File size: 6,396 Bytes
c2dba9f
1a14ea6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ec7d23d
c2dba9f
ec7d23d
 
 
 
 
 
 
 
 
 
c2dba9f
 
 
ec7d23d
 
f35eaf9
ec7d23d
c2dba9f
 
 
 
 
 
 
 
 
 
e5fdf63
c2dba9f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
af64bc3
c2dba9f
 
 
 
 
 
 
 
ee97500
dbe74d6
 
 
c2dba9f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2258b5a
 
4586d86
2258b5a
 
c2dba9f
 
 
 
 
 
 
e226ee2
bc629e5
e226ee2
c2dba9f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
61b93b4
c2dba9f
 
 
 
 
 
 
4586d86
c2dba9f
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
---
language: 
- multilingual
- ab
- af
- am
- ar
- as
- az
- ba
- be
- bg
- bi
- bo
- br
- bs
- ca
- ceb
- cs
- cy
- da
- de
- el
- en
- eo
- es
- et
- eu
- fa
- fi
- fo
- fr
- gl
- gn
- gu
- gv
- ha
- haw
- hi
- hr
- ht
- hu
- hy
- ia
- id
- is
- it
- he
- ja
- jv
- ka
- kk
- km 
- kn
- ko
- la
- lm
- ln
- lo
- lt
- lv
- mg
- mi
- mk
- ml
- mn
- mr
- ms
- mt
- my
- ne
- nl
- nn
- no 
- oc
- pa
- pl
- ps
- pt
- ro
- ru
- sa
- sco
- sd
- si
- sk
- sl
- sn
- so
- sq
- sr 
- su
- sv
- sw
- ta
- te
- tg
- th
- tk
- tl
- tr
- tt
- uk
- ud
- uz
- vi
- war
- yi
- yo
- zh
thumbnail:
tags:
- audio-classification
- speechbrain
- embeddings
- Language
- Identification
- pytorch
- ECAPA-TDNN
- TDNN
- VoxLingua107
license: "apache-2.0"
datasets:
- VoxLingua107
metrics:
- Accuracy
widget:
- example_title: English Sample
  src: https://cdn-media.huggingface.co/speech_samples/LibriSpeech_61-70968-0000.flac
---

# VoxLingua107 ECAPA-TDNN Spoken Language Identification Model

## Model description

This is a spoken language recognition model trained on the VoxLingua107 dataset using SpeechBrain.
The model uses the ECAPA-TDNN architecture that has previously been used for speaker recognition.

The model can classify a speech utterance according to the language spoken.
It covers 107 different languages (
Abkhazian, 
Afrikaans, 
Amharic, 
Arabic, 
Assamese, 
Azerbaijani, 
Bashkir, 
Belarusian, 
Bulgarian, 
Bengali, 
Tibetan, 
Breton, 
Bosnian, 
Catalan, 
Cebuano, 
Czech, 
Welsh, 
Danish, 
German, 
Greek, 
English, 
Esperanto, 
Spanish, 
Estonian, 
Basque, 
Persian, 
Finnish, 
Faroese, 
French, 
Galician, 
Guarani, 
Gujarati, 
Manx, 
Hausa, 
Hawaiian, 
Hindi, 
Croatian, 
Haitian, 
Hungarian, 
Armenian, 
Interlingua, 
Indonesian, 
Icelandic, 
Italian, 
Hebrew, 
Japanese, 
Javanese, 
Georgian, 
Kazakh, 
Central Khmer, 
Kannada, 
Korean, 
Latin, 
Luxembourgish, 
Lingala, 
Lao, 
Lithuanian, 
Latvian, 
Malagasy, 
Maori, 
Macedonian, 
Malayalam, 
Mongolian, 
Marathi, 
Malay, 
Maltese, 
Burmese, 
Nepali, 
Dutch, 
Norwegian Nynorsk, 
Norwegian, 
Occitan, 
Panjabi, 
Polish, 
Pushto, 
Portuguese, 
Romanian, 
Russian, 
Sanskrit, 
Scots, 
Sindhi, 
Sinhala, 
Slovak, 
Slovenian, 
Shona, 
Somali, 
Albanian, 
Serbian, 
Sundanese, 
Swedish, 
Swahili, 
Tamil, 
Telugu, 
Tajik, 
Thai, 
Turkmen, 
Tagalog, 
Turkish, 
Tatar, 
Ukrainian, 
Urdu, 
Uzbek, 
Vietnamese, 
Waray, 
Yiddish, 
Yoruba, 
Mandarin Chinese).

## Intended uses & limitations

The model has two uses:

  - use 'as is' for spoken language recognition
  - use as an utterance-level feature (embedding) extractor, for creating a dedicated language ID model on your own data
  
The model is trained on automatically collected YouTube data. For more 
information about the dataset, see [here](http://bark.phon.ioc.ee/voxlingua107/).


#### How to use

```python
import torchaudio
from speechbrain.pretrained import EncoderClassifier
language_id = EncoderClassifier.from_hparams(source="TalTechNLP/voxlingua107-epaca-tdnn", savedir="tmp")
# Download Thai language sample from Omniglot and cvert to suitable form
signal = language_id.load_audio("https://omniglot.com/soundfiles/udhr/udhr_th.mp3")
prediction =  language_id.classify_batch(signal)
print(prediction)
  (tensor([[0.3210, 0.3751, 0.3680, 0.3939, 0.4026, 0.3644, 0.3689, 0.3597, 0.3508,
           0.3666, 0.3895, 0.3978, 0.3848, 0.3957, 0.3949, 0.3586, 0.4360, 0.3997,
           0.4106, 0.3886, 0.4177, 0.3870, 0.3764, 0.3763, 0.3672, 0.4000, 0.4256,
           0.4091, 0.3563, 0.3695, 0.3320, 0.3838, 0.3850, 0.3867, 0.3878, 0.3944,
           0.3924, 0.4063, 0.3803, 0.3830, 0.2996, 0.4187, 0.3976, 0.3651, 0.3950,
           0.3744, 0.4295, 0.3807, 0.3613, 0.4710, 0.3530, 0.4156, 0.3651, 0.3777,
           0.3813, 0.6063, 0.3708, 0.3886, 0.3766, 0.4023, 0.3785, 0.3612, 0.4193,
           0.3720, 0.4406, 0.3243, 0.3866, 0.3866, 0.4104, 0.4294, 0.4175, 0.3364,
           0.3595, 0.3443, 0.3565, 0.3776, 0.3985, 0.3778, 0.2382, 0.4115, 0.4017,
           0.4070, 0.3266, 0.3648, 0.3888, 0.3907, 0.3755, 0.3631, 0.4460, 0.3464,
           0.3898, 0.3661, 0.3883, 0.3772, 0.9289, 0.3687, 0.4298, 0.4211, 0.3838,
           0.3521, 0.3515, 0.3465, 0.4772, 0.4043, 0.3844, 0.3973, 0.4343]]), tensor([0.9289]), tensor([94]), ['th'])
# The scores in the prediction[0] tensor can be interpreted as cosine scores between
# the languages and the given utterance (i.e., the larger the better)
# The identified language ISO code is given in prediction[3]
print(prediction[3])
  ['th']
  
# Alternatively, use the utterance embedding extractor:
emb =  language_id.encode_batch(signal)
print(emb.shape)
  torch.Size([1, 1, 256])
```

#### Limitations and bias

Since the model is trained on VoxLingua107, it has many limitations and biases, some of which are:

 - Probably it's accuracy on smaller languages  is quite limited
 - Probably it works worse on female speech than male speech (because YouTube data includes much more male speech)
 - Based on subjective experiments, it doesn't work well on speech with a foreign accent
 - Probably it doesn't work well on children's speech and on persons with speech disorders


## Training data

The model is trained on [VoxLingua107](http://bark.phon.ioc.ee/voxlingua107/).

VoxLingua107 is a speech dataset for training spoken language identification models. 
The dataset consists of short speech segments automatically extracted from YouTube videos and labeled according the language of the video title and description, with some post-processing steps to filter out false positives.

VoxLingua107 contains data for 107 languages. The total amount of speech in the training set is 6628 hours. 
The average amount of data per language is 62 hours. However, the real amount per language varies a lot. There is also a seperate development set containing 1609 speech segments from 33 languages, validated by at least two volunteers to really contain the given language.

## Training procedure

We used [SpeechBrain](https://github.com/speechbrain/speechbrain) to train the model.
Training recipe will be published soon.

## Evaluation results

Error rate: 7% on the development dataset


### BibTeX entry and citation info

```bibtex
@inproceedings{valk2021slt,
  title={{VoxLingua107}: a Dataset for Spoken Language Recognition},
  author={J{\"o}rgen Valk and Tanel Alum{\"a}e},
  booktitle={Proc. IEEE SLT Workshop},
  year={2021},
}
```