Datasets:

License:

Update README.md with data structure description

#2
by m-wosik - opened
Files changed (1) hide show
  1. README.md +70 -76
README.md CHANGED
@@ -8,7 +8,8 @@ language:
8
  - fr
9
  - it
10
  license:
11
- - mit
 
12
  multilinguality:
13
  - monolingual
14
  dataset_info:
@@ -25,7 +26,7 @@ dataset_info:
25
  ---
26
 
27
 
28
- # MOCKS dataset
29
 
30
  ## Table of Contents
31
  - [Table of Contents](#table-of-contents)
@@ -54,11 +55,7 @@ dataset_info:
54
 
55
  ## Dataset Description
56
 
57
- - **Homepage:**
58
- - **Repository:**
59
  - **Paper:**
60
- - **Leaderboard:**
61
- - **Point of Contact:**
62
 
63
  ### Dataset Summary
64
 
@@ -72,8 +69,6 @@ MOCKS contains both positive and negative examples selected based on phonetic tr
72
 
73
  Please refer to our [paper]() for further details.
74
 
75
- [More Information Needed - add link to paper]
76
-
77
  ### Supported Tasks and Leaderboards
78
 
79
  The MOCKS dataset can be used for Open-Vocabulary Keyword Spotting (OV-KWS) task. It supports two OV-KWS types:
@@ -95,22 +90,75 @@ The MOCKS incorporates 5 languages:
95
 
96
  ## Dataset Structure
97
 
98
- ### Data Instances
99
-
100
- [More Information Needed]
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
101
 
102
- ### Data Fields
 
 
 
103
 
104
- [More Information Needed]
 
 
 
 
 
105
 
106
- ### Data Splits
107
 
108
- The MOCKS testset is split by language, source dataset and OV-KWS type. Each split is divided into:
109
- - positive examples - test examples with true keyword, 5000-8000 keywords in each subset,
110
- - similar examples - test examples with similar phrases to keyword selected based on phonetic transcription distance,
111
- - different examples - test examples with completaly different prases.
112
 
113
- Each split also contains subset of whole data to allow faster evaluation.
 
 
 
114
 
115
  ## Dataset Creation
116
 
@@ -124,59 +172,9 @@ and 48kHz sampling rate for MCV based testset.
124
  The offline testset contains additional 0.1 second at the beginning and end of extracted audio sample to mitigate the cut-speech effect.
125
  The online version contrains additional 1 second or so at the beginning and end of extracted audio sample.
126
 
127
- ### Curation Rationale
128
-
129
- [More Information Needed]
130
-
131
- ### Source Data
132
-
133
- #### Initial Data Collection and Normalization
134
-
135
- [More Information Needed]
136
-
137
- #### Who are the source language producers?
138
-
139
- [More Information Needed]
140
-
141
- ### Annotations
142
-
143
- #### Annotation process
144
-
145
- [More Information Needed]
146
-
147
- #### Who are the annotators?
148
-
149
- [More Information Needed]
150
-
151
- ### Personal and Sensitive Information
152
 
153
- [More Information Needed]
154
-
155
- ## Considerations for Using the Data
156
-
157
- ### Social Impact of Dataset
158
-
159
- [More Information Needed]
160
-
161
- ### Discussion of Biases
162
-
163
- The MOCKS testset is speaker gender balanced.
164
-
165
- ### Other Known Limitations
166
-
167
- [More Information Needed]
168
-
169
- ## Additional Information
170
-
171
- ### Dataset Curators
172
-
173
- [More Information Needed]
174
-
175
- ### Licensing Information
176
-
177
- [More Information Needed]
178
-
179
- ### Citation Information
180
 
181
  ```bibtex
182
  @inproceedings{pudo23_interspeech,
@@ -185,8 +183,4 @@ The MOCKS testset is speaker gender balanced.
185
  year={in press.},
186
  booktitle={Proc. Interspeech 2023},
187
  }
188
- ```
189
-
190
- ### Contributions
191
-
192
- Thanks to [@github-username](https://github.com/<github-username>) for adding this dataset.
 
8
  - fr
9
  - it
10
  license:
11
+ - cc-by-4.0
12
+ - mpl-2.0
13
  multilinguality:
14
  - monolingual
15
  dataset_info:
 
26
  ---
27
 
28
 
29
+ # MOCKS: Multilingual Open Custom Keyword Spotting Testset
30
 
31
  ## Table of Contents
32
  - [Table of Contents](#table-of-contents)
 
55
 
56
  ## Dataset Description
57
 
 
 
58
  - **Paper:**
 
 
59
 
60
  ### Dataset Summary
61
 
 
69
 
70
  Please refer to our [paper]() for further details.
71
 
 
 
72
  ### Supported Tasks and Leaderboards
73
 
74
  The MOCKS dataset can be used for Open-Vocabulary Keyword Spotting (OV-KWS) task. It supports two OV-KWS types:
 
90
 
91
  ## Dataset Structure
92
 
93
+ The MOCKS testset is split by language, source dataset and OV-KWS type:
94
+ ```
95
+ MOCKS
96
+
97
+ └───de
98
+ │ └───MCV
99
+ │ │ └───test
100
+ │ │ │ └───offline
101
+ │ │ │ │ │ all.pair.different.tsv
102
+ │ │ │ │ │ all.pair.positive.tsv
103
+ │ │ │ │ │ all.pair.similar.tsv
104
+ │ │ │ │ │ data.tar.gz
105
+ │ │ │ │ │ subset.pair.different.tsv
106
+ │ │ │ │ │ subset.pair.positive.tsv
107
+ │ │ │ │ │ subset.pair.similar.tsv
108
+ │ │ │ │
109
+ │ │ │ └───online
110
+ │ │ │ │ │ all.pair.different.tsv
111
+ │ │ │ │ │ ...
112
+ │ │ │ │ data.offline.transcription.tsv
113
+ │ │ │ │ data.online.transcription.tsv
114
+
115
+ └───en
116
+ │ └───LS-clean
117
+ │ │ └───test
118
+ │ │ │ └───offline
119
+ │ │ │ │ │ all.pair.different.tsv
120
+ │ │ │ │ │ ...
121
+ │ │ │ │ ...
122
+ │ │
123
+ │ └───LS-other
124
+ │ │ └───test
125
+ │ │ │ └───offline
126
+ │ │ │ │ │ all.pair.different.tsv
127
+ │ │ │ │ │ ...
128
+ │ │ │ │ ...
129
+ │ │
130
+ │ └───MCV
131
+ │ │ └───test
132
+ │ │ │ └───offline
133
+ │ │ │ │ │ all.pair.different.tsv
134
+ │ │ │ │ │ ...
135
+ │ │ │ │ ...
136
+
137
+ └───...
138
+ ```
139
 
140
+ Each split is divided into:
141
+ - positive examples (`all.pair.positive.tsv`) - test examples with true keyword, 5000-8000 keywords in each subset,
142
+ - similar examples (`all.pair.similar.tsv`) - test examples with similar phrases to keyword selected based on phonetic transcription distance,
143
+ - different examples (`all.pair.different.tsv`) - test examples with completaly different prases.
144
 
145
+ All those files contain columns separated by tab:
146
+ - `keyword_path` - path to audio containing keyword phrase.
147
+ - `adversary_keyword_path` - path to test audio.
148
+ - `adversary_keyword_timestamp_start` - start time in seconds of phrase of interest for given keyword from `keyword_path`, field only available in **offline** split.
149
+ - `adversary_keyword_timestamp_end` - end time in seconds of phrase of interest for given keyword from `keyword_path`, field only available in **offline** split.
150
+ - `label` - whether the `adversary_keyword_path` contain keyword from `keyword_path` or not (1 - contains keyword, 0 - doesn't contain keyword).
151
 
152
+ Each split also contains subset of whole data with the same field sctructure to allow faster evaluation (`subset.pair.*.tsv`).
153
 
154
+ Also, trascriptions are provided for each audio in:
155
+ - `data_offline_transcription.tsv` - transcriptions for **offline** examples and `keyword_path` from **online** scenario,
156
+ - `data_online_transcription.tsv` - transcriptions for adversary, test examples from **online** scenario,
 
157
 
158
+ three columns are present within each file:
159
+ - `path_to_keyword`/`path_to_adversary_keyword` - path to audio file,
160
+ - `keyword_transcription`/`adversary_keyword_transcription` - audio transcription,
161
+ - `keyword_phonetic_transcription`/`adversary_keyword_phonetic_transcription` - audio phonetic transcription.
162
 
163
  ## Dataset Creation
164
 
 
172
  The offline testset contains additional 0.1 second at the beginning and end of extracted audio sample to mitigate the cut-speech effect.
173
  The online version contrains additional 1 second or so at the beginning and end of extracted audio sample.
174
 
175
+ The MOCKS testset is gender balanced.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
176
 
177
+ ## Citation Information
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
178
 
179
  ```bibtex
180
  @inproceedings{pudo23_interspeech,
 
183
  year={in press.},
184
  booktitle={Proc. Interspeech 2023},
185
  }
186
+ ```