MushanW commited on
Commit
951f8cd
1 Parent(s): 9ab2761

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +80 -26
README.md CHANGED
@@ -1,26 +1,80 @@
1
- ---
2
- license: mit
3
- dataset_info:
4
- features:
5
- - name: audio
6
- dtype: audio
7
- - name: speaker_id
8
- dtype: string
9
- - name: transcript
10
- dtype: string
11
- - name: native_language
12
- dtype: string
13
- - name: subset
14
- dtype: string
15
- splits:
16
- - name: all
17
- num_bytes: 3179636720.056
18
- num_examples: 41806
19
- download_size: 3667597693
20
- dataset_size: 3179636720.056
21
- configs:
22
- - config_name: default
23
- data_files:
24
- - split: all
25
- path: data/all-*
26
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: cc0-1.0
5
+ task_categories:
6
+ - text-to-audio
7
+ - automatic-speech-recognition
8
+ - audio-to-audio
9
+ - audio-classification
10
+ dataset_info:
11
+ features:
12
+ - name: audio
13
+ dtype: audio
14
+ - name: speaker_id
15
+ dtype: string
16
+ - name: transcript
17
+ dtype: string
18
+ - name: native_language
19
+ dtype: string
20
+ - name: subset
21
+ dtype: string
22
+ splits:
23
+ - name: all
24
+ num_bytes: 3179636720.056
25
+ num_examples: 41806
26
+ download_size: 3667597693
27
+ dataset_size: 3179636720.056
28
+ configs:
29
+ - config_name: default
30
+ data_files:
31
+ - split: all
32
+ path: data/all-*
33
+ ---
34
+
35
+ # ESLTTS
36
+
37
+ The full paper can be accessed here: [arXiv](https://arxiv.org/abs/2404.18094), [IEEE Xplore](https://ieeexplore.ieee.org/document/10508477).
38
+
39
+ ## Dataset Access
40
+ You can access this dataset through [Huggingface](https://huggingface.co/datasets/MushanW/ESLTTS) or [Google Driver](https://drive.google.com/file/d/1ChQ_z-TxvKWNUbUMWnbyjM2VY3v2SKEi/view?usp=sharing) or [IEEE Dataport](http://ieee-dataport.org/documents/english-second-language-tts-esltts-dataset).
41
+
42
+ ## Abstract
43
+
44
+ With the progress made in speaker-adaptive TTS approaches, advanced approaches have shown a remarkable capacity to reproduce the speaker’s voice in the commonly used TTS datasets. However, mimicking voices characterized by substantial accents, such as non-native English speakers, is still challenging. Regrettably, the absence of a dedicated TTS dataset for speakers with substantial accents inhibits the research and evaluation of speaker-adaptive TTS models under such conditions. To address this gap, we developed a corpus of non-native speakers' English utterances.
45
+
46
+ We named this corpus “English as a Second Language TTS dataset ” (ESLTTS). The ESLTTS dataset consists of roughly 37 hours of 42,000 utterances from 134 non-native English speakers. These speakers represent a diversity of linguistic backgrounds spanning 31 native languages. For each speaker, the dataset includes an adaptation set lasting about 5 minutes for speaker adaptation, a test set comprising 10 utterances for speaker-adaptive TTS evaluation, and a development set for further research.
47
+
48
+ ## Dataset Structure
49
+
50
+ ```
51
+ ESLTTS Dataset/
52
+ ├─ Malayalam_3/ ------------ {Speaker Native Language}_{Speaker id}
53
+ │ ├─ ada_1.flac ------------ {Subset Name}_{Utterance id}
54
+ │ ├─ ada_1.txt ------------ Transcription for "ada_1.flac"
55
+ │ ├─ test_1.flac ------------ {Subset Name}_{Utterance id}
56
+ │ ├─ test_1.txt ------------ Transcription for "test_1.flac"
57
+ │ ├─ dev_1.flac ------------ {Subset Name}_{Utterance id}
58
+ │ ├─ dev_1.txt ------------ Transcription for "dev_1.flac"
59
+ │ ├─ ...
60
+ ├─ Arabic_3/ ------------ {Speaker Native Language}_{Speaker id}
61
+ │ ├─ ada_1.flac ------------ {Subset Name}_{Utterance id}
62
+ │ ├─ ...
63
+ ├─ ...
64
+ ```
65
+
66
+ ## Citation
67
+ ```
68
+ @article{wang2024usat,
69
+ author = {Wenbin Wang and
70
+ Yang Song and
71
+ Sanjay K. Jha},
72
+ title = {{USAT:} {A} Universal Speaker-Adaptive Text-to-Speech Approach},
73
+ journal = {{IEEE} {ACM} Trans. Audio Speech Lang. Process.},
74
+ volume = {32},
75
+ pages = {2590--2604},
76
+ year = {2024},
77
+ url = {https://doi.org/10.1109/TASLP.2024.3393714},
78
+ doi = {10.1109/TASLP.2024.3393714},
79
+ }
80
+ ```