mspitzna commited on
Commit
604734f
·
verified ·
1 Parent(s): e8eaec8

Add files using upload-large-folder tool

Browse files
ball_bounce/eval.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c2b9b3700c3e8b3c6a4ea54ee4e8e7e6f7a2bb8c297bd28ccc489f314026685
3
+ size 5635966
ball_bounce/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73f44301ff73102b6808f6ff698d4c20f75b3dcede7baf03086419623e39f972
3
+ size 157885979
ball_bounce/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e162198bb9e078f3f5ae9e5f77935b5e26a80f5a24362929c07300ed3a468aa1
3
+ size 4400772013
ball_roll/eval.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5d0269ba07df46b18ae07b45c21d4028e5d8eb4e06faea17d8884aa5787cc49
3
+ size 4910894
ball_roll/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5dab6232e19c6a93e6a3233f840fe79ea83f569926ea110cb73f6e46d4451f32
3
+ size 177317911
ball_roll/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a44ea2931cb69c7ab6572b22a7295656bc00f03cca5c304083f9eaa66337f9e
3
+ size 2733858105
lens_p1/eval.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e1069246dccb40616500356ae0c02955f05df0452cd60147963fb9ee8e54c0fa
3
+ size 121983
lens_p1/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d55477d86ef8805806b45905afe97defa3946ca0f210b13b404c1834a1a9f8bd
3
+ size 41675
lens_p1/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e630f98734088164ee063ac2b869ac4b2b3b091846a1024dd2d701107102508b
3
+ size 670287
lens_p2/eval.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6499f15410309a4948b25a17afc613c23c5c355eb7b45246d80b7edd70541696
3
+ size 121983
lens_p2/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9068e606c3f81d8943db24d291214d464e83b3cc389037bc9339a9d307153fdb
3
+ size 41675
lens_p2/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52881d2caf3baba771433e1c385ffa95002c4070b53a9edc7d7a6d7b91d17378
3
+ size 670287
physicsgen.py ADDED
@@ -0,0 +1,163 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ import csv
3
+ from PIL import Image
4
+ import datasets
5
+
6
+ # Define configurations for each flavor.
7
+ BUILDER_CONFIGS = [
8
+ datasets.BuilderConfig(
9
+ name="sound_baseline",
10
+ description="Physical dataset: baseline variant",
11
+ data_dir="./physicsgen/urban_sound_25k_baseline"
12
+ ),
13
+ datasets.BuilderConfig(
14
+ name="sound_reflection",
15
+ description="Physical dataset: reflection variant",
16
+ data_dir="./physicsgen/urban_sound_25k_reflection"
17
+ ),
18
+ datasets.BuilderConfig(
19
+ name="sound_diffraction",
20
+ description="Physical dataset: reflection variant",
21
+ data_dir="./physicsgen/urban_sound_25k_diffraction"
22
+ ),
23
+ datasets.BuilderConfig(
24
+ name="sound_combined",
25
+ description="Physical dataset: reflection variant",
26
+ data_dir="./physicsgen/urban_sound_25k_combined"
27
+ ),
28
+ datasets.BuilderConfig(
29
+ name="lens_p1",
30
+ description="Distortion dataset variant",
31
+ data_dir="./physicsgen/lens_distortion_p1"
32
+ ),
33
+ datasets.BuilderConfig(
34
+ name="lens_p2",
35
+ description="Distortion dataset variant",
36
+ data_dir="./physicsgen/lens_distortion_p2"
37
+ ),
38
+ datasets.BuilderConfig(
39
+ name="ball_roll",
40
+ description="Double image dataset variant",
41
+ data_dir="./physicsgen/ball_roll"
42
+ ),
43
+ datasets.BuilderConfig(
44
+ name="ball_bounce",
45
+ description="Double image dataset variant",
46
+ data_dir="./physicsgen/ball_bounce"
47
+ ),
48
+ ]
49
+
50
+ class MyPhysicalDataset(datasets.GeneratorBasedBuilder):
51
+ BUILDER_CONFIGS = BUILDER_CONFIGS
52
+ VERSION = datasets.Version("1.0.2")
53
+
54
+ def _info(self):
55
+ if self.config.name in ["sound_baseline", "sound_reflection", "sound_diffraction", "sound_combined"]:
56
+ features = datasets.Features({
57
+ "lat": datasets.Value("float"),
58
+ "long": datasets.Value("float"),
59
+ "db": datasets.Value("string"),
60
+ "soundmap": datasets.Image(),
61
+ "osm": datasets.Image(),
62
+ "temperature": datasets.Value("int32"),
63
+ "humidity": datasets.Value("int32"),
64
+ "yaw": datasets.Value("float"),
65
+ "sample_id": datasets.Value("int32"),
66
+ "soundmap_512": datasets.Image(),
67
+ })
68
+ elif self.config.name in ["lens_p1", "lens_p2"]:
69
+ features = datasets.Features({
70
+ "label_path": datasets.Value("string"),
71
+ "fx": datasets.Value("float"),
72
+ "k1": datasets.Value("float"),
73
+ "k2": datasets.Value("float"),
74
+ "k3": datasets.Value("float"),
75
+ "p1": datasets.Value("float"),
76
+ "p2": datasets.Value("float"),
77
+ "cx": datasets.Value("float"),
78
+ "distortion_path": datasets.Value("string"),
79
+ })
80
+ elif self.config.name in ["ball_roll", "ball_bounce"]:
81
+ features = datasets.Features({
82
+ "ImgName": datasets.Value("string"),
83
+ "StartHeight": datasets.Value("int32"),
84
+ "GroundIncli": datasets.Value("float"),
85
+ "InputTime": datasets.Value("int32"),
86
+ "TargetTime": datasets.Value("int32"),
87
+ "input_image": datasets.Image(),
88
+ "target_image": datasets.Image(),
89
+ })
90
+ else:
91
+ raise ValueError(f"Unknown config name: {self.config.name}")
92
+ return datasets.DatasetInfo(
93
+ description="Multiple variant physical tasks dataset.",
94
+ features=features,
95
+ )
96
+
97
+ def _split_generators(self, dl_manager):
98
+ data_dir = self.config.data_dir
99
+ return [
100
+ datasets.SplitGenerator(
101
+ name=datasets.Split.TRAIN,
102
+ gen_kwargs={"split_dir": os.path.join(data_dir, "train")},
103
+ ),
104
+ datasets.SplitGenerator(
105
+ name=datasets.Split.TEST,
106
+ gen_kwargs={"split_dir": os.path.join(data_dir, "test")},
107
+ ),
108
+ datasets.SplitGenerator(
109
+ name=datasets.Split.VALIDATION,
110
+ gen_kwargs={"split_dir": os.path.join(data_dir, "eval")},
111
+ ),
112
+ ]
113
+
114
+ def _generate_examples(self, split_dir):
115
+ if self.config.name in ["sound_baseline", "sound_reflection", "sound_diffraction", "sound_combined"]:
116
+ csv_path = os.path.join(split_dir, "meta_data.csv")
117
+ with open(csv_path, encoding="utf-8") as f:
118
+ reader = csv.DictReader(f)
119
+ for idx, row in enumerate(reader):
120
+ row["soundmap"] = os.path.join(split_dir, row["soundmap"])
121
+ row["osm"] = os.path.join(split_dir, row["osm"])
122
+ row["soundmap_512"] = os.path.join(split_dir, row["soundmap_512"])
123
+ row["lat"] = float(row["lat"])
124
+ row["long"] = float(row["long"])
125
+ row["temperature"] = int(row["temperature"])
126
+ row["humidity"] = int(row["humidity"])
127
+ row["sample_id"] = int(row["sample_id"])
128
+ row["yaw"] = float(row["yaw"]) if row["yaw"] else 0.0
129
+ yield idx, row
130
+
131
+ elif self.config.name in ["lens_p1", "lens_p2"]:
132
+ csv_path = os.path.join(split_dir, "meta_data.csv")
133
+ with open(csv_path, encoding="utf-8") as f:
134
+ reader = csv.DictReader(f)
135
+ for idx, row in enumerate(reader):
136
+ row["label_path"] = str(row["label_path"])
137
+ row["distortion_path"] = str(row["distortion_path"])
138
+ row["fx"] = float(row["fx"])
139
+ row["k1"] = float(row["k1"])
140
+ row["k2"] = float(row["k2"])
141
+ row["k3"] = float(row["k3"])
142
+ row["p1"] = float(row["p1"])
143
+ row["p2"] = float(row["p2"])
144
+ row["cx"] = float(row["cx"])
145
+ yield idx, row
146
+
147
+ elif self.config.name in ["ball_roll", "ball_bounce"]:
148
+ csv_path = os.path.join(split_dir, "meta_data.csv")
149
+ with open(csv_path, encoding="utf-8") as f:
150
+ reader = csv.DictReader(f)
151
+ for idx, row in enumerate(reader):
152
+ # Construct image path from ImgName, e.g., "DoubleImg_0.jpg"
153
+ image_filename = "DoubleImg_" + row["ImgName"] + ".jpg"
154
+ input_image_path = os.path.join(split_dir, "x", image_filename)
155
+ target_image_path = os.path.join(split_dir, "y", image_filename)
156
+ row["input_image"] = input_image_path
157
+ row["target_image"] = target_image_path
158
+ row["ImgName"] = row["ImgName"]
159
+ row["StartHeight"] = int(row["StartHeight"])
160
+ row["GroundIncli"] = float(row["GroundIncli"])
161
+ row["InputTime"] = int(row["InputTime"])
162
+ row["TargetTime"] = int(row["TargetTime"])
163
+ yield idx, row
sound_baseline/eval.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd3c0342d7e793207a0d1a02c0906c4aa7d205988d3d240af4195dfc8d375ec7
3
+ size 119262832
sound_baseline/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58384373f4c6c7ba284620be445a8a5c86674c9ee355b32cc7ee9186db3deba2
3
+ size 40109864
sound_baseline/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9069deff9b384e3c2d188af5af5b1a6c8942b1531f34464a34f65e80326ecf04
3
+ size 648507369
sound_combined/eval.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f579cc194a38c9a018103e13c5462988fc2102f4c5f463c5df1ce825d99d0fb7
3
+ size 222984612
sound_combined/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc7645396f0db7074a36bf2dbd00745435a4b2a7dc4c5e5e9735637b9d6415dc
3
+ size 75292721
sound_combined/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f831d6918d4de32c60c30967edd6897033bd3978f8b64e66c3b9ebfdab92f89d
3
+ size 1262623105
sound_diffraction/eval.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0613ad429cef5519a8574eaec99e882df295f70289d7b7b387196bf2e5206b32
3
+ size 195071676
sound_diffraction/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa6d9609b402dedec77135bddac8744d450df62baaafd2de9873020f3b725901
3
+ size 65603263
sound_diffraction/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a70b1fdb2443dafdc6caa1d021869bee688555821f53b982fa47a3a7b5e16055
3
+ size 1108509903
sound_reflection/eval.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fd8290688fe2707391924de7ce6a3c6c153921fc1f8453e628413c5fefe717d
3
+ size 187197405
sound_reflection/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c912b3e5154ba0534dbafc42bab3d141b190c91f3d891040a6adf5f03ab90aa2
3
+ size 63205053
sound_reflection/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e801031b02b05111cda20c96b49d393c96ff12fc9b4bcd35ae7b19ec4589b5d
3
+ size 1047687635