01 5.15494466e-01
-8.51132751e-01 -9.36920762e-01 -8.46396506e-01 -1.64564013e-01
-1.39499009e-02 1.35117337e-01 1.31007493e+00 -2.63744056e-01
-1.72272718e+00 4.65238094e-01 -3.10506999e-01 9.07241330e-02
2.31251210e-01 7.87749514e-02 -9.70885217e-01 2.30917901e-01
1.56951115e-01 -1.06738647e-03 8.70453179e-01 -1.24988604e+00
-2.29698941e-01 -3.52512658e-01 -1.08816355e-01 1.84006989e-01
-3.16946268e-01 3.37204486e-01 -2.72399843e-01 -7.82528818e-01
1.91617131e-01 -5.93275249e-01 -2.19169617e-01 -3.39900523e-01
-5.37242353e-01 1.73892304e-02 4.27039444e-01 -7.61519909e-01
1.37175202e+00 -2.30520177e+00 -9.95026380e-02 3.67701083e-01
-1.11128673e-01 2.77760357e-01 -1.45466685e-01 5.87425113e-01
-2.79629618e-01 1.45236552e-01 -1.86628833e-01 -2.12519839e-01
-2.66983390e-01 -4.65295583e-01 -4.48046178e-01 7.56355226e-01
-2.88461745e-01 1.04566388e-01 -6.71429753e-01 -2.06331030e-01
6.09532237e-01 7.69119680e-01 -4.67765570e-01 2.53733426e-01
1.36420116e-01 1.02389920e+00 -6.34633720e-01 5.37751496e-01
9.52115774e-01 1.84953049e-01 -3.90833497e-01 -8.96351710e-02
-7.25562572e-01 5.06490350e-01 -1.63292694e+00 1.49183214e+00
-9.45765793e-01 6.53326213e-01 4.94387507e-01 -8.82933617e-01
1.26539564e+00 7.75571764e-01 5.31812608e-01 -8.73077810e-01
2.87257158e-03 3.92584085e-01 -2.07392901e-01 -4.98520702e-01
2.91886121e-01 9.06409323e-02 7.70450458e-02 2.93145776e-01
-3.97448212e-01 -3.92952830e-01 -1.13310911e-01 -3.61388028e-01
1.18778729e+00 -4.95050371e-01 8.55100751e-02 -2.25204319e-01
8.38594854e-01 -4.13435519e-01 4.26840425e-01 7.41342843e-01
3.33603285e-02 1.13228846e+00 -6.15403652e-02 -6.81613088e-02
-7.87484467e-01 -1.28802407e+00 -8.25291038e-01 6.72652900e-01
-3.84003483e-02 -1.51361108e-01 -6.39244139e-01 2.94966191e-01
-3.65859240e-01 7.32216775e-01 -3.20527196e-01 4.26554829e-01
-5.77377677e-01 -4.09551144e-01 7.40099847e-01 3.04942727e-01
5.16976595e-01 -1.09069371e+00 -3.02164823e-01 4.29092608e-02
-6.25478745e-01 -1.06486726e+00 -2.02958435e-01 -2.03500707e-02
-3.63781273e-01 -7.06966758e-01 -1.10267520e+00 -3.67603183e-01
4.11541969e-01 5.14655471e-01 9.29105520e-01 -1.17057621e-01
-2.22690895e-01 5.34060955e-01 -6.42306089e-01 -3.32432359e-01
-3.21321756e-01 -3.07234704e-01 2.42342830e-01 2.02326000e-01
-2.25034654e-01 -6.72555208e-01 -5.41682363e-01 7.99378395e-01
-7.29097903e-01 -3.85527998e-01 -3.36741284e-02 2.81413198e-01
5.88975608e-01 3.12942535e-01 6.52305543e-01 -8.05654675e-02
3.99596184e-01 -6.76530898e-01 -7.51213968e-01 -1.53553873e-01
2.71542966e-01 -3.39401186e-01 8.01208556e-01 -3.38171005e-01
-1.33907926e+00 -4.36603963e-01 -7.66217172e-01 -3.33247691e-01
-6.42560601e-01 2.33698621e-01 -2.28039041e-01 3.90702374e-02
7.22850204e-01 2.19049498e-01 -2.66923130e-01 -8.84102643e-01
2.09224764e-02 1.39156210e+00 4.05707985e-01 -6.34015262e-01
6.29677892e-01 4.11080927e-01 -2.52539843e-01 -1.55782926e+00
-3.85325491e-01 -8.00598741e-01 -2.64958471e-01 -3.75948757e-01
5.94967544e-01 -9.58986402e-01 -2.39052638e-01 7.43274510e-01
-9.65733230e-01 -4.49031919e-01 -2.90078253e-01 8.90165925e-01
-3.83832008e-01 8.92814100e-02 -3.70024621e-01 -1.03923225e+00
-1.29677191e-01 -1.11185479e+00 1.33563459e+00 -8.44164751e-03
-1.42203212e-01 -1.03683758e+00 3.95112962e-01 4.48853582e-01
4.65220660e-01 -2.19577849e-01 4.02960986e-01 -1.17375314e-01
-2.11152598e-01 -1.11551151e-01 2.90339261e-01 4.67109829e-01
2.67951161e-01 -1.02479003e-01 -1.39426267e+00 -1.88721433e-01
2.34651953e-01 -5.92256114e-02 5.06928146e-01 8.27362776e-01
1.38960230e+00 -6.89734370e-02 -1.17234580e-01 4.57263172e-01
1.19319463e+00 4.62065935e-01 8.90754998e-01 9.73259807e-02
2.21934348e-01 6.07457876e-01 5.50674319e-01 7.14174390e-01
-1.15550980e-01 1.09251142e+00 2.36178651e-01 -1.55686438e-01
-1.67089313e-01 3.72868180e-02 2.79379576e-01 9.78911042e-01
-1.67792186e-01 -4.08467084e-01 -1.02958560e+00 5.91625571e-01
-1.03666937e+00 -1.04408360e+00 -3.10445666e-01 2.46137691e+00
5.82256615e-01 -6.17070913e-01 1.34166386e-02 6.02704823e-01
9.64730918e-01 4.03722703e-01 -2.24760339e-01 -2.35646546e-01
-1.04056448e-01 3.11711967e-01 1.81525826e-01 7.26502180e-01
-8.20829868e-01 1.29551157e-01 5.89853477e+00 9.79217827e-01
-1.21805024e+00 5.60451224e-02 3.56194049e-01 1.49963312e-02
-4.60675925e-01 -4.43954319e-01 -8.48130882e-01 4.96169657e-01
1.06029105e+00 1.48872823e-01 6.10646188e-01 4.84723061e-01
6.76122010e-01 -2.67774969e-01 -5.79286754e-01 1.11570370e+00
-1.21790104e-01 -1.08274674e+00 -4.92146075e-01 -4.58529852e-02
3.51300687e-01 2.30274901e-01 3.61445993e-01 -1.65077388e-01
-1.96271967e-02 -8.99285793e-01 8.04263711e-01 4.95591849e-01
1.19264877e+00 -5.51323175e-01 3.08578998e-01 5.79125285e-01
-1.32667625e+00 -1.50157418e-02 -4.27272707e-01 -4.32467274e-02
2.27602661e-01 9.67659295e-01 -5.33148229e-01 6.10682607e-01
1.13768959e+00 4.58031356e-01 -2.44896367e-01 1.13689148e+00
-6.78304136e-02 9.56891119e-01 -5.38734376e-01 1.10999249e-01
-1.12402856e-01 -3.12354028e-01 8.66074681e-01 9.85204458e-01
9.61400628e-01 1.79938942e-01 -3.94191504e-01 8.30313861e-01
2.21022218e-01 1.18918359e-01 -7.91395724e-01 4.54214960e-01
9.36783791e-01 1.28476286e+00 -3.49485606e-01 1.26458079e-01
-4.07323718e-01 2.17749417e-01 -3.00895631e-01 5.75461149e-01
-8.93080950e-01 -5.55420816e-01 7.83899844e-01 3.98320228e-01
3.06171149e-01 -1.56497508e-01 3.33341025e-02 -9.68560100e-01
6.93691522e-02 -8.63317192e-01 -7.68388435e-02 -1.17432141e+00
-9.89926040e-01 7.38140821e-01 2.26335466e-01 -1.58738184e+00
-3.07454616e-01 -4.44685489e-01 -6.52421951e-01 1.04381430e+00
-1.10627651e+00 -4.00052339e-01 -3.74844223e-01 6.77300632e-01
4.11763638e-01 -1.00154869e-01 1.06509280e+00 6.76820874e-01
-5.93868315e-01 4.27197069e-01 5.44814110e-01 1.22519799e-01
4.11486000e-01 -8.57337654e-01 2.90336728e-01 6.30204022e-01
1.15864396e-01 4.07003641e-01 6.36653364e-01 -2.00089179e-02
-1.19532692e+00 -1.03803754e+00 6.18738055e-01 -1.62687689e-01
4.91044670e-01 -5.52922249e-01 -9.24305797e-01 2.40438491e-01
-2.59547550e-02 4.05417919e-01 5.47155201e-01 -7.03762844e-02
-2.86408991e-01 -3.80434722e-01 -1.11280239e+00 3.85445595e-01
8.39758217e-01 -6.51546955e-01 -1.44553915e-01 2.47297019e-01
4.04271156e-01 -4.35400665e-01 -1.18715000e+00 4.30596590e-01
1.92847550e-01 -1.18655980e+00 1.27564287e+00 4.04419750e-01
9.95654687e-02 -3.69715542e-01 -5.10875165e-01 -1.49068069e+00
-4.12367545e-02 -4.71743912e-01 3.70598406e-01 1.47625124e+00
3.40507984e-01 -1.03817511e+00 3.70288283e-01 9.73704550e-03
-2.07369640e-01 -4.64206278e-01 -1.29135525e+00 -9.06126797e-01
2.00797692e-01 -9.52235162e-01 7.16369271e-01 5.53644598e-01
-4.24707346e-02 2.54172981e-02 -1.38622358e-01 6.53725803e-01
6.40059054e-01 3.68971610e-03 4.69455630e-01 -9.72272813e-01
-2.20878631e-01 -2.91285157e-01 -1.30638868e-01 -1.17439795e+00
7.46461079e-02 -5.47099769e-01 2.89115846e-01 -1.37987864e+00
-4.10727590e-01 -1.01472044e+00 -5.73002696e-02 5.84976608e-03
3.91516149e-01 3.57634276e-01 -2.16961652e-01 1.79440275e-01
-1.51776457e-02 3.38311285e-01 1.09070849e+00 1.77392941e-02
-1.70599595e-01 5.00719786e-01 -3.60641181e-01 7.77681172e-01
1.05345035e+00 -2.13050768e-01 -4.70350295e-01 -5.24640322e-01
3.64381932e-02 5.63679993e-01 5.61447620e-01 -1.35501575e+00
2.30899736e-01 1.09626673e-01 -2.95048747e-02 -7.40177631e-01
7.24330783e-01 -6.48000419e-01 5.43081760e-01 -1.61752090e-01
-3.16254556e-01 -2.96847939e-01 1.15699247e-01 2.47196600e-01
-6.11696839e-01 -2.17385709e-01 8.91140759e-01 1.92895010e-02
-4.80207622e-01 7.02432469e-02 -5.60141444e-01 2.29606092e-01
6.09288096e-01 9.01728775e-03 -1.79014385e-01 -6.80072784e-01
-5.00378489e-01 -4.08582509e-01 3.51194888e-02 2.76887566e-01
7.43107319e-01 -1.33201563e+00 -7.54259884e-01 4.67765570e-01
2.07303073e-02 2.10168093e-01 5.20555019e-01 6.66571259e-01
-3.45131040e-01 5.56192219e-01 2.51162320e-01 -6.84268117e-01
-1.28512454e+00 8.56985673e-02 5.84577978e-01 1.91410229e-01
-5.56983769e-01 9.64402735e-01 3.80454659e-01 -5.98306835e-01
3.44439715e-01 -3.78080308e-01 -5.70526481e-01 -2.26733908e-02
9.64687467e-01 7.75453389e-01 4.38245356e-01 -1.00392807e+00
-3.86668086e-01 9.07898188e-01 6.37449086e-01 -5.12685895e-01
1.18253493e+00 -3.56661171e-01 -1.71548262e-01 6.60859585e-01
1.40855622e+00 5.15485585e-01 -7.46315598e-01 -1.85757279e-01
-5.20686924e-01 -4.47938353e-01 2.14662880e-01 -5.43847084e-01
-9.49054003e-01 1.03897822e+00 4.61712807e-01 3.75375539e-01
1.42043304e+00 1.69110447e-01 4.23800677e-01 9.97044072e-02
7.55810976e-01 -7.30253398e-01 -9.16217044e-02 4.93987650e-01
1.27445722e+00 -7.06171632e-01 -5.02203703e-01 -4.26633060e-01
-3.54001135e-01 8.62356782e-01 1.04464866e-01 3.05627137e-01
9.97503221e-01 7.28312314e-01 3.76264244e-01 2.27667972e-01
-4.14463580e-01 1.56257171e-02 1.42846256e-01 7.55918801e-01
5.73709667e-01 1.91456191e-02 1.80731490e-01 5.11113226e-01
-3.98379743e-01 -3.53335202e-01 3.24168772e-01 3.48411888e-01
-3.54330927e-01 -9.67968404e-01 -1.01320791e+00 2.40079954e-01
-4.86684144e-01 -1.26719177e-01 2.65249580e-01 3.73567671e-01
-8.34188052e-03 1.68347192e+00 9.43279564e-02 -3.60877246e-01
5.14083982e-01 -4.41177398e-01 8.93573686e-02 -4.52961177e-01
-3.14459443e-01 1.42095551e-01 2.59333163e-01 -4.73529607e-01
-2.25016087e-01 -8.91342402e-01 -1.15961802e+00 -1.83511540e-01
-3.49202454e-01 5.91042399e-01 7.64846325e-01 5.10423064e-01
1.46897689e-01 4.98445928e-01 1.11320508e+00 -1.04263473e+00
-4.87247109e-01 -1.07298243e+00 -1.19214606e+00 -1.15950257e-01
5.11010587e-01 -5.91675162e-01 -8.91541243e-01 -2.11013466e-01] | [15.122405052185059, 5.791810512542725] |
13f82878-570c-412a-95ea-fcc7b6af7a9a | bayesian-inverse-contextual-reasoning-for | 2306.06403 | null | https://arxiv.org/abs/2306.06403v1 | https://arxiv.org/pdf/2306.06403v1.pdf | Bayesian Inverse Contextual Reasoning for Heterogeneous Semantics-Native Communication | This work deals with the heterogeneous semantic-native communication (SNC) problem. When agents do not share the same communication context, the effectiveness of contextual reasoning (CR) is compromised calling for agents to infer other agents' context. This article proposes a novel framework for solving the inverse problem of CR in SNC using two Bayesian inference methods, namely: Bayesian inverse CR (iCR) and Bayesian inverse linearized CR (iLCR). The first proposed Bayesian iCR method utilizes Markov Chain Monte Carlo (MCMC) sampling to infer the agent's context while being computationally expensive. To address this issue, a Bayesian iLCR method is leveraged which obtains a linearized CR (LCR) model by training a linear neural network. Experimental results show that the Bayesian iLCR method requires less computation and achieves higher inference accuracy compared to Bayesian iCR. Additionally, heterogeneous SNC based on the context obtained through the Bayesian iLCR method shows better communication effectiveness than that of Bayesian iCR. Overall, this work provides valuable insights and methods to improve the effectiveness of SNC in situations where agents have different contexts. | ['Wan Choi', 'Mehdi Bennis', 'Yoonseong Kang', 'Hyowoon Seo'] | 2023-06-10 | null | null | null | null | ['bayesian-inference'] | ['methodology'] | [ 2.43400425e-01 2.79910207e-01 -3.50806504e-01 -3.20851296e-01
-8.44348252e-01 -4.29502070e-01 1.10021555e+00 -2.08799727e-02
-4.91601408e-01 9.20581698e-01 4.31674510e-01 -5.41529655e-01
-1.13571629e-01 -9.76799846e-01 -4.89789754e-01 -5.83751619e-01
1.40044034e-01 6.80704057e-01 3.84874523e-01 1.27877071e-01
2.23226190e-01 1.28220722e-01 -1.13579583e+00 1.85100526e-01
6.28206968e-01 9.20865834e-01 2.25444764e-01 6.33378327e-01
-5.55952638e-02 1.46396840e+00 -9.94008064e-01 -4.20175016e-01
-2.69482974e-02 -3.22353661e-01 -8.93164575e-01 -7.00978458e-01
-3.88362169e-01 -6.54331565e-01 -8.65746662e-02 1.13662755e+00
1.94802046e-01 2.55886108e-01 6.61777377e-01 -1.52450752e+00
-3.92834067e-01 1.42233372e+00 -3.13405037e-01 3.50406393e-02
1.07811213e+00 -1.92034394e-01 8.10507119e-01 -2.76351929e-01
5.13648152e-01 1.93433845e+00 6.20666683e-01 4.64001864e-01
-1.21509504e+00 -9.43926632e-01 5.69214225e-01 4.10578102e-01
-1.52017117e+00 -1.49857432e-01 6.72185659e-01 -2.22154319e-01
8.06912541e-01 1.58101276e-01 6.21818960e-01 1.56826687e+00
8.41776803e-02 7.58331716e-01 1.62038839e+00 -4.33655411e-01
9.78965819e-01 2.95819193e-01 -7.75404228e-03 2.72434920e-01
2.28138864e-01 7.03310013e-01 -8.37871134e-01 -6.85754538e-01
4.06104356e-01 -1.64965257e-01 -9.74367373e-03 3.57801467e-01
-1.01411450e+00 1.02775908e+00 5.61110005e-02 3.35209891e-02
-3.10753286e-01 6.22499168e-01 4.99452390e-02 8.83224234e-02
5.57823516e-02 -9.77177322e-02 -2.38696486e-01 -3.35366398e-01
-3.54095399e-01 2.21502006e-01 1.43057489e+00 1.26460481e+00
5.42130470e-01 -1.57672092e-01 3.11637055e-02 5.89893997e-01
1.08867323e+00 9.26270366e-01 -6.29121065e-02 -1.48813796e+00
2.84557521e-01 1.48579359e-01 1.57710254e-01 -1.13232172e+00
-2.41542473e-01 -2.02621982e-01 -4.42612737e-01 -8.23241770e-02
4.13257211e-01 -2.09659398e-01 -1.86588332e-01 2.08299065e+00
4.63816524e-01 5.57332754e-01 5.70520997e-01 7.26027727e-01
7.07099199e-01 7.26140320e-01 3.37136209e-01 -1.39125645e-01
1.31856525e+00 -5.95708668e-01 -1.04703856e+00 -1.74395517e-01
2.58329064e-01 -7.20149755e-01 5.46625912e-01 2.81213254e-01
-6.51707113e-01 1.92399442e-01 -1.21145964e+00 3.36676806e-01
-4.30496752e-01 -4.46593374e-01 7.45855093e-01 9.35217917e-01
-9.51848626e-01 -2.42571294e-01 -7.73131907e-01 -2.86393613e-01
8.15410763e-02 -3.97878187e-03 2.44413361e-01 -1.83749944e-01
-1.51696408e+00 1.00082386e+00 3.65698040e-01 -2.82664578e-02
-1.46672750e+00 -5.11219442e-01 -7.76736438e-01 -1.86427832e-01
1.04321527e+00 -5.78861594e-01 1.47731125e+00 -6.21481419e-01
-2.36754489e+00 -2.82707866e-02 -2.83761770e-01 -4.87924457e-01
5.41232049e-01 -5.01503889e-03 -4.10592884e-01 3.86824310e-01
3.08525234e-01 3.36091250e-01 7.46674001e-01 -1.91654551e+00
-7.73250937e-01 -3.17638785e-01 5.90300143e-01 7.72972256e-02
6.18454516e-01 -1.11994511e-02 -4.43053067e-01 -4.22418207e-01
-4.79827262e-02 -1.02437139e+00 -3.32797803e-02 -4.18155581e-01
-4.31439728e-01 -4.68194366e-01 9.59753752e-01 -4.72996593e-01
9.49600816e-01 -1.67153084e+00 -7.03720003e-02 6.51769817e-01
8.35692957e-02 -6.18231781e-02 -1.76021364e-02 4.86925304e-01
7.41934001e-01 -4.44028899e-02 -1.11226380e-01 -2.91196883e-01
3.64438444e-01 4.57060456e-01 -3.60962749e-01 3.51915419e-01
-3.89624894e-01 5.67227602e-01 -9.88947809e-01 -6.43091977e-01
2.31948286e-01 7.46213496e-01 -7.44141519e-01 2.14799732e-01
-5.64358950e-01 6.60407960e-01 -6.99877381e-01 5.05898714e-01
6.01827025e-01 -3.12931359e-01 8.94960999e-01 -1.11654133e-01
1.36890277e-01 1.62639350e-01 -1.11428642e+00 1.25140095e+00
-8.97124171e-01 2.74214506e-01 2.12927222e-01 -5.83163798e-01
6.56082034e-01 3.89164716e-01 -1.26774132e-01 -6.56231940e-01
2.96909064e-01 2.16574538e-02 -3.33567709e-01 -2.63440520e-01
2.61731237e-01 -2.07429394e-01 -3.67612690e-01 9.93640482e-01
-3.59044939e-01 -9.94570255e-02 -1.49983481e-01 7.64952242e-01
8.78459513e-01 1.60220951e-01 4.08425421e-01 -5.10711893e-02
6.47592604e-01 -2.50373811e-01 7.07564116e-01 1.36023998e+00
-4.83426332e-01 -2.83318520e-01 4.70198333e-01 -1.31473571e-01
-2.92504996e-01 -1.25493813e+00 7.15418607e-02 1.07155216e+00
7.65289605e-01 -3.47142875e-01 -8.87180448e-01 -7.14997292e-01
-1.42570138e-01 1.29164755e+00 -4.68691468e-01 -1.48206830e-01
-3.33412886e-01 -4.75342423e-01 7.73619115e-01 5.21974921e-01
9.12078559e-01 -5.38436770e-01 -6.97276294e-01 9.79901999e-02
-7.04562604e-01 -1.57642496e+00 -2.46965170e-01 -4.75661427e-01
-2.56995022e-01 -1.11989093e+00 8.98901969e-02 4.70357053e-02
3.35904449e-01 2.28358537e-01 8.79149497e-01 5.54173365e-02
2.82552183e-01 1.00308216e+00 -6.48382187e-01 -2.46306017e-01
-7.62796104e-01 -1.62417263e-01 -1.87644903e-02 -5.62931001e-02
6.37807846e-01 -4.46019709e-01 -4.27977353e-01 5.85783839e-01
-7.47937918e-01 -7.67713115e-02 3.69730204e-01 4.70504820e-01
9.84083638e-02 3.47977489e-01 9.38359261e-01 -9.17312801e-01
7.91518211e-01 -5.83857179e-01 -8.72575641e-01 3.32518727e-01
-7.55894244e-01 6.09576181e-02 4.63343710e-01 -4.47451591e-01
-1.97286606e+00 -5.09737551e-01 3.76935862e-02 1.96241602e-01
-1.95523873e-01 6.00969613e-01 -8.91959146e-02 2.08743051e-01
3.46259475e-01 -1.21051393e-01 -4.85741943e-02 -7.50658335e-03
4.67161149e-01 8.26136291e-01 2.26125225e-01 -1.18583608e+00
3.31053823e-01 8.83998632e-01 -2.02753633e-01 -6.51775479e-01
-1.13923764e+00 7.15581328e-02 1.60373420e-01 -6.07033968e-01
1.05797744e+00 -9.05593812e-01 -1.51036131e+00 2.38831446e-01
-1.27284849e+00 -6.19543135e-01 3.29665065e-01 9.54042494e-01
-6.75570250e-01 3.65502328e-01 -5.81805587e-01 -1.31834352e+00
2.53287017e-01 -1.52907038e+00 8.37412655e-01 1.42591432e-01
-2.62620747e-01 -1.24358892e+00 -8.85293707e-02 8.13438356e-01
4.61136013e-01 -3.77830006e-02 1.02458739e+00 -6.24084175e-01
-1.06881297e+00 -1.52573943e-01 -3.32925290e-01 -2.97979891e-01
-1.87635407e-01 -1.98413819e-01 -9.26393807e-01 -3.04609895e-01
-1.48811555e-02 -1.95978247e-02 3.25447530e-01 3.75714242e-01
5.59246659e-01 -5.73068619e-01 -3.87699217e-01 5.85082471e-02
1.44848895e+00 7.52233267e-01 4.15360242e-01 1.56218320e-01
1.78161040e-01 4.94949818e-01 4.89746153e-01 7.32408524e-01
1.19274557e+00 7.98409998e-01 4.10822123e-01 6.38912737e-01
-1.94680184e-01 -5.86065471e-01 5.20569324e-01 5.11090100e-01
-1.32052094e-01 -2.40785792e-01 -7.95973480e-01 1.89064890e-01
-1.99545348e+00 -1.08770430e+00 2.71317482e-01 1.93917954e+00
8.11354756e-01 -1.60667207e-02 -1.79802641e-01 -2.31311724e-01
1.06987524e+00 -3.76174673e-02 -5.16049266e-01 -3.13422322e-01
1.33955941e-01 -4.39733535e-01 2.26452574e-01 1.24824810e+00
-4.74026322e-01 9.21423435e-01 6.57338715e+00 8.27311635e-01
-2.68520445e-01 6.07221305e-01 2.05732882e-01 1.68195158e-01
-5.34971535e-01 4.04601097e-01 -9.70962703e-01 7.32582211e-01
1.04230964e+00 1.69713080e-01 1.02059662e+00 5.91465533e-01
2.06789508e-01 -8.45415056e-01 -1.18256080e+00 8.07443857e-01
9.56798345e-02 -1.27214885e+00 2.87882835e-02 1.49301633e-01
7.80441284e-01 -2.00358823e-01 -1.12146311e-01 5.51349759e-01
1.43955731e+00 -6.62558734e-01 1.05672920e+00 5.89769542e-01
2.97640413e-01 -7.61007130e-01 9.94435787e-01 3.23090643e-01
-1.03801394e+00 -3.88903320e-01 3.49619798e-02 -7.42965564e-03
3.66406798e-01 5.01245677e-01 -7.05551088e-01 5.45373619e-01
7.90940404e-01 3.94543767e-01 2.27858895e-04 1.45755589e-01
-5.43069243e-01 3.69085699e-01 -3.96014780e-01 -3.00549448e-01
1.10756151e-01 -3.61764729e-01 8.07339311e-01 9.84798133e-01
-1.64358169e-01 2.28533089e-01 3.24061900e-01 1.14417660e+00
2.14477882e-01 -5.70755303e-01 -2.98690319e-01 3.35054994e-01
1.44817293e+00 6.59858227e-01 -7.14334190e-01 -5.96947789e-01
-3.35634649e-01 2.83665687e-01 2.67726392e-01 7.67095745e-01
-9.75865841e-01 7.48932064e-02 1.87677190e-01 -6.56233251e-01
3.44882384e-02 -5.27759176e-03 -1.35090411e-01 -8.38351667e-01
-4.21968341e-01 -8.25588763e-01 5.63564420e-01 -6.71041429e-01
-1.32149303e+00 3.65168244e-01 6.37644649e-01 -5.72933495e-01
-5.21808207e-01 -2.48220846e-01 -3.27256829e-01 3.46809626e-01
-1.68829620e+00 -1.20427561e+00 -6.91241249e-02 7.10760415e-01
3.17035586e-01 -2.77784765e-01 7.46388316e-01 -3.97800207e-02
-6.68610573e-01 3.33565950e-01 -1.76272824e-01 -2.97216505e-01
4.00739163e-01 -8.61915588e-01 -6.85958028e-01 7.16985345e-01
-5.49255967e-01 9.06814456e-01 8.11038852e-01 -8.68222415e-01
-1.57473910e+00 -9.39390302e-01 5.07897913e-01 -2.67916530e-01
5.47864139e-01 2.71065980e-02 -1.00924022e-01 1.01450872e+00
1.89966097e-01 -5.76415241e-01 1.01765478e+00 1.46281272e-01
-8.41261744e-01 -7.66619071e-02 -1.59260738e+00 8.75368714e-01
7.13035405e-01 -9.77478504e-01 -7.68010020e-01 -8.30086097e-02
9.51906025e-01 1.53962895e-01 -1.00313365e+00 1.53203785e-01
7.66905904e-01 -1.09990168e+00 9.03605700e-01 1.24008678e-01
-1.51809107e-03 -6.50186360e-01 -9.27972734e-01 -1.05417800e+00
3.60550135e-01 -5.79413056e-01 -3.60056520e-01 1.08955181e+00
5.07818222e-01 -1.20798504e+00 2.65107960e-01 8.18908811e-01
3.16872567e-01 -6.18820004e-02 -1.01472998e+00 -7.69107699e-01
-1.85141817e-01 -9.56127703e-01 9.68074203e-01 8.00087988e-01
1.44897595e-01 1.55980498e-01 5.47213592e-02 7.35569119e-01
1.19380832e+00 1.18085794e-01 6.38237894e-01 -1.06349623e+00
-3.25003326e-01 -1.84201285e-01 7.47487023e-02 -1.01134026e+00
6.91880047e-01 -6.60795689e-01 1.10406294e-01 -1.37693274e+00
3.95324230e-01 -5.28992534e-01 2.90033996e-01 8.23650509e-02
1.94000825e-01 -1.28097177e-01 2.64387459e-01 1.89695165e-01
-9.12997425e-01 6.05548024e-01 8.36514652e-01 -8.25542435e-02
4.00306508e-02 3.28713655e-02 -5.78395426e-01 9.91241574e-01
7.27990627e-01 -4.42060947e-01 -5.65069199e-01 -2.44755417e-01
5.59670031e-01 4.22141850e-01 7.31182456e-01 -5.73953032e-01
7.29237199e-01 -2.84729391e-01 -2.12231994e-01 -8.64026129e-01
6.03828609e-01 -1.13158882e+00 4.27962005e-01 5.01513362e-01
-6.27935886e-01 -4.77755159e-01 -3.41878325e-01 1.21930385e+00
6.54544309e-02 -3.93697619e-01 7.61509001e-01 -3.31250519e-01
-4.28160042e-01 -4.87013191e-01 -1.16012120e+00 3.88734676e-02
9.52549040e-01 1.85241234e-02 -6.24129415e-01 -8.81044149e-01
-7.16057003e-01 2.48687163e-01 3.79441306e-02 1.32398933e-01
8.30952108e-01 -1.23965383e+00 -3.39144319e-01 -8.66333023e-02
-4.82482724e-02 -1.44234762e-01 1.86612442e-01 8.81142139e-01
-2.14523971e-01 5.72707772e-01 5.71911693e-01 -6.22065008e-01
-6.39589489e-01 5.54618239e-01 2.55351543e-01 -2.34032661e-01
-3.12799484e-01 4.81759638e-01 -1.62278451e-02 -8.12322915e-01
5.99209130e-01 -6.05703779e-02 -8.49218369e-02 -1.84783727e-01
6.75550878e-01 9.47090745e-01 -5.38743794e-01 -5.86136162e-01
-3.04362625e-01 3.43182474e-01 -4.42612283e-02 -6.85692012e-01
8.19975555e-01 -7.89484441e-01 -2.56638616e-01 2.97365367e-01
7.90153682e-01 -1.25097170e-01 -1.38530445e+00 -6.47573531e-01
1.20507039e-01 -6.05917454e-01 4.10207778e-01 -1.21311498e+00
-8.95488322e-01 1.47272155e-01 3.36460710e-01 3.25494498e-01
6.35320961e-01 2.11132422e-01 3.73910576e-01 5.07692814e-01
1.09722805e+00 -1.42996204e+00 1.57230169e-01 5.23030102e-01
6.34607792e-01 -1.06024110e+00 2.52572987e-02 -5.39488733e-01
-6.99561954e-01 7.05769658e-01 3.34538430e-01 2.77797788e-01
1.03910589e+00 4.95902538e-01 -1.34300301e-02 -2.04449460e-01
-9.75127637e-01 -1.69948578e-01 -7.01313674e-01 1.04835927e+00
-2.46543363e-01 4.06677514e-01 -5.60407303e-02 5.56331933e-01
-2.37626836e-01 -1.61672235e-01 7.62167513e-01 9.15733397e-01
-8.19237158e-02 -7.71129906e-01 -5.24322927e-01 -1.55914351e-01
-4.01933461e-01 5.62710129e-02 -3.22662771e-01 5.30685306e-01
-2.65858501e-01 2.13094854e+00 -6.26952797e-02 -1.57043934e-01
-1.43326506e-01 -3.03069830e-01 2.83357769e-01 -1.41155690e-01
-1.77291319e-01 1.05726443e-01 4.35423911e-01 -8.69384646e-01
-1.04226863e+00 -5.21899641e-01 -1.27627039e+00 -5.52733421e-01
-4.56505179e-01 1.82963669e-01 8.06376874e-01 1.23642778e+00
1.12389743e-01 3.88586968e-01 6.17312729e-01 -6.44277155e-01
-5.13079524e-01 -6.62247002e-01 -3.37785333e-01 -3.09470408e-02
1.79888293e-01 -1.04892468e+00 -6.57593608e-01 -2.46114329e-01] | [4.4556379318237305, 2.4382593631744385] |
2125d999-501f-4224-ab3d-1a38267062fb | graph-attention-network-for-camera | 2209.15056 | null | https://arxiv.org/abs/2209.15056v1 | https://arxiv.org/pdf/2209.15056v1.pdf | Graph Attention Network for Camera Relocalization on Dynamic Scenes | We devise a graph attention network-based approach for learning a scene triangle mesh representation in order to estimate an image camera position in a dynamic environment. Previous approaches built a scene-dependent model that explicitly or implicitly embeds the structure of the scene. They use convolution neural networks or decision trees to establish 2D/3D-3D correspondences. Such a mapping overfits the target scene and does not generalize well to dynamic changes in the environment. Our work introduces a novel approach to solve the camera relocalization problem by using the available triangle mesh. Our 3D-3D matching framework consists of three blocks: (1) a graph neural network to compute the embedding of mesh vertices, (2) a convolution neural network to compute the embedding of grid cells defined on the RGB-D image, and (3) a neural network model to establish the correspondence between the two embeddings. These three components are trained end-to-end. To predict the final pose, we run the RANSAC algorithm to generate camera pose hypotheses, and we refine the prediction using the point-cloud representation. Our approach significantly improves the camera pose accuracy of the state-of-the-art method from $0.358$ to $0.506$ on the RIO10 benchmark for dynamic indoor camera relocalization. | ['Riadh Ksantini', 'Mohamed Bouguessa', 'Mohamed Amine Ouali'] | 2022-09-29 | null | null | null | null | ['camera-relocalization'] | ['computer-vision'] | [ 1.35834485e-01 -1.53274253e-01 3.08642406e-02 -5.08653462e-01
-5.09891629e-01 -5.83638072e-01 4.15053189e-01 -4.37049307e-02
-4.01517153e-01 -5.94082512e-02 -1.08387165e-01 -2.80273050e-01
2.24313308e-02 -1.01413596e+00 -1.22057819e+00 -2.88069636e-01
2.85862416e-01 6.95601642e-01 1.64456904e-01 -1.64449632e-01
4.95084852e-01 9.00118351e-01 -1.43625927e+00 -4.36445065e-02
4.96660084e-01 1.27129579e+00 1.80271909e-01 8.29529643e-01
5.67880906e-02 5.00252664e-01 -1.66798875e-01 -1.05774336e-01
5.61376810e-01 1.00724228e-01 -5.64376652e-01 3.37586910e-01
9.50688422e-01 -3.48683238e-01 -6.70101762e-01 1.00748837e+00
2.95678616e-01 2.89873302e-01 4.46793228e-01 -1.11276984e+00
-6.03363872e-01 1.24372467e-02 -5.73092997e-01 -1.61319301e-01
5.97455502e-01 4.40082997e-02 7.38877654e-01 -1.10835993e+00
6.40224934e-01 1.13903117e+00 1.05694366e+00 1.17039129e-01
-1.20550632e+00 -5.65338314e-01 1.88692942e-01 1.85176492e-01
-1.82913649e+00 -6.72125518e-02 1.12163711e+00 -6.40108466e-01
1.38827312e+00 4.87907939e-02 9.95932460e-01 7.82395720e-01
2.69160628e-01 1.76449195e-02 6.34115756e-01 -3.69422436e-01
2.89570373e-02 -1.55895069e-01 -2.06984371e-01 1.05936456e+00
-4.17705402e-02 2.57192820e-01 -3.49835753e-01 2.69418247e-02
1.19500327e+00 3.60679418e-01 -1.55150026e-01 -1.01371646e+00
-1.23118770e+00 8.07419062e-01 1.08789456e+00 -1.38272971e-01
-9.02815312e-02 5.49277067e-01 -7.65753165e-03 1.04678608e-01
3.95603031e-01 5.35357475e-01 -4.02421027e-01 1.05604596e-01
-7.04340994e-01 3.15829962e-02 6.42689586e-01 1.06348109e+00
1.20038164e+00 -2.60647446e-01 5.16167104e-01 5.92481673e-01
4.19758350e-01 6.30040228e-01 1.80005223e-01 -1.19861841e+00
6.29438758e-01 7.74427593e-01 5.46079278e-02 -1.47098041e+00
-4.39647228e-01 -5.38870879e-02 -7.39957571e-01 2.38699913e-01
6.50685057e-02 1.95607081e-01 -8.34010363e-01 1.38326681e+00
3.58848691e-01 5.36721945e-01 -3.02779347e-01 9.43403780e-01
6.75446928e-01 5.80678403e-01 -5.14600098e-01 5.43504477e-01
9.63240921e-01 -1.13359594e+00 -1.68160319e-01 -3.97882581e-01
4.54267681e-01 -7.93035567e-01 8.51983249e-01 4.55510877e-02
-8.93592954e-01 -8.81261885e-01 -1.24826598e+00 -4.60168898e-01
-5.52952945e-01 1.14783384e-01 4.32918489e-01 3.31004083e-01
-1.35078633e+00 6.14402831e-01 -9.15729821e-01 -3.48636180e-01
1.22461468e-01 6.98577344e-01 -5.69002151e-01 -1.82050154e-01
-6.62952662e-01 9.57927287e-01 2.94397771e-01 1.95529848e-01
-6.65263712e-01 -5.72849274e-01 -1.40388167e+00 7.62575269e-02
1.70971021e-01 -1.14538896e+00 9.24242914e-01 -9.14496243e-01
-1.48713195e+00 1.10070658e+00 -4.75635864e-02 -2.32758015e-01
4.14410055e-01 -2.66554356e-01 1.21470466e-02 -1.33170515e-01
8.05056989e-02 8.88781786e-01 8.26873064e-01 -1.38960814e+00
-6.57974303e-01 -4.80722845e-01 2.14342311e-01 4.61412966e-01
1.84074044e-01 -5.35419345e-01 -9.10031736e-01 -4.97447282e-01
5.34524858e-01 -1.23552847e+00 -3.94499391e-01 4.41626608e-01
-3.96967918e-01 -7.59546785e-03 9.06945169e-01 -4.88123298e-01
7.28106081e-01 -2.12779760e+00 3.91722798e-01 4.10906434e-01
1.89608648e-01 -2.69303769e-01 -1.14709221e-01 2.41604298e-01
-2.80312926e-01 -2.24590391e-01 -6.79607466e-02 -7.52650380e-01
-4.78879102e-02 1.61582336e-01 -2.53358692e-01 7.93753028e-01
1.81613281e-01 7.94928908e-01 -8.41616869e-01 -8.63087699e-02
9.10300255e-01 7.52331138e-01 -7.67259359e-01 4.45212245e-01
-1.43151343e-01 2.48901933e-01 -6.88836426e-02 4.70049351e-01
8.39733481e-01 -2.92377740e-01 -2.59841289e-02 -4.68203783e-01
-1.40207201e-01 2.33141840e-01 -1.37556851e+00 2.39729905e+00
-6.48265481e-01 7.26339877e-01 -2.62212873e-01 -7.76257694e-01
1.06954050e+00 -2.22239673e-01 5.56888878e-01 -5.18038571e-01
3.11162651e-01 -4.00133580e-02 -4.98738319e-01 -2.51728535e-01
6.72027111e-01 2.37774685e-01 -2.30195001e-01 5.02000675e-02
1.04739666e-01 -6.44876301e-01 -3.41082335e-01 -1.32962108e-01
9.36142385e-01 3.86024684e-01 1.22614339e-01 -1.03270628e-01
6.05063081e-01 9.98289213e-02 2.29874715e-01 2.65758991e-01
3.29293340e-01 9.15811002e-01 7.48826638e-02 -9.82582211e-01
-1.18505716e+00 -9.46363151e-01 2.11571798e-01 6.70646191e-01
8.74693871e-01 -3.29245031e-01 -7.10098803e-01 -4.81895596e-01
1.17262296e-01 5.42955220e-01 -7.60000825e-01 -2.48717293e-01
-6.69807255e-01 -1.73035283e-02 1.22541875e-01 7.14124262e-01
6.15404010e-01 -5.63695550e-01 -7.21638858e-01 -5.07293716e-02
-1.31577000e-01 -1.34824455e+00 -8.72998238e-01 4.30023015e-01
-6.89721942e-01 -1.33516574e+00 -5.40613271e-02 -9.56821382e-01
1.01839757e+00 4.93861049e-01 1.13685882e+00 1.37741089e-01
-1.33202225e-01 5.52815199e-01 -6.85546920e-02 -2.03474090e-01
-1.50089443e-01 1.54740708e-02 1.51901543e-01 -7.22295716e-02
4.01991755e-01 -5.66202819e-01 -5.97329319e-01 4.70815152e-01
-5.44206142e-01 3.10709059e-01 2.54971564e-01 5.79509079e-01
8.68054569e-01 -3.78634483e-02 -7.30757058e-01 -5.73894382e-01
1.93987116e-02 -1.06505051e-01 -1.15212023e+00 2.18316931e-02
-4.93524730e-01 2.15973437e-01 5.26675344e-01 -4.23589021e-01
-4.70180303e-01 8.68034184e-01 -2.36412197e-01 -1.05741787e+00
-1.72175720e-01 2.49596864e-01 -1.07009530e-01 -5.26296258e-01
5.65399170e-01 -4.21804003e-02 -2.26051837e-01 -4.53731149e-01
5.07009685e-01 3.85895222e-01 9.05464232e-01 -5.07558525e-01
1.25223839e+00 6.10746801e-01 1.24099702e-01 -5.45873761e-01
-8.23180377e-01 -7.01319396e-01 -1.26795602e+00 -1.55125692e-01
1.15566969e+00 -1.14964080e+00 -7.35062957e-01 3.68790627e-01
-1.51023412e+00 -5.83820105e-01 -1.89485162e-01 4.58745152e-01
-8.04317355e-01 2.44991779e-01 -3.27622503e-01 -1.78003639e-01
-2.53833652e-01 -1.33852744e+00 1.59354067e+00 4.94722016e-02
-4.98713963e-02 -8.71413887e-01 3.41174066e-01 2.32373074e-01
1.01337031e-01 5.42994559e-01 6.54618800e-01 1.13413811e-01
-9.91717041e-01 -3.02841634e-01 -3.80291313e-01 1.64699197e-01
-3.87775823e-02 8.61006454e-02 -9.72345591e-01 -3.71760368e-01
-1.46383226e-01 9.59736854e-02 5.64771652e-01 3.32003862e-01
1.34919167e+00 -4.66206148e-02 -3.88946146e-01 1.42287028e+00
1.77057850e+00 8.81237388e-02 6.56244874e-01 4.71990317e-01
1.24805629e+00 2.03257382e-01 2.63438016e-01 2.27054298e-01
6.92451298e-01 9.55979466e-01 9.90337253e-01 -2.05611527e-01
-1.36578530e-01 -7.16337204e-01 2.00928256e-01 7.52067447e-01
-3.33385766e-02 -4.89676520e-02 -9.97107565e-01 2.94582307e-01
-1.79199064e+00 -5.23723841e-01 -1.92511128e-03 2.17817092e+00
2.60183543e-01 1.84955761e-01 -2.38378048e-01 -1.39923930e-01
6.59127414e-01 1.43470734e-01 -5.57747781e-01 -3.70560825e-01
2.50489593e-01 1.35096192e-01 8.41114879e-01 7.66389191e-01
-1.26154602e+00 1.05105019e+00 5.57596684e+00 1.94152743e-01
-1.37736964e+00 -1.58709094e-01 3.26747149e-01 1.42617673e-01
3.32195982e-02 8.96154717e-02 -6.46544397e-01 1.74311519e-01
6.28297329e-01 2.67540902e-01 6.92963243e-01 1.22322655e+00
-1.22940779e-01 9.46412161e-02 -1.42234802e+00 1.51316690e+00
4.50209498e-01 -1.60819018e+00 -1.45879295e-02 2.41475761e-01
6.88711226e-01 5.33053756e-01 -1.20503917e-01 6.32025972e-02
3.61676157e-01 -1.07327759e+00 1.02135670e+00 5.12297094e-01
1.15966868e+00 -6.57091320e-01 6.88677311e-01 3.50121737e-01
-1.47027087e+00 -3.14325355e-02 -5.54097891e-01 -1.67717844e-01
-4.15899567e-02 1.40273392e-01 -8.40329707e-01 3.45099241e-01
8.73704135e-01 9.20826077e-01 -7.07090557e-01 9.59921420e-01
-3.00739914e-01 -2.21149191e-01 -5.28580010e-01 2.48676464e-01
2.92290777e-01 -2.66322970e-01 2.82258540e-01 9.19521511e-01
4.90458846e-01 -1.12028204e-01 3.68979990e-01 9.29943860e-01
-3.13318908e-01 -3.68321627e-01 -9.19218838e-01 5.63901901e-01
3.21057528e-01 1.19152880e+00 -7.67250001e-01 -1.41964853e-01
-3.94138396e-01 1.36463094e+00 6.01458609e-01 1.88412622e-01
-9.37641442e-01 -2.97164351e-01 7.00794637e-01 2.86184430e-01
5.61095297e-01 -6.32637501e-01 -2.54769415e-01 -1.10582316e+00
1.93824398e-03 -5.00406086e-01 4.39273715e-02 -1.17501044e+00
-9.87024307e-01 7.27210224e-01 -1.99357048e-01 -1.29555094e+00
-1.74846023e-01 -7.51361847e-01 -4.99943316e-01 7.97405481e-01
-1.31881881e+00 -1.23370528e+00 -9.35815156e-01 6.85238123e-01
5.32025635e-01 1.88195735e-01 8.90432894e-01 2.97719657e-01
-3.39816421e-01 3.14100653e-01 -1.14054218e-01 4.72034633e-01
4.47548151e-01 -1.26051652e+00 8.17483544e-01 8.53876591e-01
4.80279177e-01 5.36664426e-01 3.79121363e-01 -3.22677255e-01
-1.69008744e+00 -1.29763281e+00 6.84356570e-01 -9.39849377e-01
4.94489521e-01 -6.81460023e-01 -4.93804872e-01 9.30880249e-01
-8.54940340e-03 4.37111259e-01 1.58425301e-01 -5.82390279e-02
-5.66025555e-01 -2.39373565e-01 -9.74836707e-01 4.29261535e-01
1.24590397e+00 -7.76870549e-01 -4.79696482e-01 3.07075500e-01
1.00692070e+00 -1.12579000e+00 -9.85048652e-01 2.17979953e-01
4.78804499e-01 -9.16903317e-01 1.28459287e+00 -2.00547442e-01
4.16924804e-01 -5.77635169e-01 -4.99927640e-01 -1.17764688e+00
-4.58502591e-01 -3.49333733e-01 1.42685219e-03 5.80024838e-01
1.99056268e-01 -2.91376740e-01 9.68045235e-01 5.83533585e-01
-3.39944035e-01 -7.41729558e-01 -9.80970860e-01 -3.83530557e-01
-2.17172265e-01 -6.36718631e-01 6.57357693e-01 8.97901118e-01
-5.32844126e-01 4.17940944e-01 -1.17314003e-01 6.30662858e-01
4.62863594e-01 3.25093985e-01 1.19175673e+00 -1.04471719e+00
-1.52150705e-01 -3.49899173e-01 -7.99066067e-01 -1.53094137e+00
3.03333372e-01 -8.97662640e-01 8.58192146e-02 -1.58729076e+00
-3.24329771e-02 -5.15242755e-01 -3.71016711e-02 2.67906636e-01
1.98452413e-01 4.22225833e-01 1.74716055e-01 4.51385118e-02
-6.21836364e-01 4.48597640e-01 9.23395872e-01 -3.39279711e-01
-2.59273738e-01 -1.86007246e-01 -3.59762818e-01 8.92022431e-01
5.75081587e-01 -5.17890513e-01 -2.22317964e-01 -1.03034890e+00
3.28534693e-01 -2.01665191e-03 7.00752616e-01 -1.29847133e+00
5.25616050e-01 -3.37194614e-02 6.86401248e-01 -9.18595433e-01
6.81545973e-01 -1.38713980e+00 3.58893126e-01 4.14067179e-01
-4.32351157e-02 5.78781366e-01 1.96909040e-01 6.43389940e-01
-8.78597051e-02 8.09335187e-02 7.48575509e-01 -1.52538702e-01
-8.60657692e-01 7.28578448e-01 1.79999620e-01 -3.52684617e-01
8.75899196e-01 -6.07671261e-01 7.54409805e-02 -2.89982438e-01
-4.84673262e-01 6.67848736e-02 1.04926276e+00 7.14986384e-01
8.28503847e-01 -1.50676858e+00 -2.97240287e-01 6.61182106e-01
2.43348032e-01 6.46663904e-01 -3.74938175e-02 5.00338137e-01
-1.22171676e+00 2.71724701e-01 -8.87190849e-02 -1.05279124e+00
-1.16360033e+00 7.61381209e-01 7.34171271e-01 -6.95363879e-02
-6.54149294e-01 1.04112542e+00 7.22160339e-02 -8.05550218e-01
3.46810758e-01 -7.01713324e-01 2.65173674e-01 -4.26691473e-01
6.48047403e-02 5.86707368e-02 1.06665753e-01 -1.06588936e+00
-4.27219063e-01 1.23973358e+00 4.62699503e-01 7.88677260e-02
1.34077322e+00 -1.86961606e-01 -2.08596870e-01 2.50368744e-01
1.67406201e+00 -1.04936108e-01 -1.50812745e+00 -3.49288881e-01
-2.56514966e-01 -6.73442245e-01 2.25087851e-01 -2.85459906e-01
-1.12620807e+00 8.78444970e-01 6.48710191e-01 -1.78925052e-01
8.96230280e-01 -7.98381194e-02 7.32771456e-01 5.69768488e-01
5.22870660e-01 -9.53235507e-01 2.04856500e-01 7.17976570e-01
9.31688547e-01 -1.24100387e+00 1.42475590e-01 -3.62288415e-01
-1.39785379e-01 1.37807214e+00 6.80520177e-01 -4.83093441e-01
8.15508306e-01 2.96905309e-01 -6.64596930e-02 -4.05645579e-01
-3.34550828e-01 4.70000617e-02 5.67766607e-01 6.72686160e-01
-3.46178710e-02 -3.78335938e-02 7.71673143e-01 -2.69931331e-02
-5.75434744e-01 -2.83102244e-01 1.46032467e-01 6.22620225e-01
-2.58751184e-01 -8.16379845e-01 -5.48640311e-01 9.07637924e-02
2.41798252e-01 3.22132185e-02 -5.08767605e-01 8.55333686e-01
4.52984005e-01 6.19676292e-01 4.95682448e-01 -7.81691194e-01
6.73794150e-01 -3.61393571e-01 5.88237524e-01 -6.47846699e-01
-3.94976437e-01 -2.93415077e-02 -3.16861451e-01 -1.01971734e+00
-3.64335716e-01 -3.66298676e-01 -1.02155018e+00 -4.45695460e-01
-2.89518535e-01 -3.54936182e-01 9.19954240e-01 5.58852851e-01
3.84572655e-01 4.03062224e-01 8.03239286e-01 -1.53993440e+00
-1.59344878e-02 -5.60005784e-01 -1.91326305e-01 4.17154163e-01
3.86393696e-01 -6.95119560e-01 -2.36516893e-01 7.48967677e-02] | [7.7256855964660645, -2.287092924118042] |
152fb007-121c-4f41-8b2f-27da0648926c | multiphase-level-set-loss-for-semi-supervised | 1904.02872 | null | https://arxiv.org/abs/1904.02872v2 | https://arxiv.org/pdf/1904.02872v2.pdf | Mumford-Shah Loss Functional for Image Segmentation with Deep Learning | Recent state-of-the-art image segmentation algorithms are mostly based on deep neural networks, thanks to their high performance and fast computation time. However, these methods are usually trained in a supervised manner, which requires large number of high quality ground-truth segmentation masks. On the other hand, classical image segmentation approaches such as level-set methods are formulated in a self-supervised manner by minimizing energy functions such as Mumford-Shah functional, so they are still useful to help generation of segmentation masks without labels. Unfortunately, these algorithms are usually computationally expensive and often have limitation in semantic segmentation. In this paper, we propose a novel loss function based on Mumford-Shah functional that can be used in deep-learning based image segmentation without or with small labeled data. This loss function is based on the observation that the softmax layer of deep neural networks has striking similarity to the characteristic function in the Mumford-Shah functional. We show that the new loss function enables semi-supervised and unsupervised segmentation. In addition, our loss function can be also used as a regularized function to enhance supervised semantic segmentation algorithms. Experimental results on multiple datasets demonstrate the effectiveness of the proposed method. | ['Boah Kim', 'Jong Chul Ye'] | 2019-04-05 | null | null | null | null | ['unsupervised-semantic-segmentation'] | ['computer-vision'] | [ 4.17125612e-01 2.91771919e-01 -2.91603923e-01 -8.32383096e-01
-6.46198690e-01 -2.64336795e-01 1.60338104e-01 3.95806022e-02
-7.45602012e-01 6.87013268e-01 -3.63903433e-01 -2.09801272e-02
-3.89699684e-03 -9.57605124e-01 -8.46438825e-01 -8.06639493e-01
4.09917504e-01 5.32753050e-01 5.12828171e-01 -7.01781735e-03
9.98759493e-02 2.69580424e-01 -1.27828038e+00 -1.70493528e-01
1.26788390e+00 1.10775995e+00 4.25954044e-01 2.10457072e-02
-4.72571403e-01 4.30533469e-01 -4.86030936e-01 -1.56121850e-01
2.79702127e-01 -5.76627851e-01 -1.08369899e+00 4.08929855e-01
1.44168824e-01 -1.36333093e-01 -5.00078127e-02 1.33922005e+00
2.73063272e-01 2.62879193e-01 5.54566979e-01 -1.01478553e+00
-3.66302639e-01 6.19103432e-01 -6.35949135e-01 -2.60896027e-01
-2.02309906e-01 -2.80611962e-01 8.55713904e-01 -6.04095399e-01
3.95098120e-01 1.05214608e+00 6.83977664e-01 3.84173155e-01
-1.06181002e+00 -5.22475362e-01 1.70450583e-01 -5.65941818e-02
-1.43050790e+00 3.10206227e-02 1.16485596e+00 -3.66141111e-01
2.27782845e-01 7.67126819e-03 5.31556726e-01 4.66113359e-01
-2.67018288e-01 1.10701954e+00 1.16043544e+00 -4.07019377e-01
2.41731867e-01 1.95579499e-01 6.17613196e-02 9.50012803e-01
9.36210006e-02 -3.04518938e-01 3.38412784e-02 2.54837155e-01
1.03275990e+00 4.87084799e-02 -2.98013389e-01 -5.39002776e-01
-9.34998393e-01 9.92278934e-01 7.39782929e-01 5.12427330e-01
-1.30105883e-01 1.26005352e-01 2.75605619e-01 -1.26927704e-01
5.76117754e-01 3.57973069e-01 -3.20634335e-01 3.84372532e-01
-1.23633206e+00 -6.68313876e-02 4.87703383e-01 6.44966602e-01
1.06816101e+00 -9.05682892e-03 -1.04492754e-02 8.91275585e-01
4.62374330e-01 5.32004684e-02 6.10031128e-01 -8.72629344e-01
6.40989542e-02 7.57906258e-01 1.19308986e-01 -9.59699512e-01
-4.79255527e-01 -4.53372270e-01 -9.56463814e-01 2.11676940e-01
5.51483035e-01 -7.47739710e-03 -1.22356379e+00 1.60221601e+00
5.15956581e-01 2.09136724e-01 -3.71259749e-02 1.03606737e+00
7.67246306e-01 5.93118966e-01 -3.99837829e-02 -2.42512971e-01
9.50848639e-01 -1.16965270e+00 -7.89344668e-01 -1.70847569e-02
6.29920006e-01 -5.66535354e-01 1.19757462e+00 3.17294985e-01
-9.77913201e-01 -5.79767942e-01 -1.00611246e+00 -2.61970051e-02
-3.00013185e-01 1.62952423e-01 7.67369390e-01 6.37098789e-01
-8.98028791e-01 7.92497575e-01 -9.95366693e-01 -1.35151297e-01
7.43064284e-01 5.17325461e-01 -3.17251049e-02 2.95366347e-01
-1.09581494e+00 5.68339407e-01 7.90799797e-01 3.54607850e-01
-5.95421672e-01 -2.63560206e-01 -9.42426503e-01 4.32492197e-02
6.50114775e-01 -2.40846291e-01 1.05675387e+00 -1.23744988e+00
-1.73805439e+00 9.27798569e-01 1.07491732e-01 -5.45514941e-01
5.19796073e-01 -1.38727680e-01 -6.86115259e-03 3.50515723e-01
2.03932911e-01 8.55658293e-01 7.64936447e-01 -1.41703153e+00
-4.39766884e-01 -2.81851262e-01 -2.97374390e-02 8.51263478e-02
-4.44348395e-01 -3.26756984e-01 -4.90298927e-01 -6.74784303e-01
4.46201295e-01 -4.58961427e-01 -4.77936655e-01 5.24727888e-02
-6.19617879e-01 -3.45756233e-01 9.87112224e-01 -4.71628457e-01
1.04553747e+00 -1.94858527e+00 -4.95884195e-02 4.03000027e-01
2.49525788e-03 3.51626992e-01 1.85446575e-01 -1.09989345e-01
2.38842711e-01 2.51722723e-01 -1.05588233e+00 -1.79433972e-01
-1.55887544e-01 2.94941604e-01 6.86412454e-02 5.25640070e-01
1.97332621e-01 7.86847293e-01 -8.24262738e-01 -8.85596871e-01
3.71579796e-01 3.93965065e-01 -3.51443261e-01 1.87008411e-01
-4.38527077e-01 7.72417843e-01 -5.80879271e-01 3.58274996e-01
6.81523800e-01 -3.84074420e-01 9.00140294e-05 -2.16593236e-01
-6.41769394e-02 -2.02377051e-01 -1.12095201e+00 2.00609541e+00
-3.89310747e-01 3.34190547e-01 2.50027448e-01 -1.73159277e+00
1.03210759e+00 6.27511963e-02 6.72912419e-01 -4.52032477e-01
5.12374818e-01 3.49177957e-01 -2.15018228e-01 -2.74053007e-01
1.58445492e-01 -3.09427708e-01 -3.27465348e-02 2.47252584e-01
-7.60465637e-02 -3.73937666e-01 4.49575037e-01 -6.99612722e-02
3.44157606e-01 3.17988425e-01 -3.05404072e-03 -3.32450151e-01
8.25302839e-01 -3.40098585e-03 6.88990355e-01 4.82599974e-01
-1.41145904e-02 7.94166684e-01 2.70759851e-01 -2.98672110e-01
-8.19904864e-01 -6.79899991e-01 -5.22243202e-01 6.02111995e-01
6.18225336e-01 6.54887632e-02 -1.35199344e+00 -8.39596868e-01
-3.36347193e-01 4.44294304e-01 -4.57064331e-01 -7.12462654e-03
-5.97218633e-01 -8.40195417e-01 4.62912321e-01 6.34600937e-01
9.72479284e-01 -1.09312701e+00 -5.41911066e-01 3.87520880e-01
-1.16153911e-01 -1.25546551e+00 -5.78136921e-01 8.73922035e-02
-1.01804090e+00 -1.00168121e+00 -9.73766446e-01 -1.15132821e+00
9.83130515e-01 8.41031745e-02 7.99131334e-01 2.75962561e-01
-1.46059275e-01 -2.84989132e-03 -3.62568855e-01 -2.23953411e-01
-2.05206960e-01 1.64602965e-01 -2.58721828e-01 4.03516948e-01
2.82071885e-02 -4.84318972e-01 -6.33337975e-01 5.31556010e-01
-1.29966056e+00 7.67860413e-02 5.46250463e-01 9.65776742e-01
1.11239266e+00 4.30891931e-01 6.28461063e-01 -1.18752980e+00
3.37171763e-01 -5.68299070e-02 -8.86297703e-01 2.18940184e-01
-6.82817578e-01 1.59901962e-01 7.06308722e-01 -2.66173869e-01
-1.14381850e+00 4.00710940e-01 -4.02311593e-01 -2.32771590e-01
-2.63862848e-01 4.99356836e-01 -3.02126527e-01 -3.03602368e-01
2.86321312e-01 1.97517425e-01 3.77188250e-02 -6.03807747e-01
3.81261647e-01 6.26598597e-01 5.47637403e-01 -4.43021715e-01
6.89148903e-01 7.57886529e-01 7.00756535e-02 -8.17465603e-01
-1.05036080e+00 -6.33644819e-01 -7.09135830e-01 2.19141692e-02
1.15350413e+00 -5.28162539e-01 -4.10953194e-01 7.99818456e-01
-8.99054289e-01 -5.72297812e-01 -3.12994063e-01 4.43649471e-01
-6.16465986e-01 6.81826651e-01 -5.56055665e-01 -6.71879590e-01
-2.51388192e-01 -1.24704897e+00 1.15063071e+00 5.43105841e-01
2.48657957e-01 -1.18856907e+00 -2.11888582e-01 5.33315897e-01
1.01282261e-01 4.87046510e-01 7.63744235e-01 -4.77529407e-01
-3.55138391e-01 -1.68462813e-01 -4.30616617e-01 6.48209691e-01
1.91529959e-01 -2.20256388e-01 -7.07307577e-01 -1.11469463e-01
6.75980523e-02 -3.85111302e-01 1.11543274e+00 7.33039021e-01
1.68007731e+00 -3.48840393e-02 -1.85128316e-01 7.86998868e-01
1.61789811e+00 2.60371298e-01 5.03337502e-01 2.29314968e-01
9.76662636e-01 5.99141121e-01 5.31272054e-01 4.96815182e-02
2.98445076e-01 5.83391845e-01 4.38478559e-01 -6.08459115e-01
1.28233120e-01 -1.41471520e-01 -1.25512764e-01 7.59416103e-01
1.50156006e-01 -9.26455036e-02 -8.02712142e-01 5.86823046e-01
-1.94579053e+00 -4.81934696e-01 -3.83608848e-01 2.17651772e+00
1.05732286e+00 3.39843631e-01 2.64285535e-01 4.02787030e-01
7.65606225e-01 4.38421145e-02 -5.18186092e-01 -1.96848065e-01
-1.28944412e-01 3.64824831e-01 5.29338062e-01 5.58525741e-01
-1.31270230e+00 1.20687461e+00 5.22740793e+00 1.16887414e+00
-1.10885096e+00 2.81241179e-01 7.99312115e-01 3.65441024e-01
-9.54343900e-02 -1.11180648e-01 -6.30144477e-01 5.83099008e-01
4.19989020e-01 2.40822747e-01 2.69732885e-02 8.63894105e-01
1.42218664e-01 -3.21195483e-01 -7.13635385e-01 8.79582882e-01
-1.89111650e-01 -1.13075352e+00 -2.14992106e-01 -2.02460200e-01
9.37900603e-01 -3.62435609e-01 -2.08913982e-01 4.51844074e-02
8.65582675e-02 -1.03410614e+00 5.60573041e-01 1.23913243e-01
6.81376457e-01 -8.29898715e-01 8.56856704e-01 3.31350833e-01
-1.15246451e+00 3.06812406e-01 -3.41858476e-01 2.01682791e-01
4.31055665e-01 8.58096540e-01 -5.02376616e-01 3.51160437e-01
4.69085664e-01 5.88255286e-01 -1.26811907e-01 1.13180947e+00
-2.92923510e-01 6.30109251e-01 -4.29556847e-01 1.14859141e-01
6.79469883e-01 -5.39509892e-01 1.67886555e-01 1.05255055e+00
-4.77510095e-02 3.78256403e-02 6.97276890e-01 9.78467286e-01
-2.37419114e-01 2.92946756e-01 -2.02608898e-01 -1.34489298e-01
-1.63566172e-02 1.21665812e+00 -1.48941100e+00 -3.19888651e-01
-2.31214985e-01 1.05077291e+00 1.41168445e-01 3.15999389e-01
-7.03960001e-01 -5.45638382e-01 -5.29815741e-02 1.84104979e-01
1.92610636e-01 -1.69558778e-01 -5.49307704e-01 -8.99183273e-01
4.45345230e-02 -2.44529337e-01 6.79869726e-02 -3.09620947e-01
-1.16680038e+00 5.18998563e-01 -2.51926165e-02 -9.96821582e-01
1.39512226e-01 -5.17523706e-01 -5.96148193e-01 4.99292105e-01
-1.62781882e+00 -1.01476943e+00 -2.97799289e-01 5.80332160e-01
6.15254879e-01 2.12182477e-01 3.86726558e-01 3.48110557e-01
-7.26737440e-01 3.84893715e-01 1.89899415e-01 3.58709663e-01
3.27670962e-01 -1.49610376e+00 1.31921768e-02 7.40520597e-01
3.48401889e-02 2.03289524e-01 3.36843997e-01 -5.17960072e-01
-7.14209020e-01 -1.19519603e+00 3.92344058e-01 2.36025438e-01
3.09997559e-01 -2.18098506e-01 -1.13959515e+00 3.35128218e-01
-3.64993475e-02 4.62528467e-02 4.69125301e-01 -2.80723810e-01
2.16707885e-01 -1.29224762e-01 -1.28405046e+00 2.30887026e-01
7.17463195e-01 -2.03925729e-01 -2.79431194e-01 6.94623291e-01
7.21050620e-01 -4.16868955e-01 -7.67442226e-01 6.15002573e-01
2.07026437e-01 -8.53485584e-01 7.98850894e-01 -3.52556378e-01
3.52948934e-01 -3.94538790e-01 2.01321080e-01 -1.03100479e+00
3.99569348e-02 -4.36514527e-01 3.18942964e-01 1.17391050e+00
2.46334374e-01 -5.76482415e-01 1.00775516e+00 3.01678598e-01
-2.96881855e-01 -1.15279567e+00 -6.79282665e-01 -8.94809842e-01
8.18063915e-02 -1.73655629e-01 3.08702946e-01 9.44796264e-01
-4.96705115e-01 1.28875852e-01 -1.30283281e-01 -1.03268825e-01
8.65931034e-01 1.75856769e-01 3.64513487e-01 -1.35148120e+00
-1.36441946e-01 -6.82366252e-01 -2.04322428e-01 -1.08910239e+00
3.32292110e-01 -8.51792932e-01 4.86210346e-01 -1.82615328e+00
2.33488828e-02 -8.01925838e-01 -3.61612588e-01 4.26545858e-01
-2.08876833e-01 4.69388604e-01 -1.66405603e-01 2.86850095e-01
-4.97583151e-01 7.56863892e-01 1.51311517e+00 -1.18453369e-01
-4.63618040e-01 7.92478696e-02 -4.07886803e-01 1.23010969e+00
9.55109179e-01 -4.87361819e-01 -4.76789981e-01 -3.28945041e-01
-1.67146757e-01 -1.54856324e-01 2.02139527e-01 -8.11043680e-01
5.36440425e-02 -1.79738566e-01 6.69736937e-02 -4.78968740e-01
1.58008218e-01 -7.74091184e-01 -2.59683222e-01 2.87580281e-01
-2.03217313e-01 -6.55632913e-01 -1.38732687e-01 4.92122233e-01
-5.51917553e-01 -5.14265776e-01 1.12406754e+00 -3.39402765e-01
-6.76105142e-01 3.87450010e-01 -1.28450200e-01 1.93201184e-01
1.04570210e+00 -3.60337615e-01 3.22851062e-01 -2.21536160e-01
-5.63599169e-01 3.36691350e-01 3.71657372e-01 -5.11813350e-02
5.30277252e-01 -1.23898160e+00 -4.66149569e-01 7.52665773e-02
-2.69613206e-01 6.89143181e-01 2.18177438e-02 9.44516301e-01
-7.72243619e-01 2.69658595e-01 -1.34643808e-01 -7.46473014e-01
-8.05234790e-01 3.70196760e-01 3.17332894e-01 -3.04844558e-01
-5.67341745e-01 9.94840920e-01 4.71587658e-01 -4.89988446e-01
2.98764855e-01 -3.66314650e-01 -2.70343751e-01 3.27426791e-02
1.68016806e-01 2.95123905e-01 -4.97177653e-02 -7.73857594e-01
-2.77999312e-01 7.59918332e-01 6.62928075e-02 -3.40912119e-02
1.16314542e+00 -4.45656702e-02 -1.49423108e-01 2.07415625e-01
1.32319653e+00 -4.18326765e-01 -1.28964472e+00 -3.62757951e-01
4.68142070e-02 -2.89408833e-01 3.90346944e-01 -4.34777915e-01
-1.46673810e+00 9.80186641e-01 6.08420193e-01 3.11084032e-01
1.29799724e+00 -1.30805135e-01 1.35367906e+00 1.54028878e-01
4.01317716e-01 -1.36865079e+00 3.10094599e-02 1.06898092e-01
4.71817464e-01 -1.46215045e+00 -1.59606650e-01 -7.20790267e-01
-3.99683744e-01 1.07291818e+00 6.04350626e-01 -2.16084585e-01
6.18940949e-01 1.64919391e-01 6.51054755e-02 -4.92721759e-02
3.21259826e-01 -3.87345254e-01 4.08815712e-01 2.86743075e-01
4.09288615e-01 8.17508698e-02 -7.75342286e-01 6.19973302e-01
3.91997993e-02 2.65738089e-02 2.88533509e-01 8.58860731e-01
-5.88600874e-01 -1.20577705e+00 -2.70757198e-01 4.15667564e-01
-5.12661874e-01 1.32662281e-01 -2.40953624e-01 6.66559041e-01
3.33074808e-01 8.73005450e-01 -1.92065034e-02 -7.73493275e-02
3.19753326e-02 -1.92491591e-01 4.84359294e-01 -6.64564967e-01
-3.35025460e-01 4.28011268e-01 -3.37094069e-01 -3.91710401e-01
-7.83724248e-01 -3.61707151e-01 -1.93030405e+00 3.30100060e-01
-6.50784433e-01 2.35780343e-01 7.19604611e-01 1.11497104e+00
-2.12926850e-01 4.02630746e-01 6.48357749e-01 -8.09755743e-01
-3.18834543e-01 -8.54778409e-01 -7.02119231e-01 5.94506323e-01
-1.01305023e-02 -7.87615478e-01 -1.69679880e-01 2.49037400e-01] | [14.447525024414062, -2.0694518089294434] |
8f5d3df3-908a-4283-9d14-f85714be85a8 | an-ensemble-of-visnet-transformer-m-and | 2211.12791 | null | https://arxiv.org/abs/2211.12791v1 | https://arxiv.org/pdf/2211.12791v1.pdf | An ensemble of VisNet, Transformer-M, and pretraining models for molecular property prediction in OGB Large-Scale Challenge @ NeurIPS 2022 | In the technical report, we provide our solution for OGB-LSC 2022 Graph Regression Task. The target of this task is to predict the quantum chemical property, HOMO-LUMO gap for a given molecule on PCQM4Mv2 dataset. In the competition, we designed two kinds of models: Transformer-M-ViSNet which is an geometry-enhanced graph neural network for fully connected molecular graphs and Pretrained-3D-ViSNet which is a pretrained ViSNet by distilling geomeotric information from optimized structures. With an ensemble of 22 models, ViSNet Team achieved the MAE of 0.0723 eV on the test-challenge set, dramatically reducing the error by 39.75% compared with the best method in the last year competition. | ['Tie-Yan Liu', 'Bin Shao', 'Xinheng He', 'Zun Wang', 'Tong Wang', 'Shaoning Li', 'Yusong Wang'] | 2022-11-23 | null | null | null | null | ['graph-regression', 'molecular-property-prediction'] | ['graphs', 'miscellaneous'] | [ 1.97135687e-01 3.64052981e-01 -2.39842728e-01 -3.76008786e-02
-5.68500459e-01 -4.20386732e-01 1.99431852e-01 3.90230417e-01
-2.68000960e-01 1.34844136e+00 2.17392445e-02 -8.52738380e-01
1.43601418e-01 -7.66348720e-01 -1.03345704e+00 -8.05411279e-01
-3.73328507e-01 4.81057405e-01 -5.24971671e-02 -3.93583089e-01
3.70604128e-01 6.96770310e-01 -6.23045146e-01 2.57720411e-01
9.64300692e-01 1.07698333e+00 -5.12351580e-02 6.75101995e-01
2.62477696e-01 9.22792912e-01 -1.65894806e-01 -5.56348026e-01
1.57656729e-01 -4.99738008e-01 -1.06564975e+00 -8.24888110e-01
8.78396213e-01 2.99342543e-01 -8.07430327e-01 1.14484549e+00
9.26272392e-01 4.27573532e-01 8.29673886e-01 -7.82899857e-01
-6.59845829e-01 7.51717985e-01 -3.42280358e-01 3.27226132e-01
3.52620095e-01 5.81465900e-01 1.14744139e+00 -6.28758192e-01
1.07352293e+00 1.09775817e+00 8.06432545e-01 5.67353308e-01
-1.44263256e+00 -7.00093031e-01 -1.08827852e-01 6.99959397e-01
-1.55993235e+00 -2.69572169e-01 5.20682871e-01 -2.57896185e-01
2.29274392e+00 -3.10798571e-03 5.80451071e-01 1.11775064e+00
7.40110159e-01 -1.19351372e-01 8.71729910e-01 -3.04270871e-02
7.01567307e-02 -4.50586706e-01 2.53369272e-01 1.18624127e+00
3.80879730e-01 4.07461643e-01 -5.86608648e-01 -1.63311303e-01
1.39324889e-01 -4.40844893e-01 -3.09327900e-01 -1.91193312e-01
-6.62743330e-01 7.74289429e-01 1.36732066e+00 -2.26150006e-01
-4.36107755e-01 4.56956387e-01 3.56598377e-01 1.99230254e-01
5.19347548e-01 1.12788045e+00 -5.90501308e-01 5.76979667e-02
-3.30324620e-01 2.23093957e-01 8.58950138e-01 9.32223439e-01
6.38324738e-01 2.31821030e-01 -4.64395620e-03 2.67391890e-01
3.77662331e-01 2.26127744e-01 -1.62357867e-01 -4.07762498e-01
7.98267066e-01 6.00869179e-01 -4.56079245e-01 -7.68516183e-01
-8.81503999e-01 -5.28027594e-01 -8.72275710e-01 -1.10467046e-01
1.64940178e-01 -2.35310122e-01 -1.27128434e+00 1.50278997e+00
9.37153175e-02 6.27116784e-02 1.23267576e-01 5.71576297e-01
1.73667479e+00 9.45144296e-01 4.73522782e-01 -2.13863894e-01
5.59715748e-01 -1.30469120e+00 -3.40759695e-01 8.27916563e-02
1.08435917e+00 -3.13697577e-01 5.78866065e-01 6.23722434e-01
-1.00114858e+00 -3.51700872e-01 -1.49370170e+00 -4.52527016e-01
-8.51655960e-01 -3.99169743e-01 1.04567003e+00 3.66817594e-01
-1.18050909e+00 1.45517123e+00 -4.95136768e-01 -1.15103222e-01
3.97784799e-01 8.78154516e-01 -6.61041260e-01 5.37650511e-02
-1.30361652e+00 1.11991370e+00 1.02235520e+00 7.94434547e-03
-1.16976368e+00 -9.43418801e-01 -7.82355130e-01 3.54380943e-02
4.36931789e-01 -6.68083608e-01 6.43790424e-01 -3.95155042e-01
-1.39804268e+00 7.45286107e-01 6.03904352e-02 -8.65979671e-01
2.07468137e-01 3.58175725e-01 -7.51318097e-01 5.74290380e-02
-2.18681216e-01 8.11311543e-01 3.20306838e-01 -7.33861208e-01
4.99509908e-02 -4.04553860e-01 2.01146845e-02 1.53764546e-01
2.91623145e-01 -4.68472928e-01 -4.27063137e-01 -1.25837764e-02
-1.85155421e-01 -9.23664510e-01 -5.56240201e-01 -9.98222232e-01
-9.68434274e-01 -2.42238477e-01 3.74717623e-01 -7.15505421e-01
1.08823001e+00 -1.45971966e+00 5.11130154e-01 4.22186702e-01
5.86789489e-01 3.22208434e-01 -5.81642985e-01 6.82441652e-01
-7.91324377e-01 2.94219851e-01 -1.20506935e-01 -5.87212369e-02
-3.38760555e-01 -3.28418672e-01 1.42184049e-01 4.84749466e-01
1.55124277e-01 1.31724668e+00 -7.68723249e-01 -2.50123233e-01
2.01005802e-01 5.97858846e-01 -9.02725101e-01 -3.87756526e-02
-4.83657837e-01 4.66342360e-01 -1.71468422e-01 6.34662032e-01
8.57743740e-01 -6.22558355e-01 5.43142557e-01 -3.95433962e-01
6.37963340e-02 7.43624628e-01 -4.26792353e-01 2.12137079e+00
6.18818998e-02 2.97243983e-01 -3.69098634e-01 -9.15470362e-01
6.89854264e-01 -1.16731105e-02 6.71414435e-01 -1.07553542e+00
1.38324529e-01 3.46664041e-02 5.36479175e-01 -4.86510880e-02
5.36412954e-01 -1.58406682e-02 2.51805633e-01 -3.52887362e-01
5.53931296e-01 -3.11315596e-01 3.83286774e-01 4.12022620e-01
1.22885060e+00 3.74410003e-01 4.07752752e-01 -5.22969007e-01
7.03657806e-01 1.71675131e-01 2.09686756e-01 4.54802901e-01
-7.42005631e-02 3.09505254e-01 7.51013100e-01 -8.52024853e-01
-8.10191691e-01 -6.60131514e-01 -2.25603983e-01 7.68980324e-01
8.17390904e-02 -1.35446882e+00 -6.29034162e-01 -8.65754306e-01
4.13018726e-02 7.60946631e-01 -4.87089723e-01 -6.49505436e-01
-3.99386555e-01 -1.03534627e+00 2.30590701e-01 1.60689816e-01
2.66869396e-01 -1.18885195e+00 4.22655791e-01 5.05386829e-01
3.06361467e-01 -1.13539672e+00 -6.04869723e-01 6.52631998e-01
-6.83074296e-01 -1.40682709e+00 2.66903755e-03 -5.68021774e-01
2.82288969e-01 -7.45287016e-02 1.40650976e+00 2.91764021e-01
-3.90450597e-01 -5.20055532e-01 -1.80511028e-02 -3.24558228e-01
-3.10138375e-01 7.15898752e-01 -1.14044636e-01 -6.08147264e-01
1.65880784e-01 -6.43127441e-01 -6.79533541e-01 -1.16547517e-01
-2.23292515e-01 2.48275563e-01 3.34792845e-02 7.32943773e-01
8.17849815e-01 -2.38304287e-01 3.40245694e-01 -9.71869111e-01
3.55860770e-01 -2.71881878e-01 -9.42596257e-01 1.81956172e-01
-1.06398678e+00 3.25861752e-01 1.05905640e+00 1.63254619e-01
-3.97356868e-01 1.72930047e-01 -4.82821018e-01 -1.81823224e-01
5.23554027e-01 8.91755283e-01 -2.16786772e-01 -7.34858394e-01
8.95240068e-01 -5.25475144e-02 -4.44095671e-01 -2.91809708e-01
4.62793320e-01 1.11394681e-01 4.40196693e-01 -5.85745752e-01
4.09039855e-01 -2.84651309e-01 8.55923235e-01 -8.24750245e-01
-8.73500884e-01 -2.17280313e-01 -5.30007005e-01 1.05152108e-01
1.07980919e+00 -1.02871239e+00 -1.43713307e+00 1.97120413e-01
-1.11706007e+00 -8.06768835e-01 2.12855741e-01 3.56109947e-01
-3.91201258e-01 4.56321776e-01 -6.84532881e-01 -1.99177861e-01
-8.12005877e-01 -1.44505680e+00 7.61146724e-01 6.08901605e-02
2.95618147e-01 -1.23217320e+00 1.95688661e-02 5.26144326e-01
1.96864128e-01 8.73518288e-01 1.11902237e+00 -6.48569345e-01
-6.52408957e-01 2.53825216e-03 -3.60816181e-01 1.36045635e-01
-1.26963824e-01 -7.66194090e-02 -8.23687077e-01 -6.07366145e-01
-7.85972297e-01 -6.95024252e-01 1.45155132e+00 5.92873454e-01
1.59435129e+00 -1.11272767e-01 -4.22481656e-01 1.34866452e+00
1.48271847e+00 2.61935741e-01 7.76054502e-01 -4.23108079e-02
1.52514029e+00 -1.37502402e-01 2.69717397e-03 1.31875207e-03
2.43262291e-01 6.62567198e-01 1.00363410e+00 -1.34318769e-01
-1.08724982e-01 -6.84958279e-01 2.68409044e-01 7.93886602e-01
-6.61975622e-01 -7.11571872e-01 -9.57823157e-01 -4.29651648e-01
-1.67467856e+00 -6.94364130e-01 -5.42506456e-01 1.90776968e+00
6.49568141e-01 2.79206514e-01 -4.05303203e-02 -4.98172253e-01
4.84433264e-01 4.76412982e-01 -8.82284701e-01 -6.78245008e-01
-1.91062123e-01 8.85686040e-01 1.15639746e+00 8.61209154e-01
-1.08500695e+00 1.53274608e+00 6.71015310e+00 1.03253448e+00
-1.16680682e+00 -1.68639928e-01 7.56969512e-01 -3.55650894e-02
-9.26051363e-02 1.36396497e-01 -7.97536492e-01 3.44013512e-01
1.64635754e+00 -1.27144292e-01 9.12587762e-01 6.18766129e-01
-4.82508205e-02 2.56731331e-01 -1.16207945e+00 1.14384055e+00
-3.42739075e-02 -2.08083582e+00 2.08613947e-01 5.81262529e-01
8.75008225e-01 6.36011362e-01 -1.05925925e-01 6.91132367e-01
3.85585845e-01 -1.92119169e+00 2.38400966e-01 4.49824631e-01
1.20452249e+00 -1.10229838e+00 4.75082099e-01 2.61337887e-02
-1.31883943e+00 5.70039630e-01 -6.34811580e-01 -5.62452804e-03
-6.54575452e-02 1.51990920e-01 -7.65605688e-01 9.94041979e-01
2.52098292e-01 1.29923320e+00 -6.49908364e-01 8.28487098e-01
-2.57092416e-01 5.60380995e-01 -9.53139886e-02 -1.81079254e-01
5.46711564e-01 -7.72815228e-01 2.62963712e-01 9.54111338e-01
-2.13262327e-02 8.05104300e-02 1.78497121e-01 1.01651716e+00
-9.69220638e-01 1.14151746e-01 -6.79182410e-01 -4.91996706e-01
1.59878835e-01 1.35724127e+00 -3.56166035e-01 -2.42408946e-01
9.13067013e-02 9.64407206e-01 7.42411792e-01 3.23158115e-01
-1.11540544e+00 -3.87893617e-01 2.65298516e-01 -4.83094826e-02
2.15021938e-01 -5.23562506e-02 2.06701979e-01 -9.14835036e-01
-5.54423213e-01 -9.68774021e-01 2.59436607e-01 -8.35662544e-01
-1.18402982e+00 5.66727519e-01 -4.99579519e-01 -3.19007277e-01
3.01778018e-01 -1.36725676e+00 -6.38653278e-01 1.19523537e+00
-1.52373993e+00 -9.39037800e-01 -5.02853394e-02 3.14894944e-01
-1.26860246e-01 -3.82268220e-01 9.40833032e-01 1.89541563e-01
-9.57971931e-01 6.39718115e-01 2.02639356e-01 -3.68806601e-01
6.16400123e-01 -1.57149220e+00 9.86957490e-01 3.09933990e-01
-2.16706283e-02 6.07215464e-01 4.92387980e-01 -8.29102814e-01
-1.96589863e+00 -1.35729313e+00 6.31299794e-01 -6.68464780e-01
8.09917331e-01 -3.94333273e-01 -7.76610494e-01 3.75157803e-01
1.98295206e-01 2.99265683e-01 5.02294183e-01 1.48765340e-01
-3.53576362e-01 2.16491312e-01 -9.65872347e-01 2.81882942e-01
1.58730292e+00 -7.10925817e-01 1.02085903e-01 9.30276334e-01
1.17035949e+00 -8.69119823e-01 -1.34496689e+00 4.58256841e-01
1.04366973e-01 -7.80419350e-01 1.14727318e+00 -1.45428586e+00
4.10516620e-01 -9.79185551e-02 -1.01738468e-01 -1.45249557e+00
-4.02033210e-01 -9.45543945e-01 -2.15448245e-01 3.45994473e-01
1.02180719e+00 -7.14003861e-01 8.91091883e-01 5.91303147e-02
-6.42014921e-01 -9.07697022e-01 -9.54056382e-01 -8.18151534e-01
4.13931310e-01 -2.20387369e-01 5.71909904e-01 8.28348696e-01
-1.45981699e-01 1.06679273e+00 -3.58190954e-01 -8.02698806e-02
4.09914523e-01 -2.61465102e-01 6.69223070e-01 -1.14722383e+00
-2.93335974e-01 -4.06833231e-01 -4.10800368e-01 -7.25338519e-01
5.25416672e-01 -1.71716619e+00 -5.68785250e-01 -1.64820981e+00
4.87221479e-01 2.18701869e-01 -6.16098344e-01 5.31377316e-01
-1.00457639e-01 7.35409558e-02 1.51164308e-01 -2.48419270e-01
-8.41610789e-01 6.80521727e-01 1.57037604e+00 -5.70437014e-01
-3.01674932e-01 -4.36433792e-01 -7.74514437e-01 1.35762751e-01
6.49886489e-01 -3.70221913e-01 -2.75655985e-01 2.67234504e-01
5.43933451e-01 1.02031320e-01 2.30020493e-01 -9.40924287e-01
-1.91570744e-01 -4.34973612e-02 4.19766396e-01 -8.10533941e-01
5.03453732e-01 -3.71719033e-01 3.15960646e-01 7.63584733e-01
-2.31051996e-01 1.13997258e-01 4.68205869e-01 5.72479129e-01
2.54650384e-01 1.35714456e-01 6.71800911e-01 -1.13721669e-01
-4.76387262e-01 1.04232991e+00 2.61683822e-01 6.30717725e-02
6.46697223e-01 7.07003698e-02 -8.12346816e-01 -1.25856876e-01
-7.70548105e-01 2.81985670e-01 4.59534168e-01 -7.93894157e-02
3.82336140e-01 -1.13596547e+00 -3.63189936e-01 -2.46268839e-01
2.15722382e-01 -1.38876766e-01 1.21968068e-01 7.30769336e-01
-9.86288190e-01 9.26125407e-01 -3.19009386e-02 -1.84858367e-01
-1.16112590e+00 6.92097008e-01 8.57950628e-01 -6.06787860e-01
-3.49770695e-01 9.61713850e-01 -1.18747450e-01 -4.92886722e-01
-1.19711608e-01 -2.70693988e-01 -3.71922441e-02 -4.19084191e-01
-5.84015511e-02 3.32476735e-01 5.16939461e-01 -7.51554668e-01
-6.81776762e-01 3.41842115e-01 -2.08595902e-01 6.30023003e-01
1.52470577e+00 7.30323255e-01 -2.54030794e-01 -2.46059075e-01
1.62441885e+00 -3.24812561e-01 -1.18499696e+00 7.17529878e-02
-8.15031826e-02 4.04743701e-01 4.99775827e-01 -1.12432718e+00
-1.04555678e+00 9.64402080e-01 6.13903761e-01 -1.29210725e-01
4.46197093e-01 -1.94155544e-01 6.37916148e-01 9.92787778e-01
2.19872534e-01 -9.85788703e-01 -1.23922359e-02 8.76625121e-01
9.09915745e-01 -1.34195697e+00 4.64703441e-01 -3.60518724e-01
-5.50807238e-01 1.20649660e+00 7.63780892e-01 -2.67310180e-02
4.27946210e-01 -5.16669631e-01 -6.47186935e-01 -9.54798460e-01
-7.73149669e-01 -1.72207862e-01 8.78454447e-01 5.94587982e-01
7.44147122e-01 4.29820955e-01 -3.41154598e-02 1.56344205e-01
-5.65175354e-01 -4.24793363e-01 3.20148438e-01 4.88455623e-01
-2.63633907e-01 -1.03383410e+00 4.46676433e-01 4.00690049e-01
-2.45527118e-01 -8.09857368e-01 -1.00933814e+00 9.38637972e-01
-2.95314901e-02 9.77330506e-01 -4.53515649e-01 -6.40024364e-01
4.34865087e-01 -4.55915369e-02 9.06752288e-01 -6.71109736e-01
-6.98865533e-01 -2.34157592e-01 4.70986128e-01 -7.70683885e-01
-1.42965659e-01 1.65231183e-01 -1.58308804e+00 -8.16802800e-01
-4.67251271e-01 3.84362668e-01 5.57267606e-01 5.78870833e-01
4.52157110e-01 8.66544127e-01 4.61511672e-01 -7.76358247e-01
-2.17286259e-01 -1.04522789e+00 -5.40906131e-01 1.43937171e-01
8.50348920e-02 -4.90222663e-01 -6.62557408e-02 -6.63737595e-01] | [5.195156574249268, 5.864029407501221] |
58c34c9d-38be-4e30-ad1d-95c52d7bc940 | orrn-an-ode-based-recursive-registration | 2305.14673 | null | https://arxiv.org/abs/2305.14673v2 | https://arxiv.org/pdf/2305.14673v2.pdf | ORRN: An ODE-based Recursive Registration Network for Deformable Respiratory Motion Estimation with Lung 4DCT Images | Deformable Image Registration (DIR) plays a significant role in quantifying deformation in medical data. Recent Deep Learning methods have shown promising accuracy and speedup for registering a pair of medical images. However, in 4D (3D + time) medical data, organ motion, such as respiratory motion and heart beating, can not be effectively modeled by pair-wise methods as they were optimized for image pairs but did not consider the organ motion patterns necessary when considering 4D data. This paper presents ORRN, an Ordinary Differential Equations (ODE)-based recursive image registration network. Our network learns to estimate time-varying voxel velocities for an ODE that models deformation in 4D image data. It adopts a recursive registration strategy to progressively estimate a deformation field through ODE integration of voxel velocities. We evaluate the proposed method on two publicly available lung 4DCT datasets, DIRLab and CREATIS, for two tasks: 1) registering all images to the extreme inhale image for 3D+t deformation tracking and 2) registering extreme exhale to inhale phase images. Our method outperforms other learning-based methods in both tasks, producing the smallest Target Registration Error of 1.24mm and 1.26mm, respectively. Additionally, it produces less than 0.001\% unrealistic image folding, and the computation speed is less than 1 second for each CT volume. ORRN demonstrates promising registration accuracy, deformation plausibility, and computation efficiency on group-wise and pair-wise registration tasks. It has significant implications in enabling fast and accurate respiratory motion estimation for treatment planning in radiation therapy or robot motion planning in thoracic needle insertion. | ['Fei Liu', 'Michael Yip', 'Dimitri Schreiber', 'Shan Lin', 'Xiao Liang'] | 2023-05-24 | null | null | null | null | ['image-registration', 'motion-estimation', 'motion-planning'] | ['computer-vision', 'computer-vision', 'robots'] | [ 8.45699236e-02 1.94025904e-01 -2.95375675e-01 -1.18884243e-01
-9.66640592e-01 -7.20292926e-01 3.05342764e-01 1.20667212e-01
-5.70878208e-01 4.11370486e-01 4.70446721e-02 -4.66216177e-01
-3.73728245e-01 -5.00654459e-01 -5.42150855e-01 -9.39406693e-01
-2.69198775e-01 9.21354175e-01 2.56429821e-01 1.07046187e-01
-2.86824673e-01 9.15504992e-01 -2.36115471e-01 6.54548034e-02
6.42638564e-01 8.41114581e-01 2.66901225e-01 9.73984480e-01
1.42325789e-01 6.59674525e-01 -9.35783014e-02 8.15180317e-02
2.68162429e-01 -3.90986681e-01 -1.07927966e+00 -2.67517745e-01
5.06161332e-01 -4.58411753e-01 -4.80673164e-01 5.64968944e-01
1.00362515e+00 1.07405208e-01 6.26721263e-01 -6.96389556e-01
-2.57482111e-01 3.90500247e-01 -4.97568429e-01 3.42705250e-01
2.97432821e-02 2.54934132e-01 1.35570556e-01 -5.87078273e-01
7.43088782e-01 5.06986558e-01 1.44598877e+00 8.36031377e-01
-9.79008973e-01 -2.77907640e-01 -5.49388826e-01 -4.88567889e-01
-1.03952432e+00 1.11736231e-01 5.71266890e-01 -7.02046931e-01
9.41705108e-01 6.44117057e-01 6.54149652e-01 5.79067647e-01
1.04428589e+00 2.48312488e-01 8.57864678e-01 6.25388622e-02
-2.00460613e-01 -4.08169419e-01 -7.58702680e-02 7.89315879e-01
1.50112942e-01 2.05267549e-01 1.96864039e-01 -2.84030288e-01
1.31416500e+00 1.31104141e-01 -5.98819137e-01 -2.66654909e-01
-1.67208087e+00 5.60144365e-01 1.04360986e+00 6.20524228e-01
-5.56148648e-01 6.14924073e-01 4.20693785e-01 -1.42889589e-01
4.64165509e-01 3.92945439e-01 -2.82487512e-01 -9.32179317e-02
-6.71244383e-01 1.09037086e-01 5.98619759e-01 5.96019387e-01
9.48650911e-02 -1.18830958e-02 -5.47460496e-01 4.92191672e-01
4.01814371e-01 4.52289134e-01 8.10709059e-01 -9.93910372e-01
3.69327277e-01 4.40615296e-01 4.47193626e-03 -6.32425368e-01
-1.05786109e+00 -3.40745956e-01 -1.35028756e+00 -7.99771026e-02
4.53660667e-01 -1.88411891e-01 -1.14331591e+00 1.46259427e+00
8.46219718e-01 3.30614269e-01 -3.64111103e-02 1.11351442e+00
1.25161302e+00 2.75095850e-01 1.11878976e-01 -4.71472412e-01
1.34431183e+00 -7.59203672e-01 -7.51717687e-01 1.25592962e-01
1.18444312e+00 -8.27497363e-01 6.26406968e-01 -2.29240790e-01
-1.60535610e+00 -4.04006779e-01 -5.30579329e-01 6.08677650e-03
3.03703964e-01 -1.89785361e-01 5.31783760e-01 5.19006252e-01
-1.13108110e+00 1.22329473e+00 -1.41368783e+00 1.79596439e-01
5.71676731e-01 9.17058885e-01 -2.52183139e-01 4.28803600e-02
-1.02319705e+00 1.06444931e+00 -6.39780685e-02 3.63697320e-01
-7.57839382e-01 -1.47647989e+00 -7.26387143e-01 -5.55811167e-01
-5.86380139e-02 -1.09730601e+00 1.35270679e+00 -2.81762153e-01
-1.36716199e+00 1.09361303e+00 -3.13245282e-02 -3.60227466e-01
8.86348903e-01 5.52054383e-02 -3.99207622e-02 5.66530675e-02
-2.08205640e-01 4.39571023e-01 3.27799708e-01 -1.02531326e+00
1.84103921e-01 -4.55705553e-01 -4.10209149e-01 4.32812124e-01
8.69379938e-02 -1.57687701e-02 -4.48454231e-01 -8.01633179e-01
4.25448835e-01 -1.40136302e+00 -5.87406576e-01 4.76546228e-01
-1.82464212e-01 6.52269050e-02 6.50828779e-01 -8.68473411e-01
7.98110664e-01 -1.59802377e+00 1.34587377e-01 9.47172102e-03
6.06127739e-01 2.65208364e-01 2.23225743e-01 -2.65538901e-01
-2.92045623e-01 2.17677906e-01 -5.97149730e-01 -1.31891444e-01
-5.20973444e-01 3.44567120e-01 2.61143804e-01 8.61708403e-01
-3.53715897e-01 1.60759795e+00 -1.00043666e+00 -7.06656218e-01
2.45290771e-01 8.34602654e-01 -4.30351496e-01 3.46040428e-01
2.17468500e-01 1.22453666e+00 -4.80547667e-01 3.18834871e-01
8.79210293e-01 -2.65407234e-01 2.94397086e-01 -6.07276499e-01
-1.38249531e-01 1.24899119e-01 -5.93366802e-01 1.91636074e+00
-7.19999373e-01 3.85231376e-01 9.86785665e-02 -7.00711131e-01
5.17252624e-01 8.42495084e-01 1.59548175e+00 -5.92168868e-01
4.27747875e-01 3.44438642e-01 2.69694805e-01 -7.11431861e-01
4.39043231e-02 -5.51715016e-01 2.23968122e-02 5.51804841e-01
-2.80635983e-01 -7.89235055e-01 -5.07605791e-01 -3.23662162e-01
1.17348564e+00 -4.79651578e-02 -1.86929349e-02 -3.37630898e-01
4.40609962e-01 1.07841276e-01 3.98543447e-01 5.02726674e-01
-3.15485626e-01 8.54925931e-01 -3.58793815e-03 -7.28207648e-01
-9.68984902e-01 -1.10629058e+00 -5.31464756e-01 1.71301872e-01
3.49865496e-01 2.41363108e-01 -8.13620925e-01 -7.27359593e-01
-1.51523817e-02 2.38101840e-01 -5.08507013e-01 -1.04029916e-01
-1.50026631e+00 -1.04907084e+00 6.37645006e-01 7.39628434e-01
2.16383308e-01 -1.02434397e+00 -6.37939215e-01 3.64957839e-01
-3.16105545e-01 -1.14749467e+00 -8.98069263e-01 1.08546324e-01
-1.59443021e+00 -1.02350616e+00 -1.04348564e+00 -8.93275678e-01
9.32828784e-01 5.57060763e-02 1.08007419e+00 2.91940302e-01
-5.80019355e-01 4.87994373e-01 1.35106906e-01 1.84470173e-02
-8.38842750e-01 -5.45148775e-02 8.70949961e-03 -6.59071386e-01
-6.17541611e-01 -3.50209564e-01 -9.48735654e-01 5.83732486e-01
-6.73517406e-01 1.68298274e-01 1.25736728e-01 7.35628128e-01
1.03102589e+00 -1.79823801e-01 -3.90757509e-02 -8.44883561e-01
5.94182968e-01 -3.26837063e-01 -2.97534794e-01 2.56556153e-01
-5.01409769e-01 -4.44480777e-02 4.53510910e-01 -6.37905955e-01
-8.49346280e-01 2.24117115e-01 -4.40757483e-01 -6.29660547e-01
1.86648861e-01 1.86568528e-01 6.86367631e-01 -7.10748613e-01
7.63742924e-01 1.43172666e-01 2.62680620e-01 -1.41774908e-01
9.05431584e-02 1.93911925e-01 6.76739514e-01 -5.33845246e-01
8.97373021e-01 5.92799008e-01 5.04243076e-01 -3.73563766e-01
-3.76827300e-01 -3.75748932e-01 -1.05459571e+00 -3.14906508e-01
1.25392997e+00 -5.93724191e-01 -1.00754070e+00 5.97981215e-01
-1.16786289e+00 -7.49997258e-01 -7.77696729e-01 8.84411812e-01
-6.85945213e-01 4.08725470e-01 -9.75640237e-01 -1.52162135e-01
-9.83205914e-01 -1.54353380e+00 9.81612980e-01 9.87497270e-02
-3.57713699e-01 -1.58083975e+00 3.19789827e-01 3.07455629e-01
7.68293560e-01 8.84670198e-01 7.44862854e-01 -3.84885877e-01
-5.07966042e-01 -2.60569066e-01 2.13705506e-02 -7.99584463e-02
4.88958329e-01 -4.19414252e-01 -5.36776781e-01 -4.80085522e-01
5.10178924e-01 -1.92473903e-02 3.75102341e-01 8.86278093e-01
1.56076717e+00 -2.01048091e-01 -4.57888693e-01 1.05770767e+00
1.26442242e+00 4.66743529e-01 4.07715917e-01 -2.44126990e-01
1.25792980e+00 2.30119169e-01 3.95486921e-01 2.21600920e-01
3.25617284e-01 8.95313323e-01 5.98887146e-01 -5.62489033e-01
-4.81669247e-01 2.72296548e-01 -8.01544264e-02 1.21917856e+00
-4.81921107e-01 -5.30080386e-02 -1.25634003e+00 2.82650441e-01
-1.47565663e+00 -5.67868710e-01 -7.22640812e-01 2.25969195e+00
1.01034808e+00 -3.07956666e-01 -2.39858359e-01 -4.14970666e-01
5.33253849e-01 -2.52611283e-02 -6.68341398e-01 -2.73859560e-01
5.11420071e-01 5.30653119e-01 7.20632017e-01 7.05428123e-01
-1.00192368e+00 3.09497863e-01 6.17235708e+00 4.32527035e-01
-1.38131607e+00 5.86203575e-01 7.09626913e-01 4.32730652e-02
-4.19361562e-01 -2.96558321e-01 -4.14675176e-01 2.77919292e-01
6.90431714e-01 2.19512526e-02 3.09165448e-01 3.08718473e-01
3.02978992e-01 8.92343521e-02 -1.13967848e+00 8.82843018e-01
-1.93831444e-01 -1.63756323e+00 -4.59853381e-01 1.14890315e-01
7.42524505e-01 3.34312499e-01 -4.08285856e-02 -1.39282182e-01
3.98872420e-02 -1.29128468e+00 1.60220459e-01 6.08255029e-01
1.12445128e+00 -3.07724833e-01 6.28022373e-01 4.68134373e-01
-1.32719481e+00 6.19858921e-01 -1.68469563e-01 6.22411072e-01
3.89151275e-01 3.75471145e-01 -1.19576919e+00 8.06899011e-01
4.13942695e-01 6.06610417e-01 -9.24848393e-02 9.66146231e-01
1.82279259e-01 2.42968842e-01 -3.48849356e-01 4.31579679e-01
8.66020918e-02 -1.67196408e-01 5.54506123e-01 1.00002611e+00
5.08309662e-01 4.02445108e-01 1.49600219e-03 7.85614669e-01
-2.87561893e-01 -2.62664974e-01 -4.05717641e-01 4.79030401e-01
6.09472059e-02 1.42438591e+00 -8.18651438e-01 -1.52583405e-01
8.33322108e-02 8.59201491e-01 -1.33111879e-01 -3.29566337e-02
-1.26756406e+00 2.40762278e-01 2.90143728e-01 4.55598205e-01
-2.75877297e-01 -2.52176762e-01 -3.13635081e-01 -8.29975307e-01
-1.54530451e-01 -2.64036983e-01 3.82496029e-01 -5.54675460e-01
-1.05637896e+00 7.32066929e-01 3.07660475e-02 -1.22745872e+00
-3.63385975e-01 -4.98567343e-01 -8.48418832e-01 9.80531156e-01
-1.30239439e+00 -9.43271339e-01 -7.47089207e-01 6.15851581e-01
2.26498157e-01 4.95348334e-01 7.64266312e-01 5.26421964e-01
-3.01795155e-01 5.54644525e-01 1.02499977e-03 2.10313410e-01
5.23166716e-01 -1.35991168e+00 5.19995868e-01 2.31848359e-01
-4.14199829e-01 3.07059765e-01 1.29407108e-01 -8.00177991e-01
-1.75383651e+00 -1.30773699e+00 6.69009566e-01 -8.05214465e-01
3.25583756e-01 1.37077838e-01 -1.02715135e+00 5.30169725e-01
-2.34947558e-02 1.07059717e+00 3.48186642e-01 -8.17390978e-01
4.96046036e-01 3.27435136e-02 -1.53712761e+00 2.19666228e-01
1.14597666e+00 -3.12300295e-01 -2.14203089e-01 8.49655032e-01
8.58691216e-01 -1.69813859e+00 -1.84302354e+00 9.35466945e-01
4.52599257e-01 -3.85972500e-01 1.47617257e+00 -2.84238696e-01
1.44714683e-01 -1.93627983e-01 4.91530031e-01 -1.03162205e+00
-2.56790638e-01 -8.53252411e-01 -1.30049437e-01 5.42749882e-01
1.55215468e-02 -7.39312351e-01 8.66916895e-01 9.25767183e-01
-5.49872458e-01 -1.23206258e+00 -1.33556747e+00 -5.69299817e-01
5.67245841e-01 -4.17852312e-01 2.76989311e-01 1.29977286e+00
-5.85631192e-01 -4.82110679e-01 5.60959019e-02 8.02654475e-02
5.89730859e-01 1.21830879e-02 3.47810447e-01 -8.47982526e-01
-1.66050002e-01 -3.36661935e-01 1.35213677e-02 -9.08691764e-01
-1.46080360e-01 -1.45326817e+00 9.96401906e-02 -1.78694701e+00
1.57827392e-01 -9.88585472e-01 -1.83819413e-01 4.04893368e-01
-1.09665357e-01 3.54423463e-01 -1.55649940e-02 5.68738461e-01
9.44295824e-02 1.26241043e-01 2.28730726e+00 -3.94340083e-02
-2.71617532e-01 2.22170144e-01 1.37158309e-03 6.77665174e-01
7.48226881e-01 -7.92739987e-01 -1.49611637e-01 -4.60764110e-01
-1.29117936e-01 7.80530155e-01 5.74498415e-01 -8.47416461e-01
3.08485955e-01 -5.05861901e-02 5.07788241e-01 -5.85462749e-01
8.71425718e-02 -9.34915602e-01 6.93413079e-01 1.14547300e+00
-2.36134708e-01 3.07044894e-01 4.70243871e-01 1.47988843e-02
5.32204993e-02 -1.98316589e-01 1.02151406e+00 -2.43073598e-01
8.39229226e-02 9.97233629e-01 -1.75459296e-01 3.58162403e-01
1.01308382e+00 -3.43187690e-01 -3.47353816e-02 1.21423222e-01
-9.04741406e-01 1.08968103e-02 2.40631014e-01 1.13387704e-01
7.25992262e-01 -1.58918607e+00 -6.35176241e-01 -7.97308236e-02
-4.93478864e-01 7.21120954e-01 6.14042640e-01 1.58322275e+00
-1.09328330e+00 2.37200469e-01 2.13813871e-01 -1.08437729e+00
-1.13214540e+00 2.05548629e-01 1.14835131e+00 -9.13152874e-01
-7.59351730e-01 8.96528423e-01 3.92035335e-01 -7.47027397e-01
-1.83598727e-01 -7.40432620e-01 1.07443236e-01 -4.60945278e-01
-1.62455142e-01 2.24144518e-01 3.33698332e-01 -8.18320930e-01
-6.58354580e-01 1.29184961e+00 1.08967476e-01 3.22718352e-01
1.16439581e+00 1.59398124e-01 -1.25694081e-01 5.35683818e-02
1.39812028e+00 -3.01963478e-01 -1.33481145e+00 -1.33842349e-01
-3.65814984e-01 -1.98700279e-01 3.26902837e-01 -7.32373893e-01
-1.44973445e+00 7.77894020e-01 9.71690059e-01 -9.66062695e-02
9.98270512e-01 3.99834067e-02 1.21076500e+00 -1.34849980e-01
-2.80386265e-02 -3.92316043e-01 1.47451222e-01 4.11189765e-01
9.78476882e-01 -1.16499710e+00 2.60339826e-01 -5.35213947e-01
-5.23788452e-01 1.21106517e+00 5.42137742e-01 -5.24632782e-02
5.91921091e-01 6.36308312e-01 1.17072515e-01 -3.22953463e-01
-2.25140929e-01 4.74594951e-01 8.25514555e-01 4.98354554e-01
8.71182084e-01 1.58352535e-02 -1.59166574e-01 5.25012612e-02
1.20167829e-01 -4.95605916e-02 1.13407530e-01 9.09863114e-01
1.13519788e-01 -1.07725441e+00 -3.76723468e-01 2.45007858e-01
-6.07229888e-01 1.15389757e-01 1.26843497e-01 9.82823789e-01
4.57912423e-02 5.33818118e-02 1.13311701e-01 4.31742035e-02
4.14363414e-01 -4.04530764e-01 1.01126575e+00 -4.76140857e-01
-1.32656467e+00 4.31187935e-02 -3.91193777e-01 -6.90384626e-01
-4.51187313e-01 -4.55063730e-01 -1.85363543e+00 -1.74488947e-01
-2.17409730e-01 -1.61931589e-01 8.27280939e-01 6.71186626e-01
1.76077798e-01 9.10107076e-01 7.60610938e-01 -1.18270206e+00
-4.97348219e-01 -3.92173707e-01 -3.84877846e-02 4.83936936e-01
3.39657813e-01 -2.87736684e-01 -1.66556314e-01 -1.69389591e-01] | [13.925491333007812, -2.586927652359009] |
f346a90a-1c2f-48ae-ae6d-38d67fbb5098 | elixir-a-system-to-enhance-data-quality-for | 2212.04061 | null | https://arxiv.org/abs/2212.04061v1 | https://arxiv.org/pdf/2212.04061v1.pdf | Elixir: A system to enhance data quality for multiple analytics on a video stream | IoT sensors, especially video cameras, are ubiquitously deployed around the world to perform a variety of computer vision tasks in several verticals including retail, healthcare, safety and security, transportation, manufacturing, etc. To amortize their high deployment effort and cost, it is desirable to perform multiple video analytics tasks, which we refer to as Analytical Units (AUs), off the video feed coming out of every camera. In this paper, we first show that in a multi-AU setting, changing the camera setting has disproportionate impact on different AUs performance. In particular, the optimal setting for one AU may severely degrade the performance for another AU, and further the impact on different AUs varies as the environmental condition changes. We then present Elixir, a system to enhance the video stream quality for multiple analytics on a video stream. Elixir leverages Multi-Objective Reinforcement Learning (MORL), where the RL agent caters to the objectives from different AUs and adjusts the camera setting to simultaneously enhance the performance of all AUs. To define the multiple objectives in MORL, we develop new AU-specific quality estimator values for each individual AU. We evaluate Elixir through real-world experiments on a testbed with three cameras deployed next to each other (overlooking a large enterprise parking lot) running Elixir and two baseline approaches, respectively. Elixir correctly detects 7.1% (22,068) and 5.0% (15,731) more cars, 94% (551) and 72% (478) more faces, and 670.4% (4975) and 158.6% (3507) more persons than the default-setting and time-sharing approaches, respectively. It also detects 115 license plates, far more than the time-sharing approach (7) and the default setting (0). | ['Srimat T. Chakradhar', 'Y. Charlie Hu', 'Oliver Po', 'Murugan Sankaradas', 'Giuseppe Coviello', 'Kunal Rao', 'Sibendu Paul'] | 2022-12-08 | null | null | null | null | ['multi-objective-reinforcement-learning'] | ['methodology'] | [-1.96577087e-02 -2.27503777e-01 -7.06533156e-03 2.53716372e-02
-8.27312291e-01 -6.81693316e-01 1.61145091e-01 -2.63468713e-01
-5.71094453e-01 5.39667189e-01 -1.66503862e-01 -7.05831647e-02
-8.10184702e-02 -7.80745924e-01 -9.55392182e-01 -8.14892530e-01
-1.08773828e-01 2.00918213e-01 3.82268488e-01 1.73304826e-02
-2.76164591e-01 2.87705332e-01 -1.67053056e+00 -1.07610956e-01
5.34236014e-01 1.30855131e+00 3.24914366e-01 9.23052371e-01
4.92808849e-01 7.91300654e-01 -8.78726602e-01 -4.58701193e-01
5.32290518e-01 1.35373110e-02 1.35890931e-01 3.10769081e-01
2.56683916e-01 -7.27809846e-01 -2.93772161e-01 9.76106524e-01
4.04169261e-01 1.15086056e-01 4.11160499e-01 -1.92872703e+00
-3.75210255e-01 3.12554151e-01 -1.01755726e+00 2.36593649e-01
2.56184608e-01 6.99538231e-01 8.14713478e-01 -2.88870096e-01
5.70890196e-02 1.21876848e+00 3.68294597e-01 4.12886441e-01
-7.79496968e-01 -1.03045130e+00 3.05622876e-01 6.45262897e-02
-1.19102001e+00 -6.00250304e-01 3.71262759e-01 -2.46188074e-01
6.79519057e-01 1.17097117e-01 4.85452205e-01 1.07524633e+00
1.31728858e-01 6.11968696e-01 7.83808172e-01 3.54298726e-02
3.55660707e-01 6.74401000e-02 -1.46967709e-01 5.04183173e-01
4.34085727e-01 1.36751488e-01 -3.97895396e-01 -1.55610740e-02
6.36796057e-01 2.87118763e-01 -4.70489264e-02 3.23621109e-02
-1.07748723e+00 6.83037698e-01 -3.19877863e-02 -4.48886812e-01
-6.05726361e-01 5.09241819e-01 1.23452924e-01 2.51063675e-01
-6.62769228e-02 2.39094943e-01 -5.00003219e-01 -4.19254750e-01
-4.19187933e-01 6.52083382e-02 4.95127201e-01 1.28335631e+00
7.37181485e-01 1.39617756e-01 -3.56931537e-01 5.86366415e-01
4.45787907e-01 1.22619665e+00 1.09354198e-01 -1.58996093e+00
7.26436734e-01 2.92701453e-01 4.10482734e-01 -7.11149693e-01
-3.29325974e-01 -1.90106198e-01 -5.77680707e-01 2.02771142e-01
2.38684535e-01 -8.29302311e-01 -4.92628932e-01 1.74679089e+00
1.82215869e-01 3.73482645e-01 1.05527773e-01 8.83182883e-01
4.17582661e-01 8.86500001e-01 1.77329227e-01 -3.65380347e-01
1.44687915e+00 -6.93938136e-01 -4.46615487e-01 -3.38695198e-01
8.62335935e-02 -6.44938707e-01 8.34808111e-01 5.92334986e-01
-1.08846736e+00 -4.77243721e-01 -1.02258813e+00 7.72799075e-01
-6.93378970e-02 1.73430666e-01 1.44189030e-01 7.34057844e-01
-9.12370026e-01 1.65412635e-01 -6.18472457e-01 -3.65699470e-01
3.73132467e-01 5.33624589e-01 4.86236513e-02 -3.08739871e-01
-8.03150654e-01 3.25980246e-01 -1.32809788e-01 -3.14587146e-01
-1.00930011e+00 -7.09058762e-01 -6.21257186e-01 9.55739804e-03
8.73938382e-01 -4.76067096e-01 1.35140955e+00 -8.64014626e-01
-1.57837939e+00 3.10684472e-01 1.85343653e-01 -4.88320708e-01
4.61432785e-01 -1.94806993e-01 -7.28637397e-01 2.86258250e-01
2.02343971e-01 8.22946668e-01 1.01252794e+00 -1.11560917e+00
-1.11793196e+00 -3.84451598e-01 3.79412204e-01 2.66765058e-01
-3.00591528e-01 1.52309254e-01 -7.56881654e-01 -3.25013667e-01
-7.19496250e-01 -1.11286056e+00 -1.18276671e-01 -5.65011166e-02
-1.88854024e-01 -8.90398473e-02 1.01734865e+00 -3.96289140e-01
9.28042710e-01 -2.43268442e+00 -3.23410481e-01 3.15317027e-02
1.91007853e-01 1.92022562e-01 -1.47714064e-01 1.53648260e-03
5.07789731e-01 -8.56185183e-02 1.76108241e-01 -4.41629946e-01
8.62632915e-02 3.04853767e-01 1.97684020e-01 4.52656686e-01
1.57570466e-01 5.26299953e-01 -8.85653913e-01 -2.05864489e-01
5.48002124e-01 2.77424157e-01 -5.17838299e-01 3.06406230e-01
-5.33757918e-03 9.46179777e-02 -3.41454566e-01 1.01000834e+00
6.91056967e-01 -2.23616511e-01 8.00255015e-02 -1.74904287e-01
-4.42319885e-02 -4.54496592e-01 -1.46359634e+00 9.59012568e-01
-5.15660524e-01 6.25222147e-01 4.13686544e-01 -6.17577136e-01
5.53732812e-01 2.51800746e-01 8.08045387e-01 -1.00324619e+00
2.56303340e-01 -1.22106202e-01 -1.97106421e-01 -6.30582869e-01
4.14525658e-01 2.07309172e-01 -2.76772112e-01 4.31821227e-01
-2.97131032e-01 5.18519044e-01 2.04312652e-01 1.88350946e-01
1.66922748e+00 -3.41193318e-01 2.13416189e-01 1.68027043e-01
2.23383516e-01 -3.42204988e-01 7.06237137e-01 8.39211166e-01
-7.76529014e-01 2.14599907e-01 2.42092893e-01 -1.92344710e-02
-9.13727701e-01 -1.42727280e+00 3.35223496e-01 1.10685861e+00
5.72705090e-01 -1.50573447e-01 -6.30859673e-01 -5.14221370e-01
2.42657423e-01 6.19211495e-01 -1.37253374e-01 -2.56283551e-01
-2.98568696e-01 -5.91834009e-01 4.20328021e-01 4.88676727e-01
8.64742577e-01 -8.97854626e-01 -1.10745132e+00 2.03791812e-01
-1.42062575e-01 -1.64592946e+00 -7.37225592e-01 -1.63711786e-01
-1.75741941e-01 -1.07806087e+00 -3.20504189e-01 3.48904729e-02
2.67425656e-01 8.88296664e-01 1.04763627e+00 -2.52849728e-01
-1.33389160e-01 1.00127089e+00 -4.21848506e-01 -6.15161598e-01
-2.66795605e-01 -2.83158690e-01 3.18605393e-01 3.78513694e-01
5.15751243e-01 -2.16000602e-01 -5.97321868e-01 6.44663870e-01
-7.26134777e-01 -4.51813579e-01 7.12341726e-01 2.06848562e-01
4.88567919e-01 4.10521537e-01 5.83298922e-01 -2.34009072e-01
4.22715962e-01 -7.87771821e-01 -1.00205350e+00 9.43343118e-02
-4.66059774e-01 -1.26280800e-01 6.23737812e-01 -5.75138986e-01
-7.62279689e-01 -7.36320857e-04 2.44247705e-01 -9.25553024e-01
-2.80215859e-01 -1.99430436e-01 -3.90290201e-01 3.11238647e-01
2.22182661e-01 -1.48426190e-01 -3.35330367e-02 2.79343333e-02
1.22131467e-01 8.60021412e-01 5.39934933e-01 -3.80013168e-01
7.73821890e-01 4.40422356e-01 -2.71112382e-01 -9.84820604e-01
-4.11976874e-01 -5.42782962e-01 2.76081026e-01 -6.89202547e-01
1.10340095e+00 -1.25108933e+00 -1.24102902e+00 5.31402051e-01
-9.75087643e-01 -4.37062919e-01 -2.29333177e-01 5.79456806e-01
-2.39610851e-01 8.45616311e-02 -3.70496631e-01 -1.00714111e+00
-1.87795475e-01 -1.37600172e+00 1.40455616e+00 5.22007525e-01
1.97774678e-01 -4.74243313e-01 -2.99774319e-01 5.21994293e-01
1.91544384e-01 9.75782871e-02 1.97498918e-01 -7.89777562e-02
-8.00127923e-01 4.24633026e-02 -3.39928210e-01 4.91611689e-01
8.45559388e-02 2.38554746e-01 -1.10737932e+00 -3.89344275e-01
-4.78411317e-01 -3.49326730e-02 3.91033083e-01 5.69174886e-01
1.02951074e+00 -2.54360974e-01 -2.61087030e-01 4.17799950e-01
1.42644024e+00 7.17605770e-01 6.42212808e-01 3.76271904e-01
4.90419120e-01 2.67003119e-01 8.25081289e-01 9.25372362e-01
5.50976694e-01 7.67086923e-01 9.68109667e-01 1.28816575e-01
2.18788445e-01 1.11233275e-02 1.18526518e+00 6.34690374e-02
-3.96682546e-02 -6.18937254e-01 -3.95807117e-01 2.19493851e-01
-1.80796587e+00 -1.04636276e+00 9.70952362e-02 2.42742443e+00
6.56495690e-02 2.57158011e-01 6.66632414e-01 -9.63533763e-04
9.23936188e-01 -1.10938825e-01 -9.26440239e-01 -1.97835922e-01
1.24527685e-01 -2.80313909e-01 1.20951164e+00 1.12574443e-01
-9.54018712e-01 5.14406085e-01 5.90613651e+00 4.35052037e-01
-8.83312523e-01 1.32885367e-01 6.26453876e-01 -6.06617153e-01
2.37066209e-01 -5.33938229e-01 -8.91609788e-01 9.04142499e-01
1.28813195e+00 -2.68734433e-02 9.18088913e-01 9.03637946e-01
5.76671779e-01 -3.17967921e-01 -9.98599887e-01 1.48292935e+00
3.44716385e-02 -1.03501225e+00 -4.25681978e-01 3.48725736e-01
6.03676021e-01 3.01523775e-01 7.71390796e-02 3.95966053e-01
7.19541848e-01 -6.38783872e-01 9.24489141e-01 3.96616191e-01
7.23949492e-01 -9.84390795e-01 7.53822565e-01 2.01368675e-01
-1.41233242e+00 -7.32740402e-01 -1.08604901e-01 6.07614033e-02
3.09448719e-01 6.25578523e-01 -5.10837197e-01 2.04916596e-01
1.19755185e+00 4.48296547e-01 -4.52062637e-01 7.42494285e-01
-1.26333851e-02 6.14773095e-01 -4.90761876e-01 -4.34165001e-02
3.70504037e-02 -2.31049895e-01 6.68471873e-01 9.33858335e-01
4.51290637e-01 1.30598426e-01 3.51993382e-01 5.73494136e-01
-4.18144435e-01 -5.19814193e-01 -5.49532115e-01 3.03333789e-01
8.51678312e-01 1.42774761e+00 -4.98446584e-01 -3.76092821e-01
-6.68171465e-01 7.74773359e-01 -2.67654568e-01 4.96161759e-01
-1.47303343e+00 -2.03818664e-01 1.27473462e+00 -6.03864994e-03
4.08952057e-01 -1.23677641e-01 2.36315951e-01 -9.49417651e-01
1.35836110e-01 -8.26381743e-01 3.97693127e-01 -1.04821515e+00
-1.29165769e+00 2.24010959e-01 3.20216715e-02 -1.31402969e+00
-8.87814313e-02 -7.85317779e-01 -4.56304133e-01 3.23583335e-01
-1.51087475e+00 -5.59378386e-01 -5.98612547e-01 7.76342571e-01
7.22774684e-01 -2.51692653e-01 1.21129658e-02 6.67281210e-01
-1.06694651e+00 6.46643877e-01 -5.50539559e-03 -4.62040119e-03
7.81369984e-01 -1.00258136e+00 -3.40638310e-02 9.10317779e-01
-3.04371715e-01 -8.94775987e-02 5.48087776e-01 -2.97629356e-01
-1.70547235e+00 -1.48137474e+00 3.23991627e-02 -3.29664171e-01
6.28144443e-01 -1.97256684e-01 -1.78499326e-01 5.98554194e-01
2.29654968e-01 1.40446708e-01 3.70202392e-01 -3.66984218e-01
2.38997862e-02 -7.11034477e-01 -1.37963831e+00 6.17984712e-01
8.68329048e-01 -1.55113503e-01 1.72658451e-02 1.10594064e-01
8.16438317e-01 -1.33382663e-01 -7.79578686e-01 -7.49419164e-03
5.54118395e-01 -1.12252498e+00 8.17919433e-01 -9.24988165e-02
6.62450269e-02 -4.34922338e-01 -5.42797506e-01 -1.17325854e+00
-1.18248008e-01 -6.92994535e-01 -3.65862608e-01 1.44508243e+00
9.87578109e-02 -7.63850808e-01 6.43163741e-01 7.08409429e-01
-9.11684334e-02 -3.69093806e-01 -9.07804310e-01 -9.86861110e-01
-5.30969441e-01 -7.61447370e-01 8.71475577e-01 3.52823466e-01
-5.60948491e-01 3.74287099e-01 -4.31542516e-01 6.77784443e-01
8.21497619e-01 -3.26520503e-01 9.27199662e-01 -9.88607705e-01
-5.59554875e-01 -2.24371240e-01 -3.99335474e-01 -9.18758988e-01
-2.35465497e-01 -3.86037566e-02 -3.46591920e-02 -1.14665556e+00
1.72433779e-01 -4.32617188e-01 -3.72216463e-01 3.81730705e-01
4.82386276e-02 2.28948668e-01 5.13405085e-01 7.61607140e-02
-1.15805936e+00 3.88102114e-01 7.45395958e-01 -3.25588286e-01
-1.68636918e-01 2.65583783e-01 -7.96858251e-01 6.76836073e-01
8.27781200e-01 -2.74368316e-01 -5.42040765e-01 -4.95128632e-01
2.28343427e-01 2.03544199e-02 4.66030478e-01 -1.31370747e+00
8.13933387e-02 -2.71327287e-01 3.60031128e-01 -3.90373617e-01
3.51014704e-01 -1.30162489e+00 2.03181118e-01 2.97649890e-01
2.01122269e-01 5.28781831e-01 1.39726460e-01 7.11088538e-01
2.64656574e-01 5.55003993e-02 8.66903245e-01 -2.93563660e-02
-9.09851730e-01 4.55349714e-01 -7.11266756e-01 8.96707773e-02
1.59211934e+00 -2.50373363e-01 -4.36146468e-01 -5.12715816e-01
-2.31656715e-01 6.47079766e-01 4.36652839e-01 5.48925877e-01
5.05813956e-01 -1.22564459e+00 -4.23273265e-01 9.71328989e-02
1.54938534e-01 -9.64685231e-02 1.65388599e-01 6.93264782e-01
-1.37300774e-01 4.29754630e-02 -1.36533901e-01 -7.66799569e-01
-1.25806546e+00 6.38726473e-01 3.61530542e-01 -1.99266940e-01
-2.21521527e-01 3.38695973e-01 8.44496768e-03 -1.98392458e-02
2.31528267e-01 -3.33982170e-01 3.66187617e-02 -4.96789739e-02
5.89844465e-01 1.03914726e+00 -1.15832657e-01 -4.64453757e-01
-3.19283932e-01 4.60011631e-01 2.95276821e-01 -7.42968991e-02
1.10602212e+00 -5.25152087e-01 5.25574446e-01 1.82763815e-01
8.58796597e-01 -5.38870990e-02 -1.77666676e+00 7.89060891e-02
-4.63453621e-01 -2.90213883e-01 5.20200990e-02 -5.93426168e-01
-1.47398603e+00 3.94147843e-01 9.33764637e-01 3.42349201e-01
1.33129537e+00 1.31872013e-01 9.09789562e-01 2.92486578e-01
7.43595362e-01 -1.32907307e+00 2.62947947e-01 -1.22361314e-02
2.60548085e-01 -1.41131032e+00 -1.00108705e-01 -1.71999410e-01
-8.47473145e-01 8.53311241e-01 8.11028242e-01 9.31876376e-02
2.58657724e-01 5.04955769e-01 -1.15684651e-01 -7.32408389e-02
-7.89469600e-01 -2.16279089e-01 -4.59377080e-01 7.48846531e-01
-3.04312140e-01 3.27603877e-01 5.59514821e-01 4.87562746e-01
2.14378819e-01 -1.80477381e-01 6.70544744e-01 8.00590158e-01
-4.40063953e-01 -6.95357203e-01 -8.30071688e-01 7.02590168e-01
-2.35945404e-01 3.60841364e-01 1.19811006e-01 7.40437746e-01
4.93456364e-01 1.71305168e+00 3.84144336e-01 -5.11982083e-01
5.46375751e-01 -5.20745754e-01 1.38023004e-01 -2.55176455e-01
-2.46741667e-01 1.51549548e-01 1.02327995e-01 -8.24817359e-01
-5.43134689e-01 -1.05659401e+00 -1.18494475e+00 -6.63112581e-01
-8.55341461e-03 -2.10620165e-01 4.89424646e-01 8.85631859e-01
5.50278544e-01 8.41987371e-01 9.04851019e-01 -6.94355905e-01
-5.15967011e-01 -5.70726991e-01 -8.43384862e-01 1.79704547e-01
4.32389200e-01 -9.42879081e-01 -5.20380557e-01 6.05428070e-02] | [8.240739822387695, -0.9315669536590576] |
7fcf9071-2d7d-4eb5-8ec4-a8405bc621d5 | fast-disparity-estimation-from-a-single | 2209.11342 | null | https://arxiv.org/abs/2209.11342v1 | https://arxiv.org/pdf/2209.11342v1.pdf | Fast Disparity Estimation from a Single Compressed Light Field Measurement | The abundant spatial and angular information from light fields has allowed the development of multiple disparity estimation approaches. However, the acquisition of light fields requires high storage and processing cost, limiting the use of this technology in practical applications. To overcome these drawbacks, the compressive sensing (CS) theory has allowed the development of optical architectures to acquire a single coded light field measurement. This measurement is decoded using an optimization algorithm or deep neural network that requires high computational costs. The traditional approach for disparity estimation from compressed light fields requires first recovering the entire light field and then a post-processing step, thus requiring long times. In contrast, this work proposes a fast disparity estimation from a single compressed measurement by omitting the recovery step required in traditional approaches. Specifically, we propose to jointly optimize an optical architecture for acquiring a single coded light field snapshot and a convolutional neural network (CNN) for estimating the disparity maps. Experimentally, the proposed method estimates disparity maps comparable with those obtained from light fields reconstructed using deep learning approaches. Furthermore, the proposed method is 20 times faster in training and inference than the best method that estimates the disparity from reconstructed light fields. | ['Henry Arguello', 'Edwin Vargas', 'Emmanuel Martinez'] | 2022-09-22 | null | null | null | null | ['disparity-estimation'] | ['computer-vision'] | [ 6.44778490e-01 -4.11955506e-01 3.46925646e-01 -4.32158142e-01
-4.35391754e-01 -5.62855899e-02 3.24942261e-01 -1.64370060e-01
-7.70099819e-01 9.05720890e-01 -3.78671825e-01 -1.26746133e-01
5.10017425e-02 -7.76720464e-01 -7.77018845e-01 -8.55026603e-01
3.79578114e-01 8.58299993e-03 1.13810204e-01 2.57244468e-01
7.63683081e-01 4.50997710e-01 -1.90482879e+00 -1.35945022e-01
8.89306962e-01 1.38361454e+00 9.93350565e-01 5.24315000e-01
-1.58942744e-01 8.86143804e-01 -2.38675401e-01 -8.66832063e-02
4.68594164e-01 -3.25418383e-01 -2.72057980e-01 1.84910521e-02
1.04710114e+00 -1.09209538e+00 -5.04533708e-01 1.26924706e+00
5.07902563e-01 7.59668350e-02 3.86501551e-02 -7.27252662e-01
-4.31498468e-01 1.00409030e-03 -3.52841467e-01 8.83751437e-02
3.36909950e-01 2.15656906e-01 4.85706359e-01 -1.01788318e+00
3.38892519e-01 6.42998159e-01 4.25857276e-01 3.30290198e-01
-8.51969361e-01 -5.25394857e-01 -4.41261232e-01 3.80982041e-01
-1.36549926e+00 -7.02664316e-01 8.17998290e-01 -5.04943132e-01
9.28836584e-01 -2.53684610e-01 8.92686009e-01 3.96643221e-01
2.33429953e-01 3.29987794e-01 1.36269903e+00 -6.30525172e-01
2.19817400e-01 -1.24443293e-01 -1.81343883e-01 6.73210442e-01
5.99830806e-01 3.82510632e-01 -4.78454828e-01 3.51806045e-01
9.07807410e-01 3.52830261e-01 -5.18865943e-01 -5.37537374e-02
-1.09811699e+00 3.59333694e-01 4.62187499e-01 2.34938860e-01
-5.18734932e-01 3.38163942e-01 -2.35209569e-01 -8.24711192e-03
3.02461207e-01 3.65000218e-01 1.29134402e-01 4.55978289e-02
-1.40045249e+00 -1.22461677e-01 3.94078255e-01 7.11607456e-01
1.39792776e+00 2.06234440e-01 2.17981100e-01 5.86445987e-01
3.35833073e-01 8.55995893e-01 2.38452628e-01 -1.16930449e+00
3.11529845e-01 3.38312447e-01 3.61239821e-01 -1.11014295e+00
-3.12771678e-01 -4.54941332e-01 -9.20229495e-01 3.85565370e-01
5.42868972e-01 -8.88795108e-02 -7.26219773e-01 1.45957756e+00
2.04931155e-01 5.59660137e-01 -3.78764831e-02 1.13636518e+00
4.99541193e-01 5.76070309e-01 -5.73941112e-01 -3.40895534e-01
7.34187603e-01 -6.89292014e-01 -6.60353720e-01 -3.12429309e-01
-4.15451005e-02 -1.16257977e+00 5.24355114e-01 6.69113457e-01
-1.31629133e+00 -6.19051039e-01 -1.27163482e+00 -3.62594306e-01
1.14885062e-01 3.59133363e-01 7.10448027e-01 2.92267591e-01
-1.11620378e+00 4.15286094e-01 -7.95543194e-01 2.12250650e-02
4.08569396e-01 4.25072283e-01 -8.21665376e-02 -6.60797596e-01
-5.73992908e-01 7.96052635e-01 2.32514799e-01 4.60881293e-01
-7.68114865e-01 -5.03059328e-01 -6.84527636e-01 -7.61206402e-03
1.04436055e-01 -7.78850257e-01 7.74547875e-01 -5.37194133e-01
-1.87942731e+00 4.85628724e-01 -4.24875319e-01 -2.90849537e-01
1.27139822e-01 -3.49950731e-01 1.87663846e-02 4.59441096e-01
4.42161923e-03 4.46378022e-01 1.13363850e+00 -9.75958228e-01
-7.55628169e-01 -3.73116255e-01 1.67972699e-01 -1.93200838e-02
-2.81077027e-02 -4.03900832e-01 -4.19746965e-01 1.35955811e-02
5.36311150e-01 -7.41540134e-01 -2.05019802e-01 1.66207731e-01
-2.53207862e-01 3.99250954e-01 5.53058267e-01 -3.73905331e-01
8.81808400e-01 -2.02762842e+00 -7.18194246e-02 -1.04598500e-01
4.07832742e-01 3.22447926e-01 -4.54545319e-02 1.39118150e-01
2.59840965e-01 -7.79549837e-01 -1.84411481e-01 -6.30146265e-01
-5.41138411e-01 6.21083640e-02 -2.50459343e-01 7.97580302e-01
-9.96324569e-02 5.45714140e-01 -8.40397596e-01 -3.66461545e-01
7.77440727e-01 6.76387966e-01 -9.68750536e-01 3.82549971e-01
-1.37978151e-01 8.75863314e-01 -2.46071368e-01 6.62536442e-01
1.18205762e+00 -5.26543558e-01 1.53977185e-01 -4.13969278e-01
-6.37605250e-01 1.56197920e-01 -1.14528584e+00 1.90586936e+00
-7.38048315e-01 9.65824068e-01 2.97100514e-01 -1.02706325e+00
9.93574977e-01 1.47975340e-01 5.53216219e-01 -8.70035470e-01
3.10457945e-01 6.33751571e-01 -3.16105574e-01 -5.86494446e-01
6.65749133e-01 -3.22103024e-01 5.77972412e-01 3.97144288e-01
-8.95824805e-02 -4.35118824e-01 1.02964267e-01 -1.81107029e-01
8.59166384e-01 1.75824538e-01 1.42841294e-01 2.05316216e-01
7.20487595e-01 -2.93517858e-01 5.88968039e-01 5.25509477e-01
-2.04712544e-02 5.45383036e-01 -3.95804018e-01 -6.97521627e-01
-9.33845878e-01 -8.47603142e-01 -3.11660290e-01 2.97762841e-01
6.71188235e-01 5.42697823e-03 -4.35540825e-01 1.28772214e-01
-1.29073381e-01 2.85637230e-01 6.83846250e-02 3.54357392e-01
-7.14789629e-01 -4.22181875e-01 7.94060752e-02 9.72250104e-03
9.52792704e-01 -5.78199446e-01 -1.10547268e+00 2.19733715e-01
-5.15187085e-01 -1.49602532e+00 -9.13633704e-02 -1.22395858e-01
-9.09003556e-01 -1.06616676e+00 -6.15121901e-01 -6.69685960e-01
8.12332273e-01 7.18797743e-01 8.21208119e-01 1.29879594e-01
-3.49902302e-01 1.55927747e-01 9.37760156e-03 -2.16743678e-01
2.21419916e-01 -3.07522625e-01 2.32120398e-02 2.92960703e-01
3.72045308e-01 -9.07667696e-01 -1.10264575e+00 1.03614345e-01
-8.58597279e-01 3.58818114e-01 8.28254163e-01 7.22469509e-01
6.70401812e-01 -1.06995881e-01 1.72936454e-01 -3.79985631e-01
7.81414062e-02 -1.90587059e-01 -1.43074131e+00 -1.53707694e-02
-6.69919252e-01 3.61747965e-02 7.02334404e-01 -6.04946241e-02
-1.08767867e+00 2.49432653e-01 -9.05717835e-02 -4.98899907e-01
-1.50114387e-01 2.91408718e-01 3.12944293e-01 -5.62179685e-01
3.61244351e-01 4.83549148e-01 7.53962323e-02 -4.87676263e-01
-6.70832992e-02 6.61254406e-01 7.69443750e-01 -1.50161877e-01
6.65973365e-01 9.82918918e-01 5.20074904e-01 -1.02185941e+00
-8.04497361e-01 -5.44486821e-01 -5.71546435e-01 -3.90463114e-01
8.51717949e-01 -1.00628924e+00 -9.51420188e-01 7.23655641e-01
-1.38628507e+00 3.07297916e-03 2.27820221e-02 1.16223323e+00
-4.31639880e-01 5.89822412e-01 -5.18385291e-01 -6.01728201e-01
-2.20198914e-01 -1.25181091e+00 1.01013565e+00 2.83443660e-01
4.54207093e-01 -6.61427021e-01 9.51440781e-02 5.18827200e-01
7.46944308e-01 -1.58671081e-01 5.24557412e-01 6.49372756e-01
-1.58944464e+00 -1.91271931e-01 -4.82368767e-01 4.15025622e-01
1.72006890e-01 -3.37407529e-01 -1.14734685e+00 -3.41712683e-01
3.44808251e-01 -2.68408388e-01 7.11886525e-01 9.11132157e-01
1.05097079e+00 1.37588382e-01 -4.31878939e-02 1.27341318e+00
2.17096233e+00 3.03959936e-01 8.54306698e-01 2.59073019e-01
6.17502153e-01 3.99696797e-01 3.68777543e-01 5.79320371e-01
3.87458593e-01 6.48715138e-01 7.37410963e-01 -1.90959256e-02
-2.54268229e-01 1.21665839e-03 5.54694533e-02 9.14891481e-01
-2.87191182e-01 -9.60277393e-02 -5.10694027e-01 3.82647514e-01
-1.35172105e+00 -1.08198023e+00 -1.70250479e-02 2.36432886e+00
5.85049272e-01 -3.03591937e-01 -6.67892754e-01 1.48138374e-01
4.44433630e-01 2.14176446e-01 -5.35277963e-01 1.39284655e-01
-9.68262330e-02 4.27570999e-01 5.52857220e-01 9.58221376e-01
-6.20494366e-01 7.45849669e-01 6.28710270e+00 3.19387257e-01
-1.78480756e+00 -3.82776000e-02 -1.50803337e-03 -3.58258843e-01
-3.46431404e-01 1.19273037e-01 -8.35280955e-01 6.25866413e-01
4.45620298e-01 2.25970000e-01 8.91186178e-01 4.38001186e-01
2.72383362e-01 -7.27934718e-01 -1.02183807e+00 1.60922742e+00
4.58436869e-02 -1.47642338e+00 -1.02919199e-01 1.68412194e-01
8.31496835e-01 3.03294957e-01 1.65882289e-01 -4.02360886e-01
-2.22513184e-01 -7.45401144e-01 3.33050460e-01 8.12822878e-01
1.15747857e+00 -2.45989442e-01 6.63095415e-01 4.61090147e-01
-8.61571252e-01 -3.46453227e-02 -6.16096556e-01 -4.70566213e-01
3.79973918e-01 1.25687289e+00 -6.61407292e-01 4.16322351e-01
5.83013892e-01 8.90692115e-01 -1.66827500e-01 1.05806005e+00
-8.75794068e-02 2.84112573e-01 -4.52605605e-01 2.05945507e-01
4.83799800e-02 -6.60202503e-01 5.02117813e-01 6.15746021e-01
8.24132860e-01 7.16910958e-02 -1.66024989e-03 9.51697171e-01
6.11437373e-02 -3.55827183e-01 -6.80531979e-01 1.85552359e-01
3.61823976e-01 1.17590308e+00 -3.84136766e-01 -2.87167013e-01
-7.43116915e-01 7.80482113e-01 1.68094903e-01 3.97083580e-01
-3.88196707e-01 -5.09931982e-01 4.44192439e-01 3.68809581e-01
2.87358254e-01 -5.22399306e-01 -2.29423374e-01 -1.39231968e+00
1.95463941e-01 -3.24201405e-01 -3.44298244e-01 -1.03431058e+00
-8.82286251e-01 6.76924288e-01 -2.20396310e-01 -1.56645644e+00
-3.40813547e-01 -6.07068717e-01 -3.21214437e-01 1.17001450e+00
-2.41754365e+00 -8.00389767e-01 -8.90953004e-01 8.76935899e-01
2.52922088e-01 -8.42433795e-02 3.96806419e-01 6.53635502e-01
-4.02683705e-01 -3.13544348e-02 2.08868116e-01 -1.91929460e-01
6.58870101e-01 -8.42725933e-01 -2.36754984e-01 1.23363483e+00
-2.87470937e-01 6.51582539e-01 4.79238540e-01 -3.77788067e-01
-1.58184493e+00 -7.03998804e-01 9.88318801e-01 2.21005633e-01
6.51883036e-02 1.58826098e-01 -5.90406179e-01 5.46191633e-01
3.35028917e-01 3.48788202e-01 4.19978440e-01 -4.51094061e-01
7.96049014e-02 -4.96884853e-01 -1.09429502e+00 1.81264117e-01
7.89508402e-01 -6.78961396e-01 -2.82331645e-01 7.87528902e-02
1.61313146e-01 -4.14272398e-01 -3.74541461e-01 2.10787416e-01
6.80272758e-01 -1.53022766e+00 9.53530967e-01 5.11243105e-01
6.62774801e-01 -6.98351324e-01 -4.27158386e-01 -1.03329122e+00
-1.26579210e-01 -5.83644509e-01 -7.01574162e-02 4.86132294e-01
-2.03013659e-01 -8.98064852e-01 8.63311648e-01 2.57755071e-01
-3.49484593e-01 -4.46089596e-01 -8.95421028e-01 -2.73929238e-01
-5.37729740e-01 -4.40229774e-01 4.57314104e-01 6.50919020e-01
-4.32913780e-01 1.76389396e-01 -5.68981767e-01 4.64303404e-01
1.26526010e+00 5.83672047e-01 8.02545190e-01 -1.19858158e+00
-4.14409280e-01 -2.65091080e-02 -4.58477706e-01 -1.93676150e+00
-3.23998556e-02 -8.03554714e-01 3.46278846e-01 -1.46051824e+00
-5.85537553e-02 -6.45183086e-01 -6.53771907e-02 -1.38515174e-01
3.58217135e-02 4.23466712e-01 1.43314332e-01 5.18389821e-01
-3.26868296e-01 4.55554456e-01 1.52903759e+00 6.61872998e-02
4.67994809e-02 -6.42537847e-02 -2.50656784e-01 6.98532522e-01
4.63523388e-01 -2.88318723e-01 -5.01443446e-01 -1.11409068e+00
5.90909183e-01 3.67252290e-01 4.86377239e-01 -1.40598285e+00
7.59613037e-01 -1.73701525e-01 3.67751926e-01 -8.31711769e-01
6.51486635e-01 -9.17059243e-01 1.27708107e-01 4.34061766e-01
1.29550561e-01 -2.68841028e-01 -1.31736472e-01 5.10708630e-01
-5.41396260e-01 -3.25074553e-01 8.90015662e-01 -1.27281055e-01
-8.33651304e-01 3.69194627e-01 -2.09239662e-01 -2.81219810e-01
7.81677425e-01 -5.02283573e-01 -2.17846170e-01 -2.84511209e-01
-2.25658491e-01 -2.74926156e-01 4.78105456e-01 -3.73954087e-01
1.04719269e+00 -1.01822221e+00 -5.07264256e-01 8.25220346e-01
-2.69879907e-01 2.69315839e-01 3.89636338e-01 9.13683951e-01
-1.19053483e+00 5.94154537e-01 -4.48166251e-01 -9.68729138e-01
-9.44075584e-01 2.90882170e-01 3.87819678e-01 8.20113644e-02
-5.88324308e-01 8.21339905e-01 -4.26374190e-02 3.77055146e-02
-4.68593799e-02 -4.26392466e-01 -2.09010765e-02 -3.77803713e-01
6.83804512e-01 5.55681169e-01 2.39961203e-02 -4.06913787e-01
-1.93671063e-01 1.17751396e+00 3.46743345e-01 -3.82259563e-02
1.39006639e+00 -4.74516541e-01 -4.04087573e-01 6.75963685e-02
1.32982266e+00 2.74340063e-01 -1.39243948e+00 -4.44150388e-01
-5.55723250e-01 -1.12880611e+00 6.04217410e-01 -2.53794640e-01
-1.31092572e+00 1.22312558e+00 6.38154805e-01 -1.25097305e-01
1.47714210e+00 -4.70164955e-01 1.08157492e+00 4.81591463e-01
8.32773268e-01 -8.47673416e-01 -6.85461685e-02 4.12785560e-01
4.01919007e-01 -1.50274146e+00 1.03372440e-01 -3.32601488e-01
7.84120783e-02 1.27977622e+00 3.03980798e-01 -7.15142787e-02
7.99820960e-01 2.74786532e-01 5.37483916e-02 -1.98292926e-01
-4.36682135e-01 -1.76478609e-01 2.85739470e-02 4.31340098e-01
3.58847499e-01 -1.73973575e-01 -2.36340299e-01 -4.22961146e-01
3.32669839e-02 6.25753760e-01 6.26809657e-01 8.25766325e-01
-5.38448155e-01 -1.02162707e+00 -2.47743070e-01 2.29460686e-01
-3.89878899e-01 -3.14333677e-01 4.03385937e-01 1.78160354e-01
4.39302117e-01 1.05615759e+00 3.04200023e-01 -3.13854992e-01
-1.19577594e-01 -3.86067927e-01 8.97770405e-01 -6.14087880e-01
4.89040986e-02 2.24756338e-02 -4.28441346e-01 -7.80060887e-01
-8.78717601e-01 -2.16735944e-01 -9.91343856e-01 -3.95141810e-01
-3.26798856e-01 -1.85810328e-01 9.18333352e-01 8.93458247e-01
2.78667271e-01 1.93347130e-03 8.62790108e-01 -1.10807467e+00
-1.02899313e-01 -7.30009079e-01 -6.07228458e-01 1.55730873e-01
7.50875294e-01 -6.39603972e-01 -5.49548328e-01 3.53072733e-02] | [9.734971046447754, -2.6083598136901855] |
bf62c3b3-e9e6-4ce9-8799-da86dbfed0b8 | effective-medical-code-prediction-via-label | 2305.05162 | null | https://arxiv.org/abs/2305.05162v1 | https://arxiv.org/pdf/2305.05162v1.pdf | Effective Medical Code Prediction via Label Internal Alignment | The clinical notes are usually typed into the system by physicians. They are typically required to be marked by standard medical codes, and each code represents a diagnosis or medical treatment procedure. Annotating these notes is time consuming and prone to error. In this paper, we proposed a multi-view attention based Neural network to predict medical codes from clinical texts. Our method incorporates three aspects of information, the semantic context of the clinical text, the relationship among the label (medical codes) space, and the alignment between each pair of a clinical text and medical code. Our method is verified to be effective on the open source dataset. The experimental result shows that our method achieves better performance against the prior state-of-art on multiple metrics. | ['Guodong Liu'] | 2023-05-09 | null | null | null | null | ['medical-code-prediction'] | ['medical'] | [ 7.46886358e-02 8.42721537e-02 -3.33092600e-01 -5.87584853e-01
-8.12823176e-01 -5.42975485e-01 6.19864240e-02 7.85613000e-01
-2.32316822e-01 3.19025844e-01 5.56543350e-01 -2.82888889e-01
-1.01142302e-01 -5.00898302e-01 -1.03743844e-01 -4.41579491e-01
2.06867427e-01 7.30975330e-01 -1.38545737e-01 1.96253538e-01
1.94190383e-01 -4.65759113e-02 -8.78280342e-01 6.48448408e-01
7.33860672e-01 9.96101558e-01 1.14122815e-01 4.91452366e-01
-1.60129055e-01 1.17978883e+00 -3.40111256e-01 -5.96023500e-01
-1.29300440e-02 -5.66321254e-01 -9.60737348e-01 2.17524897e-02
1.95434809e-01 -2.14652181e-01 -2.03794703e-01 1.34515059e+00
4.54662710e-01 -2.33186558e-01 7.97449648e-01 -8.42123210e-01
-1.25886571e+00 6.20123923e-01 -3.80806357e-01 2.60906667e-02
3.10827047e-01 -3.05041313e-01 1.11783087e+00 -9.40295994e-01
6.26070440e-01 7.49255240e-01 8.62824440e-01 6.56568468e-01
-5.08650899e-01 -4.26830441e-01 -6.81830421e-02 2.84754634e-01
-1.30844474e+00 -2.70899862e-01 4.74087059e-01 -9.83616292e-01
7.32627928e-01 2.44768158e-01 5.16675472e-01 8.28483343e-01
4.16655421e-01 5.40600598e-01 8.88903558e-01 -3.57230425e-01
2.20188275e-01 2.02495098e-01 4.19962883e-01 1.02137434e+00
8.96378085e-02 -2.16221273e-01 5.64970672e-02 -4.41132158e-01
4.50961888e-01 5.75701356e-01 -3.05842727e-01 -2.29819804e-01
-1.32118225e+00 8.76852632e-01 3.80418390e-01 2.76183814e-01
-3.60182822e-01 -2.35577434e-01 7.60848284e-01 -2.37631395e-01
4.08606976e-01 3.07046026e-01 -7.58227289e-01 -1.45229414e-01
-7.66551912e-01 -2.14849249e-01 7.36768365e-01 1.06986570e+00
8.54219347e-02 -5.23373246e-01 -4.26277727e-01 1.00711870e+00
5.21482170e-01 1.72845289e-01 7.32796609e-01 -6.22047663e-01
6.60906434e-01 9.23380136e-01 7.59725198e-02 -1.08822334e+00
-5.81815422e-01 -3.76460493e-01 -9.41333354e-01 -3.73862535e-01
1.87758803e-01 -3.05126429e-01 -9.24605608e-01 1.20751166e+00
1.87610179e-01 3.64830680e-02 1.60473928e-01 7.04020083e-01
1.21200275e+00 4.59834725e-01 3.19618940e-01 -2.78069228e-01
1.65078712e+00 -1.28095162e+00 -1.22718775e+00 -2.65075434e-02
8.99607658e-01 -7.52282262e-01 6.47106230e-01 -2.02023387e-02
-9.65758443e-01 -4.72967446e-01 -6.02864265e-01 -1.35915861e-01
-2.10240588e-01 6.63065255e-01 1.92570925e-01 2.17581078e-01
-6.41508281e-01 3.96802694e-01 -9.09139216e-01 -3.60074878e-01
4.75225270e-01 7.13433996e-02 -2.81339258e-01 2.70010829e-02
-1.10034966e+00 7.07147896e-01 4.39532220e-01 5.15887998e-02
-4.27268863e-01 -5.85076034e-01 -1.06119132e+00 3.98545265e-01
1.62387818e-01 -7.27915108e-01 1.37786078e+00 -8.26888740e-01
-9.44737852e-01 1.29957199e+00 -1.52822718e-01 -1.29871711e-01
4.04530674e-01 -7.20819607e-02 -7.41785824e-01 1.60291657e-01
2.45162398e-01 3.33916277e-01 2.01184377e-01 -9.75610733e-01
-7.10761666e-01 -4.85219151e-01 -1.08794339e-01 2.12771431e-01
-4.39251959e-01 2.29635596e-01 -5.58019936e-01 -8.79666567e-01
-1.21406093e-02 -1.08227372e+00 -3.27307165e-01 -3.50457430e-02
-5.77634037e-01 -3.27254623e-01 3.11437011e-01 -9.94169474e-01
1.65917861e+00 -2.32400036e+00 -7.09417537e-02 6.78273290e-03
4.40011173e-01 3.02908719e-01 4.52411890e-01 3.91274452e-01
-2.58943558e-01 2.32483372e-01 -2.59104073e-01 -4.12409827e-02
-2.96490610e-01 6.91294521e-02 -9.38913226e-02 4.22168702e-01
-2.54589051e-01 7.94402242e-01 -9.52726722e-01 -9.71688390e-01
-2.17381179e-01 3.78390729e-01 -6.15033269e-01 4.86615747e-01
9.87992287e-02 2.99501926e-01 -7.53776729e-01 6.70878112e-01
2.57901192e-01 -1.07352829e+00 1.78546160e-01 -2.25683123e-01
1.70015529e-01 2.32887268e-01 -6.68706238e-01 1.70335782e+00
-1.94305629e-01 2.16901153e-01 -3.45235050e-01 -7.57217824e-01
7.30791569e-01 7.73600221e-01 6.92240894e-01 -3.05518191e-02
3.06933612e-01 5.22031114e-02 9.97383222e-02 -1.26083517e+00
4.82890122e-02 9.46954638e-02 6.95179924e-02 4.30620193e-01
-1.12224698e-01 4.23890144e-01 2.25274102e-03 9.25271362e-02
9.78791535e-01 -1.29157528e-01 8.83615732e-01 -1.92145005e-01
6.85478806e-01 1.20900936e-01 7.80120850e-01 5.07843494e-01
-4.23002124e-01 6.35909021e-01 3.29522431e-01 -8.20918739e-01
-9.85751629e-01 -5.65688372e-01 -4.89487380e-01 7.32815862e-01
-4.13586423e-02 -5.30362904e-01 -8.69960845e-01 -1.12082827e+00
-1.50847986e-01 4.08973485e-01 -9.38188732e-01 -3.20371962e-03
-3.71280164e-01 -5.43003082e-01 4.88862336e-01 1.03794324e+00
2.35374480e-01 -9.27783787e-01 -6.92172527e-01 2.60808736e-01
-3.37509274e-01 -9.63762641e-01 -1.10438943e+00 2.76001776e-03
-8.41946959e-01 -1.39374530e+00 -7.45857477e-01 -1.21892858e+00
9.82603014e-01 -1.56850651e-01 1.09372365e+00 3.93920004e-01
-1.73517182e-01 2.73316115e-01 -4.56394225e-01 -4.69653368e-01
-5.11474788e-01 -1.69122368e-02 -3.71586323e-01 1.92905441e-01
7.30449677e-01 -3.30337323e-02 -8.26823115e-01 1.16521664e-01
-8.33962023e-01 1.78699940e-01 4.59057510e-01 1.02135682e+00
6.31683052e-01 -4.69737887e-01 2.03421801e-01 -1.27394557e+00
6.17758036e-01 -5.68089485e-01 -2.83572942e-01 4.54440355e-01
-8.37407291e-01 2.00888589e-02 6.65225625e-01 -8.41836110e-02
-8.33373427e-01 3.08309138e-01 -1.73289135e-01 -2.69085795e-01
-4.85739678e-01 1.01607597e+00 1.46942914e-01 3.53625715e-01
4.59719807e-01 -7.55729005e-02 -2.10782781e-01 -6.39141679e-01
1.16975214e-02 1.19600713e+00 3.50098193e-01 -1.37670442e-01
1.66275188e-01 3.14415127e-01 -1.90871358e-01 6.97854981e-02
-1.32665741e+00 -6.65147960e-01 -7.42818117e-01 1.26180192e-02
1.25599444e+00 -9.24919069e-01 -6.54804647e-01 1.36273593e-01
-1.19542336e+00 3.51536542e-01 2.55995274e-01 7.04695344e-01
-2.36358047e-01 3.70545536e-01 -9.95012581e-01 -3.63167822e-01
-5.08443058e-01 -1.28388965e+00 8.78741741e-01 9.25457776e-02
-4.19604242e-01 -1.43023372e+00 2.90499300e-01 5.94904780e-01
8.46072882e-02 2.77428448e-01 1.28509176e+00 -1.26036143e+00
4.65166084e-02 -3.78240734e-01 -3.03961009e-01 -1.99141949e-02
5.29998600e-01 1.20325074e-01 -7.09805310e-01 -1.88762769e-01
1.28026634e-01 -1.90699130e-01 5.41919351e-01 4.04071540e-01
1.65086341e+00 -5.37966073e-01 -7.02740610e-01 5.68220377e-01
1.38860428e+00 7.91377902e-01 1.94159877e-02 1.76137567e-01
1.01113415e+00 4.50844318e-01 4.15611804e-01 5.03344357e-01
6.65622234e-01 6.58073485e-01 3.16664696e-01 -2.05593601e-01
2.49067500e-01 -7.80479684e-02 -1.28567547e-01 1.34985638e+00
4.11640555e-02 -1.84082195e-01 -1.49914491e+00 7.37021983e-01
-1.87229490e+00 -8.69526029e-01 -3.76782954e-01 1.84061837e+00
1.06665397e+00 -1.96766153e-01 -2.63664603e-01 -2.73792744e-01
1.03143632e+00 -3.20905387e-01 -5.58400810e-01 -3.39359999e-01
5.14862835e-01 -2.00455233e-01 3.72753650e-01 3.45103621e-01
-1.42431617e+00 2.82620221e-01 6.89037800e+00 3.35105896e-01
-8.81051004e-01 2.66368151e-01 8.52345347e-01 5.24865277e-02
-3.60228531e-02 -4.35802370e-01 -6.75501525e-01 7.94789553e-01
8.38215709e-01 -2.02216491e-01 -5.11858538e-02 1.10207975e+00
-9.78602618e-02 3.70278507e-01 -1.44787502e+00 9.96845245e-01
3.51469994e-01 -1.29025090e+00 1.99977346e-02 3.15523073e-02
7.33749211e-01 3.28200199e-02 8.87014195e-02 -4.36101633e-04
4.97369617e-01 -1.02826178e+00 3.44893575e-01 4.47551608e-01
9.23887312e-01 -3.68793875e-01 1.18941987e+00 1.75048858e-01
-1.14875352e+00 -1.09375022e-01 7.30885938e-03 2.52557546e-01
-1.84063611e-04 4.21529561e-01 -8.43762875e-01 5.19679785e-01
7.17112660e-01 1.19622564e+00 -5.23437917e-01 1.09913170e+00
-6.47581220e-02 5.78020811e-01 4.16682094e-01 5.99480718e-02
3.97672713e-01 -6.05765767e-02 -4.25462751e-03 1.27260160e+00
4.82051194e-01 2.76030958e-01 6.58789396e-01 6.39037251e-01
-1.84726760e-01 4.38108742e-01 -5.95786154e-01 -5.87001145e-02
2.05869123e-01 1.26288044e+00 -8.33026886e-01 -6.75633729e-01
-7.93651879e-01 7.07776308e-01 2.47115880e-01 1.65851668e-01
-9.59345579e-01 -5.10395586e-01 2.97726005e-01 -2.54690409e-01
2.17794433e-01 5.68249226e-01 -3.86043370e-01 -1.24626482e+00
3.60092781e-02 -7.57076204e-01 8.48993301e-01 -9.79758978e-01
-1.46022797e+00 9.65658605e-01 -3.72704744e-01 -1.74968278e+00
-1.39511496e-01 -6.11050129e-01 -3.76171172e-02 7.47677147e-01
-1.23921442e+00 -8.05629432e-01 -3.15528423e-01 6.23147786e-01
4.24462020e-01 -4.74859536e-01 1.37972820e+00 6.35262191e-01
-6.38461113e-01 6.16013408e-01 3.39935929e-01 9.29920435e-01
9.31265175e-01 -1.30214000e+00 1.13430358e-01 3.80591780e-01
-2.17383727e-03 8.21185112e-01 2.94079661e-01 -7.45457947e-01
-7.45932519e-01 -1.22282791e+00 1.14875805e+00 -6.28998816e-01
4.99921411e-01 2.31296688e-01 -9.94207501e-01 1.01743650e+00
4.40854490e-01 2.04294711e-01 1.52567530e+00 -4.38018739e-02
-3.79449338e-01 1.54094830e-01 -1.09499276e+00 2.84454256e-01
7.03072190e-01 -4.51923698e-01 -9.79242861e-01 5.63218594e-01
6.89143836e-01 -8.53252172e-01 -1.12005532e+00 3.05351734e-01
3.84074420e-01 -5.35152256e-01 6.94458723e-01 -9.83711839e-01
1.03350472e+00 -2.51188725e-01 5.95458485e-02 -1.40415716e+00
-6.17524981e-01 8.53123292e-02 -4.12925482e-02 1.05842721e+00
5.32600999e-01 -2.93945819e-01 4.43992049e-01 9.66447890e-01
-2.36604497e-01 -9.42952037e-01 -7.04351544e-01 -2.40026951e-01
-1.34901404e-01 2.30630517e-01 5.68561137e-01 1.69235682e+00
4.21782762e-01 3.73590469e-01 -2.79193968e-01 1.71721160e-01
1.43292055e-01 2.99164176e-01 -1.16035849e-01 -1.46854115e+00
-2.57996231e-01 -4.77695972e-01 -2.11812332e-01 -5.31082273e-01
-2.54875179e-02 -1.22394812e+00 -5.78664914e-02 -1.84800196e+00
6.99642777e-01 -3.84445667e-01 -7.27362573e-01 7.78639495e-01
-5.42003632e-01 -2.20431998e-01 -1.35204628e-01 4.35155362e-01
-7.05630839e-01 1.04565360e-02 1.00922418e+00 -3.46410662e-01
7.39427805e-02 -1.28143072e-01 -7.39085734e-01 1.12126136e+00
7.40890503e-01 -9.01292264e-01 -2.41497219e-01 -6.02065682e-01
3.23510319e-01 4.64235783e-01 -8.05300772e-02 -6.47129297e-01
3.41905028e-01 -2.19708338e-01 3.66933554e-01 -4.08982426e-01
-1.77780464e-01 -1.21316338e+00 -3.13728452e-02 7.43479013e-01
-9.05700684e-01 4.74320799e-01 -8.81208032e-02 5.95143139e-01
-4.35189754e-01 -4.72519219e-01 6.58528090e-01 -4.55272138e-01
-3.19961280e-01 4.31432039e-01 -2.56755680e-01 2.51249403e-01
9.96773243e-01 1.54682606e-01 -1.82912290e-01 -9.25140977e-02
-9.28927064e-01 3.67466271e-01 2.45068356e-01 4.34452474e-01
5.77056825e-01 -1.56168413e+00 -6.32902741e-01 -1.05419524e-01
6.63906157e-01 -2.41683230e-01 2.08066300e-01 8.45959127e-01
-8.17741632e-01 6.66988432e-01 -6.46229535e-02 -4.53144938e-01
-1.63723075e+00 9.66799915e-01 4.05782759e-01 -5.13183713e-01
-6.63766682e-01 7.45449543e-01 2.31541172e-01 -4.69642699e-01
4.93966609e-01 -4.75602329e-01 -9.02154446e-01 -2.99969572e-03
6.54367805e-01 4.19230200e-02 8.50893836e-03 -6.79025769e-01
-4.39538032e-01 8.35302949e-01 -3.58819813e-01 3.18995923e-01
1.41554821e+00 8.51316378e-03 -4.01918948e-01 5.49285293e-01
1.26210368e+00 -6.31739572e-02 -6.94215119e-01 -4.40157115e-01
1.15379646e-01 -2.48175517e-01 -1.07117236e-01 -1.13363707e+00
-1.32765496e+00 9.59039211e-01 7.04762459e-01 2.30587214e-01
9.44883585e-01 -1.63424332e-02 8.84373784e-01 3.30614060e-01
4.26560966e-03 -1.14516664e+00 -1.43789411e-01 4.21375453e-01
5.66713393e-01 -1.48276329e+00 -1.92933753e-01 -2.20757619e-01
-1.02936721e+00 1.18958926e+00 6.19477093e-01 3.42974871e-01
8.81174207e-01 1.09754093e-01 6.42687917e-01 -4.56786543e-01
-6.30941093e-01 2.80444562e-01 4.68328774e-01 1.34658635e-01
9.30494785e-01 1.35668442e-01 -3.57633531e-01 7.36357749e-01
2.49969736e-01 1.52158886e-01 4.93254840e-01 9.88747656e-01
-1.41623259e-01 -9.39905703e-01 -3.73849213e-01 6.01364315e-01
-1.18935895e+00 -3.95758867e-01 -3.25173229e-01 2.30787933e-01
2.96420664e-01 8.56423855e-01 2.22355705e-02 -2.26892874e-01
1.79697528e-01 5.14007449e-01 -1.99371189e-01 -9.05198693e-01
-8.30773115e-01 1.79898560e-01 -1.77436695e-01 -2.37689838e-01
-6.31507754e-01 -6.68862879e-01 -1.51336706e+00 1.66955560e-01
-8.17385465e-02 3.65332454e-01 2.51768470e-01 8.80150437e-01
7.59690583e-01 8.32120836e-01 4.27128643e-01 2.06628188e-01
-3.24350953e-01 -8.68146539e-01 -2.21954063e-01 9.13191319e-01
4.98178810e-01 -4.56852406e-01 9.90226492e-02 7.14576900e-01] | [7.997981071472168, 6.834123611450195] |
9a66758b-0443-430f-84a8-a99f1b7e36de | target-speaker-verification-with-selective | 2103.16269 | null | https://arxiv.org/abs/2103.16269v2 | https://arxiv.org/pdf/2103.16269v2.pdf | Target Speaker Verification with Selective Auditory Attention for Single and Multi-talker Speech | Speaker verification has been studied mostly under the single-talker condition. It is adversely affected in the presence of interference speakers. Inspired by the study on target speaker extraction, e.g., SpEx, we propose a unified speaker verification framework for both single- and multi-talker speech, that is able to pay selective auditory attention to the target speaker. This target speaker verification (tSV) framework jointly optimizes a speaker attention module and a speaker representation module via multi-task learning. We study four different target speaker embedding schemes under the tSV framework. The experimental results show that all four target speaker embedding schemes significantly outperform other competitive solutions for multi-talker speech. Notably, the best tSV speaker embedding scheme achieves 76.0% and 55.3% relative improvements over the baseline system on the WSJ0-2mix-extr and Libri2Mix corpora in terms of equal-error-rate for 2-talker speech, while the performance of tSV for single-talker speech is on par with that of traditional speaker verification system, that is trained and evaluated under the same single-talker condition. | ['Haizhou Li', 'Jibin Wu', 'Wei Rao', 'Chenglin Xu'] | 2021-03-30 | null | null | null | null | ['target-speaker-extraction'] | ['audio'] | [ 1.22373864e-01 -4.48625647e-02 -3.65684368e-02 -5.44191301e-01
-1.65102673e+00 -5.00902176e-01 4.88389432e-01 -2.26149619e-01
-2.44474709e-01 1.20916590e-01 3.52441192e-01 -3.38110030e-01
2.17926502e-01 5.78221492e-02 -3.72959435e-01 -1.02481961e+00
1.28474429e-01 2.01336041e-01 -2.25999936e-01 -2.83017248e-01
-1.35632783e-01 9.33546871e-02 -1.53141177e+00 2.40677327e-01
4.87578571e-01 8.70004416e-01 2.51460344e-01 8.78800869e-01
-3.05850562e-02 3.49176377e-01 -9.45731580e-01 -4.62095529e-01
-1.48964554e-01 -2.69198179e-01 -6.11929059e-01 -6.56688809e-02
7.79875159e-01 4.11977284e-02 -2.63633400e-01 1.08818102e+00
1.16177189e+00 -9.78833213e-02 5.17394006e-01 -1.37060595e+00
-7.64114141e-01 1.28195584e+00 -5.16749918e-01 4.78047609e-01
2.84416527e-01 -1.14902683e-01 1.32243574e+00 -1.29631174e+00
-1.58541188e-01 1.72296536e+00 4.89383489e-01 8.37126493e-01
-1.31881630e+00 -1.14372706e+00 3.92291486e-01 3.13885123e-01
-1.76830316e+00 -1.26372075e+00 8.18980694e-01 -2.67061144e-01
7.84380436e-01 4.99695718e-01 -2.03819126e-01 1.51191282e+00
-2.23629743e-01 1.04569507e+00 8.52373242e-01 -3.28025788e-01
-3.92544791e-02 4.62802619e-01 5.94063580e-01 2.46330947e-01
-2.30829760e-01 3.48306656e-01 -1.04108393e+00 -1.15690365e-01
-9.79007334e-02 -5.86234570e-01 -7.08142400e-01 2.98288226e-01
-1.33938324e+00 8.36701274e-01 7.17738718e-02 4.94685948e-01
-1.94553986e-01 2.21405160e-02 5.15682936e-01 5.69018662e-01
4.82903033e-01 -1.91436678e-01 -5.06378531e-01 1.78999245e-01
-1.06389713e+00 3.79193872e-02 7.70155430e-01 9.94363427e-01
3.64988863e-01 7.56548285e-01 -4.87063587e-01 1.17208433e+00
7.18263507e-01 9.53431487e-01 7.75669813e-01 -2.34365270e-01
6.02177024e-01 -1.79848328e-01 -1.07014328e-01 -4.25799817e-01
1.59692019e-02 -7.18106210e-01 -8.49790037e-01 7.11691380e-02
-1.12746786e-02 -1.27082631e-01 -7.24550068e-01 1.88860679e+00
4.54294533e-01 3.26685339e-01 5.73143899e-01 7.80563176e-01
1.35234523e+00 8.75583649e-01 -4.91458438e-02 -2.68035263e-01
1.64512742e+00 -1.16428149e+00 -1.13624322e+00 -3.75833005e-01
2.05420241e-01 -1.02214992e+00 1.04124713e+00 2.57661968e-01
-7.63945878e-01 -7.76975274e-01 -1.10344529e+00 2.05686465e-01
-1.24110833e-01 3.13144922e-01 -1.67846978e-01 1.32549417e+00
-1.16422117e+00 -1.77875057e-01 -4.38022077e-01 3.86149101e-02
1.36901632e-01 2.99109250e-01 -3.98034871e-01 1.02458164e-01
-1.27750468e+00 9.15586710e-01 4.97338362e-02 1.68502703e-01
-1.44748187e+00 -8.97293627e-01 -1.12647212e+00 3.27921778e-01
1.02962099e-01 -1.85989887e-01 1.56172991e+00 -6.34551108e-01
-1.75438321e+00 8.39172721e-01 -7.99916387e-01 -5.34018219e-01
1.66501701e-01 -1.76971182e-01 -1.00067270e+00 -1.91723332e-01
-6.20137788e-02 1.60259202e-01 1.34858906e+00 -1.36296952e+00
-4.30322379e-01 -3.58366609e-01 -6.04825497e-01 1.40118137e-01
-3.06315303e-01 8.98746610e-01 -1.40008390e-01 -6.31806433e-01
1.04819834e-01 -6.61170840e-01 3.98398221e-01 -4.73073184e-01
-7.63743281e-01 -3.67177278e-01 1.31163764e+00 -8.55860829e-01
9.67734396e-01 -2.56513429e+00 2.59870559e-01 -9.60005149e-02
-8.24671313e-02 2.93391913e-01 -3.63255501e-01 1.88370347e-01
-3.61676633e-01 -3.74681614e-02 -3.31760868e-02 -1.08449244e+00
3.64523709e-01 -3.00976187e-01 -4.62505281e-01 5.46759248e-01
1.56558007e-01 5.14179707e-01 -3.60633522e-01 -3.72129232e-01
2.80553587e-02 8.98312449e-01 -1.47649676e-01 2.53156841e-01
4.72558051e-01 2.14168251e-01 1.05841570e-01 9.36782777e-01
7.73166120e-01 2.01888770e-01 -8.11815783e-02 -1.19185619e-01
-7.17071146e-02 6.13076806e-01 -1.44300354e+00 1.36530352e+00
-6.75129592e-01 7.61966467e-01 9.49803233e-01 -6.40884042e-01
1.05707836e+00 1.04887545e+00 -1.09710626e-01 -1.02612630e-01
1.40065029e-01 2.58169085e-01 2.92495787e-01 -2.23336518e-01
2.70550638e-01 -4.63792562e-01 -1.62661076e-01 3.45807493e-01
3.99195313e-01 -2.26072613e-02 -4.07295406e-01 2.08498135e-01
7.28461742e-01 -7.73399889e-01 2.07552493e-01 -4.13133889e-01
9.30328786e-01 -8.22031677e-01 5.03646612e-01 5.30596495e-01
-6.42316103e-01 6.09810352e-01 -1.00835487e-01 5.05217731e-01
-3.26382279e-01 -1.19855452e+00 -3.24428350e-01 1.61895537e+00
-1.33417383e-01 -3.05629522e-01 -6.71916544e-01 -3.50979120e-01
8.40618834e-02 9.05764818e-01 -4.43260282e-01 -1.26379564e-01
-4.74268079e-01 -7.61422515e-01 1.01945603e+00 3.88137788e-01
2.38827482e-01 -9.52015340e-01 3.42478245e-01 2.58930892e-01
-4.18596864e-01 -1.20617902e+00 -1.17601430e+00 2.40170836e-01
-1.77475601e-01 -4.85588044e-01 -9.57397640e-01 -1.14136791e+00
1.37080820e-02 3.06640506e-01 6.35811329e-01 -3.76219779e-01
8.57466683e-02 4.57759976e-01 -2.44413123e-01 -7.30541885e-01
-9.42601323e-01 -2.75852066e-02 6.00476742e-01 6.72195852e-01
4.42996651e-01 -3.47084492e-01 -3.20863947e-02 5.85881829e-01
-4.35023487e-01 -5.46222627e-01 4.56209391e-01 1.04555941e+00
1.48133010e-01 -1.42687589e-01 1.18606603e+00 -1.82389483e-01
5.33768535e-01 -2.39950240e-01 -5.19794941e-01 3.67095262e-01
-4.26129729e-01 -4.10351492e-02 4.51525440e-03 -7.77994931e-01
-1.09863544e+00 -4.63068634e-02 -5.15455186e-01 -5.39977610e-01
-2.65682429e-01 1.75213441e-01 -8.41308355e-01 -7.23111704e-02
3.90453070e-01 6.57405853e-01 -7.61141181e-02 -6.48915648e-01
2.11300418e-01 1.41215670e+00 6.07457101e-01 -2.40889028e-01
8.77986670e-01 -9.70679820e-02 -8.80028307e-01 -1.28090465e+00
-6.09196484e-01 -9.06237125e-01 -2.96616048e-01 -3.02866045e-02
8.07276368e-01 -1.27237153e+00 -9.44370151e-01 7.75898516e-01
-1.34126163e+00 -5.31645566e-02 1.79236561e-01 7.08685756e-01
-2.13329598e-01 4.34921354e-01 -6.34235382e-01 -1.29664397e+00
-7.93925703e-01 -1.48519230e+00 1.45034862e+00 -2.56506383e-01
-1.35770336e-01 -6.99555337e-01 1.98529214e-02 6.00646138e-01
7.03169048e-01 -6.69738412e-01 5.27224600e-01 -1.02926672e+00
-2.98747450e-01 -9.78481546e-02 1.15604267e-01 8.02675307e-01
3.79612982e-01 -2.20656842e-01 -1.90727019e+00 -6.37927771e-01
3.36800605e-01 -1.50764748e-01 9.69072402e-01 4.89017099e-01
7.72434950e-01 -3.36068183e-01 -2.52947867e-01 4.46916372e-01
9.56803799e-01 4.47575487e-02 1.66836560e-01 -2.15575397e-01
6.09526217e-01 5.44353426e-01 1.38383687e-01 6.08861297e-02
3.79837096e-01 8.45083296e-01 2.04831585e-01 1.32873267e-01
-4.82613921e-01 6.22852780e-02 1.01376033e+00 1.28879547e+00
4.97195512e-01 -3.86588782e-01 -8.42801392e-01 7.65771031e-01
-1.23228538e+00 -1.10296202e+00 1.23234624e-02 2.25524426e+00
8.14671457e-01 -1.27795801e-01 2.48750120e-01 7.39304602e-01
1.15166509e+00 4.04341340e-01 -3.73671174e-01 -2.56057322e-01
-4.09741104e-01 1.00827236e-02 1.92931473e-01 9.56291139e-01
-1.13621294e+00 8.13545763e-01 6.48535490e+00 7.89587140e-01
-1.28206086e+00 6.70447767e-01 1.99657634e-01 -2.64751256e-01
1.62183717e-02 -5.55402160e-01 -1.45399106e+00 4.25265431e-01
1.47487330e+00 -3.25596780e-01 2.44798318e-01 7.59885311e-01
1.54920056e-01 6.88252211e-01 -1.31740212e+00 1.42157221e+00
5.53177893e-01 -7.07190931e-01 -1.78945765e-01 1.30508356e-02
2.27568462e-01 3.11492980e-01 3.37503612e-01 6.23913825e-01
-7.67009240e-03 -9.56397593e-01 1.12423074e+00 -3.80689621e-01
8.06343973e-01 -6.00000918e-01 7.28088379e-01 2.43926436e-01
-1.41501462e+00 -1.09493501e-01 7.61355013e-02 6.05986357e-01
4.43306327e-01 4.42035288e-01 -1.04117119e+00 4.42926049e-01
7.40459859e-01 2.90343195e-01 -1.53965682e-01 8.25683177e-01
-2.98157156e-01 1.05399632e+00 -2.35221237e-01 2.50786189e-02
-2.33380809e-01 3.41674060e-01 9.50281620e-01 1.43108892e+00
2.84758270e-01 -2.96390027e-01 -1.35537311e-01 6.92435503e-01
-2.07699880e-01 1.63079947e-01 -2.66324580e-01 1.26928583e-01
7.43981004e-01 1.12134671e+00 1.28279254e-01 -4.78285044e-01
-2.87647486e-01 1.02081800e+00 2.45942920e-02 5.35157502e-01
-8.64172459e-01 -5.17913342e-01 1.01965237e+00 -3.58345091e-01
6.54735684e-01 3.46804075e-02 -1.79006398e-01 -1.10062277e+00
1.61508754e-01 -1.15634573e+00 2.67517209e-01 -3.35582137e-01
-1.28753126e+00 9.88642633e-01 -2.96577901e-01 -9.10984576e-01
-9.53991711e-02 -5.08634686e-01 -8.69183302e-01 1.32967162e+00
-1.84776187e+00 -1.18960261e+00 1.95490360e-01 6.34492040e-01
9.06604588e-01 -3.96967977e-01 1.12462878e+00 5.69766760e-01
-8.95957351e-01 1.34233391e+00 1.27460957e-01 2.36633405e-01
9.43393648e-01 -1.23657274e+00 5.69728017e-01 8.32497716e-01
3.72962028e-01 5.46029687e-01 7.13995814e-01 -3.74539243e-03
-1.41493964e+00 -1.03499651e+00 1.29232633e+00 -4.61013258e-01
4.79054272e-01 -8.57634842e-01 -1.13696718e+00 8.18929851e-01
4.34980184e-01 -2.21365079e-01 9.68369067e-01 5.02315760e-01
-8.01911116e-01 -3.35251749e-01 -1.18441641e+00 1.26696691e-01
5.74448526e-01 -1.00569928e+00 -9.47538018e-01 2.35444069e-01
9.93992090e-01 -1.10429667e-01 -7.74370909e-01 8.41979980e-02
3.27028096e-01 -4.00762230e-01 1.17772830e+00 -3.53445679e-01
-5.22477686e-01 -4.17554587e-01 -6.17268503e-01 -1.60836399e+00
-2.64414281e-01 -8.06908786e-01 -1.47161454e-01 1.83790362e+00
7.88643301e-01 -7.77732849e-01 1.70001462e-01 9.27482769e-02
-4.30799365e-01 2.12677475e-02 -1.49873090e+00 -1.17527997e+00
5.84145375e-02 -5.79814494e-01 7.73263991e-01 9.75127816e-01
-7.62046725e-02 7.40416586e-01 -5.17390609e-01 9.07978594e-01
9.26686645e-01 3.90144973e-03 6.06459379e-01 -1.01784408e+00
-3.75483543e-01 -5.03764272e-01 -2.92004019e-01 -9.42674816e-01
6.49320722e-01 -1.01409960e+00 4.82701004e-01 -1.17577410e+00
-1.14292257e-01 1.38142277e-02 -5.81215680e-01 4.10159647e-01
-3.63933623e-01 -1.18420823e-02 3.68166864e-02 -1.29677564e-01
-1.17944658e-01 9.31057036e-01 5.80981195e-01 -6.05355322e-01
-1.48055449e-01 3.30959439e-01 -8.34062159e-01 3.32922697e-01
6.54310405e-01 -5.08700132e-01 -1.07497558e-01 -4.00174975e-01
-8.53062749e-01 -5.42062558e-02 1.30114466e-01 -8.62813890e-01
2.82093763e-01 3.70283008e-01 -2.66835213e-01 -6.73973739e-01
8.43964219e-01 -4.74121988e-01 -3.25378865e-01 2.54217893e-01
-4.40063894e-01 -2.72093952e-01 3.78574938e-01 5.41819572e-01
-6.04138017e-01 -6.90080374e-02 9.86144185e-01 4.39226806e-01
-2.98624098e-01 1.24540396e-01 -4.62732583e-01 -1.84236899e-01
7.20731378e-01 1.51526734e-01 -2.28943065e-01 -3.31429303e-01
-7.78815329e-01 1.98912233e-01 -3.89618248e-01 7.33853579e-01
7.35515475e-01 -1.36502421e+00 -1.48882532e+00 4.20170695e-01
3.20454478e-01 -4.01152730e-01 4.43807989e-01 8.17317784e-01
5.26674628e-01 4.57878381e-01 4.91506755e-01 -8.71254742e-01
-1.98968422e+00 5.73471665e-01 5.34401119e-01 3.89158130e-01
-2.46868104e-01 1.30307150e+00 3.80466431e-01 -6.35651112e-01
7.00234532e-01 -4.38210994e-01 8.55009481e-02 6.20659702e-02
1.04475737e+00 2.96149284e-01 4.20639813e-01 -1.25216651e+00
-7.31217980e-01 3.65251631e-01 -2.90419161e-01 -3.64763290e-01
1.02803326e+00 -1.77623168e-01 2.73987591e-01 7.35188127e-01
1.35618544e+00 3.17517698e-01 -6.06515884e-01 -5.65552890e-01
-1.11793421e-01 -2.33106479e-01 4.43283349e-01 -7.46790648e-01
-8.84141088e-01 1.10750711e+00 8.25195551e-01 2.59381443e-01
7.63298750e-01 3.17322910e-01 6.14536047e-01 2.09694728e-01
2.93964259e-02 -5.77147961e-01 -1.73304394e-01 3.74011993e-01
1.31534493e+00 -1.58797383e+00 -6.09183729e-01 -4.86304611e-01
-7.26370394e-01 6.32407486e-01 3.16550195e-01 5.42077124e-01
9.14454877e-01 3.20784062e-01 5.59075415e-01 -6.33549094e-02
-7.45810688e-01 -5.11843897e-02 5.36538839e-01 5.88455200e-01
5.70804954e-01 2.21729875e-01 5.44651270e-01 7.40143061e-01
-5.23796439e-01 -8.35094094e-01 1.32428080e-01 4.49855208e-01
-4.55656737e-01 -1.00387144e+00 -8.89722288e-01 -1.02947183e-01
-5.83940685e-01 -3.57832342e-01 -4.70350325e-01 3.47002298e-01
-4.89897400e-01 1.56926990e+00 -2.27523744e-01 -2.86140114e-01
4.84679133e-01 6.13668025e-01 1.86351836e-01 -6.62666082e-01
-9.02576447e-01 3.69269311e-01 1.48664981e-01 -3.65057439e-02
-3.26606482e-01 -7.81359911e-01 -8.52108896e-01 -7.01898560e-02
-9.62577760e-01 2.83463657e-01 1.14185488e+00 7.06496477e-01
2.29282841e-01 6.91760182e-01 8.46443176e-01 -7.96134710e-01
-1.09393907e+00 -1.57406449e+00 -8.68225515e-01 5.13907932e-02
1.11793590e+00 -4.15419757e-01 -9.42151427e-01 -4.09474447e-02] | [14.376219749450684, 6.0920891761779785] |
50ec9deb-f314-40f7-895e-9c3af0ae8135 | turning-a-clip-model-into-a-scene-text | 2302.14338 | null | https://arxiv.org/abs/2302.14338v3 | https://arxiv.org/pdf/2302.14338v3.pdf | Turning a CLIP Model into a Scene Text Detector | The recent large-scale Contrastive Language-Image Pretraining (CLIP) model has shown great potential in various downstream tasks via leveraging the pretrained vision and language knowledge. Scene text, which contains rich textual and visual information, has an inherent connection with a model like CLIP. Recently, pretraining approaches based on vision language models have made effective progresses in the field of text detection. In contrast to these works, this paper proposes a new method, termed TCM, focusing on Turning the CLIP Model directly for text detection without pretraining process. We demonstrate the advantages of the proposed TCM as follows: (1) The underlying principle of our framework can be applied to improve existing scene text detector. (2) It facilitates the few-shot training capability of existing methods, e.g., by using 10% of labeled data, we significantly improve the performance of the baseline method with an average of 22% in terms of the F-measure on 4 benchmarks. (3) By turning the CLIP model into existing scene text detection methods, we further achieve promising domain adaptation ability. The code will be publicly released at https://github.com/wenwenyu/TCM. | ['Xiang Bai', 'Bo Ren', 'Deqiang Jiang', 'Wei Hua', 'Yuliang Liu', 'Wenwen Yu'] | 2023-02-28 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Yu_Turning_a_CLIP_Model_Into_a_Scene_Text_Detector_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Yu_Turning_a_CLIP_Model_Into_a_Scene_Text_Detector_CVPR_2023_paper.pdf | cvpr-2023-1 | ['scene-text-detection'] | ['computer-vision'] | [ 2.68955588e-01 -4.37723011e-01 -1.46815836e-01 -1.58074886e-01
-6.34122491e-01 -2.89555818e-01 9.05060470e-01 -3.95196714e-02
-5.32320559e-01 2.19573602e-01 3.62347126e-01 -2.88687825e-01
5.80263257e-01 -6.14134490e-01 -6.40489459e-01 -5.69645703e-01
7.09394336e-01 5.44103533e-02 5.52170634e-01 -1.68789223e-01
4.74393249e-01 -6.09182008e-03 -1.31333137e+00 4.81516719e-01
9.78224814e-01 7.71621644e-01 6.46982372e-01 4.15039569e-01
-4.14449751e-01 9.42986250e-01 -3.00802797e-01 -2.59499937e-01
1.45062104e-01 -4.87622380e-01 -5.97792983e-01 3.83240968e-01
3.49991620e-01 -5.90112269e-01 -5.41476786e-01 1.03833604e+00
4.68482465e-01 7.67530128e-02 5.92391729e-01 -9.23896015e-01
-9.11722958e-01 5.85804224e-01 -8.69989991e-01 -2.46422850e-02
1.96953103e-01 2.59650171e-01 7.65578210e-01 -1.48122895e+00
3.79919112e-01 1.18706322e+00 4.20006394e-01 4.46136266e-01
-7.28080511e-01 -6.61933899e-01 4.19204980e-01 2.31392950e-01
-1.39713299e+00 -4.94463116e-01 8.31413090e-01 -4.98380661e-01
8.88956189e-01 -1.07241748e-02 4.15947497e-01 1.25878155e+00
-8.74965191e-02 1.39101338e+00 9.71430957e-01 -7.38282800e-01
-5.47602735e-02 2.32520506e-01 1.69338927e-01 8.52190793e-01
1.88740671e-01 -1.12624027e-01 -5.84106863e-01 3.15036595e-01
5.75590253e-01 1.72550589e-01 -1.55436441e-01 -1.93159327e-01
-1.30536759e+00 7.57648706e-01 3.87787014e-01 3.51098120e-01
1.37503341e-01 4.97769564e-02 5.97079098e-01 -1.47208825e-01
5.51035702e-01 1.95643660e-02 1.36224488e-02 7.10854530e-02
-1.00616682e+00 -1.07707374e-01 4.76909518e-01 1.14934111e+00
5.99323511e-01 2.03309044e-01 -4.57005441e-01 1.06303442e+00
4.25104588e-01 7.95344949e-01 6.60325885e-01 -3.68201584e-01
7.65050709e-01 7.06506968e-01 -1.28846109e-01 -7.81096697e-01
-1.23040326e-01 -1.80033267e-01 -9.00290906e-01 -1.11942619e-01
7.76286796e-02 -1.42971545e-01 -1.13765442e+00 1.17119443e+00
1.43690616e-01 1.74371392e-01 9.81205516e-03 7.68999338e-01
9.85955894e-01 9.42455351e-01 8.96328315e-02 -7.01656714e-02
1.32361805e+00 -1.55282998e+00 -7.03746617e-01 -5.56748509e-01
5.68438232e-01 -1.02503288e+00 1.56776428e+00 3.10544312e-01
-7.98146605e-01 -5.81589401e-01 -1.03952050e+00 -3.55850667e-01
-4.45550531e-01 5.89917898e-01 5.22030115e-01 5.70481300e-01
-1.07214284e+00 -2.72545777e-02 -8.82028580e-01 -8.01036596e-01
5.54060161e-01 4.06104550e-02 -1.52358618e-02 -3.67533833e-01
-9.11823690e-01 4.60964113e-01 5.76793671e-01 -4.86298138e-03
-9.59056199e-01 -2.95443594e-01 -8.05409789e-01 1.02472857e-01
6.04979575e-01 -6.22184694e-01 1.20396674e+00 -1.01576126e+00
-1.63837302e+00 8.26747060e-01 -3.53127569e-01 -2.04945907e-01
6.11531377e-01 -5.83271801e-01 -2.38504604e-01 3.30343515e-01
1.85789779e-01 8.27854812e-01 9.20838177e-01 -1.10626996e+00
-6.27709627e-01 -7.48217329e-02 -2.07051277e-01 3.32749754e-01
-9.21587765e-01 2.54464060e-01 -1.14035106e+00 -8.14691246e-01
-1.94906935e-01 -7.45603442e-01 -1.04119200e-02 2.61916935e-01
-5.14943182e-01 -3.56454462e-01 1.09783423e+00 -5.02559423e-01
1.15063107e+00 -2.17574286e+00 -2.65719086e-01 -3.98320109e-01
1.66443512e-01 5.55220544e-01 -2.65827477e-01 7.60619104e-01
3.04965407e-01 8.20534378e-02 -2.60463208e-01 -4.71747279e-01
-1.96108390e-02 -2.53108680e-01 -5.50309896e-01 3.15563381e-01
2.47619793e-01 1.03498232e+00 -6.40559494e-01 -8.20544124e-01
5.28767288e-01 2.82283247e-01 -3.87234241e-01 1.45286247e-01
-3.52592826e-01 2.02436253e-01 -5.67056119e-01 5.79168499e-01
6.93282068e-01 -5.20607829e-01 -9.97852832e-02 -4.75888439e-02
-3.31415951e-01 3.47650461e-02 -7.80471683e-01 1.73179781e+00
-1.13284469e-01 8.72775018e-01 -9.37338918e-02 -1.07330310e+00
7.68149197e-01 1.39808983e-01 2.14531362e-01 -8.13733518e-01
2.83205599e-01 -1.45082757e-01 -4.48054582e-01 -6.75961375e-01
5.35202384e-01 8.35506842e-02 8.69833156e-02 3.47484767e-01
-1.53018579e-01 -1.15403878e-02 3.61138612e-01 5.05046189e-01
8.05440128e-01 2.17160791e-01 2.31546849e-01 -9.47412476e-02
7.26318717e-01 1.44939482e-01 2.59224057e-01 8.21479857e-01
-3.20924968e-01 5.80887020e-01 2.12362900e-01 -9.50038731e-02
-9.71087515e-01 -8.36234033e-01 -1.27460480e-01 1.21355569e+00
3.31759185e-01 -5.18558681e-01 -6.41417146e-01 -6.68718457e-01
-1.96103424e-01 5.46724677e-01 -4.47782010e-01 -1.43387122e-03
-4.65374172e-01 -6.98060334e-01 6.07046425e-01 8.47737730e-01
1.11093473e+00 -9.52334940e-01 -1.72657996e-01 -1.63002595e-01
-3.59938800e-01 -1.61798131e+00 -6.49120927e-01 -7.11993799e-02
-9.28783834e-01 -6.62622809e-01 -9.48846996e-01 -1.20854211e+00
6.47481024e-01 9.63164508e-01 6.67515516e-01 1.79666325e-01
-2.55795419e-01 5.35522819e-01 -6.52595878e-01 -5.22036135e-01
-2.13046655e-01 -5.02985641e-02 -2.30091959e-01 6.91193715e-02
4.69064981e-01 -2.05347151e-01 -6.83864951e-01 2.21998990e-01
-9.42530811e-01 6.10513926e-01 7.99118400e-01 8.83225679e-01
4.69532192e-01 -9.64177921e-02 4.27686065e-01 -8.98812711e-01
4.70311940e-01 -1.80963188e-01 -5.54111362e-01 2.42031828e-01
-5.99478781e-01 -1.99128121e-01 6.54687822e-01 -3.27016830e-01
-1.37995601e+00 2.55918115e-01 -1.36355966e-01 -3.76573831e-01
-2.36412585e-01 4.80208218e-01 -3.18695642e-02 -1.05500393e-01
4.27935004e-01 7.26808190e-01 -2.92025894e-01 -4.96537834e-01
3.91691297e-01 9.83580232e-01 4.44018245e-01 -4.59794164e-01
8.97282422e-01 7.09971666e-01 -4.44002151e-01 -1.02543080e+00
-8.84659171e-01 -8.37054551e-01 -7.33020246e-01 -1.22281294e-02
9.03695166e-01 -1.30402124e+00 -3.65026981e-01 7.57382989e-01
-9.47577536e-01 -5.40658116e-01 3.69229347e-01 5.37399173e-01
-3.04486692e-01 1.02214873e+00 -7.18203723e-01 -8.01238477e-01
-4.79475468e-01 -8.33653450e-01 1.34484541e+00 2.70947278e-01
3.64415348e-01 -9.92435575e-01 -1.94990411e-01 4.95738119e-01
3.17351609e-01 -2.66622573e-01 7.36420333e-01 -5.16456962e-01
-5.36053956e-01 -1.49458498e-01 -6.11264884e-01 3.71941894e-01
5.53094735e-03 -9.33219492e-03 -1.03508246e+00 -4.51179504e-01
-3.14323604e-02 -5.07112861e-01 1.15322959e+00 1.94841564e-01
1.20012152e+00 9.36959777e-03 -4.80672240e-01 4.45701987e-01
1.57242823e+00 8.05029869e-02 5.54727197e-01 4.66097206e-01
9.55559433e-01 3.45627964e-01 6.34254813e-01 4.46270555e-01
4.28864777e-01 5.67743182e-01 1.81784213e-01 -4.19111818e-01
-3.44199300e-01 -4.95137721e-01 6.07936263e-01 9.43783402e-01
2.59891957e-01 -4.77729350e-01 -1.08519256e+00 5.25685728e-01
-2.02135992e+00 -8.09574902e-01 -6.79683462e-02 1.89375949e+00
6.06983960e-01 2.07620949e-01 4.93569635e-02 -1.46480143e-01
8.67741406e-01 2.67715812e-01 -5.94654739e-01 5.25444224e-02
-5.37748411e-02 -1.36765137e-01 3.10156792e-01 1.72973245e-01
-1.39690363e+00 1.50247228e+00 5.81443357e+00 9.00487542e-01
-1.20681810e+00 3.95195559e-02 3.86987060e-01 5.30231707e-02
1.19038351e-01 -2.73614842e-02 -1.02533567e+00 4.87489432e-01
5.54759383e-01 -3.47252727e-01 2.04006121e-01 9.31646705e-01
5.27946174e-01 -1.24308541e-01 -9.18532193e-01 1.08997607e+00
6.46163046e-01 -1.24085212e+00 3.82291019e-01 -1.78849414e-01
8.10246408e-01 2.27457285e-01 1.50366262e-01 5.43946683e-01
9.38583985e-02 -7.66105056e-01 5.77362180e-01 1.38762996e-01
9.20665622e-01 -4.29375827e-01 5.37059247e-01 5.62321067e-01
-1.32084131e+00 -7.01808929e-02 -5.38491428e-01 -2.06716768e-02
2.09753886e-02 4.39188004e-01 -8.46506417e-01 5.38211823e-01
6.00249588e-01 1.11629486e+00 -7.86444664e-01 1.19445062e+00
-3.78022552e-01 9.46967185e-01 -1.15285352e-01 -2.05428064e-01
3.80303204e-01 -2.03726903e-01 2.85457879e-01 1.49964774e+00
1.10799387e-01 -9.98055860e-02 5.82347989e-01 7.01579809e-01
-2.46092483e-01 5.14439583e-01 -6.16111159e-01 -2.85844803e-01
2.48677820e-01 1.21327615e+00 -9.22279716e-01 -5.80302477e-01
-9.54338372e-01 1.17804968e+00 1.70866936e-01 4.90548074e-01
-8.00313771e-01 -5.73554754e-01 -8.28425512e-02 -1.13609456e-01
5.65050066e-01 -3.39590579e-01 -3.53048772e-01 -1.41402245e+00
1.41503349e-01 -8.36196125e-01 2.28356749e-01 -1.01011360e+00
-1.18143129e+00 5.14715374e-01 -2.57194489e-01 -1.35384071e+00
1.91584811e-01 -8.38432074e-01 -7.57057071e-01 6.35521352e-01
-1.68298042e+00 -1.55759394e+00 -6.18584454e-01 8.32233667e-01
1.20823729e+00 -1.97189316e-01 4.76206094e-01 2.27373213e-01
-8.69478226e-01 6.34004593e-01 3.48946750e-01 5.83444595e-01
9.83242810e-01 -9.84874725e-01 4.92282093e-01 1.26878905e+00
3.29295367e-01 4.24851924e-01 3.09327245e-01 -6.61781549e-01
-1.49403489e+00 -1.40093172e+00 5.73963821e-01 -4.48370576e-01
7.66176760e-01 -6.22617960e-01 -8.71714294e-01 7.33968973e-01
4.53921050e-01 -2.49175712e-01 4.31087941e-01 -9.93960574e-02
-4.00618911e-01 4.51912805e-02 -6.24206305e-01 8.75751555e-01
9.33837175e-01 -5.21401823e-01 -6.52032733e-01 6.27808869e-01
7.33383000e-01 -2.56221652e-01 -2.52545029e-01 2.37079516e-01
2.92902917e-01 -7.38285065e-01 7.02156425e-01 -2.02365831e-01
6.72077060e-01 -3.25455725e-01 -6.22409098e-02 -7.34401286e-01
-2.42383331e-01 -3.28808665e-01 -5.51788025e-02 1.31550789e+00
1.13107048e-01 -5.02863765e-01 6.83390796e-01 8.48369226e-02
-3.66011530e-01 -5.22318780e-01 -5.23891687e-01 -7.78131306e-01
2.95842052e-01 -4.92626488e-01 2.47594155e-02 8.52807224e-01
1.28644913e-01 6.78475499e-01 -6.51609838e-01 3.86286154e-02
3.62818241e-01 2.23089859e-01 9.04203534e-01 -9.09500301e-01
-2.12885037e-01 -4.18044537e-01 -7.06834793e-02 -1.71905112e+00
4.32948507e-02 -9.83169556e-01 3.69711757e-01 -1.70019710e+00
8.18533719e-01 2.76451111e-02 -1.77302197e-01 5.58232427e-01
-4.60677207e-01 2.47907609e-01 4.54089165e-01 4.02137399e-01
-1.02061582e+00 7.93904662e-01 1.33156633e+00 -2.59995818e-01
-6.41023219e-02 -2.10549638e-01 -7.14005411e-01 7.99259782e-01
8.51936221e-01 -2.20574185e-01 -4.15838301e-01 -6.79723501e-01
-6.47735968e-02 -1.48471177e-01 3.05869281e-01 -1.05089748e+00
4.91589963e-01 -7.95323774e-02 3.86916369e-01 -8.74165833e-01
1.71468034e-01 -5.23186147e-01 -6.49166465e-01 4.20739740e-01
-2.52655834e-01 -2.04896301e-01 4.41655934e-01 8.05724084e-01
-2.56250024e-01 -1.88176975e-01 7.62599528e-01 1.11414231e-02
-1.11736119e+00 1.76122978e-01 -4.91327614e-01 2.80190520e-02
8.94675195e-01 -1.76156193e-01 -5.93321204e-01 -3.47767442e-01
-8.20174217e-02 4.40007836e-01 5.32170177e-01 5.66020310e-01
7.70586789e-01 -9.75807011e-01 -7.11941659e-01 8.95784795e-02
4.24224764e-01 -1.77593663e-01 1.84851915e-01 9.70317960e-01
-4.97612119e-01 7.44777322e-01 -1.64489553e-03 -7.36310184e-01
-1.31094754e+00 8.57411921e-01 1.26197105e-02 -9.67410803e-02
-9.86388266e-01 6.58059359e-01 8.74486506e-01 -1.93765331e-02
3.75380635e-01 -1.63165644e-01 -1.97547913e-01 -3.51092398e-01
7.01126099e-01 8.51182938e-02 -1.62812844e-01 -3.60520244e-01
-3.05625081e-01 7.38325953e-01 -4.67445552e-01 -4.11842428e-02
1.08830798e+00 -3.22259724e-01 1.32685244e-01 4.57756668e-01
9.51881409e-01 1.43013477e-01 -1.27938533e+00 -4.86484855e-01
-4.10960838e-02 -4.48672503e-01 2.18361527e-01 -9.09269035e-01
-8.83281946e-01 1.25904870e+00 6.84488714e-01 -1.07496895e-01
1.23818970e+00 1.76742189e-02 8.63111079e-01 6.22178376e-01
1.22690193e-01 -1.15497160e+00 5.55971265e-01 6.71763599e-01
6.76566303e-01 -1.63443863e+00 6.64734468e-02 -5.12768030e-01
-8.68302643e-01 1.12412310e+00 7.49461830e-01 -8.58927593e-02
2.81498313e-01 2.57705063e-01 2.08198369e-01 -4.97598201e-02
-7.28033423e-01 -4.96601850e-01 2.22966015e-01 4.59437370e-01
5.47243834e-01 -2.50672221e-01 -3.83709431e-01 4.61276680e-01
3.81161034e-01 9.28680450e-02 5.09648442e-01 8.12814176e-01
-7.94870436e-01 -9.48246062e-01 -2.30087802e-01 2.83834249e-01
-3.65900785e-01 -5.78183591e-01 -6.31281018e-01 8.48290741e-01
-2.92007715e-01 1.01267409e+00 -1.78389877e-01 -2.77218372e-01
1.25002876e-01 -2.11993568e-02 7.10332692e-02 -8.30376923e-01
-1.71606094e-01 4.89974529e-01 -1.68526009e-01 -2.87825137e-01
-3.87075067e-01 -4.67313498e-01 -1.17895627e+00 -1.31836116e-01
-4.27489012e-01 -1.17618576e-01 4.59194243e-01 9.72025216e-01
3.52358669e-01 3.93150538e-01 5.52826285e-01 -6.80259705e-01
-3.28651398e-01 -1.08995545e+00 -2.99953789e-01 2.64502376e-01
5.87192327e-02 -5.02709687e-01 -2.81403661e-01 4.58365709e-01] | [11.752411842346191, 2.1353819370269775] |
e2ddd03c-59c8-44ce-86cc-d0e6da0e2ba3 | splenomegaly-segmentation-on-multi-modal-mri | 1811.04045 | null | http://arxiv.org/abs/1811.04045v1 | http://arxiv.org/pdf/1811.04045v1.pdf | Splenomegaly Segmentation on Multi-modal MRI using Deep Convolutional Networks | The findings of splenomegaly, abnormal enlargement of the spleen, is a
non-invasive clinical biomarker for liver and spleen disease. Automated
segmentation methods are essential to efficiently quantify splenomegaly from
clinically acquired abdominal magnetic resonance imaging (MRI) scans. However,
the task is challenging due to (1) large anatomical and spatial variations of
splenomegaly, (2) large inter- and intra-scan intensity variations on
multi-modal MRI, and (3) limited numbers of labeled splenomegaly scans. In this
paper, we propose the Splenomegaly Segmentation Network (SS-Net) to introduce
the deep convolutional neural network (DCNN) approaches in multi-modal MRI
splenomegaly segmentation. Large convolutional kernel layers were used to
address the spatial and anatomical variations, while the conditional generative
adversarial networks (GAN) were employed to leverage the segmentation
performance of SS-Net in an end-to-end manner. A clinically acquired cohort
containing both T1-weighted (T1w) and T2-weighted (T2w) MRI splenomegaly scans
was used to train and evaluate the performance of multi-atlas segmentation
(MAS), 2D DCNN networks, and a 3D DCNN network. From the experimental results,
the DCNN methods achieved superior performance to the state-of-the-art MAS
method. The proposed SS-Net method achieved the highest median and mean Dice
scores among investigated baseline DCNN methods. | ['Prasanna Parvathaneni', 'Camilo Bermudez', 'Michael R. Savona', 'Hyeonsoo Moon', 'Yuankai Huo', 'Tamara K. Moyo', 'Zhoubing Xu', 'Shunxing Bao', 'Richard G. Abramson', 'Bennett A. Landman', 'Albert Assad'] | 2018-11-09 | null | null | null | null | ['splenomegaly-segmentation-on-multi-modal-mri'] | ['medical'] | [ 5.27053559e-03 -1.25111446e-01 2.09222972e-01 -4.77557212e-01
-6.13083422e-01 -4.06224936e-01 5.19350588e-01 1.02178734e-02
-2.08388850e-01 5.71653843e-01 3.44257683e-01 6.17133453e-02
-3.26587185e-02 -5.46901882e-01 -2.43555456e-01 -8.83110523e-01
-6.77929878e-01 1.04051971e+00 1.74693689e-01 7.22853988e-02
-2.14233711e-01 7.00506330e-01 -3.61757368e-01 -2.31513023e-01
1.02845907e+00 1.12546504e+00 -1.77476361e-01 7.27892756e-01
-6.99277967e-02 4.44528311e-01 -3.56175989e-01 -1.87798113e-01
4.03923333e-01 -7.18288183e-01 -4.79404271e-01 2.18525440e-01
3.72264683e-01 -6.21792138e-01 -4.75737780e-01 9.88078296e-01
9.16513801e-01 -9.74362940e-02 7.67884195e-01 -1.09877872e+00
-6.64577901e-01 7.64976323e-01 -8.16490948e-01 6.45233631e-01
-2.26577386e-01 4.41299707e-01 -4.76348307e-03 -6.30512595e-01
3.29453409e-01 8.85600507e-01 8.38036299e-01 2.92449713e-01
-1.25416684e+00 -7.10756123e-01 -6.53920233e-01 -6.56189859e-01
-1.09153569e+00 1.52716339e-01 7.67050922e-01 -4.86129910e-01
4.07898277e-01 1.47996902e-01 9.40857351e-01 6.87756598e-01
5.85876048e-01 7.49177516e-01 1.42319596e+00 9.82346386e-02
-1.26868114e-01 -6.73587799e-01 -2.03447998e-01 9.18268263e-01
1.20769210e-01 1.50709495e-01 5.14677241e-02 -3.31472039e-01
1.47921729e+00 2.16175601e-01 -1.25545159e-01 -6.34307623e-01
-1.69169164e+00 9.65768337e-01 6.76985681e-01 2.75770277e-01
-8.55541408e-01 1.30623266e-01 7.14052856e-01 -1.84088778e-02
4.11787540e-01 2.33315319e-01 -1.37819678e-01 2.99700081e-01
-1.28613174e+00 -9.12280232e-02 5.81590116e-01 5.01442730e-01
-1.23695001e-01 4.29020911e-01 -3.75694960e-01 8.19062650e-01
2.28651613e-01 7.57352829e-01 8.24445963e-01 -4.90494907e-01
5.72368652e-02 3.56681436e-01 -4.93163854e-01 -7.47399032e-01
-9.01347101e-01 -3.38280082e-01 -1.39912188e+00 -1.17730588e-01
6.25970185e-01 -1.93967357e-01 -1.29158902e+00 1.47687149e+00
6.35959089e-01 -1.35678733e-02 -4.28603858e-01 1.42375481e+00
9.98654783e-01 1.34719178e-01 4.30419624e-01 -2.01959729e-01
1.32004750e+00 -1.08322012e+00 -5.87793231e-01 1.97557732e-01
4.48227674e-01 -6.72706008e-01 8.17623734e-01 -7.13463128e-02
-1.36647534e+00 7.45635927e-02 -7.54087508e-01 4.52369094e-01
5.25376052e-02 5.58452494e-02 8.48537862e-01 5.35048246e-01
-1.03231132e+00 5.96317232e-01 -1.31332743e+00 -2.43239075e-01
6.47753119e-01 5.49837053e-01 -3.10717940e-01 5.60387969e-02
-1.12863159e+00 7.24618971e-01 2.33430371e-01 1.49969272e-02
-1.25566113e+00 -1.14189136e+00 -8.12310576e-01 -6.67611510e-02
1.60801187e-01 -8.21577132e-01 8.45643818e-01 -6.60891652e-01
-1.37840176e+00 9.76079345e-01 5.27356327e-01 -3.67478698e-01
6.84889853e-01 3.34649563e-01 -3.66216242e-01 7.30860054e-01
7.43433312e-02 8.19088340e-01 6.79668188e-01 -1.06799066e+00
1.99572906e-01 -7.45382011e-01 -6.43522978e-01 7.68109709e-02
4.84873891e-01 2.67795920e-01 -1.15987370e-02 -1.16740906e+00
3.88658553e-01 -9.50239062e-01 -1.62107825e-01 1.98609397e-01
-5.06188035e-01 1.68094769e-01 6.40592039e-01 -1.26330340e+00
7.41900742e-01 -1.69206905e+00 -9.84360054e-02 2.85060495e-01
6.84441924e-01 2.25801483e-01 -9.67127755e-02 -2.65906364e-01
-5.63217103e-02 1.23108909e-01 -4.84057248e-01 3.20515782e-02
-6.23929873e-02 5.55441715e-02 2.65168786e-01 9.12981689e-01
-2.59694252e-02 1.39263368e+00 -8.92415941e-01 -9.43576396e-01
5.07276177e-01 6.59149051e-01 -1.43775403e-01 3.96789163e-01
1.98815241e-01 1.17170238e+00 -4.79654908e-01 9.55447733e-01
7.77272999e-01 -4.29311514e-01 9.05064568e-02 -2.40432531e-01
3.16240460e-01 -4.13203061e-01 -5.07647514e-01 1.91978097e+00
-1.27716780e-01 -1.45561025e-01 4.44110990e-01 -8.18346262e-01
8.48226488e-01 5.59554756e-01 1.06257224e+00 -8.43957186e-01
4.98498887e-01 3.37176383e-01 1.92035884e-01 -4.54381436e-01
-2.54315555e-01 -7.34658182e-01 -4.58715223e-02 7.71124780e-01
1.43806800e-01 -3.86239052e-01 1.53042048e-01 1.35080144e-01
8.68518531e-01 -2.44209126e-01 1.42976075e-01 -5.21479368e-01
6.44511104e-01 -1.71791509e-01 5.35078168e-01 4.63605165e-01
-8.60457063e-01 8.91035736e-01 6.54275298e-01 -7.90263772e-01
-1.18988204e+00 -1.57832015e+00 -3.40558738e-01 8.09519291e-01
4.87739500e-03 5.36032021e-01 -7.26195931e-01 -1.03176904e+00
2.41520166e-01 2.20476076e-01 -7.00623631e-01 2.90810704e-01
-7.28568852e-01 -1.09015298e+00 1.04233742e+00 6.70804322e-01
5.36290407e-01 -1.08904648e+00 -5.60785055e-01 2.29232848e-01
-9.12466496e-02 -7.76681781e-01 -9.69720602e-01 -8.74494612e-02
-1.25923979e+00 -1.18925905e+00 -1.60168028e+00 -6.98653340e-01
8.80302906e-01 -3.33801240e-01 1.21796954e+00 -2.44559664e-02
-6.02082074e-01 2.15973511e-01 1.34828627e-01 -5.28618656e-02
-3.99537057e-01 -3.33522737e-01 3.86169553e-02 -3.29065800e-01
5.07547520e-02 -7.29212046e-01 -1.11021352e+00 3.59170705e-01
-1.23846436e+00 7.56433159e-02 8.75255346e-01 1.07051754e+00
9.22068238e-01 -3.63277525e-01 6.88601851e-01 -7.96758473e-01
6.99184299e-01 -5.30846119e-01 -5.11151612e-01 4.22596663e-01
-7.07775652e-01 -1.25289723e-01 5.66403270e-01 -5.13628125e-01
-8.81853104e-01 -8.33862647e-02 5.91225401e-02 -3.98192972e-01
-2.17019573e-01 5.12308002e-01 4.60590988e-01 -3.86740804e-01
1.59283638e-01 5.71733952e-01 6.06179178e-01 -2.19848439e-01
1.98076576e-01 1.85344055e-01 6.92871630e-01 -5.09341240e-01
2.35322028e-01 5.55167437e-01 3.91684622e-01 -3.82738948e-01
-3.69496495e-01 -2.06884682e-01 -9.51006055e-01 -1.98944971e-01
1.04236078e+00 -5.83877325e-01 -3.79297316e-01 7.21203804e-01
-5.82098663e-01 -4.80815500e-01 -2.25592613e-01 6.67969167e-01
-5.79330206e-01 6.11519694e-01 -1.20944834e+00 -7.18076155e-02
-1.22059870e+00 -1.59617770e+00 8.74895990e-01 3.06110501e-01
3.06294542e-02 -1.47685766e+00 3.65871154e-02 3.71423155e-01
1.22165167e+00 9.72828805e-01 1.18715215e+00 -1.11902010e+00
3.94936092e-02 -2.98419803e-01 -5.25738299e-01 2.86300033e-01
3.29025388e-02 -5.67270756e-01 -1.75811052e-01 -4.31849182e-01
5.73220812e-02 -5.92296839e-01 4.66562539e-01 1.11003160e+00
1.14864588e+00 5.12189791e-03 2.92193711e-01 9.60674047e-01
1.33975661e+00 1.47477668e-02 2.49406725e-01 -1.23338461e-01
6.15831912e-01 2.00810626e-01 1.81267306e-03 6.93525374e-01
3.98810565e-01 1.65370166e-01 4.16705966e-01 -5.98582149e-01
-3.61367404e-01 -7.61553049e-02 -2.34486878e-01 1.05344224e+00
8.04434568e-02 2.14171022e-01 -1.19537783e+00 5.26373625e-01
-1.19547915e+00 -3.34733129e-01 -1.24132968e-01 2.09783888e+00
8.03430378e-01 -4.16668594e-01 3.38472366e-01 -6.10647202e-01
8.49069536e-01 2.02739879e-01 -7.24439621e-01 -1.66962042e-01
-1.46962464e-01 2.84767509e-01 6.44608140e-01 1.93806589e-02
-1.21315074e+00 2.52610594e-01 6.42648411e+00 5.60683668e-01
-1.43116212e+00 4.52520370e-01 8.62534523e-01 -2.29152709e-01
-1.39936119e-01 -4.25487965e-01 -8.02758485e-02 6.93964720e-01
6.36031866e-01 1.54832955e-02 5.97383380e-01 3.25901419e-01
4.86937575e-02 -2.22953334e-01 -8.15495133e-01 6.77360773e-01
2.19491616e-01 -8.90480280e-01 -2.53098439e-02 -5.65976910e-02
8.39096904e-01 4.26371336e-01 3.13245505e-02 2.73537070e-01
2.52806872e-01 -1.30354166e+00 2.47513894e-02 4.52281922e-01
1.29320765e+00 -6.36127949e-01 1.06832421e+00 -3.57661285e-02
-9.54576850e-01 5.95033884e-01 8.78015626e-03 7.42128491e-01
3.03367287e-01 5.66941917e-01 -8.22359324e-01 2.43439674e-01
2.52159864e-01 1.88812524e-01 -4.83379990e-01 1.13347030e+00
-2.62726754e-01 5.35328031e-01 -4.02494133e-01 3.70772064e-01
4.98708934e-01 -4.82369483e-01 5.41840971e-01 1.24799109e+00
1.74399078e-01 2.05163807e-01 3.46261263e-01 1.14546418e+00
-3.22659791e-01 4.78535667e-02 -4.07805815e-02 -1.56239688e-01
-4.69460636e-02 1.58843184e+00 -1.31066942e+00 -4.40149814e-01
-3.74713331e-01 7.39703178e-01 -2.75045067e-01 2.97431499e-01
-9.44606125e-01 -6.10359060e-03 -2.54413486e-01 2.00446308e-01
-1.15770206e-01 9.85325649e-02 -4.04000551e-01 -1.14162934e+00
-3.97806168e-01 -5.79909742e-01 6.60938144e-01 -5.72357178e-01
-1.57787049e+00 3.15428048e-01 -1.09622046e-01 -6.41728461e-01
6.70245662e-02 -3.25960740e-02 -7.53883421e-01 1.01235986e+00
-1.60980070e+00 -1.39275050e+00 -4.90764230e-01 4.56675917e-01
2.13166714e-01 -1.22642078e-01 6.61999702e-01 2.41567761e-01
-3.24498385e-01 4.02882248e-01 1.53593421e-01 6.65448010e-01
7.04592526e-01 -1.60759139e+00 -9.41984802e-02 5.47255695e-01
-8.13109457e-01 5.58061182e-01 1.53650880e-01 -9.86094415e-01
-1.37041914e+00 -1.07572663e+00 4.33141053e-01 -4.78306077e-02
7.41931915e-01 2.65143722e-01 -7.88334012e-01 7.67617822e-01
2.39340454e-01 9.04593289e-01 8.17270160e-01 -5.65139592e-01
7.56162852e-02 3.41985583e-01 -1.94856417e+00 2.99108084e-02
2.57014811e-01 -4.62484173e-02 -6.75696194e-01 2.01027453e-01
4.17632610e-01 -9.05437768e-01 -1.77902901e+00 5.84567428e-01
7.25084424e-01 -6.49952769e-01 1.16741061e+00 -4.58860606e-01
7.14790106e-01 1.65459782e-01 2.60837048e-01 -1.50454283e+00
-1.92930996e-01 -2.74230778e-01 -3.38076562e-01 7.08540916e-01
-1.39624462e-01 -5.44024587e-01 7.58404732e-01 7.74778545e-01
-2.35588476e-01 -1.03704727e+00 -9.56781685e-01 -5.03269494e-01
7.00171173e-01 3.46165746e-01 7.02887356e-01 1.13006973e+00
-3.75342518e-01 -1.71694800e-01 -8.42567012e-02 -8.34115371e-02
1.16870272e+00 1.19067356e-01 2.03953966e-01 -1.06994569e+00
1.40352711e-01 -7.42883563e-01 -2.05265522e-01 -5.36148310e-01
-1.79414630e-01 -1.22866392e+00 -3.13053787e-01 -1.52868485e+00
6.87303126e-01 -6.14680588e-01 -7.69558549e-01 2.31118590e-01
-1.87077150e-01 4.94750768e-01 2.14830190e-01 2.56344020e-01
-7.15460896e-01 5.75724185e-01 2.14523244e+00 -2.00447083e-01
-2.49306280e-02 -1.59185216e-01 -2.05789611e-01 5.32289922e-01
7.83378422e-01 -4.40590143e-01 -2.59889960e-02 -2.69275039e-01
-3.95264804e-01 8.28976631e-01 5.70778489e-01 -7.28162408e-01
1.41765431e-01 1.83794856e-01 1.01567018e+00 -7.93663800e-01
-1.63488626e-01 -7.18263149e-01 8.23031440e-02 9.23142672e-01
-3.96636248e-01 1.25330523e-01 5.69347758e-03 9.91013199e-02
-1.63988024e-01 2.16033086e-01 1.25754762e+00 -5.11322856e-01
-4.47836109e-02 8.89717638e-01 -1.93888888e-01 4.99276221e-01
9.71227467e-01 1.15007199e-01 -6.61165491e-02 -2.22529352e-01
-8.91225457e-01 5.71870089e-01 3.57814163e-01 -3.39388214e-02
8.19115937e-01 -1.36478329e+00 -9.92614985e-01 3.68401259e-01
-2.96674758e-01 2.05575302e-01 5.00565767e-01 2.06099343e+00
-1.17576170e+00 6.67047322e-01 -6.08863235e-01 -8.16629708e-01
-7.11556971e-01 3.55420530e-01 5.98500371e-01 -9.27734673e-01
-8.47646654e-01 7.66621649e-01 3.52096766e-01 -9.38109279e-01
1.06505133e-01 -2.78200537e-01 1.67937487e-01 -3.90384972e-01
3.25364351e-01 4.31785583e-01 -1.07915401e-01 -7.92962313e-01
-2.44590759e-01 2.83698350e-01 1.74886025e-02 3.77590001e-01
1.42274785e+00 -1.71329796e-01 -3.08596164e-01 -1.44592851e-01
1.05276060e+00 -4.41087037e-01 -1.27615249e+00 -4.33450639e-01
-3.67972821e-01 -3.43705684e-01 5.01105487e-01 -1.31456864e+00
-1.68541706e+00 9.70181346e-01 8.79776835e-01 -9.92099009e-03
1.07958984e+00 -1.45285830e-01 1.44422948e+00 -6.52642429e-01
3.01976234e-01 -5.48625231e-01 -2.16149613e-02 1.30581126e-01
8.23027790e-01 -1.48755777e+00 1.41833141e-01 -1.37919962e-01
-1.03545392e+00 1.31001759e+00 4.40775275e-01 -2.25836530e-01
4.68785971e-01 5.18326342e-01 2.42412269e-01 -4.85664368e-01
1.06158540e-01 3.03919405e-01 3.51447880e-01 5.19696414e-01
5.00685215e-01 3.98541242e-01 -4.01944607e-01 7.17955947e-01
3.57283920e-01 1.87470898e-01 1.24666974e-01 7.19385684e-01
-1.76357940e-01 -5.19242942e-01 -4.83876199e-01 8.55587542e-01
-9.54683840e-01 -1.50343254e-01 -1.87903047e-01 9.97075021e-01
-1.83376133e-01 2.27114648e-01 -7.78099597e-02 2.35496625e-01
-4.49201800e-02 -1.22599667e-02 5.83923876e-01 -2.66184565e-02
-1.03494668e+00 4.99372780e-01 -5.54492533e-01 -5.59268057e-01
-1.55502886e-01 -5.47711790e-01 -1.59097910e+00 -4.93163258e-01
-3.34069990e-02 -1.98521614e-01 8.11318815e-01 6.63388968e-01
3.11884489e-02 6.33384764e-01 7.88290739e-01 -7.40037143e-01
-8.88622701e-01 -1.19904864e+00 -1.04512441e+00 7.67796218e-01
5.98821826e-02 -5.47817945e-01 -2.56861657e-01 -1.45787492e-01] | [14.362770080566406, -2.4911022186279297] |
d2987e2e-79a8-4f5b-8348-6de7314c0f3e | neural-person-search-machines | 1707.06777 | null | http://arxiv.org/abs/1707.06777v1 | http://arxiv.org/pdf/1707.06777v1.pdf | Neural Person Search Machines | We investigate the problem of person search in the wild in this work. Instead
of comparing the query against all candidate regions generated in a query-blind
manner, we propose to recursively shrink the search area from the whole image
till achieving precise localization of the target person, by fully exploiting
information from the query and contextual cues in every recursive search step.
We develop the Neural Person Search Machines (NPSM) to implement such recursive
localization for person search. Benefiting from its neural search mechanism,
NPSM is able to selectively shrink its focus from a loose region to a tighter
one containing the target automatically. In this process, NPSM employs an
internal primitive memory component to memorize the query representation which
modulates the attention and augments its robustness to other distracting
regions. Evaluations on two benchmark datasets, CUHK-SYSU Person Search dataset
and PRW dataset, have demonstrated that our method can outperform current
state-of-the-arts in both mAP and top-1 evaluation protocols. | ['Shuicheng Yan', 'Karlekar Jayashree', 'Jianguo Jiang', 'Meibin Qi', 'Hao Liu', 'Zequn Jie', 'Jiashi Feng', 'Bo Zhao'] | 2017-07-21 | neural-person-search-machines-1 | http://openaccess.thecvf.com/content_iccv_2017/html/Liu_Neural_Person_Search_ICCV_2017_paper.html | http://openaccess.thecvf.com/content_ICCV_2017/papers/Liu_Neural_Person_Search_ICCV_2017_paper.pdf | iccv-2017-10 | ['person-search'] | ['computer-vision'] | [ 1.31793767e-01 6.29625050e-03 -8.69012922e-02 -2.09108680e-01
-9.40014243e-01 -4.20754761e-01 5.71620405e-01 6.97008744e-02
-8.49743783e-01 4.45386559e-01 3.19507778e-01 1.84097901e-01
-7.65509009e-02 -6.88585520e-01 -5.48621237e-01 -5.37461519e-01
6.09854013e-02 9.60846186e-01 5.22540033e-01 -2.17612609e-02
1.08823799e-01 1.88533723e-01 -1.53022718e+00 2.20380425e-02
8.48370492e-01 9.11905468e-01 4.83997375e-01 4.65182960e-01
2.56893128e-01 5.93206763e-01 -7.40382254e-01 -4.62201774e-01
3.04460555e-01 -2.16324851e-01 -1.02242851e+00 -1.07441448e-01
7.63112247e-01 -3.32072169e-01 -7.30409563e-01 1.06593621e+00
1.06757009e+00 5.91977358e-01 2.26234704e-01 -7.63808906e-01
-1.02682960e+00 3.85369509e-01 -5.86036861e-01 5.48260391e-01
7.06914246e-01 3.30519587e-01 9.92076159e-01 -1.13873100e+00
5.18622458e-01 1.30418015e+00 3.01049173e-01 8.87970805e-01
-1.12887931e+00 -6.27157092e-01 4.32420820e-01 3.10598850e-01
-1.82879066e+00 -5.89358866e-01 3.15617651e-01 -8.00791457e-02
9.58299041e-01 3.78822684e-01 5.78889191e-01 1.13828421e+00
-3.62632394e-01 1.03054810e+00 8.65237176e-01 -3.86577338e-01
1.75024316e-01 -1.92689430e-02 2.50618935e-01 7.30493724e-01
-4.35965024e-02 3.04143876e-01 -9.70403373e-01 -3.25260669e-01
5.47692597e-01 -1.25479132e-01 -6.50661886e-01 -4.46256548e-01
-1.22612906e+00 5.75993955e-01 9.00873959e-01 1.16196595e-01
-5.12919903e-01 1.53923899e-01 1.23271227e-01 2.28311941e-02
2.05552325e-01 5.36653459e-01 -7.39506111e-02 1.36777237e-01
-1.22474742e+00 4.95657444e-01 4.21628922e-01 8.62548292e-01
5.79130888e-01 -6.84750497e-01 -8.95236135e-01 9.29264247e-01
1.21069230e-01 7.34159887e-01 6.00647986e-01 -7.74946451e-01
4.90948141e-01 6.94872618e-01 4.15251762e-01 -9.00302589e-01
-2.45883241e-01 -8.39275837e-01 -6.50823414e-01 -2.55988181e-01
2.58707643e-01 2.03066736e-01 -1.28418696e+00 1.92609859e+00
4.81118262e-01 1.36632770e-01 -1.30085468e-01 1.37370133e+00
9.26019847e-01 4.48273659e-01 2.37552732e-01 2.10159242e-01
1.65119457e+00 -1.47914875e+00 -2.22414151e-01 -5.70005476e-01
1.75642401e-01 -1.75174311e-01 1.16973615e+00 4.97499704e-02
-1.22609174e+00 -7.43033051e-01 -7.94963479e-01 -2.18865633e-01
-4.88552988e-01 2.65938967e-01 2.46596545e-01 3.85530114e-01
-1.35545945e+00 9.32233334e-02 -5.31149149e-01 -6.83294296e-01
5.59857845e-01 4.42307323e-01 -4.55275655e-01 -1.78201482e-01
-1.29666281e+00 6.95380092e-01 2.73559988e-01 3.05195957e-01
-9.65499043e-01 -5.51340103e-01 -7.71387100e-01 4.22343135e-01
5.90994716e-01 -1.06942415e+00 1.33676052e+00 -6.23642027e-01
-1.01778507e+00 1.14889145e+00 -6.24023139e-01 -6.85798109e-01
5.44149399e-01 -3.68638247e-01 -2.68223137e-01 3.25840026e-01
4.70298737e-01 1.04259694e+00 7.98796356e-01 -1.04629481e+00
-6.61313891e-01 -5.78038275e-01 -4.65301424e-02 5.37865281e-01
-1.95747495e-01 1.49695337e-01 -1.41467535e+00 -5.19215763e-01
3.42826359e-02 -8.68073344e-01 -2.85979599e-01 -1.41696289e-01
-5.95005929e-01 -5.19773185e-01 3.46405804e-01 -4.80449766e-01
1.33150470e+00 -1.85120726e+00 1.96475938e-01 4.87650096e-01
4.00556237e-01 2.52742916e-01 -2.42425278e-01 1.86043948e-01
4.12480295e-01 -2.79457062e-01 5.26105426e-02 -7.99050987e-01
9.55231413e-02 1.13665813e-03 -3.66421789e-01 4.38687861e-01
-2.27781847e-01 1.35005224e+00 -8.92087758e-01 -6.36858761e-01
-6.89013973e-02 3.47064674e-01 -3.65413696e-01 2.06005678e-01
-5.38372770e-02 3.66785377e-01 -5.08682430e-01 8.70351017e-01
6.56375229e-01 -6.78601980e-01 -3.37587386e-01 2.05460504e-01
1.06671959e-01 8.69677588e-02 -8.82397115e-01 1.93109572e+00
-1.43240184e-01 3.91915977e-01 1.17613196e-01 -5.09203851e-01
6.33218348e-01 -4.24603149e-02 1.47846133e-01 -1.19182611e+00
-2.30990753e-01 -9.82613415e-02 -3.34648162e-01 -2.61878580e-01
8.04971278e-01 6.81898415e-01 -1.38847679e-01 3.41813982e-01
2.24674214e-02 8.43624175e-01 1.14367343e-01 3.29635710e-01
1.17688453e+00 -9.56010818e-02 4.71779555e-02 -2.83857763e-01
7.64385641e-01 -1.98120065e-02 4.23850358e-01 1.48131812e+00
-5.99473119e-01 5.44828176e-01 -1.09593593e-01 -5.03374040e-01
-5.30231059e-01 -1.35124409e+00 1.28767058e-01 1.64333177e+00
6.91785097e-01 -4.40173298e-01 -8.00872803e-01 -7.68221915e-01
1.25628104e-02 3.79069477e-01 -8.07956040e-01 -2.21740723e-01
-7.15469897e-01 -4.67142552e-01 6.12737238e-01 4.04889166e-01
9.98240054e-01 -1.32860959e+00 -1.02705193e+00 -5.60843535e-02
-3.78376007e-01 -9.06944096e-01 -1.07548630e+00 -1.22542679e-01
-1.50723323e-01 -9.85688746e-01 -1.21271336e+00 -1.01114118e+00
7.32579648e-01 4.35685158e-01 1.38417721e+00 2.62191236e-01
-3.22574914e-01 4.86584365e-01 -8.41992721e-02 -2.75196731e-02
1.48656398e-01 3.54968995e-01 -3.27385478e-02 -1.29865393e-01
6.26537025e-01 -1.22397393e-01 -1.22437954e+00 4.97317106e-01
-3.99865896e-01 -1.04449280e-01 6.23569310e-01 7.70782828e-01
7.23286033e-01 -2.14464497e-03 3.50495994e-01 -4.43342179e-01
7.06108093e-01 -3.13228041e-01 -5.92970133e-01 6.69663072e-01
-5.71764171e-01 6.44211546e-02 -4.32213210e-02 -4.72673118e-01
-9.11616087e-01 1.33601218e-01 1.04797669e-01 -3.24265748e-01
-1.77576151e-02 1.04744978e-01 -8.51092711e-02 -1.31302364e-02
7.96396375e-01 7.61486113e-01 -4.88698602e-01 -5.40406883e-01
3.08033019e-01 4.36625361e-01 1.19465697e+00 -5.78469038e-01
7.91397572e-01 5.10499656e-01 -4.15128380e-01 -3.14359903e-01
-7.93856859e-01 -9.17377353e-01 -4.72474039e-01 -3.02035123e-01
9.18466449e-01 -1.21508980e+00 -1.02717447e+00 2.91844755e-01
-1.06455410e+00 -2.24224806e-01 -1.57342404e-01 -2.75240876e-02
-1.55761838e-01 1.43107325e-01 -4.82681572e-01 -8.62219751e-01
-6.68623745e-01 -1.17447078e+00 1.53376973e+00 5.95689535e-01
-1.73142940e-01 -4.74736094e-01 2.70596385e-01 4.27756250e-01
5.07556319e-01 -2.10148856e-01 3.90668273e-01 -7.72048771e-01
-9.62430000e-01 -3.58249754e-01 -3.80260319e-01 -3.86899143e-01
-2.58083791e-01 -1.05912483e+00 -1.00050080e+00 -5.19021630e-01
-3.53745580e-01 -3.36578995e-01 1.39631391e+00 1.91841200e-01
1.06091571e+00 -2.35964492e-01 -9.05398190e-01 6.75828576e-01
1.19270015e+00 -1.71050683e-01 5.05224943e-01 4.62422460e-01
3.34851891e-01 3.32855135e-01 4.89836454e-01 1.31746218e-01
6.52629375e-01 9.78824258e-01 2.64421493e-01 -1.85439557e-01
-3.20660949e-01 -5.89337409e-01 8.76424983e-02 -4.58986461e-01
5.84685393e-02 -4.78626668e-01 -8.66552591e-01 7.02210426e-01
-2.12822104e+00 -1.09075844e+00 5.05650043e-01 2.39110732e+00
7.46589959e-01 -5.24270013e-02 2.63161868e-01 -5.38984895e-01
7.98437417e-01 3.07458967e-01 -5.57353556e-01 2.54166603e-01
-2.89061904e-01 9.44900885e-02 4.03375715e-01 4.51374948e-01
-1.32863271e+00 1.23357582e+00 6.25498962e+00 7.66410768e-01
-7.03088164e-01 1.99592009e-01 5.45509279e-01 -4.56687242e-01
6.61346391e-02 -2.77311772e-01 -1.04609704e+00 5.29002130e-01
5.16775906e-01 1.06031723e-01 6.24520481e-01 8.79393160e-01
6.78822920e-02 -4.19773340e-01 -1.19600928e+00 1.23598015e+00
2.90977061e-01 -1.07733321e+00 7.01600388e-02 5.31707928e-02
5.38199902e-01 1.96723074e-01 3.19707513e-01 5.57460308e-01
2.38726258e-01 -1.25106180e+00 7.60038078e-01 7.33286619e-01
4.77300167e-01 -5.23612201e-01 5.85884392e-01 4.45219487e-01
-1.24737906e+00 -2.16524720e-01 -3.59157145e-01 2.26930335e-01
1.50187790e-01 1.47264436e-01 -7.88393497e-01 1.06627904e-01
1.11848664e+00 5.77813052e-02 -1.05385685e+00 1.27729893e+00
-1.18559048e-01 8.47518221e-02 -3.83748561e-01 1.09704643e-01
9.80814546e-02 1.91304177e-01 6.86781287e-01 1.20748341e+00
1.80466115e-01 -7.54677653e-02 3.90809119e-01 1.10527563e+00
-1.60328418e-01 9.46703181e-03 -2.03768954e-01 5.78048766e-01
7.30964124e-01 9.11414862e-01 -5.06041527e-01 -4.79996026e-01
-1.51119083e-01 1.59260368e+00 7.14385033e-01 7.17569351e-01
-8.26245368e-01 -8.66227597e-02 4.30301398e-01 1.12427309e-01
5.32703936e-01 1.36756152e-01 1.46787733e-01 -1.01502895e+00
3.62509608e-01 -7.21035302e-01 8.06513846e-01 -8.60657215e-01
-9.83980596e-01 7.86017656e-01 -1.30711317e-01 -6.31667793e-01
-1.99708402e-01 -5.79969920e-02 -4.78253514e-01 1.33881938e+00
-1.42005277e+00 -1.38619804e+00 -5.14649987e-01 7.62848675e-01
6.01046860e-01 -2.19024017e-01 7.16995835e-01 2.39377633e-01
-5.09438813e-01 1.01658726e+00 -4.28256989e-01 3.62264991e-01
7.91666985e-01 -1.11841857e+00 5.27552366e-01 1.03634202e+00
2.77789891e-01 9.57409739e-01 5.48471928e-01 -7.25790083e-01
-9.15156066e-01 -8.96936595e-01 1.13718033e+00 -6.86607182e-01
1.49347991e-01 -5.64288855e-01 -7.45072246e-01 4.01596725e-01
2.54715562e-01 6.87758625e-02 3.14635396e-01 2.44452357e-01
-5.87380171e-01 5.16322665e-02 -1.06809950e+00 6.91035032e-01
1.30189836e+00 -6.56875551e-01 -5.80202341e-01 2.92261213e-01
7.53163815e-01 -5.66834509e-01 -2.38545194e-01 2.60478199e-01
5.42249680e-01 -9.89677191e-01 1.40657115e+00 -4.49977130e-01
-1.28435850e-01 -4.64837193e-01 2.79145334e-02 -1.02356172e+00
-5.82308710e-01 -6.54358864e-01 -4.36238766e-01 9.73361254e-01
4.94229168e-01 -4.13430810e-01 1.00166655e+00 8.36542606e-01
2.22095340e-01 -7.40766883e-01 -1.19010210e+00 -5.67001164e-01
-3.66824985e-01 1.60640646e-02 7.40542650e-01 2.11812586e-01
-2.57059127e-01 3.56973439e-01 -4.77509379e-01 5.83213627e-01
7.28969872e-01 2.47123599e-01 7.38830864e-01 -9.75655556e-01
-6.50366127e-01 -4.66144562e-01 -2.03106582e-01 -1.62010205e+00
1.27444118e-01 -7.31731892e-01 3.17587376e-01 -1.49479342e+00
7.04019547e-01 -2.81043202e-01 -4.93164897e-01 4.28841799e-01
-6.05896711e-01 2.21690342e-01 3.06276143e-01 5.16103923e-01
-1.29275572e+00 4.36095804e-01 9.50002968e-01 -3.81619841e-01
-3.98110449e-01 2.17954695e-01 -8.23914826e-01 2.52193898e-01
4.44684267e-01 -2.30311230e-01 -3.64983261e-01 -6.97762072e-01
-1.34526286e-02 -2.29612947e-01 9.08680081e-01 -1.08472371e+00
8.62178624e-01 3.08373570e-01 6.52966917e-01 -7.71572828e-01
5.99860668e-01 -5.90192735e-01 -2.82974601e-01 4.44159627e-01
-7.03964829e-01 1.36532843e-01 -6.92703724e-02 7.86231160e-01
-3.20882723e-02 4.41311449e-02 5.51908791e-01 -2.24637136e-01
-1.07845402e+00 5.05904555e-01 1.12533614e-01 -2.31740586e-02
8.01135838e-01 -1.58766270e-01 -4.26763296e-01 -4.68775004e-01
-8.00591648e-01 8.62118185e-01 4.94407684e-01 3.68706435e-01
6.77349031e-01 -1.17963755e+00 -5.76518595e-01 2.71598727e-01
4.85743046e-01 -2.54580043e-02 4.47578460e-01 6.47927284e-01
-1.33223444e-01 6.87425554e-01 4.27963793e-01 -6.67426229e-01
-1.31602550e+00 9.75728750e-01 6.64009869e-01 -4.08871472e-01
-6.64295673e-01 1.33748078e+00 5.39231718e-01 -2.28054509e-01
7.52718627e-01 1.31182134e-01 -1.99197710e-01 -3.86905819e-01
9.06005383e-01 1.17191344e-01 -1.99606180e-01 -8.10305595e-01
-6.44450009e-01 4.73728538e-01 -1.07004963e-01 -2.15858266e-01
7.14582145e-01 -5.22043228e-01 4.92419340e-02 -2.97732919e-01
8.88744652e-01 8.35873261e-02 -1.11219561e+00 -6.47003770e-01
1.22773111e-01 -4.89911556e-01 -5.76960184e-02 -1.17665410e+00
-8.44801009e-01 3.92750353e-01 8.47124636e-01 -1.22246191e-01
1.27244091e+00 6.23932302e-01 7.83165157e-01 5.96788883e-01
3.94162714e-01 -1.12425256e+00 -1.07274055e-02 3.14456761e-01
9.00139809e-01 -1.28822136e+00 -2.08485112e-01 -1.60757184e-01
-4.89473492e-01 5.07584751e-01 7.95918703e-01 -3.95633951e-02
2.57853657e-01 -2.80936182e-01 -1.14675775e-01 -3.64740938e-01
-7.17077017e-01 -5.84434986e-01 6.66153252e-01 6.96533024e-01
-1.11841984e-01 -6.70724362e-02 2.81122625e-01 6.31153286e-01
-2.41107941e-01 -1.74166441e-01 -3.57571751e-01 6.04626298e-01
-6.20155931e-01 -8.00979435e-01 -6.54221594e-01 1.52141422e-01
-2.10246637e-01 -3.12669277e-01 -5.85040450e-01 6.97878420e-01
2.93824852e-01 9.17914510e-01 2.22231939e-01 -1.21530458e-01
3.97511095e-01 2.78917663e-02 2.72932887e-01 -3.84882063e-01
-5.74465275e-01 -4.29876409e-02 -6.59744069e-02 -8.76891911e-01
-2.60812849e-01 -5.70184827e-01 -9.31780100e-01 -1.09637879e-01
-3.26423943e-02 2.88110793e-01 9.90548581e-02 7.56232619e-01
6.42042458e-01 2.48921156e-01 4.51814868e-02 -6.08808577e-01
-5.22474885e-01 -8.39525640e-01 -1.45571694e-01 2.82205224e-01
5.00953555e-01 -6.74248278e-01 8.53594318e-02 -3.16902548e-01] | [14.821783065795898, 0.8245379328727722] |
21e9d2ef-9b01-4b50-ad96-1a43a2f68530 | estimating-relative-diffusion-from-3d-micro | 2208.03337 | null | https://arxiv.org/abs/2208.03337v1 | https://arxiv.org/pdf/2208.03337v1.pdf | Estimating relative diffusion from 3D micro-CT images using CNNs | In the past several years, convolutional neural networks (CNNs) have proven their capability to predict characteristic quantities in porous media research directly from pore-space geometries. Due to the frequently observed significant reduction in computation time in comparison to classical computational methods, bulk parameter prediction via CNNs is especially compelling, e.g. for effective diffusion. While the current literature is mainly focused on fully saturated porous media, the partially saturated case is also of high interest. Due to the qualitatively different and more complex geometries of the domain available for diffusive transport present in this case, standard CNNs tend to lose robustness and accuracy with lower saturation rates. In this paper, we demonstrate the ability of CNNs to perform predictions of relative diffusion directly from full pore-space geometries. As such, our CNN conveniently fuses diffusion prediction and a well-established morphological model which describes phase distributions in partially saturated porous media. | ['Nadja Ray', 'Andreas Meier', 'Fabian Woller', 'Florian Frank', 'Stephan Gärttner'] | 2022-08-04 | null | null | null | null | ['parameter-prediction'] | ['miscellaneous'] | [-1.06115244e-01 -1.17894515e-01 6.93882033e-02 2.35153958e-01
-3.08483720e-01 -2.71578014e-01 7.42305458e-01 8.59886825e-01
-7.81672299e-01 9.54160869e-01 5.31566888e-02 -5.92018306e-01
-4.37003195e-01 -1.31759834e+00 -6.20601892e-01 -1.09937048e+00
-4.57097054e-01 7.14628339e-01 3.74751687e-01 -4.25779283e-01
2.42248550e-01 1.01109684e+00 -1.50397074e+00 -1.34582547e-02
7.04681158e-01 1.29933929e+00 1.63035810e-01 5.82782865e-01
-2.94845939e-01 5.53458989e-01 1.46393597e-01 9.41133797e-02
1.31007239e-01 -2.73520909e-02 -1.02699375e+00 -2.93708056e-01
3.60312834e-02 -2.37748563e-01 -7.43242204e-01 7.84227312e-01
5.51926494e-01 4.27029014e-01 1.11133361e+00 -3.48987937e-01
-4.45498288e-01 6.06048584e-01 -7.30214342e-02 5.66111147e-01
6.88428134e-02 2.01963857e-01 7.74107337e-01 -9.11981344e-01
5.95890880e-01 9.55435634e-01 6.72057331e-01 4.21323299e-01
-1.20214283e+00 -6.52711466e-02 -1.11207657e-01 -1.90004662e-01
-1.26467550e+00 -3.58921170e-01 5.78183711e-01 -5.83793998e-01
1.01123631e+00 5.32049909e-02 9.64334249e-01 8.18182647e-01
4.87008452e-01 6.63693905e-01 1.25564575e+00 -7.92823881e-02
6.37483895e-01 -4.79151867e-02 -3.01341742e-01 3.75422210e-01
3.51209283e-01 1.95503742e-01 -3.60688753e-02 1.50875553e-01
9.81995761e-01 -9.17674229e-02 -3.56079966e-01 -5.45714915e-01
-1.18095171e+00 6.90840065e-01 7.57591724e-01 3.12509477e-01
-3.63567293e-01 1.83075443e-02 3.20358753e-01 1.71411008e-01
6.57779992e-01 8.63959312e-01 -6.10795557e-01 -2.72148877e-01
-1.27990353e+00 1.02893353e+00 1.14476740e+00 6.15870178e-01
5.83587408e-01 9.84363854e-02 1.83995202e-01 5.95908284e-01
2.91614205e-01 6.58073723e-01 2.22715512e-01 -6.16975427e-01
3.34580243e-01 3.38157505e-01 1.34939104e-01 -8.41543853e-01
-5.20007789e-01 -3.95464927e-01 -1.22881949e+00 2.17879906e-01
9.25532758e-01 6.71345592e-02 -7.50234246e-01 1.10789430e+00
2.37841278e-01 -3.81377161e-01 2.81671941e-01 8.77060890e-01
7.98826754e-01 6.79212630e-01 1.61270648e-01 -3.08669377e-02
9.62298870e-01 -5.72207332e-01 -2.29561538e-01 1.85313329e-01
5.63732922e-01 -2.86873072e-01 8.02831113e-01 1.36194587e-01
-1.33231318e+00 2.07058191e-02 -9.35828865e-01 -4.89339679e-02
-7.51972020e-01 -5.48107445e-01 8.67115736e-01 2.53622979e-01
-8.63845587e-01 1.46614337e+00 -1.07271445e+00 -5.00382245e-01
6.61863804e-01 3.98176402e-01 -3.35555643e-01 -2.10787337e-02
-1.23835170e+00 6.92279637e-01 1.44020543e-01 2.04088867e-01
-8.78220022e-01 -1.14195979e+00 -6.99105144e-01 1.89171538e-01
-3.87322307e-02 -6.13695323e-01 1.19579947e+00 -2.45226726e-01
-1.52609956e+00 6.87384546e-01 7.01728612e-02 -7.06265330e-01
1.05429590e+00 1.18338451e-01 -1.62351020e-02 2.20274597e-01
-3.10039073e-01 4.14291352e-01 2.89948374e-01 -1.07791388e+00
-1.72479048e-01 -1.24519199e-01 2.55630910e-01 -1.11240253e-03
-1.69056296e-01 -4.65887785e-01 8.05895850e-02 -4.62759167e-01
4.06164765e-01 -5.81412017e-01 -8.03851664e-01 3.00569803e-01
-3.82261217e-01 9.31826085e-02 6.49603367e-01 -3.60970527e-01
8.79868567e-01 -1.62080765e+00 4.32522058e-01 2.21970901e-01
6.10224307e-01 3.67495060e-01 3.09784532e-01 6.56980336e-01
1.58650368e-01 2.30801716e-01 -7.44620919e-01 -1.06459163e-01
-1.44794554e-01 1.59668475e-01 1.28410086e-02 7.85817385e-01
4.05251324e-01 9.84907627e-01 -7.66185403e-01 -3.11214268e-01
4.58808035e-01 6.24356687e-01 -7.76457548e-01 5.45065477e-02
-3.72721404e-01 9.70380247e-01 -4.31881785e-01 6.57880068e-01
9.25576210e-01 -5.37787020e-01 -1.96911260e-01 5.36575094e-02
-6.40250742e-01 1.95054203e-01 -8.66388917e-01 1.35002589e+00
-5.50631642e-01 5.88899493e-01 3.26750934e-01 -1.35370898e+00
7.21613348e-01 1.43326551e-01 7.02554643e-01 -1.01734006e+00
2.85076201e-01 6.36357844e-01 1.61244363e-01 -6.10536873e-01
6.80182636e-01 -5.01846254e-01 3.97284061e-01 8.85996148e-02
-2.23987773e-01 -4.79625463e-01 3.37392479e-01 2.03528721e-02
1.00876653e+00 -2.68645678e-02 1.68194249e-01 -1.02171338e+00
6.93110108e-01 5.64574040e-02 -3.07350047e-02 7.56688476e-01
-1.29336402e-01 7.73889482e-01 3.08283031e-01 -5.37532926e-01
-1.67019904e+00 -1.13243091e+00 -8.55389535e-01 3.41098815e-01
3.70455086e-01 -8.75847489e-02 -6.29408836e-01 1.06369682e-01
4.44244295e-01 -2.12951303e-01 -6.62101567e-01 1.18452720e-01
-6.62291229e-01 -7.28463948e-01 5.21386266e-01 5.71237266e-01
6.62164569e-01 -1.09420764e+00 -3.88472587e-01 7.43387163e-01
4.37625527e-01 -1.01472855e+00 5.49013555e-01 3.89427811e-01
-1.26095641e+00 -1.09467661e+00 -1.25069642e+00 -4.23034251e-01
2.95976400e-01 -6.30220920e-02 9.49111998e-01 1.98704571e-01
-3.85426193e-01 1.49506986e-01 -2.03091115e-01 -1.26678213e-01
-5.68733931e-01 3.64699602e-01 -8.24816898e-03 -3.47739935e-01
-4.34634946e-02 -7.97972977e-01 -1.05049968e+00 -6.47979751e-02
-1.11346686e+00 -3.34735438e-02 4.18248296e-01 5.55062473e-01
5.18987060e-01 1.57806501e-01 3.34410936e-01 -7.64327943e-01
5.13796747e-01 -9.74512279e-01 -5.22807300e-01 -3.81661266e-01
-6.03467584e-01 1.46284312e-01 9.33259606e-01 -2.80649722e-01
-1.03006756e+00 -4.06145990e-01 -5.45597374e-01 6.23777173e-02
-5.17351985e-01 8.55236709e-01 2.35350460e-01 -3.27897310e-01
5.80212235e-01 1.16509102e-01 6.89480528e-02 -4.57498908e-01
-1.07622489e-01 4.32522148e-01 1.97481215e-01 -9.61038947e-01
6.31632566e-01 8.59417796e-01 4.39997554e-01 -1.43585610e+00
-4.54863429e-01 -2.97668368e-01 -6.49913669e-01 -1.18693396e-01
6.95328593e-01 -6.14069223e-01 -8.89010787e-01 6.51122689e-01
-8.28119397e-01 -6.92196965e-01 -5.30924857e-01 4.49282944e-01
-5.64917564e-01 3.87438774e-01 -1.08505440e+00 -9.20047522e-01
-1.81273431e-01 -1.37366843e+00 7.84596384e-01 2.66106874e-01
-1.98009983e-02 -1.62831259e+00 -1.70425206e-01 -1.09655283e-01
1.11430907e+00 4.90550369e-01 1.11505961e+00 -2.98378468e-01
-6.23366892e-01 -6.63358942e-02 -4.55448687e-01 -5.31796291e-02
-9.03148726e-02 1.36125758e-01 -9.35714960e-01 -2.28944510e-01
-2.51069784e-01 -4.76099327e-02 1.34950483e+00 8.28922212e-01
1.20174253e+00 3.99716608e-02 -4.19684082e-01 6.81997299e-01
1.57183433e+00 -1.61957890e-01 6.22803152e-01 5.67038536e-01
4.61044431e-01 5.80783606e-01 1.53180465e-01 5.58249831e-01
-1.00609269e-02 2.60245681e-01 5.48717797e-01 -2.93622434e-01
-1.74720094e-01 -4.56897169e-02 -2.39496619e-01 8.25429082e-01
-5.85440874e-01 -4.07125622e-01 -1.22604334e+00 5.94037950e-01
-1.46256876e+00 -7.33278215e-01 -3.92104685e-01 2.15055251e+00
5.95448136e-01 1.63996130e-01 -9.80530605e-02 4.43006366e-01
4.31808084e-01 2.11349621e-01 -5.59168875e-01 -1.44154176e-01
-5.25505602e-01 4.24726099e-01 9.13790286e-01 4.48917508e-01
-1.20167696e+00 6.53680801e-01 6.53938246e+00 5.01756966e-01
-1.49643290e+00 -2.42878109e-01 5.69834888e-01 4.31628078e-02
-3.76844108e-01 -1.88568562e-01 -6.48298383e-01 3.79115045e-01
8.64773870e-01 1.32017881e-02 4.96367425e-01 2.96406507e-01
3.42121392e-01 -2.96336204e-01 -8.22735727e-01 7.33025432e-01
-7.69375145e-01 -1.87214732e+00 3.01479429e-01 4.87347186e-01
7.71046877e-01 3.54397267e-01 1.83073908e-01 -1.59462783e-02
5.51240630e-02 -1.19320393e+00 8.87288809e-01 6.25219584e-01
7.22283721e-01 -3.86337370e-01 7.46419072e-01 2.72429675e-01
-1.38409269e+00 -1.93882272e-01 -6.08355999e-01 -5.09095073e-01
3.43414575e-01 9.31683958e-01 -3.77507627e-01 4.38927203e-01
7.46176124e-01 9.06290412e-01 -2.28412479e-01 1.43664324e+00
3.66978347e-01 5.40735841e-01 -5.49544692e-01 -1.67551205e-01
5.39555848e-01 -2.86494195e-01 4.92527276e-01 1.33781421e+00
4.59123313e-01 6.84369281e-02 -1.27845973e-01 1.10076272e+00
4.83307540e-02 3.17074358e-01 -8.09606433e-01 -3.49546492e-01
1.30530372e-01 1.01428676e+00 -9.45555210e-01 -1.98731363e-01
-2.58307815e-01 2.68509269e-01 3.17324728e-01 5.02565980e-01
-4.65548217e-01 -1.96852252e-01 8.46599638e-01 7.79140890e-01
3.03725630e-01 -6.04935050e-01 -5.56449033e-02 -9.96088982e-01
-2.33715937e-01 -1.50547192e-01 -7.75806829e-02 -3.57952863e-01
-1.47845852e+00 3.77735168e-01 -6.96009248e-02 -7.07010448e-01
4.57881987e-01 -1.10116208e+00 -5.28538883e-01 8.73172522e-01
-2.15610957e+00 -7.99507678e-01 -1.47972301e-01 3.97573858e-01
4.23093915e-01 1.91685483e-01 6.70259297e-01 3.18912983e-01
-2.57393628e-01 -8.03301204e-03 5.82482457e-01 3.79656325e-03
1.50322720e-01 -1.42507839e+00 3.53502810e-01 2.66179740e-01
-7.40249038e-01 6.58362448e-01 1.07612514e+00 -5.86174846e-01
-1.61340761e+00 -8.44774663e-01 5.67258000e-01 3.71952704e-03
9.44897532e-01 -9.81175974e-02 -1.26619685e+00 -6.01880588e-02
-1.91852182e-01 3.51998866e-01 3.00814837e-01 -1.96899906e-01
-1.28159150e-01 1.98113635e-01 -9.47402120e-01 6.41517639e-01
8.64559948e-01 -3.86806428e-01 -5.51427864e-02 1.61190495e-01
4.30529714e-01 -6.01170957e-01 -1.41671443e+00 7.06104219e-01
6.51523650e-01 -8.38746488e-01 1.05205703e+00 -7.78586209e-01
1.07860541e+00 9.73821133e-02 -1.08155653e-01 -1.03561318e+00
-1.34969488e-01 -2.41921172e-01 -1.35852724e-01 6.32099628e-01
6.87800407e-01 -6.00222051e-01 9.63258266e-01 5.14054477e-01
-6.62454888e-02 -1.13133717e+00 -1.09036410e+00 -4.81504977e-01
9.45673108e-01 -3.16320956e-01 4.91136163e-01 7.39274502e-01
-3.15053612e-02 -2.87596256e-01 5.35805821e-02 3.64523642e-02
4.87730771e-01 9.71926562e-03 1.37626335e-01 -1.70998478e+00
-7.45082274e-02 -7.82715619e-01 -3.07097554e-01 -1.36798465e+00
1.23666786e-01 -1.02753937e+00 7.07655177e-02 -1.58115101e+00
-8.36657956e-02 -1.02804077e+00 -2.56404221e-01 -2.80370325e-01
3.85349929e-01 3.29452336e-01 -4.47126478e-01 3.22523326e-01
-3.21409136e-01 6.68973088e-01 1.82966828e+00 -3.92285615e-01
-2.46612713e-01 -1.25242189e-01 -1.75242737e-01 4.58677977e-01
9.29914534e-01 -8.57340544e-02 5.26222251e-02 -2.32771590e-01
6.27365708e-01 3.12190294e-01 3.73737365e-01 -1.34521019e+00
2.18700707e-01 -3.56319249e-02 4.15140510e-01 -1.75813183e-01
1.89672157e-01 -6.35134399e-01 -2.06923679e-01 3.99468184e-01
-4.41314429e-01 -4.04099405e-01 4.77716148e-01 6.99272513e-01
-3.16842586e-01 -8.12100992e-02 9.05453563e-01 -5.57516813e-01
-3.90074402e-01 8.88408959e-01 -9.59863245e-01 -1.55427113e-01
5.35108626e-01 -4.53196824e-01 -2.27805153e-01 -5.19614518e-02
-1.06982839e+00 -2.02264562e-01 7.50334918e-01 -6.41655475e-02
4.94109929e-01 -9.77264166e-01 -3.63423705e-01 2.02142805e-01
-1.95126459e-02 5.72833598e-01 5.85950911e-01 1.19839072e+00
-1.22997177e+00 4.58599329e-01 -1.93171605e-01 -5.45584381e-01
-5.08990169e-01 4.15733814e-01 6.81825280e-01 -2.01265350e-01
-8.68833899e-01 6.25986934e-01 1.24649003e-01 -3.81349027e-01
-2.00491711e-01 -6.05177283e-01 -2.11866766e-01 -2.39558414e-01
1.62079692e-01 5.00510216e-01 2.84749746e-01 -5.81682265e-01
2.85783689e-02 4.23877656e-01 1.71177998e-01 2.41312236e-01
1.68639588e+00 -1.48707151e-01 -2.54084975e-01 4.57189322e-01
1.00772762e+00 -1.74198061e-01 -1.65915465e+00 -2.19319478e-01
-1.86217725e-01 -1.93784878e-01 1.81667387e-01 -2.12792039e-01
-1.30260968e+00 1.25933444e+00 1.54730156e-01 5.79361916e-01
5.20866036e-01 1.92034543e-01 8.27849567e-01 4.38247651e-01
2.94283956e-01 -8.55406523e-01 -1.58989161e-01 7.74343550e-01
5.40247202e-01 -1.18113637e+00 9.42546874e-02 -3.96832079e-01
2.59033293e-01 1.34812486e+00 3.15978378e-01 -2.14056626e-01
1.11986244e+00 6.35676265e-01 -2.01727077e-01 -4.22913611e-01
-3.35210353e-01 7.23472461e-02 -2.52735704e-01 5.05076647e-01
6.28834963e-01 -9.33606178e-03 -8.68049040e-02 1.62929799e-02
-1.80720478e-01 -2.34264433e-01 5.82612336e-01 9.90256727e-01
-5.87000608e-01 -8.21947157e-01 -8.15123841e-02 6.27481997e-01
-4.40102488e-01 -8.64017457e-02 2.09579751e-01 8.35116446e-01
-1.18239634e-01 6.80608630e-01 3.25880766e-01 1.44230589e-01
1.60613433e-01 -2.66962200e-01 5.32695174e-01 -2.01796114e-01
-3.42646688e-01 -1.65905103e-01 -1.94124088e-01 -4.08930421e-01
-7.56976664e-01 -6.64394855e-01 -1.41316986e+00 -6.88803375e-01
-1.38379440e-01 3.49279553e-01 6.80310369e-01 1.12616158e+00
-4.32079025e-02 5.78908145e-01 2.68801693e-02 -1.47893536e+00
-2.83457190e-01 -9.22048211e-01 -1.32722497e+00 3.68858606e-01
5.17025769e-01 -9.69959259e-01 -3.77006829e-01 -3.19373429e-01] | [6.42698335647583, 3.379267454147339] |
126841ae-781b-420f-85f6-74e51bfbdf75 | phase-only-image-based-kernel-estimation-for-1 | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Pan_Phase-Only_Image_Based_Kernel_Estimation_for_Single_Image_Blind_Deblurring_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Pan_Phase-Only_Image_Based_Kernel_Estimation_for_Single_Image_Blind_Deblurring_CVPR_2019_paper.pdf | Phase-Only Image Based Kernel Estimation for Single Image Blind Deblurring | The image motion blurring process is generally modelled as the convolution of a blur kernel with a latent image. Therefore, the estimation of the blur kernel is essentially important for blind image deblurring. Unlike existing approaches which focus on approaching the problem by enforcing various priors on the blur kernel and the latent image, we are aiming at obtaining a high quality blur kernel directly by studying the problem in the frequency domain. We show that the auto-correlation of the absolute phase-only image 1 can provide faithful information about the motion (e.g., the motion direction and magnitude, we call it the motion pattern in this paper.) that caused the blur, leading to a new and efficient blur kernel estimation approach. The blur kernel is then refined and the sharp image is estimated by solving an optimization problem by enforcing a regularization on the blur kernel and the latent image. We further extend our approach to handle non-uniform blur, which involves spatially varying blur kernels. Our approach is evaluated extensively on synthetic and real data and shows good results compared to the state-of-the-art deblurring approaches.
| [' Yuchao Dai', ' Miaomiao Liu', ' Richard Hartley', 'Liyuan Pan'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['single-image-blind-deblurring', 'blind-image-deblurring'] | ['computer-vision', 'computer-vision'] | [ 2.80356228e-01 -5.70119441e-01 3.51794690e-01 -7.13368282e-02
-3.97973627e-01 -5.63938916e-01 5.81900835e-01 -5.79769075e-01
-3.16781968e-01 8.48869562e-01 5.63281953e-01 9.96266156e-02
-2.92483687e-01 -2.35295549e-01 -5.59171617e-01 -1.05198801e+00
4.55357768e-02 -2.47440517e-01 1.09354340e-01 2.44661555e-01
5.00314176e-01 2.21317708e-01 -1.08704436e+00 -2.20698975e-02
1.20266116e+00 7.07819223e-01 4.93489951e-01 8.74326825e-01
3.98864686e-01 1.05516326e+00 -6.40643060e-01 7.48069910e-03
8.06178674e-02 -7.13636339e-01 -9.44404781e-01 4.77243572e-01
4.21667755e-01 -5.79120398e-01 -6.55552208e-01 1.48264802e+00
2.53450811e-01 1.28191635e-01 5.87200522e-01 -6.63919985e-01
-9.55983460e-01 2.23293215e-01 -7.63792276e-01 4.32901531e-01
3.67426276e-01 -3.83623764e-02 4.48771179e-01 -9.67454255e-01
4.43254739e-01 1.01430249e+00 6.86367631e-01 2.26503089e-01
-1.46910703e+00 -9.40381661e-02 -4.39521074e-01 5.35011709e-01
-1.50179398e+00 -7.72384524e-01 8.92399251e-01 -6.13495946e-01
1.78231761e-01 4.17286873e-01 1.74553186e-01 7.45364189e-01
3.46693844e-01 3.99961233e-01 1.51326370e+00 -5.60626626e-01
2.55726844e-01 -8.72633830e-02 1.74078181e-01 3.73230070e-01
2.69205481e-01 2.52907753e-01 -2.72034049e-01 -2.28116944e-01
1.18936038e+00 -9.70627889e-02 -1.25280583e+00 -4.11764473e-01
-1.45682704e+00 3.63839030e-01 5.49371004e-01 4.31026042e-01
-6.07682228e-01 2.47445956e-01 -1.19750760e-01 7.71939456e-02
5.89262187e-01 5.37406087e-01 -8.35084841e-02 -1.87093183e-01
-1.31372571e+00 2.18663260e-01 8.18513691e-01 5.32677174e-01
7.73132265e-01 -1.48922190e-01 -3.73563856e-01 9.16167796e-01
3.52176607e-01 4.87306446e-01 4.67779011e-01 -1.07352173e+00
4.08927947e-02 -1.46896228e-01 8.93428922e-01 -6.98891640e-01
1.78430706e-01 -2.25764781e-01 -1.08280969e+00 2.68253654e-01
7.40983963e-01 -3.59136105e-01 -1.01203787e+00 1.56467855e+00
-3.63198631e-02 7.41018355e-01 1.20148361e-01 1.38318431e+00
2.80696571e-01 6.41814351e-01 -4.42679465e-01 -5.12161076e-01
1.48519993e+00 -1.03611279e+00 -1.23423564e+00 -3.68062228e-01
-1.35059401e-01 -1.37174094e+00 5.44514656e-01 1.48288429e-01
-1.24687672e+00 -5.73679328e-01 -9.44491267e-01 6.44521862e-02
2.20770031e-01 1.99654132e-01 1.96775869e-01 4.08269703e-01
-1.23929489e+00 6.07244015e-01 -7.01197565e-01 -2.58830577e-01
7.95990080e-02 1.56799220e-02 -2.32569695e-01 -5.02965510e-01
-1.18038690e+00 1.26750016e+00 1.62669003e-01 3.37491512e-01
-6.18763149e-01 -5.86106777e-01 -7.78232634e-01 6.89675957e-02
5.62856980e-02 -8.78738344e-01 1.20236599e+00 -1.03233469e+00
-1.50100183e+00 4.48236912e-01 -5.77862859e-01 -3.54145736e-01
6.26271248e-01 -5.39619565e-01 -2.05460638e-01 2.33308941e-01
-4.14017662e-02 1.36207864e-01 1.50490093e+00 -1.45194113e+00
-1.67238563e-01 2.65613496e-02 -3.18838693e-02 2.41071373e-01
2.22593933e-01 3.99387628e-02 -4.67692196e-01 -1.01027870e+00
1.01410083e-01 -8.79032731e-01 -3.09916101e-02 -2.13520393e-01
-2.26860940e-01 5.58518946e-01 8.96583736e-01 -1.04421508e+00
1.34181440e+00 -2.33873057e+00 3.78229350e-01 -3.38696599e-01
3.35042387e-01 2.59571522e-01 1.17482960e-01 3.34006548e-01
-2.34329507e-01 -4.19283688e-01 -6.02262795e-01 -2.80587763e-01
-3.45181942e-01 -5.85015072e-03 -4.51963454e-01 9.60864842e-01
2.06208620e-02 8.23703825e-01 -1.01402402e+00 -3.70212793e-02
4.10490543e-01 6.61081195e-01 -2.19276398e-01 6.08238518e-01
2.93579698e-01 6.97142959e-01 -9.48454663e-02 1.19887076e-01
1.29405200e+00 -3.58456314e-01 -1.44206047e-01 -4.75372046e-01
-4.04628932e-01 -2.40817085e-01 -1.10202503e+00 1.64496410e+00
-4.32070047e-01 9.41538095e-01 5.26967764e-01 -5.26790798e-01
4.98906553e-01 4.58376706e-01 2.43910715e-01 -4.15051840e-02
-1.09996647e-01 1.80050239e-01 -2.83365361e-02 -6.46505594e-01
5.26633739e-01 -2.70034730e-01 3.21988553e-01 2.77759135e-01
-1.88949451e-01 -2.45096698e-01 -8.50025564e-02 -4.42170314e-02
1.04867554e+00 7.12340772e-02 2.08935127e-01 -7.44553030e-01
6.33081913e-01 -2.56413162e-01 4.20288481e-02 7.82678902e-01
-2.10345641e-01 9.52934921e-01 1.43368214e-01 -7.43070915e-02
-1.04501522e+00 -1.02281249e+00 -2.88792461e-01 2.35127419e-01
6.16401196e-01 -7.60224313e-02 -1.25209212e+00 -1.37612581e-01
-1.73826769e-01 5.99677444e-01 -6.45025194e-01 -1.51163459e-01
-4.27661717e-01 -9.46312249e-01 8.83538872e-02 1.18795805e-01
9.25105155e-01 -5.57380557e-01 -4.75676626e-01 2.44410574e-01
-5.24864376e-01 -1.15322030e+00 -1.17604935e+00 -1.65801600e-01
-6.59812093e-01 -1.12140918e+00 -1.52151954e+00 -6.89085364e-01
8.06599915e-01 6.66789174e-01 7.24730730e-01 -2.34901920e-01
-1.47291586e-01 2.73266166e-01 -5.24993427e-02 2.06459850e-01
-4.27303016e-01 -7.14660943e-01 -1.33210905e-02 5.49275935e-01
-1.06475040e-01 -4.59286600e-01 -9.10810232e-01 3.51382136e-01
-1.03251278e+00 2.53091276e-01 6.41114354e-01 9.96821284e-01
-1.72546044e-01 3.80315155e-01 1.04893707e-01 -5.31061590e-01
8.14539433e-01 -2.67619401e-01 -7.17660904e-01 1.20755598e-01
-5.25380671e-01 2.95212418e-01 2.66732067e-01 -6.25733554e-01
-1.64263892e+00 -1.50564209e-01 3.86476487e-01 -6.66795611e-01
-1.92160547e-01 3.13167930e-01 4.54978691e-03 -2.57904172e-01
6.33635283e-01 3.39870870e-01 -2.20539607e-02 -7.10990191e-01
5.17072797e-01 7.73936510e-01 9.05925393e-01 -1.84994414e-01
1.03814244e+00 5.28933108e-01 -1.07052900e-01 -1.00637615e+00
-6.62113428e-01 -6.67207539e-01 -6.33954942e-01 -4.42559719e-02
9.07657564e-01 -8.91205668e-01 -4.06553298e-01 1.18868065e+00
-1.52935100e+00 -2.11663768e-01 -1.29048020e-01 7.94534743e-01
-5.31308055e-01 7.74409652e-01 -1.03992426e+00 -7.81743228e-01
7.47938603e-02 -1.27133191e+00 8.39932740e-01 2.42765784e-01
-9.90095437e-02 -1.16903162e+00 2.63159811e-01 1.33084401e-01
8.69538844e-01 -9.53728631e-02 6.53729022e-01 4.29848552e-01
-8.97043288e-01 6.81857765e-02 -7.44423628e-01 4.25066233e-01
6.85725570e-01 -4.79695857e-01 -1.25494814e+00 -2.60879695e-01
6.98901474e-01 4.00268734e-01 9.73107219e-01 9.28528905e-01
7.57247865e-01 -5.11118233e-01 -1.87032774e-01 7.70263255e-01
1.55383468e+00 -4.44485433e-02 8.83598089e-01 1.81864157e-01
8.13196957e-01 3.78039062e-01 4.34621990e-01 1.96435884e-01
-1.21777676e-01 9.01523888e-01 1.48426414e-01 -1.65516660e-01
-4.48407620e-01 1.89557672e-01 2.37792000e-01 6.23062611e-01
-2.70359099e-01 -1.09512262e-01 -5.24538636e-01 6.97753966e-01
-1.98452747e+00 -1.00397933e+00 -3.11388522e-01 2.36339974e+00
1.11013198e+00 -3.02247852e-01 -4.75301474e-01 -1.37269765e-01
9.49637949e-01 3.81197065e-01 -2.94568449e-01 8.72805417e-02
-1.18280191e-03 -1.67057905e-02 5.96159935e-01 1.11892986e+00
-1.13320625e+00 7.04814494e-01 6.86908674e+00 6.39159203e-01
-1.06850588e+00 7.60107953e-03 3.64106774e-01 4.19155061e-01
-3.21027488e-02 9.18753967e-02 -1.50902495e-01 7.65137851e-01
6.80469930e-01 -3.52357268e-01 1.06312096e+00 3.05183500e-01
6.38121426e-01 -4.89545792e-01 -8.59190941e-01 1.22432303e+00
-7.61198401e-02 -9.77482855e-01 -4.41347539e-01 1.12051740e-01
8.42108488e-01 -2.63645411e-01 5.23517728e-02 -4.87181664e-01
1.13802001e-01 -8.88576031e-01 7.49432564e-01 9.86084878e-01
9.10571814e-01 -2.45374054e-01 8.73557448e-01 1.95632011e-01
-8.07359874e-01 2.88976192e-01 -2.45838404e-01 -2.04248533e-01
3.12563151e-01 1.09697926e+00 -4.42853957e-01 5.21533966e-01
6.05981529e-01 7.69095898e-01 -3.54365140e-01 1.43464148e+00
-3.16641390e-01 5.30232728e-01 2.15752989e-01 6.65633857e-01
-1.06384426e-01 -5.54342151e-01 8.38214874e-01 1.26095080e+00
6.01282597e-01 1.35662451e-01 -3.78702939e-01 1.14553368e+00
3.85466591e-02 -6.61574960e-01 -2.35000342e-01 2.30367020e-01
1.34565279e-01 1.11465204e+00 -4.13339704e-01 -3.67585242e-01
-3.39551955e-01 1.74125266e+00 -1.35439321e-01 9.31698918e-01
-7.09231079e-01 -3.29527259e-01 9.20754910e-01 -1.57323226e-01
4.28824544e-01 -1.92266405e-01 -9.05230269e-02 -1.59510028e+00
-1.33708864e-01 -5.51145911e-01 -2.73277700e-01 -1.30204022e+00
-1.35798299e+00 6.23125672e-01 7.82156289e-02 -1.14368021e+00
-2.22482786e-01 -4.27743137e-01 -7.01875687e-01 1.67140365e+00
-1.75289989e+00 -7.52603054e-01 -6.64083898e-01 4.98162597e-01
4.57778513e-01 4.48538840e-01 4.78572905e-01 9.76705477e-02
-1.86789185e-01 -2.28388548e-01 5.22021770e-01 -1.70211926e-01
1.09162152e+00 -1.47991526e+00 3.69640857e-01 1.26642382e+00
-3.32531303e-01 7.66731739e-01 1.28031266e+00 -6.00644231e-01
-1.25569296e+00 -9.14374948e-01 6.88877165e-01 -4.10313487e-01
8.26018929e-01 -8.12168717e-02 -1.16409957e+00 4.98263657e-01
7.45608211e-01 2.02727959e-01 7.32818320e-02 -5.03495872e-01
2.69505046e-02 1.54018164e-01 -9.99454558e-01 3.47984880e-01
6.53466046e-01 -6.22377455e-01 -8.14470351e-01 -3.16626914e-02
5.38310230e-01 -5.55253744e-01 -5.46443105e-01 1.90228373e-01
3.67938608e-01 -1.00830257e+00 1.01160741e+00 -3.29335034e-02
3.56158972e-01 -7.04852104e-01 1.76944956e-01 -1.81017959e+00
-8.60979795e-01 -8.86491239e-01 -4.03960973e-01 1.07610774e+00
-2.53372222e-01 -5.70879519e-01 3.26360822e-01 4.68745440e-01
9.60842595e-02 -1.19023412e-01 -5.71790397e-01 -5.06536663e-01
-5.61458409e-01 1.35517925e-01 2.13600472e-01 1.05774784e+00
-1.84787676e-01 2.21439898e-01 -7.83572137e-01 5.72709322e-01
1.03446913e+00 2.19552383e-01 3.38334084e-01 -7.79363811e-01
-3.84589612e-01 -2.17552885e-01 -1.06015630e-01 -1.59585905e+00
-1.98973298e-01 -1.56406417e-01 4.97361839e-01 -1.35477519e+00
3.38855416e-01 -1.19040506e-02 -1.22031711e-01 -2.58954763e-01
-6.60423875e-01 2.21625507e-01 -7.30183646e-02 7.07935631e-01
-1.21530205e-01 3.79259586e-01 1.50123858e+00 3.59730679e-03
-1.08937576e-01 1.26928285e-01 -4.89999235e-01 7.55823195e-01
3.26651424e-01 -1.37778416e-01 -2.84919143e-01 -5.46703994e-01
-3.85966361e-01 2.12012947e-01 7.55997658e-01 -8.30276668e-01
4.28114772e-01 -1.36499867e-01 3.95951629e-01 -1.48024872e-01
3.60888302e-01 -7.88783431e-01 3.56513470e-01 3.31963956e-01
-1.22207038e-01 -3.88624579e-01 9.66502056e-02 6.50115967e-01
-4.25755024e-01 -5.61218798e-01 1.02104855e+00 -1.44747287e-01
-5.48960388e-01 2.76495945e-02 -4.43404287e-01 -1.97558358e-01
6.34975076e-01 -2.53637247e-02 -5.16785085e-01 -5.92272937e-01
-7.34567940e-01 -4.08949405e-01 8.65918458e-01 8.70167091e-02
5.22985935e-01 -1.21301961e+00 -7.07653165e-01 2.30339319e-01
-2.14496881e-01 -2.91419715e-01 3.52268964e-01 1.09274471e+00
-6.60414219e-01 3.24679136e-01 -1.00796726e-02 -4.28089499e-01
-1.20716095e+00 7.76790917e-01 5.89677989e-01 -2.50776596e-02
-3.47729415e-01 7.81894088e-01 6.75559759e-01 5.46819270e-01
-1.24076322e-01 -3.61611247e-01 -9.06661525e-02 -3.40887308e-01
7.72024274e-01 5.32104015e-01 -2.31830418e-01 -8.38559866e-01
-2.22185478e-01 7.08804429e-01 1.37242302e-01 -2.20323056e-01
1.12913322e+00 -6.60458803e-01 -6.07363641e-01 1.68692529e-01
1.17163408e+00 4.73117858e-01 -1.74519682e+00 -3.13407063e-01
-3.94011587e-02 -9.35003579e-01 4.99299526e-01 -6.75877512e-01
-8.38025272e-01 6.81013048e-01 7.96569347e-01 3.81171018e-01
1.34607100e+00 -1.29872482e-04 4.64827627e-01 -4.45182592e-01
1.77206680e-01 -4.24147725e-01 -2.21817289e-02 3.14481586e-01
1.04921949e+00 -1.06901944e+00 9.65481550e-02 -6.19794607e-01
-2.71806628e-01 1.14344645e+00 -1.27888005e-02 -1.41149744e-01
5.78779161e-01 1.86731979e-01 1.15646578e-01 4.10182104e-02
-1.96361512e-01 -1.50787786e-01 6.56603754e-01 3.94284785e-01
4.16038275e-01 -1.20555006e-01 -1.13053530e-01 4.24789488e-02
2.69594371e-01 3.98799539e-01 7.76432037e-01 6.68955445e-01
-4.41851437e-01 -9.49115515e-01 -9.51447368e-01 -5.01247533e-02
-5.78685880e-01 -3.41504812e-01 -1.05808517e-02 1.42504320e-01
5.11195883e-02 9.99709249e-01 -1.12895556e-01 1.17779218e-01
8.85265693e-02 -1.52416766e-01 7.40927815e-01 -3.32862020e-01
1.44884646e-01 4.72641736e-01 -1.92550495e-01 -4.07749861e-01
-5.99120975e-01 -6.54259086e-01 -5.09175360e-01 -2.16225445e-01
-5.23495793e-01 2.71301955e-01 4.16827977e-01 7.71273434e-01
1.41444504e-01 4.46675986e-01 6.31676614e-01 -1.04822159e+00
-4.53585774e-01 -1.24661124e+00 -8.32005560e-01 5.28749228e-01
1.07609379e+00 -4.12798166e-01 -7.91039169e-01 5.78422248e-01] | [11.591912269592285, -2.729902982711792] |
dc0cd569-8a95-46a4-84aa-3c391f82e58d | self-attention-on-multi-shifted-windows-for | 2207.04403 | null | https://arxiv.org/abs/2207.04403v1 | https://arxiv.org/pdf/2207.04403v1.pdf | Self-attention on Multi-Shifted Windows for Scene Segmentation | Scene segmentation in images is a fundamental yet challenging problem in visual content understanding, which is to learn a model to assign every image pixel to a categorical label. One of the challenges for this learning task is to consider the spatial and semantic relationships to obtain descriptive feature representations, so learning the feature maps from multiple scales is a common practice in scene segmentation. In this paper, we explore the effective use of self-attention within multi-scale image windows to learn descriptive visual features, then propose three different strategies to aggregate these feature maps to decode the feature representation for dense prediction. Our design is based on the recently proposed Swin Transformer models, which totally discards convolution operations. With the simple yet effective multi-scale feature learning and aggregation, our models achieve very promising performance on four public scene segmentation datasets, PASCAL VOC2012, COCO-Stuff 10K, ADE20K and Cityscapes. | ['Qiang Wu', 'Jian Zhang', 'Zhibin Li', 'Litao Yu'] | 2022-07-10 | null | null | null | null | ['scene-segmentation'] | ['computer-vision'] | [ 5.52903593e-01 -1.53186157e-01 -1.58625200e-01 -7.92515934e-01
-4.90306556e-01 -4.85115707e-01 4.99714822e-01 1.11767583e-01
-4.93512183e-01 3.31850260e-01 4.94332798e-02 -3.27229232e-01
-8.48442465e-02 -8.88409138e-01 -9.28720713e-01 -5.38702905e-01
1.55639380e-01 1.76920667e-01 6.38562024e-01 -6.42656311e-02
3.31906945e-01 4.63450760e-01 -1.73408961e+00 6.19794428e-01
9.45421696e-01 1.39581668e+00 6.66806161e-01 8.49373877e-01
-5.45073628e-01 9.68109608e-01 -3.02267343e-01 -4.88949001e-01
2.06809625e-01 -3.06917101e-01 -1.25788236e+00 3.93460304e-01
5.61851859e-01 -8.17739218e-02 -7.97942951e-02 1.19541097e+00
9.09029916e-02 1.60749674e-01 6.94645107e-01 -1.11019313e+00
-5.71077645e-01 4.00418609e-01 -5.45977473e-01 2.78052837e-01
-6.98756725e-02 -1.15505517e-01 1.10069311e+00 -5.09241760e-01
4.82247114e-01 1.17467749e+00 5.37944853e-01 1.95618317e-01
-1.00528085e+00 -2.76692808e-01 4.20292348e-01 7.32652903e-01
-1.17976451e+00 -9.10324603e-03 7.47958779e-01 -4.86000836e-01
7.77569711e-01 3.99572462e-01 8.51293981e-01 6.21781111e-01
3.00142244e-02 1.07516408e+00 1.37172925e+00 -3.06526721e-01
1.05481260e-01 2.89202780e-01 3.10167611e-01 6.73089564e-01
-1.02225490e-01 -4.29120898e-01 -2.76671886e-01 4.13136929e-01
8.31047237e-01 5.00680685e-01 -1.66561715e-02 -2.92764455e-01
-1.03402090e+00 9.00770426e-01 1.03588498e+00 3.66295516e-01
-1.75503507e-01 1.35820374e-01 3.37552786e-01 5.64728826e-02
5.15077114e-01 2.99432099e-01 -6.34859264e-01 1.39124230e-01
-9.25127268e-01 2.38646455e-02 3.67407411e-01 9.80822086e-01
1.13698757e+00 -2.70138204e-01 -2.64533848e-01 1.04535878e+00
3.37490648e-01 4.31762367e-01 4.25583392e-01 -9.03060317e-01
2.51641870e-01 6.96946681e-01 -2.45610774e-01 -9.77459431e-01
-4.45403308e-01 -3.13628435e-01 -8.55245531e-01 6.89589158e-02
2.97634959e-01 1.98421434e-01 -1.52369535e+00 1.24968827e+00
3.81918132e-01 2.99473494e-01 -1.06239744e-01 7.80682445e-01
1.12855637e+00 8.45859051e-01 3.79296601e-01 1.78194135e-01
1.35627091e+00 -1.27751529e+00 -4.70988482e-01 -4.09277499e-01
3.47896457e-01 -7.39160776e-01 1.14965570e+00 2.19643325e-01
-7.91845918e-01 -9.21898782e-01 -8.58927667e-01 -5.39958000e-01
-7.44970262e-01 2.60683477e-01 9.93796527e-01 3.14680010e-01
-8.13114524e-01 5.84581256e-01 -7.10101724e-01 -5.35540223e-01
8.91720951e-01 2.68051445e-01 -1.68529391e-01 -3.11587632e-01
-7.65690863e-01 6.98300242e-01 6.17829859e-01 1.31520689e-01
-7.92108297e-01 -5.67261994e-01 -1.01785958e+00 2.27459683e-03
1.71058610e-01 -3.66980642e-01 1.05229545e+00 -1.31101048e+00
-1.23540080e+00 1.05231881e+00 -3.46005797e-01 -5.07733166e-01
2.94606000e-01 -2.26132929e-01 -2.48956308e-01 2.36421958e-01
3.80960643e-01 1.08580410e+00 8.75747025e-01 -1.23955154e+00
-1.00255513e+00 -1.73840269e-01 2.84665406e-01 2.75877327e-01
-2.83820838e-01 -8.57398063e-02 -7.63502717e-01 -4.79250878e-01
2.93270409e-01 -6.63144529e-01 -6.55515611e-01 -1.75706431e-01
-6.24337912e-01 -3.23201627e-01 8.04192603e-01 -6.26851141e-01
7.97982275e-01 -2.00261760e+00 7.61215165e-02 2.06404433e-01
1.05061151e-01 2.31690466e-01 -9.69766974e-02 -2.79480163e-02
7.31745213e-02 3.34934406e-02 -5.17448485e-01 -1.71696931e-01
-2.62490779e-01 3.01739931e-01 -4.88230824e-01 1.99692816e-01
3.70003521e-01 1.20570576e+00 -6.30034506e-01 -7.12377131e-01
6.15972459e-01 2.80239761e-01 -6.18293941e-01 3.99580091e-01
-3.74334157e-01 2.72597641e-01 -5.50445735e-01 6.38323009e-01
8.42190146e-01 -4.58458185e-01 -1.91576764e-01 -4.26992655e-01
-2.11936086e-01 1.34984016e-01 -9.96749878e-01 1.71991658e+00
-3.39589000e-01 6.72834575e-01 -2.32264102e-01 -1.35873830e+00
7.61561692e-01 -2.64325738e-01 3.68566006e-01 -1.00680435e+00
1.84977576e-01 8.31832960e-02 -2.65721947e-01 -6.31466329e-01
4.96281624e-01 2.44582747e-03 -2.24700972e-01 -4.14031930e-02
3.08458865e-01 -3.85928690e-01 3.24711353e-01 2.87691206e-02
7.06399858e-01 1.44864187e-01 1.97994247e-01 -2.67237961e-01
5.41584611e-01 2.63053983e-01 6.17952704e-01 7.20376909e-01
-1.53017566e-01 8.19187582e-01 4.63786095e-01 -5.74188590e-01
-9.71137822e-01 -1.03866744e+00 -2.66237915e-01 1.26249897e+00
5.24495304e-01 -2.30894789e-01 -9.25625980e-01 -7.40679383e-01
-1.10486425e-01 4.35613722e-01 -7.82422781e-01 6.71994314e-03
-4.20731753e-01 -7.27313101e-01 1.93632990e-01 6.63597465e-01
9.09549952e-01 -1.25986588e+00 -6.96208358e-01 1.19498439e-01
-1.96369767e-01 -1.24000347e+00 -2.31277794e-01 4.68191981e-01
-6.72807217e-01 -1.02244902e+00 -7.58602500e-01 -1.04942966e+00
7.14063883e-01 4.83443707e-01 9.83003557e-01 -1.02492541e-01
-7.43701696e-01 1.93328664e-01 -3.91132444e-01 -3.15400839e-01
2.21717954e-02 1.98795855e-01 -5.99847555e-01 2.66506195e-01
3.60245496e-01 -2.95585245e-01 -6.20122731e-01 1.42594576e-01
-9.17368412e-01 5.17766833e-01 7.22678781e-01 6.63788438e-01
1.06343281e+00 -3.85039374e-02 2.15895280e-01 -1.17505467e+00
1.81624237e-02 -3.92720342e-01 -5.42190731e-01 4.76822525e-01
-2.61720598e-01 -5.97475618e-02 7.30633020e-01 -1.27885282e-01
-1.10539341e+00 4.09985662e-01 -4.51910973e-01 -3.45235914e-01
-4.60329056e-01 2.93394715e-01 -2.98897624e-01 -1.27335846e-01
2.14715049e-01 5.25223553e-01 -4.32898611e-01 -5.44814229e-01
7.04520941e-01 6.14824474e-01 5.51803648e-01 -3.43758464e-01
5.57298481e-01 5.32983482e-01 -1.21514417e-01 -1.09267926e+00
-1.27886188e+00 -7.20010817e-01 -1.02261400e+00 -4.91272770e-02
1.40204489e+00 -9.60709095e-01 -4.49641854e-01 6.46841407e-01
-9.36867774e-01 -4.55109477e-01 -3.62090021e-01 1.02662846e-01
-5.78672886e-01 2.06436574e-01 -5.23012578e-01 -3.16712379e-01
-9.62712243e-02 -1.24357784e+00 1.15147603e+00 7.14363754e-01
2.64324993e-01 -9.62074697e-01 -2.15101838e-01 4.76973087e-01
4.49897647e-01 1.39337540e-01 1.03478551e+00 -4.64132220e-01
-8.03945780e-01 1.20717734e-01 -8.23324800e-01 4.71053362e-01
1.15624592e-01 6.06661476e-02 -1.05746043e+00 5.87610789e-02
-2.59290278e-01 -5.14651477e-01 1.27389073e+00 6.62919164e-01
1.87476385e+00 -9.48765352e-02 -2.05875441e-01 1.05639720e+00
1.64945567e+00 1.27417192e-01 6.63596749e-01 2.69163638e-01
1.14036107e+00 6.93406165e-01 6.42653108e-01 1.60249382e-01
6.94154143e-01 3.78629684e-01 4.95551884e-01 -5.26989877e-01
-2.01003820e-01 -2.70517856e-01 2.19082292e-02 7.69978464e-01
3.05482503e-02 1.28557086e-01 -7.13840961e-01 5.68389893e-01
-1.74333572e+00 -8.47860098e-01 -2.48712420e-01 1.68155837e+00
7.42170155e-01 8.76791552e-02 -1.22320116e-01 -1.15092613e-01
5.06667495e-01 3.10926616e-01 -6.25439525e-01 -4.80407506e-01
-2.51417190e-01 3.23272049e-01 6.76010847e-01 2.25714326e-01
-1.62210512e+00 1.36356974e+00 6.15487289e+00 1.01346898e+00
-1.40646994e+00 8.96123871e-02 1.08260405e+00 2.53865480e-01
-2.25252301e-01 -1.63992271e-01 -9.02835190e-01 3.29768479e-01
7.28032708e-01 1.88516095e-01 2.62537539e-01 9.94805217e-01
-2.17738152e-01 -2.20031932e-01 -8.06946814e-01 9.69201386e-01
1.03689440e-01 -1.54626429e+00 1.27680078e-01 -2.03036770e-01
8.84895802e-01 2.51259804e-01 2.00762704e-01 3.57065290e-01
3.81256968e-01 -1.21960139e+00 7.39758193e-01 4.29985702e-01
6.33255839e-01 -6.54415965e-01 4.51170534e-01 1.12483926e-01
-1.40772176e+00 -1.86677679e-01 -6.80123806e-01 1.90769777e-01
7.39661306e-02 4.55615133e-01 -4.82240587e-01 3.05466950e-01
1.17294228e+00 1.17783010e+00 -1.07279813e+00 1.32463229e+00
-2.18372941e-01 6.83458686e-01 -2.09964052e-01 -1.06934300e-02
5.04189134e-01 -1.61843702e-01 -1.79757178e-01 1.40823138e+00
9.36536789e-02 -4.34071906e-02 2.92003095e-01 9.20572221e-01
-1.38708148e-02 1.56061426e-01 -3.92015576e-01 1.06788956e-01
-4.40466814e-02 1.49020648e+00 -1.27501869e+00 -4.54361886e-01
-5.47508419e-01 1.27907574e+00 4.13620234e-01 4.08775210e-01
-9.54702139e-01 -4.50310081e-01 6.62644625e-01 -5.79784475e-02
6.75642312e-01 -1.01006329e-01 -4.57405388e-01 -1.03766251e+00
-1.40971005e-01 -5.30665636e-01 3.14140975e-01 -6.41454518e-01
-1.08974636e+00 6.40441298e-01 2.17751376e-02 -1.06679630e+00
5.97919896e-02 -7.78642237e-01 -6.88736379e-01 6.54839218e-01
-1.94758606e+00 -1.50501072e+00 -4.13536638e-01 8.73214066e-01
8.33710074e-01 2.00209454e-01 5.01598835e-01 2.07380041e-01
-4.57167178e-01 2.71026313e-01 3.05794477e-01 2.17691258e-01
2.59588629e-01 -1.46167088e+00 3.30472648e-01 7.91215897e-01
5.33273995e-01 2.16770545e-01 4.04628783e-01 -3.51431638e-01
-9.66928422e-01 -1.49417758e+00 6.30093396e-01 -2.07073092e-01
3.22502047e-01 -6.01359844e-01 -9.05184209e-01 5.62394619e-01
2.57751495e-01 4.46958750e-01 6.27008617e-01 -4.65077646e-02
-3.86906743e-01 -2.25799456e-01 -9.16785359e-01 5.20921886e-01
1.06112278e+00 -4.53629315e-01 -5.64831793e-01 3.74971509e-01
9.48278248e-01 -2.54251778e-01 -6.64096832e-01 3.96798879e-01
2.47691855e-01 -1.01728237e+00 1.11125243e+00 -5.71762264e-01
6.01656914e-01 -3.93925428e-01 -3.63321632e-01 -1.10010457e+00
-3.47613961e-01 3.34913954e-02 5.25998235e-01 1.28076255e+00
3.23625952e-01 -3.58593017e-01 7.95467019e-01 4.20192152e-01
-1.31966904e-01 -7.85133362e-01 -8.40943933e-01 -2.69154668e-01
2.57033911e-02 -5.20957053e-01 4.49275225e-01 7.11955428e-01
-6.26958787e-01 2.12808460e-01 -2.03792572e-01 6.57216758e-02
6.36843741e-01 6.63108528e-01 6.34859860e-01 -1.19760549e+00
-1.29773840e-01 -4.06117409e-01 -4.78011042e-01 -1.37595761e+00
8.81529897e-02 -9.09070969e-01 1.05286166e-01 -1.86537337e+00
3.78814697e-01 -5.34663916e-01 -4.31271255e-01 4.44172323e-01
-2.05915868e-01 4.89300281e-01 3.26636344e-01 -5.88194691e-02
-1.11134529e+00 6.05863750e-01 1.35450757e+00 -3.45528454e-01
4.24060933e-02 6.50497228e-02 -7.61112452e-01 8.99755538e-01
6.05024576e-01 -2.91322589e-01 -4.45452660e-01 -4.50093657e-01
-1.36877254e-01 -3.40547144e-01 4.50340062e-01 -1.05596578e+00
2.00509399e-01 -4.08246219e-01 7.10063279e-01 -8.76226187e-01
3.34154248e-01 -9.64580715e-01 -1.52984053e-01 2.83733040e-01
-4.90674645e-01 -2.58257657e-01 1.80166572e-01 4.36584979e-01
-4.36803162e-01 -2.03756124e-01 8.48623514e-01 -3.36566508e-01
-1.54357660e+00 4.36206788e-01 -1.86055183e-01 1.45693973e-01
1.15262043e+00 -3.42168570e-01 -2.05274940e-01 4.09951173e-02
-6.80164576e-01 3.69116902e-01 2.20084310e-01 5.94643414e-01
8.03836644e-01 -1.19403851e+00 -3.62972915e-01 4.00801122e-01
2.17516527e-01 3.93162161e-01 5.76653957e-01 5.24059832e-01
-6.57903314e-01 4.47115868e-01 -4.34639633e-01 -8.82818222e-01
-1.22631168e+00 3.96252304e-01 2.48383403e-01 -1.07608534e-01
-5.58302820e-01 1.20173597e+00 5.06714880e-01 -4.34838086e-01
2.55875606e-02 -6.84234142e-01 -5.54132700e-01 1.45181030e-01
4.52132851e-01 -4.24061529e-02 -4.25884267e-04 -9.08697784e-01
-1.97895423e-01 1.05502164e+00 -2.30188787e-01 3.35163623e-01
1.33165145e+00 -3.75161201e-01 -2.41406068e-01 4.70018059e-01
1.64627194e+00 -4.17647898e-01 -1.49918187e+00 -3.89819592e-01
1.58451542e-01 -5.14837623e-01 -7.98749477e-02 -6.08570814e-01
-1.24120104e+00 1.20207882e+00 6.64623618e-01 1.50264934e-01
1.26018059e+00 2.36711875e-01 7.99174428e-01 2.86055535e-01
2.03515738e-01 -1.22980893e+00 1.04915194e-01 6.26028657e-01
6.07565284e-01 -1.58532679e+00 -1.66215003e-01 -5.43284595e-01
-9.07404959e-01 1.08534110e+00 7.69166529e-01 -2.76784033e-01
7.60974407e-01 -1.14292227e-01 9.46586058e-02 -8.30831826e-02
-3.61012131e-01 -8.05208981e-01 4.89581347e-01 4.95863199e-01
2.24405482e-01 1.24871701e-01 6.47024065e-02 6.07325971e-01
-1.26045138e-01 -2.86914885e-01 2.49331519e-01 5.87541878e-01
-7.25725591e-01 -9.08727288e-01 -3.06655709e-02 7.39010513e-01
-5.18427789e-01 -1.75956145e-01 -1.03809364e-01 5.62912226e-01
3.87191355e-01 6.97261155e-01 3.01515937e-01 -2.74943948e-01
2.96461552e-01 -1.23756699e-01 2.88189173e-01 -6.84090436e-01
-4.53083605e-01 5.69668300e-02 -3.77670944e-01 -7.92273581e-01
-6.03099823e-01 -6.26455426e-01 -1.42927635e+00 2.12119520e-01
-1.83036793e-02 -9.20450985e-02 8.04216981e-01 1.03203189e+00
1.63451925e-01 7.75848806e-01 6.13146961e-01 -8.13065886e-01
-7.82245696e-02 -6.81373537e-01 -6.69264853e-01 6.88693106e-01
3.29532921e-01 -4.97044563e-01 -4.04769033e-02 2.99065292e-01] | [9.609635353088379, 0.37397605180740356] |
ce7915b5-7cb5-4bbe-899b-04d2d024e994 | investigation-of-multilingual-neural-machine | null | null | https://aclanthology.org/2022.wat-1.9 | https://aclanthology.org/2022.wat-1.9.pdf | Investigation of Multilingual Neural Machine Translation for Indian Languages | In the domain of natural language processing, machine translation is a well-defined task where one natural language is automatically translated to another natural language. The deep learning-based approach of machine translation, known as neural machine translation attains remarkable translational performance. However, it requires a sufficient amount of training data which is a critical issue for low-resource pair translation. To handle the data scarcity problem, the multilingual concept has been investigated in neural machine translation in different settings like many-to-one and one-to-many translation. WAT2022 (Workshop on Asian Translation 2022) organizes (hosted by the COLING 2022) Indic tasks: English-to-Indic and Indic-to-English translation tasks where we have participated as a team named CNLP-NITS-PP. Herein, we have investigated a transliteration-based approach, where Indic languages are transliterated into English script and shared sub-word level vocabulary during the training phase. We have attained BLEU scores of 2.0 (English-to-Bengali), 1.10 (English-to-Assamese), 4.50 (Bengali-to-English), and 3.50 (Assamese-to-English) translation, respectively. | ['Sivaji Bandyopadhyay', 'Partha Pakray', 'Riyanka Manna', 'Sahinur Rahman Laskar'] | null | null | null | null | wat-2022-10 | ['transliteration'] | ['natural-language-processing'] | [ 2.43111491e-01 -8.56902003e-02 -3.05272788e-01 -4.58972871e-01
-1.39352584e+00 -8.51688147e-01 8.01221311e-01 -2.61026263e-01
-5.77422500e-01 1.34724391e+00 2.32949466e-01 -9.25008416e-01
4.37452316e-01 -7.09319711e-01 -9.91299391e-01 -3.30094516e-01
5.97130001e-01 9.16145384e-01 -4.92727935e-01 -5.81992149e-01
5.89687265e-02 2.32979044e-01 -5.70411086e-01 5.31577766e-01
1.05820394e+00 4.42274153e-01 2.54659265e-01 4.53797996e-01
-2.75024831e-01 2.77466565e-01 -2.75845379e-01 -7.52196431e-01
4.13567334e-01 -1.13217628e+00 -1.14777231e+00 -4.81331110e-01
2.33452618e-01 -9.86010879e-02 1.87390670e-01 1.01307535e+00
3.93476576e-01 -4.22844708e-01 5.76155543e-01 -8.65465283e-01
-1.08902836e+00 7.46571183e-01 -4.26401258e-01 6.91506825e-03
1.44335136e-01 4.39153984e-03 9.63549733e-01 -1.37844491e+00
9.05038178e-01 1.05841517e+00 4.54262882e-01 5.38472295e-01
-1.09840226e+00 -4.17188019e-01 -5.13832748e-01 -4.84160818e-02
-1.23981154e+00 -5.03317535e-01 1.91969588e-01 -3.57773542e-01
1.24052489e+00 2.54496008e-01 3.15196425e-01 1.21666932e+00
6.48268044e-01 4.74195480e-01 1.52637994e+00 -7.87127018e-01
-4.82490435e-02 3.03525120e-01 -3.78002077e-01 3.62135291e-01
5.88428974e-02 6.02125563e-02 -7.66727865e-01 1.33974239e-01
5.57616711e-01 -4.75700498e-01 2.25906894e-02 5.77127039e-01
-1.72264230e+00 9.08816636e-01 2.31453717e-01 6.52037203e-01
-6.61629260e-01 -4.97453548e-02 5.45345306e-01 9.25102472e-01
7.50255823e-01 7.40144074e-01 -8.23977351e-01 -4.53389615e-01
-8.17002773e-01 6.84025213e-02 8.66804481e-01 1.16372049e+00
7.43994117e-01 2.45575681e-01 2.12998949e-02 9.36097383e-01
-3.23864035e-02 7.61943996e-01 7.23845243e-01 -5.22009313e-01
9.45319355e-01 1.98120520e-01 1.37153059e-01 -6.19467199e-01
-5.38550392e-02 -3.52806032e-01 -9.89568472e-01 -3.45092893e-01
3.90757412e-01 -4.35688525e-01 -8.02051306e-01 1.71055341e+00
1.08784296e-01 -6.49587214e-01 5.11627495e-01 9.34562981e-01
4.12028581e-01 1.04835618e+00 -4.67517525e-02 -3.98284942e-01
1.27068913e+00 -1.31514072e+00 -6.01315022e-01 -3.64412844e-01
6.77104771e-01 -1.38272488e+00 1.21772921e+00 5.24592921e-02
-1.33783519e+00 -4.75082934e-01 -7.14211762e-01 -5.61358519e-02
-5.26009440e-01 3.69753391e-01 2.39548177e-01 3.66791338e-01
-1.28760755e+00 5.17045498e-01 -6.14080608e-01 -8.83463800e-01
-1.25882015e-01 3.79400551e-01 -7.13362992e-01 -1.71512008e-01
-1.57653213e+00 1.36621046e+00 4.28454578e-01 1.78892046e-01
-6.68093801e-01 -2.46278524e-01 -5.35274088e-01 -3.17184895e-01
-1.36313587e-01 -6.28849149e-01 1.21046102e+00 -1.36985958e+00
-1.79771721e+00 1.22643435e+00 -2.95630693e-01 -4.40692872e-01
6.45940721e-01 -2.54214674e-01 -5.40760458e-01 -9.00105163e-02
3.85191739e-01 5.60161948e-01 4.63530362e-01 -6.86572790e-01
-5.39876044e-01 -3.80154133e-01 -2.50201464e-01 2.98045725e-01
-1.47616386e-01 9.10442710e-01 -6.26558289e-02 -6.59836650e-01
1.31849065e-01 -1.19258881e+00 -3.88947576e-02 -4.08766568e-01
-2.69141287e-01 -8.95899087e-02 3.43487710e-01 -1.27277255e+00
8.14600527e-01 -1.75181568e+00 3.75924885e-01 -2.73149878e-01
-3.51804107e-01 4.00211364e-01 -2.70682216e-01 9.50810254e-01
-4.20918651e-02 2.73013353e-01 -3.18970025e-01 -2.95164943e-01
-1.97613053e-02 6.22918122e-02 -3.73637885e-01 2.90411681e-01
4.61179465e-01 1.30138791e+00 -7.76240885e-01 -3.53836000e-01
-2.95741260e-01 1.78427577e-01 -5.61413094e-02 3.77397329e-01
-7.64886886e-02 6.52913749e-01 -3.33824694e-01 7.65455365e-01
5.67021430e-01 9.44728032e-02 1.33687437e-01 5.88916279e-02
-4.68038172e-01 6.07783735e-01 -2.22065747e-01 1.80452263e+00
-7.96892643e-01 6.91736042e-01 -8.36344436e-02 -8.36061358e-01
1.02540612e+00 5.57515502e-01 1.11929528e-01 -8.29844892e-01
5.26638776e-02 1.05530322e+00 3.16300154e-01 -2.98636079e-01
7.26639390e-01 -4.76709723e-01 -3.01317126e-01 6.03832245e-01
1.62207857e-01 -2.16931805e-01 1.94279835e-01 -2.84754634e-01
7.43563652e-01 5.21920085e-01 4.20695007e-01 -5.29539227e-01
2.11022660e-01 5.55824101e-01 4.86168474e-01 2.34322846e-01
-3.67516018e-02 6.03334785e-01 2.21713632e-01 -5.97869515e-01
-1.56080496e+00 -7.79427409e-01 1.02012455e-01 1.11589444e+00
-4.52667773e-01 -1.03385776e-01 -1.01817095e+00 -4.26794589e-01
-6.36015117e-01 8.22501063e-01 -3.78775984e-01 -6.57298341e-02
-1.07968473e+00 -7.04146504e-01 9.66966689e-01 1.57258883e-01
6.41456664e-01 -1.23583210e+00 -1.07101679e-01 5.62847435e-01
-5.93532145e-01 -1.28349280e+00 -7.45567501e-01 2.01814339e-01
-6.61261618e-01 -3.35785300e-01 -1.07740653e+00 -1.09918046e+00
4.85064179e-01 -1.62674949e-01 1.09875059e+00 -3.17380220e-01
2.10261613e-01 -3.12404156e-01 -4.20721412e-01 -3.36596876e-01
-9.75866377e-01 4.80284691e-01 3.31351250e-01 -9.89050046e-03
6.06704116e-01 -3.41564804e-01 -1.87835783e-01 2.91426361e-01
-6.19789779e-01 4.65883344e-01 9.34403181e-01 9.02434766e-01
5.38706064e-01 -9.69288349e-01 7.50668287e-01 -7.48532414e-01
6.49298191e-01 -3.57375920e-01 -3.99441868e-01 5.10458112e-01
-5.44714749e-01 -1.67119041e-01 1.16278327e+00 -3.93616378e-01
-8.06247771e-01 -4.25060391e-02 -1.57423183e-01 -5.98290283e-03
2.07607057e-02 7.95201957e-01 -2.11966321e-01 1.44938529e-01
7.05443025e-01 6.56931579e-01 -2.56388426e-01 -1.85609117e-01
4.57735687e-01 1.05094504e+00 4.61434901e-01 -6.21864378e-01
9.05820966e-01 -2.11127743e-01 -2.85892218e-01 -4.79963720e-01
-4.35566992e-01 -3.13556846e-03 -9.57256138e-01 1.12883270e-01
1.15602195e+00 -1.08326936e+00 1.79260373e-01 6.33446157e-01
-1.70530307e+00 -4.14824754e-01 -1.92655995e-02 7.26292908e-01
-5.44634938e-01 -2.44496122e-01 -9.81305659e-01 -1.84116572e-01
-9.01161611e-01 -1.21361411e+00 1.01047349e+00 -6.91588223e-02
-4.71929520e-01 -8.00967991e-01 2.37790927e-01 3.99834991e-01
8.21645319e-01 1.34368986e-01 1.05106151e+00 -8.79638314e-01
-3.78835112e-01 -3.42229493e-02 -3.53910029e-01 5.34573495e-01
2.78126210e-01 -2.39712358e-01 -2.85591245e-01 -1.85469985e-01
-9.38444808e-02 -5.85358143e-01 3.31109643e-01 -1.15057938e-01
1.81059524e-01 -2.61844635e-01 9.99155194e-02 3.11973155e-01
1.28593671e+00 2.29748726e-01 6.01883054e-01 5.13879299e-01
5.15661240e-01 4.04632032e-01 6.66625321e-01 -1.58316791e-01
5.03365457e-01 7.03073025e-01 -2.24054471e-01 -3.41596723e-01
-1.42356157e-01 -3.37024510e-01 7.39134014e-01 1.63050044e+00
-1.30392522e-01 -1.98697582e-01 -1.02675915e+00 6.97507620e-01
-1.50458813e+00 -4.47679043e-01 -3.78798336e-01 2.02428007e+00
1.35708642e+00 -7.72134513e-02 -2.25467443e-01 -4.63099092e-01
7.64479518e-01 -2.21085295e-01 -2.62201399e-01 -1.09615946e+00
-2.87721574e-01 4.65723723e-01 4.15493637e-01 5.53427517e-01
-6.11481965e-01 1.69588757e+00 5.61240768e+00 7.06679583e-01
-1.51689947e+00 4.43528712e-01 7.16784954e-01 4.48905408e-01
-2.55038530e-01 2.45729133e-01 -7.84672678e-01 3.17539781e-01
1.53440845e+00 -3.24542373e-01 7.71068692e-01 4.39739227e-01
2.61157811e-01 2.59888768e-01 -1.18751287e+00 5.56866169e-01
-3.20674405e-02 -1.29983819e+00 2.61429965e-01 7.66037181e-02
9.82809186e-01 4.96337891e-01 -1.85891241e-01 5.30179918e-01
1.21757023e-01 -1.06006634e+00 7.54402578e-01 3.33447486e-01
1.28608930e+00 -6.88103139e-01 7.85171032e-01 4.06796008e-01
-8.86307836e-01 5.40266633e-01 -5.69903493e-01 -4.75377738e-02
2.84134269e-01 5.10802209e-01 -8.64444733e-01 8.29889536e-01
3.89039099e-01 3.50848168e-01 -1.75589725e-01 2.63809443e-01
-3.15090716e-01 8.39294255e-01 -9.21510682e-02 -2.50782818e-01
5.39272547e-01 -7.58566499e-01 3.32254529e-01 1.26360404e+00
7.61438310e-01 -1.25782520e-01 4.65520546e-02 8.31932843e-01
-4.72406387e-01 6.17436230e-01 -6.83950961e-01 -4.39371079e-01
1.10734038e-01 1.07563770e+00 -5.10788143e-01 -4.44306672e-01
-3.21777880e-01 1.54190779e+00 3.11991066e-01 3.59400451e-01
-7.01188207e-01 -5.86420596e-01 4.00902212e-01 -1.25061437e-01
-4.33217436e-02 -4.64113981e-01 -1.86157361e-01 -1.15789711e+00
1.97485089e-01 -1.40859687e+00 -4.05894190e-01 -7.83259928e-01
-1.12841463e+00 1.28913772e+00 -3.48046124e-01 -1.20553756e+00
-5.44256091e-01 -4.98791844e-01 -4.71531302e-01 1.46466088e+00
-1.19891763e+00 -1.68083131e+00 3.92582744e-01 3.10839832e-01
8.33326161e-01 -6.53167665e-01 1.19042873e+00 5.71051300e-01
-2.86511689e-01 6.17248535e-01 5.43845057e-01 2.94909447e-01
1.04242516e+00 -8.86175692e-01 9.81637657e-01 9.10809398e-01
1.56598285e-01 8.41712296e-01 5.14587045e-01 -5.87568760e-01
-1.54237449e+00 -1.31461430e+00 1.97216022e+00 -3.03516239e-01
8.09864819e-01 -4.52162027e-01 -6.20588422e-01 6.63329244e-01
8.76966953e-01 -3.84836107e-01 6.34428740e-01 -2.52453864e-01
-1.01735830e-01 6.05383217e-02 -9.73604083e-01 7.79085875e-01
7.00622976e-01 -7.16704726e-01 -5.12530923e-01 7.29811430e-01
1.16389227e+00 -4.38816100e-01 -8.84925604e-01 1.66565880e-01
5.17307281e-01 -2.67288238e-01 4.25936937e-01 -1.04406476e+00
1.01590991e+00 -2.04754144e-01 -5.06785810e-01 -1.62063169e+00
5.74527532e-02 -8.67786705e-01 5.69219530e-01 1.22331679e+00
9.28372562e-01 -6.92593753e-01 2.76563257e-01 2.96127647e-01
-2.22514734e-01 -6.54354572e-01 -1.24944353e+00 -8.43711674e-01
8.03055644e-01 1.10009678e-01 5.29740572e-01 1.22315705e+00
-2.00997487e-01 7.53127277e-01 -7.17155278e-01 -2.40677893e-01
2.39858925e-02 2.25133792e-01 8.03704977e-01 -6.59431219e-01
-3.83715928e-01 -2.56559372e-01 7.20979869e-02 -8.45819712e-01
1.79478019e-01 -1.33156025e+00 8.05272833e-02 -1.51940393e+00
5.96873499e-02 -7.42796659e-02 -1.85381174e-02 4.78649169e-01
-3.45342085e-02 4.67361212e-01 1.30633816e-01 3.66256654e-01
5.79470843e-02 3.94461811e-01 1.24038041e+00 -7.97111019e-02
8.71558785e-02 -2.75157839e-02 -5.88870049e-01 1.28220379e-01
1.05844283e+00 -7.73246467e-01 1.95665896e-01 -8.92450035e-01
4.05950636e-01 1.82463810e-01 -1.86041951e-01 -5.97989440e-01
3.13798040e-02 -3.88594687e-01 3.89862508e-02 -3.55060756e-01
1.27724692e-01 -5.03971636e-01 2.48294771e-01 5.74531257e-01
-4.61062968e-01 7.54516006e-01 4.75356355e-02 -2.27029651e-01
-5.32972276e-01 -1.92541167e-01 6.49778247e-01 -4.24396932e-01
-2.68191844e-01 1.94102556e-01 -5.80042720e-01 8.93500745e-02
8.25533628e-01 1.11572169e-01 -1.97117001e-01 -2.53825247e-01
-2.99053490e-01 -1.63353577e-01 3.41349036e-01 5.14233708e-01
3.63180071e-01 -1.44951332e+00 -1.42602706e+00 3.87177877e-02
6.23249449e-02 -5.39407670e-01 -4.54978555e-01 9.70819890e-01
-8.63510966e-01 6.16676033e-01 -4.65590477e-01 -4.13535893e-01
-8.74104559e-01 1.88157633e-01 1.85682103e-01 -4.64296758e-01
-4.79217693e-02 7.06481516e-01 -2.41067588e-01 -1.19582045e+00
-4.41126317e-01 -1.14754699e-01 4.45074052e-01 -2.80032665e-01
3.14606100e-01 2.46339776e-02 3.50006044e-01 -8.99977863e-01
-2.74674207e-01 5.50631523e-01 -1.24126151e-01 -4.49573278e-01
1.17267609e+00 3.43481153e-02 -7.66040146e-01 5.16117156e-01
1.35409045e+00 1.04968391e-01 -2.16816604e-01 -3.06429267e-01
2.37424985e-01 -4.22963500e-02 -3.87639731e-01 -1.06077814e+00
-5.60477197e-01 1.08996415e+00 4.07218307e-01 -3.17612410e-01
9.14036572e-01 -3.97537112e-01 1.20732880e+00 5.92965961e-01
6.41255796e-01 -1.21604693e+00 -4.06811267e-01 9.80846882e-01
1.05881572e+00 -1.44401193e+00 -3.55607450e-01 6.42478466e-02
-7.25890279e-01 1.23575187e+00 3.80753040e-01 1.55048981e-01
2.04129755e-01 9.04431045e-02 5.45682192e-01 3.61850291e-01
-6.81654215e-01 1.30404189e-01 2.08235249e-01 3.08107972e-01
9.06811476e-01 3.61718297e-01 -8.05971622e-01 3.25297326e-01
-4.77313280e-01 8.00227225e-02 2.48050839e-01 7.75236368e-01
-6.39049560e-02 -1.66374016e+00 -3.05668861e-01 1.63694426e-01
-6.73069000e-01 -7.39723802e-01 -6.71317518e-01 7.44905949e-01
-1.18654087e-01 9.71005976e-01 -3.64399999e-01 -2.95732141e-01
1.91184431e-01 1.51570767e-01 2.91869134e-01 -6.05686009e-01
-1.09263217e+00 8.38724151e-02 2.90669590e-01 -1.30183756e-01
-2.71491706e-01 -4.19479132e-01 -1.09652567e+00 -4.35962230e-01
-1.68164626e-01 4.76920485e-01 1.19716096e+00 1.04397261e+00
3.70573133e-01 1.47823840e-01 6.01767838e-01 -5.31880558e-01
-6.88560307e-01 -1.30339527e+00 -1.66401774e-01 -5.70828132e-02
-5.73889948e-02 2.94200033e-01 8.89860168e-02 1.24734625e-01] | [11.549581527709961, 10.368590354919434] |
66fc5bdc-7e21-4e01-800a-bfbd51f2d2a5 | group-activity-recognition-by-using-effective | null | null | https://ieeexplore.ieee.org/abstract/document/9031366 | https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9031366 | Group Activity Recognition by Using Effective Multiple Modality Relation Representation With Temporal-Spatial Attention | Group activity recognition has received a great deal of interest because of its broader applications in sports analysis, autonomous vehicles, CCTV surveillance systems and video summarization systems. Most existing methods typically use appearance features and they seldom consider underlying interaction information. In this work, a technology of novel group activity recognition is proposed based on multi-modal relation representation with temporal-spatial attention. First, we introduce an object relation module, which processes all objects in a scene simultaneously through an interaction between their appearance feature and geometry, thus allowing the modeling of their relations. Second, to extract effective motion features, an optical flow network is fine-tuned by using the action loss as the supervised signal. Then, we propose two types of inference models, opt-GRU and relation-GRU, which are used to encode the object relationship and motion representation effectively, and form the discriminative frame-level feature representation. Finally, an attention-based temporal aggregation layer is proposed to integrate frame-level features with different weights and form effective video-level representations. We have performed extensive experiments on two popular datasets, and both have achieved state-of-the-art performance. The datasets are the Volleyball dataset and the Collective Activity dataset, respectively | ['AND XU LIU', 'Dong Wang', 'Meng Jian', 'Lifang Wu', 'HENG FU', 'Dezhong Xu'] | 2020-03-10 | null | null | null | ieee-access-2020-3 | ['group-activity-recognition'] | ['computer-vision'] | [ 1.75308064e-01 -6.12457335e-01 -4.50539798e-01 -3.02734494e-01
-3.21871877e-01 -1.31450649e-02 4.97093886e-01 9.34142023e-02
-4.93093550e-01 4.94356245e-01 4.41119611e-01 2.46559769e-01
-2.75236815e-01 -8.91088903e-01 -5.19560158e-01 -9.27676320e-01
-7.10935891e-02 -7.51721784e-02 5.89583158e-01 -9.93023440e-02
5.03544271e-01 2.44174004e-01 -1.65170693e+00 2.31948137e-01
8.72854054e-01 1.10986853e+00 7.21248835e-02 5.93119681e-01
-6.34493381e-02 1.28771269e+00 -3.36030811e-01 2.36727875e-02
3.53471888e-03 -6.91739678e-01 -6.20177865e-01 6.14844739e-01
4.05869603e-01 -3.50481361e-01 -8.12579870e-01 7.40079105e-01
5.85825920e-01 7.31403887e-01 5.60239553e-01 -1.14694393e+00
-4.09273684e-01 -2.32950523e-02 -7.50218809e-01 7.53288150e-01
5.83436549e-01 5.55635989e-01 8.95124197e-01 -5.94562411e-01
6.04433596e-01 1.28639972e+00 4.95830849e-02 2.13846818e-01
-9.19773698e-01 -6.27725005e-01 4.59892601e-01 9.08924401e-01
-1.29792058e+00 -3.22390944e-01 1.06959748e+00 -5.78929305e-01
6.38358116e-01 1.03356153e-01 1.13287151e+00 6.62105322e-01
2.04949737e-01 1.21964157e+00 6.44974232e-01 -1.79148853e-01
-1.19614862e-01 -4.38544542e-01 2.49409199e-01 8.51889431e-01
-2.74719074e-02 -6.57715797e-02 -7.15067267e-01 2.29419261e-01
1.04681385e+00 3.86551648e-01 -4.23222005e-01 -2.60419130e-01
-1.36956215e+00 7.75055528e-01 4.16510969e-01 3.81642222e-01
-3.94747257e-01 1.77635282e-01 4.95835721e-01 3.92834544e-02
5.47050893e-01 -1.10179901e-01 5.00559658e-02 -3.10010314e-01
-5.48052311e-01 2.65411288e-01 3.10389310e-01 6.70999587e-01
8.57299745e-01 -1.40332863e-01 -6.32366061e-01 9.52493727e-01
4.10442382e-01 8.85468572e-02 5.91119587e-01 -1.01498485e+00
6.55724525e-01 8.87702525e-01 -2.16367081e-01 -1.51374459e+00
-3.04897010e-01 -3.43633175e-01 -8.39123189e-01 -6.37367666e-02
1.56814232e-01 1.65568143e-01 -5.77266216e-01 1.52000082e+00
5.29285491e-01 8.50738466e-01 -1.94055960e-01 1.18634236e+00
8.91084671e-01 8.28069389e-01 1.51108027e-01 -6.30070150e-01
1.43239558e+00 -1.23303223e+00 -8.20029378e-01 1.91230085e-02
5.93280554e-01 -6.01461351e-01 8.37454319e-01 4.08168852e-01
-1.14935601e+00 -8.98192108e-01 -1.00408721e+00 -2.64493406e-01
-1.62780080e-02 1.22517459e-01 7.38120615e-01 1.05333738e-01
-4.69895720e-01 5.35963953e-01 -8.49112630e-01 -2.47913435e-01
6.75817370e-01 2.30868712e-01 -3.03151071e-01 -1.77156284e-01
-9.47126031e-01 4.99891758e-01 4.10333991e-01 4.13225412e-01
-7.19968975e-01 -1.60919815e-01 -9.76316214e-01 -4.35760394e-02
4.83637482e-01 -7.18963683e-01 8.66597712e-01 -1.00167131e+00
-1.60485291e+00 6.18795514e-01 -2.23900959e-01 -6.03857497e-03
3.71049762e-01 -1.87938377e-01 -5.31260610e-01 4.22566265e-01
7.23071173e-02 3.00527453e-01 5.68458140e-01 -6.34035468e-01
-8.99374843e-01 -4.65486974e-01 1.73294634e-01 7.19385862e-01
-3.70227367e-01 9.85361636e-02 -7.80756176e-01 -7.48276412e-01
3.77974302e-01 -6.64056301e-01 -2.52999723e-01 6.71732277e-02
-8.46398249e-02 -6.01091623e-01 9.62845922e-01 -6.04427278e-01
1.45615184e+00 -2.11130285e+00 5.91372550e-01 -3.90970632e-02
1.84916764e-01 2.70568162e-01 -4.28234264e-02 1.04417861e-01
1.31174624e-01 -1.90604508e-01 -1.00115620e-01 -9.89260003e-02
-4.67579603e-01 2.72110373e-01 8.71650279e-02 5.80432236e-01
2.03880996e-01 7.57787764e-01 -9.69429195e-01 -9.69472289e-01
6.65067554e-01 1.92594081e-01 -5.73400736e-01 3.28936517e-01
-1.01189919e-01 6.71390593e-01 -9.34299171e-01 4.78704810e-01
5.50806820e-01 -6.58127144e-02 -1.26575202e-01 -3.75148267e-01
-3.66161019e-01 -2.08966378e-02 -1.31783450e+00 2.02291369e+00
-2.27505058e-01 4.88774508e-01 -2.51105070e-01 -1.36131775e+00
7.97136903e-01 2.85589192e-02 9.84424412e-01 -7.54523337e-01
1.76855713e-01 -1.47776246e-01 1.64707676e-02 -1.01558626e+00
3.62006724e-01 2.14755446e-01 3.35437357e-02 2.77694464e-01
8.45945999e-02 2.33694687e-01 4.31763142e-01 2.07128063e-01
9.70398009e-01 3.87362987e-01 9.69553590e-02 -2.49184128e-02
1.12219048e+00 -2.40353465e-01 9.72997367e-01 3.04774702e-01
-3.50307286e-01 5.16757548e-01 3.27318579e-01 -6.67220414e-01
-5.75603783e-01 -6.88475251e-01 8.38283971e-02 9.09822941e-01
8.42067182e-01 -4.47616339e-01 -4.09949154e-01 -5.82967937e-01
-3.12918216e-01 1.38901561e-01 -5.77452838e-01 -5.10839343e-01
-9.08154547e-01 -7.15806663e-01 5.64678535e-02 6.96628094e-01
9.87787008e-01 -1.33434629e+00 -4.48801398e-01 4.28497285e-01
-4.78896201e-01 -1.03326738e+00 -7.89092779e-01 -5.84787250e-01
-8.84105921e-01 -1.14117801e+00 -6.05675161e-01 -7.84611642e-01
5.41856825e-01 4.96254057e-01 6.42135859e-01 3.96798819e-01
-4.66951102e-01 2.62844682e-01 -4.67104942e-01 -2.91195493e-02
3.69328916e-01 -3.33555811e-03 -5.63969575e-02 6.54308915e-01
3.45891148e-01 -6.99515343e-01 -9.48796213e-01 4.91030335e-01
-8.82897735e-01 1.24663830e-01 6.15282595e-01 7.63492823e-01
7.20341086e-01 -2.02428386e-01 2.09960267e-01 -3.88538688e-01
3.30065340e-01 -3.31484735e-01 -2.39882439e-01 1.77413806e-01
-8.07699636e-02 -1.60043672e-01 3.24377477e-01 -6.07012093e-01
-1.11790323e+00 -8.76043141e-02 4.11919802e-02 -5.89554191e-01
8.04778375e-03 4.56357509e-01 -4.13371772e-01 8.34753830e-03
2.55220592e-01 5.01192510e-01 -1.84479561e-02 -3.70281130e-01
2.31016695e-01 5.11676610e-01 5.76583207e-01 -6.01663530e-01
4.85675544e-01 5.83161056e-01 1.34997606e-01 -7.43998766e-01
-9.43376303e-01 -7.85348594e-01 -6.62254035e-01 -6.15950644e-01
1.27040493e+00 -8.46352875e-01 -9.89497721e-01 8.59174430e-01
-1.12306583e+00 6.82051033e-02 -1.88814372e-01 1.05121994e+00
-6.65675640e-01 7.70399928e-01 -7.23573387e-01 -6.55999780e-01
-1.68930203e-01 -1.20115268e+00 1.03251219e+00 7.31064916e-01
3.61202061e-01 -8.05066109e-01 7.66060129e-02 5.89890480e-01
-9.94545519e-02 4.19952720e-01 3.97797823e-01 -1.00266732e-01
-9.50960934e-01 2.27721632e-02 -2.68957734e-01 3.81181031e-01
2.23380074e-01 2.56738644e-02 -5.50820649e-01 -1.51818946e-01
-1.35944501e-01 -1.33726478e-01 1.02803612e+00 6.03368521e-01
1.61119366e+00 -8.87401178e-02 -4.12215799e-01 7.69060910e-01
1.04138720e+00 3.41335237e-01 9.31075692e-01 1.39089748e-01
1.03212917e+00 4.66290146e-01 9.03693140e-01 4.83717352e-01
4.47622746e-01 9.79228854e-01 2.78647095e-01 -1.21997064e-02
-5.30083589e-02 -1.63463727e-01 3.62502813e-01 1.09139431e+00
-7.63465703e-01 -6.65633678e-02 -3.26300472e-01 2.72412986e-01
-2.43296170e+00 -1.47216916e+00 -3.54083747e-01 2.05003333e+00
4.43178296e-01 1.08466588e-01 2.80094326e-01 -4.45985235e-02
8.01095903e-01 4.45533365e-01 -5.36040127e-01 2.23196652e-02
-9.64190140e-02 -2.72487938e-01 1.92282964e-02 2.38764480e-01
-1.25494182e+00 7.47993588e-01 5.20159531e+00 1.08155203e+00
-8.96626115e-01 1.25484213e-01 6.33428574e-01 -2.20727593e-01
5.28804809e-02 1.76779166e-01 -5.34690380e-01 7.16004014e-01
2.08863348e-01 -9.44704860e-02 1.87888786e-01 5.01702785e-01
4.69600320e-01 -2.41185352e-01 -7.75318265e-01 1.23239243e+00
2.74055958e-01 -1.32922757e+00 4.34451662e-02 2.16831103e-01
4.29100305e-01 -5.09619713e-01 -3.47201645e-01 3.43147546e-01
-2.27436364e-01 -6.71653032e-01 5.57030559e-01 1.17111564e+00
3.62380862e-01 -7.38430321e-01 5.76346457e-01 3.66263747e-01
-1.70302713e+00 -1.86794639e-01 -3.50171000e-01 -3.42754304e-01
4.33691144e-01 3.75894666e-01 1.07705422e-01 9.32360351e-01
5.85253537e-01 1.46408010e+00 -3.57725918e-01 1.28341341e+00
-1.33334279e-01 5.73578238e-01 -2.62317866e-01 -1.29672363e-01
1.45996913e-01 -5.41722238e-01 5.73750794e-01 9.03734684e-01
2.07516365e-02 6.12902701e-01 7.56691337e-01 4.90224659e-01
1.43291131e-01 2.92259187e-01 -1.94800586e-01 2.65570521e-01
9.90032181e-02 1.36870992e+00 -6.48799002e-01 -4.86805975e-01
-5.06072402e-01 9.02024448e-01 2.49720886e-01 2.71923512e-01
-9.49496269e-01 -4.05189842e-01 6.33948863e-01 9.86090228e-02
1.55156761e-01 -3.06629479e-01 2.34381601e-01 -1.61817980e+00
2.12408379e-01 -5.16328931e-01 7.21319675e-01 -7.45485783e-01
-1.09476554e+00 1.31944761e-01 9.92774516e-02 -1.69811940e+00
-3.09541691e-02 -3.13921422e-01 -9.93119836e-01 6.52655542e-01
-1.30258071e+00 -1.12242556e+00 -7.22418189e-01 7.74275601e-01
8.40662837e-01 -1.68199688e-01 1.42779604e-01 6.86677039e-01
-1.00396931e+00 3.22067082e-01 -2.54021913e-01 3.34543973e-01
4.67225879e-01 -7.17904747e-01 -2.71134228e-01 8.51327300e-01
1.31530821e-01 2.90489316e-01 2.54956514e-01 -5.13092935e-01
-1.34370065e+00 -1.07203770e+00 4.22526598e-01 -2.23651871e-01
3.22376668e-01 -4.90400195e-02 -8.48778486e-01 4.96672988e-01
9.44316760e-03 4.81903166e-01 4.43444490e-01 -1.56856835e-01
1.90004304e-01 -3.20678025e-01 -7.08736479e-01 4.25429672e-01
1.51362026e+00 -2.42997244e-01 -5.73299468e-01 3.59597564e-01
6.45872533e-01 -3.60544741e-01 -8.93344879e-01 4.50821042e-01
5.46378136e-01 -9.24072146e-01 8.91523600e-01 -6.46189928e-01
4.83423471e-01 -7.49957740e-01 8.52537379e-02 -9.43330348e-01
-7.29451478e-01 -4.25507843e-01 -3.11078012e-01 1.31955147e+00
-2.90281892e-01 -3.47312182e-01 7.35404074e-01 2.37059876e-01
-2.79471010e-01 -8.88153851e-01 -8.93851995e-01 -5.18198431e-01
-5.15517056e-01 -3.30249012e-01 3.61997217e-01 9.57615495e-01
-3.65071208e-03 4.59109277e-01 -5.03015935e-01 -2.00025082e-01
4.89570349e-01 2.55784631e-01 9.26563621e-01 -1.15249574e+00
-4.10903364e-01 -5.67520201e-01 -1.06452966e+00 -1.40514934e+00
3.66741717e-02 -6.65413737e-01 -1.93485901e-01 -1.60681248e+00
3.67850095e-01 -1.49170578e-01 -4.15847719e-01 2.19874159e-01
-3.31351072e-01 2.05659956e-01 1.35568336e-01 4.00087774e-01
-1.03535092e+00 9.22413945e-01 1.70047998e+00 -2.28678763e-01
-2.62035251e-01 5.39989676e-03 -4.04066592e-01 7.66101420e-01
4.14954245e-01 -1.94391403e-02 -4.13030326e-01 -3.63930166e-01
-2.57508695e-01 2.53947645e-01 5.63715577e-01 -1.18504310e+00
3.29639524e-01 -4.83140230e-01 4.34962302e-01 -6.29007518e-01
4.40109551e-01 -4.44229841e-01 -1.23939224e-01 3.77755761e-01
-2.97000378e-01 -2.05231622e-01 -2.73398727e-01 8.36795092e-01
-4.76959229e-01 3.34645882e-02 5.57718694e-01 -4.25461642e-02
-1.05021775e+00 9.74376559e-01 -2.28124648e-01 5.13222404e-02
1.21902204e+00 -4.65988338e-01 -9.77237150e-02 -2.71405101e-01
-7.68438637e-01 5.40314734e-01 1.02097146e-01 6.01830661e-01
7.19622850e-01 -1.73844266e+00 -7.77650297e-01 2.40757823e-01
2.23748922e-01 2.41142392e-01 8.64738524e-01 1.03150797e+00
-4.88077641e-01 -6.80160597e-02 -3.27275038e-01 -8.84117067e-01
-1.15149450e+00 4.62804824e-01 3.38670939e-01 -2.48151422e-01
-7.07575619e-01 6.84206069e-01 3.84533346e-01 1.58660650e-01
-5.02701625e-02 -3.19175571e-01 -6.62214756e-01 5.58675714e-02
7.17750251e-01 5.35225987e-01 -3.21491539e-01 -1.04946113e+00
-2.82570750e-01 1.03504360e+00 3.21317650e-02 4.27466854e-02
1.13962865e+00 -1.89936399e-01 -7.74690658e-02 3.73816222e-01
1.22483504e+00 -3.67816895e-01 -1.50706518e+00 -4.17144299e-01
-4.45380539e-01 -8.44495773e-01 1.22555643e-01 -4.91265245e-02
-1.35646009e+00 1.14740634e+00 5.82061291e-01 9.01076347e-02
1.27777803e+00 -4.46650460e-02 8.83090436e-01 9.83462855e-02
2.61702955e-01 -1.20675850e+00 5.33949494e-01 2.95362324e-01
6.76637590e-01 -1.00577700e+00 2.22351849e-01 -5.59161544e-01
-5.93458891e-01 1.07779694e+00 9.83646452e-01 -4.05237585e-01
5.52685857e-01 -3.41037720e-01 -3.57094526e-01 -1.76081210e-01
-6.78748429e-01 -2.93379992e-01 5.50615847e-01 4.91380274e-01
2.96219200e-01 -2.20510587e-01 -7.46435225e-01 6.59325302e-01
4.03270751e-01 2.29382724e-01 2.31794462e-01 8.45560074e-01
-5.73290884e-01 -9.64953601e-01 -1.89188421e-01 5.46583533e-01
-2.12346137e-01 3.18249375e-01 1.08645238e-01 4.83248085e-01
5.11964977e-01 8.49552810e-01 2.79223293e-01 -5.90895653e-01
4.63512957e-01 -3.96460116e-01 5.85401893e-01 -5.58889687e-01
-3.21408570e-01 1.84200630e-01 -7.04232045e-03 -7.52486110e-01
-1.01384318e+00 -8.84196401e-01 -1.25255299e+00 -1.37715280e-01
-3.63391101e-01 1.01581961e-01 4.11563218e-02 1.14454973e+00
2.26823717e-01 6.84603333e-01 8.08500230e-01 -1.02907968e+00
-2.99643502e-02 -8.81067812e-01 -6.51656032e-01 8.01520646e-01
5.09340987e-02 -1.05120909e+00 -3.97591740e-02 2.10091814e-01] | [8.401277542114258, 0.6018940210342407] |
b5464ef4-3250-4fb2-8f51-15a6dadffed3 | monocular-3d-object-detection-with-depth-from | 2207.12988 | null | https://arxiv.org/abs/2207.12988v2 | https://arxiv.org/pdf/2207.12988v2.pdf | Monocular 3D Object Detection with Depth from Motion | Perceiving 3D objects from monocular inputs is crucial for robotic systems, given its economy compared to multi-sensor settings. It is notably difficult as a single image can not provide any clues for predicting absolute depth values. Motivated by binocular methods for 3D object detection, we take advantage of the strong geometry structure provided by camera ego-motion for accurate object depth estimation and detection. We first make a theoretical analysis on this general two-view case and notice two challenges: 1) Cumulative errors from multiple estimations that make the direct prediction intractable; 2) Inherent dilemmas caused by static cameras and matching ambiguity. Accordingly, we establish the stereo correspondence with a geometry-aware cost volume as the alternative for depth estimation and further compensate it with monocular understanding to address the second problem. Our framework, named Depth from Motion (DfM), then uses the established geometry to lift 2D image features to the 3D space and detects 3D objects thereon. We also present a pose-free DfM to make it usable when the camera pose is unavailable. Our framework outperforms state-of-the-art methods by a large margin on the KITTI benchmark. Detailed quantitative and qualitative analyses also validate our theoretical conclusions. The code will be released at https://github.com/Tai-Wang/Depth-from-Motion. | ['Dahua Lin', 'Jiangmiao Pang', 'Tai Wang'] | 2022-07-26 | null | null | null | null | ['monocular-3d-object-detection'] | ['computer-vision'] | [ 1.13272488e-01 -9.53487083e-02 -9.18877050e-02 -2.24828377e-01
-6.67957842e-01 -7.62801468e-01 4.50886488e-01 -4.23003286e-01
-3.35052580e-01 3.65546316e-01 3.54850069e-02 -1.72543049e-01
4.09514271e-03 -5.23093402e-01 -8.56730700e-01 -7.01133192e-01
3.68224025e-01 3.03645372e-01 4.58772510e-01 8.62306803e-02
5.52083015e-01 6.88153863e-01 -1.74714744e+00 6.79621994e-02
5.96146882e-01 1.12638795e+00 7.66120911e-01 6.74085379e-01
2.47177973e-01 5.51614583e-01 -2.03943387e-01 -6.86388686e-02
8.00237358e-01 -1.47212157e-02 -4.40479994e-01 3.66327554e-01
6.59158170e-01 -1.02808321e+00 -4.41089362e-01 1.04462075e+00
4.92353559e-01 -1.18381359e-01 4.47727293e-01 -1.20549870e+00
-2.82956690e-01 -2.07488760e-01 -7.88695931e-01 -1.61792673e-02
8.29909801e-01 3.09471726e-01 7.83789694e-01 -1.29654860e+00
7.29562104e-01 1.28239655e+00 4.51736242e-01 5.10543406e-01
-8.15131724e-01 -2.18455359e-01 2.16621786e-01 1.47975996e-01
-1.39085877e+00 -6.28042221e-01 8.94793749e-01 -5.97245336e-01
8.82107854e-01 1.56618863e-01 5.67006946e-01 9.57383394e-01
8.77798796e-02 7.40114570e-01 1.00673461e+00 -4.72324640e-01
5.21394722e-02 2.13846475e-01 -1.89092696e-01 5.74162662e-01
3.85411978e-01 3.16751778e-01 -6.12207294e-01 3.10855836e-01
1.19621181e+00 1.88445181e-01 -4.43334252e-01 -9.44879949e-01
-1.39888477e+00 4.95226026e-01 4.38106358e-01 -1.18931137e-01
-1.84845820e-01 1.18303753e-01 -1.00120373e-01 1.36268109e-01
2.82429755e-01 3.56097847e-01 -3.75230730e-01 -1.31442264e-01
-2.99045801e-01 1.82729691e-01 4.05436993e-01 1.26669598e+00
8.93972933e-01 -3.35862041e-01 4.58667815e-01 4.26115811e-01
3.31790775e-01 6.06164038e-01 1.44088402e-01 -1.36221027e+00
6.63514435e-01 5.95395684e-01 4.51202124e-01 -1.05291951e+00
-4.20197666e-01 -2.13030934e-01 -5.12087882e-01 5.47602355e-01
7.52805948e-01 1.00527257e-01 -5.01553118e-01 1.38734400e+00
5.12376726e-01 -3.09377491e-01 -2.94926018e-02 1.36185348e+00
6.56810522e-01 2.39840493e-01 -8.04555953e-01 -3.14459354e-02
1.12653172e+00 -8.19234490e-01 -3.47835511e-01 -6.61653161e-01
5.34394860e-01 -6.54350996e-01 9.05605614e-01 5.05168140e-01
-1.03056133e+00 -3.88445318e-01 -1.09016609e+00 -3.65338504e-01
-1.32665575e-01 2.06144601e-01 7.13930607e-01 4.50009942e-01
-9.74326372e-01 2.04531640e-01 -8.00646186e-01 -4.85795140e-01
1.43480092e-01 4.14193451e-01 -5.42803109e-01 -2.93672889e-01
-6.15525842e-01 1.08860350e+00 2.50271827e-01 3.06839406e-01
-6.03620887e-01 -4.41795081e-01 -9.84786391e-01 -3.88993204e-01
6.93277478e-01 -9.37261522e-01 1.30426681e+00 -5.17226756e-01
-1.48240221e+00 1.02861941e+00 -4.05198336e-01 -1.39860939e-02
8.50432098e-01 -3.55736732e-01 3.70115846e-01 4.74065185e-01
-7.26604369e-03 7.54879236e-01 5.55651903e-01 -1.55820048e+00
-6.48460269e-01 -7.96925902e-01 5.46403408e-01 6.77075148e-01
-4.30475920e-03 -4.39242691e-01 -5.66653967e-01 -1.07884876e-01
8.54433775e-01 -9.19973791e-01 -1.66196346e-01 5.39024711e-01
-3.85638922e-01 1.72322035e-01 6.30688190e-01 -2.62527585e-01
6.37515962e-01 -1.93677211e+00 2.28247389e-01 -1.65095985e-01
3.97057623e-01 -1.83233514e-01 5.10175638e-02 2.93705761e-01
3.47003400e-01 -3.71869296e-01 2.24193588e-01 -5.31757236e-01
-2.51383722e-01 7.40420595e-02 -1.86808288e-01 8.00500035e-01
8.70656967e-02 7.94175088e-01 -9.12996650e-01 -3.07557791e-01
7.15558350e-01 3.37480187e-01 -6.29002571e-01 2.81386495e-01
3.89372557e-03 6.16871357e-01 -4.24965203e-01 1.00571191e+00
9.48099852e-01 -1.91233456e-01 5.56076653e-02 -4.07470375e-01
-4.04582471e-01 2.10903019e-01 -1.34523189e+00 1.90097558e+00
-3.64141524e-01 5.33881128e-01 6.59464896e-02 -7.37946689e-01
8.44295263e-01 -2.62935720e-02 4.21280980e-01 -5.74610054e-01
2.03545600e-01 4.62872714e-01 -2.36896187e-01 -5.18714726e-01
5.40784299e-01 8.29743817e-02 1.19435720e-01 -9.14464295e-02
-1.75577462e-01 -5.55677712e-01 -2.23088354e-01 1.41742909e-02
1.00129795e+00 4.01173294e-01 5.12699187e-01 -8.19925815e-02
3.36334825e-01 -9.23684016e-02 5.27846456e-01 5.98093033e-01
-2.72979856e-01 9.36305881e-01 2.15205625e-01 -3.12919348e-01
-1.06893921e+00 -1.15178990e+00 -1.63643524e-01 2.66667068e-01
7.97261059e-01 -1.06451884e-01 -3.70780051e-01 -3.83677453e-01
2.63945907e-01 1.63326621e-01 -4.42376345e-01 9.38399434e-02
-2.98734605e-01 -2.85115153e-01 -1.33814216e-01 6.41435146e-01
4.02215838e-01 -4.75668579e-01 -1.20277107e+00 -1.24855742e-01
-2.33454317e-01 -1.55076551e+00 -2.70365983e-01 8.70078430e-02
-8.56967449e-01 -1.12933290e+00 -6.74374521e-01 -4.66741413e-01
6.56076908e-01 1.10259044e+00 7.36002386e-01 -2.05890164e-01
-1.38943717e-01 6.42519355e-01 -3.05894315e-01 -2.94916570e-01
1.86468698e-02 -3.34496140e-01 4.07038718e-01 -2.13746354e-01
3.47852916e-01 -6.67024374e-01 -1.02561462e+00 6.07980669e-01
-5.68562329e-01 4.20957327e-01 6.83036566e-01 5.40414631e-01
5.11113942e-01 -3.57431293e-01 1.16886199e-01 -2.44293883e-01
-2.37219214e-01 -2.39411354e-01 -9.77840662e-01 -1.31087601e-01
-3.24904799e-01 -2.11534739e-01 2.18522251e-01 -3.44209462e-01
-8.45010400e-01 5.41305602e-01 2.98017144e-01 -7.40587294e-01
-5.02326451e-02 7.01224357e-02 -3.60909879e-01 -3.28209937e-01
5.60928643e-01 6.05213270e-02 1.65754035e-01 -3.55513960e-01
3.00122648e-01 7.03686833e-01 6.42220438e-01 -2.92431563e-01
8.01440120e-01 9.93084550e-01 2.68467069e-01 -8.34405243e-01
-9.56867158e-01 -8.21695447e-01 -1.05194283e+00 -4.46592450e-01
7.03211725e-01 -1.20679617e+00 -1.03436553e+00 6.00481093e-01
-1.48748398e+00 -7.77796656e-02 5.05885072e-02 7.29930282e-01
-7.71559596e-01 5.82340479e-01 -3.76764715e-01 -1.10550845e+00
7.46160522e-02 -1.22190499e+00 1.34995961e+00 8.12758729e-02
-1.49503155e-02 -7.15533197e-01 -2.60969311e-01 4.51620221e-01
-5.89363426e-02 2.31405005e-01 2.39935502e-01 1.67152211e-01
-1.24389243e+00 -1.98509023e-01 -3.87949377e-01 1.93548530e-01
7.83389062e-02 -2.78526783e-01 -1.32744873e+00 -2.45667487e-01
2.51731843e-01 -2.56429881e-01 6.66426122e-01 4.30096477e-01
9.18202877e-01 1.98254809e-02 -2.55717218e-01 8.22013438e-01
1.58825266e+00 4.97030467e-02 5.48983514e-01 6.02685630e-01
8.97944868e-01 7.85940707e-01 7.87498236e-01 5.85133374e-01
6.33273661e-01 1.07275546e+00 9.90842044e-01 1.93263650e-01
-5.17827123e-02 -3.01262200e-01 3.37902158e-01 5.94097853e-01
-1.82298645e-01 -2.01572552e-01 -9.03803647e-01 4.55085039e-01
-1.79623294e+00 -6.98130548e-01 -2.23955601e-01 2.43753242e+00
4.36655492e-01 1.34410352e-01 -3.25977765e-02 2.08907083e-01
4.50203419e-01 -9.32184905e-02 -6.83799386e-01 1.56104416e-01
-2.93185562e-01 -5.43708801e-01 7.62007058e-01 6.78617001e-01
-8.61932158e-01 6.98796272e-01 5.34098434e+00 2.97552735e-01
-9.51519847e-01 -1.45706594e-01 2.76180118e-01 -3.22262913e-01
-2.45153427e-01 1.23558618e-01 -1.02699172e+00 2.50828415e-01
1.76262662e-01 -5.62338810e-03 3.74558449e-01 8.77595544e-01
1.96797758e-01 -5.45475900e-01 -1.46285081e+00 1.49732482e+00
1.82860449e-01 -1.01178718e+00 -7.62358978e-02 3.08795244e-01
6.09967470e-01 1.95399985e-01 -1.67435855e-02 -2.76350468e-01
-2.54184067e-01 -6.31601930e-01 1.04031491e+00 3.89240056e-01
8.27605009e-01 -3.59739453e-01 6.16761565e-01 7.32463717e-01
-1.11659193e+00 -2.51389176e-01 -6.58032298e-01 -4.79282171e-01
1.85398132e-01 6.91131473e-01 -8.01563382e-01 6.15394473e-01
6.15820527e-01 8.47467959e-01 -4.61850166e-01 1.01836729e+00
-1.40671790e-01 -3.81396145e-01 -5.08021891e-01 1.89092681e-01
1.65977199e-02 -1.63103491e-01 7.18081415e-01 6.26535594e-01
5.62419355e-01 2.15181217e-01 2.13455618e-03 8.28253508e-01
2.16074407e-01 -3.36829424e-01 -8.56288791e-01 4.48596239e-01
5.71837842e-01 1.14751005e+00 -6.33120239e-01 2.03722771e-02
-5.96954584e-01 1.08290970e+00 2.12776765e-01 2.11352378e-01
-4.70764518e-01 -1.06302842e-01 6.87191665e-01 2.79061675e-01
2.53300130e-01 -5.22303820e-01 -4.76533443e-01 -1.52279699e+00
7.09955096e-01 -4.80413914e-01 -7.07531348e-02 -1.17661798e+00
-1.00470912e+00 4.00753289e-01 2.20458284e-02 -1.75129688e+00
-3.78937215e-01 -1.08445847e+00 -1.46132767e-01 7.35837817e-01
-1.69213963e+00 -9.70052898e-01 -7.31414974e-01 4.01684314e-01
5.80799162e-01 3.53625268e-01 5.82632005e-01 1.23198643e-01
-1.05110407e-01 3.59823883e-01 -5.17440587e-02 -2.72923827e-01
6.62373364e-01 -1.10057259e+00 2.26709351e-01 9.46420372e-01
-1.50787979e-01 4.62161541e-01 6.33959115e-01 -3.28357875e-01
-2.07013631e+00 -5.40081978e-01 6.16275549e-01 -1.19531155e+00
3.99885744e-01 -6.61043763e-01 -6.16851509e-01 5.37418485e-01
-4.09889579e-01 1.13003805e-01 -2.03037504e-02 -2.82573074e-01
-3.85715365e-01 -8.15464333e-02 -1.06396437e+00 4.75881249e-01
1.45049870e+00 -4.47479635e-01 -4.02396113e-01 1.99774429e-01
7.53721058e-01 -8.29581261e-01 -4.93320584e-01 4.65526968e-01
9.50099826e-01 -1.41662979e+00 1.14304006e+00 1.52737781e-01
5.05469680e-01 -4.61802930e-01 -6.27210259e-01 -9.69299316e-01
7.78711438e-02 -5.27088702e-01 -1.87649056e-01 7.70395041e-01
8.76265839e-02 -7.85160542e-01 8.96524847e-01 5.52832842e-01
-1.41635716e-01 -6.86692119e-01 -1.01781416e+00 -1.01969814e+00
-2.04905719e-01 -5.66348791e-01 2.56782919e-01 6.66889846e-01
3.80878616e-03 2.79466659e-01 -3.21526766e-01 4.72083271e-01
8.50813448e-01 3.32946807e-01 1.12520647e+00 -1.19463551e+00
-3.90356034e-01 -2.49528497e-01 -7.29387522e-01 -1.88656735e+00
-3.19564581e-01 -3.84164751e-01 1.97374880e-01 -1.27057600e+00
2.51522332e-01 -2.00651050e-01 1.98183835e-01 -6.00534566e-02
9.00597125e-02 2.82922387e-01 3.11340064e-01 4.88708436e-01
-4.71970797e-01 4.98202443e-01 1.45505881e+00 3.48728240e-01
-4.29515988e-02 3.41878533e-02 -5.42215466e-01 8.93217683e-01
6.25303805e-01 -1.11972503e-01 -3.92628431e-01 -7.41927326e-01
3.77968252e-01 3.45565379e-01 7.98327625e-01 -9.11956668e-01
2.35946015e-01 -1.41928986e-01 4.62839901e-01 -8.61296952e-01
8.96131873e-01 -9.65567827e-01 -1.27922088e-01 2.90435016e-01
1.16230331e-01 -1.21234961e-01 -4.12635412e-03 5.51346540e-01
-1.66017897e-02 -9.71050784e-02 6.42638266e-01 -2.36042738e-01
-8.82333815e-01 3.41546804e-01 -3.19876373e-02 -2.52884418e-01
9.54261005e-01 -8.02305877e-01 -4.11386639e-01 -5.95476449e-01
-3.35142970e-01 2.10940227e-01 1.06139827e+00 4.32990760e-01
8.74529183e-01 -1.25563800e+00 -3.94642264e-01 3.16815674e-01
4.78267699e-01 4.36367869e-01 2.35605776e-01 1.03480530e+00
-6.18984520e-01 6.55752838e-01 -3.08701415e-02 -1.03366661e+00
-1.21704888e+00 5.12302876e-01 2.92902082e-01 3.93943071e-01
-6.50792658e-01 6.93774998e-01 7.71265149e-01 -4.47586775e-01
2.69863218e-01 -5.46716213e-01 3.26664507e-01 -3.25019807e-01
4.99239206e-01 3.71451855e-01 -1.37355365e-02 -5.26766419e-01
-4.69884038e-01 1.12479711e+00 1.75071895e-01 -1.91143721e-01
1.18609750e+00 -7.14242339e-01 1.26179680e-01 5.69820881e-01
1.20975208e+00 -6.68126494e-02 -1.79947412e+00 -2.60430366e-01
-2.52438992e-01 -1.03584945e+00 -5.23090782e-03 -4.55698282e-01
-6.90071702e-01 1.07835269e+00 4.22528565e-01 -3.74769010e-02
1.07652187e+00 2.31197998e-01 4.39340532e-01 5.81953287e-01
6.86280251e-01 -7.88555205e-01 2.13167429e-01 5.59023201e-01
9.80708182e-01 -1.65280104e+00 2.58803256e-02 -7.73712099e-01
-3.78194392e-01 1.23755825e+00 8.08042407e-01 3.98403034e-02
2.76523441e-01 1.29961550e-01 7.25639462e-02 -1.01104945e-01
-5.36527514e-01 -1.72588989e-01 2.22139820e-01 6.62633598e-01
-8.72527622e-03 -2.13625029e-01 2.36405239e-01 2.07929716e-01
-6.92006201e-02 -1.10972263e-01 7.06042051e-01 9.41071689e-01
-5.70015192e-01 -7.84700036e-01 -4.27251250e-01 1.09335245e-03
-4.18074988e-03 1.34773448e-01 -3.32088739e-01 8.15953314e-01
-5.87551780e-02 9.63419437e-01 5.39505035e-02 -4.76763785e-01
4.30075318e-01 -4.09611285e-01 1.01443541e+00 -6.25232339e-01
2.28412554e-01 6.37090653e-02 -1.42567620e-01 -9.40795958e-01
-5.56573391e-01 -7.45081961e-01 -1.00026572e+00 -2.68050998e-01
-4.59998488e-01 -4.74409610e-01 9.11180317e-01 7.33963788e-01
3.25800568e-01 -6.95140883e-02 8.14180911e-01 -1.54309595e+00
-7.69631922e-01 -7.40025878e-01 -4.79176313e-01 1.44027144e-01
6.84684157e-01 -9.17130053e-01 -7.83549488e-01 -1.93535015e-01] | [8.227338790893555, -2.501697301864624] |
87fa3d8d-7b97-41ce-80ea-789e5e1ae48c | protein-secondary-structure-prediction-using-1 | 1611.01503 | null | http://arxiv.org/abs/1611.01503v1 | http://arxiv.org/pdf/1611.01503v1.pdf | Protein Secondary Structure Prediction Using Deep Multi-scale Convolutional Neural Networks and Next-Step Conditioning | Recently developed deep learning techniques have significantly improved the
accuracy of various speech and image recognition systems. In this paper we
adapt some of these techniques for protein secondary structure prediction. We
first train a series of deep neural networks to predict eight-class secondary
structure labels given a protein's amino acid sequence information and find
that using recent methods for regularization, such as dropout and weight-norm
constraining, leads to measurable gains in accuracy. We then adapt recent
convolutional neural network architectures--Inception, ReSNet, and DenseNet
with Batch Normalization--to the problem of protein structure prediction. These
convolutional architectures make heavy use of multi-scale filter layers that
simultaneously compute features on several scales, and use residual connections
to prevent underfitting. Using a carefully modified version of these
architectures, we achieve state-of-the-art performance of 70.0% per amino acid
accuracy on the public CB513 benchmark dataset. Finally, we explore additions
from sequence-to-sequence learning, altering the model to make its predictions
conditioned on both the protein's amino acid sequence and its past secondary
structure labels. We introduce a new method of ensembling such a conditional
model with our convolutional model, an approach which reaches 70.6% Q8 accuracy
on CB513. We argue that these results can be further refined for larger boosts
in prediction accuracy through more sophisticated attempts to control
overfitting of conditional models. We aim to release the code for these
experiments as part of the TensorFlow repository. | ['Akosua Busia', 'Navdeep Jaitly', 'Jasmine Collins'] | 2016-11-04 | null | null | null | null | ['protein-secondary-structure-prediction'] | ['medical'] | [ 6.61387682e-01 2.35930234e-01 2.13315114e-02 -8.64762306e-01
-8.57495487e-01 -4.22398150e-01 3.93940389e-01 1.12900734e-01
-8.37509990e-01 9.71540570e-01 1.35690182e-01 -8.13591242e-01
2.98533291e-01 -3.51368457e-01 -9.59570587e-01 -7.70804167e-01
-9.68505964e-02 3.24275970e-01 1.55532330e-01 -2.81283468e-01
1.28991798e-01 6.34807706e-01 -1.13655710e+00 6.92074299e-01
5.53848088e-01 8.07907164e-01 2.08624095e-01 1.12969255e+00
6.22996651e-02 1.07631218e+00 -5.57379723e-01 -3.66015345e-01
1.23153418e-01 -3.10004950e-01 -1.26468349e+00 -8.47632959e-02
3.72347564e-01 -2.70760000e-01 -2.73498476e-01 7.43472099e-01
5.42770624e-01 2.19870964e-03 4.13685143e-01 -6.44318998e-01
-7.66042411e-01 3.27292293e-01 -2.87330538e-01 2.86455184e-01
1.45980179e-01 4.57781374e-01 1.13204038e+00 -7.42204547e-01
7.29827702e-01 1.04682159e+00 1.00928855e+00 6.59023821e-01
-1.87909400e+00 -5.14380038e-01 3.28533649e-02 2.10019663e-01
-1.03997374e+00 -5.60357153e-01 2.60758400e-01 -3.84119064e-01
1.83693147e+00 7.36774430e-02 3.58910233e-01 1.28488886e+00
4.16905463e-01 5.85842311e-01 1.01332903e+00 -3.65949005e-01
3.43671031e-02 -3.36444348e-01 2.01383829e-01 7.99973667e-01
-2.68422574e-01 6.47315234e-02 -3.29663217e-01 -5.35175204e-01
3.79705071e-01 -1.52127489e-01 -2.19708115e-01 -2.29053289e-01
-1.19163454e+00 9.57806468e-01 4.37270314e-01 1.19712323e-01
-3.67502123e-01 1.03916511e-01 3.94676477e-01 5.12666345e-01
6.02737069e-01 6.66723490e-01 -1.24105990e+00 -1.63472623e-01
-8.92983258e-01 2.40546241e-01 9.50414717e-01 5.74659884e-01
8.07776451e-01 -1.79310232e-01 -3.50784734e-02 9.23820019e-01
1.60736546e-01 3.53443262e-04 5.82763493e-01 -8.01307678e-01
3.72942351e-02 2.07052305e-01 3.98835093e-02 -3.87644947e-01
-7.09074616e-01 -3.57883245e-01 -5.86534739e-01 2.92150408e-01
3.69870365e-01 -3.08928281e-01 -1.37671888e+00 1.80133033e+00
3.52484137e-02 1.26438081e-01 2.71810740e-01 8.03692162e-01
6.14374936e-01 6.61542952e-01 4.01807666e-01 -7.85626024e-02
1.19447589e+00 -8.18342090e-01 -2.72373497e-01 -5.34238704e-02
9.45505559e-01 -9.47580636e-01 9.72608626e-01 4.10424232e-01
-9.48885620e-01 -4.89830613e-01 -1.34901989e+00 -3.35604846e-01
-3.30911040e-01 1.59633979e-01 6.65046692e-01 4.38397080e-01
-1.18040216e+00 1.07971597e+00 -1.15229273e+00 -4.46706474e-01
3.58633846e-01 7.04586148e-01 -4.02438939e-01 2.31526673e-01
-1.23395848e+00 1.08167958e+00 4.94121969e-01 -1.55576453e-01
-7.58834243e-01 -6.13321126e-01 -5.43276966e-01 1.43633872e-01
-8.66411403e-02 -5.48769951e-01 1.64670384e+00 -1.36512351e+00
-1.58782923e+00 1.04156411e+00 -2.58641303e-01 -8.79679561e-01
7.30645508e-02 -9.28774476e-02 -2.82127678e-01 1.12492763e-01
-1.62224203e-01 8.86820734e-01 5.38049340e-01 -5.88779747e-01
-5.60208321e-01 -7.93286040e-02 4.88343425e-02 -8.06726068e-02
-6.43297508e-02 2.07216024e-01 -1.17480233e-01 -4.71219391e-01
-4.98616658e-02 -9.85169649e-01 -4.86027926e-01 -1.07349284e-01
-2.89569885e-01 -2.50032246e-01 4.34052825e-01 -6.93921506e-01
6.69025719e-01 -1.89541209e+00 1.02329463e-01 -6.37022872e-03
2.51370877e-01 6.62797570e-01 -4.75093395e-01 3.32164347e-01
-6.16172910e-01 6.22715726e-02 -3.46226841e-01 -2.02359959e-01
-2.36337095e-01 3.30191553e-01 -9.10297185e-02 4.98191953e-01
7.68146396e-01 9.83177543e-01 -4.76411015e-01 1.24379903e-01
-5.35788797e-02 7.97497571e-01 -7.86144078e-01 2.33310789e-01
-5.19528270e-01 3.13265443e-01 -1.56435117e-01 2.60853857e-01
6.70340836e-01 -7.01489627e-01 5.38361371e-01 -2.02715725e-01
4.02633213e-02 7.63836861e-01 -4.29309636e-01 1.79739308e+00
-1.58250779e-01 5.47887683e-01 1.04428321e-01 -1.24225771e+00
8.20380330e-01 3.02203923e-01 7.19850838e-01 -4.47608978e-01
-3.41661163e-02 9.41232666e-02 4.52343702e-01 -3.72458607e-01
2.70691186e-01 -3.58664542e-01 4.59619403e-01 1.16941407e-01
5.16343892e-01 3.21390390e-01 8.33172575e-02 2.09060907e-01
1.50548637e+00 4.64942694e-01 1.86291218e-01 -4.95273471e-01
4.67869759e-01 2.47316901e-02 6.30787373e-01 5.89434505e-01
-2.84947097e-01 6.11451566e-01 6.94790840e-01 -8.30501139e-01
-1.46974993e+00 -8.26003730e-01 -2.65979230e-01 1.52179539e+00
-5.88368535e-01 -5.66579521e-01 -7.86715865e-01 -8.31789553e-01
3.08576692e-02 3.67843419e-01 -5.53871155e-01 -2.89790779e-01
-6.80818260e-01 -1.45750844e+00 7.45524168e-01 5.66158533e-01
1.98362339e-02 -1.23940992e+00 -3.14776301e-01 5.70543110e-01
-1.55725507e-02 -8.30953121e-01 -1.60302296e-01 1.17794442e+00
-6.93233073e-01 -1.19120181e+00 -6.43922806e-01 -9.54334140e-01
3.57377172e-01 -3.45635891e-01 1.34619319e+00 3.17563027e-01
-2.85904318e-01 -1.52194336e-01 -2.84318954e-01 -1.31226510e-01
-7.41390526e-01 4.85191017e-01 -3.05204578e-02 -2.10003674e-01
8.48691046e-01 -6.48837268e-01 -6.98508501e-01 7.15709180e-02
-8.99113357e-01 1.67661756e-01 7.08350897e-01 1.15044141e+00
5.65726221e-01 -6.31138384e-01 6.11680031e-01 -1.15922999e+00
5.15809178e-01 -4.15471435e-01 -5.04358947e-01 1.14786953e-01
-6.73690975e-01 5.98980188e-01 7.31218994e-01 -3.41187298e-01
-9.25803363e-01 6.25801563e-01 -9.60471928e-01 -9.64295790e-02
-3.11430067e-01 5.59117436e-01 1.82956487e-01 -1.73183247e-01
9.29300606e-01 2.41358310e-01 2.16241419e-01 -5.85408032e-01
3.30714941e-01 6.78539515e-01 4.59680706e-01 -5.61058760e-01
1.86702639e-01 8.57781917e-02 -7.37154111e-02 -6.63616478e-01
-9.99972641e-01 -4.36099231e-01 -6.37568355e-01 5.10667145e-01
1.02402496e+00 -8.98138940e-01 -9.93612409e-01 4.61971968e-01
-1.13888514e+00 -5.83677769e-01 8.53765905e-02 3.40372175e-01
-7.13285744e-01 4.65483516e-01 -1.11894512e+00 -2.30164230e-01
-4.05458003e-01 -1.24569750e+00 9.83984172e-01 -1.14309117e-01
-1.88694239e-01 -7.20554292e-01 2.63519287e-01 3.12260687e-01
5.21294355e-01 -8.56995583e-02 9.96696174e-01 -1.12794495e+00
-3.14014763e-01 1.83068365e-01 -2.25731954e-01 5.66213489e-01
4.70103882e-02 -2.21788082e-02 -9.99947667e-01 -5.86262047e-01
-7.18881115e-02 -7.29275644e-01 1.27597392e+00 3.35879445e-01
1.17575490e+00 -1.04313783e-01 -2.06046253e-01 9.44131255e-01
1.26635456e+00 1.95054546e-01 6.59868836e-01 4.29675639e-01
5.49831629e-01 2.24588916e-01 2.31811121e-01 2.21291274e-01
-2.99823545e-02 5.07822871e-01 3.60587209e-01 -2.94328749e-01
-2.84235682e-02 7.68270418e-02 2.22754821e-01 7.18975186e-01
-7.32678687e-03 -1.79215446e-01 -9.22387838e-01 2.43304819e-01
-1.68470299e+00 -8.33149791e-01 -1.26145650e-02 1.78984475e+00
1.26323760e+00 2.60349244e-01 7.51133487e-02 -2.96376109e-01
4.30529118e-01 3.76906898e-03 -8.66097152e-01 -6.13504410e-01
-1.65273413e-01 7.21315920e-01 7.79839993e-01 5.72641492e-01
-1.34237027e+00 1.05192435e+00 7.23394775e+00 5.92123449e-01
-1.24969649e+00 -9.35121328e-02 1.03438652e+00 -8.59139860e-02
5.38760908e-02 -5.48052602e-02 -8.82314265e-01 2.92768419e-01
1.63542128e+00 3.87936115e-01 4.94144976e-01 7.36117065e-01
1.19329685e-04 2.05307692e-01 -1.14548147e+00 5.42317212e-01
-1.59197375e-01 -1.61981714e+00 -2.00726256e-01 -2.31375638e-03
5.14809310e-01 7.75557518e-01 -9.20925587e-02 3.31799895e-01
7.35068679e-01 -1.23098361e+00 1.60737127e-01 3.69104922e-01
7.85815656e-01 -8.59659791e-01 6.59166694e-01 1.74951717e-01
-8.02807987e-01 1.87196001e-01 -7.41601765e-01 -1.63892493e-01
-3.10172848e-02 4.96824056e-01 -1.10833287e+00 1.57720968e-01
6.93073988e-01 6.86626792e-01 -5.33742785e-01 5.61793566e-01
-6.62188828e-02 9.08637464e-01 -1.56583935e-01 -1.10623866e-01
3.75376850e-01 -4.19199131e-02 9.06315446e-02 1.47783446e+00
-2.68067628e-01 1.43728212e-01 1.09403439e-01 6.45447314e-01
-2.92595834e-01 1.36702225e-01 -1.52221933e-01 -1.92915127e-02
-1.68358907e-02 1.16878712e+00 -4.78670686e-01 -5.60170591e-01
-7.01667666e-01 1.03694427e+00 9.14387167e-01 4.33879018e-01
-7.84144938e-01 -5.79807341e-01 1.13514280e+00 -9.13597196e-02
5.91543555e-01 -1.96744233e-01 -1.32905273e-02 -1.21774220e+00
-2.59105742e-01 -1.20771539e+00 1.55114278e-01 -7.00281799e-01
-1.40628242e+00 7.29257286e-01 -6.16066992e-01 -4.15863752e-01
-2.46324733e-01 -1.09850347e+00 -1.28686517e-01 1.07166648e+00
-1.40055335e+00 -8.97177637e-01 4.46076244e-01 2.88022995e-01
3.43316525e-01 -2.56921470e-01 1.10607207e+00 2.37540305e-01
-3.32103401e-01 5.26688814e-01 4.78183568e-01 1.61004767e-01
8.39175045e-01 -1.40164840e+00 1.11791527e+00 6.44183874e-01
1.66062683e-01 7.92882860e-01 7.09290743e-01 -4.68744516e-01
-1.04749727e+00 -1.14489770e+00 1.05381024e+00 -5.80550790e-01
6.84851348e-01 -5.68442941e-01 -1.23822916e+00 9.97940421e-01
2.52864748e-01 4.27729525e-02 8.21282804e-01 3.48406613e-01
-4.36496466e-01 2.55410790e-01 -9.83025312e-01 3.65881562e-01
1.03550744e+00 -6.85559571e-01 -6.74444735e-01 5.78755021e-01
9.56319332e-01 -2.73447692e-01 -1.00522435e+00 4.99524534e-01
5.48470914e-01 -9.65477347e-01 9.80068088e-01 -1.36231256e+00
4.31015372e-01 -6.22483119e-02 -1.78491652e-01 -1.23092008e+00
-6.90552533e-01 -5.39454281e-01 2.22613320e-01 6.57682300e-01
7.62836635e-01 -6.46185398e-01 1.12208247e+00 5.04680395e-01
-3.54254633e-01 -1.03892648e+00 -7.62751400e-01 -4.82987463e-01
5.47774434e-01 -1.35024562e-01 3.00975472e-01 8.30056429e-01
-1.37128131e-02 6.47253811e-01 -3.81080121e-01 -8.81564096e-02
1.07641369e-01 -1.29419804e-01 4.83536035e-01 -1.02920806e+00
-6.00466192e-01 -1.62733361e-01 -4.06525582e-01 -1.30597472e+00
4.59516853e-01 -1.09583855e+00 1.38462365e-01 -1.00531590e+00
3.98869604e-01 -4.89942264e-03 -7.10437536e-01 7.92253017e-01
-2.27486521e-01 4.13601786e-01 -8.71852636e-02 5.41163981e-02
-6.36380732e-01 3.80670935e-01 7.98752129e-01 -1.73709750e-01
-1.68940201e-02 -1.42463386e-01 -4.89504129e-01 5.20140469e-01
1.10750592e+00 -4.85036254e-01 -1.25586659e-01 -2.29742721e-01
4.71848287e-02 -1.57984763e-01 1.98172584e-01 -8.98417234e-01
-7.00112730e-02 -8.08206853e-03 7.35774636e-01 -3.61470133e-01
4.09715921e-01 -4.81351137e-01 -4.41507660e-02 7.15317249e-01
-8.04490566e-01 1.54045105e-01 2.88326859e-01 4.68796641e-01
-1.70652699e-02 -3.46965790e-02 9.92175817e-01 -1.90091580e-01
-4.17722791e-01 2.78384477e-01 -8.63789380e-01 -1.25805676e-01
5.81130743e-01 2.79429376e-01 -2.87516683e-01 -1.02762073e-01
-1.21319687e+00 -1.18561551e-01 5.16801775e-01 1.79221317e-01
3.00576001e-01 -8.80019069e-01 -7.73888171e-01 4.45813864e-01
3.74029465e-02 -5.91095567e-01 -1.31719172e-01 6.03534400e-01
-8.30895007e-01 6.72883391e-01 -3.00499052e-01 -6.29568160e-01
-1.40555656e+00 6.22320414e-01 6.50903046e-01 -2.82542676e-01
-5.78382790e-01 1.06861675e+00 1.79333761e-01 -8.12577963e-01
7.73616061e-02 -4.15304601e-01 9.23880488e-02 -4.58653331e-01
3.85593921e-01 -2.34036267e-01 4.77222979e-01 -5.98639488e-01
-3.91207457e-01 7.07575157e-02 -5.84366322e-01 2.29766697e-01
1.51850843e+00 2.59693056e-01 -1.09912172e-01 1.08319029e-01
1.52206123e+00 -3.07738066e-01 -1.61960864e+00 -1.86420277e-01
3.57951283e-01 1.09817833e-01 -1.62195131e-01 -1.24233794e+00
-7.63998985e-01 7.24471927e-01 7.45565832e-01 -6.79590926e-02
9.27140653e-01 -1.08874984e-01 7.18806565e-01 7.86987066e-01
-6.52567744e-02 -8.28250945e-01 -1.72250748e-01 9.06526208e-01
4.33404505e-01 -1.37004375e+00 -6.53900951e-02 3.36915031e-02
-3.31950545e-01 1.17115152e+00 4.94779050e-01 -2.74138451e-01
5.61389685e-01 4.30566072e-01 3.26429605e-01 -1.76611602e-01
-1.30166698e+00 -4.27789986e-02 8.34074840e-02 5.49488664e-01
1.23890114e+00 8.76874253e-02 -3.07846785e-01 3.70425552e-01
-1.09458692e-01 -3.09157604e-03 2.98599482e-01 9.66757119e-01
-5.27615547e-01 -1.43786728e+00 3.75167690e-02 6.49557710e-01
-9.75088716e-01 -5.69473028e-01 -5.49642503e-01 3.86318892e-01
-7.76093379e-02 8.15016687e-01 -4.57114987e-02 -3.11637193e-01
1.31364211e-01 5.31598568e-01 3.86713028e-01 -6.44153357e-01
-7.84968495e-01 7.82766566e-03 2.86161959e-01 -7.47562647e-01
-5.76465309e-01 -4.93430912e-01 -1.39104271e+00 -4.27831143e-01
-1.59676686e-01 1.30533129e-01 6.23473406e-01 9.34573531e-01
6.84996665e-01 6.14109099e-01 1.11696832e-01 -8.16258430e-01
-9.30111110e-01 -1.18171787e+00 -2.58153528e-01 5.33686161e-01
4.94571507e-01 -2.21083328e-01 -9.70757306e-02 2.60690331e-01] | [4.726266384124756, 5.642025947570801] |
8904091f-11eb-44a5-855a-856535a3957a | benchmarking-of-lightweight-deep-learning | 2110.1227 | null | https://arxiv.org/abs/2110.12270v1 | https://arxiv.org/pdf/2110.12270v1.pdf | Benchmarking of Lightweight Deep Learning Architectures for Skin Cancer Classification using ISIC 2017 Dataset | Skin cancer is one of the deadly types of cancer and is common in the world. Recently, there has been a huge jump in the rate of people getting skin cancer. For this reason, the number of studies on skin cancer classification with deep learning are increasing day by day. For the growth of work in this area, the International Skin Imaging Collaboration (ISIC) organization was established and they created an open dataset archive. In this study, images were taken from ISIC 2017 Challenge. The skin cancer images taken were preprocessed and data augmented. Later, these images were trained with transfer learning and fine-tuning approach and deep learning models were created in this way. 3 different mobile deep learning models and 3 different batch size values were determined for each, and a total of 9 models were created. Among these models, the NASNetMobile model with 16 batch size got the best result. The accuracy value of this model is 82.00%, the precision value is 81.77% and the F1 score value is 0.8038. Our method is to benchmark mobile deep learning models which have few parameters and compare the results of the models. | ['Huseyin Uvet', 'Mehmet Erhan Guvenilir', 'Yegor Samoylenko', 'Mucahit Kalebasi', 'Abdurrahim Yilmaz'] | 2021-10-23 | null | null | null | null | ['skin-cancer-classification'] | ['medical'] | [-8.57034847e-02 -3.43618751e-01 -1.92219764e-01 -4.79832590e-02
-7.61634290e-01 -3.05409223e-01 4.49458390e-01 7.14868456e-02
-7.64280438e-01 8.50733995e-01 1.33935332e-01 -1.23715661e-01
7.91039690e-02 -7.97430098e-01 -2.10074455e-01 -8.93695712e-01
9.47824419e-02 -1.40679345e-01 4.54555601e-01 -1.07543282e-01
1.95828632e-01 1.05673008e-01 -9.45360005e-01 4.92204875e-01
9.70272005e-01 9.68831182e-01 -2.36556325e-02 1.08042216e+00
-1.90025419e-01 7.07707405e-01 -4.81092125e-01 -5.98877013e-01
1.74534425e-01 -1.97262540e-01 -9.18127656e-01 -2.42023647e-01
3.06343526e-01 -1.97605804e-01 -2.86086857e-01 9.30840731e-01
8.43912899e-01 -2.74064600e-01 5.91670752e-01 -1.08957112e+00
-5.15878499e-01 5.63432500e-02 -8.36202264e-01 1.94562986e-01
-1.22552887e-02 -1.68282330e-01 3.05414081e-01 -4.98399943e-01
5.29157698e-01 7.38580644e-01 9.87700284e-01 9.80983555e-01
-2.19392836e-01 -6.07200265e-01 -2.05263421e-01 4.57119465e-01
-1.37718058e+00 -1.07641466e-01 1.73268154e-01 -2.65019327e-01
6.91051126e-01 4.42913383e-01 9.29566562e-01 1.26579058e+00
5.53759217e-01 5.65876663e-01 1.41573668e+00 -5.41695118e-01
1.54230341e-01 3.21189880e-01 -3.95779759e-02 7.27916718e-01
2.91273952e-01 -2.95446932e-01 -2.07178771e-01 4.59213257e-02
6.94261193e-01 1.97070017e-01 -5.96360639e-02 2.12040097e-01
-4.93981391e-01 7.12513626e-01 5.74346244e-01 4.89975363e-01
-3.69797200e-01 2.13127166e-01 6.99976385e-01 3.08160126e-01
5.59581518e-01 -2.35108614e-01 -4.31490719e-01 -1.08541414e-01
-3.63034993e-01 7.63354227e-02 6.15700066e-01 3.98203462e-01
8.65890831e-02 -4.11567301e-01 3.26360017e-02 1.17480004e+00
3.78302544e-01 2.15430155e-01 9.79461312e-01 -2.80185431e-01
1.83537081e-01 8.06693792e-01 -3.23780477e-01 -7.38231719e-01
-4.33406740e-01 -2.91885406e-01 -1.11928785e+00 7.03832433e-02
5.01464486e-01 -3.80847335e-01 -1.42120194e+00 1.22081947e+00
4.38775688e-01 4.51145500e-01 2.44191498e-01 7.21483707e-01
1.05055988e+00 4.43026066e-01 5.45954823e-01 2.02118978e-01
1.29648244e+00 -1.21940267e+00 -5.19725084e-01 1.88437283e-01
8.39578271e-01 -9.55952048e-01 8.87749732e-01 5.88181794e-01
-6.48519933e-01 -2.72253692e-01 -9.78095949e-01 8.23083147e-02
-7.71338999e-01 -2.97693769e-03 9.40347433e-01 1.09354186e+00
-1.21141315e+00 2.83042997e-01 -7.38660872e-01 -1.16657233e+00
6.12872243e-01 3.07172894e-01 -5.01573741e-01 -2.43055761e-01
-1.21217072e+00 8.64943683e-01 2.55120695e-01 -3.63480709e-02
-9.41004872e-01 -3.72455359e-01 -4.14677570e-03 -5.24824917e-01
4.74995673e-02 -5.26992798e-01 1.09885097e+00 -1.30630529e+00
-1.26049805e+00 1.07706261e+00 3.83732587e-01 -4.14772063e-01
6.45947933e-01 -1.83053553e-01 -7.19937861e-01 3.93203227e-03
-1.68587565e-01 6.58447802e-01 4.06204283e-01 -9.04882193e-01
-8.74495089e-01 -4.76778179e-01 2.28988886e-01 4.48429674e-01
-7.52098382e-01 8.54119062e-02 -7.79640138e-01 -2.31030941e-01
-4.04060870e-01 -1.07524705e+00 -3.59074235e-01 1.30707234e-01
-4.00155574e-01 -9.30307284e-02 6.65384471e-01 -9.00381327e-01
1.25748348e+00 -1.97886753e+00 -2.16655910e-01 2.82525003e-01
1.45019278e-01 6.87280834e-01 -1.66674644e-01 4.16968226e-01
-1.03584759e-01 4.73609120e-01 1.58843771e-02 2.72526834e-02
-6.19700849e-01 5.46296276e-02 6.56968653e-01 1.39841557e-01
-1.74428120e-01 5.95177889e-01 -6.27101481e-01 -6.70768261e-01
2.03802865e-02 7.55065382e-01 -2.75533676e-01 6.59586862e-02
2.04137981e-01 1.19316086e-01 -3.34862679e-01 9.88720357e-01
8.29971433e-01 -1.83849186e-01 1.18909612e-01 -1.15753643e-01
2.23040745e-01 -5.87389052e-01 -8.87411892e-01 1.46241248e+00
-3.67712677e-01 4.44754660e-01 4.62202029e-03 -5.47521412e-01
5.93201280e-01 3.83958787e-01 6.11950576e-01 -7.31505394e-01
2.40700960e-01 1.03336029e-01 -1.49785867e-02 -1.07446671e+00
2.37550497e-01 -8.94639865e-02 3.10879618e-01 -2.92000342e-02
-1.32042721e-01 2.36037523e-01 2.85815634e-02 1.75069243e-01
1.25881922e+00 -2.39656165e-01 3.04540187e-01 5.21415146e-03
5.88926017e-01 8.05774182e-02 2.16361046e-01 1.96473330e-01
-5.39286673e-01 5.29472888e-01 3.51005912e-01 -5.92318833e-01
-7.45948970e-01 -8.65377307e-01 -1.96034521e-01 8.89928579e-01
-5.22520319e-02 -3.24801236e-01 -1.30038309e+00 -9.05671239e-01
-2.66731024e-01 -1.05046026e-01 -9.56362426e-01 4.00243662e-02
4.24148031e-02 -1.13320649e+00 8.35530221e-01 5.41190624e-01
1.22980094e+00 -8.64866853e-01 -1.87122777e-01 -2.48286664e-01
-1.11432329e-01 -7.07761049e-01 -1.93248004e-01 -4.80671711e-02
-7.73482800e-01 -1.67415702e+00 -1.12526011e+00 -9.73399997e-01
7.12587357e-01 2.39641055e-01 5.64131796e-01 4.91006076e-01
-7.20806658e-01 1.41034573e-01 -5.39986193e-01 -7.67623603e-01
-2.01761007e-01 3.01553249e-01 -1.84666321e-01 5.87233678e-02
6.18871093e-01 1.50926948e-01 -7.22381473e-01 -3.45983058e-02
-1.07135463e+00 2.48636976e-02 8.00064981e-01 8.51186395e-01
5.09098411e-01 2.21357644e-01 6.89499736e-01 -1.10050714e+00
8.28767180e-01 -7.18052864e-01 -3.29275429e-02 3.04866612e-01
-5.18159568e-01 -5.82313597e-01 2.41061613e-01 -3.32750142e-01
-1.10957134e+00 6.30447567e-02 -4.00694579e-01 1.15652315e-01
-2.50105888e-01 5.76109946e-01 1.90297231e-01 -2.11668089e-01
8.46727312e-01 -1.94730341e-01 1.82315394e-01 -3.18090737e-01
-9.41556618e-02 1.15131056e+00 1.37173936e-01 8.38550460e-03
3.02526236e-01 2.48420149e-01 -2.73233414e-01 -1.03676391e+00
-6.71128273e-01 -4.45712119e-01 -3.06658775e-01 -6.44799113e-01
1.14063025e+00 -9.63717759e-01 -1.30242273e-01 1.42903888e+00
-7.18176425e-01 -5.25123596e-01 4.06924307e-01 3.52615923e-01
1.60311088e-01 2.69693255e-01 -8.38000774e-01 -7.75617361e-01
-6.87405169e-01 -8.94680619e-01 5.99302948e-01 7.75838971e-01
-8.87492523e-02 -1.33196354e+00 1.68023601e-01 5.81129611e-01
8.02030921e-01 5.41867256e-01 6.29387140e-01 -4.63119626e-01
-5.34902476e-02 -6.26251101e-01 -3.87715727e-01 5.19271851e-01
3.84501070e-01 2.87484884e-01 -1.17753291e+00 -2.15435565e-01
-3.34559053e-01 -3.56703430e-01 9.20008779e-01 3.92447174e-01
1.46037555e+00 -4.20888104e-02 -5.80127180e-01 4.08328682e-01
1.80131435e+00 4.16995913e-01 1.09562552e+00 5.58287024e-01
6.21178627e-01 2.92413503e-01 3.27440411e-01 -1.54440412e-02
3.77823740e-01 1.47174880e-01 6.06044352e-01 -3.87189239e-01
-2.31295168e-01 -5.15886284e-02 9.30661708e-02 7.12473512e-01
-5.18883765e-01 -1.44640341e-01 -9.85019445e-01 4.48624164e-01
-1.37266028e+00 -7.10492551e-01 -4.35543507e-01 2.04344916e+00
8.29875767e-01 4.51541692e-02 1.76301211e-01 1.58677176e-01
5.97204268e-01 -1.27584934e-01 -4.62569416e-01 -4.04807150e-01
8.96944776e-02 3.86826456e-01 6.35480165e-01 3.51829231e-01
-1.30959606e+00 9.03875291e-01 6.22806978e+00 8.58673573e-01
-1.54952395e+00 1.35761753e-01 1.03913629e+00 2.13702396e-02
1.98878422e-01 -4.65364605e-01 -6.27402365e-01 5.38922429e-01
9.88968492e-01 5.46481647e-02 7.36277774e-02 8.32349002e-01
-1.84863314e-01 -5.71339309e-01 -3.92640948e-01 9.60827947e-01
2.29766563e-01 -1.17762959e+00 -1.41696081e-01 2.18390003e-01
8.15295637e-01 1.28887951e-01 3.48600894e-02 3.36845160e-01
6.88032731e-02 -1.25153589e+00 -2.66826481e-01 8.43387544e-01
7.90638864e-01 -7.52759576e-01 1.32363296e+00 2.21401334e-01
-8.62138212e-01 -1.62376180e-01 -1.95598051e-01 1.15426406e-01
-3.83648545e-01 4.71336782e-01 -7.45474219e-01 4.68342781e-01
1.06878829e+00 5.12500405e-01 -9.62722600e-01 1.45475388e+00
2.43904278e-01 7.01759100e-01 -2.91740447e-01 -6.30076468e-01
1.78796634e-01 5.33025935e-02 -2.06970617e-01 1.22463024e+00
2.91147053e-01 -1.37768567e-01 -2.62516230e-01 -2.31722236e-01
-9.13759097e-02 3.06171268e-01 -4.96982634e-01 7.76752904e-02
2.78323799e-01 1.69136274e+00 -7.07908511e-01 -1.36426970e-01
-4.38192636e-01 1.00383103e+00 -3.99563760e-02 4.74550501e-02
-1.08385968e+00 -4.28633273e-01 3.08216304e-01 1.29018113e-01
-2.75817305e-01 1.98423266e-01 -2.11888790e-01 -8.25622320e-01
-2.26469338e-01 -8.70665491e-01 6.01647615e-01 -6.89102113e-01
-1.39646327e+00 5.12662590e-01 -2.17853263e-01 -8.79340053e-01
3.85609567e-01 -9.25051033e-01 -9.75933731e-01 7.56530702e-01
-1.45778477e+00 -1.21703935e+00 -1.04407501e+00 7.18071163e-01
5.38905680e-01 -3.86520863e-01 1.29321575e+00 5.54004133e-01
-7.89329410e-01 8.69798660e-01 8.30835328e-02 2.74461836e-01
9.31652486e-01 -1.15470266e+00 -5.67076951e-02 4.62081522e-01
-4.03353155e-01 4.54119086e-01 -1.02582149e-01 -6.14655197e-01
-1.15757573e+00 -1.24813545e+00 5.05802214e-01 -3.92125845e-01
4.86312062e-01 -4.67746742e-02 -6.45789027e-01 2.43644357e-01
5.72382390e-01 -1.11587234e-01 1.18018234e+00 -5.63395768e-03
-1.22728467e-01 -3.04273546e-01 -1.51703453e+00 5.65025508e-01
4.84809250e-01 -1.66102603e-01 2.20463946e-01 4.95226711e-01
2.66695052e-01 -6.09240592e-01 -8.97161305e-01 6.57636374e-02
7.71636188e-01 -9.61505890e-01 6.76703036e-01 -5.92013061e-01
5.72851717e-01 4.42305095e-02 -2.13271007e-01 -1.28723371e+00
-7.27712288e-02 1.29716679e-01 4.55508456e-02 1.00092936e+00
4.85974848e-01 -6.00822210e-01 1.20736039e+00 4.95848417e-01
7.88515583e-02 -1.47697723e+00 -5.77536285e-01 -3.31435829e-01
3.28090817e-01 -2.32348568e-03 3.90809655e-01 1.00926542e+00
-3.66299719e-01 2.01833636e-01 -2.15330422e-01 -2.21570954e-01
4.01955128e-01 -9.41112280e-01 6.75122976e-01 -9.25246119e-01
4.16838974e-02 -1.11477941e-01 -3.46888542e-01 -2.66843021e-01
-6.24336064e-01 -5.75551391e-01 -5.34309447e-01 -1.82035732e+00
4.10656571e-01 -4.78387177e-01 -6.56422555e-01 7.17030823e-01
-2.11554870e-01 5.08309245e-01 -6.29948406e-03 -1.43000126e-01
-4.09764498e-01 -2.80095011e-01 1.28974473e+00 -2.24966154e-01
-9.93142575e-02 1.28384214e-02 -9.10664260e-01 7.45314837e-01
1.39307463e+00 5.14945425e-02 -3.31862062e-01 -4.32912976e-01
1.36892453e-01 -3.26497227e-01 2.68356770e-01 -1.35586154e+00
3.18588823e-01 -2.40263373e-01 1.02506804e+00 -3.38744551e-01
3.32914829e-01 -7.58203804e-01 1.21333256e-01 6.88255966e-01
-1.42059803e-01 -2.95825273e-01 1.89086556e-01 4.68673706e-01
-6.73043281e-02 -2.71066457e-01 9.76599097e-01 -2.62667775e-01
-1.12740171e+00 5.67543507e-01 -3.33202094e-01 -2.46380478e-01
1.51662195e+00 -4.22306120e-01 -3.47715795e-01 -1.24665819e-01
-6.36265159e-01 1.62298277e-01 1.47062212e-01 4.72190559e-01
5.38737714e-01 -1.39615977e+00 -6.13815486e-01 -4.54754271e-02
1.32229283e-01 -6.40931576e-02 6.10192001e-01 9.25779164e-01
-9.56704199e-01 1.09476149e-01 -4.74318117e-01 -2.93314248e-01
-1.66684818e+00 -3.45647335e-02 7.02913284e-01 -3.96326512e-01
-1.60784528e-01 1.20463288e+00 -3.74742627e-01 -4.34816629e-01
3.91190290e-01 -1.42134234e-01 -5.18293202e-01 -6.14486896e-02
7.10528731e-01 5.36680818e-01 1.24928884e-01 -3.01635772e-01
-3.62917721e-01 6.90495312e-01 -5.29170036e-01 1.67854294e-01
1.20922673e+00 3.45790952e-01 -1.20206006e-01 2.04507336e-02
1.34293759e+00 -7.61349499e-02 -7.64527023e-01 3.74647826e-01
-2.95145035e-01 -3.45852375e-01 4.92940098e-02 -1.34189308e+00
-1.30476022e+00 7.62102067e-01 1.35080588e+00 2.07917899e-01
1.39526808e+00 -3.57947260e-01 9.07996655e-01 1.18913360e-01
1.86498329e-01 -1.40895247e+00 2.83408314e-01 4.52629626e-01
6.11737013e-01 -1.28614080e+00 -1.16056882e-01 -3.49505305e-01
-7.06856668e-01 9.66629744e-01 1.05862319e+00 -2.24767074e-01
8.67482305e-01 4.26106930e-01 5.51364183e-01 4.79235360e-03
-5.59620380e-01 7.35875312e-03 -2.27492258e-01 1.02400827e+00
7.44609714e-01 1.82417482e-01 -6.82528257e-01 5.86251199e-01
5.94471805e-02 4.37431067e-01 4.54048663e-01 8.57307553e-01
-4.91092145e-01 -1.24104261e+00 -2.15892717e-01 8.58935654e-01
-9.35447454e-01 1.14048406e-01 -6.93266094e-01 9.53414619e-01
3.38412553e-01 8.48322511e-01 -2.30195165e-01 -6.26234472e-01
1.18188515e-01 -1.37241349e-01 4.35118973e-01 -2.99743384e-01
-5.99184215e-01 -3.06032389e-01 2.75846690e-01 -2.54936039e-01
-3.84690583e-01 -1.26618534e-01 -1.27679479e+00 -3.79327774e-01
-3.79959941e-01 1.05095983e-01 1.05480611e+00 5.74584723e-01
-1.20373778e-02 5.25307059e-01 4.72902328e-01 -1.79959610e-01
-2.60423213e-01 -1.11478925e+00 -5.15124798e-01 3.44661534e-01
-1.43766642e-01 -4.10878301e-01 -7.69237429e-02 1.21290088e-01] | [15.66567325592041, -2.972818613052368] |
980ba6b8-fac0-475f-923b-c785bb521163 | 3d-reconstruction-from-public-webcams | 2108.09476 | null | https://arxiv.org/abs/2108.09476v2 | https://arxiv.org/pdf/2108.09476v2.pdf | 3D Reconstruction from public webcams | We investigate the possibility of 3D scene reconstruction from two or more overlapping webcam streams. A large, and growing, number of webcams observe places of interest and are publicly accessible. The question naturally arises: can we make use of this free data source for 3D computer vision? It turns out that the task to reconstruct scene structure from webcam streams is very different from standard structure-from-motion (SfM), and conventional SfM pipelines fail. In the webcam setting there are very few views of the same scene, in most cases only the minimum of two. These viewpoints often have large baselines and/or scale differences, their overlap is rather limited, and besides unknown internal and external calibration also their temporal synchronisation is unknown. On the other hand, they record rather large fields of view continuously over long time spans, so that they regularly observe dynamic objects moving through the scene. We show how to leverage recent advances in several areas of computer vision to adapt SfM reconstruction to this particular scenario and reconstruct the unknown camera poses, the 3D scene structure, and the 3D trajectories of dynamic objects. | ['Cenek Albl', 'Konrad Schindler', 'Tianyu Wu'] | 2021-08-21 | null | null | null | null | ['3d-scene-reconstruction'] | ['computer-vision'] | [ 2.05739364e-01 -2.24663451e-01 2.03471050e-01 -2.62756824e-01
-5.98673522e-01 -1.20802069e+00 7.84568548e-01 -1.73425287e-01
-2.29627877e-01 1.81190461e-01 1.10636735e-02 -1.10910840e-01
-4.70692944e-03 -2.90537179e-01 -8.85004580e-01 -5.83749712e-01
1.99036807e-01 9.00805891e-01 8.34991932e-01 -5.44210970e-02
2.07050979e-01 6.41477466e-01 -1.62817991e+00 3.75982635e-02
-3.91579717e-02 7.82783449e-01 6.04736805e-01 1.14293182e+00
-7.50289438e-03 6.40501738e-01 -1.90149635e-01 -2.92691350e-01
4.92378056e-01 -1.00056864e-01 -6.82461679e-01 9.94592845e-01
8.16045880e-01 -4.52646136e-01 -5.91732264e-01 9.08235431e-01
2.34923661e-02 2.15571135e-01 9.39136595e-02 -1.12209880e+00
2.66391009e-01 -4.82216060e-01 -5.97983956e-01 3.09804291e-01
8.41281950e-01 7.59803131e-02 5.31017065e-01 -7.70163178e-01
9.58986878e-01 1.08687270e+00 6.22716606e-01 8.42120051e-02
-1.29897964e+00 3.70935164e-03 3.89877439e-01 1.14628322e-01
-1.13496387e+00 -7.66013741e-01 7.59515107e-01 -6.06290460e-01
7.36105442e-01 2.30830103e-01 8.21674466e-01 1.41710389e+00
-4.79703434e-02 2.26759359e-01 7.50402749e-01 -3.41555089e-01
2.22880602e-01 2.60771453e-01 -3.05525772e-02 3.81612867e-01
3.14651281e-01 -1.30773578e-02 -7.96743751e-01 -2.57741690e-01
1.01691449e+00 4.95885402e-01 -3.31253022e-01 -9.52151120e-01
-1.57828891e+00 3.97833735e-01 -2.05655575e-01 -4.35576262e-03
-5.00205532e-03 -1.16374239e-01 -4.58812118e-02 3.42567950e-01
1.85676411e-01 1.21832229e-02 -6.18602574e-01 -3.21960330e-01
-6.77498043e-01 2.24888921e-01 1.03940058e+00 1.49463665e+00
8.00679147e-01 -2.24001825e-01 9.24889863e-01 3.08588326e-01
1.38906360e-01 7.77329862e-01 1.84461743e-01 -1.44235265e+00
7.18968570e-01 2.27889985e-01 3.90620589e-01 -1.04417956e+00
-2.91278154e-01 1.71878397e-01 -6.18178248e-01 4.28814888e-01
6.98059618e-01 1.07192740e-01 -3.61632138e-01 1.35233676e+00
7.72044420e-01 4.78814423e-01 -2.93844491e-01 8.41758072e-01
3.66249055e-01 3.08195561e-01 -9.52791214e-01 -3.59086335e-01
1.14489794e+00 -7.11997688e-01 -4.83360887e-01 -7.03115761e-01
-1.96578298e-02 -1.00960481e+00 7.20664859e-01 5.76866031e-01
-1.14077032e+00 -5.15903771e-01 -1.00247002e+00 -5.58807254e-02
-7.53590316e-02 -3.47936511e-01 3.54764342e-01 2.62934983e-01
-8.27886999e-01 5.44908106e-01 -1.28896427e+00 -7.66423762e-01
-7.56015480e-02 2.55434155e-01 -7.73036659e-01 -3.58075976e-01
-2.53177047e-01 9.21668828e-01 -7.07155764e-02 2.48022117e-02
-7.63839066e-01 -3.81935686e-01 -7.48675883e-01 -3.13916445e-01
6.82843685e-01 -5.90413034e-01 1.49956131e+00 -8.15444350e-01
-1.38808060e+00 1.03704143e+00 -4.00363714e-01 1.18998745e-02
6.99521184e-01 -3.64368290e-01 -1.99200258e-01 1.67947397e-01
-1.20314650e-01 -1.42340101e-02 8.37217152e-01 -1.41383243e+00
-7.43657291e-01 -7.90005863e-01 1.42945066e-01 4.75075334e-01
1.54002190e-01 -1.32828832e-01 -9.09148395e-01 1.48831261e-02
6.52531266e-01 -1.15555811e+00 -3.42256248e-01 2.96784908e-01
7.89768174e-02 6.25786543e-01 1.23216498e+00 -3.93237382e-01
4.40221697e-01 -2.15301704e+00 3.74763995e-01 -9.31321383e-02
1.76032320e-01 -9.74396318e-02 1.68283150e-01 5.06829321e-01
1.25690207e-01 -4.64606404e-01 -3.70006226e-02 -6.68681026e-01
-4.01758105e-01 5.65542579e-01 -3.96893889e-01 8.47350657e-01
-3.31975460e-01 3.37011814e-01 -9.09601390e-01 -1.85686663e-01
5.65711260e-01 3.27310145e-01 -5.15149117e-01 4.08258289e-01
-2.96125710e-01 9.40561712e-01 -3.20624441e-01 4.12737638e-01
8.22254717e-01 -4.71901894e-01 4.76543576e-01 1.60850003e-01
-2.34043643e-01 1.60330474e-01 -1.83957803e+00 2.15004444e+00
-2.80858904e-01 1.03024995e+00 2.78616250e-01 -8.00600588e-01
5.82774639e-01 3.73496711e-01 6.27648294e-01 -2.52859414e-01
-1.25010293e-02 -3.67582813e-02 -4.70701247e-01 -7.74893522e-01
4.54562396e-01 -5.48296981e-02 2.52912790e-01 2.15037689e-01
6.76336512e-02 -5.31212032e-01 -1.38915017e-01 2.12115914e-01
1.32025361e+00 2.90009946e-01 4.23842907e-01 1.99506775e-01
1.87246233e-01 8.94784033e-02 4.74279851e-01 5.82151949e-01
5.49527071e-02 1.23442101e+00 2.04275444e-01 -5.09835005e-01
-1.22138894e+00 -1.12157178e+00 -5.00933379e-02 3.21814597e-01
6.41773343e-01 -4.31318313e-01 -3.63040030e-01 -4.42013532e-01
-1.57040715e-01 2.17441365e-01 -3.16976100e-01 3.17171663e-01
-5.73657751e-01 -3.50953162e-01 -2.03072578e-01 2.00554922e-01
1.91367000e-01 -5.10329306e-01 -1.07433963e+00 1.76900655e-01
-4.90403682e-01 -1.77281833e+00 -5.30792296e-01 3.24282721e-02
-1.24289167e+00 -1.44551432e+00 -3.68068904e-01 -4.15226579e-01
6.29118502e-01 1.19974303e+00 1.43468595e+00 -3.65057439e-01
-4.37745266e-02 1.11721730e+00 -3.07222724e-01 -2.89653182e-01
-2.37935543e-01 -5.92140496e-01 1.87108368e-01 2.48227045e-01
2.23726287e-01 -1.11161435e+00 -4.16782230e-01 5.57752609e-01
-7.11191833e-01 2.75935996e-02 2.87097711e-02 2.42684811e-01
7.09104240e-01 1.33254617e-01 -5.60906291e-01 -7.82718718e-01
-3.61058205e-01 -4.53834563e-01 -1.11614859e+00 2.29597390e-02
-3.75537388e-02 -3.79457057e-01 4.42711711e-01 -7.97184169e-01
-1.00135052e+00 5.11306942e-01 3.14351439e-01 -1.00912297e+00
-5.72349846e-01 -6.92402720e-02 -4.42366958e-01 1.69354141e-01
5.83693862e-01 -6.60841465e-02 -5.48640732e-03 -7.92844415e-01
2.05307722e-01 2.48767883e-01 8.42150748e-01 -2.36075789e-01
9.39170480e-01 1.25968122e+00 1.34677842e-01 -1.10989988e+00
-7.38402605e-01 -9.13739204e-01 -1.27812338e+00 -1.67373851e-01
7.30750442e-01 -9.47810829e-01 -6.61652386e-01 4.29625273e-01
-1.42086864e+00 -3.27996790e-01 -4.60326821e-01 6.32257044e-01
-7.05591440e-01 5.81176221e-01 -7.13855326e-02 -8.34667861e-01
5.40152311e-01 -1.10255313e+00 1.45196962e+00 -5.04176654e-02
-1.36730626e-01 -1.00380635e+00 4.18813527e-01 3.63751948e-01
-1.18150584e-01 3.96422476e-01 2.08813593e-01 -1.73701391e-01
-1.05526161e+00 -4.21372354e-01 3.08000326e-01 1.68337766e-02
1.84431434e-01 8.86422619e-02 -1.09389305e+00 -2.05676004e-01
7.02179730e-01 2.62260109e-01 2.13692516e-01 4.34470028e-01
5.58749974e-01 -1.69684842e-01 -2.50563771e-01 7.50085115e-01
1.63171530e+00 1.93337172e-01 3.00627887e-01 2.65264690e-01
8.20050776e-01 8.65751684e-01 4.33722824e-01 5.31459808e-01
4.61541474e-01 1.20519006e+00 7.85316050e-01 3.11539501e-01
9.55073759e-02 -1.65776417e-01 4.77560192e-01 1.02084136e+00
-1.35034651e-01 -1.25283182e-01 -8.95101130e-01 3.97644162e-01
-1.88735592e+00 -9.98305559e-01 -5.49570918e-01 2.71821356e+00
1.66093782e-01 1.07674330e-01 4.47730683e-02 -3.92377675e-02
4.94938135e-01 2.22389624e-01 -6.87143803e-01 4.12133664e-01
-1.99516192e-01 -4.45313007e-01 6.13864601e-01 5.93184173e-01
-9.18437302e-01 4.92809057e-01 6.23508739e+00 7.83470366e-03
-7.67724693e-01 1.81232482e-01 -4.80587706e-02 -3.34545434e-01
-4.85442542e-02 5.88000298e-01 -9.23102379e-01 3.53657275e-01
9.30305660e-01 1.98506668e-01 6.27162397e-01 6.81839645e-01
9.09585580e-02 -3.56054395e-01 -1.47959697e+00 1.45278513e+00
2.10936174e-01 -1.21921921e+00 -2.74932235e-01 4.18984085e-01
6.51088953e-01 4.05290842e-01 -2.81253755e-01 -4.62811381e-01
5.15061021e-01 -4.13825095e-01 9.29311574e-01 6.23041332e-01
5.53526700e-01 -2.99793959e-01 3.15790921e-01 9.97555017e-01
-1.17949665e+00 3.48766632e-02 -4.56402689e-01 -7.80844986e-02
5.81508815e-01 5.46932697e-01 -3.47851038e-01 4.28196162e-01
9.41751599e-01 9.70299602e-01 -5.43594480e-01 9.88807261e-01
2.25941941e-01 1.62528560e-01 -6.14900231e-01 6.17557049e-01
-9.16803628e-02 -5.62864840e-01 8.35377812e-01 6.83571517e-01
5.38581491e-01 3.28645110e-01 1.45034522e-01 3.26842606e-01
2.36439735e-01 -6.83921218e-01 -1.18118501e+00 4.70073521e-01
1.93444818e-01 1.14900970e+00 -7.60867476e-01 -1.52836874e-01
-9.89221454e-01 1.04080665e+00 -1.87264234e-02 3.58200014e-01
-5.90223849e-01 2.83587456e-01 7.71642089e-01 3.99829447e-01
4.32772100e-01 -8.66863251e-01 1.33166596e-01 -1.81865203e+00
3.48170996e-01 -8.43062103e-01 2.87953943e-01 -1.04283357e+00
-1.04859030e+00 2.49363095e-01 1.43145114e-01 -1.74105048e+00
-2.64684200e-01 -4.81985807e-01 -4.17207330e-01 2.51140416e-01
-1.00201845e+00 -8.12974095e-01 -5.87852359e-01 8.88892889e-01
1.11667109e+00 1.05285048e-01 5.98983705e-01 4.56755869e-02
-1.18340954e-01 -2.94416875e-01 5.71494043e-01 -3.15041602e-01
5.26918232e-01 -1.14169836e+00 5.24221778e-01 9.73737359e-01
8.25558722e-01 2.38346249e-01 8.63386393e-01 -3.68580192e-01
-2.07692719e+00 -6.53178334e-01 5.79806149e-01 -1.42976248e+00
5.87409616e-01 -9.02995884e-01 -9.46900725e-01 1.09950793e+00
-1.58744142e-03 3.07588100e-01 2.24995673e-01 -1.11618444e-01
-3.58296305e-01 -7.71890804e-02 -7.13168621e-01 3.96597147e-01
1.25705028e+00 -5.56338906e-01 -4.65745807e-01 5.44657767e-01
6.60438418e-01 -7.84771919e-01 -5.07463574e-01 3.77604272e-03
6.64039373e-01 -1.29096794e+00 1.26404858e+00 -3.37267220e-01
9.59497467e-02 -3.21498811e-01 -7.55351245e-01 -8.93778920e-01
9.43129733e-02 -8.36217165e-01 -4.09225851e-01 1.02520430e+00
-2.24697590e-01 -3.61995190e-01 8.07192326e-01 7.50544310e-01
2.35950202e-01 2.60577928e-02 -9.90058303e-01 -8.89108062e-01
-4.01580304e-01 -6.50135040e-01 3.19675505e-01 8.08644950e-01
-4.39273030e-01 5.26291132e-01 -5.60800731e-01 5.33236325e-01
9.83874321e-01 2.86002010e-01 1.35978603e+00 -1.54160690e+00
-6.77901745e-01 9.05195847e-02 -4.87889260e-01 -1.50632942e+00
-1.43567979e-01 -3.47851604e-01 4.91766557e-02 -1.08101130e+00
1.22237228e-01 2.06379965e-02 4.07223076e-01 -3.53518277e-01
2.63745308e-01 -1.83951169e-01 2.24363104e-01 5.54751933e-01
-7.73113132e-01 1.89183533e-01 8.37877691e-01 1.72262400e-01
-1.45317316e-01 4.43732768e-01 -7.82203451e-02 1.22012794e+00
2.33401060e-01 -4.89431381e-01 -5.80857217e-01 -8.71872306e-01
3.12746942e-01 5.71062624e-01 5.93555272e-01 -8.61916244e-01
4.54970002e-01 -4.16977853e-01 3.74290705e-01 -8.07793081e-01
7.38369584e-01 -1.53924990e+00 8.08922529e-01 -1.40068248e-01
7.04114661e-02 3.22617322e-01 2.40105577e-03 1.11113858e+00
-1.04521558e-01 -1.92333177e-01 5.38546503e-01 -6.77183330e-01
-6.30250633e-01 4.92877424e-01 -5.32166123e-01 1.26173139e-01
8.77763748e-01 -4.49273646e-01 9.64675769e-02 -6.15766525e-01
-7.81878352e-01 -9.70308930e-02 9.62718189e-01 5.42881906e-01
5.22247374e-01 -1.22449434e+00 -2.43056536e-01 3.08934838e-01
1.05710208e-01 7.15931654e-01 2.64959246e-01 8.08752596e-01
-3.88100833e-01 2.14072704e-01 1.03673197e-01 -1.09663081e+00
-1.55043161e+00 6.52597964e-01 2.33417168e-01 3.24029685e-03
-1.17443895e+00 4.27543610e-01 4.40751821e-01 -5.44969440e-01
2.02515408e-01 -1.32463366e-01 1.74244076e-01 -1.14504427e-01
5.77858150e-01 5.51231503e-01 2.05812499e-01 -8.68417144e-01
-2.44069725e-01 1.00259411e+00 3.01197648e-01 -4.70447063e-01
1.48356199e+00 -7.73464262e-01 7.83735067e-02 9.71675932e-01
1.29795516e+00 2.30516881e-01 -1.89838171e+00 -5.49573183e-01
-5.28993569e-02 -9.18957055e-01 -3.80678207e-01 5.82541674e-02
-8.86588097e-01 1.10537195e+00 4.24577475e-01 2.55025297e-01
9.75788713e-01 2.87759304e-01 5.41657865e-01 5.72752714e-01
8.41818750e-01 -7.59349465e-01 1.15496270e-01 6.64518416e-01
7.52466738e-01 -1.32611740e+00 1.48599848e-01 -2.01815471e-01
-4.73867536e-01 1.19740427e+00 3.25891763e-01 -3.65221873e-02
6.61024451e-01 4.24816012e-01 3.04933898e-02 -2.70940870e-01
-8.99396956e-01 6.65141866e-02 -7.29827732e-02 7.94995010e-01
-3.39449942e-01 -3.84302586e-01 8.57023120e-01 -5.97866252e-02
-2.03062236e-01 -5.03281653e-01 8.49717855e-01 1.12214625e+00
-2.81440079e-01 -9.66240525e-01 -7.49805152e-01 4.99737486e-02
-1.46747351e-01 4.08605814e-01 -3.29428047e-01 8.22789609e-01
-6.46681488e-02 1.06608403e+00 1.83406010e-01 -1.96395114e-01
6.12852395e-01 -4.49345708e-02 7.12889433e-01 -5.03753841e-01
-1.95880234e-03 4.23878044e-01 -1.75338373e-01 -9.98783052e-01
-7.31479704e-01 -1.28586841e+00 -5.52120149e-01 -3.33363891e-01
-3.86370957e-01 -3.68616551e-01 8.07734907e-01 9.05283272e-01
1.46524519e-01 -2.17136424e-02 7.85703719e-01 -1.42359269e+00
-2.29026258e-01 -4.67527092e-01 -6.95382893e-01 5.09363830e-01
9.98344481e-01 -6.20975554e-01 -5.75527549e-01 7.06252635e-01] | [8.25352954864502, -2.4404947757720947] |
6bdc1c27-8373-4014-aed9-ff4f0409e8a3 | content-differences-in-syntactic-and-semantic-1 | null | null | https://aclanthology.org/N19-1047 | https://aclanthology.org/N19-1047.pdf | Content Differences in Syntactic and Semantic Representation | Syntactic analysis plays an important role in semantic parsing, but the nature of this role remains a topic of ongoing debate. The debate has been constrained by the scarcity of empirical comparative studies between syntactic and semantic schemes, which hinders the development of parsing methods informed by the details of target schemes and constructions. We target this gap, and take Universal Dependencies (UD) and UCCA as a test case. After abstracting away from differences of convention or formalism, we find that most content divergences can be ascribed to: (1) UCCA{'}s distinction between a Scene and a non-Scene; (2) UCCA{'}s distinction between primary relations, secondary ones and participants; (3) different treatment of multi-word expressions, and (4) different treatment of inter-clause linkage. We further discuss the long tail of cases where the two schemes take markedly different approaches. Finally, we show that the proposed comparison methodology can be used for fine-grained evaluation of UCCA parsing, highlighting both challenges and potential sources for improvement. The substantial differences between the schemes suggest that semantic parsers are likely to benefit downstream text understanding applications beyond their syntactic counterparts. | ['Ari Rappoport', 'Daniel Hershcovich', 'Omri Abend'] | 2019-06-01 | null | null | null | naacl-2019-6 | ['ucca-parsing'] | ['natural-language-processing'] | [ 4.56424534e-01 4.52247530e-01 -1.60346314e-01 -6.76682949e-01
-8.53257418e-01 -9.31946695e-01 7.37341106e-01 5.25854945e-01
-3.54470611e-01 6.41316652e-01 5.86423874e-01 -6.63869083e-01
-2.76978642e-01 -8.22937548e-01 -4.37930375e-01 -6.65395677e-01
2.02411398e-01 3.78688723e-01 6.39218867e-01 -3.98274213e-01
4.77281541e-01 3.29006106e-01 -1.64109862e+00 5.07495344e-01
6.98286355e-01 8.22270632e-01 3.16656977e-01 3.75487179e-01
-8.71032596e-01 6.87254727e-01 -7.26923943e-01 -6.82828307e-01
-1.58362031e-01 -6.49829626e-01 -1.42800498e+00 -1.14458069e-01
1.05746612e-01 4.18151289e-01 3.99430752e-01 1.24509490e+00
3.04106712e-01 -1.41618356e-01 5.63980579e-01 -9.72471356e-01
-6.00133300e-01 7.77609706e-01 -2.96802461e-01 3.36001098e-01
3.44049424e-01 -3.87854785e-01 1.54874861e+00 -3.70871693e-01
8.88135254e-01 1.53605461e+00 5.76690614e-01 6.18323803e-01
-9.95222867e-01 -1.80112496e-01 3.95580888e-01 3.53348665e-02
-1.06426489e+00 -4.08054501e-01 5.68045080e-01 -3.90112221e-01
9.91239607e-01 3.42532307e-01 2.22377479e-01 9.50068772e-01
-1.61181673e-01 6.36672437e-01 1.41743457e+00 -9.05065954e-01
2.18405038e-01 7.43983388e-02 5.84996819e-01 3.20595115e-01
2.76866227e-01 -1.49499401e-01 -2.98571825e-01 -2.01520398e-01
4.90495563e-01 -8.39695275e-01 -2.80876040e-01 -1.77370831e-01
-9.72143888e-01 1.21846652e+00 1.52523041e-01 7.14735270e-01
8.86787623e-02 -1.18364505e-01 7.33350754e-01 1.66782856e-01
2.80976057e-01 2.97580957e-01 -7.85832644e-01 -2.24409699e-01
-3.86922628e-01 3.61089915e-01 8.42952371e-01 9.57720339e-01
5.14383733e-01 -3.17081362e-01 2.30710834e-01 1.01577389e+00
2.57432252e-01 -2.41538975e-02 4.59958464e-01 -8.70259345e-01
6.07536674e-01 5.68825901e-01 2.73192078e-02 -7.71093905e-01
-3.75891685e-01 -6.17429987e-03 -2.12855428e-01 -1.31089404e-01
7.42257237e-01 -4.40091640e-03 -4.66890275e-01 2.03829813e+00
3.94291848e-01 -4.31099772e-01 2.85130978e-01 8.09439719e-01
9.19984818e-01 4.90151614e-01 7.75948346e-01 -3.09623808e-01
2.00385904e+00 -6.42440259e-01 -7.64967382e-01 -3.21942031e-01
1.02410400e+00 -9.03016031e-01 1.09242713e+00 -1.65811330e-01
-1.15471721e+00 -2.85039395e-01 -7.13253081e-01 -3.95916671e-01
-6.87258422e-01 -1.90235108e-01 9.52639282e-01 9.10520732e-01
-1.04008317e+00 4.93698239e-01 -6.74260139e-01 -7.53138721e-01
6.57883212e-02 2.39342064e-01 -2.40562946e-01 1.73372701e-01
-1.27215648e+00 1.03049946e+00 6.08282924e-01 1.61121234e-01
1.33698285e-01 -3.52987468e-01 -9.54513907e-01 -1.75142828e-02
3.42034280e-01 -4.97939169e-01 1.25331724e+00 -1.31723988e+00
-1.20135689e+00 1.42653942e+00 -2.99143136e-01 -1.36020184e-01
2.54737645e-01 -2.13035613e-01 -4.77058589e-01 7.61759409e-04
4.16339010e-01 5.39620578e-01 2.29846179e-01 -1.28799057e+00
-1.00625861e+00 -5.86089313e-01 4.11364734e-01 3.89319867e-01
2.10468054e-01 8.81612718e-01 -2.82989532e-01 -7.61683702e-01
4.66850609e-01 -8.06341231e-01 7.25418478e-02 -3.28967452e-01
-3.04318458e-01 -6.46732628e-01 7.11742222e-01 -5.49151242e-01
1.43413901e+00 -2.14401817e+00 1.65538117e-01 -2.14293674e-01
-2.89353818e-01 1.21119730e-01 1.37533307e-01 6.46162152e-01
-4.04389650e-01 4.04520631e-01 -4.90441412e-01 -2.17770025e-01
1.59263775e-01 6.84501171e-01 -2.82896370e-01 2.51303792e-01
3.51487815e-01 7.79297233e-01 -7.34508932e-01 -5.06120145e-01
4.09897082e-02 3.17419827e-01 -3.22402865e-01 -2.03249156e-02
-2.73056716e-01 2.90451944e-01 -7.97712445e-01 7.09879100e-01
5.69805861e-01 -1.55065497e-02 6.00077689e-01 -2.42476776e-01
-4.28655922e-01 9.67920363e-01 -1.35762680e+00 1.49714565e+00
-1.77935064e-01 3.69847149e-01 2.85796344e-01 -1.24482703e+00
6.96926534e-01 3.34192365e-01 -7.81267658e-02 -6.91583693e-01
2.23789722e-01 4.72814202e-01 2.77968258e-01 -3.77138734e-01
5.76891720e-01 -5.98591805e-01 -4.86558437e-01 3.61910462e-01
3.39111760e-02 -4.02133502e-02 1.18749544e-01 2.63776742e-02
9.18292761e-01 4.19513017e-01 6.76857948e-01 -8.50547731e-01
6.09198332e-01 1.89290211e-01 7.60573030e-01 4.31797355e-01
-4.89332825e-01 5.09711862e-01 9.84084547e-01 -3.06288779e-01
-7.77693927e-01 -7.77751148e-01 -5.62189281e-01 1.36181664e+00
3.05671513e-01 -4.79583025e-01 -9.59065139e-01 -8.33516598e-01
-3.70815516e-01 9.05969024e-01 -7.22240269e-01 3.73329192e-01
-1.04960370e+00 -1.06441569e+00 7.38433301e-01 6.95081294e-01
2.31919616e-01 -1.23356056e+00 -9.06711459e-01 1.94244742e-01
-3.16690147e-01 -1.31373811e+00 1.82105035e-01 5.49852192e-01
-8.21193635e-01 -1.27942920e+00 -1.30240709e-01 -1.05210268e+00
3.51349384e-01 1.91058025e-01 1.28847599e+00 2.62077391e-01
3.28333169e-01 3.77973497e-01 -7.75426507e-01 -4.64293212e-01
-6.16765380e-01 -1.90623589e-02 -4.42801327e-01 -4.44405556e-01
5.50957084e-01 -2.98460633e-01 -2.92533875e-01 1.11884087e-01
-9.98314619e-01 -4.81300503e-02 1.04982674e-01 6.18384600e-01
3.03861588e-01 -7.96886310e-02 5.82888603e-01 -1.54714930e+00
6.14431977e-01 -5.30410647e-01 -4.51966792e-01 2.95692563e-01
-3.37003618e-01 2.46677771e-01 2.73703694e-01 1.43013403e-01
-1.58361864e+00 -2.06661701e-01 -5.25513828e-01 5.14739096e-01
-5.52383661e-01 4.06682193e-01 -5.56132495e-01 3.17655414e-01
3.79213184e-01 -2.64968127e-01 -3.32842588e-01 -5.90506673e-01
5.09217441e-01 4.70682919e-01 4.47115719e-01 -1.13166583e+00
4.11345601e-01 4.29558396e-01 -2.12949619e-01 -8.59793007e-01
-8.93426597e-01 -3.33402723e-01 -8.50876033e-01 3.45639884e-01
1.33627164e+00 -7.44503558e-01 -1.66627273e-01 2.28045031e-01
-1.54893661e+00 -1.18519612e-01 -1.67724237e-01 9.39911082e-02
-3.78430605e-01 6.73890054e-01 -9.24785852e-01 -5.07152438e-01
-6.36968389e-02 -1.21637917e+00 1.13705754e+00 4.40900773e-02
-5.19542515e-01 -1.37853611e+00 -1.59396306e-02 4.80006844e-01
9.35240984e-02 2.22146124e-01 1.67663038e+00 -9.38854635e-01
-1.82037964e-01 1.92618892e-01 -4.89004672e-01 1.71845898e-01
2.53761053e-01 -2.60989331e-02 -1.14411604e+00 1.10292003e-01
1.03061289e-01 -1.72948554e-01 4.46084231e-01 3.01492631e-01
5.85775435e-01 4.23828624e-02 -2.99122274e-01 2.04450205e-01
1.54244208e+00 2.98770785e-01 7.69092083e-01 6.79664671e-01
5.07082820e-01 1.37283587e+00 6.32960260e-01 -2.15979721e-02
5.76456249e-01 5.77453852e-01 3.87062341e-01 9.90818739e-02
-2.49917373e-01 5.37565770e-03 2.59703755e-01 8.58510911e-01
-1.29612595e-01 -2.58091033e-01 -8.68127823e-01 5.95721841e-01
-1.82541060e+00 -6.57050312e-01 -7.11753249e-01 1.93104899e+00
5.94404995e-01 1.46380484e-01 3.65615897e-02 1.47457123e-01
1.03721082e+00 3.82060736e-01 1.47773325e-01 -7.86005020e-01
-4.24973577e-01 4.93293732e-01 1.65151343e-01 6.20114982e-01
-1.05988014e+00 1.33522570e+00 6.38006639e+00 7.47546375e-01
-8.75124693e-01 1.60513714e-01 4.83985394e-01 8.03854764e-01
-4.32865888e-01 5.37115157e-01 -8.12443495e-01 3.31983805e-01
9.17388260e-01 2.24739358e-01 -2.01654464e-01 6.23832643e-01
-7.95164481e-02 -2.35411897e-01 -1.07478380e+00 4.52484280e-01
-1.13732010e-01 -1.23876905e+00 2.65524630e-02 -1.37289807e-01
2.64062345e-01 -9.47622582e-02 -6.10192955e-01 2.40811720e-01
3.65652442e-01 -8.65113735e-01 1.04974890e+00 -4.27619547e-01
4.25591409e-01 -4.72917765e-01 7.72187054e-01 3.42793688e-02
-1.44339120e+00 1.88844234e-01 -3.00621331e-01 -3.09500873e-01
2.63285667e-01 1.59115419e-01 -4.94077280e-02 8.82252276e-01
6.53653443e-01 2.82290876e-01 -4.60477173e-01 3.60805362e-01
-4.84523714e-01 7.35391200e-01 -8.24727491e-03 2.59955395e-02
4.21684027e-01 -5.13566673e-01 4.64935124e-01 1.58469605e+00
3.16820778e-02 3.44904989e-01 -4.70989048e-02 6.79280102e-01
3.90975475e-01 4.50166821e-01 -4.65645015e-01 1.63680822e-01
4.23303872e-01 1.01442993e+00 -1.42351472e+00 -1.87490165e-01
-8.80218685e-01 9.76452649e-01 5.49398363e-01 1.66131154e-01
-7.78270662e-01 -1.06325053e-01 7.45465219e-01 -5.08797094e-02
4.37310010e-01 -5.18239141e-02 -3.96930605e-01 -1.05923903e+00
1.57926768e-01 -7.18219280e-01 9.30484474e-01 -6.13385975e-01
-1.33425832e+00 6.97823048e-01 2.40223229e-01 -5.60415447e-01
6.09279200e-02 -9.22155499e-01 -5.80731809e-01 8.94135714e-01
-1.75324738e+00 -1.22483313e+00 1.45487562e-01 3.15077513e-01
6.98723793e-01 4.16873068e-01 1.19156730e+00 1.95514411e-01
-4.78675067e-01 1.90467045e-01 -4.05335039e-01 2.33671591e-01
4.42130983e-01 -1.44309855e+00 3.34135175e-01 8.50583494e-01
1.42382970e-03 5.93791783e-01 7.63264477e-01 -4.42915618e-01
-9.83037233e-01 -7.75256395e-01 1.47737277e+00 -7.11224854e-01
8.61246943e-01 -3.39879125e-01 -1.19307399e+00 8.64727259e-01
4.08138365e-01 -1.03955351e-01 8.05307031e-01 4.82925564e-01
-4.46312994e-01 4.90090817e-01 -1.09206438e+00 5.42108655e-01
1.12648106e+00 -5.08701682e-01 -1.11266923e+00 6.92281500e-02
8.44282269e-01 -2.13124365e-01 -7.87683547e-01 2.41388276e-01
2.76220709e-01 -1.30901492e+00 5.35233021e-01 -7.31636345e-01
4.20150280e-01 -1.73675150e-01 -4.71195132e-01 -8.62733603e-01
-1.40015125e-01 -2.57187575e-01 6.58793032e-01 1.71191692e+00
3.81488204e-01 -7.68787503e-01 4.73413855e-01 7.87130773e-01
-4.76826757e-01 -3.68882835e-01 -9.79104042e-01 -4.77284610e-01
6.86463952e-01 -7.73621559e-01 4.08035785e-01 1.19132125e+00
1.63049325e-01 6.95426583e-01 3.04901451e-01 1.19843848e-01
1.96434572e-01 2.67442882e-01 2.35337526e-01 -1.36647713e+00
-2.04114348e-01 -5.75978100e-01 -3.42327774e-01 -9.51255739e-01
3.96271706e-01 -8.33840132e-01 -9.38594118e-02 -1.48712742e+00
7.72097111e-02 -7.94987261e-01 3.57806347e-02 2.44420782e-01
-1.20991856e-01 -4.13355008e-02 1.61235943e-01 2.46099845e-01
-3.26500922e-01 2.48868227e-01 8.03355098e-01 5.86935103e-01
-4.12939042e-02 -3.13738853e-01 -1.05653727e+00 1.16737306e+00
7.15678036e-01 -3.89239132e-01 -3.75708312e-01 -6.22307420e-01
3.09613168e-01 8.23063999e-02 8.50348622e-02 -3.48261505e-01
-9.12159383e-02 -2.95326680e-01 -1.80157438e-01 1.02645308e-02
1.21731125e-02 -7.16547072e-01 -9.02333111e-02 3.03145587e-01
-2.09485888e-01 4.11644667e-01 2.11060464e-01 3.34735185e-01
-4.02946323e-01 -6.82639599e-01 6.65988743e-01 -5.83038449e-01
-1.01169229e+00 -4.28225994e-01 -4.38843548e-01 3.93419355e-01
7.97770143e-01 -4.95617688e-01 -4.04191285e-01 1.99834198e-01
-7.67965913e-01 -1.22312590e-01 4.13386881e-01 4.97215390e-01
1.01890666e-02 -9.87434626e-01 -3.80476505e-01 -1.72309324e-01
1.54931635e-01 1.10914744e-01 6.91847503e-02 6.70917273e-01
-6.34712279e-01 4.56014156e-01 1.65003613e-01 -3.18279237e-01
-1.32922876e+00 3.36296588e-01 1.80973381e-01 -2.82813936e-01
-5.69204628e-01 7.15984643e-01 6.84873581e-01 -5.37981451e-01
4.14353237e-02 -4.63750571e-01 -4.22814071e-01 2.81388432e-01
1.86873883e-01 2.05857605e-01 1.02036819e-01 -1.27736473e+00
-4.88227069e-01 5.80627322e-01 1.32538348e-01 7.69626126e-02
1.10577893e+00 -4.35525984e-01 -4.78749365e-01 4.68293458e-01
9.53200340e-01 2.18927890e-01 -6.48268759e-01 -1.29365206e-01
6.88714683e-01 -1.25110254e-01 -3.58001649e-01 -7.93843210e-01
-6.38803124e-01 7.52844572e-01 2.61030704e-01 4.94169086e-01
9.63445604e-01 5.68725824e-01 5.59222400e-01 -2.04486653e-01
3.85527730e-01 -1.06829226e+00 -5.15413344e-01 5.32395303e-01
7.09936500e-01 -1.06873965e+00 -1.47856638e-01 -1.16938150e+00
-6.93873584e-01 1.22072363e+00 4.67968434e-01 -2.60247663e-02
4.55568135e-01 2.43452549e-01 2.83599049e-01 -5.25293946e-01
-5.53312302e-01 -5.32406628e-01 -6.60348013e-02 5.61775267e-01
1.00347459e+00 2.53620476e-01 -9.91397262e-01 7.28811800e-01
-3.86934161e-01 -6.51442468e-01 3.86877000e-01 1.20570779e+00
-4.12981749e-01 -1.65518248e+00 -3.38418394e-01 -7.44358525e-02
-7.81073570e-01 -2.27069780e-01 -4.25594568e-01 1.38275337e+00
4.37221229e-01 9.06476974e-01 1.75347403e-01 3.97415221e-01
4.42022681e-01 2.97985226e-01 5.03571272e-01 -8.45396578e-01
-6.70183897e-01 2.07314730e-01 6.49192035e-01 -3.68989766e-01
-9.91801620e-01 -9.65011120e-01 -1.43907130e+00 1.22551173e-02
-4.01573360e-01 4.02158111e-01 5.50419807e-01 1.30136800e+00
5.17780334e-02 2.61542737e-01 -6.62570745e-02 -3.66427273e-01
-3.56158525e-01 -7.30714023e-01 -5.43313682e-01 6.80593789e-01
-1.79488882e-01 -7.15060592e-01 -4.36009973e-01 1.07928634e-01] | [10.335277557373047, 9.475459098815918] |
a7a95292-9b35-44dd-ba31-840285f6874b | improving-negation-detection-with-negation | null | null | https://openreview.net/forum?id=kVeV2zg8EV | https://openreview.net/pdf?id=kVeV2zg8EV | Improving negation detection with negation-focused pre-training | Negation is a common linguistic feature that is crucial in many language understanding tasks, yet it remains a hard problem due to diversity in its expression in different types of text. Recent works show that state-of-the-art NLP models underperform on samples containing negation in various tasks, and that negation detection models do not transfer well across domains. We propose a new negation-focused pre-training strategy, involving targeted data augmentation and negation masking, to better incorporate negation information into language models. Extensive experiments on common benchmarks show that our proposed approach improves negation detection performance and generalizability over the strong baseline NegBERT (Khandelwal and Sawant, 2020). | ['Anonymous'] | 2022-01-16 | null | null | null | acl-arr-january-2022-1 | ['negation-detection'] | ['natural-language-processing'] | [ 3.82071733e-02 -1.15917273e-01 -6.20656133e-01 -6.95269287e-01
-4.66875583e-01 -8.05512846e-01 6.69807076e-01 4.35609937e-01
-7.58888602e-01 1.17562973e+00 2.54980057e-01 -4.15429980e-01
4.73540008e-01 -8.78712654e-01 -6.97036386e-01 -7.63496831e-02
2.25125059e-01 3.98132712e-01 3.09667140e-01 -9.16181087e-01
2.15179585e-02 3.55274469e-01 -1.25038290e+00 6.91876709e-01
1.03969026e+00 6.31549239e-01 -3.47233862e-01 3.92718315e-01
-5.77402592e-01 1.05360401e+00 -6.91937089e-01 -7.82024384e-01
3.34700160e-02 -3.61486912e-01 -6.45481169e-01 -7.06637502e-01
4.75122303e-01 -1.14421826e-02 -6.53284118e-02 1.46032965e+00
3.78749132e-01 -1.02338895e-01 6.33691549e-01 -1.07665634e+00
-9.53978121e-01 9.18791115e-01 -5.49044251e-01 3.69334042e-01
4.79783982e-01 -5.59580233e-03 1.39922798e+00 -1.11252105e+00
6.74865186e-01 1.51602876e+00 7.95121253e-01 8.98793697e-01
-8.72712016e-01 -8.59336436e-01 3.68851244e-01 3.07607591e-01
-1.14707398e+00 -2.16557473e-01 7.20075905e-01 -6.39783889e-02
1.57743347e+00 -1.58589810e-01 4.25669581e-01 1.27140296e+00
3.84246588e-01 9.90390122e-01 1.01228189e+00 -6.21231437e-01
-7.55954487e-03 4.91367653e-02 2.44164467e-01 5.79844356e-01
5.88235974e-01 -2.67505437e-01 -7.57344365e-01 -2.46227965e-01
1.00580104e-01 -3.61408472e-01 -3.22543919e-01 7.33129457e-02
-1.10290980e+00 9.29669440e-01 3.14400308e-02 4.81959581e-01
-3.28283221e-01 -7.80065656e-02 9.95416284e-01 7.57139385e-01
5.48121989e-01 3.65725487e-01 -1.20328128e+00 -1.75327525e-01
-7.02647448e-01 5.91015816e-01 1.19275224e+00 8.85145724e-01
6.02512300e-01 -9.53246355e-02 -1.95887670e-01 7.07840204e-01
1.96369991e-01 7.32472599e-01 4.43124443e-01 -5.62211633e-01
7.52103627e-01 9.23616409e-01 -9.17594805e-02 -6.91849291e-01
-3.94426197e-01 -3.32519710e-01 -9.15187478e-01 -2.41980821e-01
3.40658337e-01 -3.86736065e-01 -8.98984075e-01 1.92560315e+00
8.43763351e-02 -5.07174656e-02 6.57605290e-01 4.32271779e-01
1.06477630e+00 5.30564308e-01 3.39583993e-01 -8.74931067e-02
1.28951526e+00 -7.72880435e-01 -1.25862694e+00 -8.92447889e-01
1.22470129e+00 -6.96636379e-01 1.19579375e+00 6.87797606e-01
-9.25787687e-01 -1.92978665e-01 -1.10009313e+00 -2.37886161e-01
-7.41444349e-01 -1.17341317e-01 8.97251725e-01 7.68998146e-01
-5.52442074e-01 1.61662742e-01 -4.70668584e-01 -3.38529229e-01
5.01229405e-01 1.99576005e-01 -6.67651951e-01 -4.80352342e-01
-1.91855991e+00 1.13369727e+00 9.40502346e-01 1.80120736e-01
-3.02649260e-01 -5.64094722e-01 -1.25046730e+00 -1.48116559e-01
5.89908540e-01 -4.63272065e-01 1.26467943e+00 -7.66667247e-01
-1.11136651e+00 1.08479750e+00 -4.44882214e-01 -6.78459167e-01
3.46913368e-01 -6.40047193e-01 -6.81758165e-01 -2.31002182e-01
9.96890198e-03 6.26386464e-01 4.13925320e-01 -8.77643049e-01
-6.63771331e-01 -4.39685695e-02 2.11221278e-01 2.74205297e-01
-4.35894072e-01 3.51578325e-01 -2.70229816e-01 -5.72261214e-01
-1.23896278e-01 -4.13370222e-01 -1.23121895e-01 -1.23987667e-01
-3.86180580e-01 -7.55509615e-01 8.99244487e-01 -3.57015610e-01
1.18880641e+00 -1.87816620e+00 -1.76784083e-01 -1.15732811e-02
-1.10352851e-01 7.35371828e-01 -2.25269809e-01 4.24809724e-01
-1.36243388e-01 3.31120253e-01 -2.89261699e-01 -2.53013432e-01
1.21600002e-01 8.30376923e-01 -4.55125421e-01 2.14777261e-01
6.34502292e-01 1.15544367e+00 -1.11815524e+00 -4.33174878e-01
-1.02563888e-01 3.21784556e-01 -6.93482578e-01 -1.16284728e-01
-6.89381003e-01 5.61639443e-02 -1.33322045e-01 1.18529403e+00
8.21756780e-01 3.55596244e-02 2.95738339e-01 -2.29317755e-01
2.81549424e-01 5.83237350e-01 -1.17648041e+00 1.44016612e+00
-7.76089281e-02 3.90154362e-01 9.91960987e-02 -1.14836991e+00
8.24467540e-01 3.88084471e-01 -3.02108496e-01 -7.35909581e-01
2.25207150e-01 4.80697453e-01 2.76405662e-01 -5.02140164e-01
4.32513773e-01 -5.29958785e-01 -3.81219119e-01 1.16233453e-02
2.39093184e-01 -4.73871171e-01 6.99152946e-01 4.88096587e-02
1.12492144e+00 -1.56403229e-01 7.54809141e-01 -2.95973688e-01
1.06059146e+00 8.19163397e-02 9.94457901e-01 8.26122940e-01
-5.35193861e-01 2.70574272e-01 8.43607068e-01 -2.29948282e-01
-4.91408676e-01 -7.19657481e-01 -1.51867317e-02 1.20575202e+00
-2.80289799e-01 -4.78746325e-01 -2.58049518e-01 -1.17402303e+00
1.78113759e-01 9.20464098e-01 -6.22893274e-01 -3.13692063e-01
-9.38372433e-01 -7.51279414e-01 1.10620499e+00 7.38039136e-01
5.82513034e-01 -1.41392553e+00 1.77032515e-01 1.84917018e-01
-2.38372996e-01 -1.46159601e+00 -3.26845236e-02 7.09175646e-01
-6.79981172e-01 -1.06282485e+00 -1.77616149e-01 -1.12123227e+00
3.93052310e-01 -4.21021760e-01 1.49808717e+00 -1.21341634e-03
2.12263748e-01 2.77967397e-02 -3.11839283e-01 -8.12996447e-01
-6.50902987e-01 1.20978691e-01 3.86834480e-02 -6.37911677e-01
1.05309284e+00 -1.33259058e-01 4.22233380e-02 -2.17445508e-01
-9.52067494e-01 -6.52457058e-01 6.21099412e-01 1.13554907e+00
5.46576500e-01 3.29458974e-02 8.13991427e-01 -1.38917196e+00
1.01519942e+00 -3.94497395e-01 -4.19398487e-01 2.85094440e-01
-2.47443169e-01 -5.67316003e-02 8.12545538e-01 -3.62145603e-01
-9.62425172e-01 -2.24002481e-01 -4.36903328e-01 4.56051296e-03
-1.32669017e-01 1.00724924e+00 -4.48440969e-01 -3.77959609e-02
5.81829429e-01 1.30140096e-01 -2.96655208e-01 -2.55842239e-01
5.12348294e-01 3.75491112e-01 4.45024222e-01 -5.92096329e-01
6.13282859e-01 3.59049648e-01 -8.47459882e-02 -9.01230812e-01
-1.48447740e+00 -4.14846718e-01 -6.64981663e-01 5.26679277e-01
4.00337040e-01 -1.12676144e+00 -4.53621626e-01 6.58987999e-01
-1.44270897e+00 -1.36352703e-01 -3.11048150e-01 3.38732898e-01
7.08330497e-02 4.62054789e-01 -8.16355526e-01 -6.25773787e-01
-6.19957566e-01 -4.76519734e-01 7.27501571e-01 1.20000288e-01
-3.42417240e-01 -1.30185235e+00 2.12818325e-01 2.77776361e-01
2.12539449e-01 -1.35643572e-01 1.19829106e+00 -1.31274986e+00
-3.10706394e-03 -2.97266126e-01 -2.47409001e-01 1.09648192e+00
1.64107263e-01 -1.74292058e-01 -7.90189147e-01 -8.91209990e-02
1.42156273e-01 -8.59682977e-01 1.05348349e+00 1.39787355e-02
8.35137486e-01 -1.31608561e-01 -2.35639170e-01 1.80458128e-01
1.22014630e+00 -1.97857857e-01 5.00360906e-01 3.58904749e-01
5.70436180e-01 3.98679554e-01 8.55287075e-01 5.74903786e-02
4.38602030e-01 -2.27827802e-01 3.65233421e-01 -6.81327879e-02
1.03906251e-01 -9.45206881e-02 4.97138709e-01 7.17155278e-01
4.99254018e-01 -6.81283474e-01 -1.10825574e+00 8.18002701e-01
-1.75592613e+00 -8.34735334e-01 -1.61505684e-01 1.49493468e+00
1.36488760e+00 7.87787735e-01 -5.58716178e-01 2.02872276e-01
3.45332265e-01 6.16853647e-02 -4.40522254e-01 -7.18700230e-01
-1.02112949e+00 4.72400069e-01 2.80429542e-01 5.63011348e-01
-1.32666016e+00 1.35587466e+00 7.11584663e+00 7.51940370e-01
-9.92459416e-01 3.11480556e-02 2.09336281e-01 2.44119257e-01
-3.64101976e-01 -1.59162417e-01 -1.20583153e+00 -6.26486614e-02
6.83712542e-01 -3.05426538e-01 -1.08591035e-01 6.63743615e-01
-9.43514146e-03 -9.30818245e-02 -1.31086111e+00 6.58671200e-01
4.71880585e-01 -1.10699892e+00 3.19493860e-01 -5.59543967e-01
8.56484294e-01 4.30306852e-01 -1.90637380e-01 8.96347284e-01
2.44274303e-01 -1.21758449e+00 3.64734501e-01 2.23836809e-01
2.57275760e-01 -8.97536874e-01 1.50857079e+00 6.60638750e-01
-9.32082891e-01 2.24628914e-02 -2.79356152e-01 -5.27413309e-01
1.85327157e-01 8.31400216e-01 -8.46707582e-01 4.47611213e-01
4.29495603e-01 8.33623767e-01 -5.24422884e-01 4.86483604e-01
-8.87097597e-01 8.60191047e-01 -4.73346174e-01 -3.25040311e-01
2.66021967e-01 1.05649807e-01 2.76253492e-01 1.69354904e+00
-1.51545882e-01 -6.17176183e-02 3.43254417e-01 7.06642210e-01
-5.96581101e-01 4.39872384e-01 -6.90928280e-01 -3.94392461e-01
3.29370290e-01 1.03791821e+00 -3.87522757e-01 -4.79757130e-01
-1.03748977e+00 7.13001430e-01 4.46595550e-01 2.79318959e-01
-6.38528287e-01 -6.24408424e-01 7.98509479e-01 -1.65269449e-01
5.79686761e-01 -1.56655774e-01 -2.54994959e-01 -1.39228988e+00
3.52658391e-01 -1.16350520e+00 5.15866160e-01 -4.30471063e-01
-1.92628014e+00 2.98090339e-01 -1.86650872e-01 -7.03908145e-01
-2.86798000e-01 -1.03962946e+00 -4.23850715e-01 9.53205049e-01
-2.21191263e+00 -1.25360858e+00 2.12834567e-01 6.09320879e-01
1.87040851e-01 -2.73289569e-02 1.01069987e+00 4.57005739e-01
-3.08400750e-01 7.48852551e-01 -2.63016939e-01 6.25665963e-01
1.08724344e+00 -1.32447195e+00 4.49889958e-01 9.20555651e-01
-5.63762933e-02 8.76363218e-01 7.91759610e-01 -9.39821124e-01
-1.33944988e+00 -1.09186387e+00 1.74453151e+00 -4.02273118e-01
1.06121802e+00 -5.19481957e-01 -1.40733504e+00 9.16674495e-01
4.46173728e-01 2.88467914e-01 7.54830360e-01 2.92345166e-01
-6.16689146e-01 1.12674624e-01 -1.35497808e+00 6.53748930e-01
9.29644346e-01 -8.61158669e-01 -1.26717675e+00 4.35348272e-01
7.56992936e-01 -4.57384646e-01 -4.26853716e-01 8.91642392e-01
8.51952285e-02 -6.67388499e-01 6.37027502e-01 -9.03983295e-01
2.90956289e-01 -2.79525518e-01 -3.13701153e-01 -9.42315102e-01
2.58397967e-01 -4.33091968e-01 -2.78195709e-01 1.32504654e+00
6.54119551e-01 -7.58109391e-01 8.91235769e-01 1.82630166e-01
5.08917831e-02 -7.12695360e-01 -8.85893881e-01 -7.23078489e-01
6.16628051e-01 -6.93860412e-01 1.17449708e-01 1.15965807e+00
1.16789386e-01 7.15556145e-01 -9.67056490e-03 1.54764220e-01
2.29300395e-01 -2.19355837e-01 4.05715317e-01 -1.23441696e+00
1.42568991e-01 -3.05548549e-01 -4.74048465e-01 -1.14597571e+00
7.94295549e-01 -9.94626760e-01 -2.68371291e-02 -1.58933532e+00
3.69729660e-02 6.55118823e-02 -3.87779862e-01 1.03942621e+00
-3.52472782e-01 2.56299347e-01 -2.88917005e-01 -4.76471007e-01
-6.79351032e-01 8.09121370e-01 1.08009017e+00 -3.31851691e-01
-5.43671101e-03 -3.79752964e-01 -8.05134118e-01 1.23706329e+00
5.11666417e-01 -5.75402558e-01 -2.81186074e-01 -3.80983412e-01
8.50784838e-01 -5.95600069e-01 1.20592892e-01 -6.95320487e-01
2.70700544e-01 -1.89943075e-01 2.86417991e-01 -9.99610364e-01
3.27995896e-01 -5.98748207e-01 -9.59477603e-01 6.36074841e-01
-1.18736297e-01 2.51977026e-01 7.11606681e-01 4.22056764e-01
-6.22150242e-01 -4.30534601e-01 4.90607649e-01 -1.28942162e-01
-8.71571183e-01 -1.30860120e-01 -3.10185850e-01 9.00925159e-01
6.87091947e-01 9.66746956e-02 -4.03494984e-01 -1.56963363e-01
-4.59961712e-01 5.34146190e-01 1.50285587e-01 4.56003428e-01
5.52194238e-01 -9.67146337e-01 -9.15295899e-01 3.08679223e-01
1.53790966e-01 1.18322149e-01 -2.01533034e-01 5.21812439e-01
-3.75815511e-01 4.83683765e-01 2.29670346e-01 -5.08812547e-01
-1.29441261e+00 1.65519655e-01 3.79931182e-01 -6.19138539e-01
-2.40545005e-01 1.24772668e+00 -1.42191604e-01 -1.09000635e+00
3.49823505e-01 -8.65327537e-01 -4.26655531e-01 1.36761338e-01
5.39904416e-01 -2.01384932e-01 4.87678528e-01 -4.21926230e-01
-7.90463388e-01 2.83948213e-01 -4.22799557e-01 2.07829818e-01
1.11839294e+00 3.01846147e-01 -4.85440284e-01 6.11708462e-01
8.51034522e-01 2.51830131e-01 -2.68251598e-01 -5.03518462e-01
4.42532957e-01 -1.44510586e-02 -1.81927204e-01 -1.15983772e+00
-6.63963497e-01 7.03245699e-01 8.54975507e-02 -7.27137700e-02
1.10722625e+00 -1.24468058e-01 7.86671579e-01 1.28817666e+00
1.02159932e-01 -1.03420818e+00 -1.20199487e-01 1.62648940e+00
7.06015706e-01 -1.67182887e+00 -4.13138792e-02 -7.34014809e-01
-5.33393323e-01 9.95993137e-01 7.98792720e-01 1.65442247e-02
8.33566487e-01 6.93093657e-01 4.65695798e-01 -1.88764513e-01
-1.10521829e+00 -1.11400768e-01 6.32200390e-02 4.29378480e-01
9.06049132e-01 -2.42306486e-01 -5.84739745e-01 9.94444013e-01
-7.65925273e-02 2.22619861e-01 4.46700096e-01 1.47624624e+00
-2.72057831e-01 -1.50245166e+00 2.84038056e-02 3.82327676e-01
-7.87514627e-01 -6.15937531e-01 -5.69400489e-01 1.00837278e+00
3.70012373e-01 8.10279548e-01 -3.53098661e-01 1.59824297e-01
7.55254090e-01 5.06366670e-01 4.48574185e-01 -1.06608748e+00
-8.80574346e-01 -2.40611553e-01 4.22492981e-01 -3.79627317e-01
-5.31583071e-01 -4.83462423e-01 -1.66483307e+00 -2.06264272e-01
-4.75251138e-01 2.18039304e-01 2.21103132e-01 1.34933364e+00
6.36129007e-02 3.96874547e-01 -2.75932193e-01 -5.64848147e-02
-8.11682284e-01 -1.03631890e+00 -4.57117796e-01 4.32648152e-01
4.39681649e-01 -3.35028172e-01 -6.00356758e-01 -1.72892615e-01] | [10.431546211242676, 9.250840187072754] |
28b43352-ac2d-4740-b584-4269950ca745 | people-penguins-and-petri-dishes-adapting | 1711.05586 | null | http://arxiv.org/abs/1711.05586v1 | http://arxiv.org/pdf/1711.05586v1.pdf | People, Penguins and Petri Dishes: Adapting Object Counting Models To New Visual Domains And Object Types Without Forgetting | In this paper we propose a technique to adapt a convolutional neural network
(CNN) based object counter to additional visual domains and object types while
still preserving the original counting function. Domain-specific normalisation
and scaling operators are trained to allow the model to adjust to the
statistical distributions of the various visual domains. The developed
adaptation technique is used to produce a singular patch-based counting
regressor capable of counting various object types including people, vehicles,
cell nuclei and wildlife. As part of this study a challenging new cell counting
dataset in the context of tissue culture and patient diagnosis is constructed.
This new collection, referred to as the Dublin Cell Counting (DCC) dataset, is
the first of its kind to be made available to the wider computer vision
community. State-of-the-art object counting performance is achieved in both the
Shanghaitech (parts A and B) and Penguins datasets while competitive
performance is observed on the TRANCOS and Modified Bone Marrow (MBM) datasets,
all using a shared counting model. | ["Noel E. O'Connor", 'Kevin McGuinness', 'Ciara E. Keogh', 'Suzanne Little', 'Mark Marsden'] | 2017-11-15 | people-penguins-and-petri-dishes-adapting-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Marsden_People_Penguins_and_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Marsden_People_Penguins_and_CVPR_2018_paper.pdf | cvpr-2018-6 | ['object-counting'] | ['computer-vision'] | [ 1.40086859e-01 -3.70260507e-01 1.39993086e-01 -3.17643881e-02
-3.07725042e-01 -2.41248712e-01 9.49907541e-01 7.46604502e-01
-1.24325037e+00 1.00939703e+00 -2.38143638e-01 -1.23114534e-01
3.88816953e-01 -7.08922327e-01 -4.22020823e-01 -6.65529788e-01
4.38903719e-02 9.18257535e-01 5.16154230e-01 8.24865922e-02
3.76417816e-01 9.33676243e-01 -1.44190967e+00 3.24206382e-01
4.09154683e-01 7.34824717e-01 1.00633621e-01 1.27138698e+00
-2.59531438e-01 1.08470833e+00 -7.51231551e-01 -2.43003741e-01
4.37268950e-02 -2.25626707e-01 -7.18455255e-01 -1.30655244e-01
7.20994174e-01 -1.09643526e-01 1.19202524e-01 6.86630309e-01
6.18626416e-01 -1.00896247e-01 1.05645084e+00 -1.00187063e+00
-3.40178519e-01 -3.98380868e-02 -5.00810444e-01 7.28585839e-01
-2.47651003e-02 9.58594829e-02 2.54136443e-01 -7.70182550e-01
5.55418313e-01 9.34740365e-01 9.49928999e-01 5.59364378e-01
-1.06849182e+00 -6.08701468e-01 -4.95557129e-01 6.36816844e-02
-1.26024961e+00 -1.22792445e-01 4.02838364e-02 -4.37279373e-01
1.09240556e+00 2.49700367e-01 8.18977773e-01 8.40742767e-01
2.51479894e-01 5.51167786e-01 1.17735732e+00 -7.94626057e-01
5.50845325e-01 2.12774411e-01 -2.00069115e-01 7.83488810e-01
6.16715074e-01 -2.27172449e-01 -2.62522519e-01 -6.41468391e-02
1.03063619e+00 -1.58184711e-02 9.23255906e-02 -2.65895903e-01
-1.27151024e+00 8.18341494e-01 4.19077814e-01 6.73771381e-01
-1.37678221e-01 4.68356162e-01 9.14138556e-01 -1.52483746e-01
8.17923486e-01 2.08994344e-01 -5.07832289e-01 6.00673109e-02
-1.09755874e+00 4.51105863e-01 8.60106885e-01 8.15220594e-01
4.27178174e-01 4.29787226e-02 -6.99518740e-01 9.15566266e-01
2.24681711e-03 3.55907649e-01 5.56123674e-01 -7.54556358e-01
1.10544050e-02 7.07805216e-01 6.45709084e-03 -6.86656654e-01
-8.02912116e-01 -5.28583825e-01 -9.90651369e-01 4.08726513e-01
9.01907742e-01 3.24153364e-01 -1.25505269e+00 1.33422768e+00
5.44183075e-01 4.22879428e-01 -2.66838789e-01 4.33567971e-01
9.44281518e-01 1.55877665e-01 4.50514108e-01 -8.09686705e-02
1.66236782e+00 -6.96337283e-01 -1.57385930e-01 1.55724315e-02
8.23717833e-01 -5.75008392e-01 6.04386806e-01 2.91459322e-01
-1.07920849e+00 -3.16284060e-01 -1.07645655e+00 -2.90670902e-01
-9.36810553e-01 1.21645391e-01 5.48925102e-01 7.79638231e-01
-9.78380740e-01 3.03812206e-01 -8.58203590e-01 -9.15077329e-01
9.44532573e-01 4.24592137e-01 -2.46276632e-01 1.13781719e-02
-5.27665973e-01 1.13120949e+00 3.85921061e-01 -5.29012561e-01
-8.09694409e-01 -8.48474562e-01 -6.25344753e-01 1.26199611e-02
-1.37387574e-01 -8.94515455e-01 1.03029382e+00 -8.04097772e-01
-1.11608422e+00 1.48133326e+00 1.80412650e-01 -7.12347567e-01
9.95741963e-01 5.36938608e-01 -3.32918428e-02 1.51650518e-01
3.71814847e-01 8.81286979e-01 4.64095622e-01 -1.10340667e+00
-9.96416926e-01 -3.22846055e-01 -3.74675244e-01 -2.55564839e-01
-2.05346480e-01 6.61639646e-02 -4.79797810e-01 -7.11728454e-01
-5.65296233e-01 -5.98723114e-01 -8.22596550e-02 1.79604694e-01
-1.96145773e-01 -1.98636949e-01 6.65965915e-01 -4.69550461e-01
7.15596855e-01 -1.51144660e+00 -8.91854465e-02 2.20157858e-02
3.91187787e-01 4.20946807e-01 -9.76235718e-02 9.93478149e-02
2.53601149e-02 -3.39538574e-01 -3.99627686e-01 -7.58277655e-01
-4.58251834e-01 3.73280287e-01 5.76194286e-01 9.34009850e-01
2.72208422e-01 8.61656845e-01 -8.85094702e-01 -9.64521110e-01
3.63792837e-01 6.78929627e-01 -4.31713194e-01 -3.35163534e-01
-1.50685728e-01 5.55469871e-01 1.89499825e-01 9.12121952e-01
8.60696435e-01 -2.72744149e-01 -8.96982923e-02 2.88666546e-01
-1.27854645e-01 -6.89368367e-01 -8.36130500e-01 1.17573023e+00
-4.03495520e-01 5.60378253e-01 -8.45205337e-02 -8.91695321e-01
7.85406768e-01 -3.56094874e-02 5.43087006e-01 -8.29627872e-01
6.37124836e-01 4.65906084e-01 -9.03417245e-02 -3.52051556e-02
4.66152161e-01 -5.04312217e-01 7.91728869e-02 -3.16653401e-02
5.87410390e-01 -1.54286742e-01 7.85970271e-01 7.68599659e-02
1.12019753e+00 -3.09347510e-01 8.76697958e-01 -5.97879827e-01
9.55006242e-01 2.28390843e-01 2.28701815e-01 6.10223114e-01
-5.72597623e-01 7.89868772e-01 4.73794103e-01 -8.19031477e-01
-1.48107994e+00 -7.26109624e-01 -5.07181287e-01 1.07468927e+00
-4.55722779e-01 2.76818573e-01 -9.24603105e-01 -4.44681972e-01
1.59440190e-01 4.36766326e-01 -1.14384210e+00 2.75347024e-01
-7.09557652e-01 -8.15133095e-01 9.20717776e-01 7.95102775e-01
5.01994610e-01 -1.17371607e+00 -7.78115034e-01 2.56088078e-01
9.00849104e-02 -1.17882895e+00 -3.51204544e-01 5.46510696e-01
-7.92407990e-01 -1.47675037e+00 -1.32421815e+00 -8.99891078e-01
5.34224510e-01 -2.44898051e-01 1.33489692e+00 5.33588707e-01
-6.70350552e-01 3.00247073e-01 -1.23927251e-01 -1.04767132e+00
-6.76095486e-01 1.84908465e-01 -1.96659073e-01 -3.05602163e-01
6.34875238e-01 -7.85121247e-02 -4.11915183e-01 1.43607752e-02
-1.03125548e+00 -3.01119536e-01 1.65772140e-01 9.22107875e-01
6.27023399e-01 -4.88037825e-01 4.64241177e-01 -1.06090128e+00
3.66919458e-01 -2.84825832e-01 -7.15047777e-01 9.00676325e-02
-3.70620012e-01 -3.19176018e-01 6.23064160e-01 -6.28847897e-01
-6.41062975e-01 3.86838019e-02 8.61002132e-02 -3.24332148e-01
-2.25425318e-01 -8.29450134e-03 4.31604385e-01 -4.41809684e-01
8.59724820e-01 2.40997925e-01 -6.02127379e-03 -1.36646196e-01
-9.07298774e-02 3.76788467e-01 9.82685685e-01 -2.96458483e-01
4.45786119e-01 9.23828065e-01 6.45286679e-01 -8.26306462e-01
-4.28965122e-01 -1.04294169e+00 -7.26363420e-01 -2.68041611e-01
1.04327512e+00 -7.53533483e-01 -6.08507276e-01 9.70771313e-01
-1.16827226e+00 -4.89286840e-01 -4.62386280e-01 3.24056774e-01
-6.91904783e-01 2.40198940e-01 -7.68786788e-01 -8.09884071e-01
-4.16737556e-01 -7.56443083e-01 1.24592185e+00 2.40611896e-01
-3.25970203e-02 -1.25305879e+00 6.02331936e-01 9.37924348e-03
4.77353722e-01 5.00792563e-01 9.63566005e-01 -7.86578059e-01
-1.79285586e-01 -5.77769637e-01 -5.48098564e-01 4.10730511e-01
-2.01330215e-01 7.11593777e-02 -1.08383107e+00 -2.83082724e-01
-1.53003335e-01 -1.51195467e-01 1.18777049e+00 7.26829886e-01
1.15571260e+00 1.35693580e-01 -3.94064039e-01 7.71346390e-01
1.76604843e+00 -9.73487571e-02 9.90295410e-01 6.53699636e-01
4.34592158e-01 1.87980801e-01 -7.97959864e-02 5.40583968e-01
2.47418880e-01 3.88849646e-01 4.90696520e-01 -2.35128120e-01
-3.29516530e-01 1.54528812e-01 -5.27200520e-01 5.05962670e-01
-5.52852273e-01 -3.25613439e-01 -1.09633565e+00 8.20003152e-01
-1.47086871e+00 -8.31613660e-01 -3.60045165e-01 2.07058144e+00
5.58666825e-01 -7.77867287e-02 4.64740247e-01 3.00762892e-01
6.84756219e-01 -4.82575268e-01 -1.73006684e-01 -5.69549501e-01
-1.86477900e-01 6.93065464e-01 1.11855948e+00 2.19767377e-01
-1.25436413e+00 5.71359277e-01 6.46909189e+00 1.04714358e+00
-9.59615171e-01 3.23362142e-01 9.74686801e-01 -2.87893713e-02
5.73299408e-01 -5.29716492e-01 -6.69044912e-01 4.46275949e-01
7.84414291e-01 2.40699738e-01 2.33004272e-01 5.24462104e-01
-2.33645931e-01 -5.32709956e-01 -9.41866219e-01 1.06116128e+00
2.70878911e-01 -1.53172886e+00 -1.14701264e-01 1.20629005e-01
6.65034592e-01 1.74196154e-01 -1.85565591e-01 6.00757241e-01
2.38291860e-01 -1.13907182e+00 6.86301649e-01 7.64848053e-01
1.11540699e+00 -8.59394550e-01 1.28776228e+00 3.95318002e-01
-1.09656405e+00 -8.91456753e-02 -4.78857160e-01 -3.42475511e-02
-1.80685148e-01 2.31805414e-01 -1.07563138e+00 3.58454920e-02
7.51800239e-01 3.14721912e-01 -9.14494038e-01 1.38765895e+00
5.99002838e-01 2.24438772e-01 -4.11988109e-01 -2.39909992e-01
1.82018265e-01 1.74854755e-01 3.17778736e-01 1.79067588e+00
2.85034627e-01 -2.43302360e-01 -2.42949605e-01 6.70984507e-01
-3.04922998e-01 2.10867628e-01 -3.13795656e-01 2.21137628e-01
1.15624115e-01 1.49364555e+00 -1.39700902e+00 -6.51108742e-01
-3.24810684e-01 7.65347064e-01 5.27904570e-01 2.43083704e-02
-8.65230441e-01 2.45668087e-02 2.21213013e-01 2.45347485e-01
3.70614916e-01 2.00082168e-01 -5.64264059e-01 -8.41382802e-01
-4.72146690e-01 -4.86588150e-01 6.13935292e-01 -4.86792594e-01
-1.43180287e+00 1.30118936e-01 1.07893340e-01 -1.10519242e+00
1.70938954e-01 -1.10010338e+00 -5.84439337e-01 7.26049364e-01
-1.41209030e+00 -1.34009743e+00 -5.38475692e-01 5.16252756e-01
3.91710937e-01 -5.69811821e-01 8.54511678e-01 2.71788150e-01
-1.53768763e-01 6.14244342e-01 2.55504727e-01 2.19700009e-01
4.49009001e-01 -1.38567221e+00 1.05280697e-01 4.36747193e-01
-1.11896656e-01 5.08602299e-02 4.99611676e-01 -4.81055200e-01
-4.51330274e-01 -1.31535208e+00 9.98992324e-01 -6.76831961e-01
2.47057319e-01 -3.83907557e-01 -7.78860271e-01 3.09919804e-01
-7.99671561e-02 6.20984316e-01 4.91884738e-01 -4.18508023e-01
-3.08558762e-01 2.27444053e-01 -1.90399587e+00 2.75744796e-01
4.89202887e-01 -3.19894046e-01 1.60830989e-02 4.12433505e-01
1.50580361e-01 -4.67655540e-01 -6.04319334e-01 1.71295196e-01
4.92321402e-01 -1.08068049e+00 1.09095776e+00 -4.38724786e-01
5.09203553e-01 -2.02629834e-01 -1.29001006e-01 -8.59280169e-01
-4.29322779e-01 2.46604830e-01 -1.47167057e-01 9.10911202e-01
-6.21428154e-02 -7.03450739e-01 1.14314914e+00 -4.49653566e-02
5.48538566e-02 -6.58340514e-01 -1.44029999e+00 -8.19539428e-01
5.90448141e-01 -2.50253886e-01 5.01734912e-01 8.84149492e-01
-2.73787737e-01 4.68282923e-02 8.64400044e-02 -2.53408074e-01
6.02472901e-01 -7.60019302e-01 7.95594752e-01 -1.28309083e+00
-1.22121803e-01 -6.63077950e-01 -1.07679367e+00 -2.20510349e-01
-2.75519669e-01 -1.00187886e+00 -1.49042040e-01 -1.32589126e+00
7.27019548e-01 -2.37680525e-01 -2.67027795e-01 1.54597774e-01
-1.41701633e-02 8.30316663e-01 3.72329921e-01 1.57570720e-01
-5.68774045e-01 -2.70871427e-02 1.24980319e+00 -3.04065377e-01
2.47704014e-01 -1.43893018e-01 -2.02153981e-01 6.77461803e-01
6.97452128e-01 -6.42588854e-01 2.96010464e-01 -1.71694949e-01
8.66995007e-02 -1.74033761e-01 9.04795051e-01 -1.44914198e+00
2.30030939e-01 8.24316442e-02 9.92873549e-01 -7.03577399e-01
1.66857973e-01 -6.52546406e-01 -7.86458477e-02 8.53451192e-01
-8.06441307e-02 -1.07586987e-01 5.19901574e-01 6.79539919e-01
1.32423326e-01 -4.15001690e-01 1.52312434e+00 -5.25176823e-01
-3.47819418e-01 2.41341248e-01 -5.15844405e-01 1.36487246e-01
1.04444659e+00 -6.71175301e-01 -4.18514878e-01 4.93147857e-02
-4.80730504e-01 -4.01927941e-02 7.02026904e-01 -2.51045197e-01
1.99136034e-01 -1.31370318e+00 -1.01981950e+00 2.13784128e-02
4.40874249e-01 1.27738982e-01 6.53481260e-02 9.06693816e-01
-1.30675972e+00 5.34231782e-01 -4.48312968e-01 -8.15617561e-01
-1.24671805e+00 5.41412890e-01 6.54982328e-01 -8.03773880e-01
-3.10411096e-01 1.02855217e+00 1.63066350e-02 -6.41264439e-01
4.44011530e-03 -7.16845095e-01 -4.42122102e-01 -5.34798242e-02
6.13974273e-01 8.61113012e-01 4.11254525e-01 -7.02893198e-01
-4.60882574e-01 1.92262486e-01 1.50576964e-01 2.54664361e-01
1.33573425e+00 2.83065468e-01 -6.71513081e-02 4.27238435e-01
1.22614646e+00 -2.96985298e-01 -1.06435621e+00 1.48023954e-02
6.54546451e-03 -3.10106665e-01 -1.11842923e-01 -8.48260343e-01
-9.97388542e-01 7.53836155e-01 9.41843629e-01 1.09438546e-01
8.23938131e-01 -7.49458298e-02 4.34886098e-01 9.51150134e-02
4.32077050e-01 -1.27951705e+00 1.17278516e-01 7.41625309e-01
5.86336970e-01 -1.24325323e+00 2.54101872e-01 2.80090012e-02
-2.54398674e-01 1.25646496e+00 4.93873596e-01 -4.46652323e-01
3.06019843e-01 5.85492969e-01 -1.10477708e-01 -3.68006647e-01
-3.97108644e-01 -2.01311588e-01 -6.07972369e-02 1.12141693e+00
4.36432034e-01 1.15592472e-01 -2.62425870e-01 -7.54596889e-02
1.91984713e-01 5.33518374e-01 5.75293124e-01 8.42074275e-01
-3.55541110e-01 -6.37419641e-01 -5.87803960e-01 8.80877376e-01
-6.96196079e-01 -6.39745295e-02 -2.47981891e-01 1.29185569e+00
5.53985298e-01 3.79202485e-01 4.72796857e-01 4.57748622e-01
1.04489543e-01 -1.12384424e-01 8.74220848e-01 -6.25481904e-01
-8.74754667e-01 -2.05202356e-01 -1.92693874e-01 -4.22255211e-02
-6.09437346e-01 -5.80605805e-01 -1.21103704e+00 -5.49361527e-01
-4.87009794e-01 -6.46782875e-01 5.48896551e-01 9.19400275e-01
-5.47640741e-01 6.08594000e-01 -2.07225263e-01 -1.01400077e+00
-1.14377625e-01 -1.24627399e+00 -8.91934872e-01 4.53198016e-01
2.24189982e-01 -7.10685074e-01 -7.26578906e-02 1.35784909e-01] | [14.774142265319824, -3.2057323455810547] |
f092c586-a32c-4569-812b-2236a7b5e811 | hidden-state-approximation-in-recurrent | 2212.09008 | null | https://arxiv.org/abs/2212.09008v1 | https://arxiv.org/pdf/2212.09008v1.pdf | Hidden State Approximation in Recurrent Neural Networks Using Continuous Particle Filtering | Using historical data to predict future events has many applications in the real world, such as stock price prediction; the robot localization. In the past decades, the Convolutional long short-term memory (LSTM) networks have achieved extraordinary success with sequential data in the related field. However, traditional recurrent neural networks (RNNs) keep the hidden states in a deterministic way. In this paper, we use the particles to approximate the distribution of the latent state and show how it can extend into a more complex form, i.e., the Encoder-Decoder mechanism. With the proposed continuous differentiable scheme, our model is capable of adaptively extracting valuable information and updating the latent state according to the Bayes rule. Our empirical studies demonstrate the effectiveness of our method in the prediction tasks. | ['Dexun Li'] | 2022-12-18 | null | null | null | null | ['stock-price-prediction'] | ['time-series'] | [-2.45812893e-01 -8.36540163e-02 -4.49799180e-01 -2.39857629e-01
-1.10471174e-02 1.99344307e-01 8.01744580e-01 -1.79211333e-01
-5.77781200e-01 8.54687095e-01 3.91635954e-01 -2.72927016e-01
-6.19109422e-02 -1.06067550e+00 -8.34894001e-01 -7.53547728e-01
-1.70190230e-01 3.87340337e-01 3.41920137e-01 -1.61452472e-01
3.96977872e-01 3.58401775e-01 -1.39960730e+00 -1.41288638e-01
5.47919273e-01 1.21121812e+00 5.20389080e-01 2.78516471e-01
-3.69080484e-01 1.25468385e+00 -4.95856315e-01 -1.57835856e-01
2.99462341e-02 -2.18926117e-01 -3.57064396e-01 -4.00639921e-01
-6.26401126e-01 -5.30833185e-01 -6.04339421e-01 1.12240171e+00
7.35306516e-02 2.37393558e-01 3.70406091e-01 -8.87702227e-01
-7.82493412e-01 8.12153697e-01 -3.39163721e-01 2.34660640e-01
-2.46324852e-01 -5.11975512e-02 7.12339520e-01 -6.62958741e-01
2.50330746e-01 1.24309480e+00 6.37479544e-01 4.11646903e-01
-6.73702538e-01 -7.75476873e-01 5.27781904e-01 4.60724413e-01
-1.05857408e+00 -3.17161173e-01 6.72630191e-01 -1.68710053e-01
1.16185760e+00 -4.88118708e-01 7.94412255e-01 1.28606439e+00
7.77780175e-01 9.38724637e-01 8.13199461e-01 -2.33285084e-01
2.86069721e-01 7.86976144e-02 1.03230812e-01 5.02444923e-01
1.98832512e-01 4.67244983e-01 -5.20916283e-01 -9.83090326e-02
8.89599025e-01 6.87086403e-01 1.07394591e-01 -5.00289686e-02
-1.36477709e+00 8.35704982e-01 3.49560887e-01 2.61922032e-01
-7.44209170e-01 4.85276550e-01 3.26573342e-01 6.16143458e-02
6.55904830e-01 6.42824024e-02 -6.07738853e-01 -3.76049310e-01
-8.65234971e-01 1.23524047e-01 8.32280874e-01 6.75940692e-01
4.20093894e-01 4.78154540e-01 -7.31146857e-02 5.26787162e-01
4.80034024e-01 4.60757256e-01 1.11799216e+00 -5.89418352e-01
2.64481217e-01 9.19630378e-02 5.81300437e-01 -8.67550731e-01
-4.28351849e-01 -7.14951634e-01 -1.00281715e+00 3.41757503e-03
5.36669455e-02 -5.34514904e-01 -8.20266187e-01 1.72096694e+00
2.25577112e-02 7.64811754e-01 4.44779634e-01 6.05531633e-01
2.38874450e-01 1.31943321e+00 7.76262507e-02 -4.56646442e-01
9.45468605e-01 -9.45457339e-01 -1.05045402e+00 -3.81598294e-01
1.11071602e-01 -3.48523080e-01 5.10262549e-01 4.45904940e-01
-8.75908852e-01 -6.46625221e-01 -9.87888634e-01 2.48537809e-01
-2.02596828e-01 2.23847598e-01 6.84769154e-01 2.09766418e-01
-8.10782015e-01 9.79242623e-01 -1.46397364e+00 -1.46749258e-01
4.55713496e-02 3.04647863e-01 3.57154936e-01 5.53305447e-01
-1.57427132e+00 1.10640657e+00 7.92953610e-01 5.71151912e-01
-8.29486072e-01 -1.85950771e-01 -4.52510715e-01 3.34251493e-01
1.83900863e-01 -5.00938475e-01 1.48323393e+00 -7.38274395e-01
-2.04294157e+00 -1.15439780e-01 -3.29561830e-01 -1.04700065e+00
2.70755410e-01 -4.34472859e-01 -5.71133196e-01 -2.82542229e-01
-1.54535189e-01 3.35441798e-01 7.24924564e-01 -4.44244146e-01
-9.84302521e-01 -9.79509205e-02 -2.95140952e-01 8.25891569e-02
-4.31545049e-01 -1.74657226e-01 -1.34268612e-01 -7.19688773e-01
2.27576122e-01 -8.02408278e-01 -4.56486374e-01 -4.37978745e-01
-7.29613155e-02 -5.19762993e-01 5.99117815e-01 -5.64093828e-01
1.16070950e+00 -2.20315504e+00 -2.96707992e-02 1.97096467e-02
-1.13283172e-01 1.34453341e-01 3.87195766e-01 5.65179169e-01
1.81776747e-01 -1.40009865e-01 1.27720848e-01 -4.57470298e-01
1.33935079e-01 4.83820021e-01 -9.53209281e-01 2.82766938e-01
-6.28852993e-02 9.55483735e-01 -9.09912825e-01 -3.33202928e-01
1.14702448e-01 3.76164317e-01 4.47853133e-02 9.69771817e-02
-3.97876471e-01 1.72142550e-01 -5.66071570e-01 4.19151820e-02
2.86543608e-01 -4.40908134e-01 1.41234091e-02 2.00522989e-01
-4.40009624e-01 4.28818643e-01 -1.04046190e+00 1.31352389e+00
-4.29796487e-01 5.51161468e-01 -4.51663792e-01 -8.84200335e-01
1.20156169e+00 3.73700380e-01 3.50237787e-01 -6.24717176e-01
1.01211019e-01 2.42984399e-01 -9.19586793e-02 -1.64141655e-01
6.79604828e-01 -5.54659724e-01 6.94683641e-02 5.24642944e-01
-1.68433748e-02 6.68763220e-01 -7.62623847e-02 -1.84883460e-01
6.43688083e-01 2.53419638e-01 3.91737521e-01 1.73401475e-01
1.49033740e-01 -2.58971781e-01 8.67321014e-01 8.65534902e-01
-4.13046032e-02 7.70668834e-02 2.25636780e-01 -6.30576372e-01
-9.48035777e-01 -9.15785849e-01 -6.26851469e-02 6.94303870e-01
3.70905846e-02 -1.07108824e-01 -1.17133416e-01 -2.25667104e-01
-1.15617469e-01 1.02941906e+00 -3.78722817e-01 -4.57519472e-01
-4.75421429e-01 -8.27341020e-01 4.56950068e-01 5.99064469e-01
7.83969343e-01 -1.56757689e+00 -7.91851223e-01 8.56296003e-01
4.36142609e-02 -8.78397107e-01 -4.37544622e-02 3.30785125e-01
-1.07860076e+00 -3.33026379e-01 -6.07340038e-01 -7.48577774e-01
1.22840978e-01 -1.87310994e-01 6.27953231e-01 -2.85601467e-01
5.20708561e-01 -1.32825732e-01 -1.77902743e-01 -4.53989863e-01
-2.67519802e-01 2.23784119e-01 2.38169461e-01 1.99255869e-01
3.90738845e-01 -7.93675601e-01 -5.28658152e-01 1.13043092e-01
-7.66153276e-01 2.93485671e-01 8.22276473e-01 8.05403292e-01
5.54494083e-01 3.97522897e-01 8.54788303e-01 -7.59609342e-01
6.87859654e-01 -6.49599850e-01 -8.67370903e-01 2.65835911e-01
-7.98576653e-01 3.81473869e-01 9.06956017e-01 -6.49753869e-01
-1.38525653e+00 -1.04492806e-01 -2.77221471e-01 -4.91275728e-01
1.45258278e-01 9.03869808e-01 2.11748719e-01 4.41834837e-01
1.71334052e-03 7.97487080e-01 -1.66487813e-01 -5.51819980e-01
3.46827917e-02 4.86018121e-01 4.84210342e-01 -3.10110420e-01
4.54092592e-01 3.26641649e-01 9.94700938e-03 -4.24646378e-01
-1.01374698e+00 3.74011360e-02 -2.38331109e-01 -6.21963758e-03
6.22715831e-01 -8.35269272e-01 -9.14333522e-01 6.85206175e-01
-1.16912234e+00 -2.53012180e-01 -3.29890132e-01 9.50074494e-01
-4.61098850e-01 1.40913740e-01 -1.11768866e+00 -1.13811350e+00
-3.36944282e-01 -8.42880249e-01 7.10615516e-01 4.22457576e-01
2.38199368e-01 -1.19080770e+00 3.26182991e-01 -6.66538119e-01
4.24631774e-01 -8.38252082e-02 7.45884657e-01 -6.92440331e-01
-6.38618588e-01 -3.59285802e-01 3.50756124e-02 3.59626293e-01
5.18852733e-02 -2.46025681e-01 -6.73906982e-01 -5.22373766e-02
4.86780286e-01 5.17544039e-02 1.08626020e+00 4.75142986e-01
9.28851008e-01 -5.09840071e-01 -3.97377253e-01 4.46219265e-01
1.17693448e+00 6.75468028e-01 6.78638458e-01 3.89304072e-01
4.03990567e-01 2.42869169e-01 5.91295183e-01 6.68992996e-01
3.81770372e-01 3.12907577e-01 2.78646052e-01 4.30024236e-01
4.96148795e-01 -8.25251579e-01 6.66371465e-01 1.20737636e+00
5.31186126e-02 -1.64578974e-01 -8.63473296e-01 2.82257318e-01
-2.23813081e+00 -1.26618087e+00 3.21979403e-01 1.97808349e+00
7.66181588e-01 5.83576858e-01 -4.96364415e-01 -2.44687647e-01
7.47674882e-01 3.74524862e-01 -9.11007881e-01 -1.73920423e-01
8.29797462e-02 -6.69986457e-02 6.68516219e-01 3.06539804e-01
-1.05469322e+00 1.00942659e+00 6.50975418e+00 7.89010525e-01
-1.30232012e+00 1.19807422e-01 4.89729851e-01 1.09599553e-01
-1.46188885e-01 9.09348130e-02 -1.34140766e+00 8.39624643e-01
1.42663205e+00 -2.92538851e-01 2.77241826e-01 8.59836698e-01
4.76539969e-01 6.42207712e-02 -6.07839346e-01 9.99957681e-01
-2.78587729e-01 -1.46759582e+00 2.62710303e-01 1.18987449e-01
5.71948707e-01 3.07808608e-01 2.89685071e-01 5.95202327e-01
5.25408566e-01 -7.92333364e-01 9.33524370e-01 1.25683570e+00
2.02317074e-01 -1.04724419e+00 8.79566133e-01 9.95531261e-01
-1.08567178e+00 -3.83954704e-01 -7.46960223e-01 -5.56085229e-01
7.09250629e-01 7.47361243e-01 -7.40785241e-01 3.14744532e-01
5.24367154e-01 1.09323597e+00 -2.39754811e-01 9.85226333e-01
-2.87075758e-01 7.91614056e-01 -4.43159252e-01 -5.37079096e-01
4.30416614e-01 -4.14151549e-01 4.55028862e-01 6.24462664e-01
5.11701286e-01 -1.41693242e-02 1.67124331e-01 7.51810610e-01
1.15600400e-01 -2.34270841e-01 -2.90180951e-01 -1.38305068e-01
3.72634262e-01 7.03294694e-01 -6.90465689e-01 -4.88400131e-01
-2.68752038e-01 7.39229143e-01 3.72256517e-01 3.67491186e-01
-9.45324302e-01 -3.41269702e-01 2.84755737e-01 -2.87465692e-01
6.58333182e-01 -4.70531136e-01 9.60002933e-03 -1.33462572e+00
9.08012036e-03 -2.67737836e-01 5.85240573e-02 -8.02830577e-01
-1.29251337e+00 6.25240028e-01 1.87211614e-02 -1.18537426e+00
-7.31808603e-01 -5.15850484e-01 -6.05492234e-01 8.40713978e-01
-1.68091524e+00 -7.75384247e-01 4.00074005e-01 3.31885815e-01
4.82328147e-01 -2.70567864e-01 6.92665637e-01 4.94960733e-02
-7.28755891e-01 -6.63819164e-02 6.93145931e-01 1.72288164e-01
2.58342564e-01 -9.77287948e-01 7.41507769e-01 8.56135428e-01
3.26039195e-01 8.15422058e-01 7.65617669e-01 -8.97920251e-01
-1.07600522e+00 -1.19715989e+00 8.28894317e-01 1.77495614e-01
9.89718854e-01 -1.52306437e-01 -9.26202714e-01 1.04017425e+00
-1.69904493e-02 -1.79478973e-01 1.24190912e-01 6.97839707e-02
1.36347920e-01 -2.39965335e-01 -6.31535113e-01 5.63250542e-01
6.68821990e-01 -3.10168445e-01 -7.33912528e-01 -2.46329531e-02
9.51038182e-01 -2.92788535e-01 -6.27310038e-01 2.98471451e-01
7.05566108e-01 -7.58595705e-01 5.94796240e-01 -5.30810535e-01
2.08876327e-01 -2.14998901e-01 6.76552802e-02 -1.24524128e+00
-4.66333538e-01 -4.61817592e-01 -6.04203761e-01 9.89513457e-01
3.76802415e-01 -1.17502952e+00 8.41387570e-01 3.83287221e-01
-1.04040541e-01 -8.75404775e-01 -1.02966058e+00 -6.45286977e-01
-1.42202169e-01 -3.89499336e-01 7.59292424e-01 3.25604409e-01
-1.55497596e-01 3.20540637e-01 -8.44306827e-01 3.38715494e-01
4.03324515e-01 2.05517113e-01 3.39658618e-01 -1.25895417e+00
-6.09929264e-01 -3.36808920e-01 -1.86388835e-01 -1.76390159e+00
4.54477906e-01 -3.52159083e-01 1.44236639e-01 -1.51261735e+00
1.94918513e-02 -3.61183971e-01 -8.13947797e-01 1.76313549e-01
2.49552056e-02 -4.74652529e-01 1.05360784e-01 5.72759986e-01
-7.39246190e-01 1.06256974e+00 9.48322296e-01 8.87312889e-02
-1.90531746e-01 5.08219242e-01 -1.98993891e-01 8.06131661e-01
1.03939390e+00 -5.87148070e-01 -3.07611048e-01 -3.48014474e-01
5.45935631e-01 3.44475031e-01 2.57719189e-01 -1.10913455e+00
5.07329524e-01 -1.96819827e-01 4.73423928e-01 -8.80155206e-01
6.13102198e-01 -7.49370873e-01 3.05187792e-01 7.94089377e-01
-3.30235630e-01 2.84692824e-01 -1.11264460e-01 1.02354419e+00
-3.91784161e-01 -4.38975185e-01 3.38161677e-01 -2.34342903e-01
-8.30680788e-01 4.81275946e-01 -5.72448850e-01 -3.52720320e-01
7.75806069e-01 4.25229594e-02 -7.64793605e-02 -6.16942465e-01
-7.01729178e-01 2.13844687e-01 -1.98761541e-02 3.93517166e-01
7.23798692e-01 -1.33092654e+00 -2.87639946e-01 4.73756604e-02
-4.20772821e-01 -2.73126185e-01 1.40224144e-01 6.80369735e-01
-2.52512544e-01 6.81259871e-01 -7.37947598e-02 -4.16951835e-01
-3.74747097e-01 5.95875263e-01 4.80589151e-01 -6.01494193e-01
-8.77915978e-01 2.31548950e-01 8.74255300e-02 -9.20307115e-02
3.95439535e-01 -5.72730720e-01 -5.76772213e-01 -8.49231258e-02
6.90945923e-01 1.93119645e-01 -2.41763994e-01 -2.63882756e-01
1.90694649e-02 3.34410556e-02 -1.36047527e-01 -3.34819794e-01
1.62207663e+00 -3.22873652e-01 -7.16346726e-02 1.32370090e+00
9.03791189e-01 -4.20311600e-01 -1.53138661e+00 -3.59111577e-01
2.68342912e-01 4.83029298e-02 1.90669239e-01 -3.84798348e-01
-8.92585039e-01 8.49106431e-01 4.89549637e-01 3.70942652e-01
6.61544919e-01 -3.80988389e-01 1.23828030e+00 6.64902747e-01
6.93496287e-01 -8.54480386e-01 -3.52018148e-01 8.96581829e-01
2.68344581e-01 -8.26188803e-01 -2.16994151e-01 4.57403958e-01
-6.06520712e-01 1.19842887e+00 2.93696702e-01 -4.60944772e-01
1.04840255e+00 1.71183735e-01 -2.83049732e-01 1.34806275e-01
-1.20806694e+00 -1.78701652e-03 -1.90032080e-01 1.32567614e-01
9.89509225e-02 1.54674962e-01 -3.13386926e-03 8.04720104e-01
-3.35682631e-01 2.38789290e-01 5.40135026e-01 7.90320575e-01
-7.17622399e-01 -9.75937963e-01 -2.04236344e-01 6.55402124e-01
-5.12231827e-01 -1.05037712e-01 3.86370867e-01 5.18428564e-01
-7.69590735e-02 6.25842988e-01 3.47313911e-01 -1.65896654e-01
1.80878833e-01 1.78588748e-01 4.17799614e-02 -5.04210472e-01
-2.63222486e-01 1.71174943e-01 -2.15265259e-01 -3.62018675e-01
-3.90103996e-01 -7.24982202e-01 -1.37440896e+00 -2.82503605e-01
-2.71531373e-01 4.73684400e-01 4.78993773e-01 1.15275490e+00
3.75472963e-01 5.80162287e-01 5.48758090e-01 -7.30713010e-01
-9.80517685e-01 -1.15048659e+00 -7.34860957e-01 -2.36788467e-01
5.19167781e-01 -7.26941705e-01 -2.04939768e-01 -1.92822769e-01] | [7.029153823852539, 3.4575259685516357] |
438f0700-4db4-4152-8430-c1260acd316d | dynamic-texture-recognition-using-pdv-hashing | 2111.12315 | null | https://arxiv.org/abs/2111.12315v2 | https://arxiv.org/pdf/2111.12315v2.pdf | Dynamic Texture Recognition using PDV Hashing and Dictionary Learning on Multi-scale Volume Local Binary Pattern | Spatial-temporal local binary pattern (STLBP) has been widely used in dynamic texture recognition. STLBP often encounters the high-dimension problem as its dimension increases exponentially, so that STLBP could only utilize a small neighborhood. To tackle this problem, we propose a method for dynamic texture recognition using PDV hashing and dictionary learning on multi-scale volume local binary pattern (PHD-MVLBP). Instead of forming very high-dimensional LBP histogram features, it first uses hash functions to map the pixel difference vectors (PDVs) to binary vectors, then forms a dictionary using the derived binary vector, and encodes them using the derived dictionary. In such a way, the PDVs are mapped to feature vectors of the size of dictionary, instead of LBP histograms of very high dimension. Such an encoding scheme could extract the discriminant information from videos in a much larger neighborhood effectively. The experimental results on two widely-used dynamic textures datasets, DynTex++ and UCLA, show the superiority performance of the proposed approach over the state-of-the-art methods. | ['Jiawei Li', 'Heng Yu', 'Jianfeng Ren', 'Ruxin Ding'] | 2021-11-24 | null | null | null | null | ['dynamic-texture-recognition'] | ['computer-vision'] | [-2.00040434e-02 -8.53553951e-01 -3.90561342e-01 -1.16419960e-02
-7.55188465e-01 -6.86862320e-02 1.47686556e-01 9.55497772e-02
-2.66577780e-01 5.25247276e-01 1.44837955e-02 6.39374405e-02
-2.01514930e-01 -1.04812455e+00 -4.44568038e-01 -1.34248078e+00
-1.71523318e-01 2.12030217e-01 9.15929794e-01 -6.32140264e-02
6.21127903e-01 6.41519785e-01 -1.83952975e+00 5.65112889e-01
2.42355421e-01 1.65788627e+00 2.60400236e-01 1.80342585e-01
-2.32872218e-01 7.32630908e-01 -2.83307195e-01 1.07893191e-01
2.13791102e-01 -1.15735114e-01 -4.94348615e-01 6.02362454e-02
3.98844063e-01 -3.06765974e-01 -6.54653549e-01 1.02069938e+00
5.94832599e-01 2.69459844e-01 4.98355329e-01 -9.57516074e-01
-7.20063627e-01 -1.63524058e-02 -8.03589225e-01 5.44025719e-01
2.68196017e-01 -1.51829094e-01 6.30817592e-01 -1.05148697e+00
8.54607582e-01 1.36250126e+00 7.91862845e-01 -1.00050323e-01
-1.11505318e+00 -7.13024795e-01 -4.71939832e-01 7.49463916e-01
-1.95935893e+00 -3.43765855e-01 7.38494515e-01 -2.41343603e-01
8.21191251e-01 3.64788145e-01 6.96159720e-01 8.27956498e-01
6.05365515e-01 4.71830368e-01 1.47123337e+00 -3.58458012e-01
1.32771709e-03 -1.73875257e-01 2.46411622e-01 1.01280153e+00
3.43425246e-03 -2.41263825e-02 -6.16572320e-01 -3.81674767e-01
9.51308429e-01 1.83502525e-01 -3.66645157e-01 -4.86722946e-01
-1.16559172e+00 8.98994327e-01 1.05567232e-01 4.45773959e-01
-1.93398237e-01 -1.47062540e-01 6.54972374e-01 3.80993485e-01
1.06666990e-01 -3.79155874e-01 -1.01184599e-01 -4.17517364e-01
-7.15342283e-01 4.87173423e-02 5.75502276e-01 4.42820132e-01
1.00478756e+00 -2.94336349e-01 -1.08411066e-01 1.17408681e+00
1.36904418e-01 9.06286359e-01 9.51922536e-01 -4.51213062e-01
3.76187176e-01 2.55009562e-01 -7.71839023e-02 -2.15970230e+00
2.84963194e-02 4.46776152e-01 -8.52148175e-01 -1.44976437e-01
3.67927670e-01 7.00022638e-01 -6.99179590e-01 1.18641388e+00
5.74859738e-01 1.27097502e-01 -1.51306465e-01 9.57851410e-01
7.02135801e-01 1.12979114e+00 -4.20803398e-01 -3.97192508e-01
1.34812009e+00 -8.21127951e-01 -7.00368106e-01 2.31632948e-01
2.74449080e-01 -8.37923944e-01 1.10080051e+00 4.71914411e-01
-4.11887169e-01 -7.96846926e-01 -9.78424072e-01 2.39516106e-02
-2.49248341e-01 2.01191008e-02 4.34349597e-01 5.33515096e-01
-7.13454962e-01 5.30882239e-01 -8.54086220e-01 -3.42481196e-01
1.34949505e-01 2.01195598e-01 -5.94796181e-01 -3.20489377e-01
-1.10296488e+00 4.16336119e-01 3.32060903e-01 5.94250672e-02
-6.50064290e-01 8.27595964e-02 -9.66030121e-01 3.46987210e-02
1.71921819e-01 6.02370024e-01 3.57946217e-01 -4.33338821e-01
-1.35434783e+00 6.21144354e-01 -3.60234886e-01 -6.71846047e-02
-3.75711098e-02 3.75116020e-01 -7.23743677e-01 5.07691205e-01
7.43815303e-02 1.66166991e-01 1.32034767e+00 -7.60254443e-01
-5.78877389e-01 -3.43644500e-01 -6.07547283e-01 3.94761600e-02
-6.41388357e-01 -1.67692378e-01 -4.93715525e-01 -7.16685116e-01
6.18494749e-01 -7.67487288e-01 1.67852119e-01 1.12257116e-01
6.48846924e-02 -2.60815293e-01 1.56063712e+00 -7.11609960e-01
1.48337531e+00 -2.69139004e+00 1.29457965e-01 5.99330425e-01
1.85016338e-02 9.46487710e-02 -2.12919429e-01 3.50576520e-01
1.16919912e-01 -2.91297197e-01 9.78229046e-02 3.16184849e-01
-8.88761207e-02 4.51005280e-01 -5.38458407e-01 6.31633759e-01
-2.83234626e-01 3.74806315e-01 -6.80680931e-01 -9.37263191e-01
2.44885430e-01 3.57173413e-01 -2.05196962e-01 -1.70966998e-01
2.38791242e-01 1.07380755e-01 -6.38415217e-01 8.30238998e-01
1.01574886e+00 -8.04195553e-02 2.27540910e-01 -5.35760820e-01
-1.71783760e-01 -2.61856049e-01 -1.45302296e+00 1.25675035e+00
-1.48866385e-01 5.78107595e-01 -2.84404188e-01 -1.26590443e+00
1.37112212e+00 -1.58746883e-01 5.49741864e-01 -1.05097675e+00
2.49043163e-02 4.22855973e-01 -3.57998639e-01 -7.28978693e-01
4.30255145e-01 6.86292946e-02 -2.30615169e-01 1.89671770e-01
-5.49584553e-02 2.17041537e-01 1.18260786e-01 -3.50581169e-01
1.02395642e+00 -9.38671753e-02 1.26526505e-01 -2.35186949e-01
9.19530332e-01 -1.42144084e-01 6.95684493e-01 7.14936972e-01
-3.93095464e-01 3.71875942e-01 5.12727797e-01 -7.59505868e-01
-9.39562082e-01 -8.04234743e-01 -7.10223496e-01 7.95166135e-01
6.22238219e-01 -4.91429448e-01 -3.00801307e-01 -6.22481346e-01
3.36279541e-01 -2.47164652e-01 -5.58820605e-01 -1.53755054e-01
-7.38827586e-01 -5.96548378e-01 4.02249485e-01 1.43770248e-01
9.01171803e-01 -1.17965686e+00 -4.12356228e-01 2.41060019e-01
-2.85275757e-01 -8.88672173e-01 -5.80663323e-01 9.19625349e-03
-7.28359222e-01 -8.36091757e-01 -7.33620703e-01 -1.31616545e+00
3.33031923e-01 5.01301408e-01 3.44546169e-01 -7.01216608e-02
-5.62348545e-01 1.00413822e-01 -6.98485196e-01 5.33584714e-01
2.80664526e-02 -5.24794579e-01 2.28808016e-01 4.20114219e-01
6.00277722e-01 -3.96037191e-01 -5.83748102e-01 7.44235396e-01
-7.39508510e-01 -2.94098526e-01 5.72049499e-01 1.41363990e+00
1.25712228e+00 7.05007076e-01 1.23417020e-01 -2.64738768e-01
2.13107273e-01 -3.25087517e-01 -6.23462975e-01 2.91575193e-01
-1.74739629e-01 3.21026333e-02 7.10463405e-01 -8.64437461e-01
-6.98383510e-01 -1.60112187e-01 -4.77465093e-02 -4.93354768e-01
1.28042772e-01 6.98005557e-01 -1.37481214e-02 -7.76952386e-01
4.51838255e-01 1.04857624e+00 1.66800454e-01 -4.78907049e-01
-8.30331221e-02 9.30867434e-01 3.48702073e-01 -6.82062984e-01
4.29555297e-01 5.73058248e-01 9.57435369e-02 -9.97895122e-01
-1.83414295e-01 -3.34171534e-01 -3.92873347e-01 6.50361180e-02
5.56025088e-01 -6.15651011e-01 -6.17808282e-01 8.84964883e-01
-5.39411008e-01 -1.57739446e-01 -2.06553251e-01 4.22762811e-01
-5.44656277e-01 1.00666034e+00 -8.32661331e-01 -4.98340756e-01
2.89700609e-02 -1.33555651e+00 9.65812325e-01 1.35832936e-01
3.47392887e-01 -6.53214157e-01 8.20667818e-02 1.12436572e-02
4.37897325e-01 2.08981559e-01 1.05426943e+00 -1.47171974e-01
-3.84783119e-01 -2.76794612e-01 -4.17454749e-01 2.15141952e-01
1.98761091e-01 -1.76963627e-01 -4.48545694e-01 -5.69411755e-01
3.21168005e-01 -5.05769968e-01 7.81198859e-01 1.26454815e-01
1.71962821e+00 -3.61411273e-01 -4.20426607e-01 7.68004596e-01
1.59964561e+00 4.00428653e-01 9.31260943e-01 4.37832773e-01
5.68931222e-01 6.97113425e-02 1.16366506e+00 7.08259225e-01
2.64550239e-01 8.79342139e-01 -1.21413618e-01 1.15037888e-01
-1.80350542e-01 -1.92351222e-01 3.50941837e-01 1.22204804e+00
1.33828551e-01 8.33555609e-02 -7.70411551e-01 4.50109184e-01
-1.81420004e+00 -1.00204813e+00 1.13583639e-01 2.18818617e+00
1.08014464e+00 -1.56924650e-02 -2.18184978e-01 3.56744230e-01
9.46141839e-01 5.08596480e-01 -2.21104696e-01 -4.12793070e-01
-3.27783763e-01 3.89580578e-01 4.59615320e-01 7.06885159e-02
-1.24159563e+00 8.23210716e-01 5.30650473e+00 1.86178207e+00
-1.64163160e+00 2.30330095e-01 5.49512327e-01 5.79847217e-01
3.59118193e-01 -1.31027838e-02 -8.87328029e-01 8.95822048e-01
6.18588924e-01 3.13626051e-01 5.24840653e-01 8.41537237e-01
-1.67813048e-01 -3.72821093e-01 -4.39849645e-01 1.65870428e+00
1.45749003e-01 -1.12443471e+00 1.45970345e-01 1.57789782e-01
5.31012774e-01 -3.20433050e-01 3.02328944e-01 2.66840190e-01
-4.55391049e-01 -8.64328861e-01 5.75411975e-01 5.93668640e-01
1.16574430e+00 -7.35148132e-01 8.19335997e-01 9.30735767e-02
-1.66210246e+00 -2.56410360e-01 -1.21603775e+00 2.01365262e-01
-2.34663352e-01 6.08898342e-01 -8.04122165e-02 2.58425564e-01
1.11923349e+00 8.88962090e-01 -5.14565825e-01 1.04714084e+00
3.90123367e-01 4.44673717e-01 -5.42189062e-01 -1.58957109e-01
1.74265474e-01 -2.17068225e-01 3.00712019e-01 1.10629416e+00
6.54706538e-01 2.84378409e-01 4.75153387e-01 7.08585680e-02
3.03704381e-01 6.19948447e-01 -5.12297511e-01 -2.03273296e-01
6.40895605e-01 1.16635442e+00 -7.73937643e-01 -4.16506201e-01
-4.13736552e-01 1.10746109e+00 2.23722801e-01 -3.37554663e-02
-7.63361990e-01 -9.83210325e-01 2.81662464e-01 -6.70815334e-02
9.66796875e-01 -3.33087176e-01 3.24262828e-01 -1.22893965e+00
1.67871803e-01 -1.15238369e+00 4.09694582e-01 -4.95271444e-01
-1.35105109e+00 4.86600548e-01 -3.67066532e-01 -1.60859072e+00
4.08278257e-01 -6.52404547e-01 -1.26924202e-01 6.86591744e-01
-1.16968441e+00 -1.12828732e+00 -4.47876543e-01 1.01360524e+00
1.96971461e-01 -2.87703663e-01 9.04407024e-01 4.09211487e-01
-3.11971664e-01 7.70873606e-01 7.86031425e-01 3.87947448e-02
9.44522083e-01 -5.00182390e-01 -2.99926549e-01 3.72482985e-01
-5.63395657e-02 3.67825568e-01 1.33175060e-01 -7.31054664e-01
-1.82520914e+00 -6.96695566e-01 6.30186617e-01 9.40121710e-02
7.50488162e-01 -2.05804467e-01 -1.07104552e+00 4.21150029e-02
-4.18501407e-01 4.99874443e-01 5.95930040e-01 -4.01953042e-01
-5.28735757e-01 -6.21188462e-01 -1.23393154e+00 1.95909441e-01
8.07904541e-01 -7.10190117e-01 -3.88051838e-01 3.88310492e-01
2.44647384e-01 -6.08150244e-01 -1.17106044e+00 3.84819180e-01
8.36828232e-01 -9.04618084e-01 1.03055441e+00 -2.11305334e-03
2.09168002e-01 -4.86311138e-01 -7.15414405e-01 -6.87009931e-01
-5.88374674e-01 -9.42056403e-02 -3.05581316e-02 1.03581250e+00
-3.48708421e-01 -7.65163898e-01 4.39197034e-01 -3.38124812e-01
3.05174828e-01 -8.85664523e-01 -1.44489682e+00 -8.83965194e-01
-4.40694273e-01 2.80846115e-02 5.62546253e-01 9.92019296e-01
1.88130230e-01 -3.74404490e-01 -7.26140738e-01 8.13735053e-02
6.56796873e-01 5.86915970e-01 5.61920226e-01 -7.73586810e-01
-3.72249961e-01 -1.26033530e-01 -1.16183734e+00 -1.33523095e+00
-2.02067941e-02 -6.70896351e-01 2.19946489e-01 -7.53377080e-01
3.59892219e-01 -9.12631214e-01 -5.49451232e-01 6.56687021e-01
1.78945780e-01 6.77217782e-01 -2.69852459e-01 5.70926845e-01
-6.21764660e-01 6.56809926e-01 1.35609221e+00 -3.63733858e-01
1.54185360e-02 -6.72366202e-01 -2.81867106e-02 3.21585596e-01
3.40014517e-01 -4.79853511e-01 -5.58522530e-02 -7.08647594e-02
-3.43098849e-01 3.90123963e-01 2.25927725e-01 -1.12921739e+00
1.40294641e-01 -3.79518837e-01 5.96028924e-01 -6.22211695e-01
3.05473298e-01 -6.34789050e-01 9.25930515e-02 6.59139156e-01
2.41696730e-01 -3.80625427e-02 1.32825315e-01 6.28815830e-01
-6.17755890e-01 -5.72637804e-02 8.73515248e-01 2.31679112e-01
-1.22567964e+00 4.40548629e-01 -3.55809033e-01 -4.73547041e-01
9.59843695e-01 -4.46001142e-01 -3.99060756e-01 7.00174225e-03
-7.24067330e-01 -2.22440138e-01 6.20625973e-01 3.10910016e-01
1.06414139e+00 -1.78023458e+00 -2.48115063e-01 6.22348607e-01
2.13863283e-01 -5.57641149e-01 5.29264212e-01 9.72565472e-01
-9.03879225e-01 2.42467448e-01 -6.72490954e-01 -8.47788334e-01
-1.31452751e+00 6.61181450e-01 -1.14764576e-03 -2.01465979e-01
-9.26880419e-01 6.97140276e-01 5.03973402e-02 -7.94846937e-02
-4.98218611e-02 1.46991266e-02 -2.56893814e-01 6.96522370e-02
5.72142899e-01 1.87434107e-01 1.13974079e-01 -1.01092911e+00
-4.09587979e-01 1.19564188e+00 -7.93785751e-02 3.68743874e-02
1.11237884e+00 -2.83555984e-01 -6.07011497e-01 2.78112054e-01
1.62387514e+00 2.00939655e-01 -8.76225293e-01 -4.60600287e-01
-3.29570651e-01 -1.26011789e+00 -3.98970093e-04 -1.34909704e-01
-1.02475071e+00 8.15151632e-01 1.13666630e+00 4.09555472e-02
1.21877539e+00 -2.72751600e-01 1.35245514e+00 5.34915388e-01
9.20816839e-01 -9.89631534e-01 3.08659822e-01 7.15671182e-01
7.08412051e-01 -9.38735306e-01 -1.84751302e-02 -4.28809911e-01
-4.58231539e-01 1.38470209e+00 4.26038772e-01 -3.55739862e-01
8.99381042e-01 1.21209994e-01 -1.01472192e-01 -2.95474995e-02
-1.89958006e-01 2.13164032e-01 1.38435349e-01 5.50788701e-01
-1.95943177e-01 -2.29407549e-02 -6.36753261e-01 3.40095133e-01
6.96168542e-02 -1.53033912e-01 2.13381261e-01 1.24065411e+00
-6.73538148e-01 -1.29165137e+00 -6.79062545e-01 5.36464870e-01
-2.01665655e-01 5.41657396e-02 3.12989116e-01 4.81691062e-01
2.76290119e-01 4.93153304e-01 2.42931306e-01 -8.35343957e-01
-1.62104979e-01 -2.23297000e-01 5.35151780e-01 -1.13237351e-01
1.44446701e-01 2.30891958e-01 -3.35153461e-01 -8.07949901e-01
-2.11998582e-01 -6.31338656e-01 -1.09839606e+00 -3.82841408e-01
-2.72613972e-01 2.22719267e-01 3.47221404e-01 5.69869101e-01
2.03952491e-01 -2.11436719e-01 9.10262764e-01 -8.19902837e-01
-5.02357662e-01 -5.88152051e-01 -1.11023951e+00 6.08322620e-01
2.61740386e-01 -1.07082963e+00 -2.80697256e-01 -7.93071538e-02] | [10.562355995178223, -0.30651214718818665] |
582908ad-a758-4bc9-bbdd-c0cd1871d610 | a-versatile-scene-model-with-differentiable | 1602.03725 | null | http://arxiv.org/abs/1602.03725v1 | http://arxiv.org/pdf/1602.03725v1.pdf | A Versatile Scene Model with Differentiable Visibility Applied to Generative Pose Estimation | Generative reconstruction methods compute the 3D configuration (such as pose
and/or geometry) of a shape by optimizing the overlap of the projected 3D shape
model with images. Proper handling of occlusions is a big challenge, since the
visibility function that indicates if a surface point is seen from a camera can
often not be formulated in closed form, and is in general discrete and
non-differentiable at occlusion boundaries. We present a new scene
representation that enables an analytically differentiable closed-form
formulation of surface visibility. In contrast to previous methods, this yields
smooth, analytically differentiable, and efficient to optimize pose similarity
energies with rigorous occlusion handling, fewer local minima, and
experimentally verified improved convergence of numerical optimization. The
underlying idea is a new image formation model that represents opaque objects
by a translucent medium with a smooth Gaussian density distribution which turns
visibility into a smooth phenomenon. We demonstrate the advantages of our
versatile scene model in several generative pose estimation problems, namely
marker-less multi-object pose estimation, marker-less human motion capture with
few cameras, and image-based 3D geometry estimation. | ['Christian Theobalt', 'Hans-Peter Seidel', 'Nadia Robertini', 'Helge Rhodin', 'Christian Richardt'] | 2016-02-11 | a-versatile-scene-model-with-differentiable-1 | http://openaccess.thecvf.com/content_iccv_2015/html/Rhodin_A_Versatile_Scene_ICCV_2015_paper.html | http://openaccess.thecvf.com/content_iccv_2015/papers/Rhodin_A_Versatile_Scene_ICCV_2015_paper.pdf | iccv-2015-12 | ['occlusion-handling'] | ['computer-vision'] | [ 2.56012797e-01 -3.75967077e-03 3.24913770e-01 -8.70492458e-02
-6.10346079e-01 -6.36551380e-01 4.82119888e-01 -1.80189759e-01
-9.28157195e-02 5.41807473e-01 -2.70056933e-01 7.59434998e-02
-1.95275843e-02 -7.82907963e-01 -8.65602851e-01 -8.50611210e-01
2.70987064e-01 9.17180359e-01 2.42806047e-01 1.35062099e-01
1.51492789e-01 9.25653279e-01 -1.68261015e+00 -5.99308729e-01
8.20793152e-01 9.89803910e-01 2.53876209e-01 8.70221496e-01
1.00286379e-01 2.50640810e-01 -5.24697125e-01 -5.23903489e-01
3.29499036e-01 -2.66985625e-01 -3.69573444e-01 6.71619356e-01
7.15675652e-01 -2.48787239e-01 -1.55167691e-02 1.07019222e+00
3.05866808e-01 7.99690560e-02 9.86684144e-01 -1.06710124e+00
-5.24095714e-01 -5.73855758e-01 -5.93831480e-01 -3.03302377e-01
6.16516471e-01 -1.49390012e-01 6.52228951e-01 -1.02943408e+00
7.86041677e-01 1.23367476e+00 6.68681562e-01 2.02711239e-01
-1.27815628e+00 1.92197874e-01 -1.76390827e-01 -2.68916428e-01
-1.63967669e+00 -2.35196203e-01 9.88962114e-01 -6.01400137e-01
5.26717305e-01 7.31696010e-01 8.47152054e-01 6.10164583e-01
4.37156439e-01 4.03137773e-01 1.07516813e+00 -4.63375300e-01
4.61417586e-02 3.21657807e-01 -4.10957932e-02 9.85773742e-01
4.66486126e-01 4.04210873e-02 -2.37462327e-01 -3.66230547e-01
1.20992184e+00 1.43083125e-01 -5.41701674e-01 -1.06376910e+00
-1.12802470e+00 6.48401618e-01 1.20883659e-01 -1.74688458e-01
-2.73849040e-01 1.52012601e-01 -4.05393749e-01 -1.13278925e-01
6.43052280e-01 3.67320478e-01 1.73437387e-01 1.75711121e-02
-5.09374380e-01 3.56984973e-01 9.86753106e-01 1.11612070e+00
1.11387801e+00 2.24120006e-01 1.37853727e-01 5.60132325e-01
6.28214717e-01 1.16077125e+00 -1.67148799e-01 -1.29458487e+00
-1.29040154e-02 5.37173867e-01 3.24597687e-01 -1.25753486e+00
-1.26333967e-01 -5.19710958e-01 -5.76461852e-01 6.75274849e-01
4.19296026e-01 1.47922695e-01 -5.61435938e-01 1.47027719e+00
9.19736922e-01 2.47871745e-02 -1.70128644e-01 1.02381277e+00
8.37607563e-01 5.82433164e-01 -5.86093426e-01 -4.11696017e-01
1.35493231e+00 -5.43558121e-01 -6.83963537e-01 -1.19867623e-01
3.51005346e-02 -1.02664614e+00 8.14471066e-01 3.58922631e-01
-1.45877326e+00 -2.82766640e-01 -9.63775337e-01 -2.76452273e-01
1.25965759e-01 6.58620000e-02 5.15444100e-01 7.85234213e-01
-9.48249876e-01 3.96679580e-01 -7.55074024e-01 -3.43577601e-02
-1.06912125e-02 2.32973292e-01 -2.64927834e-01 -5.87792806e-02
-4.11289066e-01 9.79257166e-01 -2.57461697e-01 1.48681939e-01
-8.27061355e-01 -6.27896905e-01 -1.12367523e+00 -2.42774725e-01
4.12532687e-01 -1.07964540e+00 8.64418507e-01 -5.61872900e-01
-1.74809027e+00 1.30051613e+00 -4.37236190e-01 2.65683413e-01
6.69705868e-01 -2.62084901e-01 1.12608448e-01 2.89145142e-01
-1.62608474e-01 1.35714203e-01 1.12772346e+00 -1.72684693e+00
1.72342524e-01 -5.23593843e-01 6.38880581e-02 5.55239975e-01
1.86409861e-01 -1.24291010e-01 -5.10546923e-01 -2.49449044e-01
5.21251798e-01 -8.81349027e-01 -2.56141961e-01 4.33269769e-01
-3.20857882e-01 1.67545602e-01 7.66891897e-01 -5.07246375e-01
7.65168190e-01 -2.02944970e+00 3.97006541e-01 2.30932117e-01
3.93776566e-01 -1.43155366e-01 1.76771656e-01 2.48356998e-01
2.84835547e-01 -2.63315201e-01 -2.89170980e-01 -6.76479578e-01
-4.48392555e-02 2.95518011e-01 -1.28021717e-01 1.04830134e+00
9.65087488e-02 8.05639923e-01 -7.09328234e-01 -5.47071040e-01
5.13766766e-01 9.03842211e-01 -6.01820648e-01 3.93606514e-01
-6.78380877e-02 5.78725636e-01 -5.30578971e-01 8.40008914e-01
9.68193710e-01 -2.98870206e-01 -2.35021293e-01 -2.60555148e-01
-3.21896046e-01 -2.31720954e-01 -1.49132955e+00 1.49153674e+00
-4.60754454e-01 5.14085114e-01 3.77560645e-01 -5.66293657e-01
1.09068394e+00 2.15976909e-01 4.93061125e-01 -1.45603836e-01
2.69459337e-01 1.65044844e-01 -6.31070435e-01 -4.08630997e-01
5.44781148e-01 -2.09358916e-01 1.73818648e-01 1.90643594e-01
-3.00432056e-01 -8.54199052e-01 -2.77054757e-01 -1.06101975e-01
6.05656207e-01 2.79814988e-01 3.44572961e-01 -2.40267202e-01
4.57480937e-01 -9.10196453e-02 4.06087816e-01 4.28497791e-01
2.55697608e-01 1.21457410e+00 8.50012749e-02 -2.72672117e-01
-1.04871964e+00 -1.32028341e+00 -4.58890945e-01 1.85968414e-01
7.55513310e-01 -9.29096565e-02 -7.79782355e-01 1.41887560e-01
-2.89044958e-02 4.85278875e-01 -3.61321896e-01 2.06240993e-02
-6.70563281e-01 -5.60617685e-01 -2.08335593e-01 -1.64315000e-03
2.38032028e-01 -5.85914075e-01 -7.86274254e-01 1.36414804e-02
-1.13939032e-01 -1.18365729e+00 -5.99763334e-01 -2.91072071e-01
-9.42222297e-01 -1.21630073e+00 -8.57268751e-01 -6.29721105e-01
1.01024354e+00 4.03287977e-01 1.30153525e+00 3.69633362e-02
-4.41719472e-01 8.74143004e-01 8.32348987e-02 -1.64369598e-01
-4.22815800e-01 -6.23447716e-01 -3.27330381e-02 3.39311242e-01
-2.90774614e-01 -6.38977826e-01 -7.07117617e-01 6.19772911e-01
-6.38506353e-01 1.04574136e-01 1.60359755e-01 5.03071785e-01
1.04922152e+00 -2.55180418e-01 -2.59811044e-01 -4.00062621e-01
1.89539149e-01 -8.27756077e-02 -1.00606203e+00 3.39357853e-01
-1.98252589e-01 -5.39911427e-02 5.95683008e-02 -4.60051507e-01
-1.10892582e+00 3.18333358e-01 1.19982921e-01 -8.07110012e-01
-5.73599115e-02 4.32026153e-03 -2.85983801e-01 -5.68181157e-01
5.41801035e-01 3.58552396e-01 2.95574009e-01 -4.98090208e-01
2.79125750e-01 2.81564981e-01 7.39144921e-01 -7.11412549e-01
1.09389317e+00 9.58384991e-01 5.61125875e-01 -1.35336161e+00
-5.05848408e-01 -5.36465049e-01 -7.03262031e-01 -5.78418195e-01
9.77741182e-01 -7.12008536e-01 -1.05205202e+00 4.40596282e-01
-1.43236363e+00 2.19700448e-02 -4.29174453e-01 5.87745428e-01
-8.08827817e-01 7.21442699e-01 -2.90378541e-01 -1.24488962e+00
-1.02100626e-01 -1.28063858e+00 1.56369519e+00 9.44223329e-02
-4.98280339e-02 -1.19275141e+00 1.03661492e-01 4.61498857e-01
2.09253088e-01 7.00736046e-01 5.83963692e-01 4.30378199e-01
-1.16566277e+00 -3.30202669e-01 9.72919837e-02 1.99739471e-01
-4.11062427e-02 2.69734591e-01 -9.70631599e-01 -3.03858072e-01
6.01428747e-01 4.64936672e-03 2.59652764e-01 7.21487761e-01
7.44228125e-01 -9.15107355e-02 -3.35923523e-01 1.09416795e+00
1.53257203e+00 -8.60208869e-02 6.23270035e-01 -8.89110379e-03
8.53645205e-01 5.43060541e-01 5.94997644e-01 5.29462159e-01
3.15489113e-01 1.05968654e+00 7.90674031e-01 -1.22893132e-01
-2.88836151e-01 -1.97053198e-02 3.24220508e-01 8.29646051e-01
-5.17435491e-01 -2.36277267e-01 -4.15659845e-01 1.66605130e-01
-1.48256886e+00 -6.40439510e-01 -6.88103378e-01 2.70237756e+00
4.54139233e-01 -4.30602640e-01 -1.34575412e-01 -2.17091173e-01
6.89635456e-01 -1.35454118e-01 -4.13792312e-01 -1.09012499e-01
-3.18008214e-01 -5.28292805e-02 3.64629418e-01 1.00132322e+00
-8.19200277e-01 4.91453320e-01 7.02759171e+00 6.00397289e-01
-8.42746019e-01 1.62301645e-01 2.80116707e-01 -3.67291495e-02
-8.17618668e-01 6.92895725e-02 -8.34489524e-01 2.13101432e-01
2.78667241e-01 -7.25912228e-02 1.94889992e-01 8.56252551e-01
5.30004837e-02 -3.21364671e-01 -1.04876077e+00 1.33976173e+00
4.79245245e-01 -1.16168642e+00 -4.97031957e-02 3.74587387e-01
6.98776603e-01 -2.23065466e-01 1.23410217e-01 -5.23845136e-01
-1.32323548e-01 -7.38078237e-01 1.06993365e+00 6.95106745e-01
8.14270139e-01 -5.64359248e-01 1.39115080e-01 4.57327962e-01
-1.05632210e+00 5.73906124e-01 -4.58582640e-01 1.22456260e-01
6.70544922e-01 8.20263803e-01 -5.47290742e-01 4.84854609e-01
3.49534839e-01 3.60374242e-01 -1.32137820e-01 1.31582785e+00
-1.76411703e-01 -2.15231162e-02 -5.49021959e-01 -1.01122975e-01
-2.89928615e-01 -9.68297005e-01 1.24165761e+00 7.40096688e-01
5.81470966e-01 1.15074262e-01 2.49465033e-01 1.30068302e+00
3.20503652e-01 -1.81028880e-02 -7.29962528e-01 4.72544402e-01
7.54538029e-02 1.26352262e+00 -7.43851602e-01 -7.61223491e-03
-2.92107582e-01 1.02370751e+00 9.56607237e-02 5.50101995e-01
-8.88102531e-01 8.16548541e-02 6.25279963e-01 4.81322974e-01
9.50755030e-02 -5.46447217e-01 -1.89459324e-01 -1.19242811e+00
3.56599957e-01 -3.52359593e-01 -1.66253269e-01 -1.03586006e+00
-9.99139190e-01 5.92336655e-01 2.23181039e-01 -1.37035739e+00
-2.13159159e-01 -7.37619102e-01 -4.01078403e-01 9.70969617e-01
-1.40993810e+00 -9.57469106e-01 -4.99319315e-01 6.00444496e-01
2.98586935e-01 3.03303570e-01 7.70973563e-01 1.41549036e-01
-2.21794963e-01 1.96781978e-01 1.49670299e-02 -4.01802629e-01
2.06283107e-01 -1.22095537e+00 1.13602187e-02 8.21313620e-01
1.22362629e-01 5.35813868e-01 8.75333011e-01 -5.88299692e-01
-1.90817904e+00 -7.54017115e-01 5.08293390e-01 -8.41750562e-01
1.85588241e-01 -4.27930355e-01 -7.50352025e-01 5.08407295e-01
-1.10262088e-01 5.26451133e-02 2.94860065e-01 -2.85578609e-01
1.34390919e-02 1.51128769e-01 -1.17018032e+00 4.50617164e-01
1.05782914e+00 -3.36285681e-01 -4.00567025e-01 5.36266446e-01
4.78011638e-01 -9.01756227e-01 -8.34799230e-01 2.60161489e-01
4.73587304e-01 -9.33362901e-01 1.34557283e+00 8.26381221e-02
3.12184636e-02 -4.86834228e-01 -2.16265306e-01 -1.02805936e+00
-2.28814766e-01 -1.16041458e+00 -2.67131627e-01 7.38662183e-01
3.86736542e-02 -7.77912021e-01 7.89259017e-01 5.62096655e-01
-4.83312249e-01 -6.89501345e-01 -1.04091251e+00 -9.16382551e-01
-4.05817658e-01 -3.39154482e-01 3.10901403e-01 6.07424796e-01
-5.19807577e-01 6.06801547e-02 -4.12627399e-01 5.56238592e-01
1.25074029e+00 5.16754091e-01 9.72547054e-01 -1.34351957e+00
-3.69319946e-01 -1.93993822e-01 -5.57832241e-01 -1.56386936e+00
-7.01391995e-02 -7.45659947e-01 1.04511470e-01 -1.34155202e+00
2.41737142e-01 -4.91107553e-01 6.10530615e-01 -2.90974140e-01
8.79286528e-02 3.20095837e-01 -1.15158923e-01 3.14082861e-01
-2.95686066e-01 6.91396236e-01 1.67720771e+00 1.81489304e-01
-7.71887302e-02 2.61920869e-01 -2.48595476e-01 9.28555250e-01
1.28566310e-01 -2.82305688e-01 -3.71538281e-01 -5.29946864e-01
4.19478536e-01 2.21774429e-01 7.22572446e-01 -7.74302602e-01
2.75070295e-02 -2.42681414e-01 1.61503956e-01 -7.69805610e-01
9.80455101e-01 -9.92379129e-01 8.26661766e-01 1.43886179e-01
2.92043924e-01 1.06231766e-02 -2.31590196e-01 5.54788768e-01
1.98637638e-02 -5.16343713e-01 8.11072946e-01 -2.77396947e-01
-2.93230832e-01 5.30594587e-01 2.87609491e-02 -2.05744207e-02
1.08815372e+00 -8.54693890e-01 -5.98981716e-02 -4.93997097e-01
-7.82791018e-01 -2.75407940e-01 1.01108718e+00 4.18840069e-03
9.48616922e-01 -1.43126380e+00 -6.63100421e-01 4.32963312e-01
-7.86805600e-02 4.49257314e-01 1.19686991e-01 9.03933048e-01
-9.66196656e-01 3.86283584e-02 2.70649314e-01 -1.16630077e+00
-1.43938112e+00 3.60049039e-01 5.49747229e-01 2.38952398e-01
-7.75849700e-01 8.78457069e-01 5.97702384e-01 -4.13412333e-01
-2.43691448e-03 -3.11932147e-01 2.56568730e-01 -3.30023974e-01
3.31106901e-01 5.31057894e-01 -1.83109809e-02 -9.61351573e-01
-2.11630657e-01 1.31369555e+00 5.34867942e-01 -1.08658962e-01
1.26072538e+00 -3.06204200e-01 -2.99556732e-01 4.67895925e-01
1.12936342e+00 3.60765576e-01 -1.55714500e+00 -5.80724329e-02
-7.90396988e-01 -9.47482109e-01 1.58365059e-03 -2.03888848e-01
-1.02907538e+00 6.56075180e-01 2.81398743e-01 2.42794722e-01
6.66211605e-01 3.10098082e-01 4.97221708e-01 1.53727248e-01
5.52352130e-01 -7.13543534e-01 1.74333960e-01 3.10940832e-01
1.27300882e+00 -9.77468312e-01 2.25759283e-01 -1.13366771e+00
-2.11401492e-01 9.15731013e-01 2.95048296e-01 -2.17326134e-01
6.78632557e-01 3.50600660e-01 -8.76632854e-02 -4.27907705e-01
-1.22638531e-01 -1.90532096e-02 6.43346727e-01 7.48891592e-01
1.21848106e-01 -2.52072862e-03 4.36313860e-02 -1.83135599e-01
-3.33896816e-01 -5.87221086e-01 4.96228546e-01 7.20730722e-01
-5.17048478e-01 -7.58742094e-01 -8.05593550e-01 -1.26418233e-01
-1.59566998e-01 1.76079273e-01 -9.83884186e-02 6.53659344e-01
-1.32643133e-01 6.03572905e-01 2.04567328e-01 2.39497364e-01
4.59330082e-01 -3.08085650e-01 1.01998258e+00 -7.27574527e-01
6.36593774e-02 3.85061204e-01 -1.00416131e-01 -5.97309351e-01
-5.33874989e-01 -8.03340375e-01 -1.05084908e+00 -1.37139201e-01
-6.67633116e-01 -4.73551974e-02 8.70392203e-01 7.08682656e-01
1.21543847e-01 1.53094366e-01 7.49041796e-01 -1.19921923e+00
-4.00201321e-01 -4.46923226e-01 -7.41799355e-01 5.11426568e-01
4.56162423e-01 -1.09325743e+00 -6.71476424e-01 1.87828746e-02] | [9.456892967224121, -2.961338520050049] |
68ebd322-e759-4916-8462-88734bc46f99 | faking-fake-news-for-real-fake-news-detection | 2203.05386 | null | https://arxiv.org/abs/2203.05386v2 | https://arxiv.org/pdf/2203.05386v2.pdf | Faking Fake News for Real Fake News Detection: Propaganda-loaded Training Data Generation | Despite recent advances in detecting fake news generated by neural models, their results are not readily applicable to effective detection of human-written disinformation. What limits the successful transfer between them is the sizable gap between machine-generated fake news and human-authored ones, including the notable differences in terms of style and underlying intent. With this in mind, we propose a novel framework for generating training examples that are informed by the known styles and strategies of human-authored propaganda. Specifically, we perform self-critical sequence training guided by natural language inference to ensure the validity of the generated articles, while also incorporating propaganda techniques, such as appeal to authority and loaded language. In particular, we create a new training dataset, PropaNews, with 2,256 examples, which we release for future use. Our experimental results show that fake news detectors trained on PropaNews are better at detecting human-written disinformation by 3.62 - 7.69% F1 score on two public datasets. | ['Heng Ji', 'Yejin Choi', 'Preslav Nakov', 'Kathleen McKeown', 'Kung-Hsiang Huang'] | 2022-03-10 | null | null | null | null | ['news-generation'] | ['natural-language-processing'] | [ 9.22466218e-02 3.99102926e-01 -4.06409442e-01 -1.42693684e-01
-5.57354927e-01 -7.88293362e-01 1.16437697e+00 8.59130099e-02
-2.07409739e-01 7.33772337e-01 6.54422462e-01 -5.65645278e-01
5.73121309e-01 -7.81770349e-01 -7.94468224e-01 -2.47476995e-01
2.18909591e-01 3.83726269e-01 1.77753732e-01 -6.44342124e-01
5.72915077e-01 1.33239448e-01 -1.00276601e+00 6.65327489e-01
9.12555993e-01 4.44178939e-01 -4.92694646e-01 4.71336722e-01
5.24599664e-02 1.37426221e+00 -1.18137538e+00 -9.62228000e-01
-4.15784605e-02 -9.15013134e-01 -7.14854360e-01 9.57177300e-03
5.32748699e-01 -7.78542280e-01 -5.32095671e-01 1.16282129e+00
2.58180916e-01 -5.11262417e-01 8.29652369e-01 -1.07311678e+00
-1.05783975e+00 9.26897883e-01 -3.98398578e-01 3.29728365e-01
5.11388719e-01 3.02107096e-01 7.77520359e-01 -7.05364287e-01
9.31153178e-01 1.54922795e+00 6.86032295e-01 8.31512809e-01
-8.96978974e-01 -7.21410513e-01 -1.74339846e-01 1.94998048e-02
-8.76718044e-01 -5.91963410e-01 8.11787486e-01 -6.91558421e-01
5.38619280e-01 2.37209037e-01 5.67822158e-01 2.04812503e+00
4.67244238e-01 8.62441659e-01 1.34241593e+00 -4.61504996e-01
2.02352718e-01 3.90308410e-01 9.06760991e-02 7.93465555e-01
7.05534220e-01 1.55237496e-01 -5.01014113e-01 -6.40645981e-01
3.89314741e-01 -5.18808126e-01 -1.92742795e-01 1.44881010e-01
-1.23832738e+00 1.40991938e+00 4.11907107e-01 2.97049642e-01
-2.42647663e-01 1.07141091e-02 6.62913322e-01 2.40625888e-01
6.98155642e-01 8.81741464e-01 1.84147552e-01 -2.58153789e-02
-8.70135188e-01 5.33057094e-01 8.97330761e-01 7.50683546e-01
1.02299079e-01 2.22028136e-01 -2.75469542e-01 3.82700086e-01
8.03947076e-02 6.43694162e-01 8.11184525e-01 -4.84201759e-01
6.90651655e-01 3.15980196e-01 5.26443422e-01 -1.86113799e+00
-6.54518381e-02 -4.36286449e-01 -5.93345881e-01 -2.53478497e-01
4.13062215e-01 -1.77076459e-01 -5.56037426e-01 1.40697181e+00
4.96641435e-02 -1.98037490e-01 -3.74476798e-02 9.45936024e-01
7.91801572e-01 6.48657858e-01 1.68850739e-02 -6.92349002e-02
1.46298385e+00 -9.87087369e-01 -8.26210916e-01 -6.47072971e-01
8.12345743e-01 -7.31619596e-01 8.68891895e-01 1.61663473e-01
-7.97171652e-01 -4.00481932e-02 -1.14479971e+00 1.36141092e-01
-3.86111408e-01 1.04876623e-01 4.46884334e-01 7.31292605e-01
-4.96311933e-01 4.03373599e-01 -5.19458413e-01 -1.98168516e-01
7.03508437e-01 -4.23058480e-01 -1.29363313e-01 2.07270861e-01
-1.53427434e+00 1.26727855e+00 3.66320074e-01 3.34576033e-02
-1.09538531e+00 -9.96218324e-02 -6.91645145e-01 -1.98891386e-01
3.32219154e-01 -3.47082436e-01 1.39566660e+00 -1.31136441e+00
-1.16354239e+00 1.04599369e+00 1.29965886e-01 -7.78422415e-01
1.19346225e+00 -2.42559239e-01 -6.34848416e-01 2.81453580e-01
4.19007748e-01 9.03880447e-02 1.28679240e+00 -1.30162776e+00
-8.69255066e-02 -1.83877036e-01 -1.22451097e-01 -1.75029069e-01
-3.81888717e-01 4.09072578e-01 2.23710224e-01 -1.10419285e+00
-3.77148800e-02 -8.67478788e-01 1.14262529e-01 -2.46252641e-01
-1.04779160e+00 2.70658284e-02 7.58683085e-01 -9.71607447e-01
1.24432433e+00 -1.84880555e+00 -3.67343664e-01 3.15919369e-02
5.27394831e-01 5.79318047e-01 -6.66909069e-02 7.59642065e-01
2.74818778e-01 4.41509426e-01 -6.85428679e-02 6.08756579e-02
-3.59122711e-03 -2.40363240e-01 -9.05860841e-01 7.77930439e-01
3.20671141e-01 1.06540787e+00 -1.20559716e+00 -4.21442300e-01
-3.98696601e-01 1.11431323e-01 -4.66764838e-01 5.05363569e-02
-3.52211028e-01 2.43303195e-01 -8.14440608e-01 4.04857755e-01
5.46409726e-01 -4.02006030e-01 2.20356196e-01 3.81335728e-02
3.27679142e-02 9.23886061e-01 -3.40843379e-01 7.54546523e-01
9.64047271e-04 1.00135815e+00 1.11309229e-04 -7.75381804e-01
7.66596675e-01 2.05209985e-01 -3.19179565e-01 -5.53735554e-01
4.66041088e-01 3.65448147e-01 -5.57109043e-02 -6.87215149e-01
6.94687247e-01 -2.07286045e-01 -2.51446128e-01 8.66830885e-01
-2.71522373e-01 -1.07921474e-01 -7.89826363e-02 5.80334067e-01
1.01054239e+00 -3.12060505e-01 5.10743976e-01 -1.13894880e-01
2.38257766e-01 4.73137885e-01 1.01904199e-01 1.20392656e+00
-4.02603000e-01 3.11676204e-01 7.29249656e-01 -5.38547873e-01
-1.27997243e+00 -6.77061737e-01 1.61903083e-01 8.34729791e-01
1.55859202e-01 -2.91196167e-01 -8.02940845e-01 -1.09223425e+00
-7.43919471e-03 1.10382354e+00 -6.71921670e-01 -3.01758617e-01
-7.57865250e-01 -6.08173966e-01 1.17092288e+00 2.07612868e-02
5.62446475e-01 -1.09824359e+00 -6.54903173e-01 3.91485155e-01
-4.07428324e-01 -1.25611889e+00 -2.92768538e-01 -4.63522047e-01
-4.91363227e-01 -1.06938505e+00 -4.65451807e-01 -7.41424620e-01
6.89424276e-01 3.78659785e-01 9.69899058e-01 4.67540056e-01
-1.37554510e-02 -2.82325327e-01 -4.10063148e-01 -3.57702196e-01
-1.38614726e+00 -4.87676710e-02 2.76487637e-02 -2.38931790e-01
2.56797343e-01 4.57531661e-02 -2.61326194e-01 1.89485192e-01
-9.69602227e-01 3.52212578e-01 6.09971881e-01 1.10523582e+00
-5.51993191e-01 -2.25423455e-01 6.69573367e-01 -1.20984721e+00
1.10524464e+00 -7.26680100e-01 -2.89725125e-01 3.68085615e-02
-5.54538250e-01 -8.35375115e-03 8.76698196e-01 -6.17282093e-01
-1.01222324e+00 -4.98208523e-01 -1.29572958e-01 1.65237889e-01
-6.99763745e-02 4.64289457e-01 5.09061992e-01 1.59764901e-01
1.29377210e+00 4.60039526e-01 2.12517485e-01 -2.40151674e-01
3.79733533e-01 1.09312892e+00 4.49573189e-01 -3.30114543e-01
1.06332469e+00 5.66197991e-01 -7.27511585e-01 -7.70450473e-01
-1.26200771e+00 9.68047604e-02 -3.35425586e-02 2.40496229e-02
4.97664124e-01 -9.18391883e-01 -3.26432675e-01 7.54592240e-01
-1.64401078e+00 -5.59908561e-02 3.09843779e-01 2.36304790e-01
-3.01387280e-01 6.99368834e-01 -1.22217381e+00 -7.40544438e-01
-3.82900745e-01 -8.85428905e-01 7.16473699e-01 -3.88354123e-01
-4.25042331e-01 -8.56104732e-01 5.75598739e-02 6.09263241e-01
5.58709621e-01 4.02714103e-01 8.71310174e-01 -1.07955623e+00
-1.81499317e-01 -5.19601822e-01 -2.89972991e-01 2.93282688e-01
5.80074862e-02 -2.77877171e-02 -7.90267706e-01 -2.13610128e-01
2.39561111e-01 -7.52002358e-01 6.52663469e-01 -1.54886141e-01
5.80883384e-01 -1.31628108e+00 -3.50387096e-01 -2.58232117e-01
9.92773533e-01 -2.97282904e-01 5.77327132e-01 5.80954492e-01
3.95708233e-01 7.20829487e-01 4.21163946e-01 3.80548447e-01
2.97466606e-01 6.55765057e-01 3.96120071e-01 1.39989555e-01
7.00459480e-02 -8.17651570e-01 6.17218971e-01 5.79049110e-01
2.65817612e-01 -5.46121240e-01 -8.15915585e-01 4.99278814e-01
-1.60079217e+00 -1.50483716e+00 -3.64014179e-01 1.73263383e+00
9.42384005e-01 5.08814096e-01 2.72685617e-01 -7.48603046e-02
1.02183914e+00 5.21316767e-01 -1.43496946e-01 -4.96837705e-01
-2.09386885e-01 -3.03440154e-01 5.69803596e-01 5.34531534e-01
-1.14348757e+00 1.22478688e+00 6.72376156e+00 8.16358268e-01
-1.21551108e+00 2.07777217e-01 6.64527357e-01 2.62743920e-01
-3.97797734e-01 -1.54936805e-01 -6.60051107e-01 9.14441705e-01
8.00488770e-01 -2.01235473e-01 2.96383321e-01 9.00470078e-01
5.64189315e-01 1.78362414e-01 -6.76820755e-01 4.57724601e-01
4.54582483e-01 -1.63627326e+00 2.41920695e-01 -8.93274397e-02
7.46559799e-01 -5.53258369e-03 3.38588236e-03 2.85715878e-01
5.40278077e-01 -9.96178389e-01 1.34978485e+00 -1.05069846e-01
3.80878806e-01 -2.42142320e-01 8.96615386e-01 8.66281867e-01
7.02888668e-02 5.73934838e-02 -1.99617863e-01 -3.59579712e-01
1.78886935e-01 6.55703425e-01 -1.28005409e+00 4.23857942e-02
7.03077614e-02 5.32392919e-01 -6.69414639e-01 4.02318567e-01
-6.45411849e-01 1.08814263e+00 4.80511412e-02 -6.62457883e-01
4.72098172e-01 2.24461272e-01 6.53667629e-01 1.40358424e+00
1.17197111e-01 -1.70558482e-01 -4.43863459e-02 1.07067120e+00
-1.90581501e-01 -8.96078646e-02 -9.20146704e-01 -6.46685302e-01
5.88238239e-01 9.05947208e-01 -4.98053521e-01 -6.64228797e-01
-7.48964325e-02 1.08155847e+00 3.37830365e-01 1.99716076e-01
-1.09792793e+00 -1.34680703e-01 2.02506348e-01 4.01036054e-01
8.79980996e-02 -2.13726968e-01 -2.17753053e-01 -1.49064541e+00
-7.93346111e-03 -1.33679068e+00 -2.08289144e-04 -6.64181769e-01
-1.63611877e+00 7.40635097e-01 -2.24956825e-01 -1.10121512e+00
-3.83981884e-01 -5.33560932e-01 -6.10661447e-01 5.69111645e-01
-1.16331184e+00 -1.21977651e+00 -1.22619748e-01 2.93142349e-02
4.98369008e-01 -1.27481282e-01 5.82085133e-01 -8.76846164e-02
-4.11164910e-01 4.72222745e-01 -2.92134583e-02 5.93007743e-01
7.07133710e-01 -5.99307537e-01 9.38481212e-01 8.04541051e-01
-6.70269653e-02 7.12749362e-01 1.30296755e+00 -9.55923378e-01
-1.15498126e+00 -8.94660532e-01 1.28490973e+00 -7.56287038e-01
1.20449531e+00 -5.66203773e-01 -7.77328908e-01 7.58491695e-01
2.22336262e-01 -3.56913567e-01 2.70983428e-01 -3.51159424e-01
-8.74098063e-01 8.10247302e-01 -1.35501158e+00 8.19432139e-01
9.54301298e-01 -5.38908303e-01 -9.23824012e-01 7.48448610e-01
4.94546056e-01 -4.16862309e-01 -7.74535984e-02 -7.21670389e-02
5.32677591e-01 -9.80781198e-01 6.86519623e-01 -1.15434003e+00
1.16784596e+00 2.40637753e-02 2.75094658e-01 -1.30949426e+00
-2.84175813e-01 -5.20278096e-01 -1.76216736e-01 8.75655591e-01
4.84141439e-01 -8.42503369e-01 6.91649497e-01 1.00733243e-01
1.70813173e-01 -4.45420355e-01 -6.10879481e-01 -7.86779404e-01
2.44266197e-01 -4.56960276e-02 2.06170991e-01 1.40117669e+00
2.56817847e-01 5.31617105e-01 -8.77168596e-01 -4.10552174e-02
4.32857305e-01 1.13566667e-02 7.62590349e-01 -8.44904363e-01
-3.31464320e-01 -3.57779503e-01 -3.08406740e-01 -1.15909231e+00
2.97391951e-01 -6.44520760e-01 7.89277330e-02 -1.04994607e+00
3.19857091e-01 -2.90098749e-02 5.28018117e-01 2.63286382e-01
-1.14439957e-01 4.83054042e-01 2.11704727e-02 5.23557961e-01
-3.51075321e-01 5.01481414e-01 1.31019831e+00 -3.23078275e-01
1.85160428e-01 -1.64559737e-01 -9.49895024e-01 8.84217203e-01
8.57307315e-01 -8.25654685e-01 1.10546909e-02 -3.93725127e-01
3.62706870e-01 -1.15200408e-01 6.34864330e-01 -5.38448095e-01
-1.32367775e-01 -3.67350191e-01 1.31691203e-01 -2.25883931e-01
8.06730166e-02 -8.33905935e-02 -2.07495794e-01 1.00542343e+00
-6.87723219e-01 1.17905438e-01 -1.96096987e-01 7.18539417e-01
-8.12113881e-02 -3.68879020e-01 8.26352239e-01 -4.89961714e-01
-2.32401848e-01 -3.16833138e-01 -8.90674829e-01 4.02703464e-01
8.25463831e-01 5.27902767e-02 -1.11439764e+00 -8.88445139e-01
-1.67024836e-01 -3.87823492e-01 6.44260228e-01 4.16751295e-01
4.19979066e-01 -1.08405614e+00 -1.05111110e+00 -1.58563647e-02
2.50684410e-01 -8.09608519e-01 -1.66493997e-01 6.44810617e-01
-8.00398529e-01 4.88052815e-01 -1.65010884e-01 -9.11760181e-02
-9.18177485e-01 5.44378042e-01 3.94024998e-02 -1.53058395e-01
-5.57849586e-01 5.18557072e-01 -1.17472380e-01 -3.01513344e-01
-3.07660222e-01 9.66136381e-02 6.52498156e-02 -2.26427734e-01
6.87705159e-01 2.41921872e-01 -4.08076867e-02 -9.56086397e-01
-2.51020998e-01 -4.01309311e-01 -3.79820526e-01 -9.74968225e-02
1.06157005e+00 1.77570403e-01 -1.45107344e-01 8.31765980e-02
9.49950635e-01 6.73816800e-01 -6.99842989e-01 -1.25532061e-01
1.81895271e-01 -6.46514833e-01 -2.36421928e-01 -1.01993775e+00
-4.74423319e-01 4.06305939e-01 -1.13853067e-01 7.40926981e-01
1.80162877e-01 7.59757906e-02 9.10636544e-01 3.65140080e-01
3.72216612e-01 -8.88901651e-01 4.34418052e-01 5.67436934e-01
1.08307052e+00 -1.29482102e+00 1.16961561e-01 -4.84841734e-01
-7.77367949e-01 9.55110610e-01 5.51513255e-01 -3.68288219e-01
-1.40154604e-02 -1.36893392e-01 2.73063451e-01 -3.91135693e-01
-5.91442287e-01 5.52841127e-01 4.11039926e-02 3.11441571e-01
5.13221920e-01 1.69346184e-01 -8.73457491e-01 4.18102354e-01
-5.22426486e-01 -2.58167684e-01 1.08995128e+00 8.76030862e-01
-6.03256583e-01 -5.41349530e-01 -5.55397689e-01 5.03928781e-01
-6.94733679e-01 -1.81815192e-01 -1.06615758e+00 6.85478032e-01
-3.82854521e-01 1.03877473e+00 -3.16478997e-01 -4.46492910e-01
-1.78645074e-01 -1.10019073e-01 1.81149438e-01 -6.04453146e-01
-7.47248113e-01 -3.70305002e-01 8.09734702e-01 -2.99681276e-01
-1.57547548e-01 -3.16851079e-01 -6.74315989e-01 -6.92674696e-01
-5.14239907e-01 1.67378783e-01 6.59426212e-01 1.03484571e+00
4.41976786e-01 -2.35078335e-01 6.07448936e-01 -5.96073031e-01
-1.33938146e+00 -1.21653748e+00 -1.04898639e-01 7.80482471e-01
5.27405798e-01 -4.52309638e-01 -8.38277340e-01 -5.79777397e-02] | [8.189767837524414, 10.170778274536133] |
5c2d8b09-63b2-48e3-99b9-eaf6c1c49f8a | recomif-reading-comprehension-based-multi | null | null | https://www.sciencedirect.com/science/article/abs/pii/S1566253523001057?via%3Dihub | https://www.sciencedirect.com/science/article/abs/pii/S1566253523001057?via%3Dihub | ReCoMIF: Reading comprehension based multi-source information fusion network for Chinese spoken language understanding | Spoken language understanding (SLU) plays a crucial role in the performance of dialogue systems. It usually includes slot filling and intent detection (SFID) tasks aiming at semantic parsing of utterances. At present, researchers focus mainly on English SLU tasks, while such investigations on Chinese utterances are not sufficient. In this paper, we first propose a reading comprehension based multi-source information fusion network, called ReCoMIF for Chinese SFID tasks by transforming the SLU task into a multi-turn question answering procedure comprising multiple choice for intent detection and span extraction for slot filling. Moreover, we present a TCM_CLS module with a concise architecture composed of TextCNN, MaxPooling, and feed forward network plus the [CLS] representation. Such three TCM_CLS modules are stacked in the proposed ReCoMIF network that can serve as sufficient integration of multi-source information originating from contexts, reading comprehension based queries, and hidden representations concerning intent and slot semantics. Finally, experimental results and ablation studies on three Chinese SLU datasets show that our proposed model can effectively fuse intent and slot information achieved by\ state-of-the-art performance compared with other baseline methods. | ['Yun Pan', 'Ye Wang', 'Bo Jiang', 'Bi Chen', 'Hua Zhang', 'Xiawen Song', 'Xiaohui Jia', 'Bo Xie'] | 2023-08-01 | null | null | null | journal-2023-8 | ['spoken-language-understanding', 'reading-comprehension', 'semantic-parsing', 'intent-detection', 'slot-filling', 'spoken-language-understanding'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'speech'] | [ 4.15544599e-01 5.97571731e-01 -1.62701175e-01 -5.44012666e-01
-1.13156235e+00 -1.55165359e-01 5.80551922e-01 1.60606161e-01
-6.03764594e-01 7.61522532e-01 9.40914214e-01 -5.56516469e-01
2.16354236e-01 -6.02765858e-01 -1.99037582e-01 -3.87694597e-01
4.92127568e-01 4.04276341e-01 4.15561587e-01 -6.01037323e-01
2.59067237e-01 -4.44570303e-01 -1.29822278e+00 5.95084965e-01
1.34133923e+00 1.10784960e+00 6.57261789e-01 5.66529989e-01
-9.69221354e-01 1.20688081e+00 -7.94776857e-01 6.89963326e-02
-5.60348988e-01 -6.81531906e-01 -1.38944185e+00 8.80426815e-05
-2.98823297e-01 -4.74517703e-01 -3.33313420e-02 6.81215882e-01
6.49039090e-01 1.58056572e-01 3.51537645e-01 -8.98292065e-01
-3.27486604e-01 1.22394717e+00 -1.30744502e-01 1.31493732e-01
7.73666918e-01 -2.72276372e-01 1.12673593e+00 -7.02385366e-01
5.35083532e-01 1.49410725e+00 2.18702048e-01 7.39205837e-01
-6.63042426e-01 -1.52614772e-01 5.43388963e-01 2.83401340e-01
-8.12027633e-01 -5.34267366e-01 7.08846211e-01 1.09641947e-01
1.44234478e+00 4.40321296e-01 7.48604983e-02 1.07419455e+00
-1.00316949e-01 1.72010958e+00 8.29516768e-01 -8.46606910e-01
1.46444559e-01 1.97717428e-01 9.23372746e-01 5.44995546e-01
-4.27204698e-01 -5.52367330e-01 -6.46150768e-01 8.22981447e-02
3.08585465e-01 -3.11542183e-01 -4.08472329e-01 3.19730461e-01
-1.19170022e+00 9.95683610e-01 1.52774990e-01 4.83293980e-01
-2.11211145e-01 -4.19558167e-01 6.80303216e-01 2.70268857e-01
6.95233881e-01 2.98332155e-01 -7.29876637e-01 -2.33417347e-01
-5.20058334e-01 1.47518694e-01 1.06839180e+00 1.06204545e+00
7.06208766e-01 -2.21922502e-01 -6.81171119e-01 1.28967464e+00
4.95171994e-01 3.07718575e-01 7.70174861e-01 -6.44962132e-01
8.38377059e-01 8.34683537e-01 -6.08616881e-02 -4.38252121e-01
-8.74981403e-01 -1.02605201e-01 -6.64874852e-01 -7.90191412e-01
6.42054528e-02 -6.32645071e-01 -7.24349439e-01 1.74736667e+00
3.77334654e-01 -1.60354540e-01 6.27040982e-01 9.28133428e-01
1.57968533e+00 9.02357340e-01 4.81263518e-01 -3.27156901e-01
1.88020051e+00 -1.36305022e+00 -1.29886746e+00 -4.74575251e-01
1.04393804e+00 -7.92697787e-01 9.62040544e-01 -9.39139649e-02
-9.97255087e-01 -3.41175407e-01 -9.12896037e-01 -5.18432617e-01
-4.34343785e-01 4.05681729e-01 6.62812471e-01 4.52295631e-01
-9.27221060e-01 -2.28882492e-01 -3.97567034e-01 -5.24565160e-01
-1.59780428e-01 -2.57976279e-02 -1.28176630e-01 9.85277146e-02
-1.88128662e+00 8.72084618e-01 7.43216097e-01 2.18634218e-01
-4.73616183e-01 -3.51216167e-01 -1.22258842e+00 1.46087706e-01
1.01261222e+00 -5.60455918e-01 1.75084770e+00 -2.39627630e-01
-1.59669459e+00 5.59232295e-01 -6.60959721e-01 -6.39091551e-01
-2.45432202e-02 -5.68578362e-01 -3.63515556e-01 1.03210039e-01
3.68246675e-01 6.25147820e-01 5.23297906e-01 -8.58894348e-01
-8.81762505e-01 -4.57484633e-01 1.57650292e-01 9.10538614e-01
1.27117410e-02 2.18673065e-01 -4.58754927e-01 -3.03515583e-01
4.82423633e-01 -2.66600043e-01 -6.69911131e-02 -8.60933006e-01
-6.98645830e-01 -8.95920217e-01 9.39428389e-01 -9.69286799e-01
1.84086025e+00 -1.93310273e+00 8.43982026e-02 -4.60862815e-01
1.18295148e-01 2.57456958e-01 -1.05130382e-01 7.41477370e-01
2.46222243e-01 -5.08220531e-02 -2.13538319e-01 -5.35449803e-01
2.42574960e-01 2.03985155e-01 -5.62935770e-01 -2.73824781e-01
2.74469286e-01 1.09878933e+00 -6.40136898e-01 -5.17703772e-01
3.52469295e-01 -8.64721760e-02 -4.34182197e-01 7.01158285e-01
-7.43317246e-01 3.92620385e-01 -7.38394380e-01 5.71649134e-01
4.40491706e-01 -1.05525352e-01 2.11296007e-01 -2.27870280e-03
-2.42584422e-01 9.37127054e-01 -8.59963834e-01 2.10763574e+00
-5.74392736e-01 1.32079907e-02 2.66515762e-01 -1.01180696e+00
8.24244678e-01 5.16656160e-01 9.98116210e-02 -9.37329471e-01
2.81330138e-01 1.95930168e-01 -2.47851163e-01 -5.48698246e-01
8.10391784e-01 1.10193625e-01 -6.20311558e-01 7.52855599e-01
5.15045822e-01 6.00243744e-04 1.30843893e-01 4.20994043e-01
8.39826822e-01 6.85718879e-02 4.29658473e-01 -1.55903846e-01
1.02036870e+00 -2.08443645e-02 2.81197965e-01 8.35328221e-01
-2.33857289e-01 3.54173958e-01 7.43621945e-01 -3.34574804e-02
-5.69270313e-01 -5.43273032e-01 -5.75267971e-02 1.58813369e+00
2.64685929e-01 -4.93266165e-01 -1.22886014e+00 -7.27162302e-01
-4.38678294e-01 1.04875565e+00 -4.14482713e-01 -5.25419153e-02
-5.88109553e-01 -6.85028076e-01 7.03353763e-01 3.71661574e-01
1.14130998e+00 -1.42107105e+00 -4.56348985e-01 5.34925818e-01
-8.52266967e-01 -1.37179577e+00 -3.34571838e-01 3.14836115e-01
-4.63118911e-01 -1.03118300e+00 -5.59338868e-01 -9.02896583e-01
9.65349749e-02 3.80163401e-01 1.04575372e+00 -5.97228156e-03
6.28820062e-02 3.05014223e-01 -8.63967776e-01 -4.08585012e-01
-3.49325955e-01 5.68985939e-01 -3.07389081e-01 -1.12516142e-01
6.51829004e-01 -3.19024026e-02 -3.16553921e-01 1.74091995e-01
-9.18561459e-01 5.12561977e-01 4.12270308e-01 9.48434532e-01
4.57411958e-03 -3.86871755e-01 8.46676528e-01 -1.01926649e+00
1.19824505e+00 -5.25426090e-01 -4.96723913e-02 7.00264037e-01
-2.48499155e-01 1.95369497e-01 2.96816140e-01 1.78540334e-01
-1.78459716e+00 -4.44793701e-01 -8.20633829e-01 4.53710824e-01
-5.67638457e-01 6.51821196e-01 -6.33487046e-01 5.28447986e-01
3.08534414e-01 5.77114582e-01 1.56596169e-01 -7.70149350e-01
5.58433592e-01 1.13398468e+00 3.18018138e-01 -5.22891521e-01
-1.55414060e-01 -9.42602530e-02 -8.75538945e-01 -1.30425811e+00
-1.20169067e+00 -7.65649796e-01 -6.14654005e-01 2.13876553e-02
1.23594260e+00 -9.47146297e-01 -6.33040905e-01 8.31228673e-01
-1.37735558e+00 -1.57903463e-01 2.65427330e-03 1.40951917e-01
-4.75833446e-01 6.03692949e-01 -8.52243423e-01 -1.09036934e+00
-6.02835417e-01 -1.12769759e+00 1.32857680e+00 4.99263287e-01
-2.38420859e-01 -1.08230770e+00 -2.49027252e-01 8.45401883e-01
4.88196224e-01 -5.84434688e-01 1.02926159e+00 -1.27710176e+00
-3.37462008e-01 2.86576897e-01 -4.15224195e-01 1.84692279e-01
-9.70168971e-03 -9.38064575e-01 -1.12986112e+00 2.25533351e-01
3.45457941e-01 -7.52690434e-01 1.04370201e+00 4.19356912e-01
6.82588279e-01 -3.77072006e-01 -1.21433057e-01 1.13835149e-01
9.74820256e-01 4.96406108e-01 5.00179410e-01 2.70981282e-01
4.48793650e-01 8.56450140e-01 8.50656986e-01 4.87620831e-01
9.38599288e-01 4.20679986e-01 1.20065778e-01 1.56309325e-02
4.14398760e-02 -3.81386966e-01 2.88325697e-01 1.07749963e+00
6.31810486e-01 -3.53063136e-01 -7.54644990e-01 3.79603505e-01
-2.23265982e+00 -4.25610393e-01 1.36401907e-01 1.62444699e+00
8.44953418e-01 -2.47901697e-02 -2.11334392e-01 -3.17685753e-02
4.45635140e-01 5.89113176e-01 -4.44090307e-01 -3.66685927e-01
-2.32319534e-01 -5.27701005e-02 -1.39277831e-01 7.27032363e-01
-1.32684612e+00 1.46231019e+00 5.17663813e+00 9.81767654e-01
-6.67626858e-01 2.55586535e-01 5.93998492e-01 6.84795141e-01
-5.15192091e-01 -1.38853520e-01 -1.22823322e+00 1.72122985e-01
8.76067340e-01 -6.36205915e-03 -4.47768420e-02 7.11234331e-01
4.21233550e-02 -7.13171721e-01 -5.88915229e-01 6.83708012e-01
2.11162761e-01 -1.10688102e+00 3.53567265e-02 -6.04877949e-01
9.75839645e-02 -2.43398353e-01 -4.35650945e-01 7.72637367e-01
1.31520629e-01 -8.06474209e-01 3.40068340e-01 3.06612521e-01
5.25341213e-01 -4.52857703e-01 1.08311641e+00 8.09765816e-01
-1.09080279e+00 -3.09806541e-02 -3.51728916e-01 -1.05229631e-01
5.89597106e-01 2.43637666e-01 -9.38014627e-01 8.61402750e-01
3.71992856e-01 4.63012755e-01 -3.87268454e-01 4.69887376e-01
-4.12117064e-01 6.52944803e-01 -2.80996919e-01 -5.09114861e-01
6.57388926e-01 -1.00727543e-01 4.46828514e-01 1.19922948e+00
-6.76492080e-02 5.28049290e-01 3.04579973e-01 6.53688133e-01
1.77506283e-01 4.75452125e-01 -3.33219290e-01 -6.95058927e-02
4.51755464e-01 1.03043079e+00 -6.05160296e-01 -4.30489391e-01
-5.47315836e-01 9.29480910e-01 3.13958198e-01 3.20468754e-01
-4.59379494e-01 -3.54443073e-01 6.24499083e-01 -5.18488169e-01
-1.95686873e-02 -5.27894534e-02 -3.40818882e-01 -1.49230278e+00
-4.78818119e-02 -7.32262850e-01 5.99580824e-01 -6.57405734e-01
-7.59111524e-01 8.90113533e-01 1.38754562e-01 -6.97241783e-01
-6.35815322e-01 -4.32882756e-01 -5.73813558e-01 1.12061429e+00
-1.70921600e+00 -1.17430830e+00 8.37597437e-03 5.23020566e-01
1.38375044e+00 -1.33565053e-01 1.08833480e+00 -6.60942569e-02
-8.27844322e-01 2.50231117e-01 -3.94685417e-01 1.77424341e-01
3.70777875e-01 -1.13146114e+00 4.01442379e-01 7.79212117e-01
-1.40752673e-01 6.13138437e-01 2.94628233e-01 -6.09597385e-01
-1.06330848e+00 -7.25130081e-01 1.38424253e+00 -1.18518040e-01
4.65358198e-01 -3.69295865e-01 -1.00133610e+00 6.20985687e-01
6.74950778e-01 -8.15033376e-01 7.28293538e-01 3.47743750e-01
1.64096609e-01 3.88644159e-01 -8.00253034e-01 5.58464885e-01
8.33011806e-01 -6.00204706e-01 -1.14541805e+00 2.28460953e-01
1.41387725e+00 -6.71460986e-01 -5.52798867e-01 4.67692763e-01
2.31201225e-03 -9.75979686e-01 7.08160937e-01 -6.96540117e-01
3.16003174e-01 -1.21837808e-02 -2.27470994e-01 -1.02722287e+00
1.81615785e-01 -4.33268487e-01 1.69108380e-02 1.34210384e+00
6.67904377e-01 -5.29112220e-01 3.85395855e-01 6.94875836e-01
-4.89903420e-01 -7.14236677e-01 -9.58346903e-01 1.17541716e-01
-1.88334316e-01 -7.34983742e-01 5.89682639e-01 4.46369529e-01
5.86088657e-01 1.29311538e+00 -3.38304698e-01 -1.51506767e-01
3.03924948e-01 1.23558395e-01 5.21975696e-01 -8.07818830e-01
1.07972115e-01 -2.75736094e-01 3.80522788e-01 -1.78769052e+00
4.22955751e-01 -5.39720297e-01 2.63407856e-01 -1.79533839e+00
4.12106626e-02 -3.31214190e-01 9.03059319e-02 5.35915852e-01
-5.49851179e-01 -6.03086531e-01 2.56935060e-01 1.28123283e-01
-1.04992688e+00 1.05501699e+00 1.12735868e+00 -4.26988006e-02
-3.87134075e-01 1.81884602e-01 -9.02041554e-01 5.89364588e-01
6.56247914e-01 2.84369498e-01 -6.24533236e-01 -4.50578213e-01
-3.10138673e-01 8.08389366e-01 -2.73745537e-01 -7.23823071e-01
4.49280769e-01 2.02630788e-01 -1.08625777e-01 -1.02875829e+00
2.74701238e-01 -4.13438529e-01 -7.55717814e-01 1.82473168e-01
-6.61685407e-01 -4.02131826e-01 7.99593180e-02 1.09290443e-01
-6.46542609e-01 -5.13134122e-01 2.17417493e-01 -4.42164719e-01
-1.29293799e+00 -1.52355000e-01 -6.38781965e-01 2.51358509e-01
6.31860077e-01 1.77415892e-01 -4.87750858e-01 -6.18112683e-01
-7.87822247e-01 8.61386836e-01 -4.51855391e-01 7.37615943e-01
7.82354593e-01 -7.36191690e-01 -5.64689577e-01 5.46420097e-01
1.57246187e-01 3.52425098e-01 6.80886745e-01 6.55576468e-01
-2.06194639e-01 1.17021048e+00 1.63395450e-01 -4.96048778e-01
-9.73407507e-01 1.07701279e-01 9.84872803e-02 -7.32656598e-01
-4.67055738e-01 1.06755853e+00 4.22407597e-01 -1.13063788e+00
5.24017632e-01 -4.35364038e-01 -9.80157971e-01 3.38763356e-01
7.21439064e-01 1.04499653e-01 7.50120580e-02 -5.43095946e-01
-2.57914752e-01 1.07221240e-02 -2.52182245e-01 -3.70260715e-01
7.55437911e-01 -7.71197855e-01 -3.41150880e-01 4.96784776e-01
9.33995545e-01 -4.81738895e-01 -8.09732556e-01 -5.97207069e-01
2.68196821e-01 3.87335330e-01 -2.30766699e-01 -1.09082139e+00
-3.08135659e-01 1.06249166e+00 6.91182390e-02 3.18956673e-01
1.14857376e+00 1.96413219e-01 1.29290402e+00 4.92537320e-01
3.92286748e-01 -1.25396788e+00 -1.76680669e-01 1.14639306e+00
8.80774081e-01 -1.36808538e+00 -5.46840847e-01 -5.42893767e-01
-1.04955924e+00 9.69165623e-01 9.76103425e-01 4.89685744e-01
7.24769711e-01 -1.49728544e-03 3.65821898e-01 -6.61639078e-03
-9.64305758e-01 -5.24582863e-01 2.19034612e-01 2.14048699e-01
4.85650867e-01 7.04933852e-02 -8.99824202e-01 1.23196781e+00
-2.21191019e-01 -2.38697544e-01 2.30286583e-01 1.06013000e+00
-9.65315580e-01 -1.13021863e+00 -1.84466138e-01 4.74383026e-01
-3.43170047e-01 -4.17947590e-01 -2.53231436e-01 5.18514633e-01
-2.95193285e-01 1.27492630e+00 -3.50127667e-02 -4.99174684e-01
2.62384236e-01 6.91838324e-01 -1.58461973e-01 -8.43283892e-01
-6.30069673e-01 3.46913844e-01 5.75062156e-01 -6.99200749e-01
-4.67603862e-01 -4.91882086e-01 -1.38641775e+00 3.17986250e-01
-4.67473477e-01 6.34492397e-01 5.10636270e-01 1.45434701e+00
2.12405786e-01 6.46369994e-01 5.77443719e-01 -3.86553466e-01
-4.90289152e-01 -1.55009103e+00 -5.26591837e-01 -1.15466930e-01
2.27142274e-01 -4.24007773e-01 -2.23314166e-01 -4.25426960e-01] | [12.604483604431152, 7.4475178718566895] |
1486e59b-73ce-4a1c-b463-de3cad82f42e | 3d-reconstruction-through-fusion-of-cross | 2106.14306 | null | https://arxiv.org/abs/2106.14306v1 | https://arxiv.org/pdf/2106.14306v1.pdf | 3D Reconstruction through Fusion of Cross-View Images | 3D recovery from multi-stereo and stereo images, as an important application of the image-based perspective geometry, serves many applications in computer vision, remote sensing and Geomatics. In this chapter, the authors utilize the imaging geometry and present approaches that perform 3D reconstruction from cross-view images that are drastically different in their viewpoints. We introduce our framework that takes ground-view images and satellite images for full 3D recovery, which includes necessary methods in satellite and ground-based point cloud generation from images, 3D data co-registration, fusion and mesh generation. We demonstrate our proposed framework on a dataset consisting of twelve satellite images and 150k video frames acquired through a vehicle-mounted Go-pro camera and demonstrate the reconstruction results. We have also compared our results with results generated from an intuitive processing pipeline that involves typical geo-registration and meshing methods. | ['Mostafa Elhashash', 'Xiao Ling', 'Shuang Song', 'Rongjun Qin'] | 2021-06-27 | null | null | null | null | ['point-cloud-generation'] | ['computer-vision'] | [ 3.07640612e-01 -4.26724076e-01 6.61629617e-01 -3.11608642e-01
-7.89171934e-01 -6.59154832e-01 7.10208118e-01 -3.36466432e-01
-2.25325331e-01 3.02674413e-01 2.33113356e-02 -2.49628797e-01
-1.83908939e-01 -1.23035157e+00 -7.57058322e-01 -4.85112101e-01
3.63491289e-02 8.15191805e-01 4.68601584e-01 -5.89221060e-01
4.56118137e-01 9.22380090e-01 -1.90683413e+00 -6.06400631e-02
7.33166754e-01 7.06987023e-01 4.43674684e-01 7.69791007e-01
-6.43329322e-03 -1.09583162e-01 2.36724570e-01 1.05503974e-02
8.04143190e-01 8.61948282e-02 -8.68666589e-01 7.45425642e-01
1.11255872e+00 -6.63400710e-01 -1.28635824e-01 1.12912643e+00
3.42941403e-01 7.82089215e-03 5.34239829e-01 -7.72369385e-01
7.22575709e-02 -3.12637269e-01 -9.29546356e-01 -8.25069398e-02
7.77196646e-01 -6.62560761e-02 5.68179727e-01 -1.21465218e+00
8.42888236e-01 1.29688108e+00 1.08572781e+00 -6.06703982e-02
-1.28706515e+00 -2.43370488e-01 -5.64553082e-01 -1.32635860e-02
-1.62909114e+00 -4.09605801e-01 5.35587668e-01 -6.75681293e-01
7.94272363e-01 4.66363549e-01 8.66848409e-01 1.08602643e-01
8.63426104e-02 5.37267588e-02 1.46522701e+00 -6.25107110e-01
-1.03940226e-01 -3.54806393e-01 -2.34730020e-02 3.73187184e-01
2.83447634e-02 3.80165786e-01 -4.12478626e-01 -3.70429724e-01
1.35633004e+00 3.36100966e-01 -4.13477302e-01 -3.52270961e-01
-1.32615960e+00 6.58626556e-01 2.40981832e-01 -7.78630450e-02
-4.67946559e-01 1.24192446e-01 -1.64621159e-01 3.17399174e-01
8.05991709e-01 -2.36847699e-01 -1.68033898e-01 4.31723028e-01
-1.14105833e+00 5.04704893e-01 2.68091500e-01 1.08606529e+00
1.43327916e+00 2.04445690e-01 6.83714151e-01 6.25568569e-01
7.04211235e-01 1.36000943e+00 -3.89752463e-02 -1.27771997e+00
3.81734580e-01 6.00054502e-01 2.18421891e-01 -1.20044327e+00
-2.27204904e-01 1.94581732e-01 -5.79607010e-01 5.05441308e-01
-9.82362926e-02 3.07468027e-01 -6.52123511e-01 6.49425089e-01
7.12456286e-01 2.00521171e-01 2.04746023e-01 7.93554127e-01
1.04451478e+00 5.18505454e-01 -6.95627928e-01 1.36106694e-02
1.37034786e+00 -4.90512818e-01 -2.45303378e-01 -1.27832547e-01
2.19203487e-01 -1.00224769e+00 6.73584938e-01 2.63731122e-01
-1.09544134e+00 -4.61856246e-01 -6.64224982e-01 1.69133078e-02
-5.00507094e-02 -5.25457859e-02 2.06982747e-01 1.19702049e-01
-1.42814887e+00 6.74271047e-01 -6.99391007e-01 -7.42056906e-01
-4.50200886e-02 1.52329102e-01 -7.60550320e-01 -1.96954291e-02
-7.67001390e-01 8.89848888e-01 2.84486771e-01 1.06263189e-02
-7.14056134e-01 -6.71797037e-01 -9.02825952e-01 -2.64483482e-01
6.52061254e-02 -1.07642925e+00 8.75639677e-01 -5.93381345e-01
-1.14577711e+00 1.54022324e+00 -1.82965323e-01 -2.48484954e-01
3.55884820e-01 -1.55532762e-01 -1.26002386e-01 3.86354715e-01
2.79487789e-01 3.39754105e-01 7.00273097e-01 -1.52966106e+00
-6.61616504e-01 -8.36389780e-01 -1.52566209e-01 5.66572845e-01
4.89561260e-01 2.57457048e-03 -3.25306505e-01 -3.13144207e-01
9.15832758e-01 -8.91333163e-01 -4.07611191e-01 3.75150554e-02
-2.56475151e-01 6.42302155e-01 9.97709990e-01 -7.64262795e-01
4.02912229e-01 -1.82733607e+00 1.15103573e-01 3.95150572e-01
-1.69877097e-01 -1.76008016e-01 2.11459026e-01 7.18683660e-01
9.36850347e-03 -4.73881252e-02 -4.22597110e-01 -3.22897643e-01
-6.31152570e-01 2.92779148e-01 -4.04913992e-01 9.31747973e-01
-4.02903378e-01 3.36578578e-01 -5.48839271e-01 -4.04514968e-01
8.99264574e-01 4.35586661e-01 -3.55652601e-01 1.03185840e-01
7.21436441e-02 8.44351351e-01 -2.78526068e-01 8.68063867e-01
1.45909894e+00 1.42947957e-01 -1.87070802e-01 -4.70422298e-01
-7.85373509e-01 -1.97531402e-01 -1.59781921e+00 1.86100924e+00
-4.85591531e-01 4.32465822e-01 7.20537364e-01 -5.93134463e-01
9.88569200e-01 2.90962607e-01 4.84427661e-01 -3.53514791e-01
-9.29249600e-02 3.51419628e-01 -8.71147156e-01 -6.04691446e-01
9.27587330e-01 -3.87069970e-01 2.31091589e-01 2.68460870e-01
-2.47905210e-01 -1.06870317e+00 -1.97891280e-01 1.01522803e-01
4.01069909e-01 4.80399996e-01 4.11424160e-01 -5.71752489e-01
8.76892149e-01 6.98509634e-01 1.03999689e-01 2.85516560e-01
5.32161832e-01 1.18354809e+00 -4.56114501e-01 -7.06889510e-01
-1.45232081e+00 -9.76538658e-01 -4.94155377e-01 2.46226400e-01
4.05137181e-01 -1.57173738e-01 -4.34480220e-01 2.72924781e-01
1.62237212e-01 1.97193146e-01 -3.15595031e-01 7.19244957e-01
-4.51688766e-01 -7.38013208e-01 2.26512730e-01 -5.63764311e-02
6.32587969e-01 -3.66296411e-01 -6.66007042e-01 8.13389868e-02
-3.43126208e-01 -1.08634782e+00 -4.07259949e-02 -6.30969644e-01
-1.57745981e+00 -1.27589142e+00 -5.69316089e-01 -3.05378437e-01
7.82759070e-01 1.14983225e+00 1.27950335e+00 3.58740032e-01
-1.13882348e-01 7.99619853e-01 -3.03012341e-01 -2.27636583e-02
-3.48009884e-01 -5.48600793e-01 1.52212381e-01 8.20609778e-02
-5.03311604e-02 -1.15146899e+00 -6.25640631e-01 6.14732802e-01
-1.08192432e+00 4.15389746e-01 4.84433621e-02 5.02389133e-01
1.02635086e+00 -1.64629042e-01 -6.57583475e-01 -5.48166811e-01
-2.21521512e-01 -2.57023305e-01 -1.11414838e+00 3.77166793e-02
-2.23312378e-01 -3.67081851e-01 -1.29749263e-02 5.72465301e-01
-9.93305922e-01 4.82013136e-01 -2.98083901e-01 -5.37280798e-01
-3.08943361e-01 5.55168569e-01 7.32327029e-02 -6.20798409e-01
7.28465140e-01 5.79034984e-01 5.57198346e-01 -7.95080364e-01
2.35542685e-01 8.43784153e-01 6.90221965e-01 -3.48506153e-01
1.19639218e+00 1.55949223e+00 2.93421149e-01 -1.36507273e+00
-4.51756477e-01 -9.29424644e-01 -1.14033556e+00 -4.02764320e-01
7.74041235e-01 -1.40299857e+00 -4.04908776e-01 6.65933073e-01
-1.36271489e+00 6.15488663e-02 6.17151260e-02 5.28045118e-01
-7.33784735e-01 7.67327547e-01 -1.63888380e-01 -7.25502729e-01
-5.33233881e-01 -1.17144120e+00 1.79326892e+00 9.32862982e-02
4.98531431e-01 -1.25678325e+00 3.80218714e-01 5.86436391e-01
2.36003637e-01 6.64511561e-01 1.17907360e-01 3.60014349e-01
-9.80302989e-01 1.76073045e-01 -9.23985168e-02 -1.05149718e-02
1.42899320e-01 3.12303275e-01 -9.76167500e-01 -1.54953331e-01
9.12604257e-02 7.39701986e-02 7.04888761e-01 5.99646628e-01
4.41026092e-01 5.08512966e-02 -3.05956453e-01 1.31820977e+00
2.27210379e+00 -3.76933545e-01 1.03053772e+00 7.98259914e-01
9.07799125e-01 8.44554543e-01 7.18249500e-01 5.56059301e-01
6.73037350e-01 1.07582772e+00 9.81002748e-01 -2.03586876e-01
1.09202787e-01 5.08930311e-02 1.97650477e-01 6.68934524e-01
-9.73744214e-01 4.21307862e-01 -1.19583774e+00 4.47712302e-01
-1.60078323e+00 -1.31954551e+00 -1.22509646e+00 2.42781401e+00
2.02197805e-01 -7.19753087e-01 -2.20534936e-01 -3.63233313e-02
7.01065660e-01 3.85939814e-02 -6.41173124e-03 1.00102417e-01
-4.06757444e-01 3.10313478e-02 1.03155279e+00 1.08799660e+00
-1.10436213e+00 1.09276378e+00 6.51184511e+00 4.93356436e-01
-1.09531105e+00 3.60550918e-02 -6.69693276e-02 5.87766409e-01
-6.48983538e-01 4.87579256e-01 -8.58329594e-01 -9.58453864e-02
5.96772134e-01 -1.70126826e-01 2.48290166e-01 8.11295748e-01
7.34306753e-01 -6.39789164e-01 -5.82597017e-01 1.32006955e+00
2.07564354e-01 -1.73021960e+00 2.14123815e-01 5.44759274e-01
8.63847017e-01 5.50087929e-01 -4.88990039e-01 -7.08614230e-01
2.95242876e-01 -6.70491934e-01 9.34693575e-01 7.98951209e-01
8.69882226e-01 -4.19341058e-01 3.48238915e-01 5.01578689e-01
-1.42359781e+00 5.09991646e-01 -5.20834684e-01 -2.35384956e-01
7.91498005e-01 8.88578773e-01 -6.03446782e-01 1.31648362e+00
1.03539360e+00 1.01186955e+00 -5.46669602e-01 9.63919699e-01
1.10324986e-01 -8.27935413e-02 -4.92398083e-01 1.00672841e+00
5.26527502e-02 -1.08624864e+00 7.76658654e-01 7.33885050e-01
9.12583411e-01 4.37555611e-01 7.27996826e-02 7.17855811e-01
6.54288948e-01 -1.58965718e-02 -1.21142459e+00 9.47265446e-01
2.91144311e-01 1.36528146e+00 -5.53827226e-01 -4.66496974e-01
-4.16115105e-01 8.02499712e-01 -4.39468265e-01 -6.12855889e-02
-4.58460450e-01 1.73621073e-01 5.75653791e-01 5.59190035e-01
2.41166037e-02 -7.21726656e-01 -2.23196417e-01 -1.46456110e+00
-1.72568858e-01 -4.66344357e-01 1.34325325e-01 -1.37327313e+00
-9.04106081e-01 4.90015119e-01 5.21432459e-01 -1.99638569e+00
-2.27112353e-01 -5.07256985e-01 -6.67797923e-01 1.14178371e+00
-1.54360354e+00 -1.43404698e+00 -8.69330168e-01 7.95430124e-01
4.38852310e-01 2.31537119e-01 6.78162634e-01 1.96703047e-01
2.03150973e-01 -7.44852006e-01 5.76896489e-01 -3.16872686e-01
3.42946082e-01 -9.71329749e-01 5.77478051e-01 9.47832465e-01
-3.79990824e-02 4.10884172e-01 7.26093352e-01 -7.05130577e-01
-1.74298859e+00 -1.07313251e+00 5.72323263e-01 -4.46410000e-01
2.91721851e-01 1.77587837e-01 -7.75255203e-01 8.44634652e-01
1.17804743e-01 1.90893784e-01 3.10071170e-01 -6.12691641e-01
2.34215394e-01 -1.00070953e-01 -1.35044467e+00 1.92428738e-01
1.10961580e+00 -4.11596745e-01 -8.27862680e-01 5.33604562e-01
-3.96121852e-02 -8.56833637e-01 -9.57731962e-01 4.00429219e-01
5.55021644e-01 -1.65496302e+00 1.40077889e+00 2.63665766e-01
2.63923377e-01 -8.90335619e-01 -6.33006811e-01 -1.36721420e+00
-4.92068306e-02 -2.17802852e-01 7.79353321e-01 8.32360804e-01
-1.61839545e-01 -5.46696246e-01 6.24468565e-01 2.17216745e-01
-4.64394122e-01 1.55974329e-01 -9.71447527e-01 -8.03893089e-01
-1.64130673e-01 -3.50203216e-01 6.23182833e-01 9.26163435e-01
-6.19242311e-01 -2.86656730e-02 -3.57660204e-01 7.70252645e-01
1.22464442e+00 5.98159552e-01 1.34199786e+00 -1.75773883e+00
-1.02808150e-02 2.22344398e-01 -5.47057509e-01 -1.11817670e+00
-2.46913776e-01 -6.44058228e-01 -1.81216478e-01 -1.70376408e+00
6.94933385e-02 -5.26253819e-01 1.08446372e+00 -6.16796650e-02
3.73381525e-01 5.88238180e-01 1.19059170e-02 9.35372949e-01
1.08894825e-01 4.58050370e-01 1.12675714e+00 1.59698069e-01
-3.31321321e-02 -2.15506643e-01 5.74666113e-02 1.11005557e+00
3.30008715e-01 -2.87624002e-01 3.02215144e-02 -9.81393397e-01
3.32872361e-01 3.71162921e-01 7.32582033e-01 -8.99908483e-01
7.68524930e-02 -4.27934259e-01 -9.67695378e-03 -1.21665633e+00
6.78712189e-01 -1.27501738e+00 1.18047214e+00 4.56866086e-01
6.99632049e-01 2.76280016e-01 -5.02044410e-02 3.79409522e-01
-4.04273808e-01 -3.43227178e-01 9.14832711e-01 -7.80027390e-01
-1.01948345e+00 3.87281895e-01 -2.02365696e-01 -4.14722353e-01
9.96618986e-01 -8.22432935e-01 -1.17323272e-01 -3.10523987e-01
-8.09876680e-01 1.42839372e-01 1.35430944e+00 -2.82378554e-01
9.66830015e-01 -1.45445669e+00 -1.00245905e+00 5.04235446e-01
2.45595589e-01 6.30184710e-01 5.04811883e-01 8.66411030e-01
-1.61871159e+00 2.02000126e-01 -3.54486197e-01 -1.37688887e+00
-1.63131666e+00 -1.56261787e-01 5.81542015e-01 2.15012029e-01
-8.51332426e-01 2.75606662e-01 1.02879908e-02 -8.83859992e-01
-7.52612710e-01 -2.57217139e-01 -2.92928547e-01 -1.74659654e-01
6.60703421e-01 6.38903081e-01 3.44993234e-01 -1.50553966e+00
-4.40924197e-01 1.66187632e+00 6.59147322e-01 -6.31446600e-01
1.61636317e+00 -6.50016665e-01 -4.12739247e-01 1.48215368e-01
7.31351674e-01 1.71661839e-01 -1.15918982e+00 -2.31400430e-01
-3.27088326e-01 -1.13229609e+00 3.81674856e-01 2.24181876e-01
-1.07531559e+00 8.88341010e-01 6.00639105e-01 4.32953089e-02
9.10918474e-01 3.50197069e-02 3.58311236e-01 2.51370609e-01
9.38634336e-01 -7.63870597e-01 -5.37078917e-01 7.50932515e-01
1.10810077e+00 -1.21927679e+00 5.22214115e-01 -7.87219584e-01
-2.24360242e-01 1.48833668e+00 1.69701710e-01 -2.82881141e-01
7.76413977e-01 -6.91694766e-02 9.07591805e-02 -6.30101681e-01
-1.62169471e-01 -3.08072120e-01 -5.18661402e-02 8.87578785e-01
1.22852713e-01 -2.60199249e-01 -2.28352442e-01 -5.77990294e-01
-8.38003159e-02 -3.07664331e-02 9.65350568e-01 1.02039802e+00
-7.33011782e-01 -1.19149101e+00 -1.38141334e+00 1.21775746e-01
-7.66737759e-02 -2.09133193e-01 -6.46200553e-02 6.22209489e-01
4.09171134e-02 5.79190373e-01 4.52220351e-01 -3.58828276e-01
4.36829597e-01 -4.49337602e-01 5.38086534e-01 -6.25337541e-01
-2.49152794e-01 2.09951818e-01 1.47521675e-01 -5.22826433e-01
-1.22836268e+00 -1.08708024e+00 -1.02851057e+00 -5.36015332e-01
-4.11614925e-01 8.94941669e-03 9.22923684e-01 7.53632903e-01
3.29169512e-01 -4.98781323e-01 9.38042104e-01 -1.71899557e+00
-1.13182373e-01 -8.30079615e-01 -1.03630447e+00 3.92401189e-01
2.66930580e-01 -8.16756666e-01 -6.33746505e-01 6.82301342e-01] | [8.545418739318848, -2.62018084526062] |
7ad302e8-67aa-4a0e-86fe-927228073d78 | constrained-structured-regression-with | 1511.07497 | null | http://arxiv.org/abs/1511.07497v1 | http://arxiv.org/pdf/1511.07497v1.pdf | Constrained Structured Regression with Convolutional Neural Networks | Convolutional Neural Networks (CNNs) have recently emerged as the dominant
model in computer vision. If provided with enough training data, they predict
almost any visual quantity. In a discrete setting, such as classification, CNNs
are not only able to predict a label but often predict a confidence in the form
of a probability distribution over the output space. In continuous regression
tasks, such a probability estimate is often lacking. We present a regression
framework which models the output distribution of neural networks. This output
distribution allows us to infer the most likely labeling following a set of
physical or modeling constraints. These constraints capture the intricate
interplay between different input and output variables, and complement the
output of a CNN. However, they may not hold everywhere. Our setup further
allows to learn a confidence with which a constraint holds, in the form of a
distribution of the constrain satisfaction. We evaluate our approach on the
problem of intrinsic image decomposition, and show that constrained structured
regression significantly increases the state-of-the-art. | ['Trevor Darrell', 'Philipp Krähenbühl', 'Deepak Pathak', 'Stella X. Yu'] | 2015-11-23 | null | null | null | null | ['intrinsic-image-decomposition'] | ['computer-vision'] | [ 3.03212821e-01 3.03294539e-01 -4.04271752e-01 -4.24395710e-01
-3.49987626e-01 -5.00270724e-01 6.56976521e-01 2.90644854e-01
-5.09711623e-01 6.71415567e-01 -3.44929397e-01 -8.54938757e-04
-7.43611902e-02 -8.58902335e-01 -1.10241234e+00 -9.62746441e-01
2.69168496e-01 4.50061560e-01 4.62082438e-02 2.64090568e-01
4.16144319e-02 5.86559176e-01 -1.56116152e+00 1.96142316e-01
4.98622715e-01 1.46057332e+00 3.77537787e-01 4.44127560e-01
-1.03107719e-02 8.71746719e-01 -2.83624560e-01 -5.21696731e-02
1.52445629e-01 -1.56049937e-01 -6.32148683e-01 3.81370336e-02
3.04556042e-01 1.83069997e-03 -2.41804510e-01 1.19370651e+00
-1.60786644e-01 6.55992329e-02 1.07768917e+00 -1.08494675e+00
-4.23046738e-01 2.96498269e-01 -3.13420385e-01 -2.83512801e-01
-2.03015283e-01 6.93291128e-02 1.28653419e+00 -6.61656022e-01
6.80795252e-01 8.82065356e-01 4.60804731e-01 4.42410946e-01
-1.79190564e+00 -2.43444920e-01 3.67246568e-01 7.94866085e-02
-1.16033256e+00 -1.91107333e-01 7.83076823e-01 -7.54722357e-01
6.25937819e-01 4.72645089e-02 6.05950594e-01 9.94112134e-01
2.31278345e-01 5.85952282e-01 1.07349741e+00 -5.20155847e-01
5.87650716e-01 2.70842522e-01 5.54728024e-02 8.94190371e-01
1.30846143e-01 6.52477220e-02 -6.89695001e-01 1.60327375e-01
8.21301997e-01 -1.10246219e-01 -4.39406663e-01 -4.97184485e-01
-9.70587611e-01 8.09504509e-01 6.60479784e-01 2.07344480e-02
-1.93327948e-01 3.69665265e-01 1.24874935e-01 1.86220065e-01
3.41046125e-01 3.95535380e-01 -7.41599917e-01 4.11958456e-01
-8.37849617e-01 1.32180080e-01 8.09852958e-01 5.95601141e-01
1.16303337e+00 -2.58893758e-01 1.15637807e-02 5.63720763e-01
3.46609682e-01 2.24958017e-01 -5.02205268e-02 -1.15516114e+00
2.19859824e-01 4.92874175e-01 2.55565435e-01 -1.01703751e+00
-2.80250698e-01 -6.71778560e-01 -1.10273707e+00 9.24541235e-01
8.40542614e-01 -9.65664536e-02 -8.39378595e-01 2.25583053e+00
6.59162998e-02 -6.23621307e-02 -2.78433353e-01 9.19408917e-01
2.55297095e-01 4.89898980e-01 -2.03084722e-01 -2.62703508e-01
9.04136539e-01 -5.45313776e-01 -3.30324382e-01 -3.49360526e-01
3.11025351e-01 -3.52676839e-01 7.82306194e-01 8.56942415e-01
-8.34052086e-01 -5.42272925e-01 -9.69425321e-01 -7.66467601e-02
-1.36399984e-01 3.65885794e-01 4.32033837e-01 3.65039587e-01
-8.67922485e-01 1.13875020e+00 -1.18395603e+00 -4.33344468e-02
4.26134318e-01 5.67938864e-01 -5.06686926e-01 -2.56074555e-02
-5.37327945e-01 9.76684868e-01 2.22747847e-01 3.17771405e-01
-7.00068295e-01 -6.02399409e-01 -6.81473434e-01 2.52960473e-01
3.83330703e-01 -6.43097401e-01 1.03945422e+00 -1.03720319e+00
-1.22614837e+00 9.36428070e-01 -8.63649175e-02 -4.39500809e-01
5.97023487e-01 -6.87110275e-02 1.55343264e-01 1.19819157e-01
-2.41376355e-01 7.88576365e-01 1.02311420e+00 -1.33420563e+00
-3.54086131e-01 -2.83889830e-01 1.78074896e-01 -3.05188656e-01
-6.14334224e-03 -3.37091893e-01 -3.44215721e-01 -1.26024470e-01
4.05210495e-01 -9.04991746e-01 -3.95008713e-01 7.23858953e-01
-6.62461102e-01 -2.27581430e-02 4.20190424e-01 -2.86083460e-01
6.31387651e-01 -2.03562665e+00 7.59957850e-01 2.14881673e-01
4.15833116e-01 -3.03539038e-01 1.71817839e-01 -3.22366953e-02
-1.36830717e-01 2.06198394e-01 -5.42238593e-01 -6.99205339e-01
1.74031883e-01 3.78700286e-01 -2.23259509e-01 7.46928811e-01
7.61581242e-01 6.00224197e-01 -5.18400490e-01 -3.61606508e-01
2.51880616e-01 5.35820127e-01 -5.08630872e-01 3.33570182e-01
-8.91763031e-01 6.94945276e-01 -3.82316321e-01 1.77891701e-01
4.92616147e-01 -5.29401243e-01 2.28223771e-01 -2.55572885e-01
-1.74788371e-01 4.64196764e-02 -1.19628322e+00 1.64568210e+00
-5.32824993e-01 7.67822266e-01 2.24084198e-01 -1.49910486e+00
9.34640288e-01 -4.38887104e-02 4.31720018e-01 -2.93237805e-01
3.38813663e-01 -3.96209061e-02 -1.53766398e-03 -3.66094708e-01
4.74979468e-02 -3.43993872e-01 4.20161672e-02 3.05284113e-01
2.66577959e-01 -2.66020186e-02 -1.01420164e-01 -1.51225820e-01
1.13160586e+00 3.67521495e-01 2.92961355e-02 -4.52320814e-01
3.96218061e-01 -1.51971519e-01 6.24872267e-01 7.30861127e-01
3.65748227e-01 9.44845080e-01 1.20335925e+00 -4.19575274e-01
-1.17419744e+00 -8.12513351e-01 -3.28341722e-01 5.93851268e-01
-6.61847889e-02 -1.69577777e-01 -7.14599133e-01 -4.44659054e-01
2.41072499e-03 2.09103212e-01 -9.92971003e-01 -1.03197135e-01
-3.80269587e-01 -6.33716881e-01 4.20177132e-02 2.68792897e-01
1.19166486e-01 -1.00146317e+00 -7.67179728e-01 7.07300147e-04
4.14183512e-02 -1.19287431e+00 6.42279312e-02 8.49999428e-01
-7.65628159e-01 -1.16542506e+00 -2.56861061e-01 -5.37413597e-01
8.79512191e-01 -3.42316210e-01 1.21837878e+00 2.59706765e-01
-4.78660047e-01 1.94053650e-01 9.25636515e-02 -2.96648722e-02
-3.40936571e-01 1.17275544e-01 4.08348441e-02 3.00873786e-01
-2.89788861e-02 -6.94544852e-01 -4.38666582e-01 8.11715424e-02
-8.81075680e-01 1.59798548e-01 4.24271375e-01 8.32732558e-01
8.43790412e-01 1.94497183e-01 4.33532037e-02 -8.13838422e-01
1.10527221e-02 -3.77058715e-01 -1.01525760e+00 3.72766227e-01
-2.97079712e-01 6.76295042e-01 9.69040573e-01 -6.04107618e-01
-7.31971323e-01 4.83883440e-01 -8.75044707e-03 -6.10013366e-01
-4.89131421e-01 7.02768266e-01 -2.66406149e-01 -3.22408453e-02
5.49996793e-01 -2.31031224e-01 -1.02802299e-01 -6.68749273e-01
1.98254675e-01 1.29021816e-02 6.88522100e-01 -1.00411749e+00
6.92520738e-01 5.13005853e-01 7.92074263e-01 -6.34604096e-01
-1.06678689e+00 -9.99037400e-02 -9.01728094e-01 -4.33188289e-01
9.60172653e-01 -5.58233023e-01 -1.07672513e+00 3.93254459e-01
-1.24926305e+00 -6.59717143e-01 -3.18648875e-01 4.52358574e-01
-6.81438863e-01 -9.89779383e-02 -3.04697573e-01 -9.90200162e-01
3.35481852e-01 -1.21179831e+00 8.02062929e-01 2.05876410e-01
2.80980822e-02 -9.72135127e-01 -1.96850568e-01 -9.14206356e-02
1.19971491e-01 5.88697791e-01 1.17554033e+00 -7.17675760e-02
-8.60493541e-01 -9.65614915e-02 -2.92546809e-01 4.35143113e-01
-1.30071521e-01 3.46690685e-01 -1.12559855e+00 -2.12465748e-02
5.92455417e-02 -4.92326200e-01 1.33141208e+00 6.11658812e-01
1.57265127e+00 -2.12685034e-01 -1.25676826e-01 8.43898654e-01
1.69273996e+00 -4.63654637e-01 4.28688318e-01 5.06536365e-02
5.53489625e-01 9.09005702e-01 2.43133396e-01 2.87492335e-01
-1.30685687e-01 7.74623454e-01 1.02016139e+00 1.47016227e-01
3.79338898e-02 -2.03483433e-01 2.24790052e-02 4.37875420e-01
-1.77390978e-01 -2.68611610e-01 -7.38075674e-01 1.17740244e-01
-1.91821015e+00 -6.57627165e-01 -2.02512786e-01 2.33387613e+00
8.32910120e-01 4.70008343e-01 -9.75943953e-02 2.38119379e-01
5.20273328e-01 2.06976384e-02 -9.57611918e-01 6.74921647e-02
-1.08409062e-01 3.14111412e-01 3.67431343e-01 7.96347082e-01
-9.04670835e-01 4.92903590e-01 5.97802877e+00 5.54695010e-01
-1.37695134e+00 -1.33638442e-01 9.11948085e-01 1.31335154e-01
-3.36307079e-01 1.21959314e-01 -6.83491886e-01 3.23201418e-01
5.25208414e-01 3.36764276e-01 5.87299824e-01 7.30965137e-01
-1.04691550e-01 -4.04462695e-01 -1.51171505e+00 7.98404217e-01
-2.22245738e-01 -1.40542996e+00 -3.88460010e-01 4.90620613e-01
4.59823310e-01 1.11793332e-01 2.46735454e-01 -8.75648782e-02
1.97307870e-01 -1.48922384e+00 9.20240581e-01 9.08275187e-01
1.04144704e+00 -5.35513937e-01 4.26992387e-01 7.13610590e-01
-7.56859899e-01 3.36752310e-02 -5.47455072e-01 -1.87754616e-01
-6.76935613e-02 8.69874120e-01 -3.58245164e-01 1.63118511e-01
3.83450776e-01 7.22932875e-01 -2.67344564e-01 7.85630047e-01
-3.91434103e-01 5.08373141e-01 -5.78868330e-01 -1.22653572e-02
-6.61377534e-02 -4.79126990e-01 2.45984674e-01 9.52837110e-01
1.61828771e-01 7.25279599e-02 7.16417432e-02 1.37208807e+00
-2.70356297e-01 -3.05434018e-01 -3.83474708e-01 1.54851854e-01
-1.29797593e-01 1.42604685e+00 -9.24681246e-01 1.46543756e-01
-2.33666077e-01 6.17361069e-01 8.15307677e-01 4.98537511e-01
-5.72168171e-01 2.82324165e-01 7.51657963e-01 3.49564332e-04
3.92485827e-01 -5.03924370e-01 -5.55218279e-01 -1.24914253e+00
2.82368451e-01 -4.16448325e-01 -1.49626896e-01 -6.98740482e-01
-1.17824411e+00 3.96080583e-01 -1.30532131e-01 -9.22257245e-01
-1.33290708e-01 -1.40062463e+00 -4.72997397e-01 9.14856434e-01
-1.52606416e+00 -7.50060856e-01 -2.71543473e-01 3.42799753e-01
2.35372439e-01 -4.72610258e-03 8.42221081e-01 -1.30384594e-01
-6.60516918e-01 2.93878615e-02 -8.71641263e-02 8.16204548e-02
3.71845782e-01 -1.47836018e+00 -1.94524616e-01 6.41320884e-01
3.55393052e-01 5.21278024e-01 1.10398877e+00 -2.39561647e-01
-1.41456652e+00 -7.66614139e-01 4.67773616e-01 -4.27566677e-01
6.35596752e-01 -6.25682175e-01 -8.97527695e-01 4.35432345e-01
-2.00430602e-01 6.47874534e-01 3.30186039e-01 2.74917424e-01
-6.18633926e-01 -1.40583917e-01 -9.65074539e-01 1.93430170e-01
7.15585530e-01 -5.89627683e-01 -4.39225957e-02 2.31322885e-01
5.69235265e-01 -3.46846312e-01 -6.74162149e-01 2.83856601e-01
6.16157353e-01 -1.16374242e+00 8.88782620e-01 -7.04229832e-01
1.02734268e+00 -3.40779305e-01 -4.15304065e-01 -1.06441140e+00
-1.68909892e-01 -2.88118362e-01 -3.10614675e-01 8.29054356e-01
5.74912369e-01 -2.65039295e-01 8.85532677e-01 8.30117464e-01
1.40218541e-01 -9.83658135e-01 -9.40940678e-01 -4.83445197e-01
1.56398892e-01 -8.14933538e-01 1.50677934e-01 7.14232326e-01
-4.20615137e-01 2.10094854e-01 -2.92318910e-01 5.61406851e-01
8.44574630e-01 -3.14061120e-02 3.50527942e-01 -1.57533956e+00
-8.76808882e-01 -5.77014089e-01 -5.15876472e-01 -1.05368090e+00
5.12263596e-01 -7.29339778e-01 1.45689800e-01 -1.24561334e+00
2.22312033e-01 -5.68133175e-01 -3.68409038e-01 4.97903615e-01
3.40316534e-01 2.76323497e-01 7.94677436e-02 9.21206474e-02
-3.62448364e-01 3.83288622e-01 8.56643140e-01 -1.12246051e-01
8.84933472e-02 1.38276324e-01 -2.64029175e-01 8.16921234e-01
6.83249831e-01 -6.78073704e-01 -4.75600809e-02 -4.26071554e-01
8.64251196e-01 1.29195198e-01 8.21526527e-01 -8.72366965e-01
3.61729860e-01 -3.04481387e-01 5.86652756e-01 -3.13694030e-01
4.93485719e-01 -9.31445122e-01 5.26632667e-02 2.58601397e-01
-6.16720378e-01 -5.30723214e-01 -1.46852002e-01 5.78418434e-01
-8.59738663e-02 -6.10849082e-01 8.37275505e-01 -1.63779661e-01
-1.76636234e-01 4.09646004e-01 -1.24592753e-02 -1.22223027e-01
5.22751927e-01 1.45302057e-01 -2.55428068e-02 -3.24808836e-01
-1.03551233e+00 -2.16473034e-03 5.18019378e-01 5.19971224e-03
3.52427006e-01 -9.85316277e-01 -3.91430020e-01 1.49011642e-01
-9.18835029e-02 3.57385069e-01 -1.18031211e-01 6.26890719e-01
-2.78452605e-01 1.25598893e-01 -7.57487491e-02 -8.71728718e-01
-9.00980532e-01 4.98271853e-01 6.37463033e-01 -1.09453171e-01
-6.62935913e-01 8.00105333e-01 3.76144320e-01 -2.73267657e-01
3.19108009e-01 -5.61579883e-01 -5.49175106e-02 -2.66668126e-02
2.89963096e-01 -2.50875771e-01 -1.40004070e-03 -4.08192635e-01
-1.99902818e-01 6.97069287e-01 2.92320341e-01 -3.74874204e-01
1.57902813e+00 1.32550076e-01 -3.14916670e-01 8.03693950e-01
1.19414496e+00 -2.19175160e-01 -1.88740587e+00 -3.09333861e-01
-1.01024732e-01 -3.44204992e-01 2.16502592e-01 -6.70907378e-01
-1.23081636e+00 1.17318439e+00 1.33082733e-01 5.30759335e-01
1.03242230e+00 3.40212703e-01 8.50595236e-02 4.24033999e-01
2.70330638e-01 -8.53364229e-01 8.70929658e-02 3.94593626e-01
8.47036958e-01 -1.36451006e+00 -7.42940083e-02 -4.12360579e-01
-1.91091731e-01 1.54911005e+00 4.16007698e-01 -3.71818781e-01
6.97030127e-01 5.18820941e-01 -3.29459637e-01 -8.34816322e-02
-9.38082397e-01 -1.79057106e-01 4.51144814e-01 4.82250750e-01
3.69139642e-01 5.27468659e-02 2.59753883e-01 5.33337414e-01
-1.94667071e-01 -2.61894554e-01 4.12353069e-01 3.48471582e-01
-4.36347127e-01 -1.14643741e+00 -8.07795450e-02 3.43761384e-01
-4.63070422e-01 2.15970561e-01 -2.33309865e-01 5.16214490e-01
4.72725570e-01 6.98576510e-01 5.16094379e-02 -1.71439111e-01
-5.67646846e-02 -1.26542524e-02 7.97722816e-01 -7.03519523e-01
-7.62280673e-02 8.80250242e-03 -1.57025144e-01 -4.93391156e-01
-5.78687727e-01 -6.28481984e-01 -1.06704009e+00 -1.30594010e-02
-3.44220161e-01 -1.26891553e-01 1.01144290e+00 1.31417477e+00
-3.62932801e-01 5.37483990e-01 5.40742338e-01 -9.37373877e-01
-5.02587914e-01 -6.01551354e-01 -5.44427395e-01 6.87111616e-02
9.07082021e-01 -6.99652612e-01 -5.26774406e-01 2.33925790e-01] | [9.726974487304688, 0.7985353469848633] |
956c22de-6adc-4fae-82a1-684b44c268c1 | click-here-human-localized-keypoints-as | 1703.09859 | null | http://arxiv.org/abs/1703.09859v2 | http://arxiv.org/pdf/1703.09859v2.pdf | Click Here: Human-Localized Keypoints as Guidance for Viewpoint Estimation | We motivate and address a human-in-the-loop variant of the monocular
viewpoint estimation task in which the location and class of one semantic
object keypoint is available at test time. In order to leverage the keypoint
information, we devise a Convolutional Neural Network called Click-Here CNN
(CH-CNN) that integrates the keypoint information with activations from the
layers that process the image. It transforms the keypoint information into a 2D
map that can be used to weigh features from certain parts of the image more
heavily. The weighted sum of these spatial features is combined with global
image features to provide relevant information to the prediction layers. To
train our network, we collect a novel dataset of 3D keypoint annotations on
thousands of CAD models, and synthetically render millions of images with 2D
keypoint information. On test instances from PASCAL 3D+, our model achieves a
mean class accuracy of 90.7%, whereas the state-of-the-art baseline only
obtains 85.7% mean class accuracy, justifying our argument for
human-in-the-loop inference. | ['Ryan Szeto', 'Jason J. Corso'] | 2017-03-29 | click-here-human-localized-keypoints-as-1 | http://openaccess.thecvf.com/content_iccv_2017/html/Szeto_Click_Here_Human-Localized_ICCV_2017_paper.html | http://openaccess.thecvf.com/content_ICCV_2017/papers/Szeto_Click_Here_Human-Localized_ICCV_2017_paper.pdf | iccv-2017-10 | ['viewpoint-estimation'] | ['computer-vision'] | [-6.15192652e-02 6.62211627e-02 -2.12452725e-01 -7.68615186e-01
-8.13335538e-01 -5.93654335e-01 6.04340613e-01 2.46006083e-02
-3.44113618e-01 3.47256005e-01 1.53028360e-02 -1.98796913e-01
3.03964406e-01 -7.81215370e-01 -1.24545348e+00 -1.58161610e-01
2.17112228e-01 5.29821217e-01 4.15333331e-01 2.12612912e-01
4.01663631e-01 7.15314507e-01 -1.64781249e+00 4.54851359e-01
3.88740540e-01 1.55352032e+00 3.28886062e-01 7.84238219e-01
1.28128096e-01 8.71907771e-01 -4.80624437e-01 -3.10693771e-01
4.90262836e-01 1.55949816e-01 -5.93096197e-01 -1.57150939e-01
1.25479949e+00 -8.14512670e-01 -3.97112191e-01 9.14339542e-01
3.54846627e-01 -7.61258826e-02 5.21899104e-01 -1.29174423e+00
-5.55183470e-01 -1.46805774e-02 -4.57018167e-01 1.25382394e-01
4.12553400e-01 3.77849042e-01 1.19977665e+00 -1.38087213e+00
1.01015472e+00 1.31454360e+00 5.54335415e-01 2.05314398e-01
-1.01425576e+00 -5.71441948e-01 3.23179185e-01 8.36958736e-02
-1.18343544e+00 -2.11222023e-01 6.98033988e-01 -5.55001616e-01
1.24833107e+00 1.01074748e-01 8.18981051e-01 8.15276682e-01
2.74069935e-01 1.10425723e+00 7.65761077e-01 -6.91440776e-02
1.27985388e-01 -2.02502936e-01 -1.61177322e-01 8.41051519e-01
9.60204452e-02 1.80185348e-01 -7.10092247e-01 3.32111716e-02
1.05838382e+00 2.93310225e-01 7.84037262e-02 -7.95532763e-01
-1.46584523e+00 6.08134627e-01 1.02377617e+00 -3.16595823e-01
-4.30370957e-01 5.97871363e-01 -4.16833498e-02 1.20417066e-01
5.48800290e-01 6.31733537e-01 -8.87503684e-01 -6.16573310e-03
-7.67675042e-01 7.72362053e-01 5.35927057e-01 9.45099235e-01
9.32469547e-01 -3.78326267e-01 -1.09499745e-01 4.74562436e-01
3.44149560e-01 5.55056095e-01 -3.01064588e-02 -1.50888252e+00
6.13566875e-01 8.83053780e-01 2.98080772e-01 -9.13847387e-01
-3.09819847e-01 -4.57901895e-01 -3.02409142e-01 6.47895157e-01
3.52413327e-01 2.36182913e-01 -1.07176280e+00 1.45684814e+00
4.82081801e-01 2.10727621e-02 -4.66942549e-01 1.09837723e+00
7.12501466e-01 5.02891183e-01 -9.43697914e-02 6.22017860e-01
1.17781103e+00 -1.00062919e+00 -8.74222368e-02 -4.57342774e-01
6.13563418e-01 -8.15233350e-01 1.07554448e+00 3.08031142e-01
-1.13401234e+00 -7.35088289e-01 -1.11951816e+00 -5.12478530e-01
-6.41933262e-01 2.18353063e-01 5.44577718e-01 -1.26888663e-01
-1.09250093e+00 6.23902619e-01 -7.68512726e-01 -1.11182094e-01
7.00655282e-01 1.90615103e-01 -5.08813500e-01 -3.61610681e-01
-8.40209484e-01 1.15314388e+00 5.41562624e-02 -2.98444498e-02
-1.13847792e+00 -1.04872477e+00 -1.08378756e+00 -3.89981382e-02
3.82654816e-01 -8.65194738e-01 1.53608429e+00 -6.72802091e-01
-1.05984056e+00 1.14858270e+00 -1.59094498e-01 -2.62254268e-01
8.10237586e-01 -6.83094680e-01 7.20831305e-02 2.21672386e-01
4.37851757e-01 1.46802628e+00 8.32588613e-01 -1.20866334e+00
-1.00118887e+00 -4.22190994e-01 3.94505173e-01 2.27621228e-01
4.48338658e-01 -2.70351410e-01 -7.16025770e-01 -5.40685475e-01
3.76029432e-01 -9.60379720e-01 -2.66454034e-02 7.48351634e-01
-5.94797254e-01 -3.27543199e-01 7.86954463e-01 -4.85961437e-01
5.44478476e-01 -1.96604478e+00 4.53803726e-02 -4.41892110e-02
3.78940880e-01 -1.16411164e-01 -1.99153051e-01 2.18318701e-02
-9.99206156e-02 -1.28690824e-01 2.26546884e-01 -3.87767166e-01
1.61775544e-01 -4.72904444e-02 -2.45495483e-01 2.98499554e-01
5.93284488e-01 1.33797765e+00 -9.90266740e-01 -2.53516227e-01
5.68583786e-01 5.26823103e-01 -9.78488386e-01 3.82714212e-01
-6.04107320e-01 1.40887126e-01 -4.68737900e-01 9.79679286e-01
5.79477012e-01 -5.30968249e-01 -2.94595927e-01 -5.43659866e-01
-5.50699653e-03 6.22766197e-01 -9.05914009e-01 2.00804758e+00
-3.58478516e-01 8.25088620e-01 -2.66020298e-01 -4.13940638e-01
6.70715392e-01 -1.30575091e-01 2.33284816e-01 -7.74546742e-01
-2.76967678e-02 1.77134752e-01 -2.69234836e-01 -2.24151880e-01
6.45115852e-01 4.88859028e-01 -8.40848833e-02 2.13249758e-01
2.00055361e-01 -5.14064252e-01 -1.99724272e-01 2.51980811e-01
1.08340514e+00 6.61515892e-01 1.58332318e-01 -3.22949104e-02
-3.62725481e-02 3.30522954e-02 2.10072115e-01 7.16568947e-01
-1.16027251e-01 8.17886353e-01 5.97775161e-01 -1.01458275e+00
-1.25451338e+00 -1.20411813e+00 8.17919225e-02 9.10232842e-01
2.08041057e-01 -3.78507048e-01 -5.30376375e-01 -9.07225490e-01
4.94310051e-01 6.30133927e-01 -9.20570016e-01 -4.30221297e-02
-3.82668316e-01 5.04578538e-02 1.17030703e-01 9.20471609e-01
6.01842582e-01 -7.98914611e-01 -1.05033422e+00 1.73362698e-02
3.32216658e-02 -1.17898917e+00 -2.52927542e-01 2.33416185e-01
-6.63456380e-01 -1.27403295e+00 -6.19964778e-01 -4.86527234e-01
7.23574758e-01 3.15809339e-01 1.60414970e+00 1.65303703e-02
-3.54149073e-01 2.33987361e-01 -1.12026319e-01 -3.73855412e-01
-1.07978713e-02 1.50769010e-01 -2.80773848e-01 -4.08085674e-01
5.90933383e-01 -4.28582519e-01 -9.46645200e-01 4.44465905e-01
-3.89476955e-01 4.12345678e-01 6.51341319e-01 6.18248284e-01
8.21373463e-01 -4.52139646e-01 1.22092262e-01 -6.68065071e-01
-2.36356810e-01 -1.99091405e-01 -1.01942408e+00 3.38689773e-03
-2.22645730e-01 8.86338279e-02 9.45347548e-02 -1.42716214e-01
-6.22276783e-01 5.19737601e-01 5.25293685e-02 -8.32987130e-01
-1.86644867e-01 2.34724954e-01 -2.90890634e-01 -5.04644550e-02
5.23569405e-01 -1.61095768e-01 -3.12281519e-01 -4.34470981e-01
7.35207677e-01 2.90606439e-01 7.00654566e-01 -4.68446076e-01
8.52930725e-01 4.08357829e-01 8.82427245e-02 -1.67961165e-01
-1.38302147e+00 -3.14151675e-01 -1.00483763e+00 -1.86741456e-01
1.20401323e+00 -1.36392975e+00 -7.66859651e-01 3.82366598e-01
-1.33983445e+00 -6.05354011e-01 -2.84713447e-01 3.80603254e-01
-7.12241650e-01 -4.38127786e-01 -3.97412241e-01 -4.15784508e-01
9.01378617e-02 -1.29658914e+00 1.80252171e+00 -4.78244163e-02
-3.29753637e-01 -5.23180723e-01 -3.41904491e-01 2.49210224e-01
2.60909259e-01 3.18400532e-01 8.67625594e-01 -4.84672248e-01
-1.36984885e+00 -5.51270843e-01 -6.59702361e-01 2.41706610e-01
-1.76267847e-01 -2.31584772e-01 -1.32095838e+00 1.46637280e-02
-3.00019264e-01 -5.44687748e-01 8.10158074e-01 3.59720975e-01
1.53237045e+00 1.48158055e-02 -4.05458927e-01 8.18968475e-01
1.32795644e+00 -1.33802071e-01 4.21478063e-01 2.55711824e-01
9.94808316e-01 2.34212875e-01 8.17738116e-01 1.87910110e-01
7.11801708e-01 7.18391597e-01 9.31880176e-01 -7.32526407e-02
-7.65372589e-02 -6.21728778e-01 -1.35790154e-01 2.79316455e-01
2.89341509e-01 -1.12399630e-01 -9.07628119e-01 3.53936523e-01
-1.77074087e+00 -6.46172941e-01 3.48067611e-01 2.02967048e+00
6.36647344e-01 5.15229166e-01 -2.93397427e-01 -2.37089053e-01
3.69860470e-01 3.60031933e-01 -8.15735698e-01 -2.07148911e-03
5.04147895e-02 3.84236081e-03 6.54505253e-01 5.87349236e-01
-1.32355988e+00 9.47074771e-01 6.46477413e+00 3.58883530e-01
-1.16992891e+00 -2.60544360e-01 9.11165774e-01 -3.18724155e-01
-2.25407332e-01 2.66249347e-02 -8.55980873e-01 1.88291073e-01
5.71810603e-01 3.30267966e-01 2.34043494e-01 1.30694067e+00
-1.17106162e-01 -3.16697180e-01 -1.70711684e+00 1.17958176e+00
1.69327497e-01 -1.58428013e+00 1.53751418e-01 2.53621191e-01
8.74321461e-01 5.23353577e-01 1.70967966e-01 1.42613798e-01
4.93428916e-01 -1.11506593e+00 9.34800804e-01 5.85357726e-01
9.78144109e-01 -5.19914150e-01 5.51711023e-01 2.57663310e-01
-1.00919461e+00 -3.59733822e-03 -3.62165600e-01 -2.13646784e-01
8.38802569e-03 5.29282510e-01 -1.04976654e+00 1.83430672e-01
8.19064021e-01 9.91036117e-01 -8.35262597e-01 9.33597565e-01
-4.25386429e-01 9.71617270e-03 -3.92678797e-01 4.83969375e-02
2.32271239e-01 4.19998378e-01 2.69807070e-01 7.06819415e-01
3.78088623e-01 -2.71273088e-02 1.01922199e-01 1.14735639e+00
-4.43594724e-01 -4.28145200e-01 -7.12257028e-01 1.65141031e-01
8.06695282e-01 1.19819140e+00 -5.21383882e-01 -6.15091860e-01
-4.83695090e-01 9.91336823e-01 4.61637348e-01 2.03282908e-01
-6.66957915e-01 -3.84019762e-01 8.82188797e-01 2.81498104e-01
5.74556887e-01 -3.42129558e-01 -3.37655693e-01 -1.09002078e+00
1.04197316e-01 -5.87684333e-01 1.84240974e-02 -1.61032093e+00
-1.11826444e+00 3.11424732e-01 -4.26669717e-02 -1.30250442e+00
-3.72049063e-01 -1.07634568e+00 -4.48880821e-01 1.01292133e+00
-1.56044686e+00 -1.48006904e+00 -7.63030887e-01 3.05713832e-01
5.31377912e-01 -6.48138225e-02 7.27160156e-01 6.76686615e-02
1.22007035e-01 3.55421990e-01 -4.76519853e-01 1.93451062e-01
7.08076000e-01 -1.35645521e+00 1.16878724e+00 4.98193294e-01
3.06337476e-02 5.82829475e-01 4.28846449e-01 -5.64624310e-01
-1.37732232e+00 -1.21114707e+00 9.44341481e-01 -1.31004941e+00
3.50419313e-01 -5.61517179e-01 -4.99540120e-01 8.92861664e-01
-2.45195627e-01 6.11752391e-01 2.09513113e-01 -1.01900510e-01
-7.13322639e-01 3.32817957e-02 -8.31510425e-01 6.93225920e-01
1.12539279e+00 -8.51752579e-01 -6.76901400e-01 2.62868077e-01
1.14445090e+00 -7.37660289e-01 -7.43958294e-01 4.27165359e-01
8.29250634e-01 -9.55478251e-01 1.26340580e+00 -6.13395691e-01
8.63902509e-01 -3.61503899e-01 -7.37551630e-01 -1.24710596e+00
-2.13372201e-01 -8.89206603e-02 -1.47339582e-01 4.75843489e-01
4.45864141e-01 -3.17492157e-01 9.33564723e-01 7.35560238e-01
-1.08242884e-01 -8.13694537e-01 -8.20560098e-01 -3.30418110e-01
-8.47212896e-02 -5.66015780e-01 7.79379725e-01 6.93157256e-01
-5.18232763e-01 3.64270210e-01 -1.12987146e-01 6.50886819e-02
5.38145900e-01 2.47985408e-01 1.22734129e+00 -1.32604897e+00
-1.04795083e-01 -1.61761314e-01 -7.09803939e-01 -1.46237814e+00
8.64580944e-02 -6.76262856e-01 2.49838661e-02 -1.51071000e+00
1.45259008e-01 -4.55841243e-01 -1.96177825e-01 4.94876802e-01
-2.59252220e-01 4.17176962e-01 3.97743642e-01 3.89731466e-03
-8.78809094e-01 2.53268719e-01 1.43157637e+00 -9.14817527e-02
1.68190166e-01 -2.76036084e-01 -5.27915180e-01 8.27840388e-01
2.82549679e-01 -2.75645316e-01 -2.94630945e-01 -8.36035550e-01
5.65149724e-01 4.62926226e-03 9.85082030e-01 -1.04166663e+00
1.81262523e-01 -8.49375799e-02 1.28409934e+00 -1.26313031e+00
8.10443997e-01 -9.25162494e-01 -1.96787938e-01 2.43058890e-01
-5.72478414e-01 2.31115341e-01 -5.72045967e-02 3.41978937e-01
-1.07779466e-01 3.35659921e-01 5.36399364e-01 -3.36058676e-01
-8.96785319e-01 6.28509104e-01 3.22218627e-01 -1.43430885e-02
7.51286685e-01 -1.07606448e-01 -4.72835213e-01 -4.17501718e-01
-5.37509620e-01 1.96148515e-01 8.40375364e-01 6.28821731e-01
7.24106371e-01 -1.51461816e+00 -4.32238162e-01 4.75296646e-01
4.19009209e-01 5.42063117e-01 3.84719893e-02 4.22259510e-01
-7.80007184e-01 5.63915074e-01 -2.32902154e-01 -9.87752616e-01
-8.41017544e-01 5.39703012e-01 4.40655768e-01 1.81325987e-01
-6.13906741e-01 1.06008172e+00 5.45972288e-01 -6.50157809e-01
1.84559479e-01 -6.36755645e-01 3.17709178e-01 -1.67958528e-01
4.88455445e-01 2.48138104e-02 1.00849375e-01 -4.15509224e-01
-4.85494941e-01 7.35359251e-01 4.03194083e-03 -2.50628442e-01
1.21746337e+00 1.95849970e-01 9.20319557e-02 5.29358566e-01
1.54894996e+00 -2.05405265e-01 -1.82715809e+00 -3.41443866e-01
-2.25768715e-01 -8.57795238e-01 3.92352678e-02 -1.08539557e+00
-8.43724847e-01 1.13347208e+00 5.83280444e-01 -2.62324840e-01
5.74049532e-01 1.14241257e-01 6.40729249e-01 6.46608353e-01
4.62909728e-01 -1.00033689e+00 2.62790054e-01 5.75729311e-01
9.26966131e-01 -1.56788039e+00 4.87298630e-02 -1.98882326e-01
-4.83328849e-01 9.07187760e-01 9.10597980e-01 -4.73391742e-01
7.00862765e-01 1.49376184e-01 1.86894417e-01 -4.06357080e-01
-1.08648789e+00 1.15038775e-01 5.28286874e-01 5.55728734e-01
2.98911005e-01 5.98530732e-02 7.07403183e-01 1.96074411e-01
-5.01588821e-01 1.10600457e-01 -4.55802605e-02 9.24408913e-01
-6.32814109e-01 -5.46335518e-01 -3.02881431e-02 4.85612333e-01
-2.03029975e-01 -9.80343968e-02 -4.69735175e-01 7.95471370e-01
1.06255114e-01 4.29162025e-01 4.09044206e-01 -4.37106401e-01
4.03125435e-01 2.87130307e-02 4.10988599e-01 -5.33370793e-01
-1.68688148e-01 -1.63782045e-01 4.50966917e-02 -1.21598160e+00
-9.33700651e-02 -3.30328912e-01 -9.55155671e-01 -1.49142608e-01
-4.26006652e-02 -3.40045869e-01 9.78718340e-01 8.53493392e-01
4.58227307e-01 7.04165637e-01 3.82761985e-01 -1.48289037e+00
-3.75171691e-01 -8.09595942e-01 -2.63810121e-02 4.30666625e-01
5.80049932e-01 -7.52892911e-01 -1.56648293e-01 -5.97116351e-02] | [7.626429080963135, -2.59311580657959] |
5c82cbbd-f690-473b-b30d-dc3742935962 | device-tuning-for-multi-task-large-model | 2302.1082 | null | https://arxiv.org/abs/2302.10820v1 | https://arxiv.org/pdf/2302.10820v1.pdf | Device Tuning for Multi-Task Large Model | Unsupervised pre-training approaches have achieved great success in many fields such as Computer Vision (CV), Natural Language Processing (NLP) and so on. However, compared to typical deep learning models, pre-training or even fine-tuning the state-of-the-art self-attention models is extremely expensive, as they require much more computational and memory resources. It severely limits their applications and success in a variety of domains, especially for multi-task learning. To improve the efficiency, we propose Device Tuning for the efficient multi-task model, which is a massively multitask framework across the cloud and device and is designed to encourage learning of representations that generalize better to many different tasks. Specifically, we design Device Tuning architecture of a multi-task model that benefits both cloud modelling and device modelling, which reduces the communication between device and cloud by representation compression. Experimental results demonstrate the effectiveness of our proposed method. | ['Yinsi Zhou', 'Xuanchen Hou', 'Penghao Jiang'] | 2023-02-21 | null | null | null | null | ['unsupervised-pre-training'] | ['methodology'] | [ 2.95936763e-01 -4.71968859e-01 -2.45895162e-01 -3.32977623e-01
-4.83933479e-01 -1.29435226e-01 2.57773370e-01 -8.90804827e-02
-1.07109714e-02 4.80390489e-01 6.29576817e-02 -1.98069125e-01
-1.46993801e-01 -6.23368382e-01 -5.81092834e-01 -5.73994279e-01
4.89562541e-01 2.81430870e-01 2.38901079e-01 2.22333029e-01
9.13793519e-02 2.64650047e-01 -1.73697829e+00 3.86252195e-01
9.57384169e-01 1.23267996e+00 8.96149814e-01 3.81047964e-01
-1.66961327e-01 6.88592553e-01 -5.35290539e-01 -3.12205434e-01
-2.61137158e-01 1.17881671e-01 -7.81182230e-01 -1.21983744e-01
1.73724100e-01 -2.10605502e-01 -6.44645914e-02 1.04655206e+00
9.33120787e-01 7.60645717e-02 5.96721828e-01 -1.29829919e+00
-1.26106334e+00 1.51247859e-01 -7.15083480e-01 3.75115871e-01
-2.43055835e-01 -5.98173141e-02 5.46464801e-01 -8.60270023e-01
-1.28072172e-01 1.09652925e+00 5.41116297e-01 7.19297826e-01
-6.63099110e-01 -9.49012816e-01 2.89070219e-01 3.04945260e-01
-1.42535746e+00 -4.17484283e-01 7.49604404e-01 -2.68595427e-01
1.29332483e+00 3.87730226e-02 1.57321006e-01 9.74598229e-01
1.20201729e-01 1.01661897e+00 9.25772727e-01 -3.82498622e-01
9.20171142e-02 4.47525769e-01 -7.04874247e-02 4.30778086e-01
3.02221954e-01 -4.71798301e-01 -4.01401550e-01 -1.59530461e-01
9.25064921e-01 7.55777836e-01 8.38278309e-02 -6.58099800e-02
-7.40837991e-01 6.86074734e-01 2.67623872e-01 5.46537399e-01
-6.58628702e-01 3.38087052e-01 8.12804699e-01 -2.11260375e-02
6.95950806e-01 2.39274036e-02 -7.24725187e-01 -2.12020338e-01
-7.93861151e-01 -2.11139187e-01 1.62314683e-01 1.28719568e+00
6.60083115e-01 3.09416175e-01 -2.30296835e-01 1.12177706e+00
7.44057000e-02 5.41370034e-01 9.87775564e-01 -5.36753476e-01
4.19074923e-01 6.07823670e-01 -2.02858284e-01 -8.37896228e-01
-2.12077186e-01 -6.57544732e-01 -1.30556726e+00 -3.13881099e-01
-6.40206099e-01 -3.51853907e-01 -6.96409643e-01 1.53673720e+00
8.85502696e-02 5.99423766e-01 2.62581930e-03 7.95308292e-01
9.81930614e-01 8.35003257e-01 4.59207922e-01 -3.25835973e-01
1.39832270e+00 -1.44568932e+00 -8.33645046e-01 -4.61534888e-01
6.68603480e-01 -8.59902024e-01 1.43168950e+00 3.92093629e-01
-9.78644729e-01 -1.05707455e+00 -7.94118106e-01 -1.25016138e-01
-4.13602740e-01 3.23787361e-01 1.04785717e+00 7.06541955e-01
-9.84238923e-01 4.14516807e-01 -5.99132657e-01 -4.61929232e-01
8.98179889e-01 5.66784680e-01 1.58754990e-01 -1.58548295e-01
-8.45428467e-01 5.48784137e-01 2.61515081e-01 -1.84266239e-01
-7.98133254e-01 -6.24901175e-01 -5.03905892e-01 4.91194367e-01
2.12082714e-01 -8.29413235e-01 1.19266140e+00 -1.18401217e+00
-1.56882906e+00 6.36095941e-01 -3.07358027e-01 -2.27497488e-01
-2.34063834e-01 -5.85147142e-01 -7.35534191e-01 -3.52462202e-01
-3.37808989e-02 4.74819183e-01 1.20805371e+00 -8.61975372e-01
-7.98910916e-01 -4.57451195e-01 -7.67322332e-02 2.58890748e-01
-1.21370888e+00 2.97287881e-01 -6.40647829e-01 -6.26669526e-01
-5.96373498e-01 -7.39005208e-01 -2.88201183e-01 -3.56029212e-01
-1.92738652e-01 -5.70093453e-01 1.46954119e+00 -2.96390712e-01
1.41299343e+00 -2.35166550e+00 -3.80872428e-01 -2.78855711e-01
3.19913238e-01 7.55629361e-01 -5.93841262e-02 -3.87998335e-02
4.36843559e-02 9.59264487e-02 3.97291452e-01 -4.92975831e-01
-2.88679302e-01 1.70651391e-01 -2.51870573e-01 5.78862168e-02
1.81156933e-01 9.66824532e-01 -5.68902552e-01 -5.21941781e-01
2.37977505e-01 6.80192828e-01 -5.88035226e-01 2.21390590e-01
-1.73746720e-01 3.75723273e-01 -8.63393009e-01 5.47579706e-01
7.67623007e-01 -9.12656248e-01 -1.87402904e-01 -2.51234621e-01
1.59285247e-01 1.32586345e-01 -8.36259484e-01 2.00157595e+00
-1.14300060e+00 3.28381658e-01 -9.53648090e-02 -1.19139135e+00
9.49644744e-01 3.55442733e-01 5.96558332e-01 -8.58059049e-01
2.08474174e-01 8.91921744e-02 -3.21358979e-01 -6.76676333e-01
5.51030457e-01 1.65918842e-01 -1.07997116e-02 4.58667666e-01
1.55671299e-01 3.22409093e-01 -5.26621759e-01 5.34123741e-03
7.19591379e-01 -1.05555259e-01 1.61834300e-01 -3.38139027e-01
5.37525952e-01 -2.90748477e-01 5.22457957e-01 3.49831820e-01
-4.64472733e-02 1.86147422e-01 -3.73384446e-01 -4.96295393e-01
-8.57063353e-01 -6.52232349e-01 -3.38783041e-02 1.67668462e+00
2.07439870e-01 -4.33704346e-01 -7.76677668e-01 -4.30621445e-01
-1.31777510e-01 5.59184670e-01 -3.08529317e-01 -3.70708019e-01
-4.53361154e-01 -7.76347995e-01 3.23159277e-01 8.06913793e-01
8.11440349e-01 -1.19975829e+00 -7.19697356e-01 2.91866004e-01
1.92872677e-02 -1.21512353e+00 -5.66892326e-01 1.74516290e-01
-1.03345668e+00 -4.79256004e-01 -6.64302230e-01 -1.18245161e+00
4.33513105e-01 7.31491685e-01 9.85212922e-01 2.83718854e-01
-2.61628568e-01 3.18935364e-01 -2.42431521e-01 -7.82206357e-01
1.28085122e-01 4.60265040e-01 3.54600161e-01 1.47568494e-01
6.30491555e-01 -8.05499971e-01 -5.54142296e-01 1.82688028e-01
-8.70637655e-01 7.45572299e-02 8.38010490e-01 7.68906116e-01
8.68809879e-01 3.02932411e-01 8.95177603e-01 -1.06515467e+00
8.85971248e-01 -6.75968707e-01 -3.68561000e-01 4.04009759e-01
-7.86366284e-01 -1.16482526e-01 1.07428372e+00 -8.13506603e-01
-1.08890700e+00 -2.15041582e-02 5.16337231e-02 -9.41569507e-01
-1.77225024e-01 2.18564287e-01 -2.28682935e-01 -1.38863847e-01
3.64142925e-01 4.89746273e-01 -3.99813265e-01 -7.90281534e-01
3.37947607e-01 1.31332207e+00 5.25310278e-01 -6.61462367e-01
3.07220101e-01 3.05714756e-01 -1.71748415e-01 -7.31159031e-01
-7.32530653e-01 -4.85268533e-01 -2.60717273e-01 7.88528472e-02
8.95894229e-01 -1.22070539e+00 -9.32034791e-01 2.01678425e-01
-1.23114359e+00 -1.98108509e-01 -1.61811173e-01 3.68681252e-01
-4.96516973e-01 1.40448576e-02 -3.21808189e-01 -8.53819966e-01
-1.02912652e+00 -1.18239248e+00 1.30883145e+00 6.23170495e-01
2.77297944e-01 -1.05971193e+00 -3.13884288e-01 1.66718945e-01
1.05253589e+00 -4.18005019e-01 1.05297756e+00 -5.39917707e-01
-5.66273928e-01 5.99635839e-02 -6.25908375e-01 4.18874353e-01
2.85237581e-01 -4.53852296e-01 -1.28438723e+00 -4.05798793e-01
2.58468062e-01 -5.72772920e-01 5.27040601e-01 4.95840907e-01
2.03914046e+00 -2.05190390e-01 -7.02942729e-01 6.86047912e-01
1.56396818e+00 1.13758706e-01 4.20569628e-01 6.09513968e-02
9.58247542e-01 2.63243556e-01 3.88847440e-01 4.86148387e-01
3.78446430e-01 7.12054014e-01 3.73023331e-01 -1.98527917e-01
-1.39665484e-01 -1.20297387e-01 1.51786372e-01 1.32058632e+00
-2.37975374e-01 -7.03857988e-02 -6.35277390e-01 6.86302006e-01
-1.90684164e+00 -6.50405824e-01 1.23925813e-01 2.04758358e+00
4.83848065e-01 -5.90982400e-02 -2.80378833e-02 -1.97498307e-01
8.04889321e-01 3.94609272e-02 -9.09649849e-01 -5.90177417e-01
2.24712744e-01 3.76967937e-01 3.33904028e-01 -2.95810491e-01
-1.03256226e+00 1.11071289e+00 6.17560530e+00 1.49264359e+00
-1.43312740e+00 8.77560377e-01 7.87514627e-01 -2.73639679e-01
-5.95716089e-02 -4.63506132e-01 -1.01903164e+00 6.55790031e-01
1.03207350e+00 -2.77085721e-01 5.44226110e-01 1.53634906e+00
-3.11061367e-02 2.86726713e-01 -8.57533991e-01 1.66108882e+00
1.53785825e-01 -1.43299210e+00 3.90600860e-02 -1.19362473e-01
8.84996831e-01 3.04913998e-01 3.48111838e-01 7.50688612e-01
1.06322043e-01 -8.77269268e-01 2.49668539e-01 1.21772282e-01
1.17643642e+00 -7.86836088e-01 6.28968775e-01 6.13996685e-01
-1.36991870e+00 -2.54941165e-01 -7.71457553e-01 -1.42216478e-02
-3.94801870e-02 7.42618918e-01 -3.92673045e-01 4.12191004e-01
1.04527807e+00 5.68887293e-01 -5.13696432e-01 7.53315330e-01
3.87416393e-01 5.25824845e-01 -1.07017428e-01 -2.35942930e-01
7.78622404e-02 9.29871202e-02 -1.11709982e-01 1.14986444e+00
4.97401446e-01 3.09106056e-02 3.92542511e-01 5.55013776e-01
-3.94243330e-01 4.09987748e-01 -5.91826558e-01 1.00456970e-02
6.35876656e-01 1.33526671e+00 -3.62402707e-01 -5.29531002e-01
-6.91649675e-01 1.22033823e+00 3.94520640e-01 1.59971178e-01
-9.49838102e-01 -2.67769873e-01 7.02488244e-01 -1.60811432e-02
3.27945352e-01 7.47038573e-02 -3.93302739e-01 -1.01968014e+00
1.27924114e-01 -5.85565388e-01 1.78775772e-01 -8.15803111e-01
-1.42167711e+00 8.16726506e-01 -3.68321955e-01 -1.21243513e+00
-1.60015747e-01 -4.02073681e-01 -7.53911197e-01 9.76589561e-01
-1.68067527e+00 -1.37005985e+00 -4.56945121e-01 9.86398041e-01
8.66271973e-01 -4.99885112e-01 1.03213477e+00 7.76741922e-01
-7.05182135e-01 7.74023652e-01 1.80591673e-01 -3.33618343e-01
6.62370026e-01 -6.55081272e-01 3.06106210e-01 4.47655678e-01
1.70038342e-01 6.17144823e-01 1.75704449e-01 -3.84616226e-01
-1.53848743e+00 -1.58318484e+00 4.41433966e-01 -5.87045252e-02
4.16545451e-01 -3.91819775e-01 -8.11908126e-01 5.74221313e-01
2.88291961e-01 4.23511714e-01 8.02522063e-01 2.65625745e-01
-1.63568810e-01 -3.16221505e-01 -9.43699896e-01 2.89752126e-01
1.09363806e+00 -7.71001697e-01 9.58292857e-02 8.38427782e-01
1.05560970e+00 -4.80067134e-02 -8.45100164e-01 3.16700548e-01
1.62399501e-01 -5.87894917e-01 8.79698873e-01 -8.14918697e-01
2.31545642e-01 -3.10492199e-02 -3.74130487e-01 -1.12790132e+00
-6.96368814e-01 -5.83266616e-01 -4.69057709e-01 1.41310048e+00
1.21870123e-01 -4.51734602e-01 9.84306395e-01 3.36982727e-01
-2.85034537e-01 -1.20455337e+00 -6.58600926e-01 -9.06493664e-01
-1.42084867e-01 -4.71861154e-01 9.02437449e-01 1.08445203e+00
-3.39948714e-01 8.77237558e-01 -7.08442748e-01 3.52353007e-02
2.89544225e-01 3.18697214e-01 6.67102098e-01 -1.41488516e+00
-6.29349470e-01 -3.47337097e-01 -2.27180809e-01 -1.39990103e+00
-1.85466502e-02 -7.41766512e-01 -2.23356023e-01 -1.36344707e+00
4.45608079e-01 -8.42940271e-01 -4.74957794e-01 4.54627812e-01
-2.62680411e-01 3.03495675e-02 8.28522071e-02 5.39581895e-01
-9.80783939e-01 6.73099756e-01 1.17770350e+00 -6.59060478e-02
-1.85971662e-01 7.06187710e-02 -1.07594192e+00 8.05808306e-01
7.93795109e-01 -6.09951317e-01 -7.12478161e-01 -1.03377819e+00
-1.00654274e-01 -1.90915793e-01 3.45495716e-02 -1.15553224e+00
3.23340267e-01 -2.71833669e-02 3.69835466e-01 -3.83507133e-01
4.94773000e-01 -1.10267162e+00 -9.55594257e-02 2.54717708e-01
-1.21392393e-02 2.56766051e-01 3.43669444e-01 5.51052690e-01
-3.17964666e-02 -1.43006176e-01 6.17460132e-01 1.05572544e-01
-8.71959209e-01 7.08175659e-01 1.34698510e-01 -8.04226920e-02
1.18314695e+00 4.24291454e-02 -3.52577895e-01 -2.31077343e-01
-4.53291297e-01 6.61843941e-02 4.97451201e-02 6.47221565e-01
6.19299173e-01 -1.48251605e+00 -4.73556101e-01 2.82944143e-01
1.20709635e-01 1.34497195e-01 6.93929195e-01 6.82574630e-01
4.86578010e-02 5.90451658e-01 -1.66839778e-01 -6.79746032e-01
-1.19582534e+00 9.38555837e-01 -1.68458134e-01 -5.69637656e-01
-4.03647065e-01 8.92935991e-01 5.04311740e-01 -2.40456268e-01
2.25099012e-01 -1.48088383e-02 -2.80702561e-01 -5.58361530e-01
7.97248483e-01 2.11818814e-01 2.04390898e-01 -4.67146218e-01
-2.76904881e-01 5.41771114e-01 -4.93769109e-01 6.04652762e-01
1.29261053e+00 -1.99017227e-01 -6.29797876e-02 2.50927985e-01
1.35251713e+00 -2.65201688e-01 -8.60080242e-01 -5.29993951e-01
-2.47611195e-01 -3.40700984e-01 4.28829163e-01 -5.72979152e-01
-1.37667751e+00 1.20233846e+00 9.65135694e-01 1.75446913e-01
1.56250691e+00 -2.21430790e-02 1.19359338e+00 4.94756609e-01
6.70085490e-01 -1.17597437e+00 2.30121240e-01 3.43644947e-01
4.45969939e-01 -1.23881483e+00 -1.04860432e-01 -4.09369975e-01
-7.68651187e-01 6.91578925e-01 9.48704541e-01 -1.14211954e-01
9.73053157e-01 4.04811412e-01 -3.33306640e-01 -1.50430739e-01
-8.15844119e-01 -8.79063532e-02 9.88171101e-02 9.72459614e-01
4.64410603e-01 1.78425089e-01 1.56640023e-01 1.08832121e+00
2.40911797e-01 5.43914258e-01 -1.52362380e-02 6.36377037e-01
-4.13684875e-01 -1.02701271e+00 -1.17725216e-01 8.55211139e-01
-6.92717195e-01 -3.94172907e-01 2.78280467e-01 1.63789168e-01
4.53084201e-01 8.87984574e-01 1.62801832e-01 -7.01205194e-01
1.57676905e-01 -1.32792398e-01 1.49722442e-01 -7.63076305e-01
-6.10442102e-01 7.83931240e-02 -2.26764947e-01 -4.55227584e-01
-4.23019171e-01 -2.14185804e-01 -1.09142005e+00 -3.65865320e-01
-5.32903016e-01 -2.91591845e-02 9.78481114e-01 1.00071394e+00
1.10968184e+00 1.02449512e+00 7.32714593e-01 -8.16023290e-01
-3.92523408e-01 -1.08183718e+00 -5.31739295e-01 2.73574322e-01
-1.61230206e-01 -7.63136029e-01 4.03898180e-01 7.48293400e-02] | [8.959616661071777, 3.0354721546173096] |
1f23f771-d8da-4724-b38f-7b026aa00394 | temporal-action-proposal-generation-with | 2105.12043 | null | https://arxiv.org/abs/2105.12043v1 | https://arxiv.org/pdf/2105.12043v1.pdf | Temporal Action Proposal Generation with Transformers | Transformer networks are effective at modeling long-range contextual information and have recently demonstrated exemplary performance in the natural language processing domain. Conventionally, the temporal action proposal generation (TAPG) task is divided into two main sub-tasks: boundary prediction and proposal confidence prediction, which rely on the frame-level dependencies and proposal-level relationships separately. To capture the dependencies at different levels of granularity, this paper intuitively presents a unified temporal action proposal generation framework with original Transformers, called TAPG Transformer, which consists of a Boundary Transformer and a Proposal Transformer. Specifically, the Boundary Transformer captures long-term temporal dependencies to predict precise boundary information and the Proposal Transformer learns the rich inter-proposal relationships for reliable confidence evaluation. Extensive experiments are conducted on two popular benchmarks: ActivityNet-1.3 and THUMOS14, and the results demonstrate that TAPG Transformer outperforms state-of-the-art methods. Equipped with the existing action classifier, our method achieves remarkable performance on the temporal action localization task. Codes and models will be available. | ['Hujie Huang', 'Hongxun Yao', 'Wenhao Wu', 'Haosen Yang', 'Lining Wang'] | 2021-05-25 | null | null | null | null | ['temporal-action-proposal-generation'] | ['computer-vision'] | [ 2.49654770e-01 3.75997201e-02 -6.88494146e-01 -4.23905820e-01
-1.08769119e+00 7.55951703e-02 7.84379661e-01 1.75260544e-01
-3.52298945e-01 7.69216895e-01 7.06778765e-01 1.16088636e-01
2.00009216e-02 -4.73766297e-01 -4.37531531e-01 -5.01140356e-01
-2.82733232e-01 3.55340511e-01 9.21009362e-01 -2.79657338e-02
2.90529400e-01 -1.63792998e-01 -1.45913255e+00 7.85490632e-01
6.73736989e-01 1.58018863e+00 -5.30839600e-02 3.46481025e-01
-1.68261304e-02 1.25508118e+00 -3.98867905e-01 -1.53078392e-01
-1.95670933e-01 -3.85441601e-01 -9.05543089e-01 -1.62932530e-01
6.44007027e-02 -3.26260865e-01 -2.13726267e-01 7.62504697e-01
2.96144426e-01 3.68824005e-01 5.48957944e-01 -1.34226489e+00
-2.64051855e-01 9.15435195e-01 -4.15563881e-01 4.13859963e-01
5.97142518e-01 2.83125758e-01 1.36783421e+00 -1.17834008e+00
4.94002342e-01 1.40567219e+00 7.04060197e-01 5.48129201e-01
-9.07366872e-01 -6.09909356e-01 8.42832506e-01 6.88964307e-01
-1.31263661e+00 -4.16132569e-01 5.98051131e-01 -3.42716277e-01
1.41821969e+00 -2.73349673e-01 8.20058525e-01 1.54730904e+00
5.25763273e-01 1.41965926e+00 9.50296819e-01 -9.33073685e-02
3.40400398e-01 -7.42808998e-01 4.84515578e-02 7.16301501e-01
-2.54839480e-01 1.83956549e-01 -1.09951222e+00 4.51850295e-02
5.97009778e-01 3.51354666e-02 -1.10054746e-01 -1.20986536e-01
-1.61688352e+00 4.99357283e-01 2.30244100e-01 4.79403853e-01
-3.23920697e-01 6.47211194e-01 6.52294874e-01 -2.06453323e-01
6.36704981e-01 -7.18259215e-02 -5.66843629e-01 -5.54047465e-01
-8.14817965e-01 2.34889984e-01 4.57997382e-01 1.01952970e+00
2.28163138e-01 -2.46127501e-01 -1.02779496e+00 6.17268145e-01
4.96581703e-01 -9.11422446e-03 7.67754912e-01 -1.08869278e+00
8.22098076e-01 6.56242371e-01 2.65162438e-01 -5.97448230e-01
-3.03732663e-01 -2.07691222e-01 -6.66081786e-01 -1.36004791e-01
3.60431224e-01 2.95045465e-01 -9.20675159e-01 1.85526979e+00
3.25628787e-01 7.34239697e-01 -1.40374184e-01 5.75415373e-01
7.03076720e-01 5.09515643e-01 6.44884467e-01 -3.61199737e-01
1.20500970e+00 -1.45301235e+00 -8.53716850e-01 -3.20375800e-01
7.18698502e-01 -3.81264955e-01 1.02621222e+00 4.59293336e-01
-1.11326540e+00 -7.87188172e-01 -1.07933784e+00 -1.28448218e-01
-1.65725723e-01 3.54207307e-01 6.71196580e-01 -1.94055587e-01
-8.77552152e-01 8.97373915e-01 -1.30570221e+00 -2.74808466e-01
6.12837553e-01 -5.69079863e-03 -2.51208931e-01 8.05118084e-02
-1.44134665e+00 8.18731248e-01 6.01207078e-01 3.18080008e-01
-1.28475130e+00 -3.85306239e-01 -1.06798255e+00 6.86100051e-02
6.86831117e-01 -5.28261900e-01 1.72733510e+00 -4.63723809e-01
-1.65902126e+00 6.04983747e-01 -2.84208089e-01 -7.73919642e-01
7.95217931e-01 -5.21770120e-01 -6.03746772e-01 1.79675571e-03
5.54476500e-01 8.01834106e-01 7.91832685e-01 -3.11107218e-01
-1.07682538e+00 -1.33751050e-01 -5.93820550e-02 1.63647830e-01
9.45849940e-02 4.33111712e-02 -6.24805748e-01 -9.81776893e-01
2.05603749e-01 -6.81249082e-01 -2.36291707e-01 -2.68710945e-02
-2.53442168e-01 -1.01043999e+00 5.22951365e-01 -2.79953301e-01
1.57304358e+00 -1.89808714e+00 -9.22028273e-02 -2.35044405e-01
-2.71418523e-02 1.11042134e-01 -2.51257926e-01 3.19137126e-01
-1.62815467e-01 -2.13500261e-01 -1.31071106e-01 -5.52554786e-01
2.01175377e-01 1.34686127e-01 -5.01174986e-01 3.02660257e-01
1.77673995e-01 1.13930941e+00 -1.19227803e+00 -6.75548553e-01
2.13210091e-01 2.59133756e-01 -3.65172267e-01 1.49823681e-01
-5.84364474e-01 3.20597619e-01 -7.58894682e-01 8.50498080e-01
6.84574097e-02 -3.76288354e-01 1.87656134e-01 -3.84271890e-01
-4.85160314e-02 8.21958661e-01 -9.80860770e-01 2.24082208e+00
-1.48273304e-01 3.07863355e-01 -5.22605181e-01 -9.28201616e-01
7.91729987e-01 4.88812804e-01 6.40350759e-01 -8.11024129e-01
5.49942292e-02 5.97492605e-02 -2.48941973e-01 -3.13940227e-01
4.69293594e-01 5.30778393e-02 -3.55557591e-01 4.87481117e-01
1.87957227e-01 2.51031429e-01 4.15430188e-01 2.35257074e-01
1.43267965e+00 9.79655385e-01 5.77526391e-01 -4.06972617e-02
6.10035479e-01 -3.58829141e-01 1.06317282e+00 6.71831489e-01
-8.42299819e-01 2.58511037e-01 5.71753860e-01 -5.39775133e-01
-3.92329276e-01 -1.26300693e+00 2.16543704e-01 1.27635252e+00
2.52008259e-01 -9.93441164e-01 -4.32006329e-01 -1.12631035e+00
-2.66772628e-01 6.92036271e-01 -9.00092185e-01 -3.31828803e-01
-6.61169648e-01 -3.27505320e-01 4.06424731e-01 1.29439414e+00
9.68112528e-01 -1.32876074e+00 -5.64347684e-01 4.69673097e-01
-6.06405675e-01 -1.24038839e+00 -7.62319505e-01 7.61686116e-02
-9.60706830e-01 -1.21815181e+00 -2.61207074e-01 -3.63339275e-01
1.25854775e-01 -4.49703149e-02 1.30107903e+00 -9.05955508e-02
1.64388008e-02 3.47113907e-01 -4.89219069e-01 -2.14468479e-01
4.41357344e-02 -2.23074451e-01 1.68726370e-01 -1.26113236e-01
5.64555526e-01 -7.38900900e-01 -7.21661270e-01 6.05029166e-01
-4.50205207e-01 2.52765149e-01 6.32338464e-01 8.13622534e-01
1.02415764e+00 -1.65896028e-01 5.76744497e-01 -4.50729311e-01
4.36684340e-01 -1.55110434e-01 -3.32330763e-01 4.20723289e-01
-6.34218812e-01 2.39610642e-01 5.23218736e-02 -3.26354146e-01
-1.39069045e+00 6.25440925e-02 -1.58932388e-01 -3.12276423e-01
-1.25465691e-01 5.48256159e-01 -1.93595245e-01 5.67916512e-01
4.81210738e-01 2.85426140e-01 -6.11646831e-01 -4.97207195e-01
2.03413218e-01 4.17615585e-02 6.26002848e-01 -7.74663985e-01
2.12251455e-01 5.07597387e-01 -2.23063692e-01 -6.44982159e-02
-1.64103425e+00 -4.38655496e-01 -7.53917396e-01 -3.59011203e-01
1.01374030e+00 -9.83624399e-01 -6.31742358e-01 5.90162516e-01
-1.29028058e+00 -6.22313797e-01 -4.54395741e-01 6.10971570e-01
-9.27661955e-01 3.25281739e-01 -7.76828766e-01 -6.37412131e-01
-3.07729453e-01 -9.77760613e-01 1.37711549e+00 1.51121859e-02
-1.74917147e-01 -7.93967485e-01 2.92085171e-01 3.45617592e-01
4.44101654e-02 1.59288406e-01 3.67133886e-01 -5.58591306e-01
-6.30042732e-01 -1.72239672e-02 -4.97447625e-02 3.60166997e-01
8.38190690e-02 -7.85135925e-02 -7.83703566e-01 3.50703038e-02
-1.10914752e-01 -6.45781279e-01 1.35350633e+00 4.31445241e-01
1.43584323e+00 -1.07310824e-02 -5.89461744e-01 2.55344182e-01
7.53867686e-01 1.15601271e-01 6.61267817e-01 9.69926417e-02
3.41116458e-01 1.66777834e-01 1.46096718e+00 6.68898225e-01
3.98015678e-01 9.51811075e-01 3.88867885e-01 5.41369081e-01
-2.00267714e-02 -7.12673187e-01 6.91883743e-01 5.06787956e-01
-3.79995555e-01 -1.60697162e-01 -5.63534319e-01 3.80605757e-01
-2.54058886e+00 -1.32228446e+00 2.28897661e-01 1.84134722e+00
8.75500381e-01 5.96069872e-01 1.55757397e-01 1.30863696e-01
6.74254656e-01 6.09727979e-01 -5.96072912e-01 1.03994258e-01
2.49982730e-01 7.95869455e-02 -1.84013456e-01 2.01720312e-01
-1.51836658e+00 1.06636167e+00 6.33408833e+00 1.18485498e+00
-6.06668234e-01 2.63857067e-01 5.90916574e-01 1.22011952e-01
1.55638590e-01 1.13647878e-02 -1.04093850e+00 4.68318641e-01
7.43422270e-01 -7.26554841e-02 -3.09016168e-01 9.73210633e-01
3.83626729e-01 -4.23996270e-01 -1.47594631e+00 9.51363981e-01
-1.31513506e-01 -1.48097444e+00 -1.63022466e-02 -3.71003598e-01
4.41265762e-01 4.93634269e-02 -2.36760467e-01 7.27890134e-01
4.37170327e-01 -8.10190916e-01 1.05806422e+00 8.46679091e-01
6.65811956e-01 -4.80336130e-01 5.36560416e-01 2.89846420e-01
-2.09001112e+00 -1.67084187e-01 -8.49713758e-02 -1.59404308e-01
5.38709402e-01 5.65517783e-01 -4.47831362e-01 3.71052116e-01
7.92312741e-01 1.54711676e+00 -4.63367283e-01 9.07833993e-01
-9.75221038e-01 6.06502593e-01 -1.56172533e-02 -1.39968097e-02
3.57668221e-01 7.13541657e-02 3.12709212e-01 1.05111063e+00
3.26305389e-01 1.46064699e-01 3.84010613e-01 7.91304529e-01
-6.88165873e-02 -3.13611090e-01 -1.04177915e-01 -8.63187760e-02
4.26829338e-01 1.08975565e+00 -6.17792726e-01 -6.69964254e-01
-2.92030692e-01 8.85201812e-01 4.74099755e-01 7.33528435e-02
-1.40587425e+00 3.24897990e-02 6.38072729e-01 -1.09795690e-01
4.21220332e-01 -2.24821791e-01 1.72893628e-01 -1.40172493e+00
1.62032783e-01 -5.53241789e-01 9.93578792e-01 -9.35018241e-01
-1.13110220e+00 5.02237201e-01 3.07724416e-01 -1.54112768e+00
-2.88869619e-01 -5.56392074e-01 -8.44286621e-01 2.83316821e-01
-1.23235679e+00 -1.36648810e+00 -3.36818546e-01 8.13265562e-01
1.06201279e+00 1.45359412e-01 7.60789573e-01 -2.09193043e-02
-6.51290298e-01 3.77914906e-01 -5.30582666e-01 7.88668022e-02
6.86622739e-01 -1.05700576e+00 4.04430091e-01 9.08191383e-01
3.27111691e-01 3.67292702e-01 2.55260468e-01 -8.33106935e-01
-5.90384662e-01 -1.33586907e+00 1.15290201e+00 -6.12651706e-01
9.06509697e-01 -2.57973731e-01 -5.70212424e-01 9.34233606e-01
-1.18727326e-01 4.00408626e-01 4.72559333e-01 1.17894776e-01
-5.93441248e-01 -1.41132027e-01 -6.87605560e-01 6.16454065e-01
1.74681079e+00 -5.40050387e-01 -1.04816794e+00 4.55478728e-01
9.25703764e-01 -4.80167240e-01 -7.84428358e-01 6.80245638e-01
5.36389232e-01 -1.15801501e+00 8.57565284e-01 -6.37462556e-01
5.26551843e-01 -4.27579254e-01 -9.24188346e-02 -8.75954270e-01
-3.76496583e-01 -6.43786967e-01 -6.72512949e-01 1.10257876e+00
4.37752515e-01 -1.91375703e-01 7.55660295e-01 4.42662895e-01
-5.47248304e-01 -1.14357519e+00 -1.11525607e+00 -9.24971282e-01
-4.52240735e-01 -1.12169671e+00 3.51522565e-01 4.97414529e-01
3.45197320e-01 4.41057354e-01 -4.42303360e-01 -4.18980241e-01
3.83958966e-01 1.34322032e-01 3.30774188e-01 -9.56904054e-01
-4.88669038e-01 -3.52085114e-01 -2.64799327e-01 -1.52734768e+00
3.11272740e-01 -5.19678235e-01 3.39498729e-01 -1.58154857e+00
1.95955828e-01 -2.08767399e-01 -7.10805357e-01 8.29205811e-01
-1.81940615e-01 7.11298436e-02 -1.46870106e-01 7.58722052e-02
-1.50164092e+00 1.04251266e+00 1.16035461e+00 -3.83139670e-01
-1.46404505e-01 2.43679836e-01 -2.64731884e-01 1.17881870e+00
4.56380993e-01 -4.37807590e-01 -6.88958108e-01 -2.55392253e-01
1.23423390e-01 8.49400684e-02 3.56272966e-01 -1.43381810e+00
2.88785845e-01 -5.73633969e-01 2.83517301e-01 -1.03337371e+00
6.26179099e-01 -3.83277088e-01 -1.53641060e-01 4.63667125e-01
-7.29594588e-01 -4.57463004e-02 -8.18760544e-02 1.10292363e+00
-3.69131595e-01 2.91492581e-01 6.35689199e-01 -9.52451378e-02
-1.18182337e+00 8.23977828e-01 -2.54507661e-01 3.23953986e-01
1.14089477e+00 -5.55073768e-02 -3.35208297e-01 -1.48368984e-01
-1.11504853e+00 2.24283770e-01 -1.43290773e-01 5.97338736e-01
8.25091600e-01 -1.79325175e+00 -3.36325824e-01 -3.00471157e-01
4.81309742e-01 -4.49150465e-02 2.64218122e-01 1.32195604e+00
2.59275466e-01 5.82342744e-01 2.26856973e-02 -6.37815416e-01
-1.03145194e+00 4.66923922e-01 3.78233373e-01 -8.62672210e-01
-7.49818265e-01 1.15071440e+00 2.99391329e-01 1.88071504e-01
4.94672328e-01 -8.96453202e-01 -2.38633782e-01 -9.81611311e-02
7.87400782e-01 2.02990994e-01 -1.51330724e-01 -4.93608683e-01
-6.63353086e-01 3.43044013e-01 -1.43358661e-02 -1.67274088e-01
1.27470779e+00 4.96679544e-02 1.10960171e-01 5.61190665e-01
6.62080884e-01 -4.85897064e-01 -1.61714613e+00 -4.77398127e-01
3.99044782e-01 -3.96277487e-01 -2.47008219e-01 -7.49585450e-01
-8.56869459e-01 8.47108960e-01 2.65202254e-01 -2.48830944e-01
1.07431924e+00 2.20162556e-01 8.12715650e-01 5.22282243e-01
7.00969815e-01 -1.57343268e+00 8.96349013e-01 8.42725337e-01
1.02025318e+00 -1.14282918e+00 2.82249320e-02 -2.28170797e-01
-6.50346339e-01 8.40570331e-01 1.07503939e+00 -1.15437004e-04
7.26022840e-01 -4.23402265e-02 -4.85952497e-01 -1.54123366e-01
-1.35815918e+00 -2.96296746e-01 5.22163272e-01 5.04226327e-01
6.04638219e-01 1.43311685e-03 -5.41395724e-01 1.09754360e+00
4.49792713e-01 4.40335810e-01 3.29307071e-03 9.58454370e-01
-5.30644357e-01 -1.20716918e+00 2.17189521e-01 4.06269670e-01
-2.35068977e-01 -6.79612830e-02 -1.70843586e-01 5.58706939e-01
3.96983653e-01 9.43913698e-01 -2.97583509e-02 -6.09663904e-01
3.38649571e-01 1.23032227e-01 4.75620002e-01 -6.85800552e-01
-2.65890598e-01 1.30337765e-02 5.15124679e-01 -1.60927761e+00
-9.47210073e-01 -7.22495317e-01 -1.50222838e+00 1.05253413e-01
-6.86203390e-02 2.24607065e-01 -1.39730219e-02 1.39390242e+00
3.44902396e-01 6.66565657e-01 4.11642045e-02 -8.91103029e-01
-5.96731842e-01 -1.20429814e+00 -4.62700039e-01 2.67117530e-01
-1.36198312e-01 -1.19772649e+00 -1.48566231e-01 3.37230824e-02] | [8.398733139038086, 0.4901648461818695] |
593c6b13-d515-4059-b4d7-169d220e62fd | gradient-domain-image-reconstruction | null | null | http://openaccess.thecvf.com/content_cvpr_2016/html/Shibata_Gradient-Domain_Image_Reconstruction_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Shibata_Gradient-Domain_Image_Reconstruction_CVPR_2016_paper.pdf | Gradient-Domain Image Reconstruction Framework With Intensity-Range and Base-Structure Constraints | This paper presents a novel unified gradient-domain image reconstruction framework with intensity-range constraint and base-structure constraint. The existing method for manipulating base structures and detailed textures are classifiable into two major approaches: i) gradient-domain and ii) layer-decomposition. To generate detail-preserving and artifact-free output images, we combine the benefits of the two approaches into the proposed framework by introducing the intensity-range constraint and the base-structure constraint. To preserve details of the input image, the proposed method takes advantage of reconstructing the output image in the gradient domain, while the output intensity is guaranteed to lie within the specified intensity range, e.g. 0-to-255, by the intensity-range constraint. In addition, the reconstructed image lies close to the base structure by the base-structure constraint, which is effective for restraining artifacts. Experimental results show that the proposed framework is effective for various applications such as tone mapping, seamless image cloning, detail enhancement, and image restoration. | ['Masatoshi Okutomi', 'Masayuki Tanaka', 'Takashi Shibata'] | 2016-06-01 | null | null | null | cvpr-2016-6 | ['tone-mapping'] | ['computer-vision'] | [ 6.66912258e-01 -2.79794812e-01 -1.09961659e-01 -1.83035120e-01
-1.32989064e-01 -1.70779079e-01 1.94786072e-01 -3.63431007e-01
-9.49242339e-02 8.36127639e-01 2.10499555e-01 -5.85710630e-02
-2.01710761e-01 -8.92837226e-01 -5.06404519e-01 -9.16940629e-01
4.55407977e-01 -6.64576411e-01 5.25002420e-01 -3.26662034e-01
5.43036640e-01 5.15137553e-01 -1.62626135e+00 3.00660104e-01
1.15545213e+00 1.17775536e+00 5.33048570e-01 2.58701056e-01
-3.86011720e-01 6.59646153e-01 -4.25037861e-01 1.67875871e-01
3.84433359e-01 -8.69707584e-01 -2.73393750e-01 5.13104379e-01
1.17282286e-01 -5.83504260e-01 -2.24293277e-01 1.43632627e+00
2.83345789e-01 1.21735632e-01 3.51235658e-01 -8.26906383e-01
-6.64454579e-01 -1.62837610e-01 -9.27533746e-01 3.50774489e-02
1.38582185e-01 -2.53149092e-01 4.71639454e-01 -8.16871166e-01
5.67353487e-01 1.01865721e+00 4.01204795e-01 3.24715436e-01
-1.24322224e+00 -5.38902223e-01 2.44559556e-01 9.71058458e-02
-1.37856662e+00 -3.98689985e-01 1.34771478e+00 -1.29362509e-01
3.16059589e-01 5.87637186e-01 4.99361753e-01 1.85364336e-01
3.26931238e-01 4.06902701e-01 1.54034126e+00 -7.77133703e-01
7.09299222e-02 2.10140795e-01 -8.56690332e-02 7.10341573e-01
2.30932251e-01 2.03026727e-01 -4.16606039e-01 4.02638242e-02
1.35422862e+00 -3.64152864e-02 -8.20765197e-01 -4.04979706e-01
-1.19721282e+00 3.06132942e-01 4.21414614e-01 5.41383088e-01
-2.07338676e-01 -2.81474531e-01 1.38289958e-01 2.42146999e-01
4.88366514e-01 6.26611039e-02 -2.02075895e-02 3.18433225e-01
-1.01603711e+00 2.71709487e-02 3.18626106e-01 1.03202307e+00
9.13542330e-01 3.31521451e-01 -1.33550271e-01 1.12785482e+00
3.02077979e-01 3.36423188e-01 4.17976290e-01 -9.10507798e-01
3.98923606e-01 4.04568255e-01 3.59119654e-01 -1.10479200e+00
2.76240129e-02 -4.10042644e-01 -1.04392087e+00 6.95233762e-01
2.10622519e-01 -1.34255746e-02 -8.84512603e-01 1.61091506e+00
3.94337684e-01 1.37926519e-01 -3.60604376e-02 1.16248548e+00
7.61574030e-01 1.01094663e+00 -3.87481779e-01 -6.60811782e-01
1.26292419e+00 -8.45182598e-01 -1.14023483e+00 5.97064383e-02
-2.21218750e-01 -1.17671955e+00 1.19271934e+00 3.36021572e-01
-1.19165802e+00 -7.72686481e-01 -1.45095372e+00 -1.73250616e-01
1.51989460e-01 3.27552110e-01 1.61515623e-01 5.63114047e-01
-8.90722334e-01 4.38826084e-01 -4.38674241e-01 -5.91624193e-02
-1.56188041e-01 3.33034880e-02 -2.54738629e-01 -2.42068201e-01
-8.49481225e-01 4.66334432e-01 2.03734279e-01 1.78385496e-01
-3.09380412e-01 -5.74210525e-01 -7.20454812e-01 1.55345172e-01
1.51468560e-01 -4.40436482e-01 6.96196139e-01 -1.10077703e+00
-1.74700522e+00 8.01292777e-01 -3.63139510e-01 2.18066812e-01
6.02774322e-01 1.70321733e-01 -5.21975279e-01 2.08255753e-01
-5.64478710e-02 4.20852005e-02 1.00462246e+00 -1.62712908e+00
-7.50390232e-01 -2.57534593e-01 -2.85050631e-01 5.00377476e-01
-4.30811703e-01 -1.11225754e-01 -7.59714007e-01 -8.67439568e-01
7.22249925e-01 -4.38560039e-01 5.54161593e-02 4.23790455e-01
-2.75346935e-01 5.38823724e-01 1.20130444e+00 -9.70710814e-01
1.41053045e+00 -2.45318985e+00 -7.38231167e-02 3.54590684e-01
-4.43270877e-02 1.43577307e-01 -4.81977277e-02 2.98180491e-01
-7.54863098e-02 -3.49687696e-01 -4.97469068e-01 -2.24448200e-02
-4.26446944e-01 2.99485568e-02 -1.56682968e-01 4.09584880e-01
3.18906903e-02 2.69055128e-01 -5.92064261e-01 -4.87025529e-01
3.08113188e-01 6.72542930e-01 -4.27235872e-01 3.17118466e-01
1.13184243e-01 5.96153259e-01 -4.25285250e-01 5.90933025e-01
1.22447395e+00 3.98502588e-01 1.17831789e-01 -5.16262949e-01
-6.17844760e-01 -2.74512395e-02 -1.50941753e+00 1.26968622e+00
-4.55337077e-01 3.88973325e-01 7.09173918e-01 -7.21156001e-01
1.42337811e+00 1.92687199e-01 4.16459173e-01 -1.01963758e+00
-9.80250612e-02 4.48120862e-01 -2.45842203e-01 -2.51875371e-01
3.87604296e-01 -4.62161720e-01 4.61890638e-01 2.09112734e-01
-3.96371007e-01 -1.05298676e-01 6.36562929e-02 -3.16803515e-01
4.01225299e-01 1.91239581e-01 1.74047753e-01 -6.35739207e-01
9.29367006e-01 -1.54276416e-01 8.11358035e-01 3.12333137e-01
-1.03512861e-01 7.59820998e-01 2.32874483e-01 -2.04967380e-01
-1.22667384e+00 -1.05918121e+00 -3.50883186e-01 6.27364159e-01
8.28318894e-01 1.14723638e-01 -7.53296554e-01 -5.09888772e-03
-3.25938553e-01 3.10971171e-01 -3.94725025e-01 -1.29342794e-01
-6.33981764e-01 -4.29799467e-01 5.36278002e-02 2.15414390e-01
1.36604345e+00 -9.18124557e-01 -3.90861809e-01 2.95137972e-01
-3.69847506e-01 -6.88448012e-01 -8.25882614e-01 -1.89306617e-01
-1.08799517e+00 -8.30819249e-01 -1.12700081e+00 -1.31385624e+00
9.78146851e-01 5.89649975e-01 7.65872896e-01 3.29389900e-01
-1.85267687e-01 -5.00292592e-02 -2.26193041e-01 2.21704632e-01
-1.87177017e-01 -5.45705974e-01 -4.77886051e-01 4.57106024e-01
-3.50444674e-01 -6.46830857e-01 -6.52913511e-01 5.62332451e-01
-1.16885412e+00 4.98528719e-01 5.25219679e-01 1.00562990e+00
9.38569784e-01 5.10103703e-01 2.64835954e-01 -6.00190997e-01
7.22983062e-01 1.78616375e-01 -6.84068263e-01 2.90894300e-01
-6.20343328e-01 -1.29611477e-01 6.64026320e-01 -5.23374319e-01
-1.68380141e+00 -1.06706405e-02 -5.30016124e-02 -2.09507406e-01
1.11566797e-01 2.90522307e-01 -5.75772285e-01 -3.24044913e-01
3.85354966e-01 7.25668073e-01 2.32353792e-01 -6.14082754e-01
2.19826940e-02 6.07369304e-01 8.75646830e-01 -5.20769775e-01
6.76389813e-01 5.98954916e-01 1.55811548e-01 -9.02164817e-01
-3.22908938e-01 -1.85702443e-01 -2.19841391e-01 -2.81134188e-01
6.07175350e-01 -7.44454503e-01 -4.38928008e-01 6.95435703e-01
-8.09505224e-01 -9.05170813e-02 -2.33264989e-03 5.04906058e-01
-4.64094639e-01 6.76136672e-01 -8.13787282e-01 -7.86396563e-01
-3.29151839e-01 -1.03260934e+00 6.88593984e-01 4.03090745e-01
4.11370188e-01 -7.37659156e-01 -3.80552381e-01 -4.50938046e-02
7.02809155e-01 4.70043093e-01 1.18461072e+00 6.30545795e-01
-7.00217426e-01 8.94300491e-02 -4.72058326e-01 4.55089241e-01
5.85689187e-01 -1.39068305e-01 -6.19633496e-01 -3.04622859e-01
4.37972039e-01 2.01753810e-01 7.14573860e-01 3.86197925e-01
9.95596349e-01 -2.69782305e-01 -2.42750403e-02 9.29705024e-01
1.82180548e+00 5.48731506e-01 1.20800483e+00 5.32822847e-01
2.99524724e-01 5.43999374e-01 7.42961407e-01 3.86695832e-01
-8.83247182e-02 8.06768000e-01 1.20123498e-01 -6.65691018e-01
-4.68452722e-01 -3.36338520e-01 -3.92375663e-02 5.21700501e-01
-8.81503057e-03 -2.73748562e-02 -2.30674878e-01 4.68229145e-01
-1.63226318e+00 -9.02384281e-01 -1.10621355e-01 2.35890579e+00
9.96233344e-01 1.64415658e-01 -2.56214857e-01 4.01199192e-01
1.16325307e+00 1.93341717e-01 -4.66851711e-01 -5.95834434e-01
-2.08144650e-01 -1.10737912e-01 3.07068646e-01 7.51175523e-01
-7.28153467e-01 5.48830688e-01 5.92675543e+00 1.00106585e+00
-1.47424328e+00 -2.60872185e-01 6.44807994e-01 2.86277384e-01
-5.18291831e-01 2.94321012e-02 -4.07990962e-01 5.85452616e-01
-6.66557699e-02 -3.28813672e-01 6.27005637e-01 5.93264997e-01
4.03886378e-01 -3.84868950e-01 -3.88742179e-01 9.55998361e-01
-2.71147955e-02 -1.03060782e+00 2.49334369e-02 -1.08263858e-01
8.06541979e-01 -9.47728097e-01 2.45382607e-01 -2.35443324e-01
-5.20772398e-01 -5.79330921e-01 7.29134440e-01 4.12830204e-01
1.31159854e+00 -8.57932508e-01 4.43770498e-01 2.71733761e-01
-1.31539226e+00 2.13076636e-01 -4.37595904e-01 4.45959456e-02
1.45655274e-01 1.00886500e+00 -1.30132316e-02 6.83570564e-01
6.35560572e-01 5.18489540e-01 1.93935744e-02 1.00680614e+00
-9.80022848e-02 1.82984591e-01 -1.81577906e-01 4.36221510e-01
-5.74241392e-02 -8.48707020e-01 4.33503985e-01 8.05539310e-01
4.23385084e-01 3.42159271e-01 1.81548774e-01 9.89815831e-01
2.33060613e-01 1.37439191e-01 -3.85553032e-01 3.48734677e-01
3.97121340e-01 1.00987792e+00 -7.47950971e-01 -3.12535793e-01
-4.47166055e-01 1.31658065e+00 -2.94563681e-01 5.41110873e-01
-7.06307352e-01 -9.57567275e-01 3.36323351e-01 3.31323445e-01
3.35574687e-01 -2.06796601e-01 -7.28971839e-01 -8.66308153e-01
4.04868007e-01 -7.73539841e-01 -6.46670908e-02 -8.51876795e-01
-8.17215443e-01 7.20475435e-01 -1.01288356e-01 -1.60630703e+00
1.85653642e-01 -4.00658041e-01 -6.42431319e-01 1.24321568e+00
-1.97940397e+00 -1.01200199e+00 -5.71913838e-01 9.12399113e-01
2.58101791e-01 1.78101122e-01 4.70431000e-01 4.99207437e-01
-2.31623843e-01 3.76529843e-01 3.43264401e-01 -2.36286581e-01
6.09260857e-01 -7.47699976e-01 -3.09816033e-01 1.00436747e+00
-5.61855733e-01 6.00644886e-01 6.06433272e-01 -7.21539795e-01
-1.02228475e+00 -7.97573924e-01 7.47536361e-01 5.80695570e-01
-2.21123043e-02 -1.96379349e-01 -1.04161966e+00 2.25382745e-01
1.17778011e-01 -1.97706223e-02 2.38882110e-01 -3.98677081e-01
-1.52399302e-01 -4.55409467e-01 -1.56455016e+00 7.13866055e-01
7.62856424e-01 -3.96553487e-01 -2.87338585e-01 -4.61274236e-02
4.22155410e-01 -6.86563611e-01 -6.02153122e-01 5.74498594e-01
6.50024116e-01 -1.34932959e+00 8.43047798e-01 1.84318766e-01
5.50682306e-01 -9.47725475e-01 -1.19738072e-01 -1.03494358e+00
-6.42100513e-01 -5.26885331e-01 3.96395147e-01 1.48310804e+00
2.17013121e-01 -6.80931330e-01 5.30157268e-01 4.50310618e-01
-2.26361141e-01 -5.61033666e-01 -8.45359445e-01 -6.39427960e-01
-4.05465156e-01 3.69990259e-01 4.97086167e-01 8.67339432e-01
-1.81542769e-01 -1.15135975e-01 -6.28696978e-01 3.77502114e-01
6.21612430e-01 4.82743800e-01 4.05856550e-01 -6.63621068e-01
-3.31191331e-01 -3.67847741e-01 -9.54580382e-02 -1.18699932e+00
-3.99833560e-01 -3.10044944e-01 1.18692830e-01 -1.74795520e+00
1.84009925e-01 -5.28882921e-01 -4.08594668e-01 2.06388697e-01
-1.80842102e-01 4.38248366e-01 8.68075117e-02 3.82185459e-01
3.11342001e-01 5.54249287e-01 1.77258885e+00 9.48277116e-02
-4.27178770e-01 -1.45951658e-01 -6.57746732e-01 3.27644318e-01
7.03328431e-01 2.42515281e-03 -6.38772011e-01 -4.89869654e-01
-3.87257159e-01 2.31799036e-01 7.92783275e-02 -9.44974959e-01
6.64848136e-03 -4.09605950e-01 5.37783206e-01 -3.97080451e-01
3.83568257e-01 -1.06163478e+00 3.07242423e-01 4.92051065e-01
-1.53565109e-01 -1.93547487e-01 2.44082212e-01 3.65465432e-01
-5.88715672e-01 -1.57720193e-01 1.23009670e+00 -7.91299120e-02
-8.45366180e-01 -2.13705420e-01 9.89468396e-03 -4.91245955e-01
1.01168048e+00 -7.30181813e-01 -2.30588272e-01 -4.02045965e-01
-4.66725767e-01 -1.76678151e-01 7.48074234e-01 1.53054774e-01
7.84909844e-01 -1.35963166e+00 -6.07515514e-01 7.65560210e-01
-9.01966467e-02 -3.70724648e-01 5.09270072e-01 7.10445762e-01
-7.86986411e-01 8.08460191e-02 -5.19958436e-01 -4.96834487e-01
-1.21434975e+00 6.18792951e-01 4.16162014e-01 -1.44909650e-01
-7.69047320e-01 5.70279360e-01 4.95035201e-01 -2.93634925e-02
1.13870405e-01 -2.21843123e-01 -2.88980994e-02 -6.21224403e-01
7.69487321e-01 2.48203874e-01 -2.44549401e-02 -4.76764381e-01
-5.76854423e-02 1.00857723e+00 1.39961407e-01 -2.80697852e-01
1.02203500e+00 -4.33523893e-01 -4.60880876e-01 8.29544067e-02
1.02881372e+00 4.21289176e-01 -1.35201693e+00 -1.77234724e-01
-4.68989789e-01 -9.71171677e-01 3.91981155e-01 -8.58127654e-01
-1.16884971e+00 6.30181730e-01 6.63856506e-01 9.23264995e-02
1.90814924e+00 -6.15118444e-01 7.00225651e-01 -2.42581412e-01
3.38847637e-01 -9.92501855e-01 -2.84825508e-02 6.35583103e-02
1.02984548e+00 -6.17778480e-01 1.18771724e-01 -9.84329522e-01
-4.14952219e-01 1.06449187e+00 6.53404713e-01 -8.13573301e-02
5.23778737e-01 1.81919992e-01 8.95909518e-02 2.17303887e-01
-2.84004807e-01 -8.11052546e-02 3.22519928e-01 4.96881366e-01
5.72069645e-01 -2.35343024e-01 -9.21570718e-01 1.34899095e-01
3.16999733e-01 1.82182774e-01 4.66609061e-01 1.04902840e+00
-8.26321781e-01 -1.05599272e+00 -8.01565886e-01 1.64901465e-01
-3.64184648e-01 -1.27179146e-01 5.48927076e-02 6.04734480e-01
1.55073404e-01 1.13115597e+00 4.42934595e-02 -7.24026188e-02
4.98792410e-01 -2.23698601e-01 5.18903792e-01 -2.59662896e-01
-4.21566926e-02 6.72623515e-01 -7.49255940e-02 -3.42477918e-01
-4.19705212e-01 -7.61921927e-02 -1.19783056e+00 -1.85656965e-01
-4.72151697e-01 7.17181042e-02 5.57438672e-01 3.93773854e-01
1.97643638e-01 4.03332591e-01 9.26689923e-01 -5.70691466e-01
-2.08532974e-01 -6.67384028e-01 -1.14807558e+00 3.99600893e-01
4.41822797e-01 -4.57040817e-01 -4.01141435e-01 2.88998723e-01] | [10.97653579711914, -2.3432939052581787] |
c591a93e-1031-4642-8234-3793d198889b | illumination-insensitive-binary-descriptor | 2305.07943 | null | https://arxiv.org/abs/2305.07943v1 | https://arxiv.org/pdf/2305.07943v1.pdf | Illumination-insensitive Binary Descriptor for Visual Measurement Based on Local Inter-patch Invariance | Binary feature descriptors have been widely used in various visual measurement tasks, particularly those with limited computing resources and storage capacities. Existing binary descriptors may not perform well for long-term visual measurement tasks due to their sensitivity to illumination variations. It can be observed that when image illumination changes dramatically, the relative relationship among local patches mostly remains intact. Based on the observation, consequently, this study presents an illumination-insensitive binary (IIB) descriptor by leveraging the local inter-patch invariance exhibited in multiple spatial granularities to deal with unfavorable illumination variations. By taking advantage of integral images for local patch feature computation, a highly efficient IIB descriptor is achieved. It can encode scalable features in multiple spatial granularities, thus facilitating a computationally efficient hierarchical matching from coarse to fine. Moreover, the IIB descriptor can also apply to other types of image data, such as depth maps and semantic segmentation results, when available in some applications. Numerical experiments on both natural and synthetic datasets reveal that the proposed IIB descriptor outperforms state-of-the-art binary descriptors and some testing float descriptors. The proposed IIB descriptor has also been successfully employed in a demo system for long-term visual localization. The code of the IIB descriptor will be publicly available. | ['Ce Zhu', 'Yipeng Liu', 'Xun Zhang', 'Yingjie Zhou', 'Xinyu Lin'] | 2023-05-13 | null | null | null | null | ['visual-localization'] | ['computer-vision'] | [ 2.53271490e-01 -7.07070351e-01 -3.05122703e-01 -3.87632340e-01
-6.32627547e-01 -4.18398470e-01 5.49468577e-01 3.09620619e-01
-3.85777622e-01 7.13948011e-01 -2.33279541e-01 1.63623765e-01
-5.07800221e-01 -1.03664911e+00 -3.15555364e-01 -9.79002833e-01
7.22440192e-03 5.14787696e-02 5.81851065e-01 -2.05044642e-01
5.77294409e-01 8.82232904e-01 -2.04578972e+00 -5.13430387e-02
7.14976609e-01 1.53109658e+00 2.94689536e-01 2.20770702e-01
3.05657741e-02 3.38407815e-01 -5.98804235e-01 2.53913626e-02
3.66405547e-01 -4.81057130e-02 -3.68047714e-01 1.63364917e-01
6.91223383e-01 -1.23953462e-01 -3.09516728e-01 1.16380990e+00
6.10559940e-01 2.40460202e-01 7.33622015e-01 -1.07465076e+00
-6.02860093e-01 -1.61594927e-01 -6.15526855e-01 3.07046562e-01
5.46864152e-01 -4.30710452e-06 9.40149844e-01 -9.48745191e-01
5.87775230e-01 1.01321638e+00 6.72860205e-01 -2.50175029e-01
-1.04750407e+00 -5.78934133e-01 -8.76267105e-02 5.04234970e-01
-1.95247734e+00 -2.62763828e-01 9.56849992e-01 -4.02545661e-01
9.11867201e-01 3.07344526e-01 6.92814231e-01 5.07975876e-01
4.96420175e-01 4.75455970e-01 1.41668105e+00 -2.92591125e-01
1.61738127e-01 -1.40783668e-01 1.50404014e-02 8.93479049e-01
3.35491031e-01 1.61262840e-01 -4.49694216e-01 -9.95821133e-02
7.11748779e-01 2.42384642e-01 -2.35895187e-01 -4.51722473e-01
-1.20152211e+00 6.00457907e-01 6.87906206e-01 5.22552192e-01
-4.83600855e-01 -5.75356446e-02 3.16110134e-01 3.05413306e-01
2.56967485e-01 -1.41582759e-02 6.99090958e-03 -9.50994343e-02
-1.16253304e+00 1.73315406e-01 4.32898790e-01 8.08451116e-01
1.19286406e+00 -6.98163956e-02 -1.53278366e-01 9.34374154e-01
4.81059216e-02 8.37837577e-01 6.79561675e-01 -7.02223659e-01
3.02229643e-01 8.08350503e-01 2.22270079e-02 -1.61078930e+00
-5.65375268e-01 -2.34038621e-01 -1.03620267e+00 3.79621476e-01
2.39709005e-01 5.22344828e-01 -8.70771885e-01 1.12986457e+00
3.91215146e-01 -1.81826100e-01 -7.50215352e-02 9.90794897e-01
8.63486707e-01 5.23282170e-01 -1.50262013e-01 2.49919295e-02
1.29277301e+00 -5.23812532e-01 -4.96276855e-01 -1.34670913e-01
2.50706792e-01 -9.77159321e-01 7.60013163e-01 2.44063199e-01
-7.17584372e-01 -8.44666302e-01 -1.26991045e+00 -5.08629978e-02
-6.29043639e-01 2.13088587e-01 7.10851908e-01 5.76254010e-01
-1.03209829e+00 4.41932470e-01 -6.61671400e-01 -7.45912671e-01
2.56463736e-01 3.93322796e-01 -6.02335513e-01 -2.35980853e-01
-9.59580004e-01 5.97016275e-01 2.84725010e-01 3.15903574e-01
-5.08344054e-01 -1.06247954e-01 -8.89866710e-01 1.48105491e-02
-1.09570183e-01 -2.47722685e-01 4.28493708e-01 -7.30819166e-01
-1.31749451e+00 9.27381754e-01 -1.83943063e-01 -1.32479087e-01
3.62184227e-01 4.93135422e-01 -4.09667403e-01 4.78477448e-01
3.53998035e-01 5.37936866e-01 9.16537702e-01 -8.73628616e-01
-6.80583596e-01 -5.92303813e-01 -3.96508984e-02 2.75775284e-01
-4.36591387e-01 -2.28203312e-01 -1.93470418e-01 -6.55097783e-01
5.52955568e-01 -7.81672776e-01 2.05955192e-01 4.71967936e-01
-9.86636430e-02 -1.14660248e-01 9.86293972e-01 -2.40691200e-01
1.18396318e+00 -2.13341188e+00 -1.47566793e-03 3.94333512e-01
-1.68008879e-01 3.62502873e-01 6.21324554e-02 4.86978441e-01
2.60420799e-01 -2.42363006e-01 -1.44921780e-01 -2.74277050e-02
-1.25797153e-01 2.19667748e-01 3.47455069e-02 9.28889990e-01
5.45588434e-02 6.74542964e-01 -5.99416435e-01 -9.36389148e-01
6.99549139e-01 4.62871015e-01 -7.61112943e-02 -2.71585565e-02
3.79594177e-01 2.16892257e-01 -7.04091370e-01 1.28363383e+00
1.03526974e+00 -9.53049287e-02 -2.97546625e-01 -4.57759500e-01
-3.07972312e-01 -4.52953368e-01 -1.20488489e+00 1.47610033e+00
-4.18225646e-01 7.73959160e-01 -3.66714969e-02 -1.36831558e+00
1.37411296e+00 -4.25716229e-02 4.88054693e-01 -1.23875654e+00
1.12779420e-02 3.76781672e-01 -3.69217068e-01 -4.70716387e-01
6.65899038e-01 2.31636539e-02 -1.59529239e-01 -6.70020878e-02
-1.32758662e-01 -1.68559358e-01 3.66708517e-01 -3.35662812e-01
9.41763520e-01 -2.35396251e-01 6.34485602e-01 -3.29667449e-01
9.46406484e-01 5.35493949e-04 3.68787169e-01 7.41694868e-01
-4.41562265e-01 6.55104578e-01 -1.29761726e-01 -6.04429245e-01
-8.37354243e-01 -1.01558161e+00 -7.30507135e-01 7.32691824e-01
7.78847277e-01 -3.92205305e-02 -2.81496704e-01 -2.04672664e-01
2.64688969e-01 -2.93194890e-01 -5.51090240e-01 1.42187297e-01
-3.50985616e-01 -6.20533526e-01 5.23360193e-01 4.51680958e-01
1.02416909e+00 -1.00379324e+00 -8.01103711e-01 2.66233444e-01
4.48961481e-02 -1.04728627e+00 -1.68961868e-01 1.96800549e-02
-7.69402564e-01 -1.02619910e+00 -9.66289520e-01 -9.77896988e-01
5.46122670e-01 6.23610377e-01 6.67867959e-01 1.62986621e-01
-6.33563340e-01 3.30930173e-01 -4.30819243e-01 2.98621356e-01
1.14504427e-01 7.91256595e-03 -1.33132204e-01 8.35790560e-02
3.07281792e-01 -6.15777850e-01 -9.51118708e-01 7.98623323e-01
-1.00495028e+00 -3.42896461e-01 7.05828965e-01 1.11311102e+00
8.69555831e-01 1.20302439e-01 1.49781317e-01 -2.90962487e-01
3.24844062e-01 -1.48991138e-01 -7.10865140e-01 3.90117168e-01
-3.49966705e-01 1.46302849e-01 6.05396867e-01 -3.19089860e-01
-8.74339461e-01 -1.34596899e-01 -1.45141063e-02 -8.18942301e-03
-1.85601458e-01 4.58935648e-01 -7.56984949e-02 -8.49195182e-01
4.27594483e-01 5.10271788e-01 -1.78455397e-01 -2.81651616e-01
-1.14333266e-02 9.98553216e-01 4.92607802e-01 -6.03606403e-01
7.18137085e-01 7.90813565e-01 4.76547062e-01 -1.25046837e+00
-4.84637827e-01 -7.72290885e-01 -6.04495406e-01 -1.87316954e-01
6.60549760e-01 -8.56994212e-01 -6.88847303e-01 7.23534465e-01
-8.25653017e-01 1.02581166e-01 -3.50543000e-02 3.76362205e-01
-4.79174435e-01 6.18303180e-01 -3.64335030e-01 -4.35877293e-01
-2.88510025e-01 -1.24566603e+00 1.36642635e+00 5.21430790e-01
1.45075604e-01 -1.03625894e+00 -3.01033328e-03 2.41197497e-01
5.58726430e-01 3.29760551e-01 6.55207038e-01 -4.46146540e-02
-7.18715787e-01 -5.22944987e-01 -5.34302056e-01 3.42627823e-01
2.26208210e-01 1.03371441e-01 -6.79403245e-01 -4.77304608e-01
-3.46732616e-01 -2.79478103e-01 6.79259300e-01 2.64685363e-01
9.74864542e-01 7.69927632e-03 -4.40660685e-01 8.94574046e-01
1.67432702e+00 1.63601398e-01 7.05912232e-01 6.93857849e-01
5.03849864e-01 3.20595503e-01 1.10156345e+00 6.73282504e-01
3.45560014e-01 8.94272625e-01 3.97511363e-01 -2.03779846e-01
-1.60285071e-01 -4.09534350e-02 9.77078974e-02 6.06512070e-01
-1.77898526e-01 -1.72381297e-01 -7.24924028e-01 6.07829571e-01
-1.59861517e+00 -9.77319479e-01 1.27496257e-01 2.26109934e+00
4.16133791e-01 -1.16893597e-01 -2.97445059e-01 2.96366096e-01
8.02152932e-01 5.27549267e-01 -3.25613469e-01 -4.51502413e-01
-4.87103283e-01 3.66647273e-01 8.23232770e-01 2.26894379e-01
-1.30701065e+00 7.00591564e-01 5.84770203e+00 1.19048667e+00
-1.35111356e+00 -1.47862628e-03 2.71923512e-01 4.85655487e-01
8.67241398e-02 6.85151070e-02 -8.16009104e-01 5.75232387e-01
3.89456034e-01 -1.81580216e-01 8.15306231e-02 9.05319333e-01
-5.36426082e-02 -5.21378100e-01 -6.26395941e-01 1.38572943e+00
1.45233750e-01 -1.00889397e+00 7.80612379e-02 8.01549554e-02
7.97098815e-01 -1.49162725e-01 2.33871117e-01 -6.25326112e-02
-5.47841311e-01 -9.41271722e-01 6.71780646e-01 5.28714180e-01
8.92755866e-01 -7.22611070e-01 8.97000432e-01 1.30070513e-02
-1.69291317e+00 -1.04266092e-01 -7.04239249e-01 -1.43625170e-01
-5.73385805e-02 4.69984084e-01 -3.51518512e-01 5.28985083e-01
6.80951715e-01 8.67350817e-01 -9.27530169e-01 1.24863553e+00
-1.98232830e-02 -4.73375581e-02 -4.36704934e-01 -3.07734776e-02
3.51873547e-01 -1.95507839e-01 3.78790826e-01 1.18940210e+00
6.38782203e-01 -3.34617272e-02 1.75303370e-01 5.03037870e-01
3.02087218e-01 3.60522449e-01 -6.85131311e-01 2.08212748e-01
5.89630187e-01 1.21008182e+00 -8.89559209e-01 -2.94998378e-01
-4.76625592e-01 1.14894557e+00 7.39028603e-02 2.12298095e-01
-6.26621187e-01 -7.82970428e-01 6.44386888e-01 1.38138816e-01
6.19897962e-01 -3.77701998e-01 -9.96179804e-02 -1.01773143e+00
1.92967996e-01 -4.47933763e-01 2.32818514e-01 -6.41655803e-01
-1.11023450e+00 6.55460417e-01 7.57207796e-02 -1.68036640e+00
-3.62572968e-02 -6.72808886e-01 -3.73906493e-01 6.36074662e-01
-1.60013521e+00 -1.20283878e+00 -8.45202446e-01 8.40079546e-01
1.58172637e-01 -2.00336799e-02 7.07133889e-01 4.49926257e-01
-4.12643135e-01 6.85584128e-01 5.05356789e-01 -1.19164705e-01
6.12484515e-01 -8.26438308e-01 -1.15807451e-01 7.65707433e-01
3.19757052e-02 4.52609003e-01 5.15978575e-01 -4.03043687e-01
-1.29775190e+00 -8.07959497e-01 6.26171887e-01 2.11752966e-01
5.38964748e-01 2.87360437e-02 -7.47155845e-01 5.90367503e-02
-1.25204653e-01 6.77294016e-01 2.07531944e-01 -4.46476787e-01
-1.68270379e-01 -5.64015985e-01 -1.38378501e+00 7.24765733e-02
9.44181561e-01 -6.39737368e-01 -4.00141746e-01 2.84407586e-01
-3.99082191e-02 -4.21559900e-01 -1.11113155e+00 5.50107002e-01
8.97243321e-01 -1.29742634e+00 1.15043402e+00 3.40429515e-01
-5.52524999e-02 -4.83645946e-01 -5.02741456e-01 -8.64707351e-01
-1.95603937e-01 -1.59528255e-01 2.93058544e-01 1.15013647e+00
-1.89497620e-01 -8.45189035e-01 6.07076824e-01 1.90908998e-01
1.25024602e-01 -6.20138645e-01 -1.31722987e+00 -9.40158546e-01
-3.28875482e-01 -3.92035730e-02 4.90240842e-01 5.59936345e-01
-2.19487876e-01 -2.52640039e-01 -1.77567676e-01 1.98326737e-01
8.07835698e-01 5.45951486e-01 6.07073307e-01 -1.21716702e+00
5.07206935e-03 -3.42616677e-01 -1.50744057e+00 -1.11110866e+00
5.02517708e-02 -7.27098465e-01 -1.17305189e-01 -1.29737997e+00
2.45445520e-01 -7.65483975e-01 -4.12616342e-01 3.25453222e-01
9.89458486e-02 9.78243470e-01 6.75830692e-02 4.18866009e-01
-6.23303711e-01 6.73258245e-01 1.15149701e+00 -3.88348818e-01
5.44619411e-02 -2.14177251e-01 -7.10156187e-02 5.60630202e-01
5.40449262e-01 -3.73278320e-01 -5.04307486e-02 -1.48708314e-01
-1.16648123e-01 -1.50221288e-01 5.43613374e-01 -1.40023911e+00
3.24748844e-01 -7.82219023e-02 4.77864802e-01 -7.03187883e-01
4.92485881e-01 -9.17085409e-01 2.22062439e-01 3.29592347e-01
1.50750875e-01 1.90877363e-01 1.22539029e-02 5.14543295e-01
-8.45217228e-01 -2.24265471e-01 9.54617083e-01 -3.70919742e-02
-1.20662594e+00 3.11809123e-01 -9.37328339e-02 -1.35600984e-01
1.23292470e+00 -9.09897327e-01 -4.13942635e-01 -1.37406439e-01
-3.51459920e-01 -1.89735383e-01 9.28356051e-01 1.25982255e-01
7.95845628e-01 -1.57097888e+00 -2.90757716e-01 4.76933002e-01
4.43693310e-01 -3.04292321e-01 4.70793724e-01 9.53334451e-01
-9.87483203e-01 5.55030465e-01 -6.51436269e-01 -9.26194310e-01
-1.37652850e+00 4.52806443e-01 2.29219183e-01 7.67848492e-02
-6.43571615e-01 5.26684701e-01 6.49880394e-02 9.28745791e-02
1.17690958e-01 -3.89172971e-01 -9.27616581e-02 2.03390613e-01
4.84114915e-01 5.20312130e-01 2.66560256e-01 -1.26638782e+00
-6.86906874e-01 1.55520892e+00 2.79571831e-01 2.46511593e-01
1.23541343e+00 -4.92983758e-01 -1.61211938e-01 1.43790245e-01
1.57858980e+00 -1.53981045e-01 -1.13293254e+00 -3.41561466e-01
-2.18327895e-01 -9.62755740e-01 1.11672290e-01 -1.78953961e-01
-1.18984950e+00 8.97448897e-01 1.08042026e+00 1.61315382e-01
1.35349882e+00 -1.39799595e-01 6.40569210e-01 3.57563883e-01
8.58188570e-01 -1.04960012e+00 -9.33190882e-02 1.48425445e-01
7.65512288e-01 -1.37744391e+00 3.70292395e-01 -2.63257056e-01
-3.58318388e-01 1.31889832e+00 1.64599955e-01 -2.44942695e-01
5.89259148e-01 1.26270890e-01 -6.20155483e-02 -2.12193102e-01
-1.00372829e-01 -4.17511761e-01 3.93402964e-01 6.74020886e-01
2.24956140e-01 1.52042881e-01 -5.61682522e-01 -1.63634941e-01
1.48019865e-01 -1.73709303e-01 2.23813757e-01 1.05562890e+00
-5.60331643e-01 -1.05477226e+00 -5.48292875e-01 2.76507556e-01
-2.32800081e-01 9.05558988e-02 -9.20178965e-02 1.01887989e+00
2.87189782e-01 8.11797976e-01 1.13967955e-01 -2.92019248e-01
2.99547404e-01 -5.02326012e-01 6.07298195e-01 -1.22606687e-01
-2.57753789e-01 -2.95419127e-01 -3.53886664e-01 -8.85046899e-01
-8.78761530e-01 -5.75392723e-01 -9.18997228e-01 -3.13569665e-01
-2.89286107e-01 -6.35891110e-02 6.87613666e-01 7.33603776e-01
1.79818586e-01 7.92881027e-02 7.28726745e-01 -1.16727090e+00
-2.99311608e-01 -8.07858467e-01 -8.69463742e-01 4.51514810e-01
2.93343663e-01 -1.21161091e+00 -5.09652376e-01 -2.38592654e-01] | [10.44772720336914, -0.34803661704063416] |
3153da2b-7417-4a1d-9acb-1d8a45be9e8f | latent-dictionary-learning-for-sparse | null | null | http://openaccess.thecvf.com/content_cvpr_2014/html/Yang_Latent_Dictionary_Learning_2014_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2014/papers/Yang_Latent_Dictionary_Learning_2014_CVPR_paper.pdf | Latent Dictionary Learning for Sparse Representation based Classification | Dictionary learning (DL) for sparse coding has shown promising results in classification tasks, while how to adaptively build the relationship between dictionary atoms and class labels is still an important open question. The existing dictionary learning approaches simply fix a dictionary atom to be either class-specific or shared by all classes beforehand, ignoring that the relationship needs to be updated during DL. To address this issue, in this paper we propose a novel latent dictionary learning (LDL) method to learn a discriminative dictionary and build its relationship to class labels adaptively. Each dictionary atom is jointly learned with a latent vector, which associates this atom to the representation of different classes. More specifically, we introduce a latent representation model, in which discrimination of the learned dictionary is exploited via minimizing the within-class scatter of coding coefficients and the latent-value weighted dictionary coherence. The optimal solution is efficiently obtained by the proposed solving algorithm. Correspondingly, a latent sparse representation based classifier is also presented. Experimental results demonstrate that our algorithm outperforms many recently proposed sparse representation and dictionary learning approaches for action, gender and face recognition. | ['Luc van Gool', 'Dengxin Dai', 'Lilin Shen', 'Meng Yang'] | 2014-06-01 | null | null | null | cvpr-2014-6 | ['sparse-representation-based-classification'] | ['computer-vision'] | [ 2.47670919e-01 -4.80396627e-03 -7.06409633e-01 -5.12323558e-01
-4.28861022e-01 -2.46954978e-01 3.25959027e-01 2.65608191e-01
-4.39820997e-03 4.77302372e-01 3.80321503e-01 2.48410657e-01
-1.41188011e-01 -7.31743813e-01 -2.04989001e-01 -1.11850357e+00
1.55383244e-01 5.89398146e-01 -3.73736858e-01 3.02267391e-02
2.82928228e-01 4.85525638e-01 -1.57941556e+00 1.22400768e-01
7.18150437e-01 9.81881857e-01 2.00878337e-01 -1.72502935e-01
-2.37239510e-01 7.27858663e-01 -3.02765965e-01 -1.33241003e-03
2.23821014e-01 -7.19482303e-01 -3.27594787e-01 5.44523776e-01
6.61346078e-01 2.63489112e-02 -2.20060766e-01 1.25650942e+00
4.66330618e-01 2.49564677e-01 8.37231934e-01 -1.01891220e+00
-7.24311054e-01 6.04654193e-01 -7.89128840e-01 2.85057366e-01
2.09056675e-01 -5.22525370e-01 1.15174949e+00 -1.20017719e+00
6.41366243e-01 1.23465621e+00 4.70470935e-01 2.74077713e-01
-1.26878726e+00 -1.05103898e+00 5.06082594e-01 3.26662660e-01
-1.77714467e+00 -6.36850297e-01 1.15431738e+00 -5.53900123e-01
4.77349579e-01 2.49915794e-02 7.45221317e-01 8.38597178e-01
1.07293963e-01 5.16984403e-01 1.04081702e+00 -6.85004413e-01
3.36525619e-01 2.06912965e-01 2.35490173e-01 1.05498374e+00
2.79400170e-01 -3.95706259e-02 -8.68697107e-01 -3.65117282e-01
8.83383632e-01 1.83453545e-01 -1.43999264e-01 -7.73609698e-01
-9.68272924e-01 1.29965115e+00 -7.29476288e-02 5.42024076e-01
-3.91139746e-01 -8.81625637e-02 2.04366580e-01 3.28623325e-01
3.81362855e-01 -7.99659118e-02 4.98754962e-04 3.18214118e-01
-9.97441113e-01 -7.85989463e-02 7.22514093e-01 8.91333878e-01
9.95886385e-01 3.77585500e-01 -1.98913187e-01 1.26896369e+00
5.96904397e-01 5.96031308e-01 6.83313310e-01 -7.65519917e-01
1.08634047e-01 5.82735121e-01 -2.79830992e-01 -1.76686883e+00
-9.94782224e-02 -6.03587091e-01 -9.63776946e-01 -2.93318003e-01
5.75935319e-02 3.43529463e-01 -6.07417881e-01 1.62721014e+00
4.96511728e-01 7.81702816e-01 -9.95915160e-02 8.31949294e-01
7.49174714e-01 6.71825290e-01 4.97751497e-02 -8.08312654e-01
1.17257333e+00 -7.89770842e-01 -1.06145287e+00 -1.36900634e-01
3.45311105e-01 -8.33674431e-01 5.78684151e-01 6.00987732e-01
-6.88674271e-01 -6.42477214e-01 -9.40737128e-01 3.29248421e-02
1.27696514e-01 3.68597448e-01 8.51039708e-01 6.00753784e-01
-7.25023568e-01 3.98673564e-02 -6.82824135e-01 -3.27067733e-01
2.17211559e-01 5.63539267e-01 -4.80076730e-01 -3.80935639e-01
-9.88897443e-01 4.67041612e-01 3.06661010e-01 -3.88684273e-02
-1.06705141e+00 -3.48120660e-01 -1.03062975e+00 4.50174734e-02
1.33604348e-01 -2.78329313e-01 6.05724454e-01 -9.41024721e-01
-1.36008728e+00 9.88925517e-01 -4.50262040e-01 -1.57385245e-01
-3.39314967e-01 1.84711173e-01 -4.42698717e-01 1.79617867e-01
3.35723937e-01 4.14613396e-01 1.51289630e+00 -1.37587869e+00
-5.07816970e-01 -4.54508752e-01 -1.40473217e-01 2.85177559e-01
-7.84423470e-01 -1.97881132e-01 -5.88876307e-01 -1.09795272e+00
8.00300121e-01 -7.86522686e-01 -2.02937543e-01 6.77106604e-02
8.83532166e-02 -3.71638924e-01 8.46326590e-01 -3.80960971e-01
1.55711961e+00 -2.38706589e+00 6.43165231e-01 5.48476815e-01
1.13125935e-01 -1.26766458e-01 -2.58530170e-01 3.23063374e-01
-2.71087974e-01 -6.33387923e-01 -2.86387980e-01 -5.13404906e-01
-1.82317033e-01 6.38435781e-01 -4.98118222e-01 8.21379304e-01
-3.55942249e-01 2.33550578e-01 -9.52260613e-01 -7.64380634e-01
1.12922817e-01 7.21056879e-01 -8.13983440e-01 1.92832232e-01
2.21543938e-01 5.82383454e-01 -4.31899309e-01 1.01163387e+00
6.17846429e-01 -1.62820697e-01 7.12490320e-01 -5.36370456e-01
3.11687328e-02 -1.27139419e-01 -1.65482879e+00 1.90014648e+00
-4.83150959e-01 2.80614346e-01 2.34525785e-01 -1.58406162e+00
1.47089624e+00 2.55507320e-01 7.58731902e-01 -5.75694859e-01
1.28406614e-01 3.07393372e-01 -3.06436628e-01 -3.13118637e-01
1.52445778e-01 -4.42840368e-01 5.08560687e-02 3.26441407e-01
2.56397575e-01 1.87045947e-01 1.48418350e-02 7.01983050e-02
5.87435007e-01 -3.67184728e-01 5.84486187e-01 -3.99788678e-01
8.96315932e-01 -2.79293299e-01 9.01807964e-01 5.18142641e-01
4.83065993e-02 3.75785261e-01 2.05192491e-01 -4.46491718e-01
-5.68772018e-01 -7.58748770e-01 -5.70484459e-01 1.19278812e+00
2.91041166e-01 -5.15734732e-01 -3.02976578e-01 -5.66622972e-01
2.34821856e-01 3.69329154e-01 -6.36873305e-01 -2.81761616e-01
-4.94535387e-01 -5.23489654e-01 -6.35772664e-03 1.22688845e-01
2.67347544e-01 -8.53569686e-01 8.67356881e-02 3.50494027e-01
-1.42044514e-01 -7.56103098e-01 -6.32356584e-01 2.13922516e-01
-9.75425661e-01 -9.65357721e-01 -4.50867832e-01 -1.24696290e+00
1.01771235e+00 6.14042044e-01 7.93981194e-01 2.11000741e-01
-2.12300703e-01 5.62501729e-01 -5.94047606e-01 1.80537611e-01
-6.49781823e-02 -1.09686039e-01 3.69912803e-01 7.09928870e-01
5.61244309e-01 -8.69015992e-01 -3.42715472e-01 3.67842108e-01
-7.14451671e-01 -3.74228030e-01 4.92829531e-01 1.17533135e+00
1.19489491e+00 2.30361789e-01 5.28930366e-01 -1.06195319e+00
4.35244292e-01 -8.04595947e-01 -4.92043316e-01 9.36597660e-02
-5.78829467e-01 -5.14221713e-02 3.42806607e-01 -5.55758238e-01
-7.32103586e-01 3.67915511e-01 2.20077038e-02 -6.42313719e-01
1.76332533e-01 7.41239190e-01 -4.14026022e-01 -3.16208392e-01
3.75442028e-01 4.82534796e-01 -1.09051622e-01 -5.92592299e-01
3.97035390e-01 6.57725394e-01 2.07974121e-01 -8.44012737e-01
7.38137305e-01 6.53272212e-01 -3.54506820e-02 -9.59404826e-01
-1.13893640e+00 -8.16063166e-01 -8.97593439e-01 -5.15636876e-02
5.87639153e-01 -1.28928351e+00 -1.98523045e-01 1.44878076e-02
-8.86590004e-01 1.53561279e-01 -5.48210919e-01 5.53092837e-01
-3.70930672e-01 6.52571917e-01 -1.59985080e-01 -4.94020402e-01
7.03610703e-02 -1.09662652e+00 1.08513212e+00 -1.50423031e-02
-1.82799846e-01 -1.24032903e+00 1.66417345e-01 3.14768583e-01
1.09389350e-01 4.83783968e-02 8.75912309e-01 -4.46812421e-01
-3.04980278e-01 -1.46495923e-01 -3.11853364e-02 2.96990663e-01
5.35460651e-01 -5.73685825e-01 -5.85356653e-01 -8.34333181e-01
4.25568193e-01 -3.08550686e-01 7.66358495e-01 2.71244377e-01
1.23386908e+00 -3.11633736e-01 -5.07115006e-01 8.12654018e-01
1.31918454e+00 2.41028622e-01 3.21904212e-01 -9.89045724e-02
9.01846647e-01 3.61345768e-01 6.93698466e-01 8.64486098e-01
2.08134457e-01 9.22346294e-01 1.36808576e-02 1.83084100e-01
-1.88502148e-01 -2.02469036e-01 3.33345920e-01 1.37039185e+00
1.56927258e-01 2.01506674e-01 -6.60058260e-01 5.04129231e-01
-1.73785686e+00 -1.02944744e+00 1.06667973e-01 1.90069556e+00
8.88989866e-01 -1.22890070e-01 -1.74159080e-01 3.96616280e-01
6.29259109e-01 3.77239645e-01 -4.81510311e-01 -1.89708039e-01
-1.44114405e-01 4.47665244e-01 1.29655644e-01 7.47965515e-01
-9.90202427e-01 1.02313900e+00 6.37743473e+00 1.09898496e+00
-1.09192502e+00 4.81787056e-01 2.48869702e-01 2.43406639e-01
-4.67097670e-01 1.82584524e-01 -1.14636946e+00 3.55818272e-01
3.73197585e-01 -5.53446747e-02 4.16943520e-01 1.00874889e+00
1.43147886e-01 1.04401402e-01 -1.02822900e+00 1.50179827e+00
6.72452033e-01 -1.14552951e+00 5.87519944e-01 1.42887652e-01
9.76320624e-01 -6.44042850e-01 2.67418653e-01 2.98817337e-01
-2.98218317e-02 -9.64201033e-01 3.95435184e-01 5.27302504e-01
8.40298951e-01 -6.91763282e-01 4.24104214e-01 2.72822946e-01
-1.50973380e+00 -2.79490322e-01 -7.95798182e-01 2.93714984e-04
-2.46538609e-01 8.32902789e-01 -3.25495809e-01 1.15677252e-01
2.31549665e-01 1.30794322e+00 -3.85642290e-01 7.47405171e-01
-2.75633693e-01 7.59248734e-01 -8.39766115e-03 4.89728689e-01
1.70580521e-02 -6.21555805e-01 4.57112223e-01 1.11662459e+00
3.55183125e-01 3.98309141e-01 7.76749372e-01 4.75830436e-01
-3.52356955e-02 5.00367641e-01 -6.88715160e-01 1.22277150e-02
8.41337144e-01 1.23414218e+00 -7.37181604e-01 -2.55451262e-01
-5.15750110e-01 9.94847536e-01 3.23188066e-01 2.43291691e-01
-5.97302139e-01 7.31252655e-02 6.73836529e-01 9.56488848e-02
4.86987472e-01 -5.09668589e-01 -8.81805569e-02 -1.40504551e+00
-3.86695862e-01 -1.05912936e+00 6.10225141e-01 -1.23545378e-01
-1.32138777e+00 3.27532202e-01 -2.68525607e-03 -1.40863252e+00
9.63299349e-02 -2.78616875e-01 -1.33695990e-01 6.74544334e-01
-1.67735457e+00 -1.10564697e+00 -3.14797193e-01 1.03479660e+00
6.70127153e-01 -8.47346365e-01 9.16621625e-01 6.52368009e-01
-4.99254704e-01 8.45756471e-01 2.84875453e-01 -3.92997153e-02
7.19249249e-01 -7.90954709e-01 -6.82462633e-01 4.77153242e-01
5.83099246e-01 9.66624022e-01 4.61722463e-01 -5.26487529e-01
-1.46385312e+00 -8.99274230e-01 1.02087843e+00 -5.16715273e-02
5.36376357e-01 -1.83001533e-01 -8.31200480e-01 6.55559242e-01
-1.35936096e-01 3.36327314e-01 1.29687726e+00 3.15906525e-01
-5.81622422e-01 -4.09247100e-01 -9.74270999e-01 2.66577816e-03
1.08824039e+00 -7.80057192e-01 -4.79421318e-01 6.08302891e-01
2.60882050e-01 -2.80718476e-01 -8.96448076e-01 1.47523686e-01
5.38947046e-01 -6.89904273e-01 1.00506806e+00 -2.88523108e-01
1.35773141e-02 -5.12198985e-01 -6.43566966e-01 -1.02801239e+00
-7.32356250e-01 -1.84026539e-01 -4.55103725e-01 1.24264324e+00
-2.66946137e-01 -3.01428050e-01 8.27323556e-01 -1.02429025e-01
-4.76721041e-02 -6.72672570e-01 -1.16364491e+00 -6.37452185e-01
-1.88127428e-01 -2.59350445e-02 3.44802827e-01 1.61977136e+00
-1.08807415e-01 4.73166943e-01 -7.26179302e-01 3.59086663e-01
8.97620201e-01 5.04304349e-01 4.74266827e-01 -1.42533457e+00
-3.97714913e-01 -1.22460626e-01 -5.52482188e-01 -1.31103718e+00
6.25927448e-01 -1.39682150e+00 -3.70609432e-01 -1.22084498e+00
2.75044441e-01 -9.13020968e-01 -5.30375302e-01 6.81030393e-01
9.81796384e-02 2.94618905e-01 -6.81829453e-02 6.81152940e-01
-5.60040534e-01 6.84937119e-01 1.21381342e+00 -4.99472708e-01
-1.84396446e-01 -2.87635207e-01 -8.34412098e-01 4.81907070e-01
4.29339468e-01 -7.77394235e-01 -6.48780823e-01 -3.79133791e-01
3.45446467e-02 1.31881833e-01 3.59115228e-02 -1.05669725e+00
3.15742284e-01 -3.29771549e-01 2.40667492e-01 -3.69354904e-01
4.04070824e-01 -1.09174800e+00 3.22921544e-01 4.04187471e-01
-3.10082257e-01 -3.74811113e-01 -2.01225117e-01 8.46190572e-01
-5.73583007e-01 -4.50605810e-01 1.06103849e+00 3.80367152e-02
-7.04363585e-01 5.50579250e-01 -3.20615530e-01 -2.69588411e-01
8.95452261e-01 -3.72028172e-01 4.82045978e-01 -1.71039402e-01
-1.10962641e+00 -6.76694140e-02 1.46376431e-01 4.06530499e-01
8.47155571e-01 -1.82017279e+00 -5.84486067e-01 6.39304459e-01
2.96799183e-01 -2.74487376e-01 6.46289662e-02 4.99342352e-01
-1.28573745e-01 2.65360743e-01 -2.68801779e-01 -6.45122111e-01
-1.20986879e+00 4.83972341e-01 1.59631759e-01 -1.12169191e-01
-5.63654125e-01 1.12603188e+00 1.65322736e-01 -1.89478219e-01
4.42489654e-01 5.10937385e-02 -6.64907813e-01 6.40155971e-01
3.80545408e-01 1.81586612e-02 -1.70742005e-01 -1.14892983e+00
-2.05519274e-01 1.08527756e+00 -1.18588760e-01 2.78749794e-01
1.24550724e+00 -1.38072908e-01 -5.39450049e-01 5.03336668e-01
1.24087512e+00 2.93194771e-01 -7.98316300e-01 -5.76421320e-01
-2.17258245e-01 -8.66668701e-01 3.11100483e-01 -1.47109374e-01
-1.49497128e+00 6.41240060e-01 8.16705465e-01 -2.47030959e-01
1.16643476e+00 -1.17204420e-01 7.58187175e-01 3.47677916e-01
7.00562537e-01 -8.35617185e-01 4.47536707e-01 4.77701485e-01
8.43438864e-01 -1.22884691e+00 3.94386142e-01 -6.97313488e-01
-2.31102422e-01 1.08510411e+00 4.24610645e-01 -2.84295350e-01
1.15578878e+00 -2.02064663e-01 -1.47429481e-01 -4.42852080e-01
-3.51897627e-01 -1.08093666e-02 3.70886624e-01 5.29908419e-01
4.91890937e-01 4.43698317e-02 -6.59284353e-01 5.03981471e-01
5.71475066e-02 -1.86857879e-01 1.64576005e-02 7.85578191e-01
-5.87787867e-01 -1.67385602e+00 -4.23172653e-01 2.50537038e-01
3.20758112e-02 8.96559469e-03 -3.36657315e-02 1.99584767e-01
7.16506004e-01 8.20351124e-01 1.27489017e-02 -2.25031018e-01
1.34980768e-01 -1.31568849e-01 5.38118958e-01 -1.19862437e+00
-1.30991697e-01 3.61850262e-01 -3.50171924e-01 -3.49822849e-01
-7.39546061e-01 -9.39519167e-01 -1.09885716e+00 6.10454343e-02
-4.20024484e-01 5.24197400e-01 3.52137208e-01 7.55828381e-01
4.29197922e-02 1.21399380e-01 8.73126447e-01 -6.18079901e-01
-3.32686245e-01 -5.48436403e-01 -1.03571713e+00 3.98061335e-01
1.08133361e-01 -1.22865677e+00 -3.65984917e-01 3.76647085e-01] | [12.407349586486816, 0.40259262919425964] |
7def38f3-a982-4494-819b-4f7b10b18ffc | is-domain-adaptation-worth-your-investment | null | null | https://aclanthology.org/2021.econlp-1.5 | https://aclanthology.org/2021.econlp-1.5.pdf | Is Domain Adaptation Worth Your Investment? Comparing BERT and FinBERT on Financial Tasks | With the recent rise in popularity of Transformer models in Natural Language Processing, research efforts have been dedicated to the development of domain-adapted versions of BERT-like architectures. In this study, we focus on FinBERT, a Transformer model trained on text from the financial domain. By comparing its performances with the original BERT on a wide variety of financial text processing tasks, we found continual pretraining from the original model to be the more beneficial option. Domain-specific pretraining from scratch, conversely, seems to be less effective. | ['Chu-Ren Huang', 'Yu-Yin Hsu', 'Emmanuele Chersoni', 'Bo Peng'] | null | null | null | null | emnlp-econlp-2021-11 | ['continual-pretraining'] | ['methodology'] | [-1.81215256e-01 1.28311083e-01 1.84298396e-01 -5.84812105e-01
-4.68221933e-01 -5.27312994e-01 1.06890821e+00 1.75322413e-01
-6.83938146e-01 6.66026831e-01 3.87615770e-01 -6.59625769e-01
1.50967360e-01 -6.19078219e-01 -2.55418271e-01 -1.24713190e-01
1.50207132e-01 7.87195504e-01 4.02404010e-01 -6.05306745e-01
3.71283591e-01 3.75165254e-01 -9.35353935e-01 4.93137121e-01
4.92719918e-01 9.63646412e-01 1.88854873e-01 4.25290972e-01
-5.52636325e-01 1.21879053e+00 -5.46687841e-01 -7.79555678e-01
3.12477261e-01 -4.72943306e-01 -9.44233119e-01 -4.94351238e-02
-2.19267812e-02 -2.56516397e-01 -4.01640385e-01 7.85262227e-01
2.89372563e-01 7.64901116e-02 6.26741946e-01 -7.86354005e-01
-5.92353165e-01 9.01166558e-01 -1.18681997e-01 7.01623023e-01
2.19214603e-01 -1.40999109e-01 1.13928652e+00 -1.19089282e+00
6.43914461e-01 1.10185266e+00 7.43201435e-01 3.32085013e-01
-1.27081645e+00 -2.88161397e-01 3.64366174e-02 1.55854434e-01
-9.86391604e-01 -5.51975131e-01 9.62578118e-01 -1.86700657e-01
1.68381798e+00 -2.25952074e-01 5.64032137e-01 8.40959489e-01
4.47812617e-01 9.66130972e-01 1.31665063e+00 -8.12671006e-01
-4.87514138e-02 3.62418175e-01 -6.66424260e-02 5.88634849e-01
8.57091472e-02 1.26972675e-01 -4.28445399e-01 -1.92558337e-02
6.61183000e-01 -2.75734752e-01 3.85554358e-02 -5.46340287e-01
-8.85728717e-01 9.25934255e-01 4.26623911e-01 7.73548424e-01
-5.19789636e-01 -6.02674484e-02 7.35303819e-01 8.27088058e-01
7.71980643e-01 6.53816581e-01 -5.53120077e-01 -3.07220221e-01
-1.11212313e+00 3.86345953e-01 9.56308782e-01 8.71872187e-01
3.84452581e-01 2.88300097e-01 2.95465112e-01 7.75465071e-01
2.75536686e-01 -1.08897060e-01 8.03489625e-01 -5.24593174e-01
6.85063004e-01 6.41481042e-01 1.74398839e-01 -7.15748608e-01
-2.66152292e-01 -3.08390290e-01 -3.67618889e-01 -5.09886071e-02
6.52920187e-01 -1.70544431e-01 -1.01599801e+00 1.32187116e+00
-8.77936333e-02 -3.46062630e-01 2.88321674e-01 6.61294520e-01
3.07300091e-01 6.23368144e-01 2.04136059e-01 -1.73778459e-01
9.69236135e-01 -7.32958615e-01 -5.58606684e-01 -3.83430421e-01
5.66037357e-01 -9.75202143e-01 9.15747523e-01 5.72274029e-01
-1.15059400e+00 -3.52217317e-01 -9.62357998e-01 -1.32498056e-01
-6.52535558e-01 -1.69112533e-01 6.34295166e-01 7.94344902e-01
-1.15485597e+00 6.69794798e-01 -8.90888512e-01 -5.68832695e-01
4.52703536e-02 1.97779968e-01 -1.53170153e-01 -8.98363590e-02
-1.10571635e+00 1.25568342e+00 6.29683852e-01 1.72362998e-01
-5.45285463e-01 -2.72464752e-01 -6.14605546e-01 6.13754280e-02
3.10477287e-01 -4.54191536e-01 1.68062460e+00 -1.17316544e+00
-1.79377961e+00 8.90769899e-01 6.72383010e-02 -1.07264364e+00
5.18658638e-01 -3.30718070e-01 -4.12601292e-01 1.98648006e-01
-1.13684215e-01 4.32182163e-01 8.76545608e-01 -8.03871155e-01
-6.47242486e-01 -1.06754184e-01 1.22909442e-01 6.83785826e-02
-6.28829002e-01 5.74804306e-01 2.92522535e-02 -9.84559953e-01
-2.16939926e-01 -5.41591465e-01 -4.61512320e-02 -6.54712975e-01
2.92703286e-02 -4.83585835e-01 6.55230463e-01 -7.05650210e-01
1.09946716e+00 -2.00484848e+00 -6.42400831e-02 5.23202121e-02
7.77108595e-02 5.37848651e-01 -3.44809256e-02 7.36541450e-01
-2.03478381e-01 8.84491950e-02 -1.27029583e-01 -5.95778465e-01
7.35776722e-02 3.69091541e-01 -6.43355608e-01 2.03435481e-01
5.51533043e-01 8.22317660e-01 -8.29010785e-01 -6.21771574e-01
2.06734315e-01 9.44497287e-02 -5.52517176e-01 2.35061675e-01
-3.25738698e-01 1.56239226e-01 -6.42649472e-01 5.26105821e-01
2.11987823e-01 -4.71718907e-02 1.05476610e-01 1.95058122e-01
-2.49181688e-01 9.14665699e-01 -7.51686096e-01 1.31050205e+00
-4.80005354e-01 5.85371673e-01 1.56600714e-01 -1.27764189e+00
8.66284072e-01 6.90197945e-01 3.42603534e-01 -8.64495635e-01
3.97972524e-01 2.64442801e-01 1.86386898e-01 -3.34646136e-01
6.56356931e-01 -8.73700738e-01 -1.11495830e-01 6.21828258e-01
2.25231349e-01 -2.43366167e-01 4.65498745e-01 4.07248698e-02
1.15162301e+00 3.96326512e-01 5.14843941e-01 -3.35585117e-01
4.18944001e-01 5.38955331e-01 3.23918700e-01 5.08070529e-01
-3.44201267e-01 3.60318124e-01 2.75105834e-01 -6.19513154e-01
-1.01906312e+00 -9.64934289e-01 -1.82634383e-01 1.06765068e+00
-5.73649704e-01 -5.22873998e-01 -5.49329519e-01 -6.26446009e-01
-1.64557233e-01 9.29005980e-01 -3.48738909e-01 8.61436203e-02
-8.87962997e-01 -4.94756997e-01 5.59115350e-01 6.50804341e-01
6.87921703e-01 -1.31450045e+00 -9.37386930e-01 8.24073195e-01
-7.99900517e-02 -1.03290009e+00 -1.50247991e-01 4.93443847e-01
-1.05006218e+00 -7.98408508e-01 -8.98356915e-01 -6.81210518e-01
9.74438712e-02 -1.15407176e-01 1.34493673e+00 -2.55719483e-01
1.66625455e-01 2.92989463e-01 -6.96741283e-01 -7.07726836e-01
-5.90633154e-01 3.49975258e-01 -2.68083155e-01 -2.36201495e-01
5.08037090e-01 -3.08186650e-01 -7.55612627e-02 1.77315235e-01
-1.00700486e+00 -2.95969516e-01 3.12800109e-01 9.55598652e-01
4.24803458e-02 1.37081027e-01 5.55463433e-01 -1.03838587e+00
9.90161657e-01 -2.58881003e-01 -4.57650661e-01 1.08071834e-01
-7.00543404e-01 1.61060408e-01 1.08913851e+00 -2.26700082e-01
-1.21171784e+00 -2.93429732e-01 -2.13807866e-01 -2.89087832e-01
-1.10612269e-02 1.00874901e+00 1.94339544e-01 4.94638160e-02
6.50203228e-01 2.09685534e-01 -2.78308451e-01 -6.20703876e-01
-2.42674146e-02 3.95167917e-01 3.44694823e-01 -5.35348654e-01
5.70347726e-01 1.49806058e-02 -3.67940098e-01 -9.13094282e-01
-8.35691154e-01 -3.76166999e-01 -4.80788112e-01 -2.80487891e-02
5.67290425e-01 -6.09164178e-01 1.01328216e-01 5.28089941e-01
-9.75943148e-01 -4.84564424e-01 -2.69946039e-01 2.88816720e-01
-5.00095665e-01 2.41898119e-01 -7.96063006e-01 -7.08788335e-01
-2.85849184e-01 -7.44207501e-01 6.41317725e-01 -1.55655697e-01
-3.17946255e-01 -1.36352968e+00 2.53497779e-01 6.68810084e-02
6.88381732e-01 -5.01557477e-02 8.73177528e-01 -1.12109113e+00
-2.15912014e-01 -5.20684361e-01 3.33591215e-02 5.67773819e-01
1.21664993e-01 -2.92574883e-01 -8.28781962e-01 -1.33240506e-01
4.43431199e-01 -4.75947946e-01 7.48636246e-01 -6.90987855e-02
5.45348048e-01 -9.55028385e-02 -4.85332757e-02 3.65299404e-01
1.37584794e+00 2.34412625e-01 3.98864508e-01 5.66508770e-01
1.80377498e-01 6.21385396e-01 7.26607919e-01 3.02967042e-01
3.78211826e-01 4.71297830e-01 6.40434623e-02 8.04225821e-03
2.94922650e-01 -4.06983286e-01 6.35951757e-01 8.02733779e-01
-4.14537899e-02 -3.98210257e-01 -1.25899446e+00 6.21270299e-01
-1.59907079e+00 -9.56189573e-01 3.49203676e-01 1.78935635e+00
6.14947736e-01 7.32506275e-01 3.06965709e-01 1.33462161e-01
3.33898664e-01 7.72163570e-02 -2.96005011e-01 -6.62016451e-01
-8.83555934e-02 5.15062988e-01 1.88080177e-01 1.76027566e-01
-1.07154500e+00 1.21559823e+00 7.78799820e+00 4.60801005e-01
-1.34441364e+00 -3.05989198e-03 4.87941444e-01 4.12684008e-02
-4.92072999e-02 1.01983309e-01 -5.98611057e-01 2.57571697e-01
1.53411663e+00 -3.99093926e-01 3.05217981e-01 7.39324749e-01
1.47149846e-01 -1.44017627e-02 -1.13300669e+00 5.03435373e-01
-2.32598484e-02 -1.08422756e+00 1.31416395e-01 -7.53545240e-02
1.12177230e-01 1.63268685e-01 -2.17925727e-01 5.71114898e-01
6.16939783e-01 -8.19381773e-01 1.08331430e+00 1.46233097e-01
1.92312792e-01 -6.42486870e-01 1.00450647e+00 4.36091155e-01
-1.08750105e+00 -1.31214380e-01 -5.13854146e-01 -2.96357453e-01
2.82762319e-01 2.96144485e-01 -1.15609181e+00 5.81521928e-01
6.83695436e-01 7.55290568e-01 -5.50673842e-01 7.50497282e-01
-1.16193004e-01 9.00798559e-01 -4.40868407e-01 -1.10190198e-01
5.18955231e-01 -2.05389395e-01 3.57297570e-01 1.45321727e+00
1.26801789e-01 8.99765491e-02 1.94076926e-01 5.23836195e-01
1.21758357e-01 2.42623582e-01 -8.20708811e-01 -4.96926785e-01
-7.82965198e-02 1.07868147e+00 -8.04309189e-01 -6.27010226e-01
-7.58438528e-01 8.14120412e-01 5.18716455e-01 1.17057107e-01
-5.39556861e-01 -2.50030160e-01 1.41923875e-01 3.31272304e-01
7.04003453e-01 -5.80325544e-01 -3.44723821e-01 -1.11285639e+00
-3.66543755e-02 -9.05049682e-01 7.38569260e-01 -8.15872431e-01
-1.21347129e+00 8.74161601e-01 3.42015058e-01 -1.08053124e+00
-8.72581840e-01 -8.60532582e-01 -6.08298659e-01 1.03742158e+00
-1.60484147e+00 -8.03093016e-01 4.54227060e-01 8.24607193e-01
5.06485701e-01 -4.19962078e-01 7.96586215e-01 2.24895030e-01
-3.20396036e-01 3.30597311e-01 8.66275188e-03 2.93098956e-01
6.55556142e-01 -1.35752118e+00 5.32652617e-01 9.85725045e-01
2.63230264e-01 7.85353422e-01 1.09863365e+00 -3.80077928e-01
-1.04927099e+00 -8.06200802e-01 1.39211607e+00 -3.01842570e-01
1.13920569e+00 -5.27906597e-01 -9.17492568e-01 8.72019887e-01
6.27251148e-01 -5.78603506e-01 4.02862906e-01 4.21312498e-03
-3.30066115e-01 -1.62522390e-01 -1.05495095e+00 4.84677672e-01
7.07731485e-01 -7.07215905e-01 -1.36384356e+00 -6.24195971e-02
3.31658185e-01 7.44299870e-03 -8.49603534e-01 1.34372905e-01
1.84343547e-01 -9.64351118e-01 6.78618252e-01 -5.32617688e-01
4.98244762e-01 1.14212431e-01 -5.50309904e-02 -1.38937187e+00
-3.21776420e-01 -5.48156440e-01 2.22296745e-01 1.20637238e+00
4.56138998e-01 -8.52115452e-01 7.90137589e-01 6.21028125e-01
-3.29254389e-01 -3.44995439e-01 -9.59950686e-01 -8.01980793e-01
4.99409139e-01 -3.96499097e-01 2.57331729e-01 8.70730340e-01
5.13794005e-01 6.11010134e-01 -1.27613708e-01 -4.46325064e-01
9.95306149e-02 3.55975926e-01 6.11448705e-01 -1.21410871e+00
-4.93978202e-01 -5.53419709e-01 -1.70723289e-01 -1.02006996e+00
1.00076340e-01 -8.40608478e-01 9.50757340e-02 -1.44742787e+00
-2.03130737e-01 -1.52401123e-02 -4.16921526e-01 7.14193881e-01
2.29453042e-01 1.02471381e-01 2.77792096e-01 1.94794908e-01
-2.22155511e-01 5.49834728e-01 1.02770829e+00 -9.75984558e-02
-1.48791164e-01 1.04712173e-02 -6.58514798e-01 6.06702089e-01
9.46177423e-01 -3.56021494e-01 -4.41923916e-01 -3.96777004e-01
1.71283200e-01 1.41344771e-01 -1.47066921e-01 -8.69118512e-01
1.22406751e-01 -5.61598800e-02 1.60717338e-01 -3.35163713e-01
3.94561529e-01 -7.47341216e-01 -1.49425194e-01 4.67407376e-01
-2.83710510e-01 5.66588044e-01 3.91785026e-01 1.90418854e-01
-7.50353277e-01 -3.64205688e-01 7.12462366e-01 -5.60288191e-01
-7.18926311e-01 -5.45438044e-02 -7.53098667e-01 1.66732892e-01
6.44286454e-01 -3.73890400e-01 1.36333302e-01 -1.73896000e-01
-6.01154089e-01 -1.37579903e-01 4.04054224e-01 4.51710671e-01
5.00272512e-01 -7.48822749e-01 -5.73878109e-01 5.17476536e-02
1.86625514e-02 -2.77377576e-01 -5.45139253e-01 6.41177833e-01
-5.02272487e-01 1.01600182e+00 -3.42392117e-01 -1.25330919e-02
-1.04249990e+00 7.41934955e-01 4.28636700e-01 -7.27599144e-01
-6.82170391e-01 6.01618946e-01 -1.13516100e-01 -2.71483868e-01
-4.28234190e-02 -6.40341640e-01 -2.62229145e-01 9.48000401e-02
3.02858144e-01 -3.91880013e-02 4.14988309e-01 -4.54420298e-01
-3.25403899e-01 4.79834341e-03 -4.53933030e-01 -3.94989431e-01
1.66039228e+00 -4.48145112e-03 8.55997205e-02 7.25357354e-01
6.73410416e-01 -1.86732456e-01 -1.02099311e+00 -5.31693876e-01
6.88438654e-01 -3.37949693e-01 1.49277389e-01 -6.62519395e-01
-7.32077122e-01 8.92821610e-01 5.80515563e-02 7.68491328e-01
1.30884731e+00 -2.29337066e-01 5.89952707e-01 5.63758850e-01
7.23083556e-01 -1.33809745e+00 -4.95251156e-02 1.01912677e+00
7.78263867e-01 -9.75965858e-01 -8.30358267e-02 -7.38353282e-02
-6.05176151e-01 1.29905629e+00 2.42051080e-01 -4.19528544e-01
7.73907363e-01 3.77127260e-01 2.03302413e-01 -2.37969935e-01
-9.28277194e-01 7.44674820e-03 -1.43857747e-01 3.56707633e-01
8.85340273e-01 -2.23543510e-01 -3.85216087e-01 4.20409232e-01
-3.72088283e-01 3.01545829e-01 4.74622279e-01 1.27624416e+00
-4.64008212e-01 -1.53950572e+00 -4.06548709e-01 4.81880605e-01
-6.95850849e-01 -3.72207791e-01 -7.63494909e-01 1.07702172e+00
-3.78503501e-01 9.70924795e-01 3.08735576e-03 5.51036606e-03
4.66833800e-01 6.09998047e-01 5.23758173e-01 -7.57964611e-01
-1.13271654e+00 9.09965187e-02 5.29464424e-01 -2.66242623e-01
-6.62075102e-01 -1.05282319e+00 -1.00012636e+00 -3.54444563e-01
2.40870863e-02 2.24688336e-01 3.37622136e-01 1.14331985e+00
2.28068605e-02 2.36063018e-01 2.83578932e-01 -6.82991266e-01
-9.53011930e-01 -1.20092392e+00 -5.81215978e-01 1.51120588e-01
2.75061488e-01 -4.58608925e-01 -9.00305063e-02 2.16334462e-01] | [10.667414665222168, 8.851781845092773] |
59df0fda-b690-43ff-9eb7-b3be4f985204 | investigating-the-decoders-of-maximum | 2003.03716 | null | https://arxiv.org/abs/2003.03716v1 | https://arxiv.org/pdf/2003.03716v1.pdf | Investigating the Decoders of Maximum Likelihood Sequence Models: A Look-ahead Approach | We demonstrate how we can practically incorporate multi-step future information into a decoder of maximum likelihood sequence models. We propose a "k-step look-ahead" module to consider the likelihood information of a rollout up to k steps. Unlike other approaches that need to train another value network to evaluate the rollouts, we can directly apply this look-ahead module to improve the decoding of any sequence model trained in a maximum likelihood framework. We evaluate our look-ahead module on three datasets of varying difficulties: IM2LATEX-100k OCR image to LaTeX, WMT16 multimodal machine translation, and WMT14 machine translation. Our look-ahead module improves the performance of the simpler datasets such as IM2LATEX-100k and WMT16 multimodal machine translation. However, the improvement of the more difficult dataset (e.g., containing longer sequences), WMT14 machine translation, becomes marginal. Our further investigation using the k-step look-ahead suggests that the more difficult tasks suffer from the overestimated EOS (end-of-sentence) probability. We argue that the overestimated EOS probability also causes the decreased performance of beam search when increasing its beam width. We tackle the EOS problem by integrating an auxiliary EOS loss into the training to estimate if the model should emit EOS or other words. Our experiments show that improving EOS estimation not only increases the performance of our proposed look-ahead module but also the robustness of the beam search. | ['Yen-Ling Kuo', 'Yu-Siang Wang', 'Boris Katz'] | 2020-03-08 | null | null | null | null | ['multimodal-machine-translation'] | ['natural-language-processing'] | [ 3.81403089e-01 1.19096830e-01 -1.36547700e-01 -5.13709307e-01
-1.32134235e+00 -6.53972447e-01 6.38286948e-01 -1.36542618e-01
-6.44390941e-01 8.44810367e-01 -2.20700447e-02 -8.87947977e-01
1.71834454e-01 -3.30805004e-01 -1.13036454e+00 -5.31533122e-01
3.53817821e-01 5.92554986e-01 2.38279045e-01 -1.60514846e-01
4.19041634e-01 2.03003377e-01 -8.01966846e-01 5.56948066e-01
9.76860762e-01 6.73641682e-01 7.35489309e-01 9.29351687e-01
5.56197092e-02 3.93738240e-01 -5.06464660e-01 -6.75837338e-01
1.91021010e-01 -4.93506879e-01 -8.84672046e-01 -1.45080313e-01
3.24649215e-01 -4.26819801e-01 1.98935885e-02 1.09953892e+00
5.72733104e-01 -1.96419269e-01 5.45801699e-01 -1.08326888e+00
-3.55516404e-01 7.36794531e-01 -5.16686678e-01 1.17340103e-01
4.36884522e-01 4.97869015e-01 6.90375209e-01 -1.02935719e+00
7.67809033e-01 1.47966659e+00 6.28693640e-01 5.17452538e-01
-1.02194047e+00 -6.64747119e-01 -1.16005018e-01 4.49532688e-01
-1.19210231e+00 -5.02249539e-01 2.91771859e-01 -1.83388144e-01
1.11458564e+00 1.99186683e-01 6.79810941e-02 1.24821877e+00
2.80231535e-01 7.74298728e-01 1.18934715e+00 -6.55290008e-01
9.86028165e-02 2.09512383e-01 1.84533969e-02 6.50495410e-01
-2.28508562e-01 -2.37700660e-02 -6.35062754e-01 4.38367511e-04
4.40071106e-01 -6.14558876e-01 -2.98535585e-01 1.21312521e-01
-1.54777300e+00 6.77018046e-01 3.90721485e-02 2.02414140e-01
-4.37719598e-02 1.37064457e-01 2.88061738e-01 5.24841726e-01
4.15118009e-01 5.60112953e-01 -7.09908187e-01 -6.16546988e-01
-1.15901291e+00 -8.17683712e-02 7.89469182e-01 9.95860696e-01
7.60660410e-01 -3.72800499e-01 -1.62050992e-01 8.46238375e-01
3.40082914e-01 7.79573321e-01 3.86999041e-01 -9.93368745e-01
1.13279831e+00 2.64984697e-01 1.64535299e-01 -4.86063987e-01
-3.11440289e-01 -3.34948361e-01 -4.88501728e-01 1.50165573e-01
7.29149401e-01 -3.52295905e-01 -1.01958501e+00 1.77793539e+00
8.27019215e-02 -3.91743451e-01 1.21879295e-01 9.26556289e-01
2.22558215e-01 1.00545812e+00 -2.31322736e-01 -3.93322587e-01
1.09630871e+00 -1.23606312e+00 -7.87502468e-01 -2.84495652e-01
1.17096698e+00 -1.28626585e+00 1.15503967e+00 3.52165222e-01
-1.09409118e+00 -3.92007798e-01 -1.04997718e+00 -2.83678174e-02
-1.15443558e-01 3.17291170e-01 3.37133527e-01 4.88957018e-01
-1.09101152e+00 8.57213259e-01 -7.83948481e-01 -3.72227818e-01
-1.11027829e-01 3.93117726e-01 -3.85759085e-01 -1.77268222e-01
-1.25783837e+00 1.31539738e+00 3.68590713e-01 6.46776035e-02
-6.18531585e-01 -4.31314439e-01 -7.91511953e-01 7.33370474e-03
3.73360485e-01 -4.84326869e-01 1.46305132e+00 -9.38747227e-01
-1.71252680e+00 4.75129634e-01 -6.82139516e-01 -2.95745701e-01
7.56794035e-01 -1.78287283e-01 -8.16356242e-02 -3.67233045e-02
-8.95807371e-02 8.20254862e-01 9.02080774e-01 -1.06095171e+00
-3.42775881e-01 -3.84252854e-02 -2.76107788e-02 3.76622260e-01
-2.56946236e-01 1.89797267e-01 -5.84116697e-01 -5.21838725e-01
-4.39995453e-02 -1.19252622e+00 -4.40734737e-02 -2.15354130e-01
-4.63308245e-01 -8.53306726e-02 5.37590563e-01 -9.08608019e-01
1.23977113e+00 -2.03648305e+00 3.53604555e-01 -1.01246787e-02
-4.44832861e-01 2.70480812e-01 -5.35900712e-01 4.79826421e-01
-4.98919301e-02 2.07976073e-01 -2.39531502e-01 -6.03705168e-01
-1.16795795e-02 2.65520990e-01 -1.62034169e-01 2.25762591e-01
3.57450873e-01 9.48935866e-01 -6.76536381e-01 -4.52933401e-01
-6.30320013e-02 1.22465514e-01 -4.28897917e-01 2.77250886e-01
-4.49066967e-01 4.35554087e-01 1.75644174e-01 5.04764915e-01
7.71369934e-01 -4.66814488e-01 1.50079340e-01 -4.88900058e-02
-1.00313753e-01 3.94913733e-01 -7.88397729e-01 1.72585332e+00
-5.10483384e-01 8.04678798e-01 -2.63556778e-01 -3.99492890e-01
6.44259036e-01 2.31690526e-01 -2.80640334e-01 -9.01836932e-01
-9.11282003e-02 5.79307854e-01 2.02083990e-01 -5.00425220e-01
6.53264463e-01 -7.07444772e-02 1.52287200e-01 3.94391418e-01
9.52941775e-02 -1.43734649e-01 1.50692642e-01 2.40950629e-01
9.66293335e-01 3.52639109e-01 1.13732787e-02 3.48578840e-02
4.35503900e-01 6.01081587e-02 2.57813781e-01 7.46056795e-01
6.97051808e-02 6.61566436e-01 4.66250926e-01 3.50281447e-02
-1.45507860e+00 -7.66791880e-01 -8.03118497e-02 1.03119874e+00
-1.72466075e-03 -3.85419756e-01 -9.64532495e-01 -8.46962869e-01
-3.34084570e-01 1.02994776e+00 -2.39079848e-01 -8.26259851e-02
-8.78020108e-01 -1.01747274e+00 5.43836296e-01 3.06593478e-01
3.59236002e-01 -7.78664708e-01 -1.54420748e-01 5.75559027e-02
-6.54667437e-01 -1.28981340e+00 -5.88030517e-01 2.68570006e-01
-9.10775363e-01 -5.61690688e-01 -1.09037256e+00 -5.24196148e-01
7.22231925e-01 -7.50696659e-02 8.84119034e-01 3.55993062e-02
3.31790209e-01 2.69347765e-02 -5.48204780e-01 -1.14950165e-01
-1.07781756e+00 3.94464761e-01 -1.03751689e-01 -3.72214675e-01
8.90356675e-02 -1.46051154e-01 -4.25029904e-01 6.09204948e-01
-5.67024410e-01 4.71392572e-01 7.18589425e-01 9.80906904e-01
1.21971361e-01 -3.52415919e-01 4.14107800e-01 -3.88486534e-01
4.80546057e-01 -3.01104367e-01 -5.65335333e-01 4.86156046e-01
-7.78512061e-01 4.50339109e-01 6.17976248e-01 -6.93255603e-01
-9.47053194e-01 -1.23678371e-01 -4.78053540e-01 -2.54467160e-01
1.86105266e-01 4.44943786e-01 -4.21366431e-02 -8.12622681e-02
4.42205191e-01 3.15335900e-01 -8.49990323e-02 -5.32479465e-01
2.67206818e-01 8.02503467e-01 4.04451281e-01 -4.96015996e-01
7.30443001e-01 -9.40111578e-02 -9.46858823e-02 -3.68021250e-01
-4.81103927e-01 -2.38299668e-01 -3.97275001e-01 -1.89486071e-01
7.12315679e-01 -6.63856566e-01 -6.23099327e-01 6.87622249e-01
-1.67146909e+00 -6.16067111e-01 2.16553241e-01 6.06421411e-01
-7.23069668e-01 6.25150859e-01 -7.89222360e-01 -7.40280986e-01
-3.06843996e-01 -1.50798011e+00 1.29870749e+00 -3.67417224e-02
-3.39033961e-01 -8.20618868e-01 -7.58830458e-02 4.94339794e-01
3.10270369e-01 -5.30668199e-01 1.22803676e+00 -6.23189807e-01
-6.94061875e-01 1.30244985e-01 -2.77717084e-01 3.73959184e-01
-1.45296484e-01 -2.12235987e-01 -8.56836557e-01 -3.77788275e-01
7.61336982e-02 -4.15805817e-01 8.66723418e-01 -8.41059610e-02
6.70814514e-01 -3.74730051e-01 -1.12333171e-01 5.87354600e-01
1.27199483e+00 3.50541830e-01 8.56976330e-01 5.65224528e-01
4.56141472e-01 4.41083074e-01 8.81399035e-01 9.21372101e-02
3.22114825e-01 9.44036067e-01 4.54262227e-01 1.67183504e-02
-2.33020067e-01 -2.02324554e-01 9.46426570e-01 1.27348506e+00
1.66193828e-01 -7.02916145e-01 -9.99585569e-01 3.06765735e-01
-1.78061175e+00 -5.42137682e-01 -1.53833345e-01 2.19738865e+00
1.33143997e+00 1.84379011e-01 -2.81925559e-01 -2.04348154e-02
6.82019174e-01 -1.01175182e-01 -4.59763795e-01 -6.55258596e-01
-1.89177960e-01 -1.15446165e-01 5.18699229e-01 9.40702081e-01
-6.97142422e-01 1.10055375e+00 6.40048361e+00 1.33460248e+00
-1.17291379e+00 2.71816254e-01 6.98312640e-01 3.21565047e-02
-3.44679058e-01 2.17766777e-01 -1.04947329e+00 7.37281680e-01
1.25934362e+00 2.77776331e-01 5.48530996e-01 4.89881992e-01
1.93495095e-01 -4.80095208e-01 -1.32583475e+00 7.54607260e-01
1.10809900e-01 -1.11170769e+00 3.71769443e-02 -3.73760872e-02
6.08909070e-01 1.02015175e-01 7.28902593e-02 1.92324445e-01
4.79311682e-02 -9.56776917e-01 7.50494003e-01 4.88727629e-01
9.10364509e-01 -5.01306176e-01 8.97943139e-01 7.43744612e-01
-4.96479630e-01 1.10251412e-01 -3.09764951e-01 4.82418686e-02
4.78736371e-01 6.33245826e-01 -1.37809408e+00 5.45146227e-01
2.78649122e-01 1.07278697e-01 -6.72615170e-01 8.51838171e-01
-2.86390126e-01 7.14474797e-01 -3.38033080e-01 -2.39177451e-01
2.61460364e-01 -9.00288895e-02 7.00339437e-01 1.44050658e+00
6.33289576e-01 -3.00102055e-01 -3.46658081e-01 6.40990198e-01
-1.27067059e-01 -5.94361313e-02 -2.89548784e-01 -2.33885199e-01
2.76459962e-01 9.65407789e-01 -4.61435229e-01 -5.14849246e-01
-2.18866825e-01 1.41457927e+00 3.27654690e-01 3.99483830e-01
-9.27620828e-01 -1.45960182e-01 2.18812153e-01 -2.68091615e-02
3.78874272e-01 -4.90697950e-01 -3.49245638e-01 -1.12248659e+00
1.13419406e-01 -1.21768272e+00 -1.83828011e-01 -1.24535370e+00
-8.73735428e-01 6.02566898e-01 -1.04797147e-02 -1.10473704e+00
-7.41137564e-01 -7.54046440e-01 -1.55889943e-01 1.08132577e+00
-1.54116881e+00 -9.06441033e-01 2.16004178e-01 1.58804625e-01
8.57866824e-01 1.52953146e-02 6.82510912e-01 3.29815924e-01
-3.33039016e-01 9.96605456e-01 2.48861492e-01 -1.13372549e-01
1.01887619e+00 -1.01903152e+00 5.68475366e-01 7.63416946e-01
-3.40139642e-02 6.45839036e-01 8.04712534e-01 -7.78204560e-01
-1.40126204e+00 -7.70016372e-01 1.35580826e+00 -5.36593199e-01
6.40681803e-01 -4.60213661e-01 -8.75319421e-01 5.48726201e-01
4.26374197e-01 -6.54092312e-01 1.75713167e-01 -1.10505484e-01
-3.01919043e-01 2.39897966e-01 -9.82641816e-01 6.84583962e-01
8.60543489e-01 -4.90021050e-01 -5.42462230e-01 5.09154737e-01
1.02939558e+00 -4.97021675e-01 -7.51481891e-01 6.53833926e-01
5.73483706e-01 -6.63985670e-01 5.94526410e-01 -3.84001911e-01
6.51633382e-01 -1.06302284e-01 -3.11601609e-01 -1.37660217e+00
1.12948520e-02 -7.76472688e-01 -2.34099999e-02 1.12189519e+00
9.44117129e-01 -4.21888441e-01 5.41196644e-01 4.05379206e-01
-1.67325791e-02 -7.98280299e-01 -1.12288129e+00 -8.65719199e-01
3.39778662e-01 -5.58747470e-01 2.68587649e-01 7.38944471e-01
4.11877185e-02 3.44395638e-01 -6.75415695e-01 7.59862289e-02
2.97702044e-01 -1.09448992e-01 7.19301462e-01 -4.40921634e-01
-6.21097386e-01 -9.36520100e-02 -1.53233530e-02 -1.56673813e+00
-8.24308544e-02 -8.24728370e-01 3.16498160e-01 -1.32140732e+00
5.31381667e-01 -1.78830460e-01 1.34426087e-01 5.29762805e-01
-4.09007013e-01 1.45981550e-01 5.03013909e-01 2.67351538e-01
-5.23992658e-01 4.29748833e-01 1.45366561e+00 -6.29391372e-02
8.94010961e-02 -2.00431183e-01 -1.00112736e-01 5.91603220e-01
7.13654101e-01 -7.17880070e-01 -1.62909776e-01 -7.52484024e-01
6.47045135e-01 2.48357907e-01 2.73951173e-01 -6.96258068e-01
2.98133790e-01 -7.69535229e-02 3.36488068e-01 -6.90291166e-01
4.56239313e-01 -6.53599501e-01 -1.13777541e-01 3.97677153e-01
-3.39479923e-01 2.96711981e-01 2.02223852e-01 1.33562163e-01
4.65694144e-02 -7.69788742e-01 5.14824808e-01 -9.07500088e-02
-3.49000901e-01 -3.39938045e-01 -5.34862816e-01 -2.05546483e-01
4.73008722e-01 -1.53609022e-01 -4.91904795e-01 -5.51867127e-01
-6.88370764e-01 3.14729869e-01 6.22950792e-01 3.96247625e-01
4.74445194e-01 -1.06707799e+00 -6.56459212e-01 1.75625488e-01
-1.53214410e-01 -2.92544127e-01 -1.16065510e-01 1.13474464e+00
-3.86313468e-01 4.17239606e-01 7.84526542e-02 -7.55435765e-01
-1.36574745e+00 3.40757161e-01 1.40317887e-01 -5.33612370e-01
-1.99551627e-01 7.59687483e-01 -4.19797935e-02 -5.72301626e-01
2.50112504e-01 -2.37450048e-01 1.87664852e-01 -1.59605071e-01
3.03116292e-01 3.15236509e-01 1.25563323e-01 -3.94189805e-01
-3.10739398e-01 6.04322731e-01 -2.55374312e-01 -6.98466718e-01
9.58964229e-01 -4.19159859e-01 -1.38826519e-01 4.58458751e-01
1.23464942e+00 -6.91967737e-03 -1.08300102e+00 -5.43333553e-02
1.68392867e-01 -2.96557307e-01 -4.34250385e-01 -1.25357389e+00
-4.48289901e-01 1.17902899e+00 6.31188631e-01 -3.89895469e-01
9.62993801e-01 2.39147106e-03 8.92846346e-01 6.49206877e-01
4.07615721e-01 -1.07852340e+00 1.04674146e-01 1.01145244e+00
9.57198262e-01 -1.33926260e+00 -1.48424357e-01 -3.25185090e-01
-5.65371692e-01 1.31978393e+00 4.27621961e-01 5.59576690e-01
-4.47842851e-02 3.94487739e-01 1.84180424e-01 3.07188958e-01
-9.11373317e-01 1.69941410e-01 2.89825469e-01 1.06963269e-01
2.91154116e-01 -2.54020989e-01 -4.80029851e-01 2.76762158e-01
-3.68478209e-01 9.95030627e-02 4.43264157e-01 7.27277219e-01
-4.19981271e-01 -1.30784667e+00 -4.12008286e-01 -5.63703477e-03
-3.46504688e-01 -4.93070990e-01 -3.59661311e-01 6.20070159e-01
6.35047117e-03 9.58397806e-01 -1.43498361e-01 -5.07887959e-01
-3.53891104e-02 5.27345657e-01 6.22019589e-01 -3.29247117e-01
-4.33899641e-01 2.22945049e-01 4.49393272e-01 -4.16814029e-01
-2.43154824e-01 -4.65818673e-01 -1.15653968e+00 -3.86764348e-01
-5.98889053e-01 2.07267851e-02 1.16820347e+00 1.16960895e+00
2.99831986e-01 2.66108274e-01 5.46188772e-01 -6.67870760e-01
-1.00022137e+00 -1.24150670e+00 1.88069910e-01 1.84648305e-01
3.58231306e-01 -1.90642208e-01 -4.05025840e-01 -3.38701047e-02] | [11.648837089538574, 10.249390602111816] |
918b516f-5887-4b5a-af1d-948d26e84dce | multimodal-neural-machine-translation-system | null | null | https://aclanthology.org/2021.mmtlrl-1.6 | https://aclanthology.org/2021.mmtlrl-1.6.pdf | Multimodal Neural Machine Translation System for English to Bengali | Multimodal Machine Translation (MMT) systems utilize additional information from other modalities beyond text to improve the quality of machine translation (MT). The additional modality is typically in the form of images. Despite proven advantages, it is indeed difficult to develop an MMT system for various languages primarily due to the lack of a suitable multimodal dataset. In this work, we develop an MMT for English-> Bengali using a recently published Bengali Visual Genome (BVG) dataset that contains images with associated bilingual textual descriptions. Through a comparative study of the developed MMT system vis-a-vis a Text-to-text translation, we demonstrate that the use of multimodal data not only improves the translation performance improvement in BLEU score of +1.3 on the development set, +3.9 on the evaluation test, and +0.9 on the challenge test set but also helps to resolve ambiguities in the pure text description. As per best of our knowledge, our English-Bengali MMT system is the first attempt in this direction, and thus, can act as a baseline for the subsequent research in MMT for low resource languages. | ['Petr Motlicek', 'Satya Ranjan Dash', 'Arghyadeep Sen', 'Ketan Kotwal', 'Satya Prakash Biswal', 'Subhadarshi Panda', 'Shantipriya Parida'] | null | null | null | null | mmtlrl-ranlp-2021-9 | ['multimodal-machine-translation'] | ['natural-language-processing'] | [ 3.44245315e-01 -1.83951244e-01 -2.17574555e-02 -2.01930791e-01
-1.42605746e+00 -9.01263535e-01 1.09239793e+00 -2.74011511e-02
-5.70256829e-01 8.12967122e-01 -2.68737786e-02 -4.72588927e-01
4.09948230e-01 -2.88988441e-01 -8.18386853e-01 -6.25092864e-01
6.28096879e-01 8.70316207e-01 -3.67005542e-02 -3.96618277e-01
1.85521737e-01 1.79434836e-01 -1.03204644e+00 5.71697652e-01
8.98194075e-01 6.65074825e-01 3.62833381e-01 4.92109090e-01
3.57738100e-02 4.72054064e-01 -5.46177268e-01 -9.84683573e-01
1.08091436e-01 -6.87629104e-01 -7.90907741e-01 4.47379090e-02
7.38630772e-01 -2.29575381e-01 -1.30677015e-01 8.17197025e-01
8.35640430e-01 -2.37320527e-01 6.39522314e-01 -1.12189448e+00
-9.22700226e-01 4.60881919e-01 -5.07353723e-01 -1.46594614e-01
4.46611851e-01 4.41830643e-02 8.50976884e-01 -1.43550718e+00
1.08362353e+00 1.10254776e+00 2.40171567e-01 5.90456545e-01
-1.13961339e+00 -3.48138481e-01 -3.91148895e-01 3.46492887e-01
-1.36670971e+00 -7.25400686e-01 3.99660796e-01 -2.96786129e-01
9.83589590e-01 3.45978171e-01 4.03217971e-01 1.28954422e+00
8.49075541e-02 8.90405536e-01 1.55512488e+00 -8.28923464e-01
-1.05667703e-01 4.03158307e-01 -4.99764949e-01 6.40296817e-01
-5.02172932e-02 -2.15587124e-01 -7.04157352e-01 2.24182174e-01
4.63048995e-01 -6.33505344e-01 -2.61368781e-01 -2.49030471e-01
-1.68504512e+00 7.43329883e-01 1.95676357e-01 3.77633125e-01
-1.47665322e-01 -1.35743007e-01 3.54146898e-01 5.08605421e-01
4.33828026e-01 2.98821032e-01 -1.22194402e-01 -3.68091524e-01
-1.03396487e+00 -2.60480288e-02 6.68830216e-01 9.72063422e-01
5.73275805e-01 -8.92380774e-02 -1.23267598e-01 1.07866240e+00
2.81539530e-01 1.05728745e+00 4.08320725e-01 -6.77498281e-01
1.10330570e+00 5.00450015e-01 1.27589732e-01 -8.44321549e-01
8.55509415e-02 -5.05329147e-02 -5.92099786e-01 -1.61741152e-01
4.86509115e-01 -2.78014969e-02 -1.03276145e+00 1.70281827e+00
1.26689345e-01 -6.96774602e-01 4.34714675e-01 1.09265554e+00
1.08618939e+00 8.40970933e-01 -1.54328480e-01 -3.80914301e-01
1.35109222e+00 -1.03785098e+00 -8.59066725e-01 -2.35356495e-01
6.42593801e-01 -1.58050144e+00 1.06879258e+00 4.69509996e-02
-1.16480887e+00 -2.90175021e-01 -9.21964824e-01 -1.56830519e-01
-3.80973428e-01 5.42514205e-01 1.45274624e-01 6.54223204e-01
-1.31820822e+00 -1.60322055e-01 -6.87393308e-01 -9.10302103e-01
2.95720324e-02 2.82506704e-01 -9.14862514e-01 -5.44560730e-01
-1.07838595e+00 1.39930081e+00 2.25672334e-01 1.66733801e-01
-8.46801519e-01 -8.03366229e-02 -8.42502356e-01 -4.93112594e-01
2.52946734e-01 -5.65409422e-01 1.26516593e+00 -1.11451578e+00
-1.37736368e+00 1.22821546e+00 -4.57553685e-01 -1.30130142e-01
8.54449570e-01 2.23139971e-02 -2.76453942e-01 4.09960747e-01
3.27087760e-01 9.73699927e-01 6.71208024e-01 -1.38241601e+00
-5.65036714e-01 -3.19157779e-01 -8.08133483e-02 5.52878559e-01
-1.82419434e-01 3.03355098e-01 -9.18350101e-01 -6.42290473e-01
2.27994267e-02 -1.35477114e+00 2.47607499e-01 -3.55526924e-01
-3.40297699e-01 1.07616991e-01 7.08044410e-01 -1.10749936e+00
8.04436386e-01 -1.97951138e+00 5.27482450e-01 -1.52962178e-01
-1.16321929e-01 1.77009612e-01 -2.92731583e-01 8.65725100e-01
2.46899739e-01 1.46838412e-01 -3.24868947e-01 -5.02048075e-01
4.64102067e-02 2.36431450e-01 1.23711117e-02 3.54906023e-01
3.51909518e-01 1.19057357e+00 -5.69852412e-01 -7.24094331e-01
1.46396771e-01 4.82229948e-01 1.06334835e-01 9.98774394e-02
1.27313556e-02 5.67019641e-01 -9.22793224e-02 8.33963335e-01
5.74067593e-01 7.10898563e-02 2.23926231e-01 -2.03960031e-01
-2.39504471e-01 2.13449989e-02 -5.61196923e-01 1.89475989e+00
-3.86578798e-01 1.05879998e+00 -5.78479543e-02 -8.60745847e-01
6.41283393e-01 6.71251237e-01 1.18308738e-01 -1.07141018e+00
9.05948803e-02 6.48190975e-01 1.51722968e-01 -5.35930157e-01
6.17654860e-01 -1.19909137e-01 -1.96684837e-01 4.42511588e-01
1.91434354e-01 -2.02664837e-01 5.13203025e-01 1.09342098e-01
5.26519954e-01 2.18577206e-01 -3.98764834e-02 5.31732291e-02
4.73404825e-01 6.24769628e-01 1.72420368e-01 5.82122564e-01
-7.57735670e-02 8.78740907e-01 1.62483513e-01 -1.22168891e-01
-1.36521351e+00 -9.49604452e-01 -1.34006992e-01 9.29436207e-01
-7.12636113e-02 -3.11096728e-01 -1.03510296e+00 -5.66804051e-01
-4.93749022e-01 7.25183666e-01 -4.40122485e-01 1.79757848e-01
-6.51968062e-01 -9.17007387e-01 8.05755615e-01 3.09259325e-01
7.13372767e-01 -7.64727652e-01 -1.81384742e-01 -6.03449345e-03
-9.24977243e-01 -1.59912395e+00 -5.17697871e-01 -1.74274787e-01
-7.17462361e-01 -7.64490545e-01 -1.27168763e+00 -1.08540142e+00
6.58387601e-01 3.23397666e-01 9.53881621e-01 -3.21088672e-01
7.61999413e-02 4.43243891e-01 -4.81708467e-01 -2.97026396e-01
-8.61189425e-01 5.39676473e-02 -4.27775159e-02 -5.93708977e-02
2.71994889e-01 4.69736289e-03 -3.12021255e-01 5.49672663e-01
-8.23887408e-01 5.58476031e-01 7.89097846e-01 8.23211908e-01
5.54950953e-01 -4.98101264e-01 3.61061782e-01 -3.94519061e-01
4.55091804e-01 -1.16220012e-01 -5.27989566e-01 6.66470826e-01
-5.31139553e-01 -2.47819036e-01 2.03773662e-01 -4.78722870e-01
-1.06826448e+00 -1.57359600e-01 7.72577822e-02 -1.04892448e-01
6.43960340e-03 7.17097998e-01 -8.59755352e-02 -2.13102639e-01
3.85785490e-01 5.91747701e-01 -1.20969802e-01 -3.17905843e-01
4.50890452e-01 9.15157557e-01 5.42269945e-01 -4.31717843e-01
8.24735820e-01 1.98032603e-01 -7.04567432e-02 -6.56676590e-01
-2.79457748e-01 -1.84230745e-01 -7.72030294e-01 -2.85607219e-01
1.00683558e+00 -1.13692737e+00 -8.27806890e-02 4.68223363e-01
-1.14560747e+00 -1.99060962e-01 4.28284824e-01 5.77434897e-01
-5.31195283e-01 3.45267266e-01 -6.67818785e-01 -5.68769991e-01
-3.23707968e-01 -1.55277920e+00 1.13578451e+00 -1.05039053e-01
-7.01807365e-02 -8.43564212e-01 4.84572817e-03 1.08731699e+00
3.23263109e-01 4.38562520e-02 9.82203782e-01 -4.35810506e-01
-5.52284896e-01 -2.54809052e-01 -3.93763453e-01 3.06039035e-01
5.73619716e-02 -4.75162901e-02 -8.63980889e-01 -4.56883639e-01
-2.02610806e-01 -4.90587801e-01 6.46260142e-01 -2.37859432e-02
-4.33006026e-02 -7.39554837e-02 2.09192447e-02 9.56346467e-02
1.33573949e+00 1.85041651e-01 7.13202238e-01 6.00374818e-01
7.96524405e-01 5.56427896e-01 8.33338380e-01 -1.51775241e-01
7.81512797e-01 1.10614514e+00 1.13075942e-01 -3.34665745e-01
-4.21906292e-01 -1.59861624e-01 6.01255834e-01 1.16081369e+00
-3.04894358e-01 -4.70573008e-01 -1.10158336e+00 6.58127904e-01
-1.87321818e+00 -6.75215364e-01 -1.43734410e-01 2.18036580e+00
9.89963055e-01 -4.39606696e-01 -5.25758043e-02 -2.00978145e-01
7.74469316e-01 -2.95566529e-01 7.05319121e-02 -5.18803358e-01
-5.25436997e-01 -7.46173933e-02 2.36122400e-01 3.18058878e-01
-9.59939480e-01 1.16003692e+00 5.80034256e+00 8.78033400e-01
-1.25448608e+00 3.58663052e-01 4.47468162e-01 1.39496699e-01
-2.16218252e-02 3.88691127e-02 -5.19198954e-01 1.78808972e-01
1.04510760e+00 1.41348481e-01 5.47125697e-01 1.69271186e-01
2.60821730e-01 -2.23191991e-01 -1.17618382e+00 1.15467656e+00
5.40245533e-01 -1.07321787e+00 3.05899531e-01 1.98899999e-01
8.57595801e-01 2.14879170e-01 3.35434854e-01 1.80223927e-01
-3.08520645e-01 -1.13668692e+00 1.09171963e+00 1.18090548e-01
1.03153741e+00 -6.36072099e-01 1.05068004e+00 3.12327176e-01
-7.24781930e-01 5.20191669e-01 -1.79221451e-01 2.67025858e-01
9.04117152e-02 -4.89307269e-02 -1.25853229e+00 8.06495249e-01
4.45614129e-01 3.75441462e-01 -7.73144126e-01 7.45845675e-01
-2.61342704e-01 6.34590805e-01 -1.08418696e-01 5.45891328e-03
4.53879863e-01 -9.72291604e-02 4.88487631e-01 1.38127589e+00
5.76728404e-01 -2.74476558e-01 8.77614990e-02 4.48803842e-01
-2.65928298e-01 6.17017150e-01 -6.19753003e-01 -4.22718316e-01
7.84507096e-02 1.16529953e+00 -6.14547729e-01 -3.19904357e-01
-6.81690335e-01 1.42664123e+00 1.52161300e-01 4.97781515e-01
-6.80419505e-01 -9.57220346e-02 -3.45698223e-02 -1.55299157e-01
2.53256798e-01 -4.04755384e-01 -1.02209389e-01 -1.42009914e+00
2.37263665e-01 -1.33874488e+00 9.30338427e-02 -1.13244283e+00
-1.03850412e+00 8.39301407e-01 3.67860943e-02 -1.34821701e+00
-4.91802096e-01 -4.57464993e-01 6.76027983e-02 1.07061601e+00
-1.27559614e+00 -1.77256382e+00 -9.19045974e-03 5.16810179e-01
6.59798741e-01 -3.90444547e-01 9.79783297e-01 3.42158288e-01
-4.67772663e-01 7.48811781e-01 3.33704978e-01 2.07121223e-01
1.19660807e+00 -1.06144094e+00 2.29788616e-01 9.50557709e-01
5.23383021e-01 5.01306891e-01 6.45162404e-01 -6.03908181e-01
-1.70884097e+00 -8.27982962e-01 1.20966256e+00 -6.76391661e-01
5.22559345e-01 -4.12078530e-01 -6.07277572e-01 5.82299352e-01
8.72685969e-01 -5.49480855e-01 6.39878631e-01 -2.52471417e-01
-3.81494105e-01 1.75993979e-01 -1.09885097e+00 7.35334873e-01
5.12566447e-01 -6.99541628e-01 -5.77379048e-01 3.47548991e-01
4.20027167e-01 -5.19780695e-01 -1.05548680e+00 4.75162774e-01
5.58822632e-01 -5.42831421e-01 7.11339116e-01 -2.61844963e-01
7.00853705e-01 -4.51043457e-01 -6.19631648e-01 -1.31144834e+00
1.68385208e-01 -4.42742854e-01 4.33518499e-01 1.43821943e+00
8.69444132e-01 -2.99865395e-01 2.72497177e-01 2.40689576e-01
-3.50522362e-02 -3.13553572e-01 -1.27659690e+00 -5.07734179e-01
2.16336861e-01 -3.39095801e-01 7.56097510e-02 9.81185555e-01
-7.65145421e-02 7.01933563e-01 -5.70437133e-01 8.24718773e-02
4.93510872e-01 2.46287957e-01 7.94616878e-01 -5.40774941e-01
-4.41812053e-02 -3.59923035e-01 -3.37374121e-01 -6.12711787e-01
-4.58752289e-02 -1.05853415e+00 8.65349621e-02 -1.76504827e+00
6.45943999e-01 4.66349125e-02 1.28192872e-01 5.31912982e-01
3.47663239e-02 9.06845510e-01 4.91085380e-01 5.66891134e-01
-5.10438502e-01 3.98951858e-01 1.50642943e+00 -2.79674441e-01
2.06695899e-01 -4.32250082e-01 -4.70433056e-01 2.67488450e-01
7.58026600e-01 -3.47068191e-01 -2.72414297e-01 -9.92519438e-01
1.69833407e-01 1.52935311e-01 2.01552451e-01 -4.91280615e-01
-2.18169298e-02 -1.32785231e-01 5.22980809e-01 -5.96223474e-01
4.71413910e-01 -5.39062798e-01 2.13289931e-01 1.72553569e-01
-1.53242707e-01 4.99876231e-01 4.24304783e-01 1.48930162e-01
-4.21233594e-01 -8.10039714e-02 5.16534150e-01 -2.02122986e-01
-6.60291135e-01 -2.58673280e-01 -6.48908854e-01 -5.65109588e-02
6.43274367e-01 -1.94290832e-01 -5.70530832e-01 -5.75122774e-01
-4.62126374e-01 1.54919460e-01 8.55952978e-01 6.09770179e-01
5.83789587e-01 -1.49184418e+00 -1.11508739e+00 -1.53749794e-01
4.89068478e-01 -6.61500156e-01 -9.95457396e-02 1.25207114e+00
-5.98094523e-01 7.11717784e-01 -3.85980487e-01 -8.85155737e-01
-1.54214251e+00 1.60931796e-01 6.84751123e-02 -5.60370907e-02
-2.56023884e-01 3.65360200e-01 1.00786358e-01 -4.77649927e-01
-8.04749802e-02 3.14452529e-01 -5.59143759e-02 3.36837769e-02
3.41644913e-01 8.12649727e-02 3.26733679e-01 -1.41926289e+00
-4.65486944e-01 6.20567679e-01 -6.37612045e-02 -1.02903414e+00
1.05362809e+00 -5.83888292e-01 -3.79709125e-01 5.56385636e-01
1.05509424e+00 1.05001561e-01 -4.90679026e-01 -1.31773874e-01
-1.63581491e-01 -5.25294781e-01 -7.28789195e-02 -1.41523826e+00
-8.04562628e-01 9.76753712e-01 7.94841111e-01 -1.92840368e-01
1.15687656e+00 3.39322910e-02 7.27315843e-01 5.23130774e-01
3.56163621e-01 -9.98226941e-01 -6.13119677e-02 5.08189678e-01
1.11847270e+00 -1.55610049e+00 -3.31068486e-01 -1.28612429e-01
-9.39451993e-01 1.01401293e+00 2.93394864e-01 6.76684022e-01
-3.29053551e-01 -4.07265685e-02 6.52956903e-01 1.30536571e-01
-6.56953096e-01 -3.44214857e-01 6.82350695e-01 6.14652634e-01
7.76378155e-01 4.18268070e-02 -6.10288978e-01 1.32410318e-01
-9.94055197e-02 -2.06588283e-01 4.70513374e-01 9.28631485e-01
-1.03196815e-01 -1.45194423e+00 -6.66864038e-01 -1.88953966e-01
-6.77976131e-01 -2.97026008e-01 -7.95404613e-01 9.90938008e-01
-8.58449787e-02 1.29474652e+00 -3.63554060e-01 -2.94924617e-01
3.40905547e-01 2.00875536e-01 9.42256927e-01 -3.61349702e-01
-5.15583754e-01 5.79409122e-01 4.05274212e-01 -1.43832400e-01
-8.28412652e-01 -6.71775639e-01 -7.99753249e-01 -3.48485082e-01
-2.05882296e-01 9.68149304e-02 1.08573055e+00 9.75588560e-01
7.37578571e-02 3.28809284e-02 2.59182036e-01 -6.74993455e-01
-3.31530049e-02 -1.15569806e+00 4.42625536e-03 3.60085994e-01
1.16340026e-01 -4.05894965e-01 4.42226417e-02 3.95291418e-01] | [11.456869125366211, 1.4955536127090454] |
5da41bf1-881f-4d8a-8af9-ba631048bf03 | interpretable-multi-objective-reinforcement | 1809.08343 | null | http://arxiv.org/abs/1809.08343v1 | http://arxiv.org/pdf/1809.08343v1.pdf | Interpretable Multi-Objective Reinforcement Learning through Policy Orchestration | Autonomous cyber-physical agents and systems play an increasingly large role
in our lives. To ensure that agents behave in ways aligned with the values of
the societies in which they operate, we must develop techniques that allow
these agents to not only maximize their reward in an environment, but also to
learn and follow the implicit constraints of society. These constraints and
norms can come from any number of sources including regulations, business
process guidelines, laws, ethical principles, social norms, and moral values.
We detail a novel approach that uses inverse reinforcement learning to learn a
set of unspecified constraints from demonstrations of the task, and
reinforcement learning to learn to maximize the environment rewards. More
precisely, we assume that an agent can observe traces of behavior of members of
the society but has no access to the explicit set of constraints that give rise
to the observed behavior. Inverse reinforcement learning is used to learn such
constraints, that are then combined with a possibly orthogonal value function
through the use of a contextual bandit-based orchestrator that picks a
contextually-appropriate choice between the two policies (constraint-based and
environment reward-based) when taking actions. The contextual bandit
orchestrator allows the agent to mix policies in novel ways, taking the best
actions from either a reward maximizing or constrained policy. In addition, the
orchestrator is transparent on which policy is being employed at each time
step. We test our algorithms using a Pac-Man domain and show that the agent is
able to learn to act optimally, act within the demonstrated constraints, and
mix these two functions in complex ways. | ['Moninder Singh', 'Nicholas Mattei', 'Kush Varshney', 'Francesca Rossi', 'Ritesh Noothigattu', 'Rachita Chandra', 'Djallel Bouneffouf', 'Piyush Madan', 'Murray Campbell'] | 2018-09-21 | null | null | null | null | ['multi-objective-reinforcement-learning'] | ['methodology'] | [ 7.91962147e-02 1.49139091e-01 -3.28843415e-01 -2.20616803e-01
-1.31747603e-01 -7.22724438e-01 7.52286196e-01 -8.16165507e-02
-9.23556745e-01 1.09546113e+00 -5.59213124e-02 -2.94455022e-01
-5.16712904e-01 -8.37069929e-01 -6.32840037e-01 -8.68662417e-01
-1.65245339e-01 8.14040303e-01 1.47079691e-01 -2.88210213e-01
3.29814851e-01 4.47957516e-01 -1.34915853e+00 -2.40600169e-01
7.59316802e-01 6.80222869e-01 1.49876326e-01 1.01644325e+00
4.37501639e-01 9.62873459e-01 -6.08672202e-01 2.06909299e-01
5.54758430e-01 -5.61797202e-01 -4.90347683e-01 1.98586777e-01
-4.55327511e-01 -5.34802556e-01 -5.33177517e-02 1.03818512e+00
3.32835019e-02 2.83283740e-01 6.11264288e-01 -1.36827123e+00
-4.95070100e-01 5.77612936e-01 -1.12877570e-01 -2.47445866e-01
3.26378822e-01 6.78950787e-01 9.29471672e-01 2.74083614e-01
6.07259870e-01 1.08820188e+00 -2.09210187e-01 7.17840791e-01
-1.44059408e+00 -5.84900975e-01 4.25759733e-01 1.07154265e-01
-7.77246714e-01 -2.86477119e-01 5.20107567e-01 -3.19264919e-01
7.48882473e-01 1.97673693e-01 1.03558016e+00 1.07083702e+00
1.97313502e-01 3.19411248e-01 1.29860771e+00 -5.00871062e-01
8.77774119e-01 2.24820524e-01 -3.98091525e-01 4.14521664e-01
3.07024568e-01 6.20595872e-01 -2.98914820e-01 -4.51094955e-01
6.14790618e-01 -1.75085396e-01 -1.28642976e-01 -8.13537180e-01
-1.17183471e+00 4.60955292e-01 1.61923587e-01 3.14593107e-01
-7.80453563e-01 4.87253875e-01 6.75700977e-02 6.03711307e-01
-3.84434283e-01 9.46203828e-01 -6.41418576e-01 -2.59721249e-01
-1.26366794e-01 5.38058817e-01 1.04009736e+00 4.19148654e-01
7.64557898e-01 -5.75455427e-02 1.43076628e-01 4.41681296e-01
5.62409878e-01 6.35352433e-01 1.44296706e-01 -1.65944266e+00
2.53208756e-01 4.13743913e-01 8.81277800e-01 -5.38390875e-01
-1.86490137e-02 -1.69875666e-01 3.42286192e-02 8.95804644e-01
6.24744713e-01 -4.38416898e-01 -7.03790009e-01 2.09853435e+00
5.64421952e-01 -1.97881088e-02 2.87276775e-01 9.57036793e-01
-4.45495009e-01 6.40272081e-01 1.20029122e-01 -2.84123123e-01
9.05943453e-01 -2.60797560e-01 -3.98156106e-01 -2.78746128e-01
6.32608831e-01 -1.31457731e-01 9.17735279e-01 5.17782331e-01
-8.77925038e-01 1.98170200e-01 -1.22025049e+00 6.01504564e-01
-1.71618268e-01 -4.59571868e-01 2.10053295e-01 5.03456831e-01
-7.84070909e-01 8.15574825e-01 -8.30956638e-01 -2.43514329e-01
8.53819549e-02 6.09134436e-01 -9.92115438e-02 3.51236105e-01
-1.02612364e+00 1.08525884e+00 4.70419407e-01 -8.35046098e-02
-1.32204294e+00 -7.72210360e-02 -6.28089488e-01 1.20733790e-01
9.97790813e-01 -5.39956450e-01 1.50247788e+00 -1.36037493e+00
-1.88658237e+00 4.08769578e-01 6.08057201e-01 -5.54056346e-01
5.47863662e-01 1.36840984e-01 -9.61861014e-02 -2.39564497e-02
4.96124737e-02 3.42229933e-01 7.64032185e-01 -1.35207486e+00
-8.65296602e-01 -3.41516852e-01 6.28017902e-01 5.13200521e-01
-1.87124372e-01 -3.40699032e-02 8.11835751e-02 1.00977808e-01
-4.17878181e-01 -1.38819635e+00 -5.12256324e-01 -1.45092890e-01
-2.59137601e-01 3.06149404e-02 6.14976406e-01 -1.42052203e-01
7.59092271e-01 -1.97490776e+00 1.94668964e-01 5.20462573e-01
-1.39716223e-01 -8.55191890e-03 -2.14631096e-01 3.29991043e-01
4.25061315e-01 -2.56680767e-03 -2.71412045e-01 -9.35421418e-03
3.36410582e-01 6.41904712e-01 -2.62454860e-02 4.98079151e-01
-1.29865676e-01 3.37531835e-01 -1.03390431e+00 -1.36929899e-01
1.82136849e-01 -1.19375875e-02 -7.39896059e-01 4.64264125e-01
-7.45261371e-01 6.48272038e-01 -1.05236042e+00 1.40121117e-01
-7.34677985e-02 5.82710505e-02 8.62190366e-01 7.21638680e-01
-2.31244504e-01 2.86436886e-01 -1.28627193e+00 8.36054325e-01
-4.40910310e-01 1.64008528e-01 3.99633408e-01 -9.18290615e-01
6.16753638e-01 3.73564094e-01 6.02303565e-01 -6.29611671e-01
3.92700404e-01 3.49560738e-01 5.95394254e-01 -6.47735476e-01
7.26938546e-02 -2.27902904e-01 -2.54037827e-01 8.42328668e-01
-2.61191279e-01 -2.99435288e-01 1.71667427e-01 3.14229205e-02
1.18552816e+00 5.44264197e-01 5.07032871e-01 -1.45702258e-01
4.84337509e-01 1.09423451e-01 8.62433910e-01 9.26616967e-01
-4.71659571e-01 -2.30310500e-01 8.74400675e-01 -4.88780230e-01
-1.05894935e+00 -8.09236586e-01 4.67917204e-01 1.30313408e+00
1.41906992e-01 2.04500958e-01 -5.85182965e-01 -6.85594320e-01
2.34563500e-01 8.89531970e-01 -7.65248120e-01 -1.69958130e-01
-7.20246255e-01 -3.20761830e-01 -1.07756324e-01 -7.01644551e-03
3.71012956e-01 -1.39502919e+00 -1.26481855e+00 2.13531598e-01
3.42703789e-01 -8.31227601e-01 -6.13635063e-01 2.89086998e-01
-3.75027418e-01 -1.18778694e+00 -9.30239335e-02 -1.48439109e-01
7.69289792e-01 -2.82957345e-01 6.56865120e-01 6.10707887e-02
1.16974048e-01 6.00921929e-01 -1.32481396e-01 -3.94641459e-01
-8.14991295e-01 -1.72004506e-01 2.64769226e-01 9.72885415e-02
3.23105752e-02 -5.85398436e-01 -5.05585313e-01 2.18630254e-01
-8.12629461e-01 -7.77913928e-02 3.79575610e-01 6.43662333e-01
1.46168545e-01 1.21826744e-02 6.16766036e-01 -7.81539083e-01
6.86856508e-01 -4.74393904e-01 -1.19983256e+00 2.81872302e-01
-6.60014391e-01 3.81537020e-01 7.92579889e-01 -7.11333990e-01
-9.50520635e-01 2.34779462e-01 5.72270453e-01 -4.91921827e-02
-2.82574356e-01 2.60360122e-01 -4.79946256e-01 1.02901034e-01
6.62920177e-01 1.17624365e-01 2.13978037e-01 -3.28209214e-02
3.89680594e-01 7.15691447e-01 2.77937531e-01 -1.25901842e+00
7.45579720e-01 2.20561415e-01 -8.49355459e-02 -5.01516759e-01
-4.38276976e-01 2.25944549e-01 -2.49203205e-01 -5.12177110e-01
6.89119458e-01 -3.79734606e-01 -1.10694766e+00 1.55149326e-01
-6.34696722e-01 -8.83100331e-01 -4.67420876e-01 5.00780582e-01
-9.80660379e-01 -8.05356577e-02 -8.30838978e-02 -1.28294182e+00
3.01473528e-01 -1.24564552e+00 3.00974488e-01 5.47613442e-01
-2.50674933e-01 -5.92407405e-01 1.21491909e-01 2.06645548e-01
4.16724116e-01 3.02827448e-01 7.82059371e-01 -6.86421037e-01
-9.51948106e-01 3.25453952e-02 3.94699156e-01 3.80309612e-01
2.20541403e-01 2.11101741e-01 -4.89689469e-01 -3.65947068e-01
-9.64016467e-02 -2.85012394e-01 -1.17991762e-02 2.84785837e-01
5.78082979e-01 -8.29917490e-01 -2.04067990e-01 4.93776128e-02
1.20484698e+00 9.46700037e-01 2.55178869e-01 8.06258619e-01
-2.43652437e-04 8.45146060e-01 5.73578179e-01 5.06286442e-01
4.94592875e-01 6.95830345e-01 7.76015699e-01 5.15379548e-01
5.95249593e-01 -1.89658254e-01 6.04427874e-01 -2.14089956e-02
-1.50071457e-01 1.08656287e-01 -6.94685459e-01 5.11744857e-01
-2.28801203e+00 -9.77884531e-01 6.74403787e-01 2.50567913e+00
8.73808920e-01 3.52690339e-01 5.05014658e-01 -1.87047780e-01
5.20713508e-01 -1.87815368e-01 -1.04530799e+00 -6.53455198e-01
3.01562607e-01 -2.75209904e-01 6.28884554e-01 6.71810508e-01
-6.79537714e-01 7.51814604e-01 6.14934683e+00 1.78173482e-01
-9.73991454e-01 -2.03024417e-01 4.59112942e-01 -4.41787779e-01
-2.70228744e-01 4.07662153e-01 -2.73372382e-01 4.83163297e-01
1.00530565e+00 -4.74576324e-01 1.26174259e+00 7.66971350e-01
6.57781065e-01 -4.10949469e-01 -1.27980292e+00 2.50764936e-01
-5.86242497e-01 -1.04971242e+00 -6.71441853e-01 5.32771409e-01
6.97157025e-01 1.22767258e-02 -3.77387740e-02 3.59862864e-01
1.15114307e+00 -7.19709098e-01 1.10099614e+00 4.36545342e-01
1.24676175e-01 -8.20888937e-01 3.48605722e-01 8.81777406e-01
-6.23210847e-01 -6.55287206e-01 2.18778387e-01 -5.39321065e-01
-1.26493692e-01 -1.97281465e-02 -9.10344183e-01 -4.09073308e-02
3.33098620e-01 5.83358221e-02 3.82895410e-01 7.32900858e-01
-4.76341695e-01 3.85952652e-01 -4.90776807e-01 -6.25300288e-01
2.67626137e-01 -6.14415050e-01 6.57693207e-01 5.56710184e-01
1.60182133e-01 3.43859822e-01 4.89272833e-01 1.02904463e+00
1.31407499e-01 -2.21135259e-01 -6.53616071e-01 -2.09255517e-01
6.20787740e-01 1.00548160e+00 -4.82737720e-01 -3.02816480e-01
-7.55521506e-02 3.16223860e-01 2.91195393e-01 6.80765808e-01
-6.29273236e-01 7.11840838e-02 1.15482485e+00 -9.31984782e-02
1.45255238e-01 -3.75175864e-01 -1.92077067e-02 -8.30159605e-01
-1.19987734e-01 -1.14788640e+00 2.96885550e-01 -5.26617467e-01
-9.34278131e-01 1.13742940e-01 1.01551123e-01 -8.79064143e-01
-5.95276535e-01 -3.64186555e-01 -4.92698342e-01 5.89616537e-01
-1.39431369e+00 -5.56254148e-01 3.72227609e-01 2.99402595e-01
-1.00769997e-02 -1.56386122e-01 5.31901896e-01 -2.82543302e-01
-4.99209255e-01 -3.02069753e-01 2.37710904e-02 -8.52887556e-02
3.40046376e-01 -1.27539194e+00 -2.75654614e-01 5.97016811e-01
-1.28702268e-01 4.56132531e-01 1.05562496e+00 -5.39769173e-01
-1.39925814e+00 -5.71910620e-01 3.43693942e-01 -1.89531893e-01
9.38565671e-01 -1.82892442e-01 -5.24008811e-01 7.40640342e-01
1.96072951e-01 -1.40485585e-01 3.88253778e-01 -6.03226423e-02
-1.69347361e-01 -3.81745517e-01 -1.36353898e+00 1.04233491e+00
7.09521353e-01 -2.58615017e-01 -4.48600203e-01 1.55704230e-01
7.39454269e-01 -2.25198105e-01 -5.04782438e-01 1.65629074e-01
7.23115087e-01 -8.54509413e-01 5.01289666e-01 -1.01399851e+00
1.23379298e-01 -6.37784123e-01 -2.66189694e-01 -1.54221928e+00
-2.15181097e-01 -8.53936553e-01 1.97797678e-02 7.26144433e-01
5.65585554e-01 -1.13663626e+00 7.64819622e-01 1.21475303e+00
2.71569580e-01 -6.39108300e-01 -1.06501484e+00 -8.60752821e-01
8.10597092e-02 -2.60452986e-01 8.42304528e-01 8.18960726e-01
2.88736016e-01 -6.05498590e-02 -3.51618797e-01 3.67731333e-01
6.53380692e-01 1.43477187e-01 8.84017527e-01 -7.96268284e-01
-7.46290684e-01 -4.95167434e-01 1.35776073e-01 -6.02119684e-01
2.03666687e-01 -3.63940746e-01 4.78603393e-01 -1.21842122e+00
-1.10395039e-02 -7.76295781e-01 -4.96204197e-01 5.63550949e-01
1.95775166e-01 -6.32837117e-01 5.74226975e-01 1.37361959e-01
-6.51221693e-01 4.85277653e-01 1.27108467e+00 1.18053388e-02
-5.88502049e-01 2.99554951e-02 -8.14384162e-01 6.70487761e-01
9.92036164e-01 -5.42518318e-01 -3.98659796e-01 -1.91387236e-01
3.39267552e-01 4.12778050e-01 1.02150522e-01 -7.45296001e-01
2.45552406e-01 -1.19068027e+00 -4.10111137e-02 4.26927119e-01
3.01143914e-01 -1.14352655e+00 3.03488433e-01 8.22509944e-01
-6.91581786e-01 -2.28814840e-01 -3.77691031e-01 5.67275345e-01
3.99939358e-01 -2.64400572e-01 8.01030040e-01 -2.64881283e-01
-2.84980297e-01 -1.13250585e-02 -7.63331294e-01 -2.08309308e-01
1.45972085e+00 -3.66650671e-02 -2.49931917e-01 -7.02814341e-01
-7.30303466e-01 7.37521589e-01 7.89577305e-01 2.02094719e-01
2.08609015e-01 -9.54845130e-01 -5.82314849e-01 1.87719166e-02
-1.03556849e-01 -3.34511459e-01 -1.92467541e-01 4.75340426e-01
-1.63844019e-01 1.03368774e-01 -3.10486227e-01 -1.64910138e-01
-8.78591359e-01 5.23556769e-01 8.11733842e-01 -2.84071505e-01
-2.87379444e-01 1.00806504e-01 7.61978254e-02 -6.08327806e-01
1.24188840e-01 -2.23801211e-01 -7.35966787e-02 -3.88264328e-01
3.81096393e-01 1.05446979e-01 -4.60708410e-01 -2.59408534e-01
-2.24936157e-01 2.22527742e-01 1.96354911e-01 -7.55514026e-01
1.39326775e+00 -7.94362742e-03 3.83017585e-02 1.35408625e-01
4.72416759e-01 -8.46778229e-02 -1.78595304e+00 -1.54854313e-01
1.63436264e-01 -6.17944598e-01 -1.71350077e-01 -9.36616838e-01
-6.12179756e-01 2.83027589e-02 1.94506824e-01 7.55822003e-01
1.05957830e+00 -3.17114800e-01 2.55519360e-01 4.94807452e-01
7.99450934e-01 -1.63079238e+00 1.70166180e-01 3.44554514e-01
6.06952310e-01 -8.85237455e-01 -1.31756708e-01 1.68292478e-01
-7.60032475e-01 8.10231805e-01 5.65659106e-01 -4.23917115e-01
4.32451129e-01 3.82797807e-01 4.30732854e-02 1.05926298e-01
-1.17773604e+00 -3.05667609e-01 -3.24958742e-01 7.06961393e-01
-2.80635446e-01 5.19835532e-01 -2.81512618e-01 -4.64812256e-02
-2.32568961e-02 1.10878281e-01 8.23960483e-01 1.07344699e+00
-9.32994246e-01 -1.48690581e+00 -6.92965746e-01 3.01252007e-01
-2.83047813e-03 5.52458167e-01 -4.86874133e-01 7.26511717e-01
2.75152743e-01 1.00888991e+00 -1.65375531e-01 1.43706612e-03
3.60424757e-01 -8.57120305e-02 3.49995106e-01 -6.11400068e-01
-3.97224754e-01 -2.06143525e-03 3.52573574e-01 -6.73193574e-01
-3.15818906e-01 -1.04248261e+00 -1.51319873e+00 -6.50186315e-02
5.38836978e-03 2.85623252e-01 8.08645487e-01 1.12576807e+00
1.50427416e-01 4.20335084e-01 9.17289019e-01 -5.87443829e-01
-9.80894446e-01 -4.65944082e-01 -5.87865055e-01 2.05038697e-01
5.97029328e-01 -4.88924056e-01 -3.77575666e-01 -1.99470311e-01] | [4.307559967041016, 2.0051205158233643] |
b2708e67-87b0-40e4-827b-dfb72f1b0085 | massively-parallel-reweighted-wake-sleep | 2305.11022 | null | https://arxiv.org/abs/2305.11022v1 | https://arxiv.org/pdf/2305.11022v1.pdf | Massively Parallel Reweighted Wake-Sleep | Reweighted wake-sleep (RWS) is a machine learning method for performing Bayesian inference in a very general class of models. RWS draws $K$ samples from an underlying approximate posterior, then uses importance weighting to provide a better estimate of the true posterior. RWS then updates its approximate posterior towards the importance-weighted estimate of the true posterior. However, recent work [Chattergee and Diaconis, 2018] indicates that the number of samples required for effective importance weighting is exponential in the number of latent variables. Attaining such a large number of importance samples is intractable in all but the smallest models. Here, we develop massively parallel RWS, which circumvents this issue by drawing $K$ samples of all $n$ latent variables, and individually reasoning about all $K^n$ possible combinations of samples. While reasoning about $K^n$ combinations might seem intractable, the required computations can be performed in polynomial time by exploiting conditional independencies in the generative model. We show considerable improvements over standard "global" RWS, which draws $K$ samples from the full joint. | ['Laurence Aitchison', 'Gavin Leech', 'Thomas Heap'] | 2023-05-18 | null | null | null | null | ['bayesian-inference'] | ['methodology'] | [-1.14627756e-01 2.86368668e-01 -1.96536943e-01 -5.02273619e-01
-1.32946396e+00 -3.72994810e-01 5.05378544e-01 3.62482294e-02
-4.10860866e-01 9.97238338e-01 1.62184685e-01 -3.35959196e-01
-1.78528965e-01 -8.66888940e-01 -6.30155087e-01 -7.69436717e-01
-2.37213135e-01 7.81826317e-01 4.00862582e-02 3.45578730e-01
2.57996947e-01 1.77805156e-01 -1.36094534e+00 -2.85079896e-01
5.16519070e-01 8.24057221e-01 -2.83207092e-03 1.01971269e+00
-3.15866247e-02 4.89804417e-01 -6.29032195e-01 -4.80709881e-01
7.32451752e-02 -3.03221971e-01 -6.41309202e-01 -3.55263948e-01
1.87787279e-01 -4.80049819e-01 8.48364457e-02 1.08336699e+00
2.83421040e-01 3.69546264e-01 7.27533937e-01 -1.21118736e+00
-2.95862108e-01 8.68545055e-01 -8.77520025e-01 2.47403502e-01
1.62738383e-01 3.71462554e-02 1.19541621e+00 -7.92486548e-01
1.58836767e-01 1.41123521e+00 6.23399973e-01 3.24417531e-01
-1.53228545e+00 -9.21160042e-01 2.42375091e-01 1.16191983e-01
-1.49692810e+00 -5.09387076e-01 4.62340713e-01 -2.09567443e-01
8.35711181e-01 2.04961926e-01 8.88795853e-01 9.07930195e-01
-1.79235712e-01 9.23310518e-01 9.94032860e-01 -4.36932534e-01
9.34603810e-01 -2.54295260e-01 2.50291288e-01 3.54114562e-01
3.98981810e-01 -2.23403588e-01 -8.84705544e-01 -8.26900780e-01
4.81894106e-01 1.48137212e-01 8.34600627e-02 4.89187948e-02
-8.94751370e-01 9.26921487e-01 -1.04251534e-01 -1.83103889e-01
-2.31191888e-01 8.30174625e-01 5.49634062e-02 -4.64286469e-02
6.92726493e-01 6.92249760e-02 -6.75896287e-01 -5.95621049e-01
-1.39415395e+00 5.08174181e-01 7.68968403e-01 7.95369625e-01
8.92198026e-01 -4.08387482e-02 -2.10125342e-01 6.30121350e-01
5.23068130e-01 8.42016220e-01 -6.17358200e-02 -1.43960106e+00
2.86631167e-01 -7.05753267e-02 7.26474524e-01 -3.71310622e-01
1.17720135e-01 -2.11159512e-01 -8.48491251e-01 1.23138189e-01
6.73290908e-01 -3.61590654e-01 -7.80214012e-01 2.10317111e+00
4.65123981e-01 2.97070652e-01 -2.26325005e-01 4.53254998e-01
-1.98707938e-01 7.74130046e-01 1.24287747e-01 -4.27852660e-01
1.36152029e+00 -5.78186452e-01 -6.22354150e-01 -5.37383735e-01
3.92530188e-02 -6.27876341e-01 9.65734363e-01 6.22788310e-01
-1.40818667e+00 -9.78551731e-02 -7.82004118e-01 3.61334458e-02
1.32572979e-01 -1.35507762e-01 1.01570511e+00 7.75573730e-01
-1.17491543e+00 6.15455449e-01 -1.09218919e+00 1.23223484e-01
4.78859901e-01 2.48460114e-01 5.05593829e-02 -2.90577114e-01
-9.91539061e-01 7.56943226e-01 9.24109742e-02 1.81737676e-01
-1.04175949e+00 -7.49739647e-01 -7.81680167e-01 2.72333950e-01
5.08747280e-01 -6.63618386e-01 1.44887149e+00 -2.65170038e-01
-1.38437080e+00 1.59314826e-01 -8.91168714e-01 -5.23748815e-01
4.33758318e-01 -3.42884898e-01 -2.89346725e-02 -6.81136027e-02
1.24982055e-02 3.80009800e-01 1.16905439e+00 -7.10247874e-01
-4.88648415e-01 -4.00780976e-01 1.68387458e-01 -3.95894945e-02
-5.96930236e-02 -8.64949599e-02 -6.00692749e-01 -2.82986403e-01
2.34770089e-01 -7.72025108e-01 -5.44913232e-01 6.74956366e-02
-3.59747618e-01 -4.44394261e-01 1.25667736e-01 -5.17596781e-01
1.08736157e+00 -1.94820309e+00 -6.64713010e-02 4.15165126e-01
4.29767698e-01 -1.27939776e-01 4.19514477e-02 3.42906654e-01
2.96708435e-01 -8.21898282e-02 -4.87252027e-01 -7.25181639e-01
3.40242416e-01 2.54813850e-01 -2.59619981e-01 4.37115908e-01
-2.32633743e-02 7.01683402e-01 -1.16173279e+00 -3.77088934e-01
8.47099796e-02 3.81457716e-01 -6.69794798e-01 1.28528476e-01
-3.50251794e-01 -6.40172064e-02 -2.19235614e-01 4.71214205e-01
9.06849682e-01 -5.63828230e-01 4.39117223e-01 8.64195675e-02
2.86976188e-01 3.71316969e-01 -1.53904247e+00 1.35332525e+00
-5.19611537e-01 4.01965588e-01 2.73801267e-01 -1.02401972e+00
5.27693748e-01 1.34354338e-01 2.12309405e-01 -6.41401932e-02
-1.83001086e-01 7.20004365e-02 -3.44892293e-01 -1.29526436e-01
3.80518496e-01 -5.17310560e-01 -2.41710439e-01 9.11348462e-01
-6.88677952e-02 -3.51099342e-01 3.00508648e-01 3.93520892e-01
1.25093520e+00 4.57489900e-02 4.48291957e-01 -2.46930227e-01
1.17329578e-03 -3.95147622e-01 6.70964181e-01 1.26827765e+00
-1.46321610e-01 2.99672902e-01 7.96643198e-01 -3.69294912e-01
-9.20534253e-01 -1.61071658e+00 -1.04583599e-01 1.07913840e+00
-1.13465197e-01 -6.81935012e-01 -6.53801680e-01 -5.81564963e-01
-1.40736356e-01 1.02160609e+00 -7.12817550e-01 1.58527028e-02
-2.46491641e-01 -1.12726462e+00 3.05464089e-01 5.59373796e-01
1.37443289e-01 -6.79928005e-01 -6.53380811e-01 2.44408801e-01
-2.99990773e-01 -6.69083953e-01 -2.39480495e-01 3.34843457e-01
-8.69199276e-01 -8.49478602e-01 -7.29031205e-01 2.06638709e-01
8.40156734e-01 1.58101991e-02 1.23268807e+00 -1.74978584e-01
-2.43527502e-01 2.07052186e-01 3.76815572e-02 -4.05026555e-01
-2.08252177e-01 -2.93153435e-01 -4.07787599e-02 -3.96344662e-01
6.43800318e-01 -8.20494652e-01 -6.54117286e-01 -2.88774431e-01
-7.91878402e-01 -6.33852854e-02 5.67925394e-01 7.04358101e-01
4.33411658e-01 1.90125912e-01 3.43689919e-01 -8.79285991e-01
5.55265307e-01 -7.10104585e-01 -8.28826249e-01 1.94168061e-01
-4.08036023e-01 6.03606045e-01 5.10324419e-01 -3.94854039e-01
-9.75979984e-01 -2.85641968e-01 -3.42258751e-01 -2.84927517e-01
1.77267745e-01 6.24013364e-01 2.58151978e-01 6.97198451e-01
4.58197057e-01 1.19629085e-01 -2.85840392e-01 -4.89572883e-01
4.58175242e-01 3.98909360e-01 2.76371539e-01 -8.53245974e-01
7.48599708e-01 5.41420043e-01 1.53449506e-01 -5.60088813e-01
-1.15069306e+00 -1.00078829e-01 -1.54464930e-01 1.06484793e-01
2.76296407e-01 -8.46724331e-01 -8.95229638e-01 2.04351410e-01
-9.28071260e-01 -6.06591284e-01 -6.09388769e-01 5.85630417e-01
-3.55061591e-01 3.48342329e-01 -3.78842562e-01 -1.46835017e+00
-3.00643831e-01 -8.96006465e-01 1.13812232e+00 7.11711422e-02
-6.47545516e-01 -7.23214805e-01 1.88205078e-01 2.03405932e-01
4.60438639e-01 -1.44230694e-01 1.00656748e+00 -3.31804723e-01
-4.46701467e-01 -5.14667094e-01 -1.75135240e-01 1.89146325e-01
-5.41596413e-02 7.86591396e-02 -8.42779756e-01 -2.18743011e-01
-1.23253278e-03 -2.65312523e-01 8.72879565e-01 6.63798332e-01
1.30939698e+00 -3.71936291e-01 -2.66049534e-01 7.57621005e-02
1.33071041e+00 -1.93390876e-01 4.81172442e-01 -1.11004271e-01
1.18513092e-01 1.82129338e-01 5.27680814e-01 9.77760851e-01
5.83726346e-01 3.15811574e-01 3.30213994e-01 3.50082576e-01
2.95189768e-01 -8.29402283e-02 2.31614411e-01 4.60003883e-01
-1.49012282e-01 1.34054562e-02 -7.53800869e-01 8.29484761e-01
-1.75530410e+00 -1.03371572e+00 2.02843085e-01 2.53020620e+00
1.20374262e+00 3.11876953e-01 7.10963830e-02 7.23154619e-02
6.46910787e-01 1.43545672e-01 -8.25901866e-01 -3.14819396e-01
4.33806032e-01 7.39898860e-01 4.67326880e-01 8.75346959e-01
-6.76987708e-01 6.60560668e-01 7.65219736e+00 1.22993171e+00
-5.53555310e-01 3.34586143e-01 7.34746456e-01 -8.19817901e-01
-7.10160255e-01 4.22834009e-01 -9.65665579e-01 7.93669820e-01
1.22654903e+00 -7.41784051e-02 7.31505632e-01 9.98474419e-01
6.39101714e-02 -6.29037023e-01 -1.03672147e+00 8.96077812e-01
-2.01026246e-01 -1.19207144e+00 -4.07530785e-01 1.45111933e-01
7.27768064e-01 1.03065357e-01 -1.85513258e-01 4.29853916e-01
1.13274479e+00 -8.99290383e-01 7.68943191e-01 5.11834204e-01
4.91992235e-01 -9.38570082e-01 6.07624829e-01 5.23655593e-01
-1.04677391e+00 2.68613528e-02 -6.36675000e-01 -2.76231557e-01
4.18533742e-01 1.32913446e+00 -6.25949144e-01 -2.00956166e-02
9.27205205e-01 1.86042950e-01 -2.34239116e-01 5.56204379e-01
-4.17713404e-01 9.65095580e-01 -9.35079634e-01 -2.45036274e-01
3.47150229e-02 -2.07239479e-01 2.69945800e-01 1.03334689e+00
6.14060104e-01 5.25001902e-04 -2.70605624e-01 9.02876616e-01
-7.07121417e-02 -5.45539916e-01 -1.45264536e-01 -1.26619264e-01
9.95801330e-01 1.19979000e+00 -7.58186817e-01 -7.38655210e-01
-1.37384430e-01 5.81370234e-01 4.68554080e-01 3.77444148e-01
-9.36525404e-01 -2.63794631e-01 6.72507048e-01 -5.83836287e-02
5.45657873e-01 -4.40516651e-01 -8.09378922e-02 -1.21740317e+00
1.75595284e-03 -5.48003078e-01 3.96414489e-01 -7.83863544e-01
-1.22208571e+00 -5.12980763e-03 4.94693935e-01 -7.58352339e-01
-5.87525368e-01 -1.47254705e-01 -6.59801662e-01 1.07694018e+00
-1.38761532e+00 -7.40937471e-01 1.96129918e-01 3.95843983e-01
4.51815039e-01 3.30052614e-01 8.46300781e-01 -1.91286549e-01
-4.00900424e-01 4.97119606e-01 2.56086916e-01 -4.15413707e-01
6.01083279e-01 -1.36314702e+00 5.56636393e-01 8.42448831e-01
-1.03120789e-01 1.12704504e+00 1.09008360e+00 -5.08390009e-01
-1.45431221e+00 -7.06276655e-01 1.10743737e+00 -2.66455650e-01
6.56027198e-01 -3.88574421e-01 -5.26907265e-01 8.21536064e-01
9.59835723e-02 1.33119985e-01 9.79089022e-01 5.81609786e-01
-4.60464269e-01 -1.09010629e-01 -1.31250894e+00 8.92139018e-01
8.01087737e-01 -4.64553267e-01 -4.27665651e-01 2.67456621e-01
4.07037914e-01 3.22346650e-02 -8.20859373e-01 1.04853101e-01
7.61947930e-01 -9.53759909e-01 1.08205938e+00 -2.79055327e-01
3.21423620e-01 -3.69976610e-01 -3.84562612e-01 -1.02093363e+00
-1.53679892e-01 -6.77280903e-01 -6.87366307e-01 1.02121472e+00
3.12238723e-01 -5.79648316e-01 8.87424529e-01 1.10422909e+00
3.67060304e-01 -7.20546365e-01 -1.18401611e+00 -5.35778463e-01
9.05341730e-02 -1.03127992e+00 7.12017119e-01 5.17718196e-01
-8.17095935e-02 1.63813636e-01 -4.93091255e-01 1.92107886e-01
1.14343810e+00 1.32411614e-01 8.21239412e-01 -1.22937202e+00
-8.96573007e-01 -1.34876758e-01 1.57106489e-01 -1.04140556e+00
4.88217138e-02 -4.97099370e-01 1.10494375e-01 -1.39370489e+00
5.81054270e-01 -4.80021656e-01 -2.28016332e-01 5.68106472e-01
-4.98383552e-01 3.55578840e-01 -1.43048301e-01 -2.73638442e-02
-7.91568995e-01 5.49468935e-01 8.13298345e-01 3.12634818e-02
1.63168952e-01 9.99593139e-02 -9.98530030e-01 1.02181339e+00
6.64955199e-01 -5.78974783e-01 -4.00837988e-01 -2.91380763e-01
7.41102815e-01 3.01352531e-01 3.19617629e-01 -6.98786616e-01
1.45709082e-01 -4.90217566e-01 5.93472898e-01 -9.61384058e-01
7.53484666e-01 -4.19787794e-01 2.52565235e-01 3.07211399e-01
-3.52737069e-01 -3.01152349e-01 -1.68143734e-02 7.49589086e-01
2.52395153e-01 -5.89923024e-01 8.07877779e-01 -2.96541125e-01
2.04619300e-02 2.28139758e-01 -6.28769457e-01 1.10604540e-01
6.28607154e-01 -1.30241886e-02 -3.19406576e-02 -6.17789507e-01
-7.41040111e-01 9.20094848e-02 3.12624842e-01 -2.76843011e-01
3.94011915e-01 -1.12581360e+00 -4.67284203e-01 6.27241656e-02
-2.31727436e-01 4.14211839e-01 4.33256567e-01 7.38243282e-01
-8.21634457e-02 -8.23923014e-03 3.73393595e-01 -3.84272605e-01
-9.23944652e-01 1.64606273e-01 -2.74120092e-01 -5.19882977e-01
-3.17430884e-01 1.08747780e+00 -2.48025060e-01 -2.67963797e-01
1.57460868e-02 -2.00608566e-01 3.98524374e-01 -6.60576373e-02
9.16941166e-01 7.02969193e-01 -3.48438829e-01 1.38661996e-01
-4.76776868e-01 4.04510975e-01 -1.47520348e-01 -7.96432316e-01
1.39696002e+00 -2.76463240e-01 -3.76484156e-01 5.15578985e-01
1.01761758e+00 3.18838686e-01 -1.60743999e+00 -3.92732561e-01
-2.04005718e-01 -5.45085549e-01 -7.72938505e-02 -5.25421560e-01
-7.81689823e-01 1.03036034e+00 7.59023577e-02 2.23927960e-01
9.70485449e-01 2.33273149e-01 5.69562376e-01 4.73567396e-01
5.95116973e-01 -1.13916349e+00 -1.41873389e-01 3.94119501e-01
3.38236183e-01 -1.09556580e+00 3.47838819e-01 8.80688280e-02
-3.06028813e-01 1.00565577e+00 1.36444688e-01 -1.92739233e-01
8.35061789e-01 4.01699424e-01 -5.57653129e-01 -8.03906098e-02
-9.74186718e-01 7.21001402e-02 -1.48445353e-01 4.32708293e-01
2.90616333e-01 1.95008293e-01 -9.16579664e-02 7.13879466e-01
-4.60958034e-01 9.53165442e-02 3.25804979e-01 9.38529372e-01
-5.04335403e-01 -1.27428281e+00 -5.15096545e-01 7.36125648e-01
-5.84929347e-01 -4.13165063e-01 3.97709548e-01 3.56130958e-01
-8.45286623e-02 1.07886314e+00 1.94772378e-01 7.21823275e-02
-4.08485770e-01 1.89520746e-01 5.69710195e-01 -5.12559593e-01
7.88632780e-02 2.01028779e-01 -1.17974967e-01 -4.81544226e-01
-3.03339660e-01 -8.35014999e-01 -1.00870979e+00 -7.34270513e-01
-2.01259628e-01 4.80989635e-01 7.11763442e-01 1.37201631e+00
2.32912749e-01 1.71355024e-01 3.73599708e-01 -1.05556023e+00
-7.99383879e-01 -1.05774808e+00 -6.91370487e-01 -5.65630570e-02
3.53190601e-01 -4.91758972e-01 -6.84873044e-01 -7.60975033e-02] | [7.019908905029297, 4.026686668395996] |
4f2010e2-0357-453e-a8df-71465e32d034 | counterfactual-explanations-for-machine-1 | 2010.10596 | null | https://arxiv.org/abs/2010.10596v3 | https://arxiv.org/pdf/2010.10596v3.pdf | Counterfactual Explanations and Algorithmic Recourses for Machine Learning: A Review | Machine learning plays a role in many deployed decision systems, often in ways that are difficult or impossible to understand by human stakeholders. Explaining, in a human-understandable way, the relationship between the input and output of machine learning models is essential to the development of trustworthy machine learning based systems. A burgeoning body of research seeks to define the goals and methods of explainability in machine learning. In this paper, we seek to review and categorize research on counterfactual explanations, a specific class of explanation that provides a link between what could have happened had input to a model been changed in a particular way. Modern approaches to counterfactual explainability in machine learning draw connections to the established legal doctrine in many countries, making them appealing to fielded systems in high-impact areas such as finance and healthcare. Thus, we design a rubric with desirable properties of counterfactual explanation algorithms and comprehensively evaluate all currently proposed algorithms against that rubric. Our rubric provides easy comparison and comprehension of the advantages and disadvantages of different approaches and serves as an introduction to major research themes in this field. We also identify gaps and discuss promising research directions in the space of counterfactual explainability. | ['Chirag Shah', 'John P. Dickerson', 'Keegan E. Hines', 'Minh Hoang', 'Varich Boonsanong', 'Sahil Verma'] | 2020-10-20 | null | null | null | null | ['counterfactual-explanation'] | ['miscellaneous'] | [ 4.50993001e-01 1.02078152e+00 -7.64527142e-01 -6.28760397e-01
-1.86086088e-01 -5.10930598e-01 9.84579265e-01 9.47894678e-02
3.52082849e-02 1.07932413e+00 2.65747815e-01 -1.14968562e+00
-5.07025421e-01 -5.71503460e-01 -8.11554611e-01 -3.06447268e-01
9.60928351e-02 3.53559762e-01 -6.06048226e-01 -6.10230630e-03
4.63165790e-01 3.79058659e-01 -1.57188499e+00 6.78288460e-01
9.47203696e-01 6.21632457e-01 -4.17332888e-01 4.21798825e-01
-7.53139630e-02 1.11979389e+00 -5.47918260e-01 -8.91560435e-01
3.27230990e-01 -8.02106559e-01 -1.16594231e+00 -2.84224898e-01
2.18901396e-01 -1.81411803e-01 -1.23330504e-02 7.62001932e-01
-1.07714638e-01 -2.42540911e-01 8.87323260e-01 -1.91413534e+00
-8.70495141e-01 9.04758155e-01 -1.74399331e-01 -3.38877253e-02
3.65387022e-01 2.57348925e-01 1.23827648e+00 -3.29776347e-01
3.51743758e-01 1.26234138e+00 5.29013813e-01 9.80460227e-01
-1.20939827e+00 -8.91581118e-01 3.63598257e-01 5.24985075e-01
-7.32905865e-01 -2.51315624e-01 8.27614665e-01 -5.34416437e-01
1.09786010e+00 1.00721800e+00 7.91887045e-01 1.00893986e+00
7.74919748e-01 7.30971813e-01 1.35627174e+00 -7.52762556e-01
4.69575167e-01 4.16129410e-01 9.21095386e-02 4.22056228e-01
8.79890621e-01 6.26743376e-01 -3.94339919e-01 -3.82751077e-01
3.75721365e-01 2.50891834e-01 -2.94234961e-01 -5.35473883e-01
-9.64567661e-01 1.30429232e+00 4.97202396e-01 2.98099399e-01
-4.43682522e-01 3.70567530e-01 2.02769190e-01 3.51455003e-01
3.72960657e-01 9.82582510e-01 -8.24232459e-01 1.70288205e-01
-8.58895361e-01 4.22958642e-01 8.80296052e-01 5.34530699e-01
5.05840302e-01 1.18139073e-01 4.98695262e-02 -1.22567397e-02
5.57843864e-01 4.29022074e-01 5.16548276e-01 -1.07104540e+00
2.80306399e-01 7.10351527e-01 2.58248180e-01 -8.60785306e-01
-2.27996990e-01 -2.42982626e-01 -7.44472325e-01 4.73516792e-01
2.62160659e-01 -7.58009553e-02 -5.24809003e-01 1.40915453e+00
2.35674664e-01 4.60741185e-02 2.33397469e-01 9.06136453e-01
4.68023092e-01 3.14600557e-01 2.69649625e-01 -5.33552408e-01
1.04700577e+00 -6.83943033e-01 -9.53000247e-01 -4.69482452e-01
7.20087051e-01 -5.22796810e-01 1.12298167e+00 2.74652064e-01
-6.57268465e-01 -1.50775239e-01 -1.09164894e+00 2.54394978e-01
-3.73972476e-01 -4.93498117e-01 1.36607802e+00 9.73787069e-01
-4.91351426e-01 6.64002180e-01 -5.23011446e-01 -9.86718610e-02
6.12602055e-01 4.14600790e-01 -1.12558924e-01 1.03129685e-01
-1.32575691e+00 1.29384732e+00 3.80995184e-01 -3.48553061e-02
-4.96519059e-01 -8.94708276e-01 -9.17456567e-01 2.00030461e-01
4.27523464e-01 -1.05074179e+00 1.49441040e+00 -1.46208847e+00
-9.63275194e-01 7.18393147e-01 -1.16177484e-01 -9.60218251e-01
6.79910183e-01 -2.27527916e-01 -5.77281296e-01 -5.25986433e-01
1.17321521e-01 3.58801812e-01 4.70451504e-01 -1.37370372e+00
-7.59959757e-01 -4.19456035e-01 2.41208777e-01 -6.62455782e-02
1.89160541e-01 -1.49072841e-01 6.50054157e-01 -5.11347413e-01
-2.25148261e-01 -7.74945021e-01 -5.00113368e-01 -8.15718621e-02
-5.92485726e-01 -8.65378678e-02 8.92177641e-01 -2.15975448e-01
1.44360530e+00 -1.42380679e+00 -5.09738207e-01 1.16218694e-01
4.43882227e-01 1.06534675e-01 3.64160538e-01 3.88299078e-01
-6.80949092e-01 7.84613907e-01 -2.91129649e-01 3.20963860e-01
2.64883161e-01 3.17485332e-01 -7.35885918e-01 3.56450558e-01
1.60755068e-01 1.09326899e+00 -9.03892577e-01 -2.77796566e-01
4.96744722e-01 -9.51649472e-02 -5.18938720e-01 1.73445687e-01
-2.07362324e-01 9.82789099e-02 -5.80405056e-01 3.37195218e-01
4.03575808e-01 -1.17381014e-01 4.76469576e-01 1.60381585e-01
-6.63816109e-02 5.53865790e-01 -9.61490095e-01 1.00377822e+00
-4.50028598e-01 8.48515272e-01 -6.69825554e-01 -1.01925337e+00
5.62816381e-01 5.05080819e-01 1.73585385e-01 -4.06891912e-01
1.44760698e-01 3.46318662e-01 3.78781348e-01 -7.03323662e-01
8.17302465e-02 -8.86771381e-01 -6.43024817e-02 1.00253487e+00
-4.52845007e-01 -3.54289591e-01 -3.18491668e-01 3.93504947e-02
7.09798813e-01 -1.27259597e-01 1.51242352e+00 -1.97139099e-01
3.04547042e-01 3.26709211e-01 5.05962491e-01 1.04695690e+00
-1.63635239e-01 3.59501392e-01 4.85949039e-01 -1.20562613e+00
-7.93008029e-01 -6.84091926e-01 8.10662135e-02 4.94879395e-01
-2.17591807e-01 -7.28811696e-02 -7.62055874e-01 -1.26445735e+00
3.08211237e-01 1.58996308e+00 -1.13274026e+00 -3.93615395e-01
-1.99190617e-01 -6.18855476e-01 4.05897737e-01 4.34483796e-01
1.86209574e-01 -1.05831313e+00 -1.10473228e+00 -8.82281885e-02
-1.67033553e-01 -4.55392003e-01 1.86470691e-02 1.00627929e-01
-1.10194576e+00 -1.74042201e+00 1.24517620e-01 5.40113337e-02
5.02970517e-01 3.73972952e-01 1.25095809e+00 4.52377379e-01
9.09275487e-02 1.59525022e-01 -1.92769080e-01 -1.32501972e+00
-7.11543024e-01 -3.65037203e-01 1.97441936e-01 -2.87085474e-01
8.35987747e-01 -2.39040047e-01 -4.85988468e-01 1.88488960e-01
-9.58766460e-01 3.34067196e-01 5.57909191e-01 8.60452831e-01
1.80342272e-01 -1.38010278e-01 7.02201605e-01 -1.53254926e+00
7.76151001e-01 -6.61145985e-01 -3.06489795e-01 5.38681448e-01
-1.48891902e+00 1.88368961e-01 6.19257152e-01 -1.62467167e-01
-1.21679592e+00 -2.14433014e-01 2.61856973e-01 -2.63191089e-02
-3.12719315e-01 4.62492794e-01 -3.61141592e-01 3.70912582e-01
9.77090955e-01 -3.10258180e-01 -1.89494025e-02 -1.59439653e-01
8.25225353e-01 7.97443092e-01 1.64430797e-01 -2.26474985e-01
8.12940419e-01 6.22798443e-01 -1.12086356e-01 -7.02345669e-02
-9.35942054e-01 -3.24496590e-02 -5.18383741e-01 -2.66238183e-01
4.31598812e-01 -2.66413152e-01 -7.33788908e-01 -3.49768341e-01
-1.23140466e+00 -2.25870118e-01 -6.97937906e-01 5.83283007e-01
-7.73098409e-01 -2.73553848e-01 2.43033871e-01 -1.09555900e+00
-2.71649033e-01 -7.94814885e-01 4.78135586e-01 9.51422825e-02
-1.06769896e+00 -1.22601271e+00 -9.65617411e-03 6.53419793e-01
2.38333091e-01 5.05699456e-01 1.21076179e+00 -1.12929654e+00
-3.79624367e-01 -2.27142230e-01 1.15025148e-01 6.00139387e-02
4.02093142e-01 -1.34735731e-02 -1.13978457e+00 1.68515116e-01
3.08243185e-01 1.47859678e-01 5.32902956e-01 6.36375487e-01
1.05061424e+00 -1.15378976e+00 -5.79945982e-01 -6.36371523e-02
1.25924170e+00 5.37023902e-01 4.69025910e-01 5.10409772e-01
2.96869457e-01 1.06442225e+00 9.67896044e-01 2.87577063e-01
2.78169870e-01 4.61364865e-01 6.16151094e-01 -2.76419967e-01
1.94960132e-01 -4.38279748e-01 6.63466901e-02 -5.58064319e-02
-1.76041394e-01 -4.34045680e-02 -9.30451632e-01 4.64911491e-01
-2.11412883e+00 -1.36577916e+00 -5.27548194e-01 2.01753521e+00
5.05105674e-01 8.27763751e-02 -1.88133806e-01 4.03381228e-01
5.40674865e-01 5.59626855e-02 -5.69148839e-01 -1.13858426e+00
1.55139431e-01 -7.53222480e-02 3.01572412e-01 5.27583301e-01
-8.10921848e-01 7.18456507e-01 7.38849306e+00 4.70912606e-01
-9.29446757e-01 1.27893701e-01 8.60916615e-01 -2.25922167e-01
-1.01702833e+00 2.98715413e-01 5.01120165e-02 2.25590408e-01
1.05422783e+00 -7.23125458e-01 1.54022336e-01 1.10852730e+00
8.83120954e-01 6.06808290e-02 -1.66156411e+00 5.35900414e-01
-3.56395952e-02 -1.86385143e+00 4.17401403e-01 4.68763933e-02
9.53620493e-01 -4.68544722e-01 1.02028869e-01 6.02851585e-02
4.71878499e-01 -1.43991721e+00 9.60397422e-01 2.80151188e-01
5.20514548e-01 -6.37942731e-01 1.01200390e+00 4.83517706e-01
-4.51354414e-01 -4.58871424e-01 -1.22131132e-01 -9.38236594e-01
-5.46760038e-02 5.87831616e-01 -1.14302278e+00 6.42102897e-01
4.38400269e-01 4.02025014e-01 -3.61477807e-02 6.86276972e-01
-7.95486867e-01 7.38487899e-01 3.52046400e-01 -9.59863365e-02
5.13449609e-02 8.59695077e-02 3.68721664e-01 9.99751449e-01
8.95883814e-02 2.87199169e-01 -5.61858773e-01 1.01441050e+00
1.90527998e-02 -5.07792868e-02 -1.21174514e+00 8.86602253e-02
5.03959358e-01 6.01883113e-01 -3.52407843e-01 -3.00713062e-01
-4.61299151e-01 5.75616181e-01 -7.53022954e-02 3.08683515e-01
-8.89206469e-01 1.47511527e-01 8.63721430e-01 2.71671116e-01
-4.77525085e-01 5.68885982e-01 -9.88605499e-01 -9.60165083e-01
-1.24832429e-01 -1.30006981e+00 7.12081492e-01 -8.10793161e-01
-1.03758252e+00 1.81466848e-01 1.77792400e-01 -1.10466599e+00
-6.64159715e-01 -5.15127301e-01 -8.58574271e-01 6.92819059e-01
-1.35084939e+00 -1.25235558e+00 6.68543279e-02 9.25758407e-02
5.89502096e-01 -1.47167921e-01 1.02572536e+00 -4.32254553e-01
-3.21266294e-01 2.41007954e-01 -1.69394195e-01 -3.54626447e-01
4.07365769e-01 -1.31730962e+00 5.03986776e-01 7.99625814e-01
4.58797365e-01 1.01571786e+00 1.12766874e+00 -6.80396795e-01
-9.66650307e-01 -1.04056334e+00 1.38841414e+00 -7.44706392e-01
4.27069455e-01 5.38589992e-02 -6.40020609e-01 1.15289354e+00
2.67726064e-01 -5.05631566e-01 1.19667315e+00 4.92576331e-01
-3.49593937e-01 -5.95434681e-02 -1.35110188e+00 6.78507984e-01
7.27753103e-01 -3.72853786e-01 -1.24892938e+00 2.35519186e-01
6.32131577e-01 7.46854767e-02 -2.64109105e-01 1.38138756e-01
9.07598794e-01 -1.18155503e+00 6.36920631e-01 -1.61203146e+00
6.50469005e-01 -2.31267959e-01 -2.19470765e-02 -1.36240113e+00
-1.92093015e-01 -7.89357245e-01 -2.05505207e-01 8.70365381e-01
7.48944640e-01 -9.81071591e-01 7.40324914e-01 1.36011672e+00
1.40262321e-01 -8.78179729e-01 -8.70661438e-01 -6.52029276e-01
2.39400953e-01 -8.63037884e-01 1.08754468e+00 1.34818995e+00
5.28908730e-01 1.21110216e-01 -4.15593296e-01 4.41363454e-02
4.99924481e-01 5.24395406e-01 8.57896507e-01 -1.31666791e+00
2.28687078e-02 -5.45108318e-01 -3.09508413e-01 -3.79571855e-01
2.85187960e-01 -7.94685423e-01 -3.57605100e-01 -1.58181655e+00
6.54884219e-01 -2.01875478e-01 -3.22626442e-01 6.97224498e-01
-3.62074792e-01 -1.47353217e-01 2.99126863e-01 2.47649133e-01
-1.83260858e-01 2.68671364e-01 8.87468636e-01 -2.54028171e-01
-4.31093089e-02 2.10316852e-01 -1.50424004e+00 1.09504855e+00
9.09311593e-01 -9.84714448e-01 -6.25642002e-01 -1.81979418e-01
2.53317773e-01 -2.19186664e-01 6.08241141e-01 -3.07149827e-01
-9.16557536e-02 -1.01624978e+00 1.39571816e-01 1.34279104e-02
-2.73380309e-01 -1.26921904e+00 6.80830359e-01 1.01009786e+00
-6.77613556e-01 2.38365233e-02 8.90174322e-03 5.53593755e-01
-1.05107524e-01 -3.44738275e-01 4.97072875e-01 -9.22052562e-02
-5.32261252e-01 -1.37481391e-01 -4.65242773e-01 -1.46941513e-01
1.40686965e+00 -4.08673763e-01 -3.26571763e-01 -5.79449296e-01
-2.65142143e-01 2.43694872e-01 3.64691585e-01 6.56460762e-01
5.86728454e-01 -1.22207308e+00 -5.81966341e-01 -1.63369581e-01
8.31706971e-02 -5.82199752e-01 1.23800926e-01 5.24271548e-01
-2.28154719e-01 1.00654387e+00 -2.21264556e-01 9.39994082e-02
-1.25573039e+00 6.43946588e-01 5.53007841e-01 -3.81735802e-01
-2.48783886e-01 3.75089139e-01 4.00299430e-01 -4.23789978e-01
-2.18223393e-01 -3.39757562e-01 -2.47705251e-01 -3.40235084e-01
6.66527748e-01 5.43155253e-01 -1.46625042e-01 -2.21134290e-01
-5.22208452e-01 -1.60156623e-01 6.16339855e-02 2.47650035e-02
1.34390557e+00 -1.41047137e-02 5.93850985e-02 5.82797647e-01
6.81073070e-01 -2.74541527e-01 -8.34453940e-01 3.36269021e-01
2.80552566e-01 -6.73804343e-01 -1.16362236e-01 -1.22606480e+00
-6.66060150e-01 7.77454019e-01 3.68540257e-01 5.04226267e-01
9.16784406e-01 5.55402748e-02 1.15220495e-01 2.10602283e-01
2.81861991e-01 -8.85133564e-01 -4.04799998e-01 -2.49004498e-01
1.29688179e+00 -1.39017951e+00 3.37849796e-01 -4.22875017e-01
-8.11045945e-01 9.20534253e-01 3.71149749e-01 3.24469179e-01
4.15285200e-01 2.21845675e-02 2.59001523e-01 -3.62036556e-01
-8.58055055e-01 2.12977156e-01 4.88113254e-01 6.94421470e-01
7.23919511e-01 5.85552931e-01 -7.35317588e-01 1.10311449e+00
-4.56400275e-01 1.11667082e-01 5.72063923e-01 5.63957036e-01
-1.85051218e-01 -1.12857652e+00 -4.93120909e-01 7.73759425e-01
-5.91741443e-01 -8.50855187e-02 -9.34860229e-01 1.50092411e+00
2.53502429e-01 1.41839170e+00 -1.68946609e-01 -2.69723892e-01
3.77640307e-01 1.17984802e-01 5.87996803e-02 -5.87647200e-01
-6.16133630e-01 -6.05443239e-01 3.18533063e-01 -7.22007930e-01
-4.87149805e-01 -7.45480359e-01 -1.33456612e+00 -4.45038348e-01
-4.33407038e-01 5.75134277e-01 6.16776228e-01 1.40543818e+00
1.15626052e-01 3.71809095e-01 5.13549149e-01 -2.07080722e-01
-6.16628647e-01 -7.28830099e-01 -3.71957660e-01 4.43035245e-01
3.23458940e-01 -5.69392562e-01 -4.30151612e-01 1.53566197e-01] | [8.775961875915527, 5.742010593414307] |
f9c320c0-8ff6-452b-a1a1-8ca25aa72090 | abstractive-approaches-to-multidocument | null | null | https://aclanthology.org/2022.sdp-1.24 | https://aclanthology.org/2022.sdp-1.24.pdf | Abstractive Approaches To Multidocument Summarization Of Medical Literature Reviews | Text summarization has been a trending domain of research in NLP in the past few decades. The medical domain is no exception to the same. Medical documents often contain a lot of jargon pertaining to certain domains, and performing an abstractive summarization on the same remains a challenge. This paper presents a summary of the findings that we obtained based on the shared task of Multidocument Summarization for Literature Review (MSLR). We stood fourth in the leaderboards for evaluation on the MSˆ2 and Cochrane datasets. We finetuned pre-trained models such as BART-large, DistilBART and T5-base on both these datasets. These models’ accuracy was later tested with a part of the same dataset using ROUGE scores as the evaluation metrics. | ['Dipali Dattatray Kadam', 'Onkar Rupesh Litake', 'Aditya Vyankatesh Mandke', 'Aditya Jagdish Vyawahare', 'Rahul Tangsali'] | null | null | null | null | sdp-coling-2022-10 | ['abstractive-text-summarization'] | ['natural-language-processing'] | [ 2.14016080e-01 3.47338021e-01 -4.16144371e-01 -2.24138498e-01
-1.47953844e+00 -6.60904765e-01 6.80099607e-01 8.52599800e-01
-4.78300154e-01 1.25472856e+00 1.03259766e+00 -4.19830412e-01
-3.83869678e-01 -9.84624773e-02 -4.32108074e-01 -3.98569614e-01
2.02439025e-01 7.23216653e-01 -3.00159659e-02 -2.04536766e-01
8.93436551e-01 6.83746338e-02 -1.00883365e+00 7.33401179e-01
1.22506940e+00 3.98349822e-01 1.35749429e-02 1.12532997e+00
-2.08224341e-01 7.44422793e-01 -1.04850888e+00 -5.76410651e-01
-1.95985049e-01 -5.25383472e-01 -1.03267646e+00 -1.77864596e-01
5.48299670e-01 1.50808498e-01 -4.14933860e-02 6.86839819e-01
1.07406461e+00 4.69355881e-02 8.62225533e-01 -6.44784570e-01
-4.50004876e-01 9.68383431e-01 -6.62450612e-01 4.39320326e-01
8.02091420e-01 -7.01835006e-02 1.12993526e+00 -5.00498474e-01
1.05334425e+00 1.34628606e+00 7.12935746e-01 4.00433302e-01
-1.01801801e+00 -3.94380808e-01 -2.44866446e-01 1.11101113e-01
-8.83099258e-01 -5.04027367e-01 4.37778085e-01 -4.11378473e-01
1.33010828e+00 5.27588665e-01 4.86443281e-01 1.15935254e+00
7.42104352e-01 7.23755479e-01 7.98890948e-01 -4.11706060e-01
2.86758393e-01 8.33193585e-02 5.57834268e-01 1.21441804e-01
8.33086491e-01 -6.23932421e-01 -5.72563350e-01 -5.29987454e-01
-8.03314000e-02 -4.28190470e-01 -2.95608193e-01 4.65566635e-01
-1.28667200e+00 9.87434983e-01 -1.68632597e-01 4.46582019e-01
-5.68135262e-01 -1.33681253e-01 1.02100980e+00 1.73676953e-01
8.14509571e-01 9.78279769e-01 -5.09074211e-01 -4.28289592e-01
-1.51321614e+00 6.63453221e-01 1.34001517e+00 6.90470278e-01
-5.46616502e-02 -4.81876373e-01 -6.83024287e-01 9.47637558e-01
-1.13926195e-01 2.90103316e-01 8.84411633e-01 -8.03269506e-01
8.46836269e-01 6.46598816e-01 -6.89099124e-03 -9.14646506e-01
-6.92900777e-01 -4.69240606e-01 -9.27456677e-01 -5.18197834e-01
-3.81653197e-02 -4.46766615e-01 -8.64777684e-01 1.13061070e+00
1.33130267e-01 -1.93848774e-01 4.13311183e-01 4.04913604e-01
1.55180669e+00 8.53911400e-01 6.23597181e-04 -6.06271863e-01
1.33067739e+00 -1.00284171e+00 -1.17805350e+00 -8.54438618e-02
9.74410892e-01 -1.05995190e+00 6.12494588e-01 6.70343578e-01
-1.37147260e+00 -1.42600685e-01 -1.06168473e+00 -3.91074330e-01
-3.15280586e-01 1.48778707e-01 3.94491702e-01 4.21441138e-01
-1.06824780e+00 8.18423271e-01 -5.22397459e-01 -6.09924257e-01
4.12552029e-01 6.96340203e-03 -3.92619789e-01 2.22894475e-02
-9.90213215e-01 1.33443546e+00 6.43960357e-01 -2.26472944e-01
-3.88163745e-01 -8.58108699e-01 -5.32955348e-01 3.68781760e-02
2.43485600e-01 -9.80372787e-01 1.45460844e+00 -1.18975662e-01
-1.24655437e+00 9.72682953e-01 -5.20310402e-02 -7.09520519e-01
5.67079663e-01 -4.69023883e-01 -4.36713547e-01 2.82817394e-01
3.64331663e-01 2.18322843e-01 3.07393223e-01 -8.21866989e-01
-6.36162400e-01 -2.81704277e-01 -4.34954643e-01 3.60109150e-01
1.50611326e-01 2.57703811e-01 -7.40915909e-02 -7.35072136e-01
-3.10962141e-01 -5.27470708e-01 -2.50485837e-01 -1.10999763e+00
-9.60210323e-01 -5.58488727e-01 5.12674928e-01 -9.73121166e-01
1.84592092e+00 -1.50749588e+00 2.87445039e-01 -2.39885181e-01
9.39839706e-02 3.03533614e-01 -8.75012726e-02 1.01450813e+00
-1.76959887e-01 5.29922605e-01 -3.07962596e-01 -3.77921343e-01
-2.27048159e-01 -3.99032719e-02 -3.76381576e-01 2.86189377e-01
5.27966907e-03 9.92353261e-01 -1.03165841e+00 -7.81366944e-01
-3.23937267e-01 -7.16051087e-02 -3.64091784e-01 -3.97797897e-02
-3.46051037e-01 2.34283805e-01 -5.65581620e-01 3.36967707e-01
3.72195154e-01 -2.52378315e-01 8.21916014e-02 -7.95374364e-02
-2.05355316e-01 5.97752750e-01 -7.08981872e-01 1.96090055e+00
-4.16133925e-02 5.78536391e-01 -2.63089418e-01 -1.02112603e+00
6.48750663e-01 6.43795013e-01 4.66745079e-01 -4.37599719e-01
1.00472853e-01 4.84837443e-01 1.36989757e-01 -7.88194299e-01
9.75912690e-01 2.25792490e-02 -3.50399256e-01 5.96151710e-01
5.60612567e-02 -6.25436068e-01 8.11263084e-01 5.61364472e-01
1.36345255e+00 -2.60420650e-01 9.79825854e-01 -3.75105560e-01
5.25412738e-01 4.63358939e-01 1.83734879e-01 9.81869519e-01
1.66041374e-01 9.10805821e-01 7.90404916e-01 -1.88611925e-01
-9.88945961e-01 -5.31697154e-01 -2.52040237e-01 5.86436868e-01
-5.60886443e-01 -8.00557256e-01 -6.84032917e-01 -7.74522185e-01
8.29307456e-03 1.32914662e+00 -6.24460697e-01 -2.03240067e-02
-4.81347889e-01 -9.88859177e-01 7.40298450e-01 -1.09259799e-01
8.09758902e-02 -1.16621518e+00 -7.27081895e-01 3.28724384e-01
-2.12460846e-01 -9.74430382e-01 -3.41645390e-01 2.07364947e-01
-1.06605279e+00 -1.20174623e+00 -9.43623364e-01 -5.89997172e-01
9.58361030e-02 -4.47266623e-02 1.38816714e+00 -3.24935198e-01
-3.00252587e-01 3.12259227e-01 -5.72285056e-01 -1.03516841e+00
-8.99725556e-01 5.77079296e-01 -2.46491715e-01 -8.05475891e-01
2.45163396e-01 -3.43309015e-01 -5.11229396e-01 -4.57885265e-01
-1.11223435e+00 7.04452619e-02 7.36766458e-01 7.77433932e-01
3.71333659e-01 -2.82805264e-01 1.15214241e+00 -1.29301214e+00
1.54515302e+00 -6.86117947e-01 5.03103249e-02 3.87862921e-01
-5.42658448e-01 7.20058233e-02 3.62298161e-01 -1.54250339e-01
-8.66995454e-01 -6.16100192e-01 -1.46961242e-01 2.29901791e-01
-9.54418704e-02 1.22578371e+00 3.83159190e-01 6.66274846e-01
8.60879898e-01 -5.21254353e-02 1.00119207e-02 -5.92820287e-01
4.58832443e-01 9.19838428e-01 4.18495744e-01 -2.14601085e-01
1.30535915e-01 5.17446175e-02 -1.41995788e-01 -1.07258677e+00
-1.14928460e+00 -6.54628456e-01 -1.33874267e-01 2.18737289e-01
7.78438985e-01 -6.82313502e-01 -2.51927078e-01 1.98710654e-02
-1.57271993e+00 6.83441982e-02 -5.09276271e-01 5.43187439e-01
-2.54667938e-01 6.83029115e-01 -3.39919716e-01 -4.72153723e-01
-1.24269927e+00 -9.38567162e-01 1.04244995e+00 2.28567988e-01
-9.17275906e-01 -9.58497226e-01 7.72683859e-01 4.14168060e-01
2.24529713e-01 4.33811367e-01 1.09182560e+00 -1.33037305e+00
3.50240648e-01 -5.34961820e-01 5.80862276e-02 2.52923667e-01
2.10988194e-01 5.09339459e-02 -7.15259135e-01 -8.79608318e-02
1.96922585e-01 -4.02053595e-01 1.24259412e+00 9.08875823e-01
1.02035165e+00 -5.91887712e-01 -4.60648388e-01 4.99289110e-02
1.02106380e+00 8.43595043e-02 5.93989551e-01 3.93428266e-01
3.69575083e-01 5.67789614e-01 6.02111459e-01 4.21369702e-01
3.92127126e-01 2.37278745e-01 -1.76802829e-01 1.42688781e-01
-1.96310118e-01 3.57642397e-02 1.82641849e-01 1.36500549e+00
5.21228500e-02 -4.68336165e-01 -9.84916091e-01 6.61686957e-01
-1.93972707e+00 -9.07747447e-01 -1.31174996e-01 1.75889945e+00
1.31130636e+00 9.04561207e-02 -5.62159643e-02 -9.79308560e-02
4.56965595e-01 2.44857401e-01 -3.53199184e-01 -1.03432274e+00
-2.80090481e-01 2.90870219e-01 3.21913272e-01 2.22388595e-01
-9.47376430e-01 5.74502647e-01 6.78308487e+00 1.08863521e+00
-8.49903703e-01 -9.21567436e-03 6.69521034e-01 -2.81338960e-01
-2.57067919e-01 -5.40090203e-02 -8.80679727e-01 4.81975079e-01
1.47246647e+00 -8.23823810e-01 -2.60908663e-01 4.44770247e-01
5.30555069e-01 -6.95330083e-01 -1.19908905e+00 6.21921182e-01
3.96109343e-01 -1.63410687e+00 4.48827922e-01 -2.51528826e-02
1.16229355e+00 2.38633066e-01 -3.47416013e-01 3.00467342e-01
2.39797637e-01 -1.18887460e+00 1.96291596e-01 5.97436726e-01
4.98695761e-01 -6.08373761e-01 1.13080633e+00 5.11476219e-01
-2.90663570e-01 2.42312521e-01 -4.10627991e-01 2.59160519e-01
3.10242355e-01 8.03312480e-01 -1.23264241e+00 1.15569150e+00
3.69014382e-01 9.59540188e-01 -7.32151151e-01 1.32475722e+00
-5.22920899e-02 6.28062606e-01 2.29195282e-02 -2.48978898e-01
2.60225624e-01 -1.45926431e-01 1.02550948e+00 1.64067709e+00
3.96369636e-01 -6.52472973e-02 -1.40920475e-01 3.25578272e-01
-3.87330681e-01 4.44888681e-01 -6.45898759e-01 -3.82271588e-01
2.80801237e-01 1.25281453e+00 -5.66355169e-01 -7.19671726e-01
-1.45672396e-01 5.08217752e-01 5.16933613e-02 1.59890562e-01
-4.94648844e-01 -5.53384900e-01 -7.13254586e-02 -2.88511813e-01
9.96685475e-02 3.20676744e-01 -5.43008685e-01 -1.02857184e+00
-2.32574031e-01 -1.18500304e+00 6.78067744e-01 -6.49849594e-01
-1.22856998e+00 5.69778919e-01 2.65829563e-01 -1.09315479e+00
-5.36107421e-01 -1.46381810e-01 -7.25780785e-01 7.19902992e-01
-1.40079784e+00 -6.39130712e-01 1.73405334e-01 -5.73000908e-02
8.06740463e-01 -2.68035561e-01 1.05475688e+00 5.75443879e-02
-6.25874996e-01 3.39000136e-01 5.39674401e-01 -3.06202859e-01
1.13413727e+00 -1.34853804e+00 2.55231947e-01 4.82483387e-01
-8.91972706e-02 7.68710136e-01 1.14297819e+00 -1.02471614e+00
-1.24434304e+00 -8.24558914e-01 1.28986943e+00 -5.10694027e-01
6.87155485e-01 2.88568825e-01 -9.17898774e-01 4.77796942e-01
8.47407162e-01 -1.06498301e+00 8.33209336e-01 5.51173203e-02
2.41146833e-01 2.16978595e-01 -9.49171841e-01 4.30688769e-01
4.13277715e-01 -6.49576932e-02 -1.28921854e+00 7.16016412e-01
8.13526750e-01 -6.50043488e-01 -9.80336785e-01 2.24855095e-01
4.49129194e-01 -4.91817474e-01 7.30193853e-01 -8.04041207e-01
1.07315087e+00 2.12966517e-01 2.85672605e-01 -1.85009348e+00
1.96049213e-02 -7.46291518e-01 -9.28857327e-02 1.25942957e+00
6.08427525e-01 -4.56049740e-01 4.96706098e-01 2.94213235e-01
-5.31890512e-01 -7.96528697e-01 -8.50602627e-01 -4.79801983e-01
4.86849159e-01 -1.24390975e-01 2.24678606e-01 6.75019443e-01
4.39105332e-01 9.47769463e-01 4.84765247e-02 -5.71182966e-01
3.54658425e-01 1.71533436e-01 7.98576236e-01 -1.34839392e+00
1.27495453e-01 -8.22414756e-01 1.02251552e-01 -3.73933375e-01
3.60424742e-02 -1.21231806e+00 -1.30714029e-01 -2.41266036e+00
6.67826653e-01 3.17301691e-01 -1.24877375e-02 2.00253412e-01
-4.09141719e-01 -3.13989818e-01 -9.68268141e-02 2.24625394e-01
-7.70913482e-01 3.20258588e-01 1.23379505e+00 -2.51204818e-01
-4.19570208e-01 3.71349975e-02 -1.03538585e+00 6.26754403e-01
7.89079487e-01 -5.25870919e-01 -2.27539182e-01 -1.49710745e-01
4.77781534e-01 2.36816213e-01 -2.51597345e-01 -6.07051969e-01
4.90490258e-01 9.23318118e-02 1.13440976e-01 -1.16968381e+00
-1.03369452e-01 -1.80555433e-01 -6.44431263e-02 4.16387945e-01
-7.77834117e-01 3.01310807e-01 5.46856821e-01 3.85902971e-01
-2.30688304e-01 -4.89948452e-01 3.20674181e-01 -3.30470234e-01
-2.07866460e-01 -3.38779718e-01 -3.26160371e-01 6.49774909e-01
5.86607575e-01 -3.99085768e-02 -7.17637479e-01 -3.09789598e-01
-2.62596667e-01 6.14932656e-01 -1.00336755e-02 2.35247999e-01
6.04981422e-01 -5.95954716e-01 -1.42653298e+00 -5.59407830e-01
-1.26328943e-02 1.46058366e-01 3.75801861e-01 1.09370553e+00
-6.12428188e-01 1.06826830e+00 -3.21478918e-02 -3.25950950e-01
-1.39949477e+00 2.79996634e-01 -1.22338913e-01 -1.11760259e+00
-7.05068111e-01 5.08806348e-01 -2.23401979e-01 -3.45233440e-01
1.20978855e-01 -5.24657190e-01 -7.86718249e-01 5.95171213e-01
6.58487678e-01 7.99765587e-01 4.14300352e-01 -2.48202294e-01
-2.78011709e-01 2.19844416e-01 -5.19625068e-01 -2.12388843e-01
1.74793315e+00 3.42376456e-02 -4.74882841e-01 5.11900365e-01
1.17240870e+00 9.37696248e-02 -9.88542810e-02 1.99079327e-02
5.17429173e-01 2.46424109e-01 1.30376950e-01 -1.31282568e+00
-3.56475532e-01 4.96382713e-01 -2.14098975e-01 3.83696377e-01
7.01040745e-01 2.73639951e-02 5.91367424e-01 5.17009318e-01
-3.20602894e-01 -1.29188621e+00 -1.41487390e-01 5.28000832e-01
1.45733500e+00 -1.03351593e+00 8.98895919e-01 9.83375087e-02
-9.64585006e-01 1.11347616e+00 -5.54746911e-02 -9.82819796e-02
2.34082371e-01 -6.69116378e-02 -1.82664037e-01 -5.79727292e-01
-1.05675662e+00 3.50319773e-01 6.56897366e-01 4.91794236e-02
7.19537497e-01 3.12095843e-02 -1.21801221e+00 7.56992698e-01
-4.33494240e-01 1.84825987e-01 8.83065879e-01 9.56995010e-01
-3.32161605e-01 -9.19513881e-01 -3.46153587e-01 1.11222100e+00
-1.04778767e+00 -2.69630760e-01 -6.35947466e-01 8.20202529e-01
-5.62850654e-01 1.13502908e+00 -3.07546556e-01 -6.88568577e-02
4.31062877e-01 1.76474884e-01 2.75217891e-01 -1.01685309e+00
-1.03756237e+00 1.65440917e-01 5.98550916e-01 -2.48067945e-01
-4.63211209e-01 -7.87100315e-01 -1.04959929e+00 -2.28553802e-01
-2.19222695e-01 7.33147025e-01 7.73564219e-01 9.13823485e-01
5.12091994e-01 8.39459538e-01 2.05718488e-01 -6.96099758e-01
-8.93313587e-01 -1.62843859e+00 -5.53789735e-01 1.49953485e-01
3.58053535e-01 -9.75379497e-02 -3.53725582e-01 -4.08251472e-02] | [12.376495361328125, 9.568678855895996] |
26242d73-1491-4d13-95bc-a58d86bd8070 | pefat-boosting-semi-supervised-medical-image | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Zeng_PEFAT_Boosting_Semi-Supervised_Medical_Image_Classification_via_Pseudo-Loss_Estimation_and_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Zeng_PEFAT_Boosting_Semi-Supervised_Medical_Image_Classification_via_Pseudo-Loss_Estimation_and_CVPR_2023_paper.pdf | PEFAT: Boosting Semi-Supervised Medical Image Classification via Pseudo-Loss Estimation and Feature Adversarial Training | Pseudo-labeling approaches have been proven beneficial for semi-supervised learning (SSL) schemes in computer vision and medical imaging. Most works are dedicated to finding samples with high-confidence pseudo-labels from the perspective of model predicted probability. Whereas this way may lead to the inclusion of incorrectly pseudo-labeled data if the threshold is not carefully adjusted. In addition, low-confidence probability samples are frequently disregarded and not employed to their full potential. In this paper, we propose a novel Pseudo-loss Estimation and Feature Adversarial Training semi-supervised framework, termed as PEFAT, to boost the performance of multi-class and multi-label medical image classification from the point of loss distribution modeling and adversarial training. Specifically, we develop a trustworthy data selection scheme to split a high-quality pseudo-labeled set, inspired by the dividable pseudo-loss assumption that clean data tend to show lower loss while noise data is the opposite. Instead of directly discarding these samples with low-quality pseudo-labels, we present a novel regularization approach to learn discriminate information from them via injecting adversarial noises at the feature-level to smooth the decision boundary. Experimental results on three medical and two natural image benchmarks validate that our PEFAT can achieve a promising performance and surpass other state-of-the-art methods. The code is available at https://github.com/maxwell0027/PEFAT. | ['Yong Xia', 'Zilin Lu', 'Yutong Xie', 'Qingjie Zeng'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['semi-supervised-medical-image-classification'] | ['medical'] | [ 4.91007119e-01 3.54446352e-01 -2.58354247e-01 -6.64128840e-01
-1.26368201e+00 -2.21654117e-01 2.80305892e-01 7.68908411e-02
-3.49321008e-01 9.11137819e-01 -2.21870974e-01 1.61698777e-02
-5.76510951e-02 -5.95604658e-01 -7.89600134e-01 -1.13222456e+00
3.07464868e-01 2.44353935e-01 5.06230779e-02 3.87997955e-01
-2.34164551e-01 2.60092854e-01 -1.32837486e+00 4.02578205e-01
1.06857729e+00 1.09587884e+00 -2.19166175e-01 1.26747206e-01
9.91726443e-02 8.01032007e-01 -3.41202199e-01 -5.90234518e-01
4.87121046e-01 -5.50818264e-01 -6.26044214e-01 2.87646413e-01
4.00552630e-01 -1.75949216e-01 -5.96817471e-02 1.49520350e+00
5.40108323e-01 -2.65314430e-01 7.49448478e-01 -1.29326952e+00
-5.53618491e-01 4.59522247e-01 -5.89438915e-01 -2.43208230e-01
-2.81968038e-04 1.18177399e-01 7.74137795e-01 -9.04739141e-01
5.64523160e-01 1.01274323e+00 7.56729543e-01 7.68589318e-01
-1.42947471e+00 -8.86570156e-01 -9.11604464e-02 1.52069092e-01
-1.58938599e+00 -4.60085452e-01 1.04714608e+00 -3.94938350e-01
-1.84161484e-01 3.31598014e-01 1.70980006e-01 1.23562109e+00
2.30597094e-01 8.77150595e-01 1.65060616e+00 -3.86150986e-01
2.06813917e-01 5.81678987e-01 -1.43360682e-02 8.68167639e-01
2.15917617e-01 2.84949034e-01 -2.35018656e-01 -4.36107993e-01
3.66844237e-01 8.66351575e-02 -4.79936182e-01 -6.07057452e-01
-9.73366022e-01 8.24054956e-01 4.23557937e-01 1.65736988e-01
-3.41437638e-01 -1.86616763e-01 4.34920400e-01 2.02945128e-01
6.97804570e-01 1.11713901e-01 -3.09217334e-01 4.65927660e-01
-9.52730596e-01 -6.80261180e-02 5.33873022e-01 7.38269448e-01
6.37635350e-01 -1.60248190e-01 -4.30385232e-01 9.62625265e-01
2.11175025e-01 4.53185081e-01 3.77491176e-01 -8.79021704e-01
5.23461923e-02 4.47737813e-01 6.68316754e-03 -8.48688364e-01
-1.38126075e-01 -7.43916392e-01 -1.04898250e+00 5.03691673e-01
5.66999733e-01 8.46526623e-02 -9.58840787e-01 1.63717401e+00
5.97665370e-01 4.26258981e-01 -7.55545720e-02 8.94753456e-01
6.75444186e-01 2.99158096e-01 1.25802115e-01 -5.17668068e-01
1.15257728e+00 -8.03115845e-01 -7.28961170e-01 -5.37502859e-03
4.62315023e-01 -6.03692293e-01 1.13028932e+00 5.68275154e-01
-8.08370888e-01 -5.57590187e-01 -1.01718473e+00 3.74559522e-01
-1.49580641e-02 1.41381592e-01 8.83055031e-02 8.30632925e-01
-4.53453660e-01 9.08168912e-01 -9.15570796e-01 2.47990236e-01
1.00520873e+00 9.98350903e-02 -4.20833647e-01 -3.66983622e-01
-1.11618555e+00 6.31017506e-01 3.35785329e-01 -1.45667456e-02
-8.95331681e-01 -9.01560247e-01 -7.89264262e-01 -4.04727340e-01
3.84908319e-01 -4.03010339e-01 7.67706633e-01 -1.25400078e+00
-1.22298026e+00 1.20610535e+00 1.61638081e-01 -4.74914342e-01
1.07035255e+00 -7.40548298e-02 -3.89931083e-01 3.60338360e-01
1.65476158e-01 5.84360778e-01 1.22081614e+00 -1.72292495e+00
-4.75110680e-01 -2.33860075e-01 -3.15737218e-01 -5.62767833e-02
-3.18387091e-01 -2.11127937e-01 -1.67479977e-01 -9.45857048e-01
1.74629480e-01 -1.03216553e+00 -3.30601633e-01 4.64527965e-01
-6.63106799e-01 5.48303388e-02 5.57664216e-01 -5.76918185e-01
9.08098221e-01 -2.22237992e+00 -1.07691512e-01 1.85934544e-01
2.30012357e-01 4.70499784e-01 1.52538598e-01 -2.29689524e-01
-1.43415481e-01 2.74191409e-01 -7.50465870e-01 -5.29082179e-01
-1.89809844e-01 2.90123671e-01 -9.29622278e-02 9.32957232e-01
3.18957686e-01 6.51482999e-01 -1.18901145e+00 -9.46503937e-01
3.05900216e-01 5.93505383e-01 -2.85092771e-01 2.28762329e-01
9.03131589e-02 9.07038808e-01 -5.13580561e-01 8.19426835e-01
1.02814317e+00 -1.91265509e-01 -8.30106214e-02 -3.79740417e-01
5.00806749e-01 -2.50737667e-01 -1.13907111e+00 1.41888380e+00
-3.62515658e-01 2.27498300e-02 2.00455129e-01 -9.17441249e-01
7.78283238e-01 4.07268703e-01 5.46558380e-01 -3.79670382e-01
2.93039262e-01 3.69720995e-01 -3.23967457e-01 -4.69506681e-01
-9.63053405e-02 -6.06413543e-01 7.40802661e-02 1.06996872e-01
4.81954664e-02 -5.74315935e-02 -1.42793342e-01 2.88619325e-02
8.01556349e-01 1.95447281e-01 2.30602324e-01 -1.70790479e-01
7.71836936e-01 -2.14445055e-01 1.00583863e+00 6.55266523e-01
-6.63592577e-01 9.40496683e-01 3.18001628e-01 -5.18718511e-02
-8.43775630e-01 -1.20959342e+00 -7.50723600e-01 6.14456654e-01
1.52298078e-01 6.67467788e-02 -7.88243890e-01 -1.33297813e+00
-4.39850651e-02 7.20705926e-01 -7.12829351e-01 -4.16692764e-01
-3.06161314e-01 -8.69772375e-01 5.78420818e-01 9.31088775e-02
3.73834252e-01 -1.01983249e+00 -2.40476027e-01 5.17917015e-02
-2.79640943e-01 -1.04882264e+00 -4.10455972e-01 2.94273913e-01
-6.92342281e-01 -1.16650403e+00 -7.95325279e-01 -6.72235787e-01
1.20386171e+00 -1.53053701e-01 1.03024375e+00 1.10426582e-01
-2.62183577e-01 5.60999364e-02 -6.38989687e-01 -3.57033819e-01
-1.00654244e+00 -2.96418160e-01 -5.69821056e-03 5.29171467e-01
9.63279009e-02 -4.30927813e-01 -5.99368811e-01 4.96744365e-01
-9.76466060e-01 -9.16748941e-02 6.26729488e-01 1.31401503e+00
1.11839807e+00 7.72827864e-02 7.91349411e-01 -1.48795950e+00
4.21284586e-02 -5.44235528e-01 -2.84017235e-01 3.38526338e-01
-1.01390541e+00 -2.27372721e-02 9.54807520e-01 -6.24348760e-01
-9.47587550e-01 2.28852004e-01 -2.84730792e-01 -8.34432304e-01
-3.65622401e-01 1.55728739e-02 -2.43969381e-01 -3.32030773e-01
6.80014312e-01 2.03568712e-01 2.06979319e-01 -2.91808516e-01
2.94338971e-01 5.93642533e-01 4.20669556e-01 -4.49568123e-01
9.73275900e-01 8.09844136e-01 1.02396257e-01 -3.62991601e-01
-1.30237520e+00 -4.20133322e-01 -4.86865371e-01 -3.45506430e-01
5.88158369e-01 -7.30974495e-01 -1.34509742e-01 5.75045049e-01
-5.57798505e-01 -8.74089599e-02 -6.03934109e-01 4.67326224e-01
-5.71027339e-01 6.51921511e-01 -5.08213341e-01 -7.93576419e-01
-4.11936492e-01 -1.18339491e+00 1.03504479e+00 1.30745679e-01
1.01916224e-01 -8.81091118e-01 -1.91749454e-01 4.62129921e-01
1.34902552e-01 6.87488854e-01 6.14892483e-01 -8.18559766e-01
-2.29998007e-01 -4.02006298e-01 -1.43888444e-01 9.87164378e-01
3.02021950e-01 -2.80463040e-01 -1.21476686e+00 -4.59274173e-01
3.85029793e-01 -7.53557920e-01 8.67585421e-01 4.38606560e-01
1.44004202e+00 -2.49484330e-01 -2.65957952e-01 8.12144995e-01
1.50841951e+00 -2.41569623e-01 5.21418750e-01 2.56357975e-02
6.46603823e-01 6.87246382e-01 9.84555960e-01 3.90263826e-01
1.98753718e-02 4.64580536e-01 5.66145360e-01 -2.23169222e-01
-3.42594743e-01 -2.51240253e-01 1.81535184e-01 6.56128824e-01
5.01212180e-01 -1.19045973e-01 -5.23829460e-01 4.64358330e-01
-1.60143995e+00 -7.70739317e-01 -5.13102822e-02 2.43421459e+00
1.35209179e+00 4.06771034e-01 -1.93083286e-01 4.10197020e-01
8.16410303e-01 4.88126464e-02 -6.48801386e-01 1.05310269e-01
-1.16637558e-01 1.06977813e-01 6.23805344e-01 4.25933808e-01
-1.49178088e+00 5.62696636e-01 4.83974695e+00 1.42878294e+00
-9.96612787e-01 4.83628064e-01 1.19686329e+00 7.88564309e-02
-2.26855725e-01 -2.25709915e-01 -5.73901415e-01 6.56913638e-01
7.37031698e-01 1.19019084e-01 5.93474060e-02 9.27242517e-01
4.29475829e-02 2.01799367e-02 -9.60912943e-01 9.51831579e-01
1.25774041e-01 -9.47812617e-01 -9.79294479e-02 -1.82362467e-01
6.96611166e-01 -2.54182309e-01 1.29526317e-01 1.34241536e-01
-2.83147898e-02 -8.23867798e-01 8.10788870e-01 6.49356723e-01
1.03258193e+00 -6.64438128e-01 8.96751404e-01 4.30741072e-01
-7.76902020e-01 9.12492573e-02 -3.86489034e-01 5.70767462e-01
1.38922349e-01 1.09857416e+00 -7.79412270e-01 8.08299482e-01
5.77712059e-01 7.52369881e-01 -5.78473151e-01 9.27327275e-01
-4.00096059e-01 9.53466296e-01 -1.59084663e-01 3.12555581e-01
-1.52964890e-01 -1.30107865e-01 5.70011973e-01 1.12144554e+00
1.49534447e-02 -1.07153744e-01 3.01629394e-01 8.43023121e-01
-1.30987942e-01 1.43854588e-01 -3.30689013e-01 4.93489593e-01
3.47882211e-01 1.38211405e+00 -8.12662840e-01 -2.85559565e-01
-1.73118249e-01 1.05008948e+00 6.93357289e-02 7.93694481e-02
-9.92435992e-01 -2.27542341e-01 1.35857120e-01 9.31273922e-02
2.18058862e-02 3.89611095e-01 -4.09449905e-01 -1.04641342e+00
1.71953008e-01 -7.11949766e-01 4.54342008e-01 -4.46139753e-01
-1.73637235e+00 7.49392271e-01 -1.24302812e-01 -1.74243283e+00
1.43858612e-01 -3.92419845e-01 -3.15067917e-01 6.88206851e-01
-1.76824176e+00 -1.22031784e+00 -1.99801475e-01 5.24883509e-01
4.07110095e-01 1.31878242e-01 8.24387670e-01 5.62904477e-01
-3.80349785e-01 1.00276446e+00 3.26832652e-01 1.06552392e-01
9.50352132e-01 -1.20285475e+00 -2.84974456e-01 8.05758774e-01
9.14141834e-02 1.47614792e-01 7.42615938e-01 -5.39453745e-01
-7.00369537e-01 -1.47700083e+00 4.82827187e-01 -4.61473346e-01
3.13007683e-01 -2.96183228e-01 -1.20502186e+00 2.38564089e-01
-1.84587806e-01 7.14887321e-01 7.57437289e-01 -5.48323631e-01
-2.76567906e-01 -2.63931870e-01 -1.86444712e+00 2.71233618e-01
8.61414194e-01 -3.82751286e-01 -3.83399367e-01 5.22723436e-01
6.72848046e-01 -3.12604785e-01 -9.97026563e-01 7.47385263e-01
4.02869940e-01 -9.01126027e-01 1.06786752e+00 -2.24927291e-01
3.81253302e-01 -2.67307281e-01 -2.87781786e-02 -1.19767797e+00
-1.54348075e-01 -3.24071199e-01 -1.51351113e-02 1.48260093e+00
2.97347695e-01 -5.79863250e-01 9.18959498e-01 3.86240333e-01
-8.65528807e-02 -1.08663797e+00 -1.13466179e+00 -8.61045778e-01
1.39971524e-01 -3.29932064e-01 1.42405242e-01 1.07272732e+00
-4.56865430e-01 -2.72915781e-01 -6.34304821e-01 3.04635108e-01
1.24921703e+00 4.50489204e-03 2.52796710e-01 -1.03197372e+00
-2.40655482e-01 -1.76167548e-01 -3.08575660e-01 -4.56556439e-01
2.87446618e-01 -1.12935746e+00 2.90920138e-01 -1.08207560e+00
1.64972305e-01 -8.45618784e-01 -6.60463989e-01 4.94069099e-01
-3.99538189e-01 6.74388349e-01 -6.71217591e-02 3.15882206e-01
-6.14735186e-01 6.72680616e-01 1.37528515e+00 -2.57530689e-01
1.53702229e-01 3.01143050e-01 -5.49377680e-01 8.03998888e-01
7.36873031e-01 -1.12785482e+00 -2.39562541e-01 1.35473132e-01
-2.30235577e-01 -2.04326496e-01 5.28669775e-01 -9.40860868e-01
-8.97967145e-02 -1.41828746e-01 2.62259692e-01 -2.74848312e-01
1.14366837e-01 -1.00737381e+00 9.17027295e-02 6.87265098e-01
-5.07498920e-01 -8.10230315e-01 -1.67394251e-01 7.32269168e-01
-1.48914754e-01 -5.12703121e-01 1.34319329e+00 -1.37273863e-01
-2.44217232e-01 4.39250141e-01 1.38429478e-01 2.70549148e-01
1.13241971e+00 -6.01719096e-02 1.74345300e-02 -1.60244405e-02
-8.53868067e-01 2.20343634e-01 4.19463307e-01 1.29845455e-01
7.00390279e-01 -1.34803748e+00 -9.31039572e-01 1.97673276e-01
2.91905195e-01 8.12259614e-02 3.77000272e-01 7.91589022e-01
-2.67077982e-01 -2.53308147e-01 -5.05541451e-03 -6.48620367e-01
-1.12002850e+00 7.37629890e-01 4.26580727e-01 -4.00379449e-01
-5.42184711e-01 9.16870534e-01 1.11222535e-01 -5.70395589e-01
3.07524174e-01 -1.13429151e-01 -4.89123985e-02 -1.29236624e-01
4.95704532e-01 2.60376275e-01 1.22082420e-01 -7.35833824e-01
-3.15574110e-01 5.42844772e-01 -4.71486002e-02 2.99024343e-01
1.08687079e+00 -1.38924837e-01 5.31337748e-04 5.00071764e-01
1.39032662e+00 1.05625048e-01 -1.50247395e+00 -4.99356896e-01
-1.18592367e-01 -6.94022119e-01 9.56946313e-02 -8.84307742e-01
-1.31256270e+00 6.45676076e-01 1.01050222e+00 3.00388299e-02
1.24523675e+00 5.45726568e-02 7.85275459e-01 -1.54881686e-01
3.96378875e-01 -9.64158058e-01 3.35188024e-02 -2.61101604e-01
7.85811961e-01 -1.59936643e+00 1.39701009e-01 -5.91856658e-01
-7.71877527e-01 6.86180234e-01 4.18899179e-01 -2.13863447e-01
8.73143911e-01 1.66576415e-01 1.55903518e-01 5.97375482e-02
-2.78068662e-01 8.64162892e-02 2.87350476e-01 5.52616775e-01
6.45656064e-02 1.99533597e-01 -4.33752894e-01 7.47415006e-01
2.67104268e-01 2.12964118e-01 2.88248658e-01 8.44469845e-01
-1.06909595e-01 -1.20549524e+00 -5.00472188e-01 7.16600299e-01
-8.75632167e-01 4.30373363e-02 1.43426418e-01 4.04418677e-01
4.88708138e-01 8.31239283e-01 -4.17159408e-01 -3.64664286e-01
1.80472881e-01 1.55181095e-01 2.81748056e-01 -5.68233848e-01
-4.70408708e-01 6.99431226e-02 -1.01119392e-01 -4.68231469e-01
-6.11027956e-01 -6.89135730e-01 -1.20518458e+00 1.20739222e-01
-6.39963865e-01 8.12714696e-02 3.86136383e-01 6.75592721e-01
1.43742949e-01 4.73755002e-01 1.02404785e+00 -5.74928522e-01
-9.48046029e-01 -7.61632085e-01 -6.61959052e-01 8.71054053e-01
4.48015004e-01 -7.38519967e-01 -6.67204559e-01 1.83420509e-01] | [14.682416915893555, -2.1153132915496826] |
49e2b9ca-cbf4-4dd8-9425-546ff0ea6268 | fast-inference-for-quantile-regression-with | 2209.14502 | null | https://arxiv.org/abs/2209.14502v4 | https://arxiv.org/pdf/2209.14502v4.pdf | Fast Inference for Quantile Regression with Tens of Millions of Observations | Big data analytics has opened new avenues in economic research, but the challenge of analyzing datasets with tens of millions of observations is substantial. Conventional econometric methods based on extreme estimators require large amounts of computing resources and memory, which are often not readily available. In this paper, we focus on linear quantile regression applied to ``ultra-large'' datasets, such as U.S. decennial censuses. A fast inference framework is presented, utilizing stochastic sub-gradient descent (S-subGD) updates. The inference procedure handles cross-sectional data sequentially: (i) updating the parameter estimate with each incoming "new observation", (ii) aggregating it as a Polyak-Ruppert average, and (iii) computing a pivotal statistic for inference using only a solution path. The methodology draws from time series regression to create an asymptotically pivotal statistic through random scaling. Our proposed test statistic is calculated in a fully online fashion and critical values are calculated without resampling. We conduct extensive numerical studies to showcase the computational merits of our proposed inference. For inference problems as large as $(n, d) \sim (10^7, 10^3)$, where $n$ is the sample size and $d$ is the number of regressors, our method generates new insights, surpassing current inference methods in computation. Our method specifically reveals trends in the gender gap in the U.S. college wage premium using millions of observations, while controlling over $10^3$ covariates to mitigate confounding effects. | ['Youngki Shin', 'Myung Hwan Seo', 'Yuan Liao', 'Sokbae Lee'] | 2022-09-29 | null | null | null | null | ['time-series-regression'] | ['time-series'] | [-3.61409664e-01 -3.01984191e-01 -3.15229595e-01 -3.70189279e-01
-9.12310243e-01 -4.66536254e-01 3.02368224e-01 3.54499251e-01
-6.36619449e-01 1.27280402e+00 1.49506658e-01 -7.58690238e-01
-2.98307925e-01 -1.07445681e+00 -7.31256247e-01 -7.08025336e-01
-1.60489142e-01 2.63144225e-01 -5.31953096e-01 1.16248038e-02
3.83459091e-01 2.24239483e-01 -1.27697039e+00 -6.35516942e-01
1.55732417e+00 1.11292255e+00 -2.74577022e-01 5.23649454e-01
1.25449970e-01 3.36458057e-01 -3.78506333e-01 -8.07677805e-01
4.08739001e-01 -3.08956623e-01 -1.26356870e-01 -2.76670605e-01
3.48667890e-01 -4.03774351e-01 2.15060294e-01 9.73871291e-01
6.39317214e-01 4.40800905e-01 7.11323798e-01 -1.05787921e+00
-4.89452541e-01 6.79967940e-01 -1.11408293e+00 2.65537679e-01
2.23697424e-01 2.18530409e-02 8.77669275e-01 -9.41682577e-01
2.74512351e-01 1.20914447e+00 7.87416995e-01 -6.25796691e-02
-1.34346449e+00 -9.13684487e-01 4.62082513e-02 -1.41611308e-01
-1.37991250e+00 -2.55230784e-01 4.63262737e-01 -5.31560302e-01
5.46241760e-01 3.60645324e-01 6.22224987e-01 8.36943805e-01
4.67329413e-01 1.01181000e-01 1.21875727e+00 -2.23420411e-01
6.68258011e-01 -1.73266664e-01 3.07641566e-01 3.22323859e-01
7.20880032e-01 1.55040771e-01 -5.16725957e-01 -5.99410117e-01
6.76381171e-01 2.96194792e-01 4.74387705e-01 3.29035878e-01
-8.15959513e-01 1.10904384e+00 2.40672342e-02 -2.08903000e-01
-6.56396031e-01 3.39833826e-01 4.38282669e-01 4.04759049e-01
1.09920633e+00 -1.49720788e-01 -6.40995800e-01 -3.22392911e-01
-1.09236646e+00 5.80875814e-01 7.78820097e-01 5.48310578e-01
5.90466976e-01 2.16089547e-01 -1.10133067e-01 7.59817898e-01
-1.34268790e-01 1.05258596e+00 2.22389445e-01 -1.04223132e+00
6.72476053e-01 3.71505678e-01 3.44210953e-01 -1.14690363e+00
-2.95228928e-01 -5.19771397e-01 -1.26858234e+00 -1.21673189e-01
8.29197943e-01 -8.39232802e-01 -3.59850585e-01 1.82953680e+00
5.49260497e-01 1.14436187e-01 -2.61634558e-01 5.22287190e-01
5.50767630e-02 5.58380008e-01 2.04066694e-01 -7.27216721e-01
1.39197135e+00 -2.07075089e-01 -4.68806863e-01 1.91350114e-02
3.43467206e-01 -5.59345722e-01 9.27023351e-01 3.07632506e-01
-1.35584474e+00 -5.23544788e-01 -2.57337481e-01 1.64300948e-01
-3.29490870e-01 -4.56965156e-03 8.96796048e-01 7.38617361e-01
-7.52800345e-01 6.24043524e-01 -7.21928120e-01 1.07839637e-01
4.91807491e-01 2.15793774e-01 5.65465391e-02 -2.68609822e-02
-1.15891528e+00 2.26339057e-01 -2.33236864e-01 2.23634854e-01
-3.98866057e-01 -1.20823765e+00 -7.01618791e-01 3.59472424e-01
3.65905732e-01 -7.97238588e-01 7.14391649e-01 -5.91516256e-01
-1.20520902e+00 3.98762703e-01 -4.46769834e-01 -4.24237013e-01
9.09078300e-01 -1.48638666e-01 -3.06145281e-01 -1.67073384e-01
3.84795785e-01 -1.48678213e-01 6.92293704e-01 -5.91033638e-01
-7.07056105e-01 -8.66163254e-01 -3.45130205e-01 -2.97999293e-01
-1.83343604e-01 1.38853565e-01 1.38421878e-01 -8.82431746e-01
1.10844905e-02 -7.51440465e-01 -5.91867507e-01 -5.89506686e-01
-2.87704408e-01 -2.76342869e-01 -7.15411752e-02 -8.39898586e-01
1.60620379e+00 -1.97605634e+00 -2.62870908e-01 5.97627819e-01
1.03198700e-01 -4.21131492e-01 3.66762310e-01 4.32145447e-01
1.83795735e-01 1.05427265e-01 -2.95704156e-01 -1.80739745e-01
1.95148498e-01 -3.45544845e-01 -4.92500126e-01 6.40473664e-01
-2.94683188e-01 6.24900877e-01 -7.52983868e-01 -2.84260184e-01
-4.76003923e-02 1.45893991e-01 -7.86004126e-01 -2.23248810e-01
2.42888942e-01 2.03769460e-01 -4.39187706e-01 6.67549133e-01
9.84962702e-01 -2.46342182e-01 1.16548568e-01 2.87983507e-01
-5.52218795e-01 -1.69323236e-01 -1.42137182e+00 1.10916662e+00
-6.12924278e-01 3.39441627e-01 8.45479295e-02 -1.44595015e+00
9.75065291e-01 7.42560849e-02 4.01124001e-01 -5.75780034e-01
-1.56384613e-02 5.41185558e-01 -2.48200789e-01 -3.20062667e-01
4.24017608e-01 -3.64112884e-01 -4.69747663e-01 5.74147046e-01
-4.29633617e-01 1.19389080e-01 6.45766854e-01 9.68371704e-02
8.13621342e-01 -1.94205552e-01 4.32837814e-01 -6.26612604e-01
1.27604216e-01 -2.26588964e-01 8.49231958e-01 9.20485616e-01
1.15672268e-01 1.78094748e-02 1.01189184e+00 -5.10702431e-01
-1.06426954e+00 -1.32692516e+00 -5.33455849e-01 1.19242108e+00
-3.28306645e-01 -1.21159572e-02 -6.69837415e-01 -8.21208134e-02
6.42759681e-01 6.23200059e-01 -6.73857868e-01 1.71605095e-01
-3.43334198e-01 -1.21474969e+00 1.64442480e-01 4.38238353e-01
6.00899458e-01 -5.70313752e-01 -6.88098609e-01 1.18172526e-01
-6.50533959e-02 -5.35190403e-01 -3.36589813e-01 -1.28807947e-01
-1.06567967e+00 -6.44972503e-01 -7.64305055e-01 -2.98477232e-01
8.09002519e-01 -1.19437017e-01 1.09058583e+00 -2.91721523e-01
-3.40346754e-01 -2.35175006e-02 9.39490497e-02 -5.40529966e-01
2.53637671e-01 5.94125167e-02 3.48549843e-01 2.42536366e-02
3.90891552e-01 -8.96504998e-01 -9.45232332e-01 -1.11268107e-02
-5.10655463e-01 -3.40916425e-01 3.83776248e-01 8.56749535e-01
3.71532470e-01 6.43151999e-02 1.18237424e+00 -1.07863939e+00
7.62559235e-01 -9.48682189e-01 -1.15281951e+00 -1.31341610e-02
-7.46199131e-01 -8.41071904e-02 9.52849627e-01 -5.52466214e-01
-1.21782923e+00 -5.97410560e-01 2.44390190e-01 4.08550166e-02
2.21781522e-01 1.08268678e+00 3.53697181e-01 4.39457238e-01
4.22841460e-01 -1.86861962e-01 -5.82126118e-02 -4.50369477e-01
9.67547372e-02 6.22339249e-01 3.95102799e-01 -9.40373957e-01
5.84596992e-01 5.52680969e-01 2.05034629e-01 -6.30647421e-01
-6.35640621e-01 -2.56159455e-01 -2.38628134e-01 9.16015953e-02
3.33468825e-01 -9.65359211e-01 -1.46037388e+00 3.55230391e-01
-5.18597007e-01 -2.23462194e-01 -4.60784465e-01 1.00387323e+00
-3.26147676e-01 -7.38568306e-02 -7.06305504e-01 -1.22360849e+00
-3.43999535e-01 -5.84698617e-01 7.46076703e-01 4.38307732e-01
-1.91852629e-01 -1.11372650e+00 2.96678156e-01 4.54050332e-01
3.77346903e-01 5.99356472e-01 8.64953160e-01 -3.66574258e-01
-1.83381334e-01 -5.26873231e-01 -3.15683067e-01 1.21431025e-02
-4.49635796e-02 2.35321015e-01 -5.94481468e-01 -4.28277850e-01
-1.74511611e-01 -2.23112643e-01 5.99806190e-01 1.12970424e+00
1.55161881e+00 -4.50180978e-01 -1.71093792e-01 7.11677611e-01
1.37394416e+00 -7.39050582e-02 4.15769905e-01 -9.55036655e-02
2.08305210e-01 4.61930841e-01 4.83891219e-01 1.33785665e+00
4.95848775e-01 -1.18585154e-02 -1.12466067e-01 -1.83054730e-01
7.75141716e-01 -2.17979535e-01 3.31532121e-01 7.60755241e-01
-4.81718063e-01 1.22576356e-01 -9.49237287e-01 5.66642344e-01
-1.60585558e+00 -1.32094049e+00 -9.07260031e-02 2.66414094e+00
1.10233533e+00 1.53342828e-01 3.46531421e-01 -2.00321421e-01
7.04745173e-01 5.92819527e-02 -8.09431911e-01 -6.58682764e-01
1.03746675e-01 6.34844482e-01 7.16928482e-01 4.80886519e-01
-5.40962517e-01 5.46115518e-01 6.01881123e+00 9.77997124e-01
-8.82387817e-01 -9.70166922e-03 1.16866136e+00 -4.12686765e-01
-4.94548529e-01 2.86552370e-01 -8.65307450e-01 8.07750344e-01
1.19765437e+00 -7.91140199e-01 3.72829318e-01 6.55939817e-01
8.09405982e-01 -6.55651867e-01 -6.21511340e-01 7.49637544e-01
-3.70890856e-01 -1.19570637e+00 -3.50142419e-01 4.37057048e-01
1.07761514e+00 -2.62305439e-01 2.40620255e-01 4.78262067e-01
7.92203367e-01 -1.07564402e+00 3.69862735e-01 6.42844141e-01
9.29932714e-01 -1.20290530e+00 6.53443038e-01 4.33183879e-01
-1.09750783e+00 -2.96134591e-01 -6.31516457e-01 -6.92643106e-01
8.57143924e-02 1.57698774e+00 -2.09786221e-01 4.45135087e-01
6.65901303e-01 2.81746119e-01 -1.41733825e-01 6.37838542e-01
2.90609539e-01 8.99526715e-01 -6.51233733e-01 -1.23574212e-01
1.01547390e-02 -9.29630756e-01 3.05708498e-02 8.73975635e-01
8.44119549e-01 3.90057027e-01 -3.20626870e-02 9.34783936e-01
-2.76852727e-01 3.07986736e-01 -4.10155058e-01 1.73143804e-01
8.79837453e-01 1.06575513e+00 -6.71708822e-01 -4.64631110e-01
-5.17603159e-01 3.04139078e-01 2.99392313e-01 6.15756512e-01
-8.03316474e-01 -5.77741146e-01 6.17266178e-01 7.08082020e-02
1.36767492e-01 -2.90820867e-01 -5.68806350e-01 -1.18835473e+00
8.44390318e-02 -6.66315317e-01 7.15971470e-01 -2.45572105e-01
-1.38182688e+00 -3.19016337e-01 1.38876572e-01 -6.18175030e-01
-5.66843569e-01 -1.76712826e-01 -8.19336236e-01 1.03936684e+00
-1.10446048e+00 -5.60746014e-01 -5.27535528e-02 5.01996517e-01
2.71548063e-01 3.23026814e-02 4.34787810e-01 3.18341374e-01
-9.26329672e-01 7.84457266e-01 8.23573589e-01 1.67622697e-02
4.92784709e-01 -1.09110665e+00 2.20701188e-01 7.81506121e-01
-5.69319487e-01 8.48739386e-01 7.77869523e-01 -8.85234475e-01
-1.18757474e+00 -7.01060772e-01 1.07265770e+00 -1.80782619e-04
8.06358218e-01 -4.03834492e-01 -5.03117621e-01 5.74069619e-01
-3.09918851e-01 2.70399004e-01 9.03228164e-01 6.03675306e-01
-1.13942465e-02 -5.90250552e-01 -1.38763738e+00 7.91881442e-01
8.13826382e-01 -1.59987882e-01 3.70252244e-02 1.46108761e-01
3.94779176e-01 -1.13547347e-01 -1.27424204e+00 1.60972655e-01
8.96716654e-01 -9.52060342e-01 8.97599995e-01 -6.56503797e-01
6.01429343e-01 3.93485636e-01 -1.34756774e-01 -1.26548553e+00
8.37492496e-02 -9.17606950e-01 1.15585379e-01 1.26598799e+00
2.75121152e-01 -1.09369183e+00 5.91808856e-01 9.61059690e-01
3.46297532e-01 -8.51789951e-01 -1.12143123e+00 -7.59305179e-01
5.55294394e-01 -3.83937836e-01 8.06855798e-01 1.02512729e+00
6.53235242e-02 -9.42045599e-02 -3.13304693e-01 -2.74064969e-02
1.15288305e+00 5.79943240e-01 8.88609886e-01 -1.38780940e+00
-1.39792904e-01 -3.19030762e-01 8.13958421e-02 -5.56173563e-01
1.90094382e-01 -5.09689212e-01 -4.68705446e-01 -9.17231023e-01
5.22919655e-01 -5.15988350e-01 -1.45238534e-01 -1.74223422e-03
-3.91277939e-01 -4.92960680e-04 -2.97298670e-01 -2.82023787e-01
1.58159900e-02 4.89079803e-01 8.73894691e-01 2.48747632e-01
-2.55774349e-01 2.25809723e-01 -7.02535212e-01 7.21753955e-01
9.69067156e-01 -3.40147257e-01 -2.16345340e-01 -5.58485240e-02
7.69289196e-01 5.57264686e-01 3.88170540e-01 -6.08200312e-01
1.05635643e-01 -7.72565663e-01 6.54602647e-01 -5.40191233e-01
-2.40724713e-01 -3.30519557e-01 1.05367385e-01 5.64384401e-01
-2.98889190e-01 3.49949121e-01 6.48702607e-02 4.53470647e-01
-3.18138376e-02 3.47319059e-02 6.11873507e-01 6.54944181e-02
1.13814078e-01 3.50492567e-01 -2.62640566e-01 5.57784617e-01
8.94568443e-01 1.15379632e-01 -3.21454704e-01 -4.24540788e-01
-4.24663424e-01 5.18164217e-01 1.94758728e-01 -2.93902427e-01
3.39960188e-01 -1.35271728e+00 -1.05763185e+00 2.38515511e-01
-5.11064351e-01 -4.07634266e-02 4.87568498e-01 1.12161601e+00
-3.09804529e-01 2.23186821e-01 -4.46516760e-02 -1.87171444e-01
-6.17554247e-01 2.35999376e-01 -1.52458847e-01 -3.76806945e-01
-2.92034030e-01 6.82705402e-01 1.62759181e-02 -3.07704329e-01
-8.78094658e-02 -4.71847296e-01 3.06913704e-01 6.09323561e-01
3.81064057e-01 1.16466534e+00 -2.19100714e-01 -8.06979835e-03
-2.27675021e-01 5.22863984e-01 2.08747447e-01 -2.67653525e-01
1.60624301e+00 -3.33042830e-01 -2.72608936e-01 6.84157670e-01
1.05401635e+00 5.24536252e-01 -1.26494145e+00 -1.10379336e-02
-2.25032330e-01 -6.36384428e-01 -1.66253686e-01 -4.12561119e-01
-1.02802145e+00 7.39702880e-01 1.88745975e-01 3.88324320e-01
9.60643113e-01 -3.02550673e-01 7.18028009e-01 1.86203226e-01
2.84749657e-01 -1.49310660e+00 -2.60823607e-01 2.56489843e-01
4.32788759e-01 -1.27937317e+00 5.00039697e-01 2.13350996e-01
-1.93624824e-01 8.17843795e-01 9.47501883e-02 -3.22960466e-01
8.04151654e-01 1.03199705e-01 -4.50362623e-01 -7.39548281e-02
-7.36565471e-01 2.13840306e-01 -3.82284597e-02 -1.08938038e-01
5.18010557e-01 3.63460541e-01 -9.32861328e-01 8.29909027e-01
-6.28218591e-01 1.25380754e-01 5.23229718e-01 6.60217404e-01
-3.33057910e-01 -6.01794958e-01 -5.34231603e-01 1.08534169e+00
-8.70917141e-01 -2.92272985e-01 3.32586706e-01 8.42045724e-01
-1.87769562e-01 7.61320233e-01 4.48789150e-01 2.25606501e-01
6.14756383e-02 -2.58923322e-02 2.27066092e-02 -1.66200563e-01
-2.96964824e-01 7.31266364e-02 -2.09659606e-01 -3.36897731e-01
1.48283904e-02 -1.09954500e+00 -1.02395153e+00 -1.35809803e+00
-1.33066043e-01 4.20791268e-01 5.76673806e-01 8.25341344e-01
2.86195934e-01 5.46345003e-02 1.09351468e+00 -6.45920157e-01
-9.66959298e-01 -9.20884490e-01 -1.00753105e+00 2.89385051e-01
2.48720378e-01 -5.49077868e-01 -8.46859455e-01 -9.60496068e-02] | [7.0985331535339355, 4.451025009155273] |
d20e68a0-8d27-4ef5-a2cd-b9559f311f43 | rule-based-event-extraction-for-artificial | null | null | https://aclanthology.org/2022.pandl-1.9 | https://aclanthology.org/2022.pandl-1.9.pdf | Rule Based Event Extraction for Artificial Social Intelligence | Natural language (as opposed to structured communication modes such as Morse code) is by far the most common mode of communication between humans, and can thus provide significant insight into both individual mental states and interpersonal dynamics. As part of DARPA’s Artificial Social Intelligence for Successful Teams (ASIST) program, we are developing an AI agent team member that constructs and maintains models of their human teammates and provides appropriate task-relevant advice to improve team processes and mission performance. One of the key components of this agent is a module that uses a rule-based approach to extract task-relevant events from natural language utterances in real time, and publish them for consumption by downstream components. In this case study, we evaluate the performance of our rule-based event extraction system on a recently conducted ASIST experiment consisting of a simulated urban search and rescue mission in Minecraft. We compare the performance of our approach with that of a zero-shot neural classifier, and find that our approach outperforms the classifier for all event types, even when the classifier is used in an oracle setting where it knows how many events should be extracted from each utterance. | ['Rebecca Sharp', 'Adarsh Pyarelal', 'Chen Chen', 'Yuwei Wang', 'Remo Nitschke'] | null | null | null | null | pandl-coling-2022-10 | ['event-extraction'] | ['natural-language-processing'] | [ 1.32719144e-01 5.55223227e-01 1.73975125e-01 -4.60099459e-01
-3.96766394e-01 -3.98335159e-01 7.88616896e-01 7.77322590e-01
-5.86499512e-01 8.37304413e-01 3.95487070e-01 -2.90524364e-01
-3.74252260e-01 -9.69531536e-01 -2.66701192e-01 -1.79286018e-01
-5.62251031e-01 1.03489709e+00 2.30766565e-01 -8.83117139e-01
7.85849988e-03 3.77885193e-01 -1.80174565e+00 3.21839571e-01
4.30899441e-01 7.60547221e-01 6.94007205e-04 8.80462408e-01
1.32516712e-01 1.73169506e+00 -9.69891012e-01 -1.88778847e-01
1.95333466e-01 -5.46170831e-01 -1.09241247e+00 1.04814231e-01
-3.62314820e-01 -6.08052649e-02 -1.63654432e-01 3.87844354e-01
2.67339885e-01 3.38778585e-01 5.29688239e-01 -1.33183610e+00
4.79758352e-01 9.06192064e-01 1.88663349e-01 3.31510365e-01
7.36434221e-01 2.73644745e-01 7.70725727e-01 -1.51965410e-01
6.72569931e-01 1.26357841e+00 6.66527152e-01 7.94406012e-02
-9.93114352e-01 -5.96755564e-01 2.78294738e-02 7.99093693e-02
-1.14418983e+00 -6.11508846e-01 2.70533442e-01 -5.51063359e-01
1.25160968e+00 2.73694862e-02 6.89506769e-01 9.32714581e-01
5.56634724e-01 5.12461662e-01 8.31211030e-01 -4.05204564e-01
6.42924607e-01 6.78677857e-02 3.07183355e-01 8.12070906e-01
2.56254002e-02 2.04285443e-01 -8.15258265e-01 -6.33957684e-01
3.71247470e-01 -1.68102294e-01 5.71719855e-02 1.53245762e-01
-1.51787066e+00 8.25580239e-01 -1.22996923e-02 2.05742821e-01
-8.61634552e-01 1.89579472e-01 3.96753341e-01 5.62720180e-01
3.09951276e-01 7.00473189e-01 -2.65557706e-01 -7.59032786e-01
-5.19947529e-01 6.16093099e-01 1.41937160e+00 5.02501369e-01
6.36642873e-01 -8.69272649e-02 -3.34362686e-02 6.06585026e-01
8.85510743e-02 2.01756835e-01 4.51116294e-01 -1.25193274e+00
2.53052324e-01 8.12008023e-01 3.40828419e-01 -1.18795621e+00
-9.47072208e-01 -2.02754334e-01 -2.53336698e-01 2.29733959e-01
3.30274761e-01 -6.95501685e-01 -3.23853552e-01 1.54392338e+00
4.50309008e-01 4.87945490e-02 4.67299044e-01 6.40509486e-01
5.92539847e-01 6.57585204e-01 6.50788285e-03 -3.46282691e-01
1.19349468e+00 -3.86090934e-01 -6.94200099e-01 -7.10023999e-01
6.80641294e-01 -3.70205849e-01 3.74510169e-01 5.99893153e-01
-1.07221055e+00 -3.34738672e-01 -1.02927995e+00 5.39775670e-01
-2.24955156e-01 -4.09918845e-01 9.65714514e-01 3.31318498e-01
-7.05660284e-01 4.95495439e-01 -1.09688175e+00 -7.85604775e-01
-1.31924808e-01 3.26214194e-01 -3.24129343e-01 3.07756990e-01
-1.29177403e+00 1.23134196e+00 5.11041760e-01 -2.26934999e-01
-9.78415132e-01 -2.70133317e-02 -1.11411095e+00 -8.74291956e-02
5.94261944e-01 -3.84753287e-01 1.74636912e+00 -7.10372329e-01
-1.40807867e+00 4.70341057e-01 5.99514768e-02 -8.33622456e-01
6.31298348e-02 2.48696208e-01 -2.72701293e-01 1.28930941e-01
2.31384695e-01 3.68024528e-01 4.15394038e-01 -9.52877820e-01
-1.03009081e+00 -1.83259636e-01 4.80525613e-01 3.55443895e-01
-7.34975189e-02 3.70598376e-01 1.79820985e-01 -1.50572449e-01
-1.43333584e-01 -1.03140652e+00 -3.93648237e-01 -6.37521267e-01
5.46247475e-02 -3.17941844e-01 2.66341805e-01 -5.37774980e-01
1.13667917e+00 -1.92006540e+00 2.86930203e-01 4.01050895e-01
2.91975252e-02 -2.27192864e-01 1.54233828e-01 1.01291180e+00
1.25653684e-01 -2.89535403e-01 -1.83565080e-01 -1.94676787e-01
-1.56411920e-02 4.89767969e-01 -1.80716619e-01 1.44711673e-01
7.72592276e-02 3.41219157e-01 -1.02846777e+00 -5.22532821e-01
3.02315559e-02 -1.11295067e-01 -3.22574943e-01 2.91689128e-01
-3.95118147e-01 2.57884800e-01 -4.48424518e-01 3.13597888e-01
-3.42275023e-01 -1.54016942e-01 5.74120224e-01 4.98537660e-01
-2.16059670e-01 5.22620559e-01 -1.21727443e+00 1.42258143e+00
-5.14040768e-01 7.14665890e-01 1.65727615e-01 -9.55660760e-01
9.43779767e-01 5.76303601e-01 6.11115456e-01 -2.71078408e-01
3.60896826e-01 -2.04220116e-01 2.67541587e-01 -7.16917515e-01
4.68531579e-01 -1.56091824e-01 -7.13277042e-01 1.12456286e+00
4.50447164e-02 -3.55719358e-01 4.29193765e-01 4.95133847e-01
1.79320812e+00 -2.53950477e-01 7.97941625e-01 -3.81827191e-03
1.29577637e-01 6.95382655e-01 6.36015654e-01 7.79249787e-01
-8.42191875e-02 -8.68878851e-04 5.75277567e-01 -7.01846123e-01
-5.79930663e-01 -4.64109182e-01 2.88242221e-01 1.15736091e+00
1.11472607e-03 -7.24780440e-01 -6.30083919e-01 -3.38335395e-01
-1.55644655e-01 1.05328321e+00 -5.29183149e-01 -2.80001968e-01
-3.43603760e-01 -6.64992869e-01 5.50787032e-01 2.15620577e-01
3.26342911e-01 -1.40492618e+00 -1.35258138e+00 6.23883188e-01
-5.16050875e-01 -1.11854374e+00 6.90785423e-02 5.68029344e-01
-3.52514237e-01 -1.23453665e+00 3.71107489e-01 -4.41475689e-01
3.06061059e-01 -8.80162939e-02 1.08404958e+00 6.75117522e-02
-3.53909880e-01 6.41158998e-01 -6.69063926e-01 -1.01504612e+00
-8.11146677e-01 -2.04826176e-01 3.35102737e-01 7.28824660e-02
4.23414141e-01 -5.50387502e-01 8.84419009e-02 3.52496356e-01
-6.14227414e-01 4.80449498e-02 2.80868173e-01 6.92953706e-01
-1.89104214e-01 6.55882359e-01 6.26418412e-01 -7.43331254e-01
1.10158730e+00 -6.88996434e-01 -4.55853790e-01 1.21405222e-01
-2.69468963e-01 -9.63737145e-02 2.93323308e-01 -2.00266927e-01
-1.06123233e+00 1.31883115e-01 3.48186284e-01 2.61800587e-01
-4.80359256e-01 9.30190802e-01 2.98347205e-01 1.87383249e-01
1.02270138e+00 -2.95108976e-03 2.98653185e-01 1.43924743e-01
-1.66511089e-01 1.07688367e+00 5.59774041e-01 -6.98204041e-01
5.19913614e-01 4.29518908e-01 -2.50368148e-01 -9.36102629e-01
-8.95778120e-01 -4.37945068e-01 -3.50619465e-01 -7.74783611e-01
6.94283426e-01 -9.64464545e-01 -1.21230519e+00 1.97818294e-01
-1.22707677e+00 -6.54553354e-01 -3.63182873e-01 5.93411565e-01
-7.18784451e-01 -1.56950146e-01 -5.70655644e-01 -1.27123237e+00
8.44517946e-02 -7.03228056e-01 7.85630226e-01 1.86288059e-01
-9.23113227e-01 -7.54168689e-01 2.27396548e-01 4.96058285e-01
2.25581214e-01 4.81146216e-01 6.23724043e-01 -1.15077853e+00
-1.64105028e-01 -4.71335292e-01 4.96836156e-01 1.41091505e-02
-4.11188751e-02 -2.51754105e-01 -5.99048257e-01 -1.26284182e-01
1.38597637e-01 -5.93075573e-01 2.28088185e-01 -3.83193977e-02
2.18138576e-01 -2.35104561e-01 -3.53146911e-01 -1.54373422e-01
6.84206724e-01 5.61918557e-01 3.37810844e-01 3.65187585e-01
-6.25458285e-02 1.10122526e+00 1.01388085e+00 8.75722051e-01
7.83061504e-01 6.21482790e-01 2.80840814e-01 3.87164205e-01
4.26928073e-01 9.48558375e-03 7.14326084e-01 5.13711631e-01
-2.95669615e-01 -2.25648180e-01 -1.38922274e+00 4.75008100e-01
-2.25976729e+00 -1.26386225e+00 1.71165094e-01 1.92455304e+00
7.61730254e-01 4.09323066e-01 2.22576708e-01 1.84578508e-01
3.55044067e-01 -1.26696169e-01 -2.00021341e-01 -6.02714479e-01
3.87577713e-01 1.43043816e-01 2.74993628e-01 5.21293879e-01
-8.41642678e-01 7.34601319e-01 6.54584026e+00 2.25572035e-01
-5.53109109e-01 1.15266465e-01 1.06605560e-01 -2.67136872e-01
3.68307084e-01 3.95493768e-02 -6.09221637e-01 2.06610054e-01
1.46559799e+00 -4.54217196e-01 6.99396491e-01 6.05506122e-01
4.90633130e-01 -5.79382360e-01 -1.32534015e+00 6.26451313e-01
2.19046488e-01 -1.38649893e+00 -5.81533849e-01 1.01280466e-01
3.16146314e-01 6.08414039e-02 -7.37640321e-01 5.03643990e-01
1.02468014e+00 -9.72371995e-01 8.22840154e-01 6.62808597e-01
-1.07622370e-01 -9.53405857e-01 8.86878848e-01 1.03970289e+00
-8.92875671e-01 -5.87397933e-01 5.16327471e-02 -8.05975139e-01
3.75571877e-01 2.62387753e-01 -1.42117012e+00 4.21967685e-01
6.52416229e-01 6.12901270e-01 -2.06644326e-01 6.75135374e-01
-1.26840636e-01 8.15334678e-01 -5.35972238e-01 -2.86379158e-01
1.20659493e-01 9.89761353e-02 8.21055770e-01 9.76364553e-01
-3.37550719e-03 6.24780238e-01 7.29960501e-01 4.72526819e-01
4.83525455e-01 -2.89816946e-01 -9.42296386e-01 -2.11806729e-01
5.98217428e-01 1.06201959e+00 -7.36888647e-01 -3.73825580e-01
-1.72427550e-01 4.22566801e-01 2.32234359e-01 1.45890519e-01
-5.27866364e-01 -4.49965626e-01 7.08283365e-01 1.08465746e-01
-1.17756752e-02 -3.82753551e-01 9.28786099e-02 -6.62730515e-01
-2.66556829e-01 -1.36352170e+00 5.52728832e-01 -8.08060884e-01
-9.67949510e-01 7.93494284e-01 2.56788850e-01 -1.12737906e+00
-9.24566031e-01 -2.16172397e-01 -6.75909996e-01 3.65301311e-01
-7.24479735e-01 -7.57976115e-01 -4.23386633e-01 5.73965311e-01
5.26764274e-01 -5.46031654e-01 1.09871435e+00 -9.89884436e-02
-5.49364567e-01 -1.60552979e-01 -5.78402519e-01 2.20701918e-01
4.56978172e-01 -8.87300611e-01 2.64510155e-01 6.58830225e-01
6.51584715e-02 5.04464447e-01 1.02498186e+00 -9.26306725e-01
-1.43173254e+00 -8.56199145e-01 9.11456645e-01 -6.07512236e-01
7.41031766e-01 -3.32381070e-01 -4.15555060e-01 1.11094546e+00
2.34357521e-01 -6.22223854e-01 7.55141437e-01 1.92098781e-01
2.67485112e-01 1.18718065e-01 -8.75134528e-01 5.54869533e-01
8.59941721e-01 -2.25860342e-01 -1.14057684e+00 6.73448384e-01
6.39620602e-01 -3.31601858e-01 -8.57698023e-01 2.42407784e-01
1.36377230e-01 -9.15312648e-01 5.38123190e-01 -6.37533844e-01
2.10133031e-01 -4.04093891e-01 -1.01191990e-01 -1.43911707e+00
-6.18949234e-02 -9.70071912e-01 1.18817896e-01 8.53568494e-01
4.39322054e-01 -7.71894753e-01 3.75676632e-01 8.28181565e-01
-1.24287540e-02 -1.72659457e-01 -6.49908125e-01 -6.06755614e-01
-7.60720551e-01 -7.45891452e-01 4.28391486e-01 7.46926844e-01
7.05841780e-01 6.88112319e-01 -2.21347138e-01 3.29989731e-01
5.47699571e-01 -4.42424901e-02 1.10380411e+00 -1.47394347e+00
-5.16149759e-01 4.06262204e-02 -6.28021359e-01 -3.43777269e-01
3.25571090e-01 -7.74279833e-01 4.20220017e-01 -1.51313436e+00
-4.61524427e-02 -3.57441217e-01 2.73643613e-01 7.84606576e-01
3.92685801e-01 -2.26401508e-01 9.14164037e-02 4.26469557e-02
-7.47331440e-01 1.86786920e-01 5.94010651e-01 -1.11368042e-03
-3.78404617e-01 8.93311426e-02 -6.37832701e-01 1.03662241e+00
7.08878756e-01 -6.24393702e-01 -2.39566296e-01 -5.35338894e-02
5.64814389e-01 5.61296642e-01 4.72303361e-01 -1.29856694e+00
7.91476548e-01 -4.70629364e-01 -1.00121483e-01 -2.48797938e-01
5.24235845e-01 -7.27932990e-01 3.44545960e-01 4.74790663e-01
-5.75811028e-01 5.20911552e-02 1.91608414e-01 6.09076917e-01
-4.46547717e-01 -2.80501127e-01 3.07597637e-01 -2.88964331e-01
-6.51618540e-01 -1.88481525e-01 -1.17957377e+00 -8.05096775e-02
1.51901376e+00 -8.19238275e-02 -2.79966921e-01 -9.37810123e-01
-6.81418180e-01 5.56458950e-01 2.65286624e-01 2.75205910e-01
4.48653072e-01 -8.23353171e-01 -9.62370753e-01 1.02317840e-01
2.12829024e-01 -1.04716301e-01 -1.93096176e-01 7.14892328e-01
-4.71345544e-01 9.53272358e-02 -3.01919758e-01 -3.64979386e-01
-1.24466050e+00 1.14714675e-01 3.99870098e-01 -2.01565593e-01
-6.68878138e-01 5.91400266e-01 -3.58989000e-01 -5.23868322e-01
4.29136068e-01 -1.42403409e-01 -4.05186176e-01 1.92438751e-01
8.30869555e-01 2.44678527e-01 7.70891085e-02 -5.28346717e-01
-3.79159421e-01 -9.91414040e-02 2.03574121e-01 -5.14176846e-01
1.52432227e+00 1.42574638e-01 -2.29492247e-01 8.53692949e-01
2.49568537e-01 -1.93920776e-01 -8.11239004e-01 -2.07599252e-01
2.14541048e-01 -3.51227587e-03 3.30176353e-02 -7.41953015e-01
-3.90458345e-01 3.52687061e-01 7.58642256e-02 7.02669442e-01
8.24742317e-01 7.57442415e-02 6.25840187e-01 1.05094504e+00
9.87811565e-01 -1.26124251e+00 7.68681243e-02 7.10756958e-01
8.67708981e-01 -1.18690538e+00 5.12388535e-02 -2.08861485e-01
-1.05137169e+00 1.06161606e+00 4.14991558e-01 8.32632482e-02
5.04426777e-01 5.84529459e-01 1.34211928e-01 -6.68484211e-01
-1.52126074e+00 -3.04532558e-01 -3.40684056e-01 4.81051236e-01
1.07900485e-01 1.67370290e-01 -6.69156760e-02 6.27482295e-01
-5.38309455e-01 2.91861705e-02 9.07723248e-01 1.46146369e+00
-8.17830920e-01 -8.91478419e-01 -5.29988348e-01 5.73852658e-01
-1.18862607e-01 2.66353935e-01 -6.62211895e-01 6.81397021e-01
1.85923561e-01 1.59366298e+00 2.18339592e-01 -6.03092134e-01
5.67238629e-01 2.07065389e-01 1.53308466e-01 -1.15491390e+00
-1.16697228e+00 -5.27624309e-01 9.13142085e-01 -7.80607581e-01
-5.45184493e-01 -1.09714961e+00 -1.59655023e+00 -2.99580753e-01
7.58082643e-02 4.99974757e-01 7.21881807e-01 1.16012323e+00
2.48559758e-01 6.58590972e-01 6.45716488e-01 -9.39843416e-01
-3.68402600e-01 -9.44765568e-01 -3.76937926e-01 3.97543997e-01
-1.79569311e-02 -7.41349995e-01 -2.36312881e-01 1.92262828e-01] | [4.093210697174072, 1.4680894613265991] |
fec17c6c-387e-49f9-8fb3-a0e7430d7c1c | feddwa-personalized-federated-learning-with | 2305.06124 | null | https://arxiv.org/abs/2305.06124v2 | https://arxiv.org/pdf/2305.06124v2.pdf | FedDWA: Personalized Federated Learning with Online Weight Adjustment | Different from conventional federated learning, personalized federated learning (PFL) is able to train a customized model for each individual client according to its unique requirement. The mainstream approach is to adopt a kind of weighted aggregation method to generate personalized models, in which weights are determined by the loss value or model parameters among different clients. However, such kinds of methods require clients to download others' models. It not only sheer increases communication traffic but also potentially infringes data privacy. In this paper, we propose a new PFL algorithm called \emph{FedDWA (Federated Learning with Dynamic Weight Adjustment)} to address the above problem, which leverages the parameter server (PS) to compute personalized aggregation weights based on collected models from clients. In this way, FedDWA can capture similarities between clients with much less communication overhead. More specifically, we formulate the PFL problem as an optimization problem by minimizing the distance between personalized models and guidance models, so as to customize aggregation weights for each client. Guidance models are obtained by the local one-step ahead adaptation on individual clients. Finally, we conduct extensive experiments using five real datasets and the results demonstrate that FedDWA can significantly reduce the communication traffic and achieve much higher model accuracy than the state-of-the-art approaches. | ['Di wu', 'Yipeng Zhou', 'Miao Hu', 'Jinyu Chen', 'Jiang Wu', 'Jiahao Liu'] | 2023-05-10 | null | null | null | null | ['personalized-federated-learning'] | ['methodology'] | [-3.04214478e-01 -1.56629264e-01 -2.41255626e-01 -7.89084315e-01
-7.44327486e-01 -4.74690169e-01 2.47634858e-01 -2.35195965e-01
-1.07497722e-01 3.63121033e-01 7.24812448e-02 -1.65698588e-01
-3.00773621e-01 -9.89604533e-01 -5.07937133e-01 -8.83374870e-01
1.64160758e-01 4.78220999e-01 1.61932573e-01 1.78984791e-01
3.25004831e-02 3.76701176e-01 -1.19168556e+00 4.67749685e-01
9.89921927e-01 1.30303550e+00 8.96099806e-02 6.38815165e-02
-5.55464387e-01 5.32015145e-01 -3.02427500e-01 -9.75715518e-01
4.44099098e-01 -2.11204574e-01 -6.76012218e-01 9.16531682e-03
1.74205288e-01 -6.37331486e-01 -4.48367596e-01 9.96075928e-01
5.19331455e-01 -1.09364744e-02 1.12509273e-01 -1.37635648e+00
-5.98127246e-01 1.04486668e+00 -3.39832485e-01 -1.24537744e-01
-7.11049959e-02 3.20884824e-01 9.61686850e-01 -6.49984062e-01
2.60449350e-01 1.22437608e+00 5.82922041e-01 7.02490330e-01
-1.25218534e+00 -9.59522069e-01 5.50718427e-01 3.72608870e-01
-1.17924190e+00 -4.31147844e-01 7.30337501e-01 -6.02797717e-02
2.27096796e-01 6.24356091e-01 3.49169523e-01 8.40146124e-01
-2.40131393e-01 8.98276985e-01 8.81695569e-01 -6.53244928e-02
4.06914979e-01 4.64235723e-01 1.79558650e-01 5.53471029e-01
2.33796135e-01 -1.50818139e-01 -4.27537829e-01 -7.59224296e-01
2.37951919e-01 7.26875424e-01 -4.23513919e-01 -6.71772718e-01
-8.06786239e-01 7.49945164e-01 3.08878571e-01 -6.01045601e-02
-4.21389818e-01 -3.29219773e-02 4.12524134e-01 4.18970406e-01
3.80633950e-01 -3.25312138e-01 -7.81013012e-01 1.64312288e-01
-4.51559305e-01 2.08082229e-01 8.46694052e-01 8.89650464e-01
1.39370489e+00 -2.24915087e-01 -2.39744782e-01 8.11552405e-01
5.59156656e-01 1.59859732e-01 8.15086365e-01 -9.59869623e-01
7.08823383e-01 9.50397551e-01 8.42173956e-03 -9.81100202e-01
2.33953580e-01 -4.51049745e-01 -6.46615028e-01 -1.34514406e-01
6.64113835e-02 -4.23086911e-01 -1.81867823e-01 1.80296695e+00
7.91120350e-01 3.38026285e-01 7.02804327e-02 8.65107000e-01
2.97843993e-01 5.92011690e-01 3.19236331e-02 -4.25765872e-01
9.07233119e-01 -1.09354675e+00 -4.26915497e-01 2.32627288e-01
6.75546706e-01 -5.92167556e-01 9.68380749e-01 3.66955012e-01
-8.31282675e-01 5.40479235e-02 -5.84689200e-01 4.58465278e-01
-1.35358840e-01 -2.41762236e-01 6.91556394e-01 7.30997443e-01
-6.37340724e-01 6.83629394e-01 -6.14787877e-01 -3.11987042e-01
7.40640700e-01 5.87064207e-01 -2.24101350e-01 -1.70984656e-01
-8.74177456e-01 2.00270757e-01 1.40000656e-01 -2.94708192e-01
-9.05664921e-01 -9.71833050e-01 -4.25538450e-01 3.76097798e-01
4.90243077e-01 -7.33809412e-01 1.55859280e+00 -8.24660718e-01
-1.48805094e+00 4.30461437e-01 -7.78704882e-02 -4.57022011e-01
8.02565575e-01 4.84303236e-02 -5.02896369e-01 -2.92636871e-01
-3.12607884e-01 -2.56508231e-01 7.07450628e-01 -1.49792230e+00
-1.06799936e+00 -6.40718520e-01 6.93688616e-02 -1.36620313e-01
-1.10253024e+00 1.06579997e-01 -7.26131082e-01 -4.20272619e-01
-3.23181190e-02 -6.16825879e-01 -3.17599297e-01 2.45593667e-01
-1.57791525e-01 -4.28955615e-01 1.29593945e+00 -4.58502769e-01
1.48739207e+00 -2.05080843e+00 -3.42077315e-01 4.85406727e-01
4.06633854e-01 3.52624297e-01 -2.32423812e-01 5.73172629e-01
4.29916948e-01 4.36436497e-02 -1.79396227e-01 -4.85393286e-01
2.90207744e-01 3.94130886e-01 -3.70112509e-01 3.03710580e-01
-5.91292620e-01 6.16525769e-01 -6.47924244e-01 -7.39859462e-01
-6.81752264e-02 4.30810124e-01 -6.59881771e-01 5.34433544e-01
-4.62603122e-01 8.55561197e-02 -9.82207775e-01 5.58058619e-01
1.07414150e+00 -3.27330709e-01 5.08602440e-01 -2.43589729e-01
4.15228568e-02 8.98004696e-03 -1.28711057e+00 1.43076801e+00
-7.32095897e-01 -2.14894399e-01 4.71729279e-01 -8.99125218e-01
1.07467842e+00 2.96135426e-01 7.39774525e-01 -3.90313148e-01
1.25181049e-01 2.73234785e-01 -5.10508060e-01 -4.77819830e-01
-2.93289453e-01 2.25477472e-01 1.50324285e-01 1.17428541e+00
-1.56180322e-01 7.74740219e-01 -2.72141904e-01 2.69836187e-01
1.03267097e+00 -2.28536174e-01 -2.30946943e-01 -2.12885272e-02
7.45644867e-01 -3.26943666e-01 1.06567049e+00 5.11124074e-01
-8.04357678e-02 1.51615188e-01 1.44572994e-02 -8.34615409e-01
-4.95693386e-01 -8.28349173e-01 2.36032560e-01 1.46332347e+00
3.06903899e-01 -4.45423007e-01 -8.28303397e-01 -1.19121873e+00
4.38972473e-01 6.38839960e-01 -3.14365387e-01 -2.99251139e-01
-6.23952866e-01 -6.18041098e-01 1.80097103e-01 9.21618715e-02
6.55707240e-01 -7.70359159e-01 -1.73266783e-01 4.17312324e-01
-1.35643765e-01 -5.56864917e-01 -1.05001807e+00 -1.85943455e-01
-8.64182115e-01 -1.08760905e+00 -4.68178272e-01 -6.30826831e-01
6.95544004e-01 7.51577675e-01 7.96010673e-01 2.34696522e-01
-4.17452073e-03 3.58641922e-01 -2.81823158e-01 -2.08325699e-01
-1.18967474e-01 1.84111148e-01 7.07192644e-02 9.83313560e-01
4.62087750e-01 -1.04265320e+00 -8.67066622e-01 6.93548203e-01
-1.03524065e+00 -1.51788220e-01 4.48235482e-01 7.09055722e-01
6.22513533e-01 -2.35176831e-03 7.52490640e-01 -1.26428246e+00
7.45843887e-01 -8.73007655e-01 -5.74435174e-01 7.87423849e-01
-1.22536778e+00 -6.12043217e-03 1.21728694e+00 -5.03003180e-01
-1.42031085e+00 9.56067666e-02 3.84216487e-01 -6.99973464e-01
2.40584016e-01 1.34038240e-01 -7.29162157e-01 -2.68730134e-01
3.92737389e-01 4.49041665e-01 2.12023810e-01 -1.03661966e+00
4.31746155e-01 1.12690318e+00 3.59007090e-01 -8.02332580e-01
8.82972717e-01 4.14388567e-01 -3.31602305e-01 2.42426947e-01
-6.38223171e-01 -2.11003274e-01 -2.42707264e-02 -1.00522816e-01
1.52661949e-01 -6.17357135e-01 -1.11914718e+00 3.07640702e-01
-9.42602098e-01 -1.49125263e-01 -2.52012163e-01 3.76054496e-01
-3.38920355e-01 4.84873772e-01 -4.78243083e-01 -6.84333682e-01
-1.00292468e+00 -9.18766320e-01 2.64213920e-01 4.89946872e-01
2.78279513e-01 -8.18109453e-01 -1.04097473e-02 4.54981089e-01
8.08079779e-01 3.42412218e-02 9.48261201e-01 -9.03386712e-01
-6.95149899e-01 -3.31213862e-01 -2.16655284e-01 2.21446902e-01
3.62191498e-01 -1.78289473e-01 -8.99639070e-01 -4.79276538e-01
-1.50437066e-02 -1.55885145e-01 3.09169829e-01 -3.89687210e-01
1.67892313e+00 -1.21290040e+00 -3.62533629e-01 8.45215023e-01
1.50236571e+00 1.12444803e-01 5.30668348e-02 1.75138384e-01
6.19726181e-01 4.95114028e-01 4.15831923e-01 1.03848648e+00
5.66480637e-01 6.39770567e-01 7.88224339e-01 3.82000357e-01
2.44676530e-01 -5.21527886e-01 1.94234073e-01 7.89273441e-01
3.58990669e-01 -7.83997253e-02 -4.69548583e-01 4.40112591e-01
-2.26032329e+00 -8.77525449e-01 1.25445321e-01 2.25329709e+00
8.15142989e-01 -4.98681873e-01 1.88872412e-01 -2.50287592e-01
8.72594118e-01 -1.86751073e-03 -1.09932888e+00 -4.40274805e-01
1.25244617e-01 -1.16943747e-01 5.20005286e-01 -2.07143594e-02
-5.10350764e-01 4.94226009e-01 4.92985487e+00 9.56835270e-01
-1.05224562e+00 4.52591747e-01 6.66723847e-01 -3.30925375e-01
-6.83491051e-01 1.34336293e-01 -6.72504246e-01 9.86646771e-01
6.36093616e-01 -8.39319289e-01 9.92938459e-01 1.32765687e+00
1.92900255e-01 6.88813567e-01 -1.06084919e+00 1.00107002e+00
-2.95074135e-01 -1.37952662e+00 3.27957809e-01 2.05262125e-01
8.48365009e-01 8.92942697e-02 4.21446376e-02 2.61105239e-01
7.75688767e-01 -4.18763667e-01 2.86200076e-01 7.36544311e-01
3.37393254e-01 -1.01212609e+00 4.03961569e-01 6.38670564e-01
-1.15839183e+00 -5.40284872e-01 -6.06189668e-01 5.31834185e-01
-8.59231204e-02 4.71608698e-01 -3.86747003e-01 6.72237217e-01
8.88395548e-01 2.79693842e-01 -3.41185093e-01 1.30730450e+00
4.28635120e-01 6.49690926e-01 -4.35915947e-01 7.40289390e-02
3.18505913e-02 -3.86802971e-01 1.65228680e-01 9.50963616e-01
4.23455387e-01 2.23971173e-01 1.76792920e-01 6.44701302e-01
-5.21327615e-01 5.57791352e-01 -1.07933745e-01 5.14295340e-01
9.79801178e-01 1.49310207e+00 2.26939157e-01 -2.40237370e-01
-5.36944091e-01 8.12663138e-01 6.20741904e-01 2.08555281e-01
-5.75723529e-01 -3.93991679e-01 1.07879925e+00 1.57079279e-01
2.97095835e-01 3.48828495e-01 6.44653291e-02 -1.14401269e+00
3.00941110e-01 -9.07722950e-01 9.25913870e-01 -1.81449071e-01
-1.80017602e+00 4.85655427e-01 -4.44776684e-01 -1.18798864e+00
5.01531027e-02 1.48458481e-01 -1.21734631e+00 9.20333564e-01
-1.41508973e+00 -1.15543985e+00 -6.07932031e-01 1.17760992e+00
2.30810806e-01 -2.61196554e-01 8.21500778e-01 5.39684594e-01
-8.46115351e-01 1.24373174e+00 5.48502088e-01 -1.66425094e-01
7.09137321e-01 -5.86633742e-01 7.07026497e-02 4.11209613e-01
-1.03800781e-01 7.13074744e-01 3.06662560e-01 -1.95329398e-01
-1.78993022e+00 -1.54370975e+00 7.34861851e-01 1.46006048e-01
4.75907534e-01 -3.55111472e-02 -1.14090335e+00 7.72111595e-01
-2.06164140e-02 5.36033392e-01 9.42273140e-01 -1.63351491e-01
-7.76743591e-01 -1.01710534e+00 -1.82623971e+00 4.31861103e-01
1.12133431e+00 -1.77076578e-01 1.40294209e-01 5.38995922e-01
1.08771431e+00 1.52854800e-01 -1.00104928e+00 1.51249886e-01
5.08015811e-01 -1.14952326e+00 6.29043639e-01 -9.55615163e-01
-2.47810975e-01 -1.42780945e-01 -3.47179204e-01 -1.14554417e+00
-3.73296887e-01 -1.05296063e+00 -6.93367958e-01 1.83776546e+00
1.08541995e-01 -1.25180566e+00 1.11436379e+00 1.35091996e+00
1.72363043e-01 -1.17954481e+00 -8.97678077e-01 -8.10233712e-01
-1.94935456e-01 -2.39344046e-01 1.77988803e+00 9.73356128e-01
-9.64151248e-02 -3.37756723e-01 -3.33124250e-01 1.46514982e-01
1.11457193e+00 5.32224238e-01 1.03437495e+00 -1.24269783e+00
-4.75309879e-01 -3.23250711e-01 -4.86337133e-02 -8.18827510e-01
2.32978672e-01 -9.08965945e-01 -4.66486990e-01 -1.29741967e+00
1.96671262e-01 -1.07550871e+00 -7.66154468e-01 7.98788548e-01
5.22769988e-02 -2.54548669e-01 3.78300428e-01 6.23421729e-01
-8.55619371e-01 5.65830410e-01 8.66804063e-01 -1.26663342e-01
-1.66129589e-01 4.95477229e-01 -1.10017908e+00 4.24143046e-01
1.02580178e+00 -5.20811796e-01 -4.66349810e-01 -6.68052614e-01
-2.97217876e-01 3.13197641e-04 1.60438478e-01 -7.60385394e-01
4.39517140e-01 -6.32288635e-01 -2.06493922e-02 -3.40166956e-01
-1.52519181e-01 -1.36695862e+00 5.57828903e-01 4.33119893e-01
-3.02888662e-01 -1.50256559e-01 -4.68728125e-01 6.91308677e-01
9.64028835e-02 -5.87615445e-02 6.77273870e-01 -7.21463859e-02
-3.11115295e-01 1.00555170e+00 1.43219933e-01 -2.10569054e-01
1.25078237e+00 8.80862325e-02 -3.73074353e-01 -3.35973203e-01
-4.55577075e-01 7.01595008e-01 4.73046720e-01 2.20744640e-01
3.28697890e-01 -1.53222239e+00 -6.02094889e-01 1.87182739e-01
1.32597461e-01 3.19278799e-02 5.03991663e-01 4.54826504e-01
-2.41702776e-02 1.82957664e-01 1.56511098e-01 -1.72530979e-01
-1.26250827e+00 7.36154079e-01 4.53913927e-01 -2.17622653e-01
-5.44122159e-01 6.59066379e-01 7.82263428e-02 -8.75517845e-01
5.93931139e-01 3.90101969e-01 2.30095461e-01 -1.53722942e-01
6.84063017e-01 7.58811772e-01 1.02188684e-01 -3.77057701e-01
-2.14024156e-01 2.80563712e-01 -4.76766020e-01 4.02848184e-01
1.55049455e+00 -3.28981012e-01 -1.48007318e-01 -1.96247369e-01
1.53576040e+00 -1.39558427e-02 -1.55913317e+00 -9.41983163e-01
-3.17987233e-01 -9.56706643e-01 -8.76919106e-02 -7.84742296e-01
-1.81515217e+00 4.38060015e-01 7.14906752e-01 1.18375927e-01
1.23928154e+00 -1.53103590e-01 1.37573683e+00 3.07744026e-01
8.46569717e-01 -9.13225710e-01 -2.98667192e-01 -1.59499541e-01
4.26467836e-01 -9.41279650e-01 -2.28259951e-01 -1.20133162e-01
-5.44526935e-01 9.94477272e-01 8.70251954e-01 1.60211354e-01
6.37191534e-01 -1.36428386e-01 1.70713425e-01 1.49676397e-01
-1.05848527e+00 4.52666402e-01 -3.07271868e-01 4.59412426e-01
-2.85868168e-01 1.46755815e-01 -3.56825203e-01 1.24399459e+00
1.43757135e-01 1.21007673e-01 -8.57510492e-02 8.21086228e-01
-3.64937305e-01 -1.52596784e+00 -3.51126969e-01 7.34442532e-01
-4.33694214e-01 4.42723006e-01 -2.08351672e-01 3.75413336e-03
1.58837855e-01 1.07781184e+00 -1.67978600e-01 -6.62742853e-01
3.34856629e-01 1.12921745e-01 -2.28148013e-01 -2.68751413e-01
-1.00643277e+00 -2.31858656e-01 -3.65836203e-01 -7.72252858e-01
9.47946012e-02 -5.44994175e-01 -1.20651245e+00 -7.93821454e-01
-3.90411556e-01 5.10948420e-01 6.06316924e-01 5.36793292e-01
8.70321453e-01 -9.37799066e-02 1.65941048e+00 -3.64271194e-01
-1.21500063e+00 -5.04835248e-01 -6.34246349e-01 3.97550434e-01
1.89859699e-02 -2.06038460e-01 -4.03079391e-01 -2.19840080e-01] | [5.837118148803711, 6.319024085998535] |
47ff2e79-4ee9-4d36-8520-5d6a7a933aff | hierarchical-dialogue-understanding-with-1 | 2305.00262 | null | https://arxiv.org/abs/2305.00262v1 | https://arxiv.org/pdf/2305.00262v1.pdf | Hierarchical Dialogue Understanding with Special Tokens and Turn-level Attention | Compared with standard text, understanding dialogue is more challenging for machines as the dynamic and unexpected semantic changes in each turn. To model such inconsistent semantics, we propose a simple but effective Hierarchical Dialogue Understanding model, HiDialog. Specifically, we first insert multiple special tokens into a dialogue and propose the turn-level attention to learn turn embeddings hierarchically. Then, a heterogeneous graph module is leveraged to polish the learned embeddings. We evaluate our model on various dialogue understanding tasks including dialogue relation extraction, dialogue emotion recognition, and dialogue act classification. Results show that our simple approach achieves state-of-the-art performance on all three tasks above. All our source code is publicly available at https://github.com/ShawX825/HiDialog. | ['Yang You', 'Fuzhao Xue', 'Heng Zhang', 'Jian Zhang', 'Xiao Liu'] | 2023-04-29 | hierarchical-dialogue-understanding-with | https://openreview.net/forum?id=Peb3QdR8zzP | https://openreview.net/pdf?id=Peb3QdR8zzP | tiny-papers-iclr-2023-5 | ['dialogue-understanding', 'dialogue-act-classification'] | ['natural-language-processing', 'natural-language-processing'] | [-2.09244207e-01 7.40240097e-01 -3.16588223e-01 -6.67362571e-01
-3.85457516e-01 -5.98908663e-01 9.12018239e-01 2.60071635e-01
-2.38180071e-01 7.24029124e-01 8.11625421e-01 -2.53808439e-01
4.68707621e-01 -5.30566692e-01 -7.71058723e-02 -1.99569892e-02
1.38954446e-01 7.16783583e-01 1.21768015e-02 -6.98600769e-01
1.01795934e-01 -1.52448654e-01 -7.14823246e-01 3.51768494e-01
1.05442524e+00 7.15200424e-01 -2.46471107e-01 9.51768994e-01
-4.82950389e-01 1.31858742e+00 -5.04311860e-01 -6.41265869e-01
-2.98617512e-01 -7.21684575e-01 -1.58787060e+00 1.63584635e-01
-2.21019700e-01 -6.22557998e-01 -8.21812809e-01 7.11224318e-01
4.00320560e-01 3.51273328e-01 4.91357297e-01 -1.35780370e+00
-6.27674878e-01 1.04521513e+00 -9.84254181e-02 3.42738032e-02
8.71444046e-01 3.17617208e-01 1.37239707e+00 -8.91153336e-01
5.59170127e-01 1.48802662e+00 3.27625066e-01 8.48827600e-01
-1.09803379e+00 -3.67431968e-01 3.37804079e-01 3.11922163e-01
-7.02016175e-01 -7.47893393e-01 1.00286210e+00 -6.82090968e-02
1.16815376e+00 1.91099197e-01 7.52673805e-01 1.30314827e+00
-9.31984410e-02 1.26367903e+00 8.87928367e-01 -2.92795807e-01
-8.07738081e-02 1.46050349e-01 8.87422621e-01 8.99498880e-01
-4.80353773e-01 -5.61771274e-01 -6.13515496e-01 -1.34344697e-01
2.29384795e-01 -1.72448903e-01 -3.42648000e-01 -1.44860655e-01
-8.00790071e-01 1.16926169e+00 3.57745260e-01 1.90293640e-01
-1.31183058e-01 -5.01795933e-02 9.02752936e-01 5.65895379e-01
7.48822212e-01 6.37444735e-01 -5.46110570e-01 -8.44933093e-01
-6.57179346e-03 1.68171614e-01 1.54691827e+00 7.61551857e-01
7.04004586e-01 -2.68481880e-01 -1.49117902e-01 1.25887537e+00
3.36759537e-01 -2.11750001e-01 3.46031815e-01 -1.03799677e+00
5.37991941e-01 1.00108624e+00 -2.41900444e-01 -9.17475581e-01
-6.60347283e-01 2.14886278e-01 -7.07526803e-01 -5.75656950e-01
2.95136273e-01 -6.08166754e-01 -4.50592101e-01 1.74627745e+00
6.02642000e-01 -1.54516578e-01 3.08087587e-01 7.34828532e-01
1.33665287e+00 7.57163286e-01 8.29898790e-02 5.61541431e-02
1.53327417e+00 -1.46635211e+00 -1.14300442e+00 -3.95132512e-01
1.07429433e+00 -4.28394347e-01 1.17773497e+00 -8.82930383e-02
-8.39637101e-01 -1.25726789e-01 -7.77007103e-01 -6.35458469e-01
-2.94601887e-01 -2.15056613e-01 8.08788300e-01 2.02277288e-01
-8.54922175e-01 1.88026935e-01 -6.75837994e-01 -5.34145594e-01
2.81302214e-01 6.52852654e-02 -2.34221220e-01 -7.54387602e-02
-1.73746526e+00 1.04482663e+00 2.93654680e-01 1.65396169e-01
-6.64051116e-01 -5.23650885e-01 -1.42920876e+00 3.23703848e-02
5.13469815e-01 -6.14565849e-01 1.66517127e+00 -6.68178201e-01
-2.24526429e+00 8.62893641e-01 -3.34329605e-01 -4.03373420e-01
3.06305528e-01 -4.17164207e-01 -1.29834503e-01 6.39043078e-02
-1.87881440e-01 4.28434134e-01 1.63907424e-01 -1.05432594e+00
-2.20619515e-01 -3.88629586e-01 6.30798161e-01 6.43790781e-01
-3.21094781e-01 -7.01404959e-02 -4.51211244e-01 2.58206688e-02
-1.79306641e-01 -9.59541261e-01 -1.17615387e-01 -5.37226200e-01
-8.73011887e-01 -8.27097893e-01 8.11491370e-01 -7.56095469e-01
1.41203845e+00 -1.95079052e+00 3.87359321e-01 -3.54597121e-01
6.01788938e-01 1.49632588e-01 -1.72838122e-01 8.84703040e-01
2.28244707e-01 1.45519137e-01 -1.23301536e-01 -5.45268178e-01
3.93959582e-01 1.49930432e-01 -7.75585920e-02 1.25550717e-01
2.22113639e-01 1.19946873e+00 -8.98509324e-01 -2.65493512e-01
3.35258812e-01 2.55711794e-01 -5.40214181e-01 6.36197925e-01
-4.48611647e-01 4.32171077e-01 -7.33238697e-01 2.52974898e-01
2.88521647e-01 -4.19332474e-01 6.30390406e-01 -1.20774627e-01
2.99906909e-01 1.00286174e+00 -5.29229939e-01 1.88224614e+00
-6.86174989e-01 8.22034001e-01 3.40074487e-02 -8.93648386e-01
7.65377760e-01 3.24178666e-01 9.06606689e-02 -4.86208886e-01
3.47192258e-01 -3.74119163e-01 1.53817078e-02 -5.82699001e-01
8.38436544e-01 3.14530358e-02 -5.78655958e-01 7.43568420e-01
3.22934836e-01 -3.44853967e-01 -4.65134569e-02 8.56094539e-01
1.15541458e+00 -2.33669043e-01 5.61031103e-01 3.62233780e-02
4.97622937e-01 -4.60165516e-02 4.67384428e-01 4.34650600e-01
-4.38911885e-01 1.49281407e-02 1.19600511e+00 -3.42632771e-01
-6.27633274e-01 -6.36944473e-01 2.78144121e-01 1.51572120e+00
2.19607547e-01 -8.24932039e-01 -7.61310279e-01 -1.07932758e+00
8.01381245e-02 8.79006743e-01 -5.98388374e-01 -1.87502772e-01
-4.46588308e-01 -3.98200393e-01 7.20001042e-01 2.19707385e-01
6.97084486e-01 -1.07636356e+00 4.43394855e-03 1.37579679e-01
-5.41000962e-01 -1.21201468e+00 -5.53935230e-01 -5.01736114e-03
-4.24638629e-01 -1.18129420e+00 -2.33327802e-02 -5.57606161e-01
2.89613366e-01 5.22270501e-02 1.26164103e+00 2.00879410e-01
-1.01975009e-01 5.83124757e-01 -6.81587100e-01 -6.91259652e-02
-4.99807358e-01 5.38718343e-01 -3.10191780e-01 -1.34971842e-01
6.07556701e-01 -2.62933731e-01 -4.44653600e-01 -1.33414075e-01
-5.19351184e-01 2.87904412e-01 -1.63728580e-01 1.03928423e+00
-2.64330745e-01 -5.85561454e-01 8.06234479e-01 -1.27278590e+00
1.35608613e+00 -6.24722600e-01 -4.48612645e-02 2.00581372e-01
-4.32634681e-01 1.92021951e-02 6.34013593e-01 -9.91352648e-03
-1.17062449e+00 -3.35220933e-01 -2.83890575e-01 1.10158376e-01
-2.94522762e-01 5.19581318e-01 -3.25975209e-01 4.55855012e-01
1.90068305e-01 7.13344365e-02 2.19975322e-01 -3.06000054e-01
8.62996757e-01 9.77237105e-01 4.62353379e-02 -5.82006395e-01
3.89479548e-01 5.25044696e-03 -7.71015704e-01 -9.99507070e-01
-1.13458228e+00 -4.78414357e-01 -5.93775749e-01 -2.71229237e-01
1.02908635e+00 -7.76855588e-01 -1.08025324e+00 5.45507610e-01
-1.41100228e+00 -8.99394810e-01 -6.91419048e-03 1.95672706e-01
-4.32153404e-01 4.32137072e-01 -1.27116370e+00 -7.69233763e-01
-5.10996461e-01 -8.78027737e-01 6.88126147e-01 4.65439141e-01
-7.25624502e-01 -1.55492949e+00 1.71890944e-01 8.72304320e-01
2.36554235e-01 -1.15801625e-01 9.30776536e-01 -1.31213725e+00
-1.71223521e-01 3.98088917e-02 -2.21242353e-01 2.36567259e-01
3.38995904e-01 -1.03367582e-01 -9.75677907e-01 3.25572118e-02
-1.27597421e-01 -1.08315086e+00 7.30755866e-01 -1.22620650e-01
9.43292797e-01 -5.58607280e-01 -1.67172298e-01 2.45291084e-01
7.50523150e-01 1.00398086e-01 3.99362564e-01 1.32987201e-01
8.34571600e-01 8.36386204e-01 6.85770392e-01 7.60328054e-01
1.29774153e+00 5.39639115e-01 1.13091610e-01 -5.09176515e-02
1.25738263e-01 -3.85929614e-01 4.56067026e-01 1.27356291e+00
3.73463452e-01 -4.86903697e-01 -9.38202083e-01 4.92217422e-01
-1.98616445e+00 -8.58152151e-01 1.54877920e-02 1.36273742e+00
1.27500474e+00 -2.39842404e-02 1.12152681e-01 -3.90843362e-01
5.27465165e-01 8.63426149e-01 -6.02490902e-01 -8.81006539e-01
1.95979178e-01 -7.47693926e-02 -1.73418984e-01 1.17822433e+00
-9.40537095e-01 1.53773057e+00 5.53923178e+00 3.89978528e-01
-7.95331597e-01 1.39848918e-01 7.55179346e-01 8.59555900e-02
-4.78689581e-01 1.47902355e-01 -6.30603492e-01 2.16029823e-01
8.97404671e-01 -4.29491848e-01 3.72612834e-01 6.06602788e-01
1.48034692e-01 -8.28469321e-02 -1.17703962e+00 6.14433289e-01
1.60185322e-01 -1.22337711e+00 -2.56564356e-02 -2.64090747e-01
2.62228787e-01 -4.36252281e-02 -3.07059914e-01 8.34388077e-01
8.83002579e-01 -1.06485176e+00 -1.55715123e-01 2.23365933e-01
3.89798731e-01 -7.54058182e-01 7.06699371e-01 2.61565030e-01
-9.99307990e-01 2.16245323e-01 7.20220804e-02 -2.98228890e-01
3.42229396e-01 4.08106029e-01 -1.14358974e+00 5.50795138e-01
1.40706420e-01 9.15609002e-01 -2.83360571e-01 1.41644612e-01
-6.13538742e-01 6.25829518e-01 -1.12042770e-01 -3.24376494e-01
3.17873269e-01 -3.12137008e-01 4.01263744e-01 1.38207448e+00
-4.27276403e-01 4.84550327e-01 4.18295294e-01 5.66690624e-01
-6.13810539e-01 8.26540515e-02 -7.66407549e-01 -3.53016376e-01
7.11444199e-01 1.48547614e+00 -1.46440193e-01 -5.09883881e-01
-4.82597023e-01 1.36099124e+00 8.39491367e-01 3.96546394e-01
-8.05261075e-01 -6.16304815e-01 9.76437330e-01 -4.59222019e-01
-1.03034295e-01 -2.20000148e-01 -5.65004349e-02 -1.43301988e+00
-1.64598495e-01 -9.39289153e-01 5.46431184e-01 -3.87734711e-01
-1.36199617e+00 5.99056542e-01 -3.65570903e-01 -4.31352228e-01
-4.86881226e-01 -3.30962598e-01 -9.26310718e-01 5.36986530e-01
-1.40623999e+00 -1.02863789e+00 -4.82073545e-01 5.03313541e-01
9.87710238e-01 6.84626177e-02 1.06738400e+00 -1.45171449e-01
-9.30436611e-01 6.87650919e-01 -3.15030038e-01 6.58830583e-01
8.75235677e-01 -1.42035210e+00 4.81415868e-01 1.35065645e-01
-1.22637205e-01 4.49431747e-01 5.35553157e-01 -4.75609988e-01
-1.40762222e+00 -8.11708808e-01 1.23700225e+00 -5.71092844e-01
1.03777266e+00 -6.69385672e-01 -1.11307991e+00 1.03021204e+00
9.31952298e-01 -4.61677998e-01 9.92166698e-01 6.86864614e-01
-4.26315784e-01 2.65040100e-01 -8.41354191e-01 7.10898817e-01
9.46158648e-01 -8.99966896e-01 -1.03268707e+00 3.95439208e-01
1.07402170e+00 -6.12861574e-01 -1.16343319e+00 9.27733555e-02
3.22095841e-01 -7.99335539e-01 4.85393465e-01 -1.02567685e+00
7.55195200e-01 4.08899844e-01 1.42568275e-01 -1.72212601e+00
9.28307325e-02 -8.96813333e-01 -2.73836583e-01 1.34497678e+00
6.03956163e-01 -7.54425168e-01 5.98869681e-01 9.83832479e-01
-1.74257681e-01 -9.99047756e-01 -6.38717413e-01 -2.47508690e-01
5.19270375e-02 2.82155965e-02 3.55667621e-01 1.44978380e+00
9.50734735e-01 1.24625134e+00 -3.33879173e-01 -2.16466963e-01
1.60235211e-01 2.14977145e-01 1.07383549e+00 -9.97024655e-01
-1.13949262e-01 -2.48851418e-01 -4.31788946e-03 -1.62864971e+00
9.08938944e-01 -8.41054082e-01 1.18598282e-01 -1.87342322e+00
2.02343673e-01 -2.15111941e-01 -2.05349759e-03 4.89885658e-01
-4.66036409e-01 -3.77247274e-01 2.30382785e-01 -1.58167794e-01
-1.14841628e+00 1.26322150e+00 1.24842072e+00 -1.83584839e-01
-3.99271011e-01 -2.47367561e-01 -7.40402400e-01 7.05028892e-01
1.09620535e+00 -1.65578485e-01 -5.11307895e-01 -3.81691307e-01
-1.86166689e-01 3.76001000e-01 1.05528131e-01 -2.73280174e-01
3.28013629e-01 -7.70703852e-02 -2.12734878e-01 -2.13440433e-01
6.51330948e-01 -2.23780230e-01 -6.89015925e-01 1.56682789e-01
-9.19605255e-01 -1.36865929e-04 1.12139918e-01 3.89345050e-01
-5.14334440e-01 -7.04643503e-02 4.17676747e-01 -5.34996353e-02
-6.45428419e-01 2.42239594e-01 -5.08935928e-01 6.41421199e-01
8.70352983e-01 3.36318642e-01 -7.52065659e-01 -9.86571968e-01
-8.41382623e-01 8.80680323e-01 3.49636912e-01 7.27897167e-01
5.84940434e-01 -1.09857202e+00 -6.67981505e-01 -2.12921530e-01
2.50946105e-01 1.17028423e-01 3.96643758e-01 8.75899374e-01
-1.73542500e-01 4.17505890e-01 5.15245572e-02 -1.81123123e-01
-1.52409923e+00 2.96190009e-02 4.43487942e-01 -5.35336137e-01
-6.59341812e-01 8.59144866e-01 2.13226706e-01 -1.11574399e+00
2.73208976e-01 -1.24717308e-02 -3.50627989e-01 2.98435003e-01
2.88392663e-01 1.22387417e-01 -5.34663260e-01 -3.88381928e-01
-3.42366248e-01 -9.52070951e-03 -5.21696985e-01 -6.23250902e-02
1.15592909e+00 -4.45078820e-01 -2.83666402e-01 8.38110149e-01
1.35182106e+00 -1.74710155e-01 -9.72220540e-01 -4.07362133e-01
2.39749402e-02 -1.13602258e-01 -3.14505368e-01 -6.64550662e-01
-6.89240754e-01 9.20817375e-01 -1.24769785e-01 5.82118511e-01
5.64624012e-01 2.77533233e-01 9.41314876e-01 7.24873126e-01
-1.47546995e-02 -1.30092144e+00 3.59243900e-01 1.27289987e+00
9.33587432e-01 -1.45858443e+00 -2.16437340e-01 -5.54778397e-01
-1.29619706e+00 1.03675735e+00 1.03005028e+00 1.10877186e-01
5.02824306e-01 6.68559521e-02 3.26583534e-01 -5.05147338e-01
-1.30126524e+00 -1.29149631e-01 -2.42731154e-01 2.31260210e-01
8.61897290e-01 1.42594665e-01 -2.89039403e-01 7.13552594e-01
-2.26918116e-01 -2.62655497e-01 7.12383151e-01 6.54883862e-01
-3.17878336e-01 -1.09688556e+00 4.28430170e-01 3.49524975e-01
-2.04655498e-01 -1.84631571e-01 -1.14909756e+00 8.89577687e-01
-8.36676896e-01 1.22321606e+00 2.42989399e-02 -6.31619573e-01
3.69848132e-01 6.53345764e-01 1.41360343e-01 -8.46284807e-01
-8.50144982e-01 -5.12636244e-01 1.05212379e+00 -7.40491807e-01
-2.01751053e-01 -3.59317601e-01 -1.64491141e+00 -5.92319071e-01
-9.21201333e-02 4.48928982e-01 2.55381942e-01 1.00900912e+00
5.54284573e-01 7.66159713e-01 9.11688149e-01 -3.17576289e-01
-5.04081666e-01 -1.28547001e+00 -2.85184294e-01 6.14236772e-01
2.44987026e-01 -4.77577537e-01 -4.61801082e-01 -3.68772715e-01] | [12.524874687194824, 7.948256969451904] |
00bcd24f-5659-4412-adda-524c5d5efcf2 | eegdenoisenet-a-benchmark-dataset-for-deep | 2009.11662 | null | https://arxiv.org/abs/2009.11662v4 | https://arxiv.org/pdf/2009.11662v4.pdf | EEGdenoiseNet: A benchmark dataset for end-to-end deep learning solutions of EEG denoising | Deep learning networks are increasingly attracting attention in various fields, including electroencephalography (EEG) signal processing. These models provided comparable performance with that of traditional techniques. At present, however, lacks of well-structured and standardized datasets with specific benchmark limit the development of deep learning solutions for EEG denoising. Here, we present EEGdenoiseNet, a benchmark EEG dataset that is suited for training and testing deep learning-based denoising models, as well as for performance comparisons across models. EEGdenoiseNet contains 4514 clean EEG segments, 3400 ocular artifact segments and 5598 muscular artifact segments, allowing users to synthesize contaminated EEG segments with the ground-truth clean EEG. We used EEGdenoiseNet to evaluate denoising performance of four classical networks (a fully-connected network, a simple and a complex convolution network, and a recurrent neural network). Our analysis suggested that deep learning methods have great potential for EEG denoising even under high noise contamination. Through EEGdenoiseNet, we hope to accelerate the development of the emerging field of deep learning-based EEG denoising. | ['Quanying Liu', 'Zherui Li', 'Dante Mantini', 'Chen Wei', 'Mingqi Zhao', 'Haoming Zhang'] | 2020-09-24 | null | null | null | null | ['eeg-denoising'] | ['methodology'] | [-5.95568717e-02 -4.32065606e-01 8.64515364e-01 -4.62656885e-01
-6.76162481e-01 -8.95869508e-02 6.84625804e-02 -1.58058196e-01
-3.83361667e-01 9.27934945e-01 1.64149836e-01 -4.75408211e-02
-2.82365650e-01 -3.63563210e-01 -6.21656895e-01 -9.77825105e-01
-3.86351347e-01 -1.59732237e-01 -4.45423007e-01 -3.15450847e-01
-3.93477678e-02 2.91142136e-01 -1.09667146e+00 4.78189647e-01
1.04542792e+00 1.15235448e+00 1.34212524e-01 3.70738298e-01
4.77569640e-01 4.15612966e-01 -1.30927777e+00 -3.91677856e-01
6.37521967e-02 -6.91308975e-01 -3.58836174e-01 -4.39474016e-01
9.14212540e-02 -1.81540653e-01 -7.09727883e-01 1.28708816e+00
1.17861235e+00 4.55391258e-02 4.36455816e-01 -1.13418210e+00
-7.63258040e-01 6.70809925e-01 -3.43176544e-01 7.33888984e-01
3.62371564e-01 2.96002269e-01 4.34312522e-02 -7.13947833e-01
9.47842076e-02 7.61134386e-01 1.14981842e+00 6.15430474e-01
-1.35166562e+00 -1.18775356e+00 2.66871322e-02 4.96216327e-01
-1.37417161e+00 -4.59489942e-01 6.83116138e-01 -2.82710105e-01
1.06593680e+00 2.68261015e-01 1.01354504e+00 2.10921574e+00
9.70087707e-01 3.99227440e-01 1.38904738e+00 1.34409994e-01
4.19990301e-01 -2.19922736e-01 5.12086570e-01 -3.27658728e-02
3.40383023e-01 1.52844846e-01 -9.08090472e-01 -1.01657465e-01
7.19134510e-01 -7.62790674e-03 -5.77978969e-01 4.54437524e-01
-1.02403152e+00 2.19295010e-01 4.04006541e-01 4.54994768e-01
-9.13316667e-01 1.62991062e-01 6.42738283e-01 8.07088554e-01
8.82649362e-01 7.68745124e-01 -4.91415918e-01 -3.13250631e-01
-1.25446606e+00 2.52303332e-01 8.16880584e-01 9.65962052e-01
2.19382375e-01 6.18703365e-01 -3.08438689e-01 9.07759190e-01
-2.05504701e-01 3.14603686e-01 7.12178290e-01 -7.69464731e-01
3.48670214e-01 5.76871336e-02 -1.91290051e-01 -9.20744598e-01
-6.72898352e-01 -7.63576269e-01 -1.69403839e+00 5.97324073e-02
-9.37066227e-03 -7.06283510e-01 -8.76625121e-01 1.49940109e+00
-7.54003942e-01 6.05859339e-01 -6.30077794e-02 8.87463152e-01
1.01298594e+00 5.28849542e-01 -5.69937080e-02 -3.15427482e-01
1.18939602e+00 -4.85020429e-01 -1.38596201e+00 -2.20401257e-01
1.12185434e-01 -5.91162741e-01 7.56110847e-01 1.31401026e+00
-1.35574543e+00 -5.37784040e-01 -1.21273816e+00 2.19120815e-01
-3.16529244e-01 -1.75292835e-01 6.33582234e-01 7.29240656e-01
-1.32816494e+00 8.82801950e-01 -1.01856458e+00 7.54958540e-02
6.86849117e-01 7.06210911e-01 -4.63070154e-01 7.78331906e-02
-1.40991771e+00 1.10631502e+00 1.10731415e-01 5.63384891e-01
-1.31997311e+00 -7.73614407e-01 -5.48480451e-01 2.06622183e-01
-2.28958756e-01 -7.07146704e-01 9.73015726e-01 -9.88415301e-01
-1.40147877e+00 2.71127909e-01 -2.09384039e-02 -8.05806816e-01
3.39015514e-01 -5.56100786e-01 -8.48241210e-01 -8.53025913e-02
-1.68946788e-01 2.00467408e-01 1.01559651e+00 -9.32627678e-01
-1.00216947e-01 -5.03329098e-01 -4.98960048e-01 -1.93050876e-01
-2.89340943e-01 1.77153975e-01 -1.92315914e-02 -1.05136132e+00
-6.36756711e-04 -3.45576197e-01 -4.71309712e-03 -5.79907477e-01
-4.15340900e-01 -1.21764295e-01 4.46557611e-01 -1.02829540e+00
1.17762816e+00 -2.15436172e+00 1.31588504e-01 2.59222686e-01
5.11523068e-01 3.71152401e-01 -5.16641736e-01 2.71671534e-01
-6.99570596e-01 2.44471598e-02 -1.87909424e-01 -5.14049470e-01
-9.63730067e-02 -8.66495818e-02 -1.44531250e-01 6.47960067e-01
2.56539017e-01 7.74953246e-01 -7.61991799e-01 3.99885446e-01
1.40989900e-01 9.27912831e-01 -2.86096007e-01 4.09174353e-01
5.51611125e-01 6.45467877e-01 9.19188336e-02 4.06331569e-01
1.00439382e+00 4.26597208e-01 -3.92267734e-01 -3.05200607e-01
1.98098831e-02 2.40131989e-01 -9.94535506e-01 1.93113303e+00
-2.87343502e-01 9.44073915e-01 8.30104873e-02 -8.99143338e-01
8.75437498e-01 6.88497066e-01 3.75500351e-01 -9.11782622e-01
4.74307656e-01 2.60949224e-01 2.01366335e-01 -6.61222517e-01
-1.25918642e-01 -5.80830723e-02 3.58851194e-01 2.66926944e-01
6.32669806e-01 -2.20738843e-01 2.70115416e-02 -1.51775733e-01
1.31611311e+00 -2.35748067e-01 -2.70015150e-01 -6.49793208e-01
1.98895007e-01 -7.51794934e-01 5.34838498e-01 8.82922888e-01
-1.71556070e-01 8.88586700e-01 5.78084171e-01 -6.75604701e-01
-6.15793705e-01 -9.20950055e-01 -4.70037788e-01 6.73093379e-01
-1.22727253e-01 -4.70984548e-01 -1.45385182e+00 5.38445152e-02
-2.58016586e-01 7.48095930e-01 -7.35089719e-01 -6.18646860e-01
-2.62167722e-01 -1.39962876e+00 8.22573543e-01 5.20871699e-01
5.87716103e-01 -1.29347670e+00 -2.76043266e-01 4.44057971e-01
-2.08271429e-01 -6.29553139e-01 -2.42457628e-01 8.83736014e-01
-8.02537918e-01 -9.08991098e-01 -9.11408722e-01 -7.30971634e-01
3.28656256e-01 -3.98426093e-02 1.19584155e+00 -1.45980433e-01
-1.07438408e-01 5.46172373e-02 -8.78508538e-02 -1.02262592e+00
-6.85356930e-02 -2.14034747e-02 6.20145202e-01 -1.69300675e-01
8.53312850e-01 -1.11205411e+00 -8.30987632e-01 1.59131721e-01
-8.91502678e-01 -2.68118203e-01 4.46150541e-01 9.41495717e-01
3.50497752e-01 2.39396617e-01 1.13774467e+00 -4.72634166e-01
1.49790692e+00 -5.88584065e-01 -2.39631370e-01 -9.76277664e-02
-5.32543242e-01 -4.27964509e-01 5.68562865e-01 -5.00057220e-01
-9.54186618e-01 -5.95055103e-01 -7.82147586e-01 -3.49271566e-01
-3.80884945e-01 6.32319450e-01 -4.30381782e-02 -1.69432357e-01
8.79739523e-01 3.44496071e-01 -3.27814221e-01 -6.28167808e-01
-3.72792512e-01 7.05120862e-01 8.15727532e-01 -2.87654757e-01
1.87197864e-01 -1.09827563e-01 -3.91799152e-01 -9.44264889e-01
-4.38691169e-01 2.37596594e-03 -5.54849982e-01 -1.68390125e-02
7.64789641e-01 -1.12248051e+00 -6.04299664e-01 9.79289830e-01
-1.43279648e+00 -3.95703763e-01 -6.72438890e-02 7.13966191e-01
-3.34020525e-01 7.42768645e-02 -9.22649622e-01 -5.48448145e-01
-8.37077916e-01 -1.35776806e+00 6.94298089e-01 -6.85914382e-02
-3.43627661e-01 -7.59006560e-01 1.52016357e-01 -1.59858316e-01
5.49036622e-01 3.34997892e-01 7.34997034e-01 -5.80024958e-01
-1.06525375e-02 -3.80797297e-01 4.26734686e-02 1.08358836e+00
1.33416787e-01 -3.91481668e-01 -1.41270232e+00 -5.20210981e-01
8.13777685e-01 -1.63240388e-01 8.55822444e-01 7.69716561e-01
1.68653703e+00 1.21513354e-02 1.79035097e-01 1.25653911e+00
1.15724623e+00 5.23030996e-01 1.09801304e+00 2.87521333e-01
4.18841362e-01 1.49729639e-01 -6.01969779e-01 2.77131677e-01
-4.41029705e-02 3.06076724e-02 3.27372819e-01 -2.58688509e-01
-6.82503432e-02 3.90078068e-01 2.64365733e-01 1.46833861e+00
-5.14530122e-01 -1.78827927e-01 -8.36709857e-01 3.49040031e-01
-1.32245886e+00 -6.62990451e-01 -2.09893048e-01 1.88079226e+00
8.31421375e-01 2.04628929e-02 -1.64823264e-01 4.56081420e-01
4.29233611e-01 -2.88497597e-01 -6.76507771e-01 -3.57611984e-01
-4.44697499e-01 8.67551029e-01 -3.09255533e-03 -2.50267595e-01
-9.34082925e-01 2.91295946e-01 7.27335739e+00 6.43177927e-01
-1.08002079e+00 4.61627990e-01 6.23921812e-01 -1.85207948e-01
2.93523312e-01 -1.01750875e+00 -2.27721855e-01 8.19511294e-01
1.47038138e+00 -1.70878887e-01 8.77024710e-01 3.96785885e-01
5.92618763e-01 1.01696998e-01 -1.29266644e+00 1.71803081e+00
9.73193645e-02 -1.14505577e+00 -1.37662485e-01 -2.93902457e-01
9.38715458e-01 3.91494960e-01 1.09688945e-01 3.45625132e-01
-2.66055346e-01 -1.33640933e+00 4.73097265e-01 8.26131105e-01
7.55686283e-01 -1.03195000e+00 1.41855800e+00 9.68342498e-02
-5.69558680e-01 -2.95461137e-02 -5.14540613e-01 -3.32394898e-01
-1.18617117e-01 7.52035916e-01 1.30136879e-02 6.16110325e-01
1.24740851e+00 1.06417978e+00 -6.81507587e-01 1.42014635e+00
-3.16327363e-01 8.40646625e-01 1.01196319e-01 3.19567829e-01
5.82773834e-02 -3.78402203e-01 4.51775819e-01 1.55399311e+00
4.36788529e-01 1.90712154e-01 -5.77892601e-01 9.43498313e-01
-3.78321737e-01 -3.76820654e-01 -4.65577155e-01 1.99047223e-01
1.61251038e-01 1.05598128e+00 -5.03136277e-01 -1.85701892e-01
-4.19287056e-01 1.05450022e+00 -2.19043419e-02 8.10297251e-01
-7.01051354e-01 -8.37090552e-01 9.92811203e-01 -3.22013825e-01
-4.47880268e-01 3.37596834e-02 -8.54361057e-01 -1.38299024e+00
-1.61301345e-02 -1.31954777e+00 -2.65956581e-01 -1.08079720e+00
-1.77225184e+00 1.27379274e+00 -9.60798562e-02 -7.95171440e-01
1.78782806e-01 -6.21026516e-01 -9.53108013e-01 1.21119440e+00
-1.27183211e+00 -3.14792514e-01 -5.82818091e-01 9.44820285e-01
5.18476069e-01 -3.51588666e-01 1.10274005e+00 6.40159786e-01
-1.01318014e+00 4.91046757e-01 2.99741834e-01 8.98922607e-02
8.53214741e-01 -1.21501839e+00 6.66391551e-01 8.63052309e-01
-2.33580500e-01 9.49496090e-01 8.11367393e-01 -3.64567637e-01
-1.27388167e+00 -1.28118789e+00 3.53985548e-01 -2.36054674e-01
5.08769870e-01 -7.42240429e-01 -1.37774098e+00 5.85376024e-01
9.86404896e-01 -2.78284729e-01 6.98997796e-01 2.52823904e-02
2.04015881e-01 -5.10128260e-01 -1.04386699e+00 4.85140413e-01
9.52975810e-01 -5.11924684e-01 -9.81337309e-01 3.65831226e-01
2.55625010e-01 -4.23302442e-01 -1.04413581e+00 2.42904738e-01
3.19032669e-01 -9.17289615e-01 7.42784619e-01 -3.84805977e-01
2.91492909e-01 1.44924149e-01 3.02938640e-01 -2.17011094e+00
-4.39887673e-01 -1.00485277e+00 -2.08737209e-01 7.95345604e-01
1.33062273e-01 -7.92009890e-01 4.04326618e-01 4.67196524e-01
-6.57180429e-01 -7.75947273e-01 -9.11391497e-01 -7.36244023e-01
1.99814722e-01 -6.85138643e-01 8.30965161e-01 7.06732750e-01
-6.64879158e-02 2.08290190e-01 -3.95013064e-01 1.96200177e-01
6.65678084e-01 -8.43173742e-01 2.31758311e-01 -1.31525588e+00
3.72262955e-01 -4.46293294e-01 -5.46396255e-01 -5.76282918e-01
2.11382672e-01 -7.38354385e-01 8.34989101e-02 -1.60433722e+00
3.97121496e-02 1.86086059e-01 -9.28541481e-01 4.50742066e-01
-3.11755687e-02 4.83890384e-01 -2.92828679e-01 -2.63550937e-01
-1.44741878e-01 6.35921478e-01 1.10046828e+00 -2.92981416e-01
-2.97917455e-01 -2.96863280e-02 -7.54544020e-01 8.68350625e-01
9.61106360e-01 -5.66423357e-01 -4.22843218e-01 -8.57260942e-01
-4.22508921e-03 -1.46506757e-01 3.47300351e-01 -1.42889452e+00
3.89149189e-01 5.98618031e-01 1.13573003e+00 -4.10578579e-01
2.60574162e-01 -7.27261782e-01 3.60490441e-01 2.60641575e-01
-2.04248950e-01 4.43220913e-01 5.74952781e-01 2.30511963e-01
-4.31504339e-01 6.30100369e-02 6.26072466e-01 -4.47315052e-02
-4.45464939e-01 2.85245657e-01 -8.97680044e-01 -9.64743793e-02
6.13320231e-01 -2.62999445e-01 -2.24675193e-01 -2.46729255e-01
-1.10974419e+00 1.01943631e-02 -3.46108079e-01 4.10616636e-01
1.09768343e+00 -1.19593227e+00 -8.50030839e-01 8.63562644e-01
-3.31711948e-01 -2.26941437e-01 3.51112247e-01 1.08708334e+00
-4.08170581e-01 2.52004236e-01 -8.45949292e-01 -4.88439262e-01
-8.91849518e-01 2.95378268e-01 6.32797480e-01 4.46316898e-01
-8.95840883e-01 1.00517452e+00 1.69669613e-01 1.52117212e-03
7.58339763e-01 -7.53232598e-01 -1.82793573e-01 2.87985299e-02
8.69502008e-01 6.92349911e-01 8.76845419e-01 -1.37110287e-02
-2.94829607e-01 -4.87018712e-02 1.91716418e-01 3.72681856e-01
2.01390767e+00 1.27560928e-01 -6.56017065e-01 4.32210743e-01
9.66487348e-01 -6.33660972e-01 -1.22209656e+00 3.70784432e-01
-5.79823069e-02 -4.19827774e-02 4.54082370e-01 -1.23402297e+00
-1.37524796e+00 9.73613918e-01 1.05575013e+00 1.77393138e-01
1.65663135e+00 -7.38758147e-01 7.48568833e-01 5.75216293e-01
6.15850925e-01 -9.03560698e-01 -4.63704020e-02 4.83998388e-01
1.35541630e+00 -8.67781103e-01 -2.98841596e-01 2.03843459e-01
-3.45130265e-01 1.29942298e+00 4.76845235e-01 -6.02623820e-01
8.08871210e-01 7.19176412e-01 8.92355591e-02 -3.58785897e-01
-7.38357782e-01 5.08332551e-01 3.23754787e-01 8.81923854e-01
4.92222518e-01 -8.07308257e-02 -3.13284025e-02 1.38843119e+00
-4.64347392e-01 3.18845898e-01 5.99879801e-01 4.97849315e-01
8.94027576e-02 -6.80886149e-01 -3.84096116e-01 8.78050089e-01
-7.92366147e-01 -5.32681525e-01 -9.53121334e-02 6.67334199e-01
5.15035927e-01 1.15464723e+00 -1.59274548e-01 -5.18032670e-01
8.49000871e-01 2.18020584e-02 5.23567080e-01 -5.40250838e-01
-1.15232420e+00 2.80037582e-01 -2.64048278e-01 -6.11199319e-01
-2.48867020e-01 -3.29842359e-01 -4.46940899e-01 -3.50301713e-01
-2.83967584e-01 1.77968577e-01 6.76243663e-01 8.46275806e-01
4.57574308e-01 1.30681109e+00 2.07843781e-01 -1.21258605e+00
-2.50108004e-01 -1.45832455e+00 -1.06323230e+00 1.53364688e-01
6.46191478e-01 -5.08105040e-01 -5.05944073e-01 8.30883309e-02] | [13.139178276062012, 3.4141664505004883] |
43676f40-3891-4815-9bff-ceb292d40e93 | adaptive-incomplete-multi-view-learning-via | 2208.0371 | null | https://arxiv.org/abs/2208.03710v1 | https://arxiv.org/pdf/2208.03710v1.pdf | Adaptive incomplete multi-view learning via tensor graph completion | With the advancement of the data acquisition techniques, multi-view learning has become a hot topic. Some multi-view learning methods assume that the multi-view data is complete, which means that all instances are present, but this too ideal. Certain tensor-based methods for handing incomplete multi-view data have emerged and have achieved better result. However, there are still some problems, such as use of traditional tensor norm which makes the computation high and is not able to handle out-of-sample. To solve these two problems, we proposed a new incomplete multi view learning method. A new tensor norm is defined to implement graph tensor data recover. The recovered graphs are then regularized to a consistent low-dimensional representation of the samples. In addition, adaptive weights are equipped to each view to adjust the importance of different views. Compared with the existing methods, our method nor only explores the consistency among views, but also obtains the low-dimensional representation of the new samples by using the learned projection matrix. An efficient algorithm based on inexact augmented Lagrange multiplier (ALM) method are designed to solve the model and convergence is proved. Experimental results on four datasets show the effectiveness of our method. | ['Xiaohong Chen', 'Heng Zhang'] | 2022-08-07 | null | null | null | null | ['multi-view-learning'] | ['computer-vision'] | [-3.63938302e-01 -4.25459176e-01 -2.79899091e-01 -1.91956416e-01
-4.61490870e-01 -2.18892083e-01 1.83954075e-01 -2.66651243e-01
-6.56380579e-02 4.21181798e-01 3.64872396e-01 3.84108096e-01
-4.71227735e-01 -6.72129452e-01 -4.50973541e-01 -8.95235062e-01
1.12652265e-01 4.20988470e-01 7.08973110e-02 -1.33610606e-01
2.30020419e-01 -3.03808432e-02 -1.28386366e+00 4.64075744e-01
7.07885087e-01 7.62822628e-01 2.63849616e-01 1.87534481e-01
-1.55820101e-01 7.10262775e-01 -1.21704768e-02 -2.61740535e-01
3.94996732e-01 -1.96474940e-01 -5.99605739e-01 6.05167806e-01
3.46715063e-01 -3.69792998e-01 -3.22493196e-01 1.20067787e+00
4.25882936e-01 2.56480891e-02 2.11888701e-01 -1.33804941e+00
-6.55918658e-01 5.23555517e-01 -1.12094760e+00 7.68616376e-03
3.20349842e-01 -2.99268126e-01 9.79456842e-01 -1.37756288e+00
6.49335265e-01 1.24909878e+00 4.61230278e-01 2.34762460e-01
-9.92016971e-01 -4.70150530e-01 4.28268939e-01 6.24987364e-01
-1.31339276e+00 -1.70781657e-01 1.17774498e+00 -5.08275509e-01
2.24731922e-01 2.44486958e-01 8.93949747e-01 7.56930768e-01
-1.77066177e-02 8.23492169e-01 1.49139690e+00 -1.93832308e-01
-1.52940646e-01 1.29603267e-01 1.48662582e-01 8.54679286e-01
3.97433668e-01 -2.87379235e-01 -4.50423509e-01 -1.94439560e-01
5.82054555e-01 6.32871628e-01 -4.57787991e-01 -8.14237177e-01
-1.53828907e+00 7.52069473e-01 2.75152802e-01 4.09985244e-01
-4.63372082e-01 -4.28353190e-01 5.99609613e-01 8.97835717e-02
4.61093754e-01 -1.81803241e-01 -2.69844770e-01 1.32806242e-01
-5.50614119e-01 2.82311141e-02 6.91190362e-01 8.52791190e-01
8.61130059e-01 5.46867102e-02 3.64850551e-01 9.16325867e-01
3.46737176e-01 5.57879925e-01 5.23797810e-01 -8.27547431e-01
9.52155948e-01 1.08176076e+00 -1.22761950e-01 -1.55282807e+00
-3.52788717e-01 -3.15512747e-01 -1.33995187e+00 2.31896024e-02
2.89298147e-01 6.42828271e-02 -5.31654894e-01 1.26788044e+00
6.16131842e-01 2.71055430e-01 6.00925237e-02 1.04277778e+00
7.05377221e-01 6.94710732e-01 -6.11674070e-01 -6.09933317e-01
1.23311710e+00 -9.08711553e-01 -9.08882439e-01 1.18120275e-01
4.09884095e-01 -8.57326150e-01 9.26285625e-01 7.63835728e-01
-8.21279287e-01 -4.79828835e-01 -1.15890276e+00 1.18692011e-01
-1.72463685e-01 8.59664232e-02 6.31372631e-01 3.08633804e-01
-4.97854739e-01 4.10417169e-01 -6.62533224e-01 -1.01223597e-02
1.49835825e-01 1.95174947e-01 -6.83388054e-01 -6.85830653e-01
-9.09810066e-01 6.71508551e-01 5.86566746e-01 4.47221667e-01
-6.49210870e-01 -5.09014487e-01 -5.56787193e-01 -1.90760911e-01
7.20790863e-01 -6.00761175e-01 4.39711720e-01 -6.43627286e-01
-1.15386462e+00 4.60074902e-01 -1.87659089e-03 3.49481732e-01
4.30886954e-01 -8.76302123e-02 -5.35519779e-01 2.05945879e-01
1.86119661e-01 -1.97650909e-01 9.76150453e-01 -1.57103288e+00
-4.86364275e-01 -7.32918203e-01 1.84335768e-01 5.15044451e-01
-5.89613438e-01 -3.59519750e-01 -5.89642227e-01 -3.54743481e-01
9.24759805e-01 -1.00721061e+00 -2.40047410e-01 -1.33898988e-01
-1.81257799e-01 -1.10215783e-01 1.23435378e+00 -8.66713524e-01
1.27570677e+00 -2.03853250e+00 8.57952178e-01 2.93925315e-01
5.65418184e-01 1.53922305e-01 8.17433298e-02 6.37761533e-01
-1.79049730e-01 -2.03235343e-01 -6.26662746e-02 -2.46224225e-01
-4.14269865e-01 5.36234319e-01 2.17844509e-02 5.45757234e-01
-5.88392556e-01 1.73163921e-01 -8.62074375e-01 -7.66211569e-01
2.48164281e-01 4.59222049e-01 -5.67437887e-01 2.41698250e-01
2.20211998e-01 6.40058994e-01 -6.29914224e-01 4.92080688e-01
9.79384482e-01 -3.98546547e-01 1.89308450e-01 -8.12348902e-01
-1.36972994e-01 -4.72418159e-01 -1.83467317e+00 1.86137938e+00
-4.11643565e-01 -3.33188713e-01 2.56236702e-01 -1.21589231e+00
6.95775807e-01 3.88731420e-01 8.93286049e-01 -2.86387503e-01
-1.06964251e-02 2.92358875e-01 -1.43467411e-01 -1.07046652e+00
2.01141134e-01 -2.05449685e-01 3.27712119e-01 3.84122610e-01
-1.44037977e-01 8.31434652e-02 2.72981644e-01 3.23500723e-01
4.31037813e-01 2.32402042e-01 2.99341649e-01 -4.74184304e-02
1.00691557e+00 -2.25387633e-01 9.30092454e-01 -9.63189900e-02
1.59536377e-01 4.82292533e-01 4.83872026e-01 -7.18071520e-01
-1.06623077e+00 -5.61248958e-01 7.67161548e-02 5.82076848e-01
2.19485059e-01 -5.76750815e-01 -5.51936209e-01 -8.45599413e-01
-1.52570859e-01 2.29011364e-02 -4.35741752e-01 -3.19737708e-03
-5.80501497e-01 -8.50874245e-01 -2.74883330e-01 2.98188120e-01
7.49256670e-01 -4.01602447e-01 1.63699202e-02 7.71152601e-02
-5.78137636e-01 -1.26139307e+00 -4.45034415e-01 -5.32934785e-01
-1.16457093e+00 -1.24057329e+00 -7.04509318e-01 -6.45849705e-01
1.01779234e+00 8.43962967e-01 5.48717082e-01 2.36155361e-01
8.84620845e-02 4.67824012e-01 -4.84600782e-01 9.99879315e-02
-9.58568975e-02 -2.50295162e-01 4.16597486e-01 7.34859228e-01
5.75074553e-02 -7.79397607e-01 -5.29071510e-01 4.05957460e-01
-1.04265869e+00 2.92601734e-01 6.99001849e-01 1.02736235e+00
7.84608722e-01 1.49910629e-01 3.95449698e-01 -9.51776683e-01
3.33207548e-01 -4.24066961e-01 -5.10702789e-01 4.15231884e-01
-8.52613270e-01 8.51985533e-03 8.61561298e-01 -3.40944290e-01
-9.02152240e-01 -1.30998297e-02 3.06104749e-01 -9.35853422e-01
3.22696537e-01 9.25805867e-01 -3.93616796e-01 -2.48075709e-01
6.64003491e-02 4.72948641e-01 2.97508121e-01 -6.00011528e-01
3.31356645e-01 4.91193742e-01 -5.50756454e-02 -5.22725284e-01
9.22109306e-01 7.38701224e-01 4.51540291e-01 -7.95384526e-01
-1.04698348e+00 -6.06254101e-01 -8.34411442e-01 -3.96928579e-01
6.83443666e-01 -9.52256203e-01 -7.86816955e-01 4.55693156e-01
-8.56633127e-01 5.88569462e-01 1.37852147e-01 9.45816934e-01
-3.93354744e-01 1.07616735e+00 -5.25189996e-01 -5.14815569e-01
-1.71155021e-01 -1.20775437e+00 6.26295567e-01 -1.26436278e-01
5.71726799e-01 -1.06658459e+00 1.10063188e-01 7.41569996e-01
1.00586072e-01 3.24002802e-01 7.67763257e-01 -1.96762994e-01
-6.65596545e-01 -2.02152684e-01 -1.78669721e-01 4.85074252e-01
2.29547396e-01 -1.92079589e-01 -6.09307766e-01 -6.09159350e-01
5.54096460e-01 -1.90136179e-01 5.12679219e-01 8.18787888e-02
1.16152835e+00 -6.10635757e-01 -2.61853844e-01 7.26548076e-01
1.76078546e+00 -1.39315501e-01 2.89487988e-01 2.48399258e-01
1.35651112e+00 6.86449766e-01 6.37400746e-01 5.56050539e-01
5.84691584e-01 5.14012218e-01 5.91528535e-01 1.54105037e-01
2.35725656e-01 -2.08094642e-01 8.07380453e-02 1.87038565e+00
-6.02017045e-01 3.16888243e-01 -6.74104035e-01 3.11996937e-01
-2.10554242e+00 -1.18439484e+00 -4.53885168e-01 2.28643918e+00
4.29410964e-01 -6.10847510e-02 9.78413522e-02 3.46130788e-01
7.08581805e-01 3.85524362e-01 -4.32847232e-01 2.09678143e-01
1.20313866e-02 -5.41337252e-01 3.17605972e-01 3.97646695e-01
-8.44086111e-01 3.03527206e-01 5.38630199e+00 5.51822245e-01
-9.63674664e-01 3.13234031e-01 1.15858026e-01 -3.44916899e-03
-3.91753078e-01 2.40308523e-01 -5.89476824e-01 3.22124362e-01
2.51419663e-01 -1.05504520e-01 6.74591422e-01 8.57365072e-01
2.16784149e-01 7.98792988e-02 -7.80084372e-01 1.48579788e+00
5.20952463e-01 -1.03498387e+00 2.56563783e-01 2.09257007e-01
8.09336424e-01 -2.00316057e-01 3.96344485e-03 1.14082791e-01
-2.08861694e-01 -2.79378682e-01 2.24236503e-01 7.15483904e-01
4.81448084e-01 -8.18271816e-01 5.90959966e-01 5.87180138e-01
-1.39345419e+00 -1.01984814e-01 -6.84978366e-01 -2.91749388e-02
6.85007945e-02 8.46471369e-01 -4.33849603e-01 1.23925626e+00
5.75800836e-01 1.24316072e+00 -6.37030780e-01 8.02186430e-01
1.19748540e-01 2.47324362e-01 -1.96542472e-01 3.30055922e-01
1.76181898e-01 -8.79140258e-01 6.50317669e-01 3.80773455e-01
2.73653030e-01 2.80657172e-01 7.93661535e-01 3.80168110e-01
2.88952410e-01 5.26611924e-01 -8.32324088e-01 1.94933876e-01
6.25177398e-02 1.60381877e+00 -3.65771532e-01 -4.02538151e-01
-9.51535702e-01 8.69192421e-01 4.09759849e-01 3.41844738e-01
-4.92411077e-01 9.10922959e-02 1.03953823e-01 -3.93508449e-02
1.21136472e-01 -5.48923492e-01 -3.67375463e-02 -1.80462849e+00
4.95157957e-01 -1.10039115e+00 6.14687741e-01 -4.57590252e-01
-1.46211517e+00 2.44634852e-01 8.43213052e-02 -1.79174757e+00
1.17931426e-01 -4.80014414e-01 -2.40860194e-01 6.36575222e-01
-1.35764349e+00 -1.52016032e+00 -4.71251041e-01 9.37196851e-01
3.55118483e-01 -2.82269984e-01 6.26203239e-01 6.41567528e-01
-6.91563666e-01 1.32053047e-01 2.31214553e-01 1.42799139e-01
6.15805387e-01 -1.06174386e+00 -5.48408389e-01 8.19191277e-01
-1.24600772e-02 6.13127708e-01 3.80510569e-01 -5.73586106e-01
-2.01206470e+00 -7.31683373e-01 4.47749972e-01 -1.83862582e-01
6.20765030e-01 -8.41794759e-02 -1.11268914e+00 8.83261800e-01
1.57664955e-01 3.56304049e-01 8.22798073e-01 2.40008399e-01
-4.19590622e-01 -5.27244389e-01 -9.38759446e-01 4.20478314e-01
8.53333294e-01 -2.76160717e-01 -5.43653309e-01 6.11675858e-01
4.89212155e-01 -4.22946155e-01 -1.25542939e+00 3.95219743e-01
4.81243461e-01 -1.20566845e+00 9.53306794e-01 -6.47184789e-01
2.94071168e-01 -6.07002854e-01 -4.12739724e-01 -1.40157211e+00
-5.16268313e-01 -2.59435028e-01 -2.50836790e-01 1.35432184e+00
-3.43438834e-02 -7.53621280e-01 6.92272902e-01 3.88304681e-01
-1.05750866e-01 -9.32746172e-01 -9.30213213e-01 -7.81106532e-01
-3.78055125e-01 8.87690336e-02 4.54067886e-01 1.32168913e+00
8.38921666e-02 4.75034356e-01 -8.56520653e-01 3.75801265e-01
1.13457501e+00 5.28453946e-01 6.34191096e-01 -1.21305478e+00
-3.35925221e-01 6.48755133e-02 -4.31417316e-01 -5.74591517e-01
-5.25226966e-02 -9.97631431e-01 -5.76373100e-01 -1.65686572e+00
6.00083113e-01 -3.67726177e-01 -2.75051206e-01 2.09961995e-01
-2.95050621e-01 -1.98046684e-01 1.80644438e-01 6.15511298e-01
-5.11890829e-01 5.75603604e-01 1.76529658e+00 -5.56838885e-02
2.12603360e-02 -5.66316061e-02 -4.99835670e-01 8.42344046e-01
4.29027826e-01 -2.89286137e-01 -5.13799429e-01 -4.77541029e-01
3.66109431e-01 4.05201375e-01 5.98096587e-02 -1.02031386e+00
3.00880402e-01 -3.28643590e-01 2.73133576e-01 -7.81181872e-01
4.74918306e-01 -1.23720980e+00 2.10580453e-01 4.98887062e-01
1.60507575e-01 3.65438491e-01 -4.99833405e-01 8.73607457e-01
-4.14872468e-01 -1.42161906e-01 6.75459385e-01 -5.05119443e-01
-7.34945476e-01 8.45173597e-01 3.26340109e-01 8.43546465e-02
1.06204557e+00 -7.65690729e-02 -5.41137345e-02 -1.70549944e-01
-7.14647353e-01 4.86083806e-01 2.73225397e-01 3.89258593e-01
8.91717315e-01 -1.80995595e+00 -7.80351043e-01 1.52260616e-01
2.99823493e-01 1.01696126e-01 6.36621833e-01 1.17227197e+00
-3.01627874e-01 1.23902164e-01 -2.79559314e-01 -6.97099209e-01
-1.27443230e+00 1.06959593e+00 -7.58792385e-02 -4.03499871e-01
-7.97139525e-01 2.02951610e-01 -2.03678719e-04 -4.62227404e-01
-7.45674521e-02 9.49584693e-02 -5.99917233e-01 3.27481359e-01
5.18783689e-01 6.22803569e-01 -1.47260293e-01 -8.48318338e-01
-1.03403054e-01 1.12098718e+00 -1.23780653e-01 1.19556924e-02
1.57950556e+00 -4.10189480e-01 -6.48798287e-01 8.45488548e-01
1.44974613e+00 1.33502841e-01 -8.68043005e-01 -5.00618339e-01
-2.39667237e-01 -7.80618846e-01 3.01181134e-02 -1.13198817e-01
-1.37536216e+00 8.88699412e-01 4.39628184e-01 2.11272791e-01
1.04763949e+00 -5.19702077e-01 8.59065354e-01 3.65614831e-01
5.96562684e-01 -1.10536718e+00 2.35320345e-01 2.70202994e-01
9.23906744e-01 -1.30371261e+00 5.24192333e-01 -6.70636952e-01
-7.07801104e-01 1.35530746e+00 7.68398941e-01 -1.51330903e-01
7.75053740e-01 -2.56739765e-01 -8.79895538e-02 -3.89152318e-01
-5.29495895e-01 1.37962788e-01 2.39229232e-01 2.66831636e-01
1.31600440e-01 -2.27446016e-02 -5.26701987e-01 1.20907746e-01
2.67478853e-01 -7.69500509e-02 7.26759434e-01 7.75842369e-01
-2.89881170e-01 -1.25812125e+00 -6.12598479e-01 4.95906770e-01
-3.43536139e-01 3.29610556e-01 1.06570914e-01 5.97263277e-01
9.77637842e-02 7.65171945e-01 -5.23628712e-01 -5.88105619e-01
4.12271649e-01 -1.50448801e-02 5.28690279e-01 -6.05027318e-01
-4.24941517e-02 1.99282125e-01 -2.28544518e-01 -5.26782990e-01
-8.49476039e-01 -7.54408300e-01 -1.00880480e+00 -2.39059538e-01
-4.81855303e-01 2.40563005e-01 5.70288420e-01 8.21236134e-01
2.00971752e-01 2.47517213e-01 1.08103371e+00 -7.38562226e-01
-7.77184725e-01 -6.58445895e-01 -9.16768670e-01 7.63647795e-01
1.33724868e-01 -7.48236179e-01 -4.71155435e-01 -4.30294089e-02] | [8.270506858825684, 4.612700939178467] |
446dd062-e191-40a1-80bb-9e0dc5026ca0 | sketch2cloth-sketch-based-3d-garment | 2303.00167 | null | https://arxiv.org/abs/2303.00167v1 | https://arxiv.org/pdf/2303.00167v1.pdf | Sketch2Cloth: Sketch-based 3D Garment Generation with Unsigned Distance Fields | 3D model reconstruction from a single image has achieved great progress with the recent deep generative models. However, the conventional reconstruction approaches with template mesh deformation and implicit fields have difficulty in reconstructing non-watertight 3D mesh models, such as garments. In contrast to image-based modeling, the sketch-based approach can help users generate 3D models to meet the design intentions from hand-drawn sketches. In this study, we propose Sketch2Cloth, a sketch-based 3D garment generation system using the unsigned distance fields from the user's sketch input. Sketch2Cloth first estimates the unsigned distance function of the target 3D model from the sketch input, and extracts the mesh from the estimated field with Marching Cubes. We also provide the model editing function to modify the generated mesh. We verified the proposed Sketch2Cloth with quantitative evaluations on garment generation and editing with a state-of-the-art approach. | ['Kazunori Miyata', 'Haoran Xie', 'Yi He'] | 2023-03-01 | null | null | null | null | ['model-editing'] | ['natural-language-processing'] | [ 1.39807105e-01 7.41996467e-02 2.00223044e-01 -6.70691878e-02
-3.78307521e-01 -5.41534960e-01 6.29167914e-01 -6.21146858e-01
4.47265089e-01 4.26658601e-01 -3.03807650e-02 2.74187643e-02
-2.77340990e-02 -1.24272835e+00 -7.89220035e-01 -1.00162238e-01
4.16098893e-01 8.65891457e-01 -2.15027668e-02 -2.91744620e-01
4.46734518e-01 9.98982906e-01 -1.46405077e+00 4.12738860e-01
8.27108622e-01 9.91490424e-01 2.29621708e-01 7.67405152e-01
-6.04650497e-01 1.64627701e-01 -4.07613069e-01 -5.14496803e-01
3.45886052e-01 -4.13392395e-01 -2.55806804e-01 3.11292976e-01
6.10647261e-01 -6.40027046e-01 -3.40986907e-01 6.84092700e-01
5.43786824e-01 -1.52075768e-01 9.43504632e-01 -1.19875121e+00
-8.99320185e-01 6.65685460e-02 -4.71698999e-01 -9.74145710e-01
6.21101618e-01 9.83899683e-02 3.02588552e-01 -1.48386967e+00
1.18590188e+00 1.58488166e+00 8.26755583e-01 7.38597751e-01
-1.33540952e+00 -8.21289778e-01 -1.50551675e-02 -5.98408937e-01
-1.59979546e+00 -8.77296627e-02 1.33815074e+00 -4.41861212e-01
5.96488237e-01 6.29621089e-01 1.12955999e+00 6.78380847e-01
2.36887917e-01 8.75865042e-01 9.64095056e-01 -4.06170875e-01
5.77995367e-02 -6.41590059e-02 -5.83579063e-01 1.00961006e+00
-2.35323489e-01 1.38732195e-01 -3.96963954e-01 -4.59017843e-01
1.93368959e+00 1.15166798e-01 -1.02155641e-01 -6.06045127e-01
-1.27341592e+00 4.87407446e-01 2.58791357e-01 -3.44219208e-02
-4.05177057e-01 6.27270162e-01 -1.49481907e-01 3.72396745e-02
8.71299624e-01 2.09068924e-01 -9.30532441e-02 -1.17750980e-01
-1.30710053e+00 7.87259340e-01 8.28383923e-01 1.46550739e+00
6.23403609e-01 2.52689898e-01 -5.88755906e-02 7.33368099e-01
6.35248363e-01 8.03501368e-01 -3.70054066e-01 -9.38231170e-01
3.20021152e-01 7.73425519e-01 3.22579384e-01 -1.19269073e+00
2.08705842e-01 2.04566643e-02 -9.60930526e-01 8.65002334e-01
1.64407745e-01 1.36760756e-01 -1.16581941e+00 1.01649678e+00
4.86223489e-01 4.42184135e-02 -5.18926620e-01 9.20848012e-01
1.05719709e+00 6.70125604e-01 -2.14312389e-01 2.64249206e-01
9.45722163e-01 -6.44518256e-01 -8.02354515e-01 2.23364115e-01
-8.91374722e-02 -1.10940969e+00 1.08989966e+00 5.16620278e-01
-1.56990087e+00 -7.02058911e-01 -8.67506623e-01 -2.31655911e-01
7.95930065e-03 4.16604936e-01 4.88855898e-01 4.61613297e-01
-8.80865932e-01 7.81197965e-01 -5.69781721e-01 1.69838324e-01
4.58953768e-01 7.42580891e-02 -2.20190540e-01 -8.31001028e-02
-8.77921462e-01 7.84044921e-01 -2.75782406e-01 2.39347294e-01
-7.72837996e-01 -1.06736434e+00 -7.89873481e-01 -2.28900388e-01
-6.98376149e-02 -9.40830350e-01 1.04418266e+00 -5.99099338e-01
-1.63992941e+00 1.00638103e+00 -1.08252347e-01 2.79503047e-01
1.06962097e+00 -1.47182345e-01 1.22228689e-01 -5.33917136e-02
-2.17279419e-01 6.47199512e-01 1.17928731e+00 -1.82543290e+00
1.92313910e-01 -1.61280021e-01 -8.66850242e-02 1.42234728e-01
4.62389976e-01 -4.86762077e-01 -4.51144099e-01 -1.17257750e+00
3.69770974e-01 -7.01144636e-01 -1.33773640e-01 1.03940129e+00
-4.24035639e-01 -4.77233231e-02 1.26359522e+00 -9.24901068e-01
1.05110157e+00 -1.87451887e+00 2.66797930e-01 4.49551374e-01
1.51954517e-01 9.95197371e-02 -1.02310315e-01 8.18952978e-01
2.61766106e-01 2.36886948e-01 -3.54207039e-01 -7.79712319e-01
2.61449277e-01 7.69793838e-02 -4.96996641e-01 1.36924148e-01
1.64423645e-01 1.04586685e+00 -8.32411587e-01 -5.81776679e-01
5.64219415e-01 9.19553161e-01 -4.78703618e-01 3.45425218e-01
-5.42422295e-01 3.23893547e-01 -5.15222669e-01 9.57631052e-01
1.14715552e+00 2.30983064e-01 7.14625344e-02 -4.56661701e-01
-6.09556176e-02 -3.39398921e-01 -1.45190156e+00 2.20556784e+00
-6.12490594e-01 2.49777257e-01 4.10462230e-01 -2.74744749e-01
1.56540585e+00 3.45531017e-01 3.76494318e-01 -3.60572636e-01
-1.27971277e-01 3.84328008e-01 -7.26474822e-01 -2.87501514e-01
5.25598764e-01 -2.26891711e-01 2.38713156e-02 6.68251455e-01
-3.58635485e-01 -1.29701328e+00 -4.59839106e-01 1.27823263e-01
5.43031812e-01 9.66935515e-01 -1.84096247e-01 -2.03435302e-01
1.10980749e-01 -6.25594780e-02 -7.51158819e-02 3.31376523e-01
6.17042065e-01 1.13652432e+00 2.17860699e-01 -6.63186491e-01
-1.36738837e+00 -1.31589305e+00 1.30780369e-01 1.35697722e-01
2.95767128e-01 -4.80091631e-01 -8.30964625e-01 -3.68716508e-01
3.91645938e-01 7.48971879e-01 -7.18232274e-01 1.68497249e-01
-6.67546630e-01 9.91503596e-02 4.26840991e-01 4.57052022e-01
4.93911982e-01 -1.07804024e+00 -5.79589009e-01 3.29228133e-01
1.30729377e-01 -4.94277656e-01 -8.75624895e-01 -7.43487000e-01
-9.59414899e-01 -8.84151101e-01 -1.13578403e+00 -7.76562870e-01
1.10069394e+00 -1.91541053e-02 1.29283237e+00 6.04857147e-01
-4.56233233e-01 5.59227049e-01 -2.50059329e-02 -4.54436630e-01
-6.87060654e-01 -3.86087239e-01 -3.23567957e-01 -6.78467602e-02
-4.83293563e-01 -7.18424678e-01 -7.80936778e-01 5.85715115e-01
-1.06005669e+00 9.39568043e-01 1.02694087e-01 5.38022876e-01
9.82273757e-01 -2.28118315e-01 4.30064559e-01 -4.23678666e-01
8.64774108e-01 7.54881129e-02 -3.77054989e-01 2.48604134e-01
-1.42102212e-01 8.33021030e-02 3.26188326e-01 -8.61193657e-01
-1.05771947e+00 1.33701965e-01 -3.54770482e-01 -8.03075671e-01
6.20180555e-02 1.31101936e-01 -2.85728276e-01 -1.48512870e-01
3.21111113e-01 2.67724156e-01 2.45420873e-01 -7.83698380e-01
5.79699218e-01 4.58690733e-01 2.97937125e-01 -9.41254795e-01
1.08123279e+00 5.76716602e-01 1.47058174e-01 -6.67322636e-01
-2.72095539e-02 2.97840416e-01 -8.26783478e-01 -7.12289393e-01
6.51994705e-01 -4.80874419e-01 -6.76953614e-01 7.33008504e-01
-1.73796642e+00 -5.30697882e-01 -4.21871901e-01 -6.94687814e-02
-8.01993072e-01 1.53187588e-01 -5.20706117e-01 -1.16408837e+00
-6.23160303e-01 -1.04745889e+00 1.58003056e+00 1.91328563e-02
-4.97313291e-01 -7.23618090e-01 4.26043123e-02 1.41243692e-02
5.27425230e-01 9.42079484e-01 1.02029121e+00 6.69313431e-01
-8.77354443e-01 -3.62879992e-01 -6.31452426e-02 -1.76170170e-02
2.78901875e-01 3.96070838e-01 -7.24727273e-01 2.34345764e-01
-4.83136177e-01 -6.45280331e-02 1.90553382e-01 2.46865287e-01
1.21539676e+00 -1.53687805e-01 -3.79069746e-01 6.22633815e-01
1.43817198e+00 1.93196058e-01 9.99868453e-01 -3.88890535e-01
8.92081022e-01 2.86932200e-01 6.11554027e-01 5.51291943e-01
2.21769691e-01 7.21136272e-01 3.92438471e-01 -2.41082653e-01
-5.01045048e-01 -1.10779703e+00 -1.53803825e-01 6.49479389e-01
-5.35885632e-01 -2.92182360e-02 -5.55339754e-01 2.16655999e-01
-1.57203913e+00 -7.77982116e-01 -2.97339886e-01 2.11652517e+00
8.57161701e-01 -3.31225134e-02 -1.50275350e-01 -5.91886602e-03
5.08159757e-01 -1.82585809e-02 -4.33822542e-01 -6.55685961e-01
3.01194698e-01 6.38627946e-01 2.94528506e-03 6.22616231e-01
-4.15036827e-01 7.70645976e-01 6.44989300e+00 9.41392839e-01
-1.04472220e+00 -2.59008616e-01 3.35585535e-01 1.69956118e-01
-1.04248857e+00 -3.76359597e-02 -5.49726844e-01 3.93623620e-01
1.86640471e-02 -2.00627774e-01 4.64815557e-01 8.40428948e-01
3.62730652e-01 1.69771183e-02 -1.09948349e+00 1.16188288e+00
2.36267578e-02 -1.77887118e+00 4.25954551e-01 1.15348160e-01
7.82270730e-01 -9.16904271e-01 -1.13791645e-01 -1.16269719e-02
-4.21396829e-02 -1.22617722e+00 1.28643775e+00 1.32853365e+00
1.58259881e+00 -7.52475798e-01 -4.47802618e-03 6.69458926e-01
-1.48277974e+00 7.83531249e-01 -1.56504303e-01 3.46859209e-02
5.25099635e-01 5.25180042e-01 -7.45442033e-01 4.63456094e-01
2.46089041e-01 3.70534718e-01 -1.13978796e-01 8.11278999e-01
-1.01041138e-01 2.05720305e-01 -2.80896038e-01 -2.62794513e-02
-1.85060278e-01 -5.18456578e-01 5.31630635e-01 9.81863797e-01
7.49863386e-01 2.30786592e-01 -2.88007706e-02 1.78349471e+00
-1.68011025e-01 -4.30135876e-02 -7.76993573e-01 -2.60426551e-01
4.56904531e-01 1.04214251e+00 -5.94025671e-01 -4.85145926e-01
3.14205736e-01 1.27853930e+00 -1.67174518e-01 2.63844788e-01
-8.38812947e-01 -4.72793072e-01 4.14305210e-01 7.99838543e-01
1.43448059e-02 -7.30039775e-01 -4.74716455e-01 -8.74047160e-01
2.26150993e-02 -5.11030376e-01 -6.71613038e-01 -1.50365031e+00
-1.15018666e+00 4.87999707e-01 2.35954672e-01 -1.45189404e+00
-6.14232011e-02 -3.13275337e-01 -7.21048295e-01 1.33003104e+00
-9.23697531e-01 -1.65707099e+00 -5.58312356e-01 2.10815698e-01
7.42680252e-01 2.31020570e-01 1.13764310e+00 2.59783506e-01
3.47291440e-01 3.37449133e-01 -3.71811420e-01 2.32718792e-03
3.71018499e-01 -8.87005210e-01 1.10726392e+00 2.38408998e-01
-7.87272602e-02 4.36822563e-01 4.29661632e-01 -1.26194632e+00
-1.80441558e+00 -7.97026515e-01 5.44884443e-01 -4.84433174e-01
-1.21485360e-01 -6.85824335e-01 -8.17445695e-01 3.36171716e-01
-1.02088287e-01 6.54308917e-03 -2.64859982e-02 -6.92632139e-01
1.45275518e-03 2.99081177e-01 -1.61538064e+00 7.05493152e-01
1.60323274e+00 -4.17285740e-01 -4.13946092e-01 5.89272659e-03
4.17554140e-01 -9.38993216e-01 -1.03746045e+00 3.35562527e-01
1.25976551e+00 -6.16358519e-01 1.35027516e+00 -2.40751237e-01
6.50955200e-01 -4.47964013e-01 8.38036090e-03 -1.43943417e+00
-2.59586304e-01 -7.81940758e-01 -3.78465056e-01 9.75041866e-01
-2.10354459e-02 -1.28608271e-01 1.01013744e+00 9.03326571e-01
-8.97185355e-02 -9.50322628e-01 -5.20004749e-01 -5.03176749e-01
-3.47418189e-02 -3.97709370e-01 1.03976285e+00 8.66095364e-01
-5.98229587e-01 -4.27048296e-01 -5.82910955e-01 -3.17539036e-01
8.63790274e-01 4.53234285e-01 1.13680208e+00 -1.35883415e+00
3.60388868e-02 -1.07852705e-01 -1.41156107e-01 -1.13955915e+00
-7.26922229e-02 -9.09052610e-01 -1.63224235e-01 -2.17508578e+00
-3.68957222e-01 -9.47640717e-01 8.13190222e-01 1.78340346e-01
2.03731030e-01 2.83460289e-01 3.40339035e-01 -7.15174600e-02
3.68516713e-01 5.97953856e-01 2.17388797e+00 -2.94433177e-01
-2.79127479e-01 -1.29904881e-01 -3.06557328e-01 7.76693583e-01
3.94079775e-01 -2.08139077e-01 -5.30554771e-01 -5.49894929e-01
2.52266645e-01 4.68061507e-01 6.71785593e-01 -6.32082999e-01
-2.13842884e-01 -4.54692096e-01 8.16141188e-01 -1.19769073e+00
7.15939105e-01 -9.09391224e-01 1.15600812e+00 3.25173616e-01
-8.30525607e-02 1.27882108e-01 2.76872724e-01 2.04357222e-01
2.72638887e-01 -4.72372137e-02 5.55793285e-01 -5.38951397e-01
-2.07068160e-01 5.38120687e-01 -7.31078535e-02 -4.57736015e-01
6.74599767e-01 -5.26011765e-01 1.27901763e-01 -3.97130191e-01
-8.83368194e-01 -3.22037995e-01 8.51857126e-01 3.63753974e-01
1.47442770e+00 -2.01135325e+00 -7.98463821e-01 6.33503377e-01
-3.25791717e-01 3.71734768e-01 3.89373541e-01 3.49036828e-02
-1.00616384e+00 -1.64081648e-01 -2.04507589e-01 -4.14255530e-01
-1.19131851e+00 1.86795309e-01 5.02824426e-01 -7.76722059e-02
-9.19584811e-01 5.15307188e-01 1.39990836e-01 -7.26148248e-01
-8.39171261e-02 -6.20667338e-01 5.66235602e-01 -3.23286682e-01
3.59268486e-01 3.82652521e-01 -1.51780918e-01 -3.13359112e-01
-1.32176876e-01 9.80945408e-01 6.08489275e-01 -3.39579701e-01
1.22091532e+00 3.77252489e-01 7.53965080e-02 1.66836500e-01
9.54329729e-01 1.97460219e-01 -1.40856063e+00 1.81246296e-01
-7.99636126e-01 -9.98095989e-01 -1.02630012e-01 -8.85391593e-01
-1.01399636e+00 8.59748006e-01 2.42718384e-01 -6.56844079e-02
6.02141857e-01 -2.64563769e-01 1.11111987e+00 -2.37146080e-01
8.37236881e-01 -8.27217400e-01 1.21512108e-01 9.55865383e-02
1.91483057e+00 -5.39626896e-01 1.78508580e-01 -6.80571556e-01
-3.65001768e-01 1.25417864e+00 4.33057845e-01 -5.21373093e-01
8.75395179e-01 6.23988092e-01 -3.12401772e-01 -3.97762448e-01
-3.04411322e-01 5.12416184e-01 7.72059083e-01 6.65075779e-01
2.46963158e-01 3.27414989e-01 -1.47913739e-01 4.83490914e-01
-1.19889595e-01 4.72598404e-01 1.49158075e-01 1.02539098e+00
-1.57921821e-01 -1.37737691e+00 -7.25488484e-01 3.36264372e-01
3.52952123e-01 4.56406921e-02 -5.20528078e-01 7.90903568e-01
2.61367172e-01 2.27394432e-01 9.20025036e-02 -2.21535429e-01
7.78248191e-01 5.08568771e-02 1.10696840e+00 -4.67870682e-01
-4.25574392e-01 1.28161609e-01 -6.24133795e-02 -4.01933610e-01
-2.17571691e-01 -1.34607017e-01 -1.20389068e+00 -3.74830902e-01
-5.36710881e-02 -1.80844471e-01 1.05432606e+00 3.53118867e-01
5.05152464e-01 1.24773711e-01 4.93173897e-01 -1.72626770e+00
-5.45159876e-02 -8.01857173e-01 -6.57354832e-01 4.03128147e-01
-5.74762374e-02 -8.51128340e-01 -1.51625231e-01 3.19608241e-01] | [9.102856636047363, -3.5187559127807617] |
fc9ed768-b627-488d-9f27-96ef86a0b310 | partition-hypergraphs-with-embeddings | 1909.04016 | null | https://arxiv.org/abs/1909.04016v5 | https://arxiv.org/pdf/1909.04016v5.pdf | Hypergraph Partitioning With Embeddings | Problems in scientific computing, such as distributing large sparse matrix operations, have analogous formulations as hypergraph partitioning problems. A hypergraph is a generalization of a traditional graph wherein "hyperedges" may connect any number of nodes. As a result, hypergraph partitioning is an NP-Hard problem to both solve or approximate. State-of-the-art algorithms that solve this problem follow the multilevel paradigm, which begins by iteratively "coarsening" the input hypergraph to smaller problem instances that share key structural features. Once identifying an approximate problem that is small enough to be solved directly, that solution can be interpolated and refined to the original problem. While this strategy represents an excellent trade off between quality and running time, it is sensitive to coarsening strategy. In this work we propose using graph embeddings of the initial hypergraph in order to ensure that coarsened problem instances retrain key structural features. Our approach prioritizes coarsening within self-similar regions within the input graph, and leads to significantly improved solution quality across a range of considered hypergraphs. Reproducibility: All source code, plots and experimental data are available at https://sybrandt.com/2019/partition. | ['Justin Sybrandt', 'Ruslan Shaydulin', 'Ilya Safro'] | 2019-09-09 | null | null | null | null | ['hypergraph-partitioning'] | ['graphs'] | [ 4.34948765e-02 3.43362182e-01 -1.08487323e-01 1.71250980e-02
-4.56467271e-01 -8.24342370e-01 2.21479803e-01 7.86746800e-01
1.52861234e-02 7.31091857e-01 1.04668206e-02 -2.17862293e-01
-5.59390366e-01 -1.18947935e+00 -7.90028512e-01 -6.60258591e-01
-3.58994275e-01 1.04611826e+00 3.05543870e-01 1.39523959e-02
4.77667749e-01 6.04148924e-01 -1.44000876e+00 2.68223882e-01
9.62518573e-01 4.61502761e-01 5.09904814e-04 6.32067502e-01
-4.05444771e-01 1.92988843e-01 -5.51548183e-01 -1.91431805e-01
5.23668468e-01 -3.27322453e-01 -1.34721971e+00 1.46171749e-01
2.80414581e-01 4.88934517e-01 -4.26028401e-01 1.26454270e+00
2.77925998e-01 1.65959522e-01 4.26624984e-01 -1.47633791e+00
-5.65466344e-01 8.69087160e-01 -1.01937068e+00 1.35143623e-01
5.20244718e-01 -7.94934332e-02 1.09636343e+00 -6.57882690e-01
6.39155686e-01 9.87741232e-01 5.63328803e-01 -1.51754180e-02
-1.84194362e+00 -5.46401620e-01 -3.82301002e-03 1.93917319e-01
-2.00674176e+00 7.44521767e-02 4.69399333e-01 -4.40663517e-01
8.52592349e-01 7.57307649e-01 8.28208029e-01 1.04679368e-01
4.48776990e-01 2.13183276e-02 1.17454183e+00 -2.63478726e-01
5.89133203e-01 1.46148816e-01 5.50704598e-01 4.70025897e-01
7.54764676e-01 -2.63472527e-01 -2.59764016e-01 -6.17133737e-01
4.59369928e-01 -1.59897968e-01 -5.62187731e-01 -7.82282114e-01
-1.23417127e+00 8.84073973e-01 6.12104118e-01 2.54412502e-01
-4.01879698e-01 4.50069569e-02 3.41209859e-01 2.96590686e-01
3.02403271e-01 8.13505411e-01 -6.26386404e-02 2.66824245e-01
-1.11741245e+00 4.17800575e-01 1.29500711e+00 9.20690954e-01
1.26184726e+00 -2.84855574e-01 3.10642030e-02 4.79845822e-01
-1.98012412e-01 2.43161805e-02 3.47170830e-02 -7.38332212e-01
1.98756978e-01 1.05463088e+00 -2.22100064e-01 -1.21282566e+00
-5.19366980e-01 -4.36906934e-01 -1.02570617e+00 1.38464853e-01
3.27523410e-01 2.52109736e-01 -9.59137022e-01 1.31429136e+00
7.70949960e-01 2.62186080e-01 -2.33865201e-01 1.02095151e+00
8.53978276e-01 9.93437409e-01 -1.92645818e-01 -2.59422809e-01
1.42783320e+00 -8.73335958e-01 -3.17140371e-01 -1.53633477e-02
3.94194663e-01 -7.40231812e-01 9.47866976e-01 3.13482225e-01
-1.23117650e+00 -2.77091950e-01 -1.01493764e+00 4.94780876e-02
-3.97482753e-01 -6.54088676e-01 5.23817718e-01 5.86614251e-01
-1.48584509e+00 5.18558085e-01 -4.67585206e-01 -4.13192242e-01
-8.85946453e-02 7.12478995e-01 -5.40908575e-01 -3.72856408e-01
-8.90160203e-01 5.92656195e-01 7.33683288e-01 -1.55308694e-01
-5.39583743e-01 -1.13389444e+00 -7.56498337e-01 5.17666817e-01
6.36833310e-01 -7.56111920e-01 5.49026251e-01 -5.13379037e-01
-8.57708514e-01 7.86908865e-01 -1.10925913e-01 -2.38070264e-01
1.46492451e-01 5.87825775e-01 -4.95812483e-02 1.17629685e-01
-2.92967036e-02 3.65555644e-01 7.00120568e-01 -1.41500831e+00
-3.98438990e-01 -3.60896498e-01 3.07472557e-01 1.85223266e-01
-2.94951737e-01 -3.23627651e-01 -6.47422254e-01 -3.49717826e-01
2.26025566e-01 -1.28164947e+00 -5.70141971e-01 -6.04773939e-01
-6.00792706e-01 -2.71984696e-01 4.47217941e-01 -2.63933837e-01
1.56934202e+00 -1.85786188e+00 6.53789520e-01 8.67679834e-01
8.79622757e-01 -1.18239619e-01 -1.71454713e-01 9.21632230e-01
-3.12722147e-01 3.78025562e-01 -4.14347202e-01 2.72206187e-01
4.85718809e-02 6.55092672e-02 -1.64251868e-02 9.36289251e-01
-2.54403591e-01 7.14370489e-01 -7.57772326e-01 -4.09625173e-01
9.07836258e-02 2.64124334e-01 -8.31728995e-01 1.42757416e-01
-1.08227320e-01 9.73702744e-02 -2.49235451e-01 3.64660293e-01
1.18969071e+00 -6.20599449e-01 5.12575924e-01 -2.69959718e-01
-9.99304280e-02 -1.43312454e-01 -1.67758477e+00 1.28630579e+00
-1.65886492e-01 3.07537526e-01 3.10422689e-01 -1.15048873e+00
7.94765115e-01 9.98864844e-02 7.11222649e-01 -3.21935445e-01
-8.33957046e-02 1.37379449e-02 1.00838937e-01 -7.63629749e-02
7.48906553e-01 1.45556673e-01 -1.65986836e-01 6.75857186e-01
-3.59671772e-01 -4.38893497e-01 6.98758841e-01 5.99774063e-01
1.38211572e+00 -4.65013504e-01 3.44577968e-01 -9.49613810e-01
4.48455065e-01 4.61471617e-01 5.86825550e-01 5.06891727e-01
2.79773146e-01 6.49487793e-01 9.05355334e-01 -5.28959930e-01
-1.22652864e+00 -9.94071364e-01 -2.99245626e-01 7.39200532e-01
4.54143077e-01 -9.59043145e-01 -9.67865348e-01 -3.29113394e-01
2.93940336e-01 2.97331214e-01 -8.43078732e-01 -4.16017976e-03
-5.29275775e-01 -8.07270288e-01 3.50300856e-02 4.06609401e-02
-1.28349811e-01 -7.26206243e-01 -3.15509528e-01 1.58826903e-01
-8.93810391e-03 -8.02754104e-01 -5.71167111e-01 2.00007603e-01
-6.40684307e-01 -1.37358272e+00 -4.37446594e-01 -7.86322176e-01
1.11302912e+00 5.48179626e-01 1.34244502e+00 5.02278507e-01
-4.77368921e-01 1.55552164e-01 -2.98297197e-01 1.89231440e-01
-2.40473807e-01 3.49458247e-01 -2.41142754e-02 -2.01237023e-01
1.39310315e-01 -6.63601518e-01 -5.92289090e-01 2.00589016e-01
-1.17038202e+00 -9.32148937e-03 3.46642166e-01 7.12010026e-01
1.06196570e+00 6.76273763e-01 1.25772953e-01 -1.33940125e+00
8.02973807e-01 -8.56157422e-01 -8.48793030e-01 3.49779755e-01
-7.55128682e-01 2.96605617e-01 8.50003421e-01 -1.65745512e-01
-3.55810434e-01 7.94711411e-02 2.76190579e-01 -3.97678167e-01
1.37680322e-01 8.41645479e-01 -2.36486010e-02 -3.52168888e-01
5.87068856e-01 5.79584949e-02 -1.33743033e-01 -1.82425871e-01
4.25580084e-01 1.61065266e-01 3.19476426e-01 -6.61958456e-01
9.97482419e-01 2.87965268e-01 4.05058026e-01 -7.12753296e-01
-2.90743977e-01 -5.89995205e-01 -4.63042319e-01 -5.45173474e-02
5.65018833e-01 -4.90283549e-01 -9.39466119e-01 -2.00297385e-01
-7.94898212e-01 -1.58180013e-01 -3.23317975e-01 1.01976775e-01
-2.23000243e-01 4.32269573e-01 -3.92697364e-01 -2.42383584e-01
-1.85061529e-01 -1.25729930e+00 8.09205234e-01 1.99893638e-01
-3.98885190e-01 -9.34294999e-01 3.82718533e-01 1.44086629e-01
3.22631270e-01 4.80868608e-01 1.21893191e+00 -5.63976586e-01
-7.18425095e-01 -1.52731180e-01 -3.92539054e-01 -4.27659929e-01
7.05705509e-02 1.20997600e-01 -4.47584659e-01 -8.41393828e-01
-1.98390260e-01 -2.46821940e-02 6.79361522e-01 4.00856584e-01
1.26240408e+00 -4.96914089e-01 -5.68072438e-01 8.13993216e-01
1.71392512e+00 -1.80194378e-01 5.95663607e-01 3.78883511e-01
8.88460577e-01 5.96788406e-01 3.00583839e-01 6.10843837e-01
4.49603885e-01 3.76078665e-01 3.54383856e-01 -2.03293592e-01
1.76384412e-02 1.92964152e-02 -3.54220718e-01 7.34833717e-01
1.54072478e-01 -3.43247801e-01 -1.26141584e+00 5.66249311e-01
-1.60044992e+00 -7.72537589e-01 -5.30432403e-01 2.31531310e+00
8.00573707e-01 -1.36828953e-02 2.42449924e-01 3.62728000e-01
1.04313040e+00 7.66761005e-02 -3.50576013e-01 -6.43477678e-01
2.21082941e-01 3.04878294e-01 6.89467847e-01 7.60072827e-01
-6.62760973e-01 7.46279001e-01 5.86670828e+00 7.74090052e-01
-9.33394551e-01 -1.47798225e-01 6.46020114e-01 -2.00383186e-01
-8.21268618e-01 2.92229861e-01 -5.09561062e-01 3.47782612e-01
1.04635668e+00 -8.95457208e-01 8.56236339e-01 5.18229842e-01
-2.71969400e-02 -2.22558573e-01 -1.19854057e+00 7.00488448e-01
-1.64037749e-01 -1.54344368e+00 1.26129478e-01 2.92213053e-01
1.15541816e+00 -2.44374275e-01 -1.20565571e-01 1.56201318e-01
4.91114736e-01 -1.32765317e+00 3.62256259e-01 3.86833288e-02
7.93945968e-01 -1.04337358e+00 5.59662461e-01 1.02628909e-01
-1.67320657e+00 1.98798627e-01 -4.79792029e-01 1.30454317e-01
1.85772721e-02 8.46329391e-01 -6.19293630e-01 8.03735554e-01
8.52648497e-01 3.83998305e-02 -5.50109804e-01 1.34606481e+00
2.45343059e-01 4.28579658e-01 -4.17128175e-01 2.28760019e-01
1.59742832e-01 -4.61763561e-01 6.41169727e-01 1.03699088e+00
1.64474115e-01 4.06440258e-01 5.08139491e-01 8.75119388e-01
-1.39544621e-01 2.70664543e-01 -6.16426170e-01 -1.16855703e-01
7.28800058e-01 1.35352778e+00 -1.27858508e+00 -2.67783374e-01
-2.47355133e-01 7.14688480e-01 5.80420554e-01 4.45832759e-01
-8.22113276e-01 -6.10887110e-01 5.65359890e-01 4.89309996e-01
1.93757549e-01 -1.94342107e-01 -5.27889848e-01 -8.93895030e-01
-7.46341720e-02 -1.05076778e+00 6.55354440e-01 -3.01969260e-01
-9.62704480e-01 8.98042321e-01 8.86121839e-02 -9.22646821e-01
1.76654562e-01 -1.19700037e-01 -5.27616441e-01 8.17775786e-01
-1.03121793e+00 -8.37754071e-01 -6.18142068e-01 4.42159593e-01
1.24630086e-01 4.11186606e-01 6.69555008e-01 6.46799877e-02
-4.94574040e-01 4.56682086e-01 2.37733901e-01 -5.03320932e-01
3.57760161e-01 -1.38092816e+00 3.65297526e-01 8.02457452e-01
5.79374610e-03 7.04960525e-01 9.54075754e-01 -8.15747321e-01
-1.87252593e+00 -1.01203513e+00 7.55499780e-01 -1.57349899e-01
7.60529697e-01 -5.10105133e-01 -1.39421952e+00 4.82571870e-01
2.78843313e-01 2.03497633e-01 5.13149917e-01 3.64450008e-01
-2.25695476e-01 -6.44987673e-02 -1.18586910e+00 4.10495192e-01
8.48585665e-01 -2.68895775e-01 -1.93574116e-01 4.18867290e-01
5.77863693e-01 -5.43646753e-01 -1.25548232e+00 3.12832892e-01
1.31802171e-01 -8.29504251e-01 9.48655665e-01 -4.48595256e-01
2.83112705e-01 -5.46276867e-01 6.68182373e-02 -1.48270559e+00
-8.71307373e-01 -7.73330808e-01 -7.29807168e-02 9.47906137e-01
3.65579426e-01 -6.42062426e-01 1.00487578e+00 6.75530195e-01
1.63467955e-02 -8.19999158e-01 -8.14530790e-01 -8.36717844e-01
2.07294375e-01 2.58576751e-01 8.77085626e-01 1.14192879e+00
5.40027618e-01 2.10562512e-01 9.69675556e-02 4.76710707e-01
8.08047473e-01 7.02800572e-01 8.43622744e-01 -1.33255434e+00
-1.19358219e-01 -5.38798511e-01 -4.15009528e-01 -4.10379946e-01
2.32130438e-01 -1.27494395e+00 -2.09549248e-01 -1.81381011e+00
5.08805692e-01 -7.58923650e-01 -1.14350252e-01 3.08476537e-01
-3.42469424e-01 4.54236895e-01 3.71414609e-02 1.42523184e-01
-4.75546032e-01 2.39733949e-01 1.10616815e+00 -2.15005606e-01
-3.92234921e-01 -3.71674567e-01 -6.55918479e-01 1.36480644e-01
8.41768801e-01 -6.68329418e-01 -4.90852773e-01 -2.62636542e-01
4.10736591e-01 2.64604986e-01 -2.51207147e-02 -9.02194142e-01
4.71649557e-01 -4.74954098e-01 -5.73181957e-02 -4.24814343e-01
-9.04973000e-02 -8.64564061e-01 8.90797019e-01 6.08713508e-01
-1.35690436e-01 5.00017285e-01 5.19510448e-01 5.13730466e-01
-1.49568900e-01 -1.69868767e-01 8.73592138e-01 7.44968727e-02
-5.26680589e-01 5.37852585e-01 -3.21041942e-01 3.00420702e-01
1.56910217e+00 -4.16393369e-01 -3.94731075e-01 -1.07157215e-01
-5.38641572e-01 5.32086968e-01 9.22842920e-01 -1.69490688e-02
4.55487758e-01 -1.08117139e+00 -9.16184068e-01 1.65081367e-01
3.72287892e-02 1.41994581e-01 3.31703007e-01 7.88994372e-01
-9.65859532e-01 2.92871296e-01 -1.49912700e-01 -5.00041127e-01
-1.29558837e+00 1.09761226e+00 -5.69749549e-02 -3.32487643e-01
-8.06858838e-01 7.34350860e-01 4.23445076e-01 -2.93703586e-01
5.94022050e-02 -2.40755469e-01 6.69145435e-02 -6.65698946e-03
3.48363340e-01 6.76979065e-01 1.99808568e-01 -4.48440582e-01
-5.92000008e-01 4.62102741e-01 -1.54981881e-01 3.05306435e-01
1.34950817e+00 -1.26584649e-01 -6.98540449e-01 -7.96512812e-02
1.12565565e+00 2.69993454e-01 -6.56214774e-01 -1.18914843e-01
-5.59184700e-02 -7.27043271e-01 7.55738914e-02 -2.39943132e-01
-1.27471876e+00 2.78202504e-01 1.33489534e-01 8.63991678e-01
1.16575515e+00 1.90522686e-01 6.17799103e-01 4.44549955e-02
3.18805426e-01 -7.77720749e-01 -2.32266679e-01 3.89960021e-01
7.88528442e-01 -8.67474258e-01 4.74241942e-01 -8.60797584e-01
-3.18615347e-01 9.23850894e-01 7.09401011e-01 -3.62516671e-01
6.76276267e-01 6.23435438e-01 -4.29269433e-01 -5.47985852e-01
-6.69761479e-01 -6.39677234e-03 2.53381044e-01 3.45008641e-01
2.65771985e-01 2.69622028e-01 -6.11526608e-01 3.20658416e-01
-5.72231293e-01 -3.61590505e-01 7.65987158e-01 7.95852542e-01
-3.68066102e-01 -1.03587818e+00 -6.40616715e-01 6.28360510e-01
-2.13159740e-01 -1.81742996e-01 -3.94325942e-01 7.11355209e-01
-1.31272256e-01 7.46813953e-01 1.86003938e-01 -3.52308750e-01
2.59835184e-01 -4.37202007e-01 3.29401851e-01 -7.77908385e-01
-7.17157781e-01 -2.75275916e-01 9.40166181e-04 -5.76959550e-01
1.70124635e-01 -2.61361629e-01 -1.49792683e+00 -1.15759885e+00
-3.11108142e-01 8.23198557e-01 2.74027884e-01 4.62157130e-01
5.01055837e-01 5.78238785e-01 5.65938830e-01 -8.51147354e-01
-3.00896227e-01 -3.94098788e-01 -8.35598886e-01 2.21967250e-01
9.24648345e-02 -4.54065382e-01 -3.96336257e-01 -1.96405351e-01] | [7.052209854125977, 5.211737155914307] |
76653b7f-b211-4392-86d1-b57d32451bfb | meemi-finding-the-middle-ground-in-cross | 1910.07221 | null | https://arxiv.org/abs/1910.07221v4 | https://arxiv.org/pdf/1910.07221v4.pdf | Meemi: A Simple Method for Post-processing and Integrating Cross-lingual Word Embeddings | Word embeddings have become a standard resource in the toolset of any Natural Language Processing practitioner. While monolingual word embeddings encode information about words in the context of a particular language, cross-lingual embeddings define a multilingual space where word embeddings from two or more languages are integrated together. Current state-of-the-art approaches learn these embeddings by aligning two disjoint monolingual vector spaces through an orthogonal transformation which preserves the structure of the monolingual counterparts. In this work, we propose to apply an additional transformation after this initial alignment step, which aims to bring the vector representations of a given word and its translations closer to their average. Since this additional transformation is non-orthogonal, it also affects the structure of the monolingual spaces. We show that our approach both improves the integration of the monolingual spaces as well as the quality of the monolingual spaces themselves. Furthermore, because our transformation can be applied to an arbitrary number of languages, we are able to effectively obtain a truly multilingual space. The resulting (monolingual and multilingual) spaces show consistent gains over the current state-of-the-art in standard intrinsic tasks, namely dictionary induction and word similarity, as well as in extrinsic tasks such as cross-lingual hypernym discovery and cross-lingual natural language inference. | ['Steven Schockaert', 'Luis Espinosa-Anke', 'Jose Camacho-Collados', 'Yerai Doval'] | 2019-10-16 | null | null | null | null | ['cross-lingual-natural-language-inference', 'hypernym-discovery'] | ['natural-language-processing', 'natural-language-processing'] | [-2.16825426e-01 -3.15848351e-01 -4.75186318e-01 -2.25680545e-01
-6.89654768e-01 -1.02921712e+00 8.81339550e-01 7.70665288e-01
-9.05247569e-01 3.63492668e-01 4.76675302e-01 -4.58199650e-01
1.34943798e-01 -7.84501195e-01 -5.70026338e-01 -4.86469597e-01
2.67509192e-01 6.21708393e-01 -1.09122276e-01 -5.49852490e-01
-1.63548723e-01 2.49729961e-01 -1.17030478e+00 -1.88004538e-01
9.16362345e-01 3.83832574e-01 1.35392085e-01 2.57219732e-01
-4.45435911e-01 8.75095055e-02 -1.02516852e-01 -6.71161115e-01
3.97786409e-01 -2.01222345e-01 -7.63404012e-01 -4.67720538e-01
6.80674970e-01 2.63164192e-01 -4.41305071e-01 1.28020060e+00
3.23297083e-01 4.35220189e-02 5.26954532e-01 -8.79733145e-01
-1.05581439e+00 8.01599622e-01 -3.22502047e-01 1.33041233e-01
3.04184854e-01 -3.32130454e-02 1.75611591e+00 -1.06613708e+00
8.09852421e-01 1.25997913e+00 6.64604187e-01 6.00710586e-02
-1.51178765e+00 -5.58875620e-01 1.63059101e-01 2.89915144e-01
-1.51408827e+00 -2.91368067e-01 4.59027618e-01 -6.04960024e-01
1.12099135e+00 -8.53060558e-02 5.60671270e-01 9.95184958e-01
2.66420096e-01 3.33862036e-01 9.71714377e-01 -8.59816611e-01
-1.31273821e-01 3.36140901e-01 3.93793613e-01 6.37999058e-01
5.16879261e-01 2.03590263e-02 -3.21736842e-01 -9.90838036e-02
4.23499703e-01 -9.90841538e-02 -1.49905264e-01 -7.27713466e-01
-1.62310112e+00 1.09799433e+00 2.70811081e-01 8.51832211e-01
-1.82062626e-01 -2.00067833e-02 6.26814604e-01 3.50222230e-01
4.12275881e-01 7.93267667e-01 -3.48312795e-01 1.86811499e-02
-6.00177705e-01 3.72830853e-02 6.48050725e-01 7.73477733e-01
9.60483611e-01 -7.39904195e-02 2.70928815e-02 1.06012785e+00
2.85848498e-01 6.30707204e-01 8.06461573e-01 -2.06296161e-01
3.72676253e-01 5.37510693e-01 -3.84345412e-01 -8.82136583e-01
-1.95108131e-01 -4.18733478e-01 -4.80805576e-01 -1.66529983e-01
3.94117981e-01 1.50048062e-01 -6.42408848e-01 2.13623261e+00
3.48638505e-01 1.52994776e-02 2.25730747e-01 5.77517033e-01
5.38874805e-01 5.32455981e-01 1.59482490e-02 2.03196734e-01
1.59503388e+00 -8.13985407e-01 -7.39021778e-01 -5.20311117e-01
8.17486227e-01 -1.03301346e+00 1.34816432e+00 -2.92622093e-02
-7.10198641e-01 -5.16023338e-01 -1.26436317e+00 -5.85450292e-01
-8.58517349e-01 -1.80306062e-01 4.97560441e-01 4.69092250e-01
-7.16933131e-01 2.90479243e-01 -6.35207891e-01 -5.82871377e-01
-1.16990507e-01 1.12021595e-01 -9.21283841e-01 -3.95485848e-01
-1.57944274e+00 1.51045072e+00 4.73391891e-01 -4.17828053e-01
-4.15565789e-01 -8.45527470e-01 -1.31113446e+00 -1.59580767e-01
1.86966792e-01 -4.61970001e-01 7.91446030e-01 -4.79372144e-01
-9.68873143e-01 1.08545995e+00 -1.56961098e-01 -3.28876615e-01
3.48456316e-02 -1.58179954e-01 -5.52292645e-01 -3.66801888e-01
4.04912531e-01 4.61374342e-01 4.59263235e-01 -1.00694799e+00
-5.56866825e-01 -4.53258544e-01 1.97751746e-01 3.30793351e-01
-6.72916532e-01 -1.20780565e-01 -5.03510416e-01 -8.04659247e-01
-1.25168398e-01 -1.06314480e+00 -1.29024908e-01 -2.00019658e-01
-6.77119121e-02 -3.32703292e-01 3.16492587e-01 -6.27014041e-01
1.31733119e+00 -2.15702724e+00 5.15659928e-01 2.06794307e-01
2.01506093e-01 1.77104354e-01 -5.36024332e-01 6.80988133e-01
-3.02406967e-01 1.20485559e-01 -2.34911308e-01 -4.13856834e-01
2.51515716e-01 7.58931220e-01 -1.87234595e-01 6.96522295e-01
2.39986971e-01 8.50616157e-01 -1.16857278e+00 -3.80638719e-01
2.34611586e-01 6.80448592e-01 -5.79992414e-01 1.32479379e-02
2.24451154e-01 1.62725717e-01 1.82918474e-01 1.59696892e-01
3.33811611e-01 3.29530776e-01 5.47955096e-01 -3.85812670e-01
-1.10308789e-01 7.12833047e-01 -1.18496025e+00 2.02134371e+00
-1.14447916e+00 6.36968136e-01 -4.16795403e-01 -9.42047954e-01
9.16002989e-01 3.15298200e-01 4.11032140e-01 -6.25032902e-01
1.77832413e-02 4.95488316e-01 2.85553455e-01 -1.83220759e-01
6.97231948e-01 -4.75449413e-01 -3.38196218e-01 5.74066758e-01
6.62270665e-01 -2.68580317e-02 4.86737788e-01 9.15078595e-02
6.95382118e-01 -9.29217711e-02 7.38147199e-01 -6.33388758e-01
5.25368094e-01 -2.45082036e-01 4.93812829e-01 3.22144777e-01
2.28141874e-01 2.73859531e-01 2.94484407e-01 -3.86984766e-01
-1.27891612e+00 -1.48742068e+00 -4.84110504e-01 1.28363562e+00
8.99305344e-02 -7.98659325e-01 -3.92390192e-01 -7.23367035e-01
2.67590523e-01 7.56201327e-01 -6.27794147e-01 -3.30655456e-01
-7.09496379e-01 -6.12865746e-01 7.21847475e-01 4.99990940e-01
-3.11882347e-01 -6.73260689e-01 5.70559949e-02 2.30062589e-01
-1.29753858e-01 -1.40174055e+00 -1.00661409e+00 4.67603654e-01
-5.24226248e-01 -9.82303858e-01 -2.38449991e-01 -1.06755543e+00
5.45486987e-01 -3.96117708e-03 1.24196196e+00 -2.61756867e-01
-3.04834694e-01 1.25473887e-01 -2.93922752e-01 -1.42515242e-01
-5.33853292e-01 2.59956807e-01 5.34168303e-01 1.62876025e-01
7.10279703e-01 -6.91187203e-01 1.42369617e-03 -1.19261160e-01
-1.26954496e+00 -4.80383307e-01 3.64501774e-01 1.04883349e+00
7.09444225e-01 -3.84512484e-01 3.69972318e-01 -7.81606138e-01
7.90930927e-01 -4.75965768e-01 -4.94839847e-01 3.90710920e-01
-7.24703193e-01 7.86347628e-01 7.58758903e-01 -6.71752334e-01
-3.06310415e-01 -1.58381045e-01 -9.70144793e-02 -1.99712455e-01
9.27152336e-02 7.14995980e-01 -1.95529193e-01 6.92456886e-02
6.65536761e-01 1.04369104e-01 -1.70935899e-01 -5.75780272e-01
1.09987152e+00 6.42722249e-01 7.19621837e-01 -6.71311319e-01
9.46071863e-01 8.44396725e-02 -2.20417276e-01 -8.04557264e-01
-7.11353004e-01 -7.40472496e-01 -1.10613155e+00 3.45964968e-01
9.34127629e-01 -8.96977842e-01 -2.03686908e-01 -3.17185670e-02
-1.28576887e+00 3.30010593e-01 -5.28433800e-01 7.14163661e-01
-7.50812218e-02 4.21278268e-01 -2.50748187e-01 -1.25489816e-01
-1.92684710e-01 -1.34401417e+00 9.81585383e-01 -3.09558690e-01
-6.61131084e-01 -1.62439966e+00 6.89313054e-01 -1.12084016e-01
2.22577155e-01 -5.65524958e-02 1.45047176e+00 -1.01331937e+00
5.73219843e-02 -2.29845941e-01 5.38652577e-02 5.75454772e-01
3.62703323e-01 -1.69115916e-01 -6.36397302e-01 -4.45848048e-01
-2.84704477e-01 -3.15183103e-02 7.18357980e-01 -2.53152341e-01
2.54180938e-01 -2.37332582e-01 -8.65777731e-02 7.02474833e-01
1.56680429e+00 -1.15070358e-01 2.27121934e-01 4.25322682e-01
9.80958164e-01 5.37638605e-01 1.04692742e-01 -1.05734155e-01
6.16241336e-01 9.00997937e-01 2.10274570e-02 -1.71654105e-01
-1.51067585e-01 -4.50824350e-01 5.64185679e-01 1.69279456e+00
3.39871138e-01 2.29201317e-01 -1.11337066e+00 1.11001849e+00
-1.62669015e+00 -5.85332870e-01 1.07683912e-01 2.57653308e+00
1.18173945e+00 -1.17651559e-01 -2.13495791e-01 1.46581857e-02
3.80825609e-01 3.04159641e-01 -2.48207361e-01 -6.92402482e-01
-3.37905705e-01 8.49772394e-01 6.71901226e-01 8.70713711e-01
-9.90364552e-01 1.20053482e+00 5.85560417e+00 7.21499085e-01
-1.19989216e+00 5.06125450e-01 -3.14884961e-01 1.45081922e-01
-8.21229994e-01 2.31882170e-01 -8.37958097e-01 1.65405378e-01
8.71696889e-01 -6.17021561e-01 5.24574816e-01 6.02269471e-01
-3.66496384e-01 3.34593117e-01 -1.58660817e+00 9.43354964e-01
2.67708749e-01 -1.14661527e+00 3.74256194e-01 4.95333709e-02
7.39413977e-01 3.63172978e-01 -4.09462564e-02 4.54789788e-01
3.80862266e-01 -1.16013622e+00 5.94264746e-01 2.63170656e-02
9.90713358e-01 -8.65549564e-01 7.07393885e-01 -4.11057612e-03
-1.32611942e+00 2.93259263e-01 -2.96462446e-01 2.81123757e-01
1.47749662e-01 5.97501576e-01 -6.05770171e-01 7.12499678e-01
2.17762694e-01 8.46138358e-01 -6.46225095e-01 5.69425523e-01
-4.80938166e-01 1.81125656e-01 -2.25108996e-01 1.87732399e-01
4.47936267e-01 -4.96140391e-01 5.72806180e-01 1.48942780e+00
2.10951671e-01 -6.85595512e-01 3.66674751e-01 6.95406854e-01
-2.92870224e-01 6.80421174e-01 -9.31022763e-01 -2.88968414e-01
5.46036422e-01 1.08480203e+00 -2.94607226e-02 -2.84250826e-01
-6.17555439e-01 9.06623363e-01 6.47819698e-01 7.40758330e-02
-6.22859955e-01 -6.57796562e-01 1.42072809e+00 -9.43190008e-02
1.74461335e-01 -6.57586336e-01 -1.31910920e-01 -1.45468044e+00
1.57086268e-01 -1.03944147e+00 2.11727574e-01 -1.43315762e-01
-1.54108131e+00 8.65081191e-01 -5.81087582e-02 -1.07479239e+00
-3.49445730e-01 -8.02624941e-01 -3.06744367e-01 1.01246560e+00
-1.57811725e+00 -1.20269632e+00 2.76609600e-01 5.18082082e-01
3.05271119e-01 -3.40702057e-01 1.29492068e+00 5.94833672e-01
-2.81549901e-01 9.18682575e-01 3.89480740e-01 3.71423960e-01
1.05235839e+00 -1.36808777e+00 6.03611827e-01 9.21404481e-01
1.00117064e+00 1.16253114e+00 5.01118243e-01 -3.34360123e-01
-1.50774491e+00 -9.67707157e-01 1.55098808e+00 -5.51566243e-01
1.26459253e+00 -6.22909844e-01 -9.76919770e-01 8.16426992e-01
3.31008404e-01 2.06094310e-01 9.17759418e-01 6.93332553e-01
-1.15973532e+00 -1.25056952e-01 -5.15217125e-01 7.89424002e-01
8.29031348e-01 -1.26049793e+00 -8.54087293e-01 3.83133620e-01
8.28423679e-01 -6.23671226e-02 -1.18556952e+00 3.06211114e-01
7.07329810e-01 -3.87011975e-01 9.72154319e-01 -9.16694582e-01
2.35134318e-01 -3.40251654e-01 -5.53725839e-01 -1.73053217e+00
-1.78093702e-01 -3.76787663e-01 2.98090339e-01 1.16522741e+00
5.83841205e-01 -7.74125934e-01 -2.76385006e-02 -1.76099420e-01
2.57684477e-02 -7.19332993e-01 -1.04507697e+00 -1.13115108e+00
6.99727416e-01 -5.73328614e-01 5.64511716e-01 1.41022718e+00
2.95726031e-01 8.85041237e-01 -9.45521444e-02 1.46380618e-01
4.51939166e-01 -5.18883020e-02 5.01779795e-01 -1.20060194e+00
-1.74011201e-01 -4.94097680e-01 -6.91777885e-01 -8.89388025e-01
5.54104209e-01 -1.71700549e+00 -7.21914321e-02 -1.30472922e+00
1.02005325e-01 -4.89895314e-01 -5.68598092e-01 4.71264392e-01
-2.51775175e-01 2.16864958e-01 2.64869660e-01 3.08390055e-02
-1.12162821e-01 5.87323904e-01 7.58321464e-01 -1.36720061e-01
2.05704309e-02 -7.76034057e-01 -7.17103422e-01 6.33878887e-01
5.48743784e-01 -6.10863030e-01 -2.09886894e-01 -8.46722007e-01
1.87459767e-01 -6.09315813e-01 -5.75906076e-02 -6.05349123e-01
6.07467853e-02 2.92554758e-02 -2.38653809e-01 1.01981610e-01
2.91257799e-01 -7.45512366e-01 -5.46077229e-02 1.47632897e-01
-3.24706525e-01 5.15764058e-01 1.76324278e-01 3.28460306e-01
-4.87549394e-01 -2.87976563e-01 7.94308007e-01 1.55091465e-01
-4.80561644e-01 1.96008086e-01 -1.16782844e-01 3.75392884e-01
9.25190032e-01 1.14092402e-01 -1.66731954e-01 3.65503356e-02
-4.77827191e-01 4.22445834e-02 5.97563088e-01 1.01193440e+00
1.66739494e-01 -1.73696065e+00 -7.68582761e-01 4.40182805e-01
6.31916106e-01 -5.07161140e-01 -4.02183115e-01 8.56429696e-01
-2.82603025e-01 5.58369100e-01 -2.79366553e-01 -4.15056169e-01
-1.20068753e+00 5.24245679e-01 2.26371735e-01 -5.36986589e-01
-3.36018622e-01 4.38098311e-01 2.28262290e-01 -1.14477372e+00
1.43514574e-01 -4.50796783e-01 -1.24887042e-01 3.57582599e-01
2.29158536e-01 -8.20385292e-02 2.63273001e-01 -1.12276936e+00
-5.24584770e-01 9.11890626e-01 -1.34140939e-01 -2.19567716e-01
1.23747265e+00 1.87436212e-02 -3.92994076e-01 8.81716549e-01
1.59381223e+00 5.32136619e-01 -2.69194573e-01 -6.48264110e-01
2.99137503e-01 -3.82569581e-01 -2.81224698e-02 -2.75590837e-01
-7.25888908e-01 9.30415511e-01 4.28234726e-01 -9.80762467e-02
5.97687483e-01 1.02678470e-01 8.98890734e-01 3.61711711e-01
2.31200144e-01 -8.71558011e-01 -3.02473575e-01 9.29333270e-01
6.48115098e-01 -1.13042903e+00 -7.44912624e-02 -1.46615297e-01
-4.45544094e-01 1.01546764e+00 1.55592158e-01 -1.19582713e-01
7.26504624e-01 1.71986103e-01 2.21339166e-01 1.11316312e-02
-5.70491314e-01 -4.10772860e-01 6.27344310e-01 4.99668121e-01
8.05788457e-01 2.21433446e-01 -6.45984113e-01 4.44784880e-01
-4.34029162e-01 -6.38958216e-01 2.79006839e-01 5.47549427e-01
-1.33239403e-01 -1.90494227e+00 -2.14486003e-01 9.07946676e-02
-3.47331136e-01 -6.58259213e-01 -2.49260738e-01 8.73462260e-01
3.80293339e-01 5.77041447e-01 8.27457085e-02 -4.02918339e-01
3.78895640e-01 2.60559708e-01 5.78866720e-01 -1.00641072e+00
-4.88074243e-01 -2.98577458e-01 2.85806451e-02 -4.24618155e-01
-1.18469343e-01 -5.31676471e-01 -1.05575597e+00 -1.96832895e-01
-2.97082126e-01 2.04277053e-01 8.03198040e-01 1.05454373e+00
1.62331671e-01 3.73852879e-01 3.62736821e-01 -3.55799377e-01
-5.80616057e-01 -8.90606463e-01 -4.18984532e-01 7.03351259e-01
1.96200326e-01 -5.81560552e-01 -1.65143847e-01 2.59708869e-03] | [11.011564254760742, 9.931183815002441] |
0b2f3b41-b80f-4946-9b37-dede3c4bf65e | quantifying-sample-anonymity-in-score-based | 2306.01363 | null | https://arxiv.org/abs/2306.01363v1 | https://arxiv.org/pdf/2306.01363v1.pdf | Quantifying Sample Anonymity in Score-Based Generative Models with Adversarial Fingerprinting | Recent advances in score-based generative models have led to a huge spike in the development of downstream applications using generative models ranging from data augmentation over image and video generation to anomaly detection. Despite publicly available trained models, their potential to be used for privacy preserving data sharing has not been fully explored yet. Training diffusion models on private data and disseminating the models and weights rather than the raw dataset paves the way for innovative large-scale data-sharing strategies, particularly in healthcare, where safeguarding patients' personal health information is paramount. However, publishing such models without individual consent of, e.g., the patients from whom the data was acquired, necessitates guarantees that identifiable training samples will never be reproduced, thus protecting personal health data and satisfying the requirements of policymakers and regulatory bodies. This paper introduces a method for estimating the upper bound of the probability of reproducing identifiable training images during the sampling process. This is achieved by designing an adversarial approach that searches for anatomic fingerprints, such as medical devices or dermal art, which could potentially be employed to re-identify training images. Our method harnesses the learned score-based model to estimate the probability of the entire subspace of the score function that may be utilized for one-to-one reproduction of training samples. To validate our estimates, we generate anomalies containing a fingerprint and investigate whether generated samples from trained generative models can be uniquely mapped to the original training samples. Overall our results show that privacy-breaching images are reproduced at sampling time if the models were trained without care. | ['Bernhard Kainz', 'Mischa Dombrowski'] | 2023-06-02 | null | null | null | null | ['video-generation'] | ['computer-vision'] | [ 9.31934953e-01 7.14122057e-01 -9.03137997e-02 -1.82822183e-01
-8.86861742e-01 -9.52548444e-01 4.25656706e-01 2.85129324e-02
-3.12099904e-01 7.63974011e-01 1.29142076e-01 -3.42530966e-01
2.06097569e-02 -8.92141700e-01 -9.53987598e-01 -8.69700015e-01
-3.68619747e-02 3.17396194e-01 -3.50025713e-01 4.00428206e-01
-1.39286697e-01 5.67138076e-01 -1.14899480e+00 4.24444795e-01
6.01300776e-01 8.95124137e-01 -2.00993985e-01 8.25289547e-01
8.96281824e-02 2.93305457e-01 -7.71376610e-01 -9.09134269e-01
6.84293985e-01 -6.73159540e-01 -2.20443085e-01 -6.52189702e-02
3.62127542e-01 -7.66907752e-01 -3.63810331e-01 9.93165433e-01
4.75937486e-01 -4.43190604e-01 5.09846926e-01 -1.15403700e+00
-7.44977057e-01 3.87119532e-01 -3.21509391e-01 -3.13673049e-01
2.83164948e-01 2.37393767e-01 4.77076352e-01 -4.15790051e-01
9.75782096e-01 5.88612258e-01 5.80904961e-01 9.84763265e-01
-1.36847353e+00 -8.30730736e-01 -4.41490889e-01 -4.68938231e-01
-1.35948956e+00 -5.45726776e-01 5.45876265e-01 -4.26178277e-01
1.26340523e-01 7.89183378e-01 8.46891046e-01 1.62286282e+00
1.56752869e-01 5.00041485e-01 9.09693420e-01 -2.09877625e-01
1.75968766e-01 5.19351482e-01 -6.39176786e-01 7.37591207e-01
6.21905148e-01 2.10205048e-01 -6.32238388e-01 -7.12635815e-01
1.00810528e+00 1.11383915e-01 -3.26775491e-01 -6.16073132e-01
-9.79834974e-01 6.13997817e-01 -1.43992538e-02 1.29389226e-01
-4.41566676e-01 1.96330443e-01 1.00035578e-01 1.82284132e-01
3.83359432e-01 3.96156043e-01 1.12410240e-01 1.42876506e-01
-1.00369525e+00 2.38558993e-01 5.62164187e-01 9.58256602e-01
4.44926530e-01 -2.04549134e-01 -3.18372905e-01 3.97538543e-01
1.58156708e-01 6.23725533e-01 3.18171620e-01 -8.85300696e-01
2.95371860e-01 5.93807459e-01 2.12764993e-01 -9.54655707e-01
2.77546555e-01 -4.43873525e-01 -9.30647075e-01 1.49752885e-01
5.29508591e-01 -3.67195815e-01 -8.74228179e-01 1.83863449e+00
3.24180305e-01 2.36746922e-01 5.88854626e-02 5.25987208e-01
2.28981808e-01 2.67705590e-01 4.29746434e-02 -2.34043747e-02
1.11078203e+00 -1.54923081e-01 -4.38402832e-01 1.47650406e-01
4.83979851e-01 -5.10747790e-01 5.96676230e-01 3.18962812e-01
-1.06627619e+00 -1.01791866e-01 -9.76549447e-01 2.75754243e-01
-1.63622528e-01 -1.02875054e-01 5.50130427e-01 1.26586163e+00
-9.81057286e-01 5.17248094e-01 -8.20129573e-01 -3.06311101e-01
1.17262733e+00 5.35509229e-01 -5.73452353e-01 -2.10090220e-01
-9.86859143e-01 2.53163785e-01 -1.46927416e-01 -1.44579820e-02
-9.63685095e-01 -9.61245954e-01 -6.01691902e-01 -1.90274697e-03
5.20708077e-02 -9.71580744e-01 5.96580386e-01 -9.33911383e-01
-1.08389997e+00 1.10033846e+00 5.52010946e-02 -6.62577033e-01
1.11457205e+00 9.00973454e-02 -4.98507023e-01 -7.84700084e-03
6.53562322e-02 4.87089097e-01 1.08045995e+00 -1.37445390e+00
-4.23043251e-01 -5.22126019e-01 -3.94001663e-01 -3.03842306e-01
-5.36726594e-01 -3.44956577e-01 -2.23892242e-01 -7.29498804e-01
-2.88636237e-01 -9.96729016e-01 -2.65301138e-01 4.81877297e-01
-6.31893992e-01 7.35590518e-01 6.56330705e-01 -7.20832944e-01
9.69845474e-01 -2.37263989e+00 -2.32103273e-01 7.95037746e-01
2.26385027e-01 3.55027884e-01 -5.75002469e-02 3.55537742e-01
3.05401146e-01 4.93339330e-01 -3.73199940e-01 -3.47515732e-01
-3.61511767e-01 1.15659967e-01 -4.83894140e-01 5.43758392e-01
2.49273747e-01 9.19312119e-01 -7.41755188e-01 -2.64509797e-01
-7.60271177e-02 7.25918651e-01 -6.22490644e-01 2.72598028e-01
-7.73928016e-02 9.12153602e-01 -5.28810680e-01 5.99634945e-01
6.24737442e-01 -2.78884798e-01 3.48053724e-01 1.31281585e-01
4.30860549e-01 -3.86783153e-01 -8.37831378e-01 1.67533231e+00
-1.38269424e-01 3.73660862e-01 4.79089208e-02 -4.62673992e-01
7.75890827e-01 3.44165832e-01 6.99672401e-01 -3.78212929e-01
-3.24776210e-02 1.30051598e-01 -1.08289056e-01 -3.71676505e-01
2.06467181e-01 -2.77135879e-01 -1.75576899e-02 5.75544059e-01
-2.15317756e-01 3.00500482e-01 -3.78616422e-01 1.98648438e-01
1.25761294e+00 -3.87936970e-03 -1.64040148e-01 1.84783101e-01
7.80897141e-02 -7.19879642e-02 3.48979980e-01 8.40558171e-01
-6.79665571e-03 8.74475360e-01 4.47695524e-01 -2.31362790e-01
-1.30937028e+00 -1.24551284e+00 -3.01331431e-01 5.30859292e-01
-1.18565373e-01 -1.41469583e-01 -8.54008675e-01 -7.68409789e-01
3.42245787e-01 5.85918248e-01 -9.05152738e-01 -3.87961417e-01
-2.98062891e-01 -4.88226146e-01 1.05390012e+00 2.11733475e-01
1.78409487e-01 -6.66964710e-01 -6.75449550e-01 2.29573861e-01
7.67024085e-02 -7.96051443e-01 -6.96522951e-01 -3.88723105e-01
-6.57870054e-01 -1.20480835e+00 -1.02875543e+00 -2.19218895e-01
1.18057573e+00 -2.49213353e-01 5.32753885e-01 -6.72466448e-03
-6.79646790e-01 6.96654260e-01 -2.18837820e-02 -7.42670000e-01
-9.08964753e-01 -1.66204661e-01 -1.16701119e-01 5.77462554e-01
2.77329888e-02 -5.41787922e-01 -9.44951534e-01 2.23902091e-01
-1.12627304e+00 -1.25665367e-01 5.55609703e-01 6.12435162e-01
6.49915755e-01 -4.77662176e-01 6.38118863e-01 -1.29613304e+00
6.46995068e-01 -6.40213668e-01 -4.89185125e-01 3.94794762e-01
-6.93881512e-01 1.39766842e-01 4.39769328e-01 -4.61347550e-01
-9.06718135e-01 1.94258317e-01 3.69222611e-02 -6.63321614e-01
-4.02975380e-02 1.59460619e-01 -5.79030067e-02 -2.28198305e-01
6.66264534e-01 3.16569805e-01 4.62671667e-01 -2.49664158e-01
4.02301729e-01 6.50818944e-01 6.61664248e-01 -4.55292940e-01
7.07071066e-01 8.28501284e-01 1.95408329e-01 -5.95301926e-01
-2.46124417e-01 -2.05425799e-01 -4.90604527e-02 -2.54452497e-01
7.48452783e-01 -6.41957998e-01 -6.07449055e-01 2.93090761e-01
-8.26330483e-01 1.28786996e-01 -6.75617158e-01 2.06039846e-01
-4.63617623e-01 2.96596199e-01 -2.07575828e-01 -1.02269447e+00
-3.67422104e-01 -7.98289955e-01 1.02249002e+00 -2.10426077e-02
-4.51742530e-01 -6.90872252e-01 1.78198889e-01 5.75294852e-01
6.29908860e-01 7.59358943e-01 9.23222601e-01 -6.49268687e-01
-9.35020745e-01 -8.08197498e-01 1.97477862e-01 3.58676106e-01
2.66728163e-01 -1.00489512e-01 -1.01885808e+00 -3.60458165e-01
2.16377657e-02 -1.13938712e-01 4.76259977e-01 2.49950230e-01
1.33353305e+00 -8.40385735e-01 -5.51284552e-01 7.86849141e-01
1.08178222e+00 1.69200644e-01 9.76315200e-01 -9.10398364e-02
5.53454161e-01 6.53270960e-01 9.45716426e-02 6.22002959e-01
-9.40393135e-02 2.40659595e-01 3.49363953e-01 -7.84234554e-02
1.43215090e-01 -6.82132840e-01 1.03786021e-01 3.07744127e-02
-6.09090738e-02 -3.25721562e-01 -6.17478430e-01 5.98749757e-01
-1.48077190e+00 -1.04682851e+00 2.22905248e-01 2.69408989e+00
8.27039540e-01 -1.52663752e-01 6.19948916e-02 -2.61257052e-01
7.20730305e-01 -1.66369289e-01 -8.68466556e-01 -1.65638208e-01
-2.03683469e-02 3.89271259e-01 8.41693401e-01 7.83172175e-02
-6.92927003e-01 3.31625819e-01 6.23968506e+00 4.16195422e-01
-1.00832307e+00 4.57286723e-02 1.00816762e+00 -4.32072997e-01
-7.71717429e-01 -7.99157619e-02 -3.68446857e-01 5.04620969e-01
9.98972774e-01 -3.99957478e-01 2.96120793e-01 6.41828537e-01
2.50651594e-02 1.30646050e-01 -1.17474282e+00 7.70595908e-01
-4.01207022e-02 -1.57062781e+00 2.58355558e-01 6.87468827e-01
6.91129625e-01 -3.77305210e-01 5.59286058e-01 -3.41341764e-01
3.60906720e-01 -1.10767269e+00 4.34764326e-01 9.46088552e-01
1.26195323e+00 -7.40349591e-01 5.36178350e-01 2.12375537e-01
-4.45108950e-01 1.30256921e-01 -9.88923833e-02 7.51127362e-01
5.61420098e-02 5.47047555e-01 -1.13703585e+00 4.23039019e-01
3.09661239e-01 1.02952763e-01 -3.36478859e-01 7.68111467e-01
1.67663008e-01 5.86110234e-01 -4.34464693e-01 3.23550820e-01
-2.74311453e-01 -4.88147587e-02 5.46493471e-01 7.44788289e-01
8.08725059e-01 -1.55684292e-01 -4.53155011e-01 1.13713813e+00
-3.92196417e-01 -1.07049160e-02 -9.25528228e-01 -4.79378700e-01
3.25230777e-01 8.61797810e-01 -2.50819057e-01 8.84351730e-02
-2.06161290e-01 1.17361081e+00 -1.88891396e-01 2.81653136e-01
-6.58431232e-01 -1.15356468e-01 7.44663119e-01 7.49831796e-01
9.77222398e-02 1.80350095e-01 -2.85845906e-01 -8.34717810e-01
1.58504993e-01 -8.30417454e-01 3.20717007e-01 -4.67964232e-01
-1.20015299e+00 2.13298261e-01 -2.70776898e-01 -1.18568003e+00
-4.91093606e-01 -1.53120980e-01 -3.33464652e-01 1.02890325e+00
-8.68020177e-01 -1.28867471e+00 -1.11939177e-01 5.74747443e-01
-3.00361753e-01 -3.07813823e-01 1.15072203e+00 1.14120707e-01
-2.26625606e-01 1.04632401e+00 3.76699954e-01 3.27451080e-01
6.19368970e-01 -7.20859170e-01 3.59327257e-01 9.22184110e-01
2.65506148e-01 9.35426176e-01 4.12120253e-01 -9.23060417e-01
-1.38253427e+00 -1.17093885e+00 5.26126385e-01 -6.91990256e-01
2.43252560e-01 -6.21484637e-01 -8.18967283e-01 7.47851610e-01
-2.29254365e-01 1.31325811e-01 1.17819405e+00 -3.59193772e-01
-2.72248179e-01 -1.71792775e-01 -1.65536761e+00 5.75699508e-01
9.40558255e-01 -6.68690205e-01 3.32816005e-01 1.01557292e-01
3.06230456e-01 -2.89987683e-01 -9.59809422e-01 1.49853721e-01
9.42866325e-01 -8.61504436e-01 8.20871711e-01 -7.44877100e-01
3.88675183e-01 -5.19034564e-02 5.52210212e-02 -8.38749945e-01
1.51123151e-01 -9.28652287e-01 -1.78615630e-01 1.38174212e+00
6.03785515e-01 -7.42053092e-01 1.29407752e+00 1.25953448e+00
3.50309908e-01 -5.81070304e-01 -1.18825245e+00 -6.50308251e-01
-1.16264321e-01 -3.50025058e-01 8.51577222e-01 9.16391313e-01
-3.88096571e-01 -5.17274976e-01 -6.99028432e-01 1.67967618e-01
8.75327110e-01 -3.02013814e-01 8.77924621e-01 -9.08032179e-01
-4.21959698e-01 -5.32929264e-02 -6.46612704e-01 -4.24911886e-01
-2.51356333e-01 -9.35879111e-01 -3.59225780e-01 -1.12099457e+00
2.02583298e-01 -7.24121392e-01 -2.13501498e-01 4.22649294e-01
5.97666092e-02 2.04742864e-01 -1.04054355e-03 4.20858145e-01
2.01485872e-01 1.13037653e-01 1.27434134e+00 -1.13286927e-01
-5.66885248e-02 2.62151927e-01 -1.03035796e+00 1.80667713e-01
6.62627161e-01 -7.40262985e-01 -6.02664530e-01 -3.87200713e-02
3.31405431e-01 1.43988952e-01 7.04956353e-01 -9.58689690e-01
1.70714453e-01 -3.22667882e-02 5.59053361e-01 -4.74989265e-02
8.77976269e-02 -1.08521402e+00 1.01564240e+00 5.37937820e-01
-5.13655961e-01 -3.31332326e-01 1.32156117e-03 1.04913855e+00
1.33364081e-01 -5.00537232e-02 5.61392248e-01 -1.41502142e-01
1.68705136e-02 4.77166235e-01 -1.32769853e-01 -1.46107569e-01
1.35164094e+00 -4.62137848e-01 -2.62354106e-01 -5.59859216e-01
-8.89443576e-01 -1.05042532e-01 7.90297329e-01 2.18954235e-01
6.73361123e-01 -1.15467787e+00 -7.55901575e-01 7.25894272e-01
2.42485195e-01 2.11056527e-02 4.46490049e-01 4.61275756e-01
-3.02854806e-01 -7.51888473e-03 -1.47596583e-01 -4.40064669e-01
-1.13246584e+00 5.26550889e-01 2.13767558e-01 -8.67726058e-02
-5.24909914e-01 6.14305258e-01 1.53120905e-01 1.65240243e-02
1.03454843e-01 5.86912669e-02 5.35078824e-01 -2.64605045e-01
4.69716489e-01 1.42576113e-01 -1.26268640e-02 -2.97351688e-01
-1.45542204e-01 3.60806659e-02 -2.36423627e-01 -2.24276528e-01
1.23397851e+00 4.36362326e-02 2.41713718e-01 -3.18275131e-02
1.22574127e+00 3.41154724e-01 -1.39971948e+00 1.39590904e-01
-4.32297677e-01 -7.87911773e-01 -3.73361051e-01 -8.64889801e-01
-1.30285919e+00 5.73684573e-01 8.05785418e-01 1.29299492e-01
8.92596900e-01 1.99683812e-02 7.43060589e-01 -1.38076931e-01
6.51418149e-01 -6.26948714e-01 -2.31059864e-01 -5.41249216e-01
7.35327244e-01 -9.44578707e-01 -1.22010335e-01 -2.54967958e-01
-5.95871627e-01 6.79268599e-01 -3.28293256e-02 1.24368683e-01
4.78657931e-01 3.71315598e-01 4.28283140e-02 -5.51864244e-02
-3.33998829e-01 5.36011517e-01 2.43524253e-01 1.08639526e+00
7.62272701e-02 2.64141768e-01 -1.85097083e-02 5.86968184e-01
-1.95225403e-01 2.59970933e-01 4.99988258e-01 1.00425684e+00
1.41046137e-01 -1.49473381e+00 -2.69272923e-01 8.88338447e-01
-6.92489684e-01 3.74777964e-03 -5.75455606e-01 4.68127310e-01
1.73109263e-01 4.82226908e-01 -2.28806399e-02 -1.71799198e-01
2.62511671e-01 1.86106220e-01 4.27564502e-01 -4.72216994e-01
-6.21882856e-01 -2.83388793e-01 -1.96713805e-02 -4.98165190e-01
-1.86565280e-01 -8.47633839e-01 -8.25207353e-01 -3.58605355e-01
-4.24875841e-02 -1.79668754e-01 7.72889376e-01 4.33314711e-01
8.86306345e-01 1.77269325e-01 5.56383312e-01 -2.09211946e-01
-5.45637012e-01 -2.71623522e-01 -6.08383000e-01 5.73778629e-01
4.41352218e-01 -1.55988745e-02 -2.17045069e-01 2.24989802e-01] | [6.129668712615967, 6.875210762023926] |
eaa3b048-5cd6-481e-b29a-fdde2cfab70e | parallelised-diffeomorphic-sampling-based | 2108.11775 | null | https://arxiv.org/abs/2108.11775v2 | https://arxiv.org/pdf/2108.11775v2.pdf | Parallelised Diffeomorphic Sampling-based Motion Planning | We propose Parallelised Diffeomorphic Sampling-based Motion Planning (PDMP). PDMP is a novel parallelised framework that uses bijective and differentiable mappings, or diffeomorphisms, to transform sampling distributions of sampling-based motion planners, in a manner akin to normalising flows. Unlike normalising flow models which use invertible neural network structures to represent these diffeomorphisms, we develop them from gradient information of desired costs, and encode desirable behaviour, such as obstacle avoidance. These transformed sampling distributions can then be used for sampling-based motion planning. A particular example is when we wish to imbue the sampling distribution with knowledge of the environment geometry, such that drawn samples are less prone to be in collisions. To this end, we propose to learn a continuous occupancy representation from environment occupancy data, such that gradients of the representation defines a valid diffeomorphism and is amenable to fast parallel evaluation. We use this to "morph" the sampling distribution to draw far fewer collision-prone samples. PDMP is able to leverage gradient information of costs, to inject specifications, in a manner similar to optimisation-based motion planning methods, but relies on drawing from a sampling distribution, retaining the tendency to find more global solutions, thereby bridging the gap between trajectory optimisation and sampling-based planning methods. | ['Fabio Ramos', 'Tucker Hermans', 'Weiming Zhi', 'Tin Lai'] | 2021-08-26 | null | null | null | null | ['normalising-flows'] | ['methodology'] | [ 2.52315462e-01 5.64428151e-01 -2.80519724e-01 7.88838118e-02
-5.11132598e-01 -5.52284598e-01 1.00109625e+00 -1.18071891e-01
-5.54504752e-01 8.02661657e-01 7.25782156e-01 -2.54257202e-01
-4.63831753e-01 -1.28745508e+00 -7.21850276e-01 -6.56226575e-01
-4.73610222e-01 7.41350353e-01 2.35395178e-01 -5.22917628e-01
5.47822595e-01 7.82698750e-01 -1.36834836e+00 -2.54420519e-01
6.67285979e-01 1.96485922e-01 2.34542817e-01 1.05088854e+00
1.60150409e-01 4.89039630e-01 -6.07180037e-03 2.83709705e-01
5.38038254e-01 -4.46584761e-01 -1.04263675e+00 -1.82610750e-01
6.43767864e-02 -2.13763610e-01 -4.95704234e-01 6.71076596e-01
4.41073388e-01 7.55395710e-01 9.96406257e-01 -1.14229405e+00
-3.43896329e-01 7.75614977e-02 -2.37411205e-02 1.21047676e-01
3.56018007e-01 6.72488511e-01 8.90499949e-01 -5.46007514e-01
1.09626472e+00 1.46203756e+00 7.76636958e-01 6.64303064e-01
-1.39405632e+00 2.24966034e-02 -2.08608240e-01 -2.17582643e-01
-9.64527130e-01 -3.79588097e-01 6.09083354e-01 -4.29805368e-01
1.07231390e+00 4.37678456e-01 1.04316521e+00 8.38803172e-01
8.45503211e-01 4.96504962e-01 5.96220493e-01 -2.17059284e-01
6.87992096e-01 -3.87259185e-01 -6.68454230e-01 6.41916394e-01
-1.28317729e-01 4.99646991e-01 -1.85000196e-01 -2.31781542e-01
1.23231399e+00 -1.06115311e-01 -4.29337025e-01 -1.08492684e+00
-1.42748582e+00 1.21110189e+00 7.05932379e-01 -4.33034860e-02
-4.44390476e-01 4.95080173e-01 3.86108696e-01 2.18831390e-01
2.12602451e-01 9.33045805e-01 -1.56035170e-01 -4.39178318e-01
-6.37206316e-01 8.04056585e-01 8.94435823e-01 8.55697334e-01
1.04782176e+00 -8.68930668e-02 -2.64751256e-01 3.65436345e-01
3.07829142e-01 2.78263777e-01 4.43566263e-01 -1.53945124e+00
3.55358392e-01 3.49316537e-01 1.20974891e-01 -9.77352381e-01
-6.29371703e-01 2.16929987e-01 -7.68132091e-01 7.34161675e-01
2.68827647e-01 -1.51445433e-01 -8.81040215e-01 1.83348775e+00
4.97906297e-01 -2.56202761e-02 6.95534842e-03 8.76442373e-01
-1.52885690e-01 7.40365446e-01 -1.10851295e-01 1.67701736e-01
8.63947034e-01 -7.46174753e-01 -1.29908457e-01 -2.44770169e-01
9.47801232e-01 -2.96566159e-01 1.19443500e+00 -6.61390424e-02
-1.23572111e+00 -2.61635453e-01 -1.12497795e+00 -2.06858590e-01
-1.96478322e-01 -7.23036170e-01 3.78917664e-01 4.64793354e-01
-1.53860557e+00 1.23853767e+00 -1.07623982e+00 -4.05124515e-01
5.14748812e-01 4.33831662e-01 -2.32448161e-01 1.43569186e-01
-9.40494657e-01 1.27566767e+00 3.61571789e-01 -2.78988570e-01
-9.44716692e-01 -1.05682290e+00 -1.41044199e+00 -1.45353541e-01
9.33557749e-03 -1.03966391e+00 1.08710670e+00 -2.61502832e-01
-1.78633833e+00 2.48566672e-01 -8.00815672e-02 -3.72520864e-01
7.31720924e-01 2.07286596e-01 3.55252594e-01 -1.01580955e-01
3.65655303e-01 1.30142450e+00 6.28003895e-01 -8.92527342e-01
-5.58064044e-01 3.01132686e-02 1.88729867e-01 7.02783585e-01
1.05019689e-01 -5.47860444e-01 1.17955327e-01 -1.29107475e-01
-9.53526571e-02 -1.07023323e+00 -7.90148079e-01 4.62350667e-01
-2.29350328e-01 -1.29951239e-01 9.93801832e-01 -3.60681623e-01
7.57948518e-01 -1.72878504e+00 7.04188883e-01 4.23652709e-01
1.87417656e-01 -1.85263902e-01 -2.30195478e-01 5.37526608e-01
3.00079703e-01 7.32982829e-02 -8.19001794e-01 -1.25684291e-01
1.39012441e-01 4.99833226e-01 -7.14096427e-02 7.33759522e-01
5.92541814e-01 1.07291079e+00 -1.48953724e+00 -1.84590504e-01
5.36244810e-01 5.21981478e-01 -1.04896295e+00 4.23060283e-02
-4.10412341e-01 7.79535294e-01 -5.48634827e-01 -1.66589785e-02
3.45360816e-01 4.36827719e-01 -1.75487787e-01 3.00252289e-01
-4.35233980e-01 4.06191826e-01 -1.02454269e+00 1.88995588e+00
-7.93179274e-01 7.48439729e-01 5.40661104e-02 -9.87711251e-01
9.81556118e-01 -4.15751431e-03 6.48619175e-01 -5.07324100e-01
1.06622845e-01 1.79800168e-01 -1.65990219e-01 -2.51088172e-01
7.71323085e-01 -3.35696906e-01 -2.08135322e-01 4.83502358e-01
-2.53897995e-01 -9.25000429e-01 8.06190744e-02 -5.90254329e-02
1.41156900e+00 5.52223206e-01 2.91104168e-01 -7.85595775e-01
4.07964259e-01 4.07927752e-01 2.31857643e-01 4.20117885e-01
-5.62344268e-02 6.52279019e-01 5.31737089e-01 -6.14470124e-01
-1.61818957e+00 -1.11583889e+00 -6.00073040e-02 5.87136030e-01
1.61882803e-01 -2.53505826e-01 -8.10745955e-01 -3.78879994e-01
-7.14901239e-02 5.89302361e-01 -8.08926225e-01 -2.97230989e-01
-1.20802462e+00 -3.67327958e-01 3.92529786e-01 5.00255704e-01
3.85477513e-01 -1.41976452e+00 -1.29937899e+00 5.54800928e-01
1.17867716e-01 -3.56923431e-01 -8.52901697e-01 3.19275290e-01
-8.12713265e-01 -1.04861450e+00 -7.92423487e-01 -6.86645865e-01
6.64919138e-01 -8.60645697e-02 9.18885827e-01 -2.68189777e-02
-3.12234670e-01 3.76693040e-01 3.75770815e-02 -7.19926208e-02
-7.51043081e-01 3.23527366e-01 7.52934162e-03 -5.29006660e-01
-1.34308189e-01 -9.01129544e-01 -6.36914790e-01 3.03866953e-01
-9.15730715e-01 -2.34824698e-02 3.42590749e-01 8.43298554e-01
4.80257481e-01 -1.11288466e-01 2.44639754e-01 -5.02925754e-01
8.66099238e-01 -6.80241585e-01 -4.31342602e-01 -4.07219052e-01
-2.37668976e-01 6.70499325e-01 8.79383385e-01 -3.78440738e-01
-6.40280128e-01 1.12481825e-01 -2.86584258e-01 -2.90117085e-01
-9.38710570e-02 1.62491694e-01 -5.27236275e-02 -2.56564766e-01
9.95071054e-01 8.72709602e-02 4.23671842e-01 6.65811896e-02
8.58201683e-01 2.56614685e-01 6.30649507e-01 -9.22530234e-01
7.50188708e-01 7.26160228e-01 3.58729780e-01 -7.63903797e-01
-7.29143899e-03 -3.64489347e-01 -7.30417907e-01 -1.53547823e-01
1.09231532e+00 -1.11469053e-01 -4.00527358e-01 4.98307198e-02
-1.07713342e+00 -9.92199361e-01 -1.03198731e+00 3.28383803e-01
-1.51798260e+00 1.36159554e-01 -3.00938755e-01 -6.10091150e-01
1.18938908e-01 -1.30537784e+00 1.16410089e+00 6.78591654e-02
-5.99040806e-01 -1.34432209e+00 7.07146227e-01 -6.19823754e-01
7.02693105e-01 7.56794572e-01 7.90452480e-01 -1.70848276e-02
-4.22731847e-01 1.06076695e-01 1.76687106e-01 -2.44268671e-01
2.89887130e-01 -3.41767520e-01 -6.25469208e-01 -3.87491673e-01
5.18145896e-02 1.29990578e-01 4.49834824e-01 5.82640290e-01
6.97523177e-01 -5.24657130e-01 -5.16837239e-01 7.80387580e-01
1.30441141e+00 1.79128602e-01 1.03567994e+00 7.63810277e-01
7.62668073e-01 9.22245502e-01 2.87174433e-01 2.70015687e-01
3.95907730e-01 7.91695416e-01 3.90448034e-01 9.45833996e-02
-1.29501447e-01 -5.36714911e-01 2.63831168e-01 4.65726644e-01
-8.88798460e-02 1.40187666e-01 -1.00097847e+00 7.63385713e-01
-1.96338320e+00 -9.05082524e-01 9.46116745e-02 2.17140150e+00
7.20995903e-01 -4.09238078e-02 3.51476580e-01 -7.92906135e-02
4.52338278e-01 3.85885566e-01 -5.77828646e-01 -8.06192398e-01
5.20696700e-01 3.88773441e-01 8.78925979e-01 1.03193653e+00
-1.00614178e+00 6.99674785e-01 6.43980408e+00 6.36165321e-01
-7.55301774e-01 -6.76531792e-02 4.00574744e-01 6.58311173e-02
-7.03034282e-01 2.66040117e-01 -3.72833014e-01 4.06883895e-01
1.06973076e+00 -3.67466599e-01 7.98145175e-01 7.27117836e-01
4.63141799e-01 -1.05473824e-01 -1.22266531e+00 2.98196167e-01
-3.98865849e-01 -1.68460512e+00 -5.73692881e-02 4.32157904e-01
6.74403012e-01 4.57381941e-02 -1.52979061e-01 2.52449006e-01
8.44643652e-01 -1.18380857e+00 7.41337001e-01 5.17858207e-01
7.09411919e-01 -9.76955533e-01 3.33266646e-01 6.03989899e-01
-1.43762708e+00 3.13482434e-02 -5.07516563e-01 -2.14584649e-01
6.15981579e-01 1.61719978e-01 -8.41484845e-01 4.81962800e-01
3.43089759e-01 7.21526325e-01 9.52452496e-02 9.36864555e-01
4.26717810e-02 -1.33921430e-01 -5.25387049e-01 -9.93393511e-02
6.53046131e-01 -4.74994749e-01 8.63729417e-01 9.94597912e-01
3.45818967e-01 -3.65851343e-01 2.74108410e-01 1.20851898e+00
4.27812755e-01 -1.52103767e-01 -1.23911166e+00 5.17909229e-01
3.81111801e-01 9.66819286e-01 -8.08293045e-01 -4.84873615e-02
5.97465336e-02 8.20673048e-01 3.73751432e-01 1.52598694e-01
-6.09242499e-01 -3.75874996e-01 8.55500817e-01 2.92002141e-01
2.96064585e-01 -6.34314120e-01 -2.48090222e-01 -6.45377576e-01
-2.62496829e-01 -5.12160838e-01 -1.27972737e-01 -4.30782437e-01
-8.23579133e-01 2.36685559e-01 3.97396475e-01 -1.31527615e+00
-7.51909018e-01 -3.91348273e-01 -9.19328928e-01 9.39721286e-01
-1.53839374e+00 -6.40069783e-01 -1.23542041e-01 2.90362567e-01
7.00275838e-01 1.95363060e-01 6.74842536e-01 -1.98427811e-01
8.24052617e-02 -1.81045756e-02 -1.14527717e-01 -1.36578962e-01
1.30439788e-01 -1.46439171e+00 1.06676710e+00 6.50735021e-01
-2.87812531e-01 4.63950962e-01 6.81934118e-01 -7.57104933e-01
-1.36093092e+00 -1.25805187e+00 4.13447082e-01 -7.04301655e-01
5.69287062e-01 -1.76382765e-01 -6.69986844e-01 5.58516920e-01
-7.66898915e-02 -7.06534684e-02 -1.93777233e-01 -4.08798993e-01
2.49236971e-01 6.29602909e-01 -1.50417936e+00 9.76101160e-01
1.43870223e+00 -2.77865112e-01 -3.86053681e-01 2.49975279e-01
7.23060310e-01 -6.97320104e-01 -7.94949055e-01 1.77261636e-01
3.25586140e-01 -7.71602988e-01 1.13309383e+00 -4.80963111e-01
4.29277182e-01 -3.90115499e-01 4.95027471e-03 -1.90089333e+00
-5.50919831e-01 -1.02526402e+00 8.56066719e-02 5.23182809e-01
1.91864550e-01 -8.34015727e-01 1.06628036e+00 5.17394364e-01
-5.23916543e-01 -8.98010373e-01 -1.26295841e+00 -9.77329731e-01
6.76420450e-01 -2.66437024e-01 7.63651192e-01 7.22957969e-01
6.65010214e-02 -1.41762137e-01 -1.20555021e-01 -5.86528182e-02
5.15672803e-01 -6.15624368e-01 8.68023157e-01 -8.49809289e-01
-2.68298358e-01 -7.47753501e-01 -5.81052542e-01 -1.18845451e+00
2.31888279e-01 -1.01281238e+00 5.76994836e-01 -1.65869021e+00
-2.58575678e-01 -8.51659238e-01 5.20723164e-01 2.81541944e-01
2.15793312e-01 -6.08186098e-03 3.82084101e-02 2.59270608e-01
1.41179845e-01 9.09192801e-01 1.55453908e+00 -8.83890092e-02
-6.65399611e-01 -3.31104964e-01 -4.08703268e-01 5.77292085e-01
9.37343955e-01 -4.27889109e-01 -7.69755900e-01 -1.92457467e-01
6.94383830e-02 -2.25747824e-02 4.18824583e-01 -1.07716346e+00
2.04023317e-01 -5.47577739e-01 -3.85885201e-02 -1.69503644e-01
2.06704050e-01 -5.06602645e-01 3.02527547e-01 7.25100935e-01
-4.83868122e-01 1.87970400e-01 1.47139937e-01 5.87031960e-01
3.43123108e-01 -2.95144916e-01 6.46820724e-01 -3.92805725e-01
-6.24060810e-01 4.29810435e-01 -6.05303943e-01 2.43705273e-01
1.00998795e+00 -5.84009409e-01 -8.60764682e-02 -3.59887868e-01
-4.77020413e-01 3.74517709e-01 9.57128644e-01 2.77459949e-01
5.93754709e-01 -1.52396023e+00 -6.95346713e-01 4.17962939e-01
-2.05724180e-01 7.81871796e-01 -2.21687526e-01 7.41367340e-01
-9.76348400e-01 1.84486941e-01 -4.95736450e-01 -6.99252725e-01
-2.64413834e-01 3.36271584e-01 6.29646003e-01 -2.42303267e-01
-1.17675674e+00 4.44234997e-01 1.39532372e-01 -1.01700652e+00
-3.42038929e-01 -5.83046675e-01 -4.03117575e-03 -3.36840600e-01
3.16500753e-01 6.29201174e-01 -1.80946767e-01 -4.10534590e-01
-2.63039440e-01 6.58312619e-01 4.91594017e-01 -5.57967424e-01
1.29600835e+00 5.44548593e-02 1.14702724e-01 1.02573680e-02
1.36435723e+00 -2.04174191e-01 -1.95071101e+00 3.95072818e-01
5.04192971e-02 -6.12186968e-01 -4.75398228e-02 7.19686449e-02
-7.85840333e-01 5.60999513e-01 1.40217066e-01 3.61142963e-01
4.13787186e-01 -1.05471902e-01 8.30797076e-01 2.37329364e-01
6.56569958e-01 -9.29564536e-01 -1.21489003e-01 7.37294853e-01
9.90480304e-01 -6.63613796e-01 -1.74635962e-01 -1.86184406e-01
-4.02786970e-01 1.24639022e+00 2.36376539e-01 -8.25717688e-01
4.64421839e-01 5.67036867e-01 -2.55360693e-01 -2.83003747e-01
-3.34618032e-01 3.71288918e-02 6.19688854e-02 1.26731217e+00
-4.70234677e-02 6.36498928e-02 -5.91797270e-02 -4.39927876e-01
-6.65629923e-01 -6.24431632e-02 7.30055571e-01 1.24165392e+00
-6.09143794e-01 -1.19127822e+00 -3.81908268e-01 3.69297534e-01
5.45348883e-01 1.01908393e-01 1.11565873e-01 9.92374480e-01
-1.81047037e-01 4.58956182e-01 4.23545927e-01 -1.76877216e-01
3.79226476e-01 -1.05451904e-01 4.65710610e-01 -7.32297897e-01
-1.34946734e-01 -3.92907679e-01 1.29049271e-01 -9.46654320e-01
-3.28883290e-01 -8.95179927e-01 -1.43775642e+00 -4.77128386e-01
1.38333455e-01 1.69239808e-02 4.84832674e-01 9.33831573e-01
2.61104137e-01 4.69139874e-01 5.93810678e-01 -1.91124845e+00
-5.83327472e-01 -4.22162324e-01 -2.07235992e-01 2.29624137e-01
5.89286745e-01 -8.51203084e-01 -3.87985468e-01 -2.81664878e-01] | [4.67917013168335, 0.9662708640098572] |
09b75fb3-da5f-4dbd-ad9a-58f11ffa747f | weight-poisoning-attacks-on-pre-trained | 2004.0666 | null | https://arxiv.org/abs/2004.06660v1 | https://arxiv.org/pdf/2004.06660v1.pdf | Weight Poisoning Attacks on Pre-trained Models | Recently, NLP has seen a surge in the usage of large pre-trained models. Users download weights of models pre-trained on large datasets, then fine-tune the weights on a task of their choice. This raises the question of whether downloading untrusted pre-trained weights can pose a security threat. In this paper, we show that it is possible to construct ``weight poisoning'' attacks where pre-trained weights are injected with vulnerabilities that expose ``backdoors'' after fine-tuning, enabling the attacker to manipulate the model prediction simply by injecting an arbitrary keyword. We show that by applying a regularization method, which we call RIPPLe, and an initialization procedure, which we call Embedding Surgery, such attacks are possible even with limited knowledge of the dataset and fine-tuning procedure. Our experiments on sentiment classification, toxicity detection, and spam detection show that this attack is widely applicable and poses a serious threat. Finally, we outline practical defenses against such attacks. Code to reproduce our experiments is available at https://github.com/neulab/RIPPLe. | ['Paul Michel', 'Keita Kurita', 'Graham Neubig'] | 2020-04-14 | null | null | null | null | ['spam-detection'] | ['natural-language-processing'] | [ 2.27823630e-01 -2.79673543e-02 -1.89944841e-02 -6.40351847e-02
-7.16665149e-01 -1.41887963e+00 4.18092042e-01 1.99591160e-01
-6.38477981e-01 5.89182436e-01 -2.40873083e-01 -6.67208552e-01
3.20599556e-01 -7.55394638e-01 -1.12195659e+00 -7.89371908e-01
-1.47302628e-01 7.46620968e-02 2.68951088e-01 -7.68701285e-02
4.94730145e-01 6.18768692e-01 -9.57192421e-01 2.80079722e-01
4.78624105e-01 5.18294871e-01 -2.73048043e-01 7.80399442e-01
1.08551227e-01 1.49768516e-01 -8.98377001e-01 -9.37107742e-01
6.44496262e-01 2.52056420e-01 -8.29547703e-01 -5.14920175e-01
2.84024835e-01 -2.80390143e-01 -2.47257069e-01 1.39633429e+00
4.57119465e-01 -2.82245457e-01 3.47943068e-01 -1.46366906e+00
-5.87584674e-01 7.62367785e-01 -6.35141373e-01 2.18794793e-01
1.35412648e-01 4.37583506e-01 8.66097927e-01 -3.14140052e-01
4.41835612e-01 9.14607406e-01 5.73842168e-01 1.04938805e+00
-1.50057423e+00 -1.11358595e+00 -3.75612602e-02 -2.26925403e-01
-1.18322587e+00 -4.17296797e-01 5.61659396e-01 -4.72991794e-01
8.34975064e-01 6.71183944e-01 2.87218630e-01 1.70101357e+00
2.21315309e-01 5.37872791e-01 1.03223097e+00 -3.03829193e-01
3.33150655e-01 5.85634232e-01 2.42337495e-01 7.01983452e-01
6.36128545e-01 1.90753192e-01 -4.44769919e-01 -1.17276168e+00
2.49688521e-01 -2.10089348e-02 -5.03624976e-01 -3.29876125e-01
-8.35126519e-01 1.09716022e+00 3.01703364e-01 5.13034128e-02
8.81090239e-02 3.44712675e-01 5.19800723e-01 3.58446062e-01
3.82921159e-01 1.00888705e+00 -1.06942403e+00 9.61388126e-02
-5.24592757e-01 8.62783343e-02 1.07286584e+00 5.51331878e-01
4.97925669e-01 -1.04023620e-01 3.26861531e-01 3.69522721e-01
4.18858714e-02 4.52720672e-01 5.39506495e-01 -5.43926060e-01
3.71046126e-01 7.00485632e-02 2.82304794e-01 -8.46482694e-01
-2.26298243e-01 -6.59197047e-02 -3.65854889e-01 2.77024359e-01
6.53956234e-01 -6.74994051e-01 -7.76897371e-01 1.99905288e+00
5.04211843e-01 2.72015572e-01 -1.50996298e-02 4.77758735e-01
1.71248436e-01 5.81219196e-01 3.24531257e-01 8.79083946e-02
1.43410885e+00 -5.83797336e-01 -2.07947522e-01 -1.24819107e-01
7.99310207e-01 -6.32389605e-01 1.20108104e+00 5.53578079e-01
-7.87136018e-01 2.60452539e-01 -1.21721578e+00 1.16850577e-01
-9.31701243e-01 -4.15684640e-01 8.23392689e-01 1.19135976e+00
-6.34251654e-01 8.16198349e-01 -8.32017481e-01 -2.61096925e-01
7.08063245e-01 6.37963116e-01 -4.91478413e-01 3.22189778e-01
-1.43662620e+00 9.07491207e-01 1.13031402e-01 -2.81141698e-01
-1.02748954e+00 -9.90976155e-01 -4.25364643e-01 1.49871394e-01
2.93004423e-01 -4.02836651e-01 1.00923538e+00 -5.81798613e-01
-1.25365901e+00 7.98498452e-01 1.95232585e-01 -5.59689403e-01
4.69861716e-01 -2.05624685e-01 -2.64052212e-01 9.70797688e-02
-2.91962147e-01 3.37687582e-01 1.19640267e+00 -1.08639741e+00
-2.18409941e-01 -3.43489259e-01 3.22170734e-01 -4.21583116e-01
-1.04406750e+00 3.72430563e-01 1.80664193e-02 -6.10342085e-01
-6.87925637e-01 -1.10063565e+00 -3.43625784e-01 8.09852257e-02
-8.23449969e-01 2.06488699e-01 7.83408463e-01 -4.75165218e-01
1.26888955e+00 -2.20924067e+00 -4.57456447e-02 6.44075811e-01
4.32599872e-01 7.21030116e-01 -3.00456047e-01 5.91680110e-01
-2.11220056e-01 9.68226135e-01 -2.97780782e-01 3.22170509e-03
1.33083194e-01 -6.76752776e-02 -8.48105967e-01 6.30886674e-01
3.83217423e-03 7.66345859e-01 -5.89606464e-01 8.89543295e-02
-3.12545687e-01 4.85919267e-01 -6.76390231e-01 5.60547113e-02
-3.90726835e-01 8.99967998e-02 -7.81138897e-01 4.13061798e-01
6.96369946e-01 -3.67346793e-01 1.77520484e-01 -1.54937580e-01
1.57063559e-01 3.12278807e-01 -8.13590050e-01 1.24252570e+00
-2.30665848e-01 2.07713455e-01 8.93787146e-02 -7.00387836e-01
4.05878276e-01 2.23603696e-01 1.04128599e-01 1.43242344e-01
2.68958598e-01 1.22853026e-01 -6.32389933e-02 -3.42953920e-01
-4.23349766e-03 1.32131308e-01 -2.32984379e-01 9.24519360e-01
-2.51672994e-02 -5.55956699e-02 -2.22855091e-01 4.95223939e-01
1.57711017e+00 -3.90044361e-01 -2.46758293e-02 -1.29039794e-01
3.62290204e-01 -9.11162421e-02 2.67844319e-01 8.14677656e-01
-1.57024652e-01 1.07288249e-01 8.81413639e-01 -4.45148885e-01
-1.05407596e+00 -1.00753951e+00 -2.18829259e-01 1.09139669e+00
-2.20550999e-01 -7.53202796e-01 -1.09099364e+00 -1.24513352e+00
3.97490770e-01 6.59913719e-01 -9.94250298e-01 -4.56904560e-01
-2.52808422e-01 -1.04172301e+00 1.05309606e+00 1.64298341e-01
2.13040248e-01 -7.58856654e-01 -3.83352280e-01 -5.05300276e-02
1.60393447e-01 -7.20246851e-01 -6.99013174e-01 5.26446104e-01
-6.37194633e-01 -1.24120390e+00 -3.57812911e-01 -3.64872456e-01
9.57257807e-01 1.39159784e-01 6.33598030e-01 4.82683122e-01
-4.77170229e-01 1.89618692e-01 -8.39795321e-02 -6.77840531e-01
-4.31898326e-01 3.85566264e-01 1.21471360e-01 -1.07843570e-01
6.15011990e-01 -6.81905746e-01 -4.42348033e-01 1.89592093e-01
-1.16257501e+00 -5.08499205e-01 2.30490372e-01 6.68879092e-01
8.52342993e-02 8.68676044e-03 4.70486850e-01 -1.50732315e+00
8.72023165e-01 -5.21998525e-01 -9.50490952e-01 3.29848796e-01
-5.60181141e-01 2.16988891e-01 7.16334641e-01 -9.68778729e-01
-4.07167614e-01 1.18163846e-01 -2.27671593e-01 -4.22322422e-01
-4.65729907e-02 1.66767463e-01 -3.15706730e-01 -5.72054029e-01
1.01315069e+00 -9.82674360e-02 -9.04036388e-02 -5.75425148e-01
5.32848239e-01 5.71787536e-01 -7.40969181e-02 -7.45744169e-01
1.40879285e+00 4.61911857e-01 -2.66032100e-01 -4.08232629e-01
-7.97474384e-01 -6.36250526e-02 -1.37000635e-01 3.58787000e-01
7.67281353e-01 -5.79367220e-01 -8.70374978e-01 5.55404305e-01
-1.13768017e+00 -4.75545347e-01 -1.10192401e-02 -3.07272077e-02
-6.54714182e-02 4.34355885e-01 -8.16368580e-01 -3.79404873e-01
-5.44191241e-01 -1.16969979e+00 4.75749016e-01 -2.61772387e-02
-3.16808850e-01 -9.86836553e-01 4.66024801e-02 4.26527351e-01
6.21555924e-01 1.35385275e-01 1.11739850e+00 -1.27870476e+00
-3.19958419e-01 -6.63564682e-01 7.83824921e-03 4.16743457e-01
-2.25477424e-02 1.45448476e-01 -1.33125055e+00 -4.04679507e-01
1.97916746e-01 -4.67716843e-01 6.30380988e-01 -2.23282173e-01
1.53557181e+00 -7.63527334e-01 -4.82964188e-01 6.71285510e-01
1.23102403e+00 -1.07385226e-01 5.85196972e-01 4.61956441e-01
6.98606133e-01 3.35088640e-01 -8.04833323e-02 4.23380345e-01
-2.18969136e-01 2.97771066e-01 2.77078062e-01 2.35810563e-01
7.35655904e-01 -3.33661020e-01 4.38213408e-01 2.99381316e-01
2.11097404e-01 -1.88496113e-01 -6.54747307e-01 9.63265300e-02
-1.43155348e+00 -8.13022077e-01 1.60546288e-01 2.38676262e+00
1.31053185e+00 5.04073739e-01 -7.11278394e-02 -7.68789724e-02
8.02868307e-01 1.39528126e-01 -7.00567722e-01 -8.09237897e-01
2.67451853e-01 6.89400792e-01 1.08831882e+00 6.09205246e-01
-1.24506986e+00 1.17898691e+00 6.23469639e+00 9.62719321e-01
-1.28823650e+00 4.07770663e-01 4.92843330e-01 -2.57969946e-01
-4.81958091e-01 1.68786552e-02 -8.37055743e-01 6.79113328e-01
1.24475956e+00 -3.27861696e-01 7.20426023e-01 8.44844878e-01
-1.16809092e-01 3.51422787e-01 -8.67904782e-01 3.82868350e-01
-7.50052258e-02 -1.42277443e+00 3.57172079e-02 3.09989214e-01
3.32424551e-01 6.26187176e-02 3.33906889e-01 1.29454836e-01
7.57546961e-01 -8.58385623e-01 1.75875783e-01 -1.75096560e-02
5.70934415e-01 -9.24277186e-01 2.11558476e-01 2.47063130e-01
-4.62735653e-01 -1.17605381e-01 -5.74824631e-01 3.11350524e-01
-1.16194382e-01 5.36687315e-01 -6.16454720e-01 -1.66129321e-01
5.88734150e-01 1.45838708e-01 -7.30451584e-01 8.45150352e-01
-6.00549161e-01 1.00679064e+00 -5.61873853e-01 -1.56046664e-02
-1.18190153e-02 9.38553289e-02 4.32054460e-01 1.05529773e+00
3.14375549e-03 -6.17695935e-02 -1.87671199e-01 7.84263194e-01
-4.87773806e-01 -8.27026460e-03 -7.65477180e-01 -3.98131430e-01
4.52518702e-01 1.36799169e+00 -4.27140862e-01 -1.42541215e-01
-2.80815870e-01 1.03100288e+00 4.53329265e-01 3.63935620e-01
-9.88182724e-01 -6.30925298e-01 1.06017482e+00 -3.03360187e-02
2.16887265e-01 -4.65917327e-02 -2.32865676e-01 -1.43065619e+00
-1.83065102e-01 -1.06428647e+00 4.09348816e-01 -5.95255435e-01
-1.40010476e+00 3.95985425e-01 -2.63324440e-01 -8.35079074e-01
3.69309276e-01 -6.71935201e-01 -7.16086328e-01 7.42901146e-01
-1.23648608e+00 -8.66745532e-01 5.06130934e-01 5.91047406e-01
3.34407911e-02 -9.35580730e-02 1.16691577e+00 1.20862626e-01
-7.97750413e-01 1.06457543e+00 1.21017866e-01 2.85273284e-01
8.29058647e-01 -1.03903639e+00 5.81208229e-01 8.27309012e-01
1.95297614e-01 1.06914067e+00 8.13168049e-01 -6.83198512e-01
-1.49535012e+00 -9.80008245e-01 8.63942444e-01 -8.58040452e-01
1.27306736e+00 -1.02557659e+00 -1.07797444e+00 8.19805741e-01
1.24175631e-01 -2.01416016e-02 1.18400502e+00 -3.88667919e-02
-1.07351291e+00 1.53429747e-01 -1.61781275e+00 7.46497989e-01
6.06170952e-01 -7.16155887e-01 -2.89537013e-01 9.91381526e-01
9.66226876e-01 -1.34601101e-01 -7.97568262e-01 1.98272206e-02
5.65929234e-01 -3.43253821e-01 1.09030271e+00 -1.33198416e+00
1.65981084e-01 7.02553382e-03 -2.76514105e-02 -1.18257773e+00
-7.67318681e-02 -9.94329333e-01 -2.04629689e-01 1.14949107e+00
9.38784301e-01 -1.16929889e+00 7.45014131e-01 1.10094810e+00
5.39116442e-01 -5.18103957e-01 -6.68582380e-01 -6.80021942e-01
6.51750445e-01 -2.58718818e-01 6.29958153e-01 1.15286911e+00
2.69807637e-01 -2.96780765e-02 -4.55257803e-01 6.52160048e-01
7.19554126e-01 -4.18705881e-01 6.84056699e-01 -9.98080671e-01
-5.10252118e-01 -1.41281515e-01 -2.15745136e-01 -4.65365410e-01
3.84078175e-01 -8.82197917e-01 -3.21775705e-01 -4.70548302e-01
1.44490510e-01 -3.19953233e-01 -4.96849298e-01 9.05581474e-01
-3.83618735e-02 3.46191138e-01 2.50711590e-01 2.26543531e-01
-2.07313061e-01 -8.11767504e-02 8.45742583e-01 -2.31269762e-01
5.66258244e-02 8.48800614e-02 -1.07173574e+00 6.71670079e-01
1.31990683e+00 -1.19186461e+00 -3.18633407e-01 -2.14614123e-01
5.75005114e-01 -5.51539481e-01 2.78080434e-01 -6.08867943e-01
1.66461974e-01 -7.65761286e-02 2.61446446e-01 2.27024317e-01
2.79883653e-01 -8.72230530e-01 -7.54538104e-02 5.72298586e-01
-5.74687898e-01 -6.18176237e-02 4.74870265e-01 6.30713999e-01
5.15728295e-01 -5.69234550e-01 8.65382135e-01 -1.66016266e-01
1.04719147e-01 3.73687744e-01 -5.23193598e-01 6.37376234e-02
1.09132588e+00 3.57624859e-01 -7.16785908e-01 -7.88071752e-02
-8.73791397e-01 1.86412573e-01 6.56566799e-01 2.11172715e-01
3.49481583e-01 -8.63846481e-01 -2.73190737e-01 3.96197110e-01
-3.10170092e-02 -7.33466029e-01 -5.90210892e-02 4.32981551e-01
-3.94335747e-01 -1.16309673e-02 8.42467397e-02 3.82745378e-02
-1.36047423e+00 1.04997957e+00 2.31626898e-01 -1.31381661e-01
-2.72938430e-01 9.76612687e-01 1.20764732e-01 -5.33833981e-01
3.22396368e-01 -1.44001888e-02 7.22470433e-02 -1.29475579e-01
7.30416238e-01 -3.85474749e-02 -1.18534500e-02 -7.99651295e-02
-3.81525934e-01 1.70892194e-01 -5.41241288e-01 4.17778417e-02
1.27510619e+00 2.33773395e-01 -2.24879280e-01 1.36213272e-03
1.44571102e+00 4.55762208e-01 -9.91187751e-01 4.24125418e-02
-6.55516759e-02 -6.35765910e-01 -2.08454058e-01 -1.09379947e+00
-1.21206176e+00 9.42112744e-01 4.13811207e-01 5.36694586e-01
6.96441770e-01 -1.71440706e-01 9.79826033e-01 6.85338676e-01
4.02321070e-01 -8.29668462e-01 -1.08385511e-01 2.52468169e-01
6.98195755e-01 -9.07402873e-01 -5.07336818e-02 -3.97236615e-01
-5.87652504e-01 9.23075140e-01 5.00908732e-01 -2.92310238e-01
1.07249296e+00 5.18894732e-01 1.28706411e-01 -8.84514228e-02
-7.98195362e-01 4.16309863e-01 -2.08707675e-01 5.44132888e-01
-1.20981768e-01 4.56311963e-02 -9.97559577e-02 7.26234138e-01
-1.96350574e-01 -1.36569038e-01 7.99113393e-01 9.90467489e-01
-4.32268858e-01 -1.52536738e+00 -4.05894816e-01 6.23763382e-01
-1.02401853e+00 -3.93707961e-01 -5.71235180e-01 3.96828324e-01
-1.74685910e-01 8.24369252e-01 -4.99943733e-01 -5.91275692e-01
9.71880406e-02 4.85855728e-01 2.48187140e-01 -7.27629244e-01
-9.81496334e-01 -2.79700518e-01 1.13187470e-01 -6.00625515e-01
2.47100383e-01 -4.85460430e-01 -9.11884010e-01 -7.08973706e-01
-5.62716544e-01 3.74022007e-01 7.65069306e-01 6.35542095e-01
5.06303012e-01 -1.73562288e-01 7.54975677e-01 -5.76930583e-01
-1.11642778e+00 -6.50876403e-01 -6.04716122e-01 3.71066540e-01
2.15657219e-01 -2.95208126e-01 -1.06165659e+00 1.75161604e-02] | [5.913322448730469, 7.7172369956970215] |
29cec251-3895-40d5-9577-cc3efacc6a31 | keyphrase-generation-with-correlation | 1808.07185 | null | http://arxiv.org/abs/1808.07185v1 | http://arxiv.org/pdf/1808.07185v1.pdf | Keyphrase Generation with Correlation Constraints | In this paper, we study automatic keyphrase generation. Although conventional
approaches to this task show promising results, they neglect correlation among
keyphrases, resulting in duplication and coverage issues. To solve these
problems, we propose a new sequence-to-sequence architecture for keyphrase
generation named CorrRNN, which captures correlation among multiple keyphrases
in two ways. First, we employ a coverage vector to indicate whether the word in
the source document has been summarized by previous phrases to improve the
coverage for keyphrases. Second, preceding phrases are taken into account to
eliminate duplicate phrases and improve result coherence. Experiment results
show that our model significantly outperforms the state-of-the-art method on
benchmark datasets in terms of both accuracy and diversity. | ['Xiao-Ming Zhang', 'Zhoujun Li', 'Yu Wu', 'Jun Chen', 'Zhao Yan'] | 2018-08-22 | keyphrase-generation-with-correlation-1 | https://aclanthology.org/D18-1439 | https://aclanthology.org/D18-1439.pdf | emnlp-2018-10 | ['keyphrase-generation'] | ['natural-language-processing'] | [ 3.55304062e-01 -4.99421328e-01 -3.93842995e-01 1.42276406e-01
-8.90247703e-01 -8.05139720e-01 7.52551317e-01 6.86308682e-01
-4.80592757e-01 1.05316651e+00 7.36274064e-01 -1.89383850e-01
1.83797941e-01 -8.88746321e-01 -4.64188159e-01 -4.37390268e-01
3.50968003e-01 1.25888050e-01 5.13404727e-01 -3.93117368e-01
8.38410020e-01 1.19601302e-01 -1.37980735e+00 4.93300706e-01
9.14907157e-01 4.28555638e-01 2.88875163e-01 3.41772765e-01
-6.40911520e-01 6.38879716e-01 -8.60034049e-01 -3.05186957e-01
1.60827860e-01 -5.41999936e-01 -7.60510206e-01 1.00767119e-02
3.14756989e-01 -3.16048592e-01 -2.94531405e-01 1.12770057e+00
3.13539743e-01 5.10562910e-03 5.14179707e-01 -1.06405401e+00
-3.82824689e-01 1.14882612e+00 -8.79668117e-01 2.51555264e-01
5.77637553e-01 -2.70093590e-01 1.51965928e+00 -1.17804849e+00
5.99379599e-01 1.08642459e+00 2.76359648e-01 1.85712919e-01
-1.27318060e+00 -6.64093971e-01 1.49695352e-01 9.45776552e-02
-1.32393491e+00 -1.86072484e-01 6.50208652e-01 -6.39524609e-02
8.95023882e-01 4.43459868e-01 7.53864825e-01 6.90533221e-01
3.22723478e-01 1.17358112e+00 8.74485075e-01 -5.91383934e-01
2.54204888e-02 8.75600651e-02 4.53931272e-01 2.61572421e-01
5.97010434e-01 -3.27664852e-01 -6.88255966e-01 -5.37515342e-01
3.84461164e-01 -7.92809650e-02 -2.63002932e-01 4.45290245e-02
-1.47647440e+00 7.86325753e-01 -3.10006682e-02 4.17761624e-01
-7.17653036e-01 -5.30570261e-02 4.55294371e-01 2.32480258e-01
3.07790041e-01 9.63570356e-01 -3.21487188e-01 -8.53381604e-02
-1.25744939e+00 8.36241722e-01 7.84294009e-01 8.68611157e-01
8.65420222e-01 -3.05199504e-01 -7.34801769e-01 8.11318517e-01
-1.19877204e-01 3.45000714e-01 5.01352012e-01 -3.70367169e-01
6.70220852e-01 7.64121115e-01 1.84709519e-01 -1.28416085e+00
-1.72351956e-01 -3.16633761e-01 -7.25848913e-01 -4.56609547e-01
-2.05963567e-01 -1.81607902e-01 -9.00978923e-01 1.27829933e+00
8.09157640e-02 -1.85964238e-02 3.85785960e-02 5.26035666e-01
8.24488640e-01 1.13187289e+00 -1.76632896e-01 -5.75125933e-01
1.45565891e+00 -1.04573667e+00 -8.57429087e-01 -7.72903189e-02
3.40082288e-01 -1.19777393e+00 8.68010700e-01 3.32451999e-01
-1.09118938e+00 -4.89350230e-01 -9.99139071e-01 1.88480690e-01
-2.78446138e-01 2.59316325e-01 3.79972070e-01 3.68571818e-01
-7.66158283e-01 4.52486217e-01 -3.97139281e-01 -2.31688410e-01
-1.53887803e-02 -1.42469242e-01 1.74667742e-02 9.38088149e-02
-1.32111859e+00 4.24858809e-01 8.92089427e-01 -6.38069272e-01
-4.93171066e-01 -9.39731598e-01 -4.68851358e-01 1.48524672e-01
9.04436052e-01 -5.75186253e-01 1.31767535e+00 -2.90354431e-01
-8.87384772e-01 2.80209571e-01 -5.31409085e-01 -5.45708597e-01
1.56023383e-01 -4.38230246e-01 -4.53868687e-01 2.13938117e-01
3.13062727e-01 6.14625454e-01 7.97495484e-01 -1.24836278e+00
-1.16975689e+00 2.90790737e-01 1.78707931e-02 3.38347524e-01
-6.34530067e-01 6.25850931e-02 -8.28303039e-01 -1.27035308e+00
1.55502245e-01 -8.42119575e-01 -4.31102157e-01 -6.75762951e-01
-8.51233780e-01 -2.70908803e-01 7.34617174e-01 -5.53022921e-01
2.12936139e+00 -1.77700222e+00 9.96542200e-02 4.41403419e-01
2.14905903e-01 2.81079829e-01 -2.07259551e-01 1.15895712e+00
8.37399662e-02 5.20594060e-01 -1.10789560e-01 6.65058149e-03
-2.90449321e-01 -4.75921258e-02 -1.04703438e+00 -2.21054569e-01
7.65464902e-02 8.80708814e-01 -1.01720583e+00 -6.45355403e-01
-2.38679826e-01 -9.40382034e-02 -3.97819132e-01 1.83541954e-01
-5.35313427e-01 7.63200969e-02 -5.67929864e-01 4.81417358e-01
5.40480137e-01 -2.33674675e-01 5.37969023e-02 -2.66302615e-01
-1.91192344e-01 7.19030499e-01 -1.22234511e+00 1.27536333e+00
-2.74543345e-01 1.46754012e-01 -5.70345938e-01 -4.65950936e-01
8.30848277e-01 3.05997372e-01 6.40196979e-01 -4.19134915e-01
-1.73387080e-01 2.24800840e-01 -2.38441423e-01 -1.94843728e-02
1.39231884e+00 2.62050122e-01 -2.80511975e-01 8.81706297e-01
-2.82051295e-01 -3.44520301e-01 8.96185815e-01 7.99222171e-01
1.01155472e+00 -3.51104110e-01 7.17498720e-01 -1.44248694e-01
7.07248330e-01 2.24050432e-01 4.37047392e-01 1.00851727e+00
4.85255629e-01 5.82841158e-01 6.19551241e-01 -1.78541943e-01
-1.05980754e+00 -6.96839213e-01 3.65033567e-01 8.17668736e-01
2.55836308e-01 -1.30540633e+00 -5.53521991e-01 -7.67313957e-01
1.10674296e-02 7.95447886e-01 -3.60230625e-01 -2.78626126e-03
-7.41217136e-01 -7.20769882e-01 3.51743191e-01 3.59095782e-01
1.83749869e-01 -1.01759958e+00 -3.41828316e-01 6.44765496e-01
-5.56305885e-01 -1.10540688e+00 -8.56625915e-01 -2.02600181e-01
-5.82011878e-01 -8.90512168e-01 -1.05628514e+00 -5.54300129e-01
5.66144407e-01 8.87486041e-01 1.25416768e+00 8.62175599e-02
-5.29389940e-02 -2.21688207e-02 -9.43216920e-01 -4.09348845e-01
-4.72815990e-01 5.09941518e-01 -9.59136114e-02 -2.28009745e-01
3.38373214e-01 -4.34036314e-01 -4.48406160e-01 1.56665385e-01
-1.21294808e+00 9.91626680e-02 8.01041245e-01 8.92079473e-01
7.24586487e-01 4.99564111e-01 4.57815677e-01 -8.52894068e-01
1.27310443e+00 -2.75645792e-01 -5.42995214e-01 4.55677301e-01
-9.03664768e-01 3.93132210e-01 7.39579916e-01 -4.54983711e-01
-6.83923006e-01 -3.45801681e-01 1.05393734e-02 -3.55768763e-02
2.22502813e-01 8.83929789e-01 2.46885896e-01 2.88804740e-01
4.17988181e-01 8.93913925e-01 -3.47138524e-01 -4.83368248e-01
4.77634072e-01 4.90369469e-01 3.62341195e-01 -5.49048007e-01
1.06715810e+00 2.49759063e-01 -1.12209305e-01 -8.47999930e-01
-8.13950837e-01 -9.21730280e-01 -3.30099583e-01 8.47536549e-02
3.65186036e-01 -9.78635967e-01 -8.15910026e-02 2.91855544e-01
-1.30487728e+00 5.73929787e-01 -1.88220128e-01 2.02260062e-01
-6.67908564e-02 7.60704041e-01 -4.62520719e-01 -6.05347931e-01
-8.92524302e-01 -1.05597472e+00 9.28742111e-01 3.39240581e-01
-6.24426186e-01 -4.34765756e-01 2.35414401e-01 -2.64740199e-01
1.73793882e-01 -1.96098968e-01 1.06723678e+00 -8.33659470e-01
-6.88379586e-01 -3.69219869e-01 -1.98369086e-01 9.48766544e-02
4.92301881e-01 1.99225522e-03 -9.79811400e-02 -2.49785945e-01
-4.49070334e-01 -1.68278247e-01 1.21160519e+00 1.58223342e-02
1.11189651e+00 -6.85244501e-01 -6.12500489e-01 2.93267425e-02
1.23268843e+00 1.49350941e-01 6.13850951e-01 4.06330317e-01
6.61913157e-01 5.56232870e-01 9.28113341e-01 6.99440300e-01
3.52984548e-01 6.47487283e-01 -6.88097551e-02 8.49412978e-02
1.31843593e-02 -6.35667086e-01 2.47237280e-01 1.00127399e+00
3.33449870e-01 -5.28536558e-01 -6.55162454e-01 8.85341406e-01
-1.78899598e+00 -1.04722250e+00 -2.35255390e-01 2.00720906e+00
1.18980849e+00 4.26684439e-01 2.49006569e-01 3.54836434e-01
5.24049282e-01 6.54308319e-01 4.38174419e-02 -1.52690917e-01
-2.18565434e-01 2.32150152e-01 5.22765040e-01 2.72614598e-01
-9.63654935e-01 1.31078815e+00 6.30036449e+00 1.16926551e+00
-7.48717368e-01 -3.61561388e-01 3.20738494e-01 7.42325783e-02
-7.58856416e-01 1.59734428e-01 -1.35679173e+00 4.11420792e-01
3.63684475e-01 -7.31959343e-01 1.74691796e-01 6.48604572e-01
4.02831733e-02 -2.52530187e-01 -8.32604468e-01 7.26997077e-01
1.79540500e-01 -1.45154333e+00 6.49769306e-01 -1.20641015e-01
1.05885398e+00 -5.01370668e-01 -7.89861977e-02 5.52659780e-02
3.72237533e-01 -5.14336288e-01 6.42627895e-01 2.72892445e-01
3.90475482e-01 -1.03625000e+00 6.63442552e-01 2.89319187e-01
-1.51548123e+00 2.22823218e-01 -3.16136599e-01 3.71023327e-01
3.27290833e-01 8.98261964e-01 -1.04664671e+00 7.33618319e-01
3.86319190e-01 5.04199743e-01 -4.28414524e-01 1.08993161e+00
-5.02709329e-01 6.09599233e-01 -1.90543011e-01 -4.94287580e-01
5.16386330e-01 1.06183616e-02 8.21456611e-01 1.43692291e+00
4.15310711e-01 -4.11867462e-02 4.71215993e-01 6.46091878e-01
-1.24968655e-01 4.84160215e-01 -3.78579319e-01 -4.48819339e-01
7.77018428e-01 1.30280769e+00 -7.47489154e-01 -5.53248405e-01
-2.82379210e-01 9.82718527e-01 -1.91664740e-01 2.40491226e-01
-5.19824743e-01 -7.60946274e-01 6.26450419e-01 6.09739013e-02
5.14972508e-01 -2.53868878e-01 -2.95098275e-01 -1.20899415e+00
3.27165872e-01 -1.17158687e+00 3.45495224e-01 -4.07557189e-01
-1.05348444e+00 5.91186821e-01 2.15686902e-01 -1.34085059e+00
-4.02245462e-01 -2.90202559e-03 -4.63326514e-01 8.55717540e-01
-1.46159303e+00 -8.85112405e-01 1.17700033e-01 1.80628389e-01
8.83144140e-01 -2.55231112e-01 5.97102404e-01 -1.24952853e-01
-2.26077422e-01 6.54517531e-01 3.89696434e-02 1.66366607e-01
8.03451061e-01 -1.10156906e+00 1.09014022e+00 1.25743282e+00
5.03340840e-01 9.29144323e-01 7.99214005e-01 -1.05040610e+00
-1.23187077e+00 -9.95023012e-01 1.34769404e+00 -3.93076725e-02
8.74509990e-01 -1.61910623e-01 -9.50928032e-01 7.48354793e-02
3.04917812e-01 -6.56421244e-01 6.20153010e-01 -8.42649583e-03
-5.76047778e-01 -1.50260413e-02 -5.22398055e-01 1.18555236e+00
6.79357469e-01 -4.14035976e-01 -9.37085092e-01 2.44271204e-01
1.12742817e+00 -3.02045435e-01 -3.41434985e-01 3.89543802e-01
4.86113340e-01 -5.91955125e-01 9.31253731e-01 -2.83174187e-01
7.00325131e-01 -4.81033027e-01 -6.89901114e-02 -1.49006271e+00
-3.58248860e-01 -9.25261378e-01 -4.12941307e-01 1.34922707e+00
6.23868585e-01 -3.01736563e-01 6.12393141e-01 -3.64147797e-02
2.21019119e-01 -7.27437794e-01 -4.58584785e-01 -8.08142006e-01
-1.83844104e-01 -1.74715891e-01 8.86297166e-01 6.76959217e-01
2.27173865e-01 7.09351242e-01 -6.52193367e-01 -1.53399110e-01
3.92257094e-01 4.27678585e-01 9.29077327e-01 -9.12070334e-01
-2.05021858e-01 -6.34902000e-01 1.75979853e-01 -1.37850058e+00
-1.53968513e-01 -7.29156971e-01 4.37876955e-02 -1.62628329e+00
5.32501340e-01 -2.47241408e-01 -2.30480462e-01 3.52819353e-01
-8.77992094e-01 -4.05998677e-02 2.62393624e-01 4.82897580e-01
-5.34655809e-01 4.76037443e-01 1.08725858e+00 -2.42293686e-01
-5.49543321e-01 2.05399152e-02 -1.01094568e+00 4.01067495e-01
7.35127807e-01 -6.50289893e-01 -3.09289545e-01 -1.24287233e-01
2.77851343e-01 5.84900239e-03 -5.63245937e-02 -8.62849355e-01
3.37555110e-01 -3.91847163e-01 2.83642188e-02 -1.37044692e+00
2.34358720e-02 -5.92175245e-01 -3.72869032e-03 5.43357193e-01
-5.57572305e-01 4.50556964e-01 2.27913469e-01 5.52753985e-01
-4.06819522e-01 -3.81392807e-01 3.38265777e-01 -2.04895839e-01
-7.01264679e-01 2.26372182e-01 -5.74981868e-01 2.93759052e-02
7.13669240e-01 1.40234694e-01 -2.84006804e-01 -4.18824106e-01
1.55132219e-01 2.90685296e-01 3.72665644e-01 5.26902676e-01
8.42511714e-01 -1.28916025e+00 -1.08618975e+00 7.03585371e-02
5.43313265e-01 -2.76409209e-01 -1.88752457e-01 4.26378191e-01
-3.49530756e-01 7.69421518e-01 2.48300299e-01 -1.57925457e-01
-1.49590766e+00 5.42308688e-01 -4.39053476e-01 -8.51089776e-01
-4.86587465e-01 8.44046831e-01 -4.13207486e-02 -2.10739430e-02
2.32073262e-01 -4.26909715e-01 -3.94505531e-01 4.17150855e-01
1.10222256e+00 8.49964172e-02 1.21128276e-01 -3.82251054e-01
-1.85239926e-01 4.50496107e-01 -9.61925507e-01 -3.32662404e-01
9.50784087e-01 -4.86330055e-02 -4.18814540e-01 2.73065567e-01
1.01674330e+00 4.27895725e-01 -4.52767819e-01 -6.91473126e-01
2.92263597e-01 -4.52265263e-01 -6.90086260e-02 -6.91937089e-01
-6.15090311e-01 2.94993371e-01 -1.20989524e-01 2.38872886e-01
1.23029888e+00 -7.86739588e-02 1.31519318e+00 5.59382856e-01
4.00677294e-01 -1.04703045e+00 3.65116775e-01 8.63321304e-01
8.27163219e-01 -9.38133955e-01 4.81592566e-01 -4.98049110e-01
-7.64918923e-01 1.05032754e+00 3.55944395e-01 9.56385881e-02
4.12432820e-01 1.72004983e-01 -9.58653465e-02 1.11684255e-01
-9.51990604e-01 -2.86017448e-01 6.26832366e-01 9.84581336e-02
3.60298872e-01 -1.66174233e-01 -9.52333689e-01 4.14052844e-01
-3.82688522e-01 -1.36615887e-01 5.98073363e-01 1.14199913e+00
-7.96742558e-01 -1.46808326e+00 -4.37588543e-01 5.37354708e-01
-8.26946020e-01 -6.81116939e-01 -7.17492580e-01 3.97524565e-01
-2.25504935e-01 9.83674824e-01 -1.36899188e-01 -7.06611872e-01
2.29537725e-01 -5.80145158e-02 9.13208276e-02 -8.85077775e-01
-7.24705756e-01 6.40083969e-01 1.30997568e-01 -1.94284916e-01
-1.86523899e-01 -5.30495644e-01 -9.83944058e-01 -3.61074120e-01
-4.74692225e-01 5.53063273e-01 3.02383095e-01 6.95166111e-01
3.14812183e-01 5.91852367e-01 9.56616640e-01 -4.12117928e-01
-6.66532874e-01 -1.04665303e+00 -3.33884358e-01 1.34894043e-01
2.23586351e-01 -1.54500529e-01 -8.73245858e-03 2.11488362e-02] | [12.286810874938965, 8.900248527526855] |
1a7a6c4b-acc2-476b-b9d5-ece4a6e85b0f | neural-additive-models-for-nowcasting | 2205.1002 | null | https://arxiv.org/abs/2205.10020v1 | https://arxiv.org/pdf/2205.10020v1.pdf | Neural Additive Models for Nowcasting | Deep neural networks (DNNs) are one of the most highlighted methods in machine learning. However, as DNNs are black-box models, they lack explanatory power for their predictions. Recently, neural additive models (NAMs) have been proposed to provide this power while maintaining high prediction performance. In this paper, we propose a novel NAM approach for multivariate nowcasting (NC) problems, which comprise an important focus area of machine learning. For the multivariate time-series data used in NC problems, explanations should be considered for every input value to the variables at distinguishable time steps. By employing generalized additive models, the proposed NAM-NC successfully explains each input value's importance for multiple variables and time steps. Experimental results involving a toy example and two real-world datasets show that the NAM-NC predicts multivariate time-series data as accurately as state-of-the-art neural networks, while also providing the explanatory importance of each input value. We also examine parameter-sharing networks using NAM-NC to decrease their complexity, and NAM-MC's hard-tied feature net extracted explanations with good performance. | ['Dongil Kim', 'Wonkeun Jo'] | 2022-05-20 | null | null | null | null | ['additive-models'] | ['methodology'] | [ 1.80196106e-01 1.84043720e-01 -3.29538882e-01 -5.84649265e-01
-3.63290370e-01 -1.85926619e-03 8.34648490e-01 1.45023856e-02
2.60643847e-02 9.49278712e-01 2.70950735e-01 -5.52896082e-01
-5.88283181e-01 -6.84430420e-01 -7.32645869e-01 -7.58630991e-01
-1.49421155e-01 3.92697960e-01 -5.38820177e-02 -2.43765593e-01
2.39276756e-02 4.32154268e-01 -1.78591299e+00 2.34960124e-01
1.08688593e+00 1.40706480e+00 1.35897949e-01 6.02261841e-01
-2.42299914e-01 1.03907490e+00 -6.45551741e-01 -4.60018247e-01
2.13807583e-01 -4.82925564e-01 -9.21822712e-03 -3.85078013e-01
2.70987272e-01 -2.24416643e-01 -5.53311229e-01 7.18822896e-01
1.65608376e-01 4.16332692e-01 8.55582595e-01 -1.71421635e+00
-8.60194266e-01 9.25085127e-01 -2.20101580e-01 2.07450151e-01
-5.81846416e-01 -1.05562039e-01 1.05624962e+00 -8.32023025e-01
-5.84879182e-02 1.29427123e+00 7.06663072e-01 6.07528269e-01
-1.11195910e+00 -9.13697243e-01 3.86265874e-01 5.52427888e-01
-7.70063102e-01 -9.55188349e-02 9.20338273e-01 -2.29010731e-01
1.00158679e+00 5.50533235e-01 5.88698268e-01 1.05411875e+00
5.77236652e-01 6.98727250e-01 6.34021461e-01 -1.39427409e-01
3.94630969e-01 -7.35437497e-02 1.90724701e-01 2.08392799e-01
1.66681394e-01 1.98023319e-01 -6.13268495e-01 -1.59900531e-01
8.38972270e-01 7.25776613e-01 -3.84319536e-02 7.88815841e-02
-1.13004386e+00 9.62546587e-01 6.12488449e-01 2.39826232e-01
-6.30157471e-01 5.05551040e-01 2.62956470e-01 9.06838775e-02
8.36172998e-01 3.99247289e-01 -9.61892188e-01 -9.71854404e-02
-5.56978345e-01 3.29702318e-01 5.71355104e-01 7.72288978e-01
3.80744874e-01 6.47449434e-01 -8.20491910e-02 8.55281174e-01
-5.00668259e-03 3.92096609e-01 6.00856960e-01 -8.48961055e-01
4.43277329e-01 6.52358651e-01 -1.66001946e-01 -1.14754653e+00
-6.09536111e-01 -9.13327575e-01 -1.67032027e+00 4.56705652e-02
9.42529142e-02 -1.97347790e-01 -9.22485471e-01 2.00105667e+00
-1.06796958e-01 6.28883243e-01 2.15596125e-01 7.23677516e-01
7.82451451e-01 9.77754056e-01 -4.53756116e-02 -3.99715394e-01
9.39067066e-01 -1.06354165e+00 -9.13170040e-01 -2.58367419e-01
1.55865088e-01 -1.93865106e-01 6.38023436e-01 5.29632151e-01
-8.22124541e-01 -7.83394158e-01 -8.14365864e-01 7.73448795e-02
-4.54736471e-01 2.28839442e-01 9.57575560e-01 2.97157150e-02
-8.48888338e-01 9.60570216e-01 -8.91718388e-01 8.98088440e-02
1.36073813e-01 4.97224897e-01 -1.67446464e-01 1.50141492e-01
-1.39873743e+00 6.83087230e-01 2.32341081e-01 4.01351333e-01
-7.33743310e-01 -9.64400828e-01 -6.23361766e-01 5.64827859e-01
2.37859562e-01 -6.04257822e-01 1.27680111e+00 -1.06660581e+00
-1.31004810e+00 -2.76993126e-01 -3.69648069e-01 -8.09913039e-01
3.10172737e-01 -4.17794496e-01 -8.61843288e-01 -2.45720923e-01
-1.87419709e-02 6.73849463e-01 8.87254417e-01 -1.03893936e+00
-6.33936167e-01 -2.41014332e-01 -1.10600609e-02 -2.57794708e-01
-5.38143039e-01 -3.59559089e-01 4.29780744e-02 -9.70866978e-01
2.68873006e-01 -7.11483061e-01 -6.16040349e-01 1.88086890e-02
-3.43747169e-01 -3.66803974e-01 8.43056381e-01 -5.83170235e-01
1.28140748e+00 -1.99628174e+00 1.54090717e-01 8.42078701e-02
5.38592398e-01 1.51578724e-01 -1.06048122e-01 3.59765112e-01
-3.32921743e-01 5.83556108e-02 -3.94337296e-01 -7.11906075e-01
1.00902662e-01 6.71093822e-01 -7.28300869e-01 9.83784571e-02
3.90348524e-01 8.45495164e-01 -6.30885959e-01 3.04791871e-02
2.46186972e-01 5.74955046e-01 -4.77153093e-01 3.01565379e-01
-4.37805206e-01 1.81108266e-01 -1.89059496e-01 3.37673664e-01
5.22781789e-01 -3.99327517e-01 1.26330834e-02 2.23029163e-02
9.93506797e-03 -5.88797480e-02 -1.04783642e+00 8.94304514e-01
-3.96604002e-01 9.77617085e-01 -5.80871940e-01 -1.18535137e+00
1.15370095e+00 3.99785459e-01 3.63393098e-01 -7.55709648e-01
2.76040528e-02 4.10237849e-01 2.57255167e-01 -2.38773182e-01
5.72339356e-01 -1.13909796e-01 1.14571184e-01 1.94744408e-01
-1.10421456e-01 2.81018794e-01 -1.98223233e-01 4.59839590e-03
7.78064489e-01 -3.08826208e-01 4.19731289e-01 -3.84134203e-02
3.08445066e-01 -1.40226096e-01 1.05324006e+00 7.59175956e-01
2.28293106e-01 6.75964355e-01 6.06648803e-01 -8.76358569e-01
-1.09905219e+00 -6.47295237e-01 -3.82092148e-02 1.01134825e+00
-8.67693126e-02 -2.63260931e-01 -2.89352417e-01 -3.02803367e-01
1.76163897e-01 1.04395688e+00 -1.03037417e+00 -2.54104972e-01
-4.44702744e-01 -6.62175119e-01 3.65430623e-01 1.11141145e+00
2.51107842e-01 -1.03418922e+00 -5.65915823e-01 4.83247399e-01
-8.76856446e-02 -7.44659066e-01 -9.78253037e-02 7.04247653e-01
-1.01105380e+00 -7.99924135e-01 -8.05261135e-01 -2.94708580e-01
5.72831094e-01 2.16474786e-01 8.78519237e-01 -1.66481305e-02
1.65729951e-02 -2.39898056e-01 -2.85720229e-01 -7.46604443e-01
-1.70920208e-01 -2.05116898e-01 4.16006416e-01 1.62131354e-01
1.70443371e-01 -8.32684815e-01 -5.52925587e-01 2.81226575e-01
-1.02004588e+00 3.40699106e-01 6.50019825e-01 1.04152536e+00
5.23653626e-01 6.33069426e-02 8.82843971e-01 -8.13224435e-01
6.35045826e-01 -8.97151589e-01 -4.99023080e-01 1.78906128e-01
-8.08133483e-01 1.65733427e-01 1.34792066e+00 -8.07860374e-01
-1.00041139e+00 -3.66427541e-01 8.87850374e-02 -9.06055689e-01
-3.13276984e-02 9.24185574e-01 1.04700752e-01 4.39570606e-01
4.57647741e-01 4.20487612e-01 -7.94107616e-02 -7.07225442e-01
2.07219690e-01 2.35396326e-01 8.03239405e-01 -2.78802037e-01
5.62506318e-01 1.48018524e-01 4.98201072e-01 -4.53112960e-01
-7.75533974e-01 -1.03735648e-01 -4.59520817e-01 -1.29052281e-01
5.86637259e-01 -7.72553861e-01 -7.11082160e-01 3.72410655e-01
-1.41697276e+00 -9.86010432e-02 -8.31137300e-02 7.16911793e-01
-3.53284299e-01 -1.69190973e-01 -4.30110067e-01 -1.07305372e+00
-2.06720188e-01 -9.55421507e-01 6.77940965e-01 3.59377474e-01
-1.77380756e-01 -1.11450672e+00 -2.18653962e-01 -4.66568060e-02
7.61933386e-01 4.79713082e-01 1.32131362e+00 -9.99875367e-01
-4.14275080e-01 -2.82592177e-01 -2.60875046e-01 3.17148447e-01
3.97396497e-02 1.06147207e-01 -1.00379717e+00 2.43357852e-01
4.67734523e-02 1.47647381e-01 1.00867605e+00 7.09484041e-01
1.62661445e+00 -6.69152975e-01 -2.11048096e-01 7.02057540e-01
1.20279908e+00 4.49338764e-01 3.41958940e-01 2.09441662e-01
6.86259866e-01 4.07900572e-01 4.88303989e-01 7.72215366e-01
4.15616542e-01 3.56928051e-01 1.07817590e+00 -1.23312607e-01
3.58095020e-01 -2.24226281e-01 2.17170313e-01 1.01082993e+00
-3.33897620e-01 -5.57840407e-01 -6.46099389e-01 4.08944905e-01
-2.43608999e+00 -1.03227770e+00 -6.55634940e-01 1.88957095e+00
2.16083765e-01 2.74055935e-02 -2.46916279e-01 2.60440856e-01
5.20710886e-01 1.96892068e-01 -1.04293013e+00 -4.43300277e-01
-3.96262825e-01 9.77697447e-02 3.15959156e-01 1.42347813e-01
-9.71971989e-01 3.86335731e-01 6.15263700e+00 7.13225961e-01
-1.16679871e+00 -9.47206020e-02 8.97911727e-01 -9.48015824e-02
-5.32991946e-01 -2.06643224e-01 -6.29240811e-01 4.87234503e-01
1.27085245e+00 -3.46872926e-01 2.59662837e-01 1.04023254e+00
4.03041273e-01 2.55598038e-01 -1.17719328e+00 9.14803088e-01
-5.96484318e-02 -1.50447750e+00 3.38496059e-01 1.06929922e-02
1.02098227e+00 -2.20263496e-01 3.05994362e-01 4.63993013e-01
8.38357359e-02 -1.15501213e+00 6.84514225e-01 8.31103027e-01
4.66202050e-01 -9.37628210e-01 1.00321829e+00 5.86376607e-01
-1.15499985e+00 -5.18576264e-01 -7.15858638e-01 -4.32131350e-01
1.56903304e-02 7.80935526e-01 -4.23697084e-01 5.52343786e-01
6.43035173e-01 1.02921069e+00 -3.93188745e-01 9.12815571e-01
-1.87865734e-01 8.44146430e-01 -1.44518018e-01 -2.25669384e-01
4.07136440e-01 -4.82069626e-02 4.00825918e-01 8.58874917e-01
7.26519585e-01 3.63448352e-01 -3.02693188e-01 1.02629685e+00
6.54415861e-02 -2.48395294e-01 -4.20094252e-01 -1.95816737e-02
4.16531861e-01 8.67891729e-01 -3.70525330e-01 -5.38493454e-01
-2.33443990e-01 6.16298854e-01 1.76133960e-01 4.77622181e-01
-1.04516220e+00 -2.30577826e-01 9.88031983e-01 -2.87405163e-01
2.16231540e-01 -1.03244692e-01 -5.56902885e-01 -1.04479992e+00
-9.94649082e-02 -7.25439429e-01 2.58131325e-01 -1.01268351e+00
-1.42494094e+00 8.75548124e-01 -5.69579974e-02 -1.69237018e+00
-5.95957577e-01 -6.78262889e-01 -8.89595509e-01 9.29582179e-01
-1.42512000e+00 -9.87804174e-01 -3.39507014e-01 5.51123381e-01
8.91371369e-01 -4.50727224e-01 8.29813540e-01 7.38096088e-02
-6.15069211e-01 5.33565938e-01 6.20949447e-01 -3.58090967e-01
4.03610498e-01 -1.22037971e+00 4.49278563e-01 8.31065774e-01
1.80494748e-02 7.57146239e-01 8.98872674e-01 -5.51336288e-01
-1.25912654e+00 -1.26226318e+00 9.83739316e-01 -3.53935696e-02
6.29865706e-01 -2.49337509e-01 -1.23928094e+00 5.43513119e-01
9.57729965e-02 -3.81480493e-02 6.79798365e-01 1.32448688e-01
-2.81232208e-01 -3.52424294e-01 -7.96657085e-01 6.14489675e-01
7.14717805e-01 -2.58986235e-01 -6.28170729e-01 1.82776466e-01
1.25462890e+00 -2.10978448e-01 -6.53232515e-01 4.97799784e-01
6.78011060e-01 -1.03697586e+00 8.53119314e-01 -8.61067832e-01
1.16438711e+00 -1.32902890e-01 -3.55161279e-01 -1.68674791e+00
-5.10497749e-01 -3.49703491e-01 -5.96242726e-01 9.72995877e-01
5.39965928e-01 -9.16695535e-01 5.05475700e-01 1.01491332e+00
-3.40614140e-01 -9.66251612e-01 -8.40528488e-01 -7.85667479e-01
-1.65294111e-01 -8.84525001e-01 1.02588224e+00 9.63828564e-01
-2.17612967e-01 1.62982009e-02 -8.52918923e-01 2.59468317e-01
5.99777341e-01 -5.70401270e-03 6.02953196e-01 -1.59432364e+00
-4.52328831e-01 -5.99363506e-01 -2.89839238e-01 -7.41629660e-01
1.69667333e-01 -5.27086258e-01 -7.18662739e-02 -1.51535201e+00
1.19427197e-01 -1.06890380e-01 -7.07011282e-01 5.90198755e-01
-4.10768747e-01 -1.14800617e-01 2.70354837e-01 2.22171187e-01
-9.87643376e-02 8.92737031e-01 9.17261064e-01 -1.37058452e-01
-1.14123903e-01 2.62532473e-01 -6.78593993e-01 6.08418584e-01
9.13035154e-01 -4.83776331e-01 -6.78745747e-01 -5.08704066e-01
1.24991529e-01 4.14103627e-01 5.53248942e-01 -9.24727380e-01
4.43514317e-01 -4.56539243e-01 6.43430173e-01 -9.26769435e-01
5.78219235e-01 -1.00385261e+00 5.55551767e-01 5.09075999e-01
-6.78429782e-01 5.07172048e-01 4.02703762e-01 8.01391900e-01
-4.63162154e-01 1.95417807e-01 3.82939637e-01 2.17456311e-01
-7.19406605e-01 4.91421759e-01 -2.65960157e-01 -5.41103065e-01
8.06340396e-01 -8.61780047e-02 -5.63129008e-01 -7.83697486e-01
-4.68371600e-01 2.54873633e-01 -3.90365571e-01 6.09848022e-01
8.92822325e-01 -1.48505187e+00 -6.14474893e-01 2.76892215e-01
2.84780283e-02 -7.56813511e-02 5.63318968e-01 7.02513039e-01
-3.38790789e-02 7.23982036e-01 -3.43017995e-01 -4.63156700e-01
-9.84154284e-01 5.12087941e-01 1.12964623e-01 -1.38440147e-01
-5.19586444e-01 7.29351699e-01 4.13780421e-01 -4.21448022e-01
2.62487173e-01 -6.65915489e-01 -4.15724397e-01 -4.39617373e-02
5.82802415e-01 6.06351733e-01 -2.00887352e-01 -2.31367871e-01
-1.06651299e-01 2.30133653e-01 2.23032042e-01 -1.94111764e-02
1.84545493e+00 1.16908200e-01 -3.97875197e-02 8.04549873e-01
1.12067914e+00 -5.97549915e-01 -1.28491080e+00 -3.22392881e-01
-2.46946618e-01 -2.28795543e-01 -2.48416420e-02 -8.17971110e-01
-1.14281547e+00 1.25808930e+00 3.34546357e-01 4.24357295e-01
1.31096792e+00 -3.30966532e-01 6.27670646e-01 3.85310680e-01
9.65398178e-02 -7.47180879e-01 -7.64208287e-02 5.93793631e-01
1.11304450e+00 -1.19319046e+00 -3.76813591e-01 7.36719891e-02
-6.45429611e-01 1.48476362e+00 8.33960474e-01 -8.17211941e-02
6.87542677e-01 -3.98011021e-02 -1.17732376e-01 1.63283929e-01
-1.44497490e+00 2.04189777e-01 5.47963560e-01 3.47002923e-01
2.72546679e-01 3.82897466e-01 -1.42562747e-01 1.37120295e+00
-1.70491919e-01 -1.43304482e-01 4.83987480e-01 3.36608350e-01
-2.93085486e-01 -7.10530996e-01 -3.68493825e-01 8.88755441e-01
-2.83119410e-01 -2.50416964e-01 -3.41550678e-01 9.03662920e-01
6.80161815e-04 9.63869333e-01 2.80323237e-01 -7.39438713e-01
6.48938119e-02 -8.95833448e-02 -4.04428124e-01 -2.33932793e-01
-5.97079277e-01 3.28462869e-02 -2.62586534e-01 -3.37684393e-01
-1.97345555e-01 -3.77259403e-01 -1.23346961e+00 -5.42444170e-01
-4.54294914e-03 1.80436060e-01 7.26724088e-01 1.10407376e+00
4.22061443e-01 1.01464021e+00 7.62350321e-01 -9.59156156e-01
-5.89355588e-01 -1.02380002e+00 -5.83459377e-01 2.05188200e-01
6.36678696e-01 -6.19817257e-01 -6.70682907e-01 -1.54140741e-01] | [7.014723300933838, 3.0870344638824463] |
7aa5201f-78ea-4c26-b036-89a5438d3be1 | an-end-to-end-goal-oriented-dialog-system | 1803.02279 | null | http://arxiv.org/abs/1803.02279v2 | http://arxiv.org/pdf/1803.02279v2.pdf | An End-to-End Goal-Oriented Dialog System with a Generative Natural Language Response Generation | Recently advancements in deep learning allowed the development of end-to-end
trained goal-oriented dialog systems. Although these systems already achieve
good performance, some simplifications limit their usage in real-life
scenarios.
In this work, we address two of these limitations: ignoring positional
information and a fixed number of possible response candidates. We propose to
use positional encodings in the input to model the word order of the user
utterances. Furthermore, by using a feedforward neural network, we are able to
generate the output word by word and are no longer restricted to a fixed number
of possible response candidates. Using the positional encoding, we were able to
achieve better accuracies in the Dialog bAbI Tasks and using the feedforward
neural network for generating the response, we were able to save computation
time and space consumption. | ['Jan Niehues', 'Stefan Constantin', 'Alex Waibel'] | 2018-03-06 | null | null | null | null | ['goal-oriented-dialog'] | ['natural-language-processing'] | [ 6.84290826e-02 4.39251572e-01 3.68731171e-01 -8.32097530e-01
-4.80474949e-01 -6.79428697e-01 7.32264161e-01 1.13486812e-01
-7.32945800e-01 1.00324094e+00 3.06632519e-01 -3.74783128e-01
1.84912977e-04 -8.68274033e-01 -1.86631992e-01 -1.55224234e-01
2.18279883e-01 9.79491055e-01 4.19713914e-01 -6.57507122e-01
1.71628699e-01 2.19318420e-01 -1.26168847e+00 5.11839509e-01
4.29108143e-01 7.08838999e-01 5.36871552e-01 9.62496281e-01
-2.93281913e-01 9.65096354e-01 -8.85834098e-01 -1.74344689e-01
-1.33924875e-02 -4.97053057e-01 -1.20770621e+00 -6.50856569e-02
2.74995156e-02 -7.80845523e-01 -1.76021174e-01 6.44446969e-01
6.08670354e-01 5.56162179e-01 4.23225105e-01 -9.11143482e-01
-3.44109863e-01 8.60382259e-01 5.03317237e-01 -1.83697149e-01
6.65171742e-01 1.78390741e-01 1.07759404e+00 -8.37533772e-01
6.75663471e-01 1.42036629e+00 2.53337651e-01 1.13233054e+00
-1.21721518e+00 -1.17676601e-01 5.51821589e-02 7.53905550e-02
-9.76692498e-01 -6.77982450e-01 6.00511730e-01 -1.76741332e-01
1.40691483e+00 2.49569938e-01 2.48209566e-01 1.08016086e+00
-1.50016502e-01 7.65662789e-01 8.41623902e-01 -7.47853160e-01
1.98789090e-01 3.42516005e-01 4.20951128e-01 5.18747926e-01
-5.13472676e-01 1.60763651e-01 -3.45341057e-01 -1.54028267e-01
6.36856198e-01 -4.16164160e-01 -4.65382859e-02 -3.58241610e-03
-8.52736473e-01 1.31701851e+00 3.78516287e-01 5.88747680e-01
-2.51251251e-01 -1.22986495e-01 4.16048527e-01 2.66845018e-01
3.09167147e-01 6.27726972e-01 -4.21992779e-01 -1.82366952e-01
-7.92883217e-01 6.02794588e-01 1.27265477e+00 8.60410511e-01
4.62694734e-01 1.01532340e-01 -3.26571792e-01 1.05693877e+00
3.06200266e-01 4.52797674e-02 5.54037929e-01 -9.59976017e-01
5.41678131e-01 4.71819788e-01 4.54975128e-01 -6.27281606e-01
-8.09328198e-01 8.65631178e-02 -4.08585489e-01 1.73302203e-01
9.53078926e-01 -6.37488663e-01 -8.49197268e-01 1.77567232e+00
1.04316352e-02 -5.02033293e-01 3.01681668e-01 9.78409410e-01
7.62538850e-01 7.82373786e-01 9.16868970e-02 -1.78029850e-01
1.18914986e+00 -8.86644065e-01 -9.32072163e-01 -3.49166393e-01
7.81931221e-01 -7.81544983e-01 1.21609533e+00 3.28923702e-01
-1.24608815e+00 -6.92034483e-01 -9.67092633e-01 -1.44095078e-01
-4.34779823e-01 1.42702773e-01 6.58097386e-01 5.22249818e-01
-1.32953000e+00 6.59857154e-01 -4.89795119e-01 -4.57096249e-01
-3.29162657e-01 8.32504869e-01 -2.92481184e-01 2.82586843e-01
-1.56863105e+00 1.15726733e+00 5.49548805e-01 4.80870157e-02
-4.45039093e-01 -1.61238164e-01 -8.42345417e-01 3.48755777e-01
3.32740217e-01 -5.13798416e-01 1.76838851e+00 -6.04885936e-01
-2.08629799e+00 5.32015324e-01 -1.70187399e-01 -8.14741135e-01
4.00320828e-01 -2.10939690e-01 -8.00393969e-02 -3.84013504e-02
-4.45830733e-01 1.07038569e+00 2.02899367e-01 -9.61591363e-01
-6.30791962e-01 -1.48837687e-02 4.35588449e-01 3.52426976e-01
-3.75193775e-01 1.78490877e-01 -1.39049143e-01 -2.33348414e-01
-3.57984714e-02 -1.13360214e+00 -3.53118390e-01 -4.14961278e-01
-1.94129378e-01 -6.20008647e-01 6.15094066e-01 -5.22633970e-01
1.07834685e+00 -1.92961621e+00 8.30845460e-02 -3.12137514e-01
-5.58143258e-02 6.15795374e-01 -1.33201391e-01 6.88564599e-01
2.38578945e-01 4.29180898e-02 -6.57831952e-02 -6.63972974e-01
9.84422937e-02 3.90352160e-01 -3.37498575e-01 -2.49827817e-01
1.71010226e-01 7.76921570e-01 -7.61229873e-01 -2.71677077e-01
6.16572559e-01 2.01684430e-01 -6.68727577e-01 7.83197045e-01
-6.41600609e-01 3.83994013e-01 -2.40143493e-01 -2.22846031e-01
3.15936178e-01 2.36134171e-01 3.59535456e-01 2.53336519e-01
-5.68957739e-02 1.01630151e+00 -9.55074310e-01 1.41353655e+00
-8.80452931e-01 6.09241843e-01 1.32284805e-01 -7.50112712e-01
1.13523555e+00 6.03568792e-01 1.44090131e-01 -5.29508710e-01
2.80346513e-01 2.16440588e-01 5.27180791e-01 -2.91812837e-01
7.05255032e-01 -3.52077425e-01 -2.32282236e-01 6.17780387e-01
3.96806598e-01 -2.14516968e-01 3.03287119e-01 6.56411722e-02
7.48725772e-01 -2.16797031e-02 1.91727877e-01 -1.54454663e-01
7.74555564e-01 -3.76762752e-03 3.72305885e-02 7.13026285e-01
4.03956044e-03 7.37569928e-01 4.63427246e-01 -6.34814262e-01
-9.63196754e-01 -6.93277597e-01 2.01160148e-01 1.44767332e+00
-2.01628655e-01 -3.04711163e-01 -8.16721737e-01 -6.45994008e-01
-5.12939990e-01 1.28418505e+00 -2.33961344e-01 1.83806270e-02
-1.07892323e+00 -4.57376480e-01 5.89571834e-01 4.39269930e-01
3.76175940e-01 -1.41721439e+00 -6.14329278e-01 6.71638310e-01
-3.28476965e-01 -1.26993155e+00 -2.39827484e-01 4.17480290e-01
-6.84894562e-01 -6.05200410e-01 -5.95836997e-01 -7.52577960e-01
3.80855650e-01 -2.03056723e-01 9.51123059e-01 3.91530916e-02
1.64185703e-01 5.21683879e-02 -3.69404316e-01 -3.22890908e-01
-7.85153270e-01 3.33043128e-01 -2.57662870e-02 -3.98935735e-01
3.65385711e-01 -1.41848356e-01 -2.93096304e-01 2.99417377e-01
-7.01634884e-01 7.94653669e-02 9.53474194e-02 1.11589158e+00
-2.52993166e-01 -3.79260719e-01 7.88364649e-01 -1.04631019e+00
1.28599477e+00 -1.36918992e-01 -4.39103782e-01 7.17045665e-02
-4.11705196e-01 4.73562330e-01 9.49900270e-01 -5.16065657e-01
-1.41152215e+00 3.14751357e-01 -9.05370116e-01 1.86279535e-01
-5.03701031e-01 4.12771314e-01 -2.72405408e-02 2.79131740e-01
6.26778185e-01 2.57209279e-02 -2.40310170e-02 -5.49531102e-01
3.55889887e-01 9.22042549e-01 1.36238217e-01 -5.04222214e-01
-1.19209290e-02 -3.20919275e-01 -2.75588214e-01 -8.78594041e-01
-4.41399664e-01 -2.31993765e-01 -5.79975247e-01 -2.36142471e-01
8.95363986e-01 -4.50469017e-01 -8.84601116e-01 3.41675818e-01
-1.58322680e+00 -5.49084783e-01 -9.54439268e-02 3.80579680e-01
-6.15293682e-01 3.03400755e-01 -7.38584399e-01 -1.14780307e+00
-4.19153392e-01 -1.27174139e+00 7.04082310e-01 2.52286375e-01
-7.33608842e-01 -1.10914767e+00 -2.31432244e-01 2.51580894e-01
6.31082475e-01 -3.89907777e-01 1.07398891e+00 -1.39462781e+00
-1.49652973e-01 -3.07340771e-01 6.13453202e-02 3.21670443e-01
1.13339409e-01 -3.34502876e-01 -8.82648468e-01 -3.96888144e-02
1.34759396e-01 -4.33777153e-01 3.93107653e-01 4.02855501e-02
7.03214467e-01 -4.23964500e-01 -7.62562528e-02 -1.56209856e-01
8.50649655e-01 6.13131166e-01 4.87226009e-01 -1.46982118e-01
2.13697821e-01 1.18408310e+00 7.00016499e-01 3.27574104e-01
4.12039518e-01 1.05854917e+00 3.27255279e-01 1.59602150e-01
-1.44209296e-01 -2.84198582e-01 1.77795365e-01 5.06941915e-01
1.15278736e-01 -7.24018633e-01 -9.30125833e-01 6.55036628e-01
-1.92575955e+00 -6.51479781e-01 -2.50338972e-01 2.25731778e+00
8.65849495e-01 3.42551887e-01 2.76881218e-01 -7.35745355e-02
5.70740342e-01 3.40650827e-02 -1.01710819e-01 -8.52498174e-01
2.90241957e-01 3.79097760e-01 1.40797570e-01 1.21145117e+00
-8.75862241e-01 1.34508634e+00 6.69518757e+00 3.37852955e-01
-1.19200921e+00 -6.41142577e-02 5.10017931e-01 6.57826215e-02
-5.37450090e-02 -3.25758643e-02 -9.84831452e-01 1.99504420e-01
1.28080690e+00 7.14036077e-02 4.49071109e-01 8.78354907e-01
2.64455736e-01 -1.69602424e-01 -1.22046816e+00 5.17928720e-01
-1.85724258e-01 -1.09074855e+00 -1.15329839e-01 -2.53666818e-01
7.89445862e-02 -3.53171825e-01 -3.20824623e-01 5.28263688e-01
4.73462760e-01 -1.12248302e+00 3.58490437e-01 2.31353432e-01
3.94625396e-01 -5.73517561e-01 9.95499253e-01 7.19637275e-01
-5.89679897e-01 -1.09221779e-01 -4.21442807e-01 -4.94100392e-01
3.17539781e-01 -4.38156836e-02 -1.73799646e+00 1.92712948e-01
5.99849932e-02 -2.51223177e-01 -1.35333776e-01 5.63360095e-01
-3.33274662e-01 5.01801550e-01 -4.53557014e-01 -7.26700485e-01
4.46349800e-01 4.31362167e-02 4.24012184e-01 1.29952943e+00
1.22640133e-01 2.47930497e-01 1.78694606e-01 5.62783360e-01
1.84130281e-01 8.11721981e-02 -4.74648714e-01 2.96770427e-02
3.67509693e-01 1.03482771e+00 -6.00684881e-01 -1.98108524e-01
-9.28230137e-02 1.02334023e+00 5.00993073e-01 7.15002269e-02
-5.21636665e-01 -4.99647707e-01 4.33363736e-01 8.66770893e-02
1.49822608e-02 -5.11617362e-01 -2.62913823e-01 -8.55148315e-01
-3.29449594e-01 -8.72191608e-01 1.32979676e-01 -6.25292003e-01
-9.73131537e-01 1.08654773e+00 3.16756874e-01 -5.65245807e-01
-1.16853786e+00 -9.89742100e-01 -5.94813704e-01 1.23821986e+00
-7.45895505e-01 -9.10109699e-01 1.13657080e-01 3.32649767e-01
1.02884948e+00 -1.47650629e-01 1.36968791e+00 1.54486805e-01
-7.66213089e-02 5.30233741e-01 -2.92233407e-01 1.89066619e-01
7.53234267e-01 -1.38704395e+00 4.58045900e-01 4.69564617e-01
1.22885682e-01 9.47714150e-01 1.13013422e+00 -3.50578606e-01
-8.32872987e-01 -4.80981529e-01 1.46485734e+00 -1.62845060e-01
4.58188355e-01 -7.02007413e-01 -9.10997570e-01 5.16590893e-01
5.70314050e-01 -4.85208899e-01 5.15378118e-01 2.30633229e-01
1.96271628e-01 2.28589535e-01 -1.18062353e+00 8.54524493e-01
7.08034158e-01 -2.91509449e-01 -7.78154492e-01 4.13146675e-01
7.16420352e-01 -6.86996818e-01 -4.71896827e-01 2.45935753e-01
5.25810063e-01 -8.56626153e-01 7.85072505e-01 -6.78174198e-01
2.84597814e-01 7.65355825e-02 -2.36851033e-02 -1.46229279e+00
7.41027296e-04 -4.83750880e-01 4.02458049e-02 1.23650873e+00
6.91979885e-01 -5.60096622e-01 8.27988625e-01 1.24151134e+00
-2.37497285e-01 -6.11822724e-01 -9.80681241e-01 -3.14332277e-01
2.76734501e-01 -1.87485874e-01 4.58110899e-01 2.28177831e-01
4.85104144e-01 9.36948299e-01 -7.20591903e-01 -2.58547664e-01
-3.03992890e-02 -3.37984748e-02 5.97420514e-01 -1.11246109e+00
-4.42746162e-01 -1.81745380e-01 4.86461297e-02 -1.57711720e+00
2.73797303e-01 -4.60400999e-01 5.92058539e-01 -1.59645247e+00
-4.17238623e-01 -4.12225813e-01 1.41430572e-01 5.62690616e-01
-1.89406872e-01 -4.48843911e-02 5.45796573e-01 -1.00424647e-01
-9.82263759e-02 3.12770367e-01 8.47223639e-01 4.75554988e-02
-3.78393412e-01 4.56437886e-01 -3.20694089e-01 6.54960513e-01
1.20339394e+00 -3.41832697e-01 -6.18449867e-01 -4.43092257e-01
-6.96789026e-02 5.27003169e-01 -3.97932008e-02 -8.88183415e-01
4.03684944e-01 -6.89205155e-02 1.44101307e-01 -5.88722825e-01
8.47640395e-01 -7.69393921e-01 -2.42083356e-01 4.51819241e-01
-9.04743969e-01 -1.39698476e-01 2.77183533e-01 3.40469368e-02
-2.93220907e-01 -9.63960767e-01 7.12302208e-01 -3.87093544e-01
-5.39198995e-01 -3.13880771e-01 -1.00906539e+00 -4.48861688e-01
8.04164231e-01 2.29470823e-02 -1.18562460e-01 -9.03805614e-01
-9.43790913e-01 2.11345524e-01 -8.13776255e-03 5.94911218e-01
5.12136579e-01 -8.56240869e-01 -5.08937061e-01 8.21042508e-02
-2.73142874e-01 -1.70459881e-01 4.55636010e-02 1.18788145e-01
-5.41506588e-01 9.56051648e-01 -3.28941733e-01 -2.27967829e-01
-1.59799325e+00 3.18451911e-01 4.61250693e-01 -6.09265029e-01
-2.38704965e-01 9.15982425e-01 -1.22329546e-02 -8.12550724e-01
4.31892455e-01 -1.84136808e-01 -5.58794856e-01 1.35629876e-02
4.02107149e-01 6.63563535e-02 1.01658165e-01 -4.89494264e-01
-2.27720857e-01 -1.44826278e-01 -2.15994090e-01 -6.80513561e-01
1.13660836e+00 -4.14732005e-03 1.92697793e-01 4.65337425e-01
8.95129561e-01 -1.16240785e-01 -9.64627624e-01 -1.12781741e-01
2.19773263e-01 -1.51735097e-01 -2.39327118e-01 -9.67748642e-01
-4.46057469e-01 1.17140412e+00 3.68463337e-01 6.16888642e-01
8.89416814e-01 -3.09164196e-01 8.58092904e-01 8.02713752e-01
3.55384827e-01 -1.03177607e+00 -2.81577520e-02 1.15710831e+00
9.28892136e-01 -1.28806102e+00 -3.38814646e-01 -4.64325130e-01
-8.10056329e-01 1.54732919e+00 8.71395290e-01 -3.58160436e-02
1.60219401e-01 2.96719640e-01 3.35167021e-01 1.70823544e-01
-9.26800430e-01 -2.68340498e-01 8.05961043e-02 5.74246824e-01
1.01041698e+00 -7.92205855e-02 -8.37408841e-01 6.00399554e-01
-4.56686884e-01 -2.28995517e-01 6.17442012e-01 6.03117466e-01
-5.91298401e-01 -1.72234464e+00 -1.36252150e-01 1.88330606e-01
-4.73403841e-01 -1.70777053e-01 -8.24882507e-01 7.24745154e-01
-1.41238973e-01 1.48397601e+00 -4.77966443e-02 -4.66335952e-01
4.74152178e-01 6.26383245e-01 3.38032156e-01 -9.45393920e-01
-1.01123011e+00 -8.27389136e-02 6.90815568e-01 -1.76250115e-01
1.03494078e-02 -3.32709879e-01 -1.40300632e+00 -1.20294333e-01
-4.35001940e-01 4.07225549e-01 6.75551057e-01 1.07864118e+00
2.21990630e-01 5.07044554e-01 4.05025333e-01 -7.93280244e-01
-9.45015311e-01 -1.52227676e+00 -7.85555765e-02 1.80039600e-01
9.82761532e-02 -4.14923579e-01 5.79107553e-03 -4.96873967e-02] | [12.868128776550293, 7.925685882568359] |
32a2b673-2238-409a-aecf-28a58221feaa | adaptive-superpixel-for-active-learning-in | 2303.16817 | null | https://arxiv.org/abs/2303.16817v1 | https://arxiv.org/pdf/2303.16817v1.pdf | Adaptive Superpixel for Active Learning in Semantic Segmentation | Learning semantic segmentation requires pixel-wise annotations, which can be time-consuming and expensive. To reduce the annotation cost, we propose a superpixel-based active learning (AL) framework, which collects a dominant label per superpixel instead. To be specific, it consists of adaptive superpixel and sieving mechanisms, fully dedicated to AL. At each round of AL, we adaptively merge neighboring pixels of similar learned features into superpixels. We then query a selected subset of these superpixels using an acquisition function assuming no uniform superpixel size. This approach is more efficient than existing methods, which rely only on innate features such as RGB color and assume uniform superpixel sizes. Obtaining a dominant label per superpixel drastically reduces annotators' burden as it requires fewer clicks. However, it inevitably introduces noisy annotations due to mismatches between superpixel and ground truth segmentation. To address this issue, we further devise a sieving mechanism that identifies and excludes potentially noisy annotations from learning. Our experiments on both Cityscapes and PASCAL VOC datasets demonstrate the efficacy of adaptive superpixel and sieving mechanisms. | ['Jungseul Ok', 'Suha Kwak', 'Sehyun Hwang', 'Minhyeon Oh', 'Hoyoung Kim'] | 2023-03-29 | null | null | null | null | ['superpixels'] | ['computer-vision'] | [ 3.70753020e-01 3.10488850e-01 -2.09952235e-01 -6.65492773e-01
-1.18608308e+00 -8.16002131e-01 6.65951595e-02 1.93237424e-01
-7.54876792e-01 8.54141831e-01 -4.84495282e-01 1.42363990e-02
3.49404454e-01 -9.07835424e-01 -8.40435624e-01 -8.57933402e-01
3.88723731e-01 3.04683268e-01 9.04978096e-01 4.06475514e-01
3.64241824e-02 5.68152368e-02 -1.57304764e+00 2.08216473e-01
1.40516090e+00 1.00038528e+00 4.60070461e-01 3.15170974e-01
-4.06341076e-01 5.82605898e-01 -5.75831532e-01 -2.47466043e-01
3.78020972e-01 -1.61829814e-01 -8.88459206e-01 4.22885090e-01
7.41059244e-01 -4.06894028e-01 4.83473033e-01 1.24322104e+00
2.00828001e-01 1.02969594e-01 1.57041743e-01 -1.04231942e+00
-3.31167996e-01 7.15498686e-01 -7.23305345e-01 9.88760889e-02
-9.57172439e-02 1.50167510e-01 1.11782551e+00 -8.41905951e-01
4.90936697e-01 8.06802630e-01 6.29475534e-01 4.75744009e-01
-1.28471577e+00 -7.43025064e-01 5.08690476e-01 -1.16380051e-01
-1.48880398e+00 -2.15834305e-01 7.46293783e-01 -2.86617547e-01
1.73663288e-01 4.15679514e-01 8.14617336e-01 8.00938189e-01
-5.95408201e-01 1.33836484e+00 1.11054516e+00 -3.88004601e-01
6.45336211e-01 1.96410075e-01 3.59311283e-01 6.95357382e-01
2.75462359e-01 -4.48713839e-01 -5.26763976e-01 -8.71295109e-02
9.81151104e-01 1.09966900e-02 -3.93717527e-01 -5.44521689e-01
-1.02766156e+00 5.92961609e-01 4.48947728e-01 2.55168825e-01
-2.77370334e-01 2.96022773e-01 9.41542238e-02 -1.93348274e-01
6.74038470e-01 2.02897280e-01 -6.26358807e-01 1.04598023e-01
-1.25615907e+00 -2.68411785e-01 4.51100290e-01 1.04641342e+00
1.34960556e+00 -4.16656911e-01 -2.60621548e-01 9.06864583e-01
4.17102695e-01 1.24040417e-01 5.18807434e-02 -1.32559550e+00
4.24187690e-01 1.06596541e+00 4.63303983e-01 -6.88344002e-01
-1.09012209e-01 -5.51287234e-01 -3.75926793e-01 2.81532198e-01
5.26152849e-01 -7.17466772e-02 -1.08224690e+00 1.33970881e+00
7.41070271e-01 3.97131324e-01 -4.57509160e-01 1.18654561e+00
7.29829371e-01 3.76366526e-01 3.20023537e-01 -3.20359290e-01
9.75659847e-01 -1.29428661e+00 -4.74775910e-01 -2.75451481e-01
6.72568500e-01 -4.68358994e-01 1.38924444e+00 3.76470566e-01
-9.71934497e-01 -6.79022849e-01 -1.01394498e+00 -1.66141212e-01
-2.73623407e-01 4.61081237e-01 6.96802318e-01 7.34801710e-01
-1.00273120e+00 3.77084613e-01 -9.98043180e-01 -5.94019704e-02
6.38275802e-01 3.71595770e-01 2.49016471e-02 1.54135421e-01
-9.75930512e-01 1.13481790e-01 5.82124591e-01 1.38281107e-01
-8.61237824e-01 -7.38941669e-01 -7.52296746e-01 -2.33140457e-02
8.45002890e-01 -3.44127774e-01 1.29907489e+00 -1.55151069e+00
-1.34239376e+00 8.63420546e-01 -1.96578071e-01 -2.12408945e-01
6.41510427e-01 -3.50554258e-01 1.09703422e-01 1.61144495e-01
2.95379966e-01 9.55799997e-01 8.05052340e-01 -1.54339898e+00
-9.06167269e-01 -2.68349022e-01 3.55477482e-01 2.84338981e-01
-6.05610967e-01 -1.73481315e-01 -1.04228890e+00 -4.92341697e-01
4.69995141e-01 -6.70326531e-01 -4.75714713e-01 3.65170121e-01
-4.54166085e-01 -4.01611358e-01 7.97605693e-01 -2.20192537e-01
1.24045384e+00 -2.13397813e+00 -2.15374112e-01 8.13053250e-02
2.52266496e-01 2.59679914e-01 7.62652755e-02 -2.69498587e-01
4.32888418e-01 1.47551596e-01 -5.21195769e-01 -6.20640278e-01
-2.66089231e-01 2.68530637e-01 -6.19094670e-02 1.88324928e-01
1.15607463e-01 8.28872204e-01 -1.25575149e+00 -1.05943131e+00
2.46671498e-01 3.74246567e-01 -4.09422189e-01 1.11639239e-01
-5.38961768e-01 5.99372089e-01 -5.98503768e-01 8.34959447e-01
8.81226480e-01 -4.75345314e-01 1.30395545e-02 -1.84293494e-01
-4.22408313e-01 1.18715763e-01 -1.27153254e+00 1.92667472e+00
-2.29356736e-01 2.44407341e-01 2.35928923e-01 -9.53342497e-01
8.38418603e-01 1.24465905e-01 4.64883298e-01 -4.13646847e-01
2.70697288e-02 3.40373784e-01 -6.78688824e-01 -3.61665994e-01
4.09219086e-01 4.42507356e-01 8.55810046e-02 5.05725205e-01
1.50711061e-02 5.66876344e-02 3.20520610e-01 8.09779391e-02
9.01626945e-01 5.44159234e-01 -2.17664782e-02 -2.68840015e-01
4.85174537e-01 1.21102557e-01 9.64088023e-01 9.05041218e-01
-5.91089487e-01 8.27271879e-01 2.25746736e-01 -3.36973190e-01
-8.75848830e-01 -9.64318395e-01 -2.80222222e-02 1.38416600e+00
6.29769385e-01 -3.16565156e-01 -1.07072580e+00 -1.07801735e+00
-2.61277020e-01 4.46844459e-01 -6.12065196e-01 3.80550236e-01
-5.20882487e-01 -8.53615344e-01 3.46048623e-01 7.39572704e-01
8.94477904e-01 -8.10536742e-01 -7.38119602e-01 1.80048496e-01
-3.46907973e-01 -9.61910486e-01 -5.79549968e-01 2.67069668e-01
-9.35443282e-01 -1.01068306e+00 -7.44753420e-01 -8.14564288e-01
1.02183187e+00 4.32238400e-01 1.05497885e+00 1.05587177e-01
-1.63167626e-01 2.07019955e-01 -5.79364598e-01 -3.53097618e-01
4.14428711e-02 2.41480693e-01 -5.70703268e-01 2.24412009e-01
4.10090506e-01 -2.01094434e-01 -8.85718763e-01 4.51432228e-01
-8.91370833e-01 2.80507326e-01 5.21072090e-01 5.46491861e-01
1.20972610e+00 -1.74585357e-02 1.77331612e-01 -1.26048875e+00
-5.04738744e-03 -2.97854513e-01 -9.46798205e-01 4.92167473e-01
-4.19898063e-01 -2.23327532e-01 4.31677699e-01 -3.30629200e-01
-1.27482855e+00 6.55752122e-01 2.49493614e-01 -3.21604908e-01
-2.71008760e-01 1.78392112e-01 -2.24589631e-01 -5.33819869e-02
5.44300556e-01 1.44895971e-01 -5.13775647e-01 -5.86921871e-01
2.95289457e-01 7.81941593e-01 4.38310504e-01 -4.21988338e-01
5.54654062e-01 8.06716859e-01 -6.58548713e-01 -4.32811379e-01
-1.35619462e+00 -8.60177755e-01 -9.10402596e-01 -3.97211611e-01
8.39328587e-01 -9.40335631e-01 -3.96532685e-01 4.04797971e-01
-1.05330670e+00 -6.68641806e-01 -3.83933485e-01 2.82885700e-01
-2.65804261e-01 4.65800971e-01 -4.78929907e-01 -8.83870602e-01
-2.82634288e-01 -1.00036204e+00 1.33815515e+00 5.42330742e-01
-1.39539003e-01 -7.68502355e-01 -1.05522074e-01 7.31314778e-01
7.03680292e-02 1.65479675e-01 3.03503066e-01 -2.36034393e-01
-1.04208708e+00 -4.52816710e-02 -4.25792098e-01 4.39098209e-01
2.07588300e-01 -1.28156729e-02 -1.17308998e+00 -2.12632060e-01
-2.39253461e-01 -4.87755537e-01 1.04781425e+00 1.81446210e-01
1.51743090e+00 -4.63560559e-02 -3.85994434e-01 5.05749881e-01
1.52051151e+00 1.07520387e-01 3.23630571e-01 3.44948113e-01
8.40483844e-01 5.20454526e-01 1.03808355e+00 3.13366383e-01
3.22453678e-01 4.08080310e-01 5.50542831e-01 -5.27410030e-01
-1.58840418e-01 -1.46433130e-01 1.69480294e-01 4.75172788e-01
-8.62573683e-02 -1.90880850e-01 -8.82637620e-01 6.37155116e-01
-1.95956624e+00 -5.04626513e-01 -3.04366529e-01 2.26097202e+00
1.22010088e+00 2.92697132e-01 -6.35808259e-02 2.97280997e-01
8.52780938e-01 4.85671312e-02 -7.91048110e-01 3.22563648e-01
-3.67440209e-02 1.93061158e-01 6.59185588e-01 5.12765527e-01
-1.44749832e+00 1.20123136e+00 5.57571936e+00 9.75589335e-01
-9.55357552e-01 4.16225851e-01 8.91455293e-01 8.14798623e-02
-2.68692553e-01 9.79158655e-02 -8.64230633e-01 6.01453543e-01
1.22261889e-01 5.01205504e-01 -7.52048101e-03 1.11600292e+00
2.30480418e-01 -5.85929751e-01 -8.72255802e-01 7.95927882e-01
-1.26055554e-01 -1.01983976e+00 -1.72492430e-01 -3.35667044e-01
1.09492421e+00 2.33001448e-03 -2.03438461e-01 -1.07451707e-01
5.60027242e-01 -6.61168396e-01 7.84576952e-01 2.25006416e-01
5.82684875e-01 -4.48735535e-01 7.29294837e-01 4.42603618e-01
-1.20305300e+00 -3.92913446e-02 -4.31414098e-01 1.13230869e-01
4.52964082e-02 9.10559952e-01 -7.19207227e-01 2.96622455e-01
1.02173960e+00 5.04547000e-01 -6.37696862e-01 1.23875499e+00
-6.26933813e-01 8.59322429e-01 -5.06827354e-01 8.19551945e-02
2.57302254e-01 -3.18650544e-01 1.65181011e-01 1.05239236e+00
2.53206305e-03 7.91513026e-02 4.87799346e-01 1.16302514e+00
-4.56753522e-02 9.66351703e-02 1.41801178e-01 2.73541361e-01
8.45947385e-01 1.52538979e+00 -1.26905966e+00 -4.94118989e-01
-4.56420749e-01 1.30101740e+00 4.43324834e-01 5.90246141e-01
-9.58367884e-01 -2.06050143e-01 6.85263425e-02 7.90122300e-02
3.87001783e-01 -2.33596802e-01 -5.93389988e-01 -1.01590145e+00
1.33094668e-01 -2.30589062e-01 3.46390694e-01 -6.66844487e-01
-9.26304162e-01 2.60585219e-01 -2.33081132e-01 -1.27144909e+00
4.63370949e-01 -1.07938156e-01 -4.81617361e-01 5.15660882e-01
-1.56119609e+00 -1.29781842e+00 -9.13151741e-01 4.50680375e-01
6.94953144e-01 6.01102829e-01 5.22155583e-01 4.36700255e-01
-8.02852869e-01 5.36554873e-01 -6.73247799e-02 3.06884766e-01
6.58713877e-01 -1.42616522e+00 4.51403745e-02 8.03077340e-01
2.09636062e-01 3.51367921e-01 3.70354235e-01 -5.99782765e-01
-6.08724117e-01 -1.40785992e+00 6.54116392e-01 -3.50379854e-01
3.31277102e-01 -4.29749548e-01 -8.42913568e-01 6.17215872e-01
-1.81966528e-01 3.06752920e-01 6.82257950e-01 -1.27338186e-01
-9.55573767e-02 -3.19692165e-01 -9.34959710e-01 3.77695739e-01
1.11809635e+00 -2.91664749e-01 -1.68587074e-01 3.53246242e-01
7.61023760e-01 -2.96163708e-01 -6.65505052e-01 5.93976915e-01
2.42105633e-01 -1.10085642e+00 8.49659920e-01 8.39869902e-02
5.14672101e-02 -7.65859246e-01 1.76959693e-01 -8.11613202e-01
-1.96584523e-01 -3.65147471e-01 1.51779994e-01 1.51060534e+00
5.95058501e-01 -3.71496975e-01 1.20910811e+00 7.73904979e-01
-2.46716514e-01 -7.50696301e-01 -7.32265294e-01 -7.25284040e-01
-3.45776200e-01 -3.87825042e-01 4.95252252e-01 9.89039838e-01
-3.30823213e-01 -3.58768702e-02 1.60696544e-02 2.64285713e-01
8.19963992e-01 2.87482440e-01 6.66248262e-01 -1.29228723e+00
-2.80321091e-01 -3.86945643e-02 5.04453257e-02 -1.35121763e+00
-1.59544736e-01 -4.57933575e-01 5.11586130e-01 -1.52415729e+00
3.95195901e-01 -1.02640307e+00 -5.25009096e-01 8.71528268e-01
-5.65479517e-01 8.04426551e-01 -2.19876096e-01 5.74432373e-01
-1.27710247e+00 2.59894669e-01 1.37041795e+00 -2.64971793e-01
-4.69396204e-01 -9.11213681e-02 -3.80498707e-01 9.85066652e-01
7.98831522e-01 -3.69008631e-01 -5.81830978e-01 -6.31718218e-01
1.67703241e-01 -4.82233405e-01 3.86498004e-01 -9.38774765e-01
3.49858522e-01 -3.11128765e-01 3.07968199e-01 -6.10229492e-01
1.87839985e-01 -6.78662479e-01 -4.87246923e-02 1.61912143e-01
-4.24874336e-01 -7.96776891e-01 -2.27381542e-01 6.66335225e-01
-3.14639866e-01 -3.05328816e-01 8.38285029e-01 -3.79009038e-01
-9.95284140e-01 3.43401641e-01 -5.34811243e-02 -1.31875306e-01
1.36687756e+00 -6.22306049e-01 -2.48025227e-02 1.26892194e-01
-6.92691743e-01 5.01353741e-01 7.35884309e-01 5.04188724e-02
2.75337547e-01 -9.64049995e-01 -2.63979763e-01 4.14522225e-03
2.31072053e-01 8.55251074e-01 1.30486622e-01 9.75532293e-01
-6.00826681e-01 8.04152414e-02 2.50920922e-01 -8.42656970e-01
-1.37391973e+00 3.29198033e-01 2.65997499e-01 1.63335964e-04
-3.25042695e-01 1.38785875e+00 4.32117283e-01 -2.86166459e-01
4.84347939e-01 -3.53288084e-01 -2.45818958e-01 8.82194042e-02
2.40899712e-01 4.04454529e-01 -2.61390004e-02 -3.51566941e-01
-2.49045372e-01 5.44015467e-01 1.05507476e-02 3.59011218e-02
1.01636744e+00 -4.47562665e-01 -1.71073154e-01 5.60458064e-01
8.46206307e-01 -4.29335907e-02 -1.85610211e+00 -2.27118731e-01
1.67081669e-01 -5.09424329e-01 1.81783319e-01 -6.67406082e-01
-1.30777895e+00 7.10427344e-01 5.59893668e-01 2.66479760e-01
1.17548060e+00 2.38716036e-01 9.14732397e-01 5.37161119e-02
6.21837378e-01 -1.60847998e+00 1.52705446e-01 3.50675657e-02
2.49490097e-01 -1.44173992e+00 -7.55439326e-02 -1.11099589e+00
-3.08435321e-01 8.77122641e-01 1.02717876e+00 -3.64023261e-03
3.65617692e-01 1.81372628e-01 3.06378603e-01 -1.54321715e-02
-2.32936606e-01 -3.84580404e-01 6.72514811e-02 3.51716191e-01
3.94293427e-01 1.12797946e-01 -6.82047963e-01 4.73548561e-01
2.26746231e-01 1.64702654e-01 2.25977078e-01 1.01520431e+00
-7.35669255e-01 -1.13183749e+00 -3.75741035e-01 2.41147682e-01
-2.41631061e-01 9.42553766e-03 -4.74710107e-01 4.44878459e-01
7.70535827e-01 9.51535821e-01 2.70877063e-01 -2.86610480e-02
-7.84592852e-02 -9.09541175e-02 2.35099465e-01 -9.46062207e-01
-3.91282380e-01 2.70843029e-01 -8.09300691e-02 -6.66165531e-01
-8.83302152e-01 -5.11956275e-01 -1.48509455e+00 3.29697996e-01
-9.19257998e-01 1.90322071e-01 3.21346015e-01 8.47557664e-01
2.92087913e-01 2.54921854e-01 4.92116272e-01 -6.16973996e-01
-2.58181453e-01 -8.23796272e-01 -6.14813864e-01 5.01253605e-01
3.65159474e-02 -4.86479282e-01 -2.64594615e-01 4.39223558e-01] | [9.486536026000977, 0.6836819648742676] |
6edb98c2-1bfb-43ad-acd4-d0062f0bdf7e | improving-prosody-for-cross-speaker-style | 2303.07711 | null | https://arxiv.org/abs/2303.07711v1 | https://arxiv.org/pdf/2303.07711v1.pdf | Improving Prosody for Cross-Speaker Style Transfer by Semi-Supervised Style Extractor and Hierarchical Modeling in Speech Synthesis | Cross-speaker style transfer in speech synthesis aims at transferring a style from source speaker to synthesized speech of a target speaker's timbre. In most previous methods, the synthesized fine-grained prosody features often represent the source speaker's average style, similar to the one-to-many problem(i.e., multiple prosody variations correspond to the same text). In response to this problem, a strength-controlled semi-supervised style extractor is proposed to disentangle the style from content and timbre, improving the representation and interpretability of the global style embedding, which can alleviate the one-to-many mapping and data imbalance problems in prosody prediction. A hierarchical prosody predictor is proposed to improve prosody modeling. We find that better style transfer can be achieved by using the source speaker's prosody features that are easily predicted. Additionally, a speaker-transfer-wise cycle consistency loss is proposed to assist the model in learning unseen style-timbre combinations during the training phase. Experimental results show that the method outperforms the baseline. We provide a website with audio samples. | ['Zhongyuan Wang', 'Xiaorui Wang', 'Ying Zhang', 'Hao Che', 'Peng Yang', 'Chunyu Qiang'] | 2023-03-14 | null | null | null | null | ['prosody-prediction', 'speech-synthesis'] | ['natural-language-processing', 'speech'] | [ 3.24113011e-01 -1.71533480e-01 -3.90767813e-01 -6.45854115e-01
-9.93169487e-01 -6.81339264e-01 2.48799920e-01 -3.30328405e-01
-3.82541269e-02 5.66447258e-01 6.17920339e-01 1.16487443e-01
4.14679378e-01 -3.78225982e-01 -4.36114639e-01 -7.34804153e-01
4.70600754e-01 8.01423043e-02 -2.54589081e-01 -3.49346101e-01
-3.06769218e-02 1.53577864e-01 -1.32744396e+00 5.71142733e-01
7.52916396e-01 1.12601376e+00 3.56901973e-01 7.18788445e-01
-2.94101030e-01 6.14233077e-01 -8.66569340e-01 -1.24848939e-01
3.30939051e-03 -1.17090082e+00 -5.45963883e-01 5.17623499e-02
2.76659012e-01 -1.50932521e-02 -1.41319811e-01 9.88872170e-01
6.93946481e-01 2.19680592e-02 6.34862006e-01 -1.08943987e+00
-5.09774268e-01 7.08681703e-01 -2.67270207e-01 6.85555302e-03
3.23224992e-01 3.65954340e-02 1.27426696e+00 -9.99751985e-01
2.73523301e-01 1.64105976e+00 6.03333950e-01 8.38440597e-01
-1.57758379e+00 -9.49301720e-01 1.69772074e-01 2.65625298e-01
-1.32438648e+00 -8.30447555e-01 1.63854122e+00 -3.13773602e-01
4.82937783e-01 5.13345420e-01 6.31607115e-01 1.36608028e+00
7.55093293e-03 8.36827993e-01 9.67372417e-01 -5.41380405e-01
1.61791757e-01 5.55677831e-01 -9.75134671e-02 1.80468425e-01
-5.72965026e-01 2.11579472e-01 -1.07478309e+00 -8.21430013e-02
5.38346708e-01 -5.87024331e-01 -5.26389301e-01 -2.23090321e-01
-1.15476477e+00 7.59827495e-01 1.62932426e-01 3.14941347e-01
-1.60300121e-01 -1.34652078e-01 7.73923635e-01 6.55376673e-01
6.20736837e-01 4.38119084e-01 -5.71145833e-01 -3.13337743e-01
-1.01062560e+00 9.25108939e-02 7.63755739e-01 9.62059498e-01
6.43133640e-01 6.65277004e-01 -2.63240635e-01 1.34184396e+00
2.58476734e-01 4.84013289e-01 6.72026873e-01 -9.39875126e-01
4.27115679e-01 1.57926455e-01 -1.31731927e-02 -6.07217014e-01
-2.49840822e-02 -3.58288407e-01 -7.99796164e-01 3.67163233e-02
3.23107332e-01 -1.23232387e-01 -4.07600492e-01 2.13496566e+00
1.42551690e-01 7.63014629e-02 5.49259260e-02 9.00283098e-01
4.92625237e-01 1.03157938e+00 3.37344892e-02 -6.13453865e-01
1.41896844e+00 -9.32600498e-01 -1.11239958e+00 -3.71763796e-01
2.02474773e-01 -9.77285624e-01 1.56967258e+00 4.21098083e-01
-1.14919508e+00 -1.05708480e+00 -1.09750056e+00 -4.74941963e-03
1.06165864e-01 2.91002512e-01 -1.95978303e-02 8.83077502e-01
-5.80500126e-01 5.94259202e-01 -4.81157869e-01 5.36470041e-02
-6.30436139e-03 1.91260204e-01 1.13429539e-01 3.52861166e-01
-1.40612853e+00 6.26875579e-01 4.30594027e-01 -1.75898090e-01
-5.92228830e-01 -1.02897763e+00 -7.76251912e-01 7.61715323e-02
-1.52320296e-01 -5.24596572e-01 1.27799618e+00 -1.54796720e+00
-2.15202665e+00 6.16850317e-01 -5.83654702e-01 -1.64864510e-01
2.49655366e-01 -1.03181407e-01 -7.06547022e-01 7.10846186e-02
-2.04205468e-01 6.30301297e-01 1.16346979e+00 -1.28785229e+00
-5.20497441e-01 -1.62679031e-01 -6.03306174e-01 5.04795909e-01
-7.16098785e-01 2.20043436e-01 -2.51453698e-01 -1.13257813e+00
-2.44903434e-02 -9.97791350e-01 3.53251696e-01 5.01644276e-02
-1.58014625e-01 -2.29159445e-01 8.10087204e-01 -8.54731619e-01
1.46233201e+00 -2.47001743e+00 5.48937380e-01 -6.08247556e-02
-2.79365450e-01 1.10016041e-01 -2.01804146e-01 1.88674927e-01
-1.02609456e-01 -2.36860842e-01 -5.70237748e-02 -6.58764958e-01
2.55950447e-02 1.81923389e-01 -6.05404615e-01 1.12043604e-01
2.66957998e-01 4.40147966e-01 -7.14781225e-01 -3.52425724e-01
-1.08434092e-02 5.07810235e-01 -7.09749103e-01 5.29610276e-01
-1.28563747e-01 8.39485765e-01 -3.55488621e-02 1.78499267e-01
3.93093765e-01 3.13701332e-01 2.90825427e-01 -4.43262517e-01
1.63810968e-01 9.24625039e-01 -1.00759709e+00 1.87433612e+00
-8.83413553e-01 6.20505273e-01 1.62724435e-01 -6.46410763e-01
1.34558988e+00 7.59285033e-01 2.14424878e-01 -4.39870447e-01
-8.61804932e-03 2.11885735e-01 1.97158173e-01 3.06062195e-02
2.58271813e-01 -6.35263443e-01 -2.08064079e-01 2.87572592e-01
1.52890176e-01 -4.45361972e-01 -3.96263748e-01 -4.78660226e-01
4.25798118e-01 2.69444846e-02 3.45726311e-01 -4.29783255e-01
7.10872233e-01 -4.48576748e-01 1.05625308e+00 2.15433806e-01
-2.98285365e-01 7.01302409e-01 2.10341200e-01 -1.08188376e-01
-1.03053272e+00 -1.17055190e+00 -1.38123101e-02 1.37303734e+00
-1.09858193e-01 -4.70247447e-01 -9.31413829e-01 -3.12520027e-01
-3.07823539e-01 1.04984736e+00 -3.69136602e-01 -4.18984026e-01
-8.55174720e-01 -8.77573788e-02 6.55902743e-01 5.19341409e-01
1.17919602e-01 -1.18730736e+00 8.07671845e-02 4.05933589e-01
-4.90687162e-01 -8.46656621e-01 -1.24941707e+00 6.24629110e-02
-7.37775803e-01 -3.37224066e-01 -5.70138276e-01 -9.17965233e-01
9.25587565e-02 3.15435864e-02 9.62875962e-01 -5.01540542e-01
2.15135828e-01 -2.64586229e-02 -2.93783754e-01 -5.23776233e-01
-1.02393603e+00 5.74715324e-02 4.50371683e-01 2.74193376e-01
3.80463719e-01 -7.50764668e-01 -4.37854171e-01 3.95466268e-01
-4.71406430e-01 1.88009858e-01 8.77675489e-02 1.13662100e+00
5.41262805e-01 2.87004933e-02 1.05195332e+00 -5.41289151e-01
6.32099450e-01 -2.01952830e-01 -2.46481836e-01 1.88972503e-02
-5.70218265e-01 1.02965035e-01 1.08682990e+00 -9.66201842e-01
-1.39793956e+00 -7.00102607e-03 -2.94544905e-01 -7.12101758e-01
-2.23771706e-01 7.89740235e-02 -7.33118057e-01 3.31669271e-01
6.27237976e-01 4.99711305e-01 8.76474530e-02 -6.12063348e-01
4.25771832e-01 8.96237016e-01 4.40581948e-01 -6.34177566e-01
6.58435583e-01 -6.10031784e-02 -4.56275702e-01 -1.10536885e+00
-9.69275177e-01 -2.96725988e-01 -4.62374002e-01 -2.68885735e-02
7.26103187e-01 -1.31125128e+00 -2.98116118e-01 4.78961617e-01
-1.34620190e+00 -3.11233252e-01 -4.62459356e-01 5.61233640e-01
-9.90869045e-01 1.60953775e-01 -8.16742420e-01 -8.22955012e-01
-4.11432981e-01 -1.06558657e+00 1.03048456e+00 -5.85862622e-02
-7.94973910e-01 -1.04207599e+00 1.39482945e-01 3.89556199e-01
3.32658440e-01 -2.26703599e-01 9.71027136e-01 -4.57032979e-01
-8.58599171e-02 2.79054493e-01 2.04933941e-01 8.92589211e-01
6.57019258e-01 -3.45718175e-01 -1.43945336e+00 -4.04553533e-01
6.89551830e-01 -2.00809628e-01 4.70089465e-01 1.95363417e-01
1.20126414e+00 -6.42030835e-01 1.15985587e-01 5.35645843e-01
7.47820556e-01 3.96002471e-01 4.34927285e-01 -3.44318867e-01
9.32813764e-01 7.36350179e-01 4.90530133e-01 2.89419174e-01
2.88434446e-01 9.54725623e-01 -3.81074518e-01 1.13632288e-02
-6.16466165e-01 -5.32408416e-01 9.25902367e-01 1.55798662e+00
3.96267831e-01 -1.98812946e-03 -3.38195354e-01 4.59661305e-01
-1.32433593e+00 -9.71184552e-01 2.81319082e-01 2.13574600e+00
1.50944042e+00 2.90736407e-01 3.26008648e-01 4.42151219e-01
7.54949749e-01 2.88179338e-01 -6.18316650e-01 -5.97195745e-01
-9.81706604e-02 9.73165706e-02 -1.95291087e-01 6.44701600e-01
-5.92044771e-01 9.37851310e-01 5.86949062e+00 1.25016427e+00
-1.46977520e+00 1.58294022e-01 5.53318560e-01 -2.58963674e-01
-5.43136179e-01 -1.58238858e-01 -8.02688539e-01 7.33525693e-01
1.19660401e+00 -4.04192895e-01 7.97387302e-01 7.92639852e-01
5.29627144e-01 5.67385077e-01 -1.57469726e+00 1.08778417e+00
1.18848093e-01 -9.71636057e-01 5.88566624e-03 -1.60820976e-01
5.17940044e-01 -4.08636957e-01 8.35522339e-02 4.15327370e-01
-3.63144875e-01 -6.88544333e-01 1.10846245e+00 3.54331672e-01
1.06612086e+00 -8.91001701e-01 9.74878073e-02 3.36141646e-01
-1.42478919e+00 -1.57702155e-02 -1.34890139e-01 6.32050857e-02
1.57164916e-01 3.10874760e-01 -9.47921038e-01 3.70782524e-01
3.76901567e-01 5.53891003e-01 -1.69524699e-01 4.37229186e-01
-1.57445341e-01 1.13412583e+00 -1.83790512e-02 3.69676016e-02
-3.68268132e-01 -2.13880748e-01 8.41310680e-01 1.37093973e+00
3.40427697e-01 -1.24497972e-01 1.01337679e-01 1.11674488e+00
-7.71227702e-02 2.41718903e-01 -4.05006111e-01 -6.14265054e-02
7.74230003e-01 9.22405779e-01 2.69938679e-03 -2.27074772e-01
-3.49990547e-01 1.15416813e+00 1.03901044e-01 3.72668982e-01
-6.18939161e-01 -3.26554120e-01 1.06751621e+00 -4.75760549e-02
1.56132534e-01 -2.23467499e-02 -5.31506658e-01 -1.05324018e+00
-3.96846496e-02 -1.13342190e+00 7.33223110e-02 -7.35976160e-01
-1.46638465e+00 7.83293664e-01 -3.46445888e-01 -1.44418478e+00
-5.70019603e-01 -4.10478979e-01 -8.50746572e-01 1.32355678e+00
-1.25282836e+00 -1.17105019e+00 1.67545751e-01 4.30871040e-01
1.20356297e+00 -3.73905987e-01 1.16285753e+00 2.35050485e-01
-3.80710542e-01 1.04332483e+00 -1.23370059e-01 -1.23681044e-02
1.09578228e+00 -1.20407128e+00 3.62603813e-01 5.33787131e-01
9.97529253e-02 4.37946469e-01 9.22077358e-01 -2.11395189e-01
-1.15448081e+00 -1.15860975e+00 9.68909264e-01 -2.38025486e-01
7.05373883e-01 -7.03166425e-01 -1.20674860e+00 3.05757493e-01
3.87445122e-01 -2.52850085e-01 1.04171097e+00 2.28800550e-01
-4.39594567e-01 -5.54149985e-01 -7.71500885e-01 6.74048424e-01
8.39232564e-01 -1.11340320e+00 -7.51899838e-01 -1.14616632e-01
9.16048288e-01 -2.32682750e-01 -8.84671807e-01 8.20292681e-02
6.61912501e-01 -7.03110874e-01 7.90698767e-01 -3.81569475e-01
1.76978514e-01 -2.67488748e-01 -3.47165436e-01 -1.78697610e+00
-2.85672098e-01 -7.87271738e-01 -1.86455652e-01 1.71974969e+00
3.47308248e-01 -1.72970429e-01 3.80721241e-01 2.15165198e-01
-2.17821673e-01 -3.24614137e-01 -9.35692668e-01 -1.00745559e+00
3.78758103e-01 -2.94138491e-01 6.37472272e-01 9.48582053e-01
1.26300931e-01 8.90680552e-01 -6.45021975e-01 -2.29094885e-02
4.62797463e-01 1.22637793e-01 6.06414795e-01 -1.10312772e+00
-7.36455321e-01 -5.23799777e-01 7.01836795e-02 -1.13205361e+00
5.75848639e-01 -8.16695452e-01 2.60254383e-01 -6.86776578e-01
-1.38167933e-01 -3.50088298e-01 -5.30656278e-01 9.85244066e-02
-3.44118088e-01 7.60854688e-03 3.95029157e-01 1.22754529e-01
-5.86068742e-02 1.00438201e+00 1.36605227e+00 -2.10624427e-01
-5.42918861e-01 3.14705968e-01 -6.63370073e-01 6.43617749e-01
9.30024803e-01 -4.81876463e-01 -6.11878932e-01 -5.68615161e-02
-3.41672093e-01 5.80884874e-01 -4.18822654e-03 -8.83472979e-01
-3.46356817e-02 -2.67675579e-01 1.73211873e-01 -3.74166489e-01
8.45517874e-01 -6.17708027e-01 1.51976928e-01 1.88757271e-01
-7.44041741e-01 -3.85717303e-01 2.75785834e-01 4.30820346e-01
-5.50214767e-01 -1.13297500e-01 1.08005905e+00 1.87111005e-01
-1.74331784e-01 2.66981032e-02 -2.48503432e-01 3.27450559e-02
4.34465498e-01 -2.36271545e-02 6.41663447e-02 -4.86535966e-01
-6.64323568e-01 -2.26146966e-01 2.25867823e-01 8.16214800e-01
3.50106537e-01 -1.93583715e+00 -8.93044770e-01 4.58119601e-01
2.75022894e-01 -3.33812505e-01 3.32285494e-01 4.09606993e-01
2.15247422e-01 1.34430483e-01 -1.76733077e-01 -5.92846930e-01
-1.33148468e+00 4.64283228e-01 3.33266377e-01 2.58964878e-02
-2.54029751e-01 8.03126216e-01 6.46877348e-01 -3.94739687e-01
2.44190991e-01 -2.20116168e-01 1.49644613e-01 2.33859017e-01
6.96896851e-01 2.76261359e-01 -3.30765955e-02 -6.92337751e-01
-2.44710609e-01 4.46756333e-01 -4.34166044e-02 -3.75396818e-01
1.03484261e+00 -5.34149349e-01 5.19284680e-02 1.23039854e+00
1.41472292e+00 3.66328746e-01 -1.53078520e+00 -4.79414523e-01
-2.14087561e-01 -2.88233757e-01 3.11576314e-02 -7.11921453e-01
-6.70997798e-01 1.12219596e+00 6.30616307e-01 1.18885152e-01
1.39443111e+00 -2.36772105e-01 8.54974210e-01 9.56634898e-03
1.46040156e-01 -1.29744339e+00 2.34320387e-01 5.46903372e-01
1.27011132e+00 -1.03358078e+00 -5.60432494e-01 -2.85882890e-01
-1.09879446e+00 1.11588240e+00 5.51053584e-01 1.31027758e-01
5.54740191e-01 1.25770226e-01 3.83042216e-01 5.23577213e-01
-8.61977339e-01 3.14988911e-01 7.37728775e-01 4.08511460e-01
7.00127661e-01 2.82397032e-01 5.85719496e-02 8.50195110e-01
-6.74226403e-01 -4.01918441e-01 1.47161409e-01 2.46687099e-01
-4.39064652e-01 -1.50922620e+00 -4.07125324e-01 -2.80157160e-02
-2.17650190e-01 -1.48732826e-01 -3.80292952e-01 1.03813432e-01
-8.54923353e-02 1.06866002e+00 7.36572817e-02 -5.41416526e-01
3.87670547e-01 6.04720712e-01 4.90719229e-01 -6.38383925e-01
-5.47795653e-01 6.66506767e-01 1.27693608e-01 -3.37169707e-01
-1.17979728e-01 -6.99010789e-01 -9.98719335e-01 -2.11990774e-02
-2.12594524e-01 1.34520546e-01 5.96363783e-01 7.86251366e-01
2.16564104e-01 7.61763573e-01 1.27820587e+00 -9.03725266e-01
-8.07800651e-01 -1.09674430e+00 -7.64844596e-01 5.75240672e-01
6.74917996e-01 -3.38952839e-01 -4.16648418e-01 3.56111467e-01] | [14.967974662780762, 6.545884609222412] |
9ad6a790-fd37-43ca-92a0-b7ee071346a8 | n-ltp-a-open-source-neural-chinese-language | 2009.11616 | null | https://arxiv.org/abs/2009.11616v4 | https://arxiv.org/pdf/2009.11616v4.pdf | N-LTP: An Open-source Neural Language Technology Platform for Chinese | We introduce \texttt{N-LTP}, an open-source neural language technology platform supporting six fundamental Chinese NLP tasks: {lexical analysis} (Chinese word segmentation, part-of-speech tagging, and named entity recognition), {syntactic parsing} (dependency parsing), and {semantic parsing} (semantic dependency parsing and semantic role labeling). Unlike the existing state-of-the-art toolkits, such as \texttt{Stanza}, that adopt an independent model for each task, \texttt{N-LTP} adopts the multi-task framework by using a shared pre-trained model, which has the advantage of capturing the shared knowledge across relevant Chinese tasks. In addition, a knowledge distillation method \cite{DBLP:journals/corr/abs-1907-04829} where the single-task model teaches the multi-task model is further introduced to encourage the multi-task model to surpass its single-task teacher. Finally, we provide a collection of easy-to-use APIs and a visualization tool to make users to use and view the processing results more easily and directly. To the best of our knowledge, this is the first toolkit to support six Chinese NLP fundamental tasks. Source code, documentation, and pre-trained models are available at \url{https://github.com/HIT-SCIR/ltp}. | ['Ting Liu', 'Wanxiang Che', 'Yunlong Feng', 'Libo Qin'] | 2020-09-24 | null | https://aclanthology.org/2021.emnlp-demo.6 | https://aclanthology.org/2021.emnlp-demo.6.pdf | emnlp-acl-2021-11 | ['lexical-analysis', 'semantic-dependency-parsing'] | ['natural-language-processing', 'natural-language-processing'] | [-7.37029761e-02 2.28998348e-01 -9.91785526e-02 -6.30659878e-01
-9.58458543e-01 -8.18675220e-01 1.80755481e-01 1.89854831e-01
-6.41569018e-01 6.77243114e-01 1.78410143e-01 -7.12180853e-01
3.13050508e-01 -5.46878219e-01 -6.43722415e-01 -5.37159920e-01
4.00698483e-01 4.67113912e-01 3.11501950e-01 -1.01079404e-01
1.11326575e-01 -5.79531826e-02 -9.77650642e-01 3.98099005e-01
1.10629237e+00 6.60481334e-01 8.05092096e-01 4.80547398e-01
-5.12138605e-01 6.99459970e-01 -5.88029027e-01 -6.26780272e-01
-1.88849822e-01 -1.87555358e-01 -1.09153521e+00 -6.79160714e-01
-1.05882540e-01 -1.45825043e-01 1.73764005e-01 1.14922655e+00
4.51330185e-01 1.63517743e-01 2.33801454e-01 -6.64806783e-01
-1.08807325e+00 9.85616386e-01 -1.92043483e-01 2.78638542e-01
2.90051907e-01 -1.39773130e-01 1.01681030e+00 -8.18095386e-01
4.85950589e-01 9.36671078e-01 3.62851709e-01 9.03664172e-01
-7.13036299e-01 -1.13263929e+00 5.24934828e-01 -7.26650134e-02
-9.57298875e-01 -1.75030574e-01 4.01592195e-01 -2.08913952e-01
1.23413825e+00 -4.35481183e-02 4.43453908e-01 1.12856472e+00
-7.99404681e-02 1.14766121e+00 1.20049930e+00 -6.45529211e-01
7.89357722e-03 -8.43831897e-02 5.98542988e-01 6.74322784e-01
2.86265165e-01 -4.55742598e-01 -4.87059087e-01 1.75624862e-01
7.18472540e-01 -5.81215024e-02 -1.07139610e-01 4.94161755e-01
-1.16652513e+00 5.68097413e-01 -5.65777682e-02 7.19460547e-01
-1.24109119e-01 1.13468237e-01 4.17808026e-01 1.15614913e-01
6.92515075e-01 3.25887054e-01 -1.23496401e+00 -4.53327179e-01
-6.04933619e-01 -2.31663257e-01 1.09238505e+00 1.33929276e+00
7.23557174e-01 7.03846961e-02 -4.41364162e-02 1.00068665e+00
3.69941473e-01 3.58474582e-01 5.90723813e-01 -9.36908543e-01
5.75936317e-01 5.73847830e-01 -1.94350943e-01 -9.09906626e-02
-3.16726476e-01 -1.53898269e-01 -5.05913675e-01 -4.38090712e-01
6.84941471e-01 -6.96824729e-01 -1.00227427e+00 1.79001069e+00
7.47890919e-02 1.78050533e-01 2.97796637e-01 5.80431640e-01
1.09169829e+00 6.66758478e-01 7.15095639e-01 -1.10629052e-01
1.70230711e+00 -1.09813714e+00 -8.21312129e-01 -4.74309772e-01
9.45631742e-01 -9.67401326e-01 1.23046374e+00 3.06192398e-01
-1.08414125e+00 -4.33031738e-01 -6.40649199e-01 -4.05646145e-01
-8.29773962e-01 3.42391551e-01 6.98943973e-01 4.77601051e-01
-1.10980678e+00 2.91774422e-01 -1.01025712e+00 -4.08007473e-01
2.05596864e-01 1.47832289e-01 -2.96977550e-01 -2.75890648e-01
-1.45840585e+00 9.07773554e-01 7.06628203e-01 3.77643034e-02
-7.08329439e-01 -6.92703426e-01 -9.41599429e-01 4.45913561e-02
6.62615716e-01 -2.60378271e-01 1.64405847e+00 -9.21102524e-01
-1.65451455e+00 9.76666868e-01 -4.09289926e-01 1.13057457e-01
-2.06711158e-01 -7.35050201e-01 -2.41312265e-01 5.89365885e-02
3.30301464e-01 6.54141068e-01 1.27494976e-01 -7.82655358e-01
-6.57330275e-01 -3.08293998e-01 8.74357894e-02 3.15205216e-01
-3.99030089e-01 7.51017749e-01 -9.26894248e-01 -7.24518716e-01
-1.92048952e-01 -8.66278291e-01 -2.87167251e-01 -6.37005150e-01
-4.49645877e-01 -6.47644758e-01 5.81658185e-01 -1.06605482e+00
1.28324723e+00 -2.05926943e+00 -2.53262401e-01 -3.35196465e-01
-8.99328440e-02 6.40465617e-01 -4.04479876e-02 5.12399971e-01
-8.42723250e-02 6.34108007e-01 -2.36789972e-01 -5.05186260e-01
-6.99394867e-02 4.23334777e-01 2.10108176e-01 -1.69290543e-01
1.50030315e-01 1.03317201e+00 -1.05274212e+00 -4.21484917e-01
-6.49472103e-02 5.00366449e-01 -2.14007899e-01 1.93362623e-01
-3.86766285e-01 5.22836387e-01 -8.34642351e-01 5.68793118e-01
4.63359505e-01 -3.73887748e-01 4.22134131e-01 3.16110730e-01
-4.83664811e-01 7.32970238e-01 -8.62923563e-01 1.85398901e+00
-3.83228362e-01 3.08281243e-01 2.32315376e-01 -7.15660095e-01
7.79970348e-01 8.53026390e-01 3.70223559e-02 -4.64248180e-01
2.00252697e-01 4.03800786e-01 -3.50046251e-03 -3.34313929e-01
2.71717608e-01 3.04069489e-01 -2.63599962e-01 6.19934738e-01
4.15222645e-01 1.51407076e-02 3.25045198e-01 1.88263804e-01
9.44958329e-01 5.49659550e-01 2.91501075e-01 -4.12893564e-01
3.03399175e-01 3.02216768e-01 8.09485078e-01 5.45478225e-01
-9.12927166e-02 3.78196239e-01 4.96945947e-01 -1.13918290e-01
-7.22087979e-01 -8.51036191e-01 -1.78445149e-02 1.64402282e+00
-3.21464151e-01 -4.73890066e-01 -1.06596816e+00 -7.38590777e-01
-4.37030077e-01 1.02502656e+00 -3.46486837e-01 3.54683638e-01
-8.51304471e-01 -7.29275644e-01 8.95559847e-01 8.07712615e-01
7.36112833e-01 -1.60899043e+00 -3.40711981e-01 2.74872988e-01
-3.13690513e-01 -1.34262025e+00 -5.93040764e-01 5.76972604e-01
-7.55707741e-01 -9.55552757e-01 -7.21350431e-01 -1.45516431e+00
6.15291715e-01 -2.83844047e-03 9.82256174e-01 1.38580322e-01
2.01425895e-01 3.45844418e-01 -6.20837688e-01 -6.75005436e-01
-1.51969358e-01 1.77856967e-01 -2.89903164e-01 -6.58697188e-01
8.14996362e-01 -3.04919690e-01 -2.93743491e-01 -6.69881254e-02
-7.63515592e-01 4.05254453e-01 6.19465172e-01 6.24233782e-01
7.11719334e-01 -2.62702078e-01 4.85015452e-01 -1.11909890e+00
6.45877481e-01 -4.04293269e-01 -5.12072384e-01 4.42532569e-01
-4.23547983e-01 -1.91402599e-01 7.26747572e-01 -3.45977426e-01
-1.62116694e+00 5.90764396e-02 -5.30262828e-01 -7.84527697e-03
-7.28847027e-01 7.47443318e-01 -3.95507991e-01 4.56687331e-01
6.17855638e-02 4.20281440e-01 -5.02346337e-01 -9.07037139e-01
5.05846202e-01 5.23469925e-01 5.07680893e-01 -9.10256743e-01
2.36364201e-01 4.73076738e-02 -6.81370556e-01 -7.48610973e-01
-1.34536099e+00 -5.85473776e-01 -9.93208528e-01 1.58684894e-01
1.49010980e+00 -1.14361215e+00 -5.99548101e-01 7.37437010e-01
-1.31322658e+00 -9.08578157e-01 -2.49308487e-03 4.85902667e-01
-1.44965917e-01 3.16107064e-01 -1.16808462e+00 -5.25877535e-01
-6.33951068e-01 -1.02689958e+00 8.64976287e-01 5.87276518e-01
-4.35027480e-02 -1.21532512e+00 -1.41786247e-01 4.99008119e-01
2.30803058e-01 -2.00177640e-01 1.00937295e+00 -1.36000502e+00
-2.67082870e-01 8.36118162e-02 -2.39549085e-01 6.35358155e-01
2.63732839e-02 -6.14884831e-02 -9.94012237e-01 5.42243272e-02
-8.04274529e-02 -3.38812798e-01 6.10011339e-01 4.68347758e-01
1.23967624e+00 -9.52763930e-02 -3.50209922e-01 4.98058051e-01
1.21665919e+00 5.10620534e-01 3.14740598e-01 3.10469806e-01
8.26624393e-01 5.62768221e-01 4.62894559e-01 -9.76813659e-02
8.06771576e-01 2.48399109e-01 -1.14894040e-01 -1.46124646e-01
3.18484940e-02 -2.55088061e-01 4.41245735e-01 1.18826282e+00
-1.25535756e-01 -2.51700699e-01 -1.26181948e+00 5.45754015e-01
-1.91038024e+00 -4.80544150e-01 -2.29028702e-01 1.82637978e+00
1.18617690e+00 -4.61997017e-02 -2.88591832e-01 -4.75528896e-01
5.69704771e-01 -2.76645347e-02 -2.64813423e-01 -6.20831490e-01
-5.72079271e-02 5.60175121e-01 3.87931913e-01 4.91473913e-01
-1.15695322e+00 1.79836941e+00 5.15654564e+00 8.36794496e-01
-9.80290890e-01 5.35077095e-01 6.18081808e-01 3.81810457e-01
-1.69466674e-01 2.00943187e-01 -1.28436029e+00 3.39667976e-01
1.09567797e+00 -5.76614253e-02 2.11543486e-01 9.16639209e-01
1.54820055e-01 -1.54298678e-01 -7.14929104e-01 3.27808857e-01
-1.29788071e-01 -1.17825019e+00 -2.37175703e-01 -7.63502195e-02
4.91408736e-01 5.67273796e-01 -3.74721378e-01 5.56195855e-01
9.79242921e-01 -6.82243228e-01 6.29811525e-01 1.06187962e-01
8.09792042e-01 -3.66082013e-01 6.49930716e-01 5.18373251e-01
-1.39936090e+00 1.44616038e-01 -1.85712621e-01 -9.80574861e-02
4.34229940e-01 5.45012832e-01 -3.31675887e-01 6.32300496e-01
1.06458187e+00 6.98409498e-01 -5.33384323e-01 7.24569499e-01
-1.01433754e+00 1.17251325e+00 -2.13760242e-01 -2.74790853e-01
2.31750488e-01 -2.36503914e-01 2.66963184e-01 1.67273283e+00
1.18524164e-01 5.14085233e-01 4.06105161e-01 5.67425787e-01
-4.68058325e-02 3.04806948e-01 -3.22063357e-01 -1.65664181e-01
6.46808684e-01 1.27340162e+00 -9.06068444e-01 -5.63463092e-01
-8.10899377e-01 9.33664322e-01 5.81001341e-01 5.62001526e-01
-5.92150986e-01 -5.75868607e-01 4.69986945e-01 -3.58312517e-01
4.14190441e-01 -5.33370316e-01 -3.47992420e-01 -1.32892501e+00
-1.30435988e-01 -7.37158179e-01 5.01418889e-01 -8.70106936e-01
-1.12644660e+00 6.26045227e-01 2.68021047e-01 -4.34530139e-01
-3.20762210e-02 -8.69541943e-01 -8.03340614e-01 1.06318641e+00
-1.52339876e+00 -1.48993301e+00 2.35252932e-01 6.65135026e-01
7.76075780e-01 9.90044326e-02 1.10056472e+00 2.07384512e-01
-8.45063090e-01 4.42813307e-01 -1.25536397e-01 6.98852777e-01
6.88791275e-01 -1.43493438e+00 5.22857308e-01 1.03073132e+00
-2.62682606e-02 1.04440355e+00 4.89812978e-02 -1.04500043e+00
-1.16214883e+00 -1.00679648e+00 1.45227194e+00 -4.36846226e-01
7.97376931e-01 -7.00886190e-01 -1.00427639e+00 1.21316481e+00
4.94255096e-01 -2.49841228e-01 1.01757622e+00 2.08349615e-01
-2.44121611e-01 3.14322889e-01 -6.77788079e-01 5.25462329e-01
9.38702762e-01 -4.34811771e-01 -7.80910492e-01 3.12746853e-01
1.20994794e+00 -3.97182703e-01 -9.37564909e-01 1.68435439e-01
4.28760439e-01 -4.70541418e-01 5.72373986e-01 -4.01317626e-01
1.71960145e-01 -9.20794979e-02 2.19027717e-02 -1.09620023e+00
-3.45407993e-01 -5.75656831e-01 1.90280974e-01 1.58849299e+00
9.48513567e-01 -8.10637832e-01 5.01718402e-01 5.18783808e-01
-7.09319413e-01 -6.78485513e-01 -8.51184249e-01 -5.63115716e-01
5.00107706e-01 -7.20472515e-01 3.34664315e-01 1.13674259e+00
2.12712154e-01 4.04693663e-01 1.94053142e-03 2.18468502e-01
3.41312021e-01 -2.18879014e-01 1.99715689e-01 -1.17909396e+00
-2.32610434e-01 -3.80212843e-01 6.32600963e-01 -1.09021163e+00
2.89252639e-01 -9.13685083e-01 1.26685023e-01 -1.76985931e+00
7.15122595e-02 -5.66885948e-01 -3.85611683e-01 1.19816267e+00
-4.76558030e-01 -1.61644340e-01 2.88105965e-01 1.62855372e-01
-8.89405131e-01 1.94357708e-01 1.29354692e+00 5.40712833e-01
-3.14784944e-01 1.71045586e-02 -1.07189226e+00 8.50534737e-01
1.07100904e+00 -7.02131212e-01 -1.05996944e-01 -1.04256916e+00
1.29199147e-01 -1.61760330e-01 -1.63554788e-01 -6.06915236e-01
4.40105110e-01 -1.59680948e-01 3.56753647e-01 -3.74592483e-01
2.00371131e-01 -3.33294332e-01 -3.20251852e-01 2.90221125e-01
-1.66242927e-01 2.83899218e-01 5.50503433e-01 1.77259650e-02
-7.22243711e-02 -4.16605890e-01 4.51585293e-01 -6.69656038e-01
-8.30974877e-01 1.40963435e-01 -7.46438026e-01 3.04487556e-01
8.80348742e-01 2.08831932e-02 -6.13474309e-01 -1.11991003e-01
-6.41963422e-01 5.59742212e-01 6.27501681e-02 5.70321202e-01
3.49090368e-01 -5.69375932e-01 -5.47352791e-01 6.23788219e-05
-1.13248691e-01 4.82901722e-01 8.65949839e-02 7.87369907e-01
-3.42254907e-01 7.08392739e-01 9.34081823e-02 -1.42449960e-01
-1.08001518e+00 2.70111859e-01 -2.24620160e-02 -5.98304808e-01
-4.84351873e-01 9.82005775e-01 2.93921620e-01 -8.63477051e-01
1.82778433e-01 -4.33110178e-01 -3.93351316e-01 -9.19764787e-02
4.91358131e-01 2.19007432e-02 -1.16023049e-01 -3.83157223e-01
-4.10783082e-01 4.00724262e-01 -2.53609508e-01 -1.66165680e-01
1.32709050e+00 -2.23649926e-02 -3.48317206e-01 4.33318019e-01
7.27515042e-01 1.71877742e-02 -1.04478955e+00 -2.46338725e-01
3.05261016e-01 4.06125218e-01 -2.23728374e-01 -1.29199767e+00
-8.22441697e-01 9.93574500e-01 -2.25096345e-02 -1.54422089e-01
1.08287251e+00 2.34496444e-01 8.51934314e-01 4.24804300e-01
2.68693000e-01 -1.20219207e+00 -1.28930420e-01 1.19827068e+00
3.86641890e-01 -1.08226407e+00 -2.91184932e-01 -5.64763844e-01
-9.39474523e-01 9.83000576e-01 1.03083551e+00 1.21831484e-01
1.00149322e+00 4.38777566e-01 4.60336745e-01 -2.06812978e-01
-7.61175394e-01 -4.25833166e-01 3.66930552e-02 5.78719199e-01
9.38665450e-01 1.12553865e-01 -5.15293419e-01 1.21085393e+00
-1.95392951e-01 7.28626624e-02 1.65390059e-01 1.19670022e+00
-3.18667442e-01 -1.51120448e+00 4.88558859e-02 2.65985429e-01
-1.05996275e+00 -7.10537970e-01 -3.26575637e-01 8.83362651e-01
1.58906505e-01 1.07586622e+00 -5.16982339e-02 2.68263668e-02
8.26222375e-02 4.75894719e-01 -1.94044430e-02 -1.19537044e+00
-1.06595862e+00 4.12534475e-01 2.71688938e-01 -3.31676990e-01
-2.46556655e-01 -6.09382451e-01 -1.62700069e+00 1.52730912e-01
-2.11933851e-01 3.97784770e-01 6.97943032e-01 1.11239314e+00
2.86559165e-01 4.29031253e-01 -1.61415234e-01 -6.30046070e-01
7.04336017e-02 -1.12267673e+00 -2.77548343e-01 -3.38930376e-02
-2.84280688e-01 -9.93066803e-02 -4.38329130e-02 1.61460444e-01] | [10.426592826843262, 9.886438369750977] |