Files changed (1) hide show
  1. README.md +160 -1
README.md CHANGED
@@ -8,4 +8,163 @@ source_datasets: []
8
  task_categories:
9
  - automatic-speech-recognition
10
  task_ids: []
11
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8
  task_categories:
9
  - automatic-speech-recognition
10
  task_ids: []
11
+ ---
12
+
13
+ # Dataset Card for Voxpopuli
14
+
15
+ ## Table of Contents
16
+ - [Table of Contents](#table-of-contents)
17
+ - [Dataset Description](#dataset-description)
18
+ - [Dataset Summary](#dataset-summary)
19
+ - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
20
+ - [Languages](#languages)
21
+ - [Dataset Structure](#dataset-structure)
22
+ - [Data Instances](#data-instances)
23
+ - [Data Fields](#data-fields)
24
+ - [Data Splits](#data-splits)
25
+ - [Dataset Creation](#dataset-creation)
26
+ - [Curation Rationale](#curation-rationale)
27
+ - [Source Data](#source-data)
28
+ - [Annotations](#annotations)
29
+ - [Personal and Sensitive Information](#personal-and-sensitive-information)
30
+ - [Considerations for Using the Data](#considerations-for-using-the-data)
31
+ - [Social Impact of Dataset](#social-impact-of-dataset)
32
+ - [Discussion of Biases](#discussion-of-biases)
33
+ - [Other Known Limitations](#other-known-limitations)
34
+ - [Additional Information](#additional-information)
35
+ - [Dataset Curators](#dataset-curators)
36
+ - [Licensing Information](#licensing-information)
37
+ - [Citation Information](#citation-information)
38
+ - [Contributions](#contributions)
39
+
40
+ ## Dataset Description
41
+
42
+ ### Dataset Summary
43
+
44
+ ### Example usage
45
+
46
+ HyVoxPopuli contains labelled data for 18 languages. To load a specific language pass its name as a config name:
47
+
48
+ ```python
49
+ from datasets import load_dataset
50
+
51
+ voxpopuli_croatian = load_dataset("Edmon02/hyvoxpopuli")
52
+ ```
53
+
54
+ ### Supported Tasks and Leaderboards
55
+
56
+ * automatic-speech-recognition: The dataset can be used to train a model for Automatic Speech Recognition (ASR). The model is presented with an audio file and asked to transcribe the audio file to written text. The most common evaluation metric is the word error rate (WER).
57
+
58
+ Accented English subset can also be used for research in ASR for accented speech (15 L2 accents)
59
+
60
+ ## Dataset Structure
61
+
62
+ ### Data Instances
63
+
64
+ ```python
65
+ {
66
+ 'audio_id': '20180206-0900-PLENARY-15-hr_20180206-16:10:06_5',
67
+ 'language': 11, # "hr"
68
+ 'audio': {
69
+ 'path': '/home/polina/.cache/huggingface/datasets/downloads/extracted/44aedc80bb053f67f957a5f68e23509e9b181cc9e30c8030f110daaedf9c510e/train_part_0/20180206-0900-PLENARY-15-hr_20180206-16:10:06_5.wav',
70
+ 'array': array([-0.01434326, -0.01055908, 0.00106812, ..., 0.00646973], dtype=float32),
71
+ 'sampling_rate': 16000
72
+ },
73
+ 'raw_text': '',
74
+ 'normalized_text': 'poast genitalnog sakaenja ena u europi tek je jedna od manifestacija takve tetne politike.',
75
+ 'gender': 'female',
76
+ 'speaker_id': '119431',
77
+ 'is_gold_transcript': True,
78
+ 'accent': 'None'
79
+ }
80
+ ```
81
+
82
+ ### Data Fields
83
+
84
+ * `audio_id` (string) - id of audio segment
85
+ * `audio` (datasets.Audio) - a dictionary containing the path to the audio, the decoded audio array, and the sampling rate. In non-streaming mode (default), the path points to the locally extracted audio. In streaming mode, the path is the relative path of an audio inside its archive (as files are not downloaded and extracted locally).
86
+ * `raw_text` (string) - original (orthographic) audio segment text
87
+ * `normalized_text` (string) - normalized audio segment transcription
88
+ * `gender` (string) - gender of speaker
89
+ * `speaker_id` (string) - id of speaker
90
+ * `is_gold_transcript` (bool) - ?
91
+ * `accent` (string) - type of accent, for example "en_lt", if applicable, else "None".
92
+
93
+ ### Data Splits
94
+
95
+ All configs contain data in three splits: train, validation and test.
96
+
97
+ ## Dataset Creation
98
+
99
+ ### Curation Rationale
100
+
101
+ [More Information Needed]
102
+
103
+ ### Source Data
104
+
105
+ The raw data is collected from [Գրքասեր site ](https://grqaser.org/am)
106
+
107
+ #### Initial Data Collection and Normalization
108
+
109
+ [More Information Needed]
110
+
111
+ #### Who are the source language producers?
112
+
113
+ [More Information Needed]
114
+
115
+ ### Annotations
116
+
117
+ #### Annotation process
118
+
119
+ [More Information Needed]
120
+
121
+ #### Who are the annotators?
122
+
123
+ [More Information Needed]
124
+
125
+ ### Personal and Sensitive Information
126
+
127
+ [More Information Needed]
128
+
129
+ ## Considerations for Using the Data
130
+
131
+ ### Social Impact of Dataset
132
+
133
+ [More Information Needed]
134
+
135
+ ### Discussion of Biases
136
+
137
+ [More Information Needed]
138
+
139
+ ### Other Known Limitations
140
+
141
+
142
+ ## Additional Information
143
+
144
+ ### Dataset Curators
145
+
146
+ [More Information Needed]
147
+
148
+ ### Licensing Information
149
+
150
+ [More Information Needed]
151
+
152
+ ### Citation Information
153
+
154
+ Please cite this paper:
155
+
156
+ ```bibtex
157
+ @inproceedings{wang-etal-2021-voxpopuli,
158
+ title = "{H}y{V}ox{P}opuli: A Large-Scale Multilingual Speech Corpus for Representation Learning, Semi-Supervised Learning and Interpretation",
159
+ author = "Edmon, Sahakyan",
160
+ booktitle = "Գրքասեր site audio and pdf books",
161
+ month = feb,
162
+ year = "2024",
163
+ address = "Online",
164
+ publisher = "Researcher"
165
+ }
166
+ ```
167
+
168
+ ### Contributions
169
+
170
+ Thanks to [@polinaeterna](https://github.com/polinaeterna) for adding this dataset.