TianYuan
|
ec571bb0d1
|
Update README.md
|
2 years ago |
Hui Zhang
|
e0081b7e50
|
[vec][spk] add speechbrain ecapa-tdnn result
|
2 years ago |
KP
|
19fd46f57b
|
Fix rir download. test=doc
|
3 years ago |
|
TianYuan
|
e233c4e449
|
Merge pull request #2079 from Honei/develop
[vector]add convert.sh
|
3 years ago |
xiongxinlei
|
d15883e3dc
|
add convert.sh
|
3 years ago |
|
KP
|
bf056c013d
|
Refactor paddleaudio to paddlespeech.audio
|
3 years ago |
|
xiongxinlei
|
07c0d7d7cc
|
remove old vector model info, test=doc
|
3 years ago |
|
xiongxinlei
|
7afbdbefad
|
update the vector model, test=doc
|
3 years ago |
|
xiongxinlei
|
c5fe181405
|
update the paddlespeech_client asr_online cli, test=doc
|
3 years ago |
|
xiongxinlei
|
4c56e4d42c
|
update the voxceleb readme.md, test=doc
|
3 years ago |
|
xiongxinlei
|
cb9beabace
|
fix the sv ecapa-tdnn cpu training, test=doc
|
3 years ago |
|
Hui Zhang
|
91e24b0480
|
format code
|
3 years ago |
|
xiongxinlei
|
df503e97c1
|
update the voxceleb readme.md, test=doc
|
3 years ago |
|
xiongxinlei
|
d1935d8552
|
add vector necessary note, test=doc
|
3 years ago |
Honei
|
48e0177767
|
Merge pull request #1630 from Honei/vox12
[vec]voxceleb convert dataset format to paddlespeech
|
3 years ago |
|
xiongxinlei
|
a8244dc5b0
|
update the note, test=doc
|
3 years ago |
|
xiongxinlei
|
d064c8196e
|
update the speaker verification model, test=doc
|
3 years ago |
|
xiongxinlei
|
38e4e9c893
|
refactor voxceleb2 data download, test=doc
|
3 years ago |
|
xiongxinlei
|
acebfad7b7
|
change the vector csv.spk_id to csv.label, test=doc
|
3 years ago |
|
xiongxinlei
|
57c11dcab0
|
add some annotations, test=doc
|
3 years ago |
|
xiongxinlei
|
30b5b3cb9e
|
add vector csv dataset format, test=doc
|
3 years ago |
|
xiongxinlei
|
5b05300e53
|
train process add new voxceleb and rirs dataset, test=doc
|
3 years ago |
|
xiongxinlei
|
9944fec3d4
|
convert rirs noise to csv file
|
3 years ago |
|
xiongxinlei
|
ec24a169ee
|
convert jsonfile to csv file
|
3 years ago |
|
xiongxinlei
|
ed7113f320
|
change the vector output to numpy.array
|
3 years ago |
|
xiongxinlei
|
ef1bc5e815
|
vector cli output dim info, test=doc
|
3 years ago |
|
xiongxinlei
|
1fdb36f757
|
add mode emb dim info, test=doc
|
3 years ago |
|
xiongxinlei
|
ad2caf2ccb
|
add speaker verification demo and doc, test=doc
|
3 years ago |
|
xiongxinlei
|
e2684e71f2
|
refactor the data prepare process
|
3 years ago |
|
xiongxinlei
|
5221c2797f
|
add voxceleb dataset and trial info, test=doc
|
3 years ago |
|
xiongxinlei
|
d28ccfa96b
|
add vector cli component, test=doc
|
3 years ago |
|
xiongxinlei
|
506d26a957
|
change the code style to s2t code style, test=doc
|
3 years ago |
|
xiongxinlei
|
7eb8fa72a1
|
convert save_freq to save_interval, test=doc
|
3 years ago |
|
xiongxinlei
|
311fa87a11
|
add some comments to the code
|
3 years ago |
|
xiongxinlei
|
8ed5c287a3
|
add vox2 data into VoxCeleb class
|
3 years ago |
|
xiongxinlei
|
584a2c0e39
|
add ecapa-tdnn config yaml file
|
3 years ago |
|
xiongxinlei
|
993d6783d7
|
remove unused code, test=doc
|
3 years ago |
|
xiongxinlei
|
0e87037f2c
|
refactor to compilance paddleaudio
|
3 years ago |
|
xiongxinlei
|
4473405f82
|
merge develop to vox12, test=doc
|
3 years ago |
|
xiongxinlei
|
60d73bb7bd
|
add state 0 to prepare the voxcele data and augment data
|
3 years ago |
|
xiongxinlei
|
14efbf5b15
|
check extract embedding result, test=doc
|
3 years ago |
|
xiongxinlei
|
7db7eb8993
|
add extract audio embedding api, test=doc
|
3 years ago |
|
xiongxinlei
|
2d89c80e6f
|
add waveform augment pipeline, test=doc
|
3 years ago |
|
xiongxinlei
|
016ed6d69c
|
repair the code according to the part comment, test=doc
|
3 years ago |
|
xiongxinlei
|
97ec01260b
|
add speaker verification using cosine score, test=doc
|
3 years ago |
|
xiongxinlei
|
1f74af110b
|
add training log info and comment, test=doc
|
3 years ago |
|
xiongxinlei
|
4648059b5f
|
add training process for sid, test=doc
|
3 years ago |
|
xiongxinlei
|
7668f61422
|
add sid dataloader for training, test=doc
|
3 years ago |
|
xiongxinlei
|
6af2bc3d5b
|
add sid loss wraper for voxceleb, test=doc
|
3 years ago |
|
xiongxinlei
|
57c4f4a68c
|
add sid learning rate and training model
|
3 years ago |
