mirrors/icefall
1
0
Fork 0
mirror of https://github.com/k2-fsa/icefall.git synced 2025-09-10 17:44:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Dongji Gao 2023-09-23 21:28:12 -04:00 committed by GitHub
parent 53623d713e
commit 1ea86de1da
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
egs/librispeech/WSASR/README.md
View File

@ -67,7 +67,7 @@ otc_token="<star>"
``` ```
This script adds the 'otc_token' ('\<star\>') and its corresponding sentence-piece ('▁\<star\>') to 'words.txt' and 'tokens.txt,' respectively. Additionally, it computes SSL features using the 'wav2vec2-base' model. (You can use GPU to accelerate feature extraction). This script adds the 'otc_token' ('\<star\>') and its corresponding sentence-piece ('▁\<star\>') to 'words.txt' and 'tokens.txt,' respectively. Additionally, it computes SSL features using the 'wav2vec2-base' model. (You can use GPU to accelerate feature extraction).
### Making synthetic errors to the transcript [optional] ### Making synthetic errors to the transcript (train-clean-100) [optional]
``` ```
sub_er=0.17 sub_er=0.17
ins_er=0.17 ins_er=0.17
@ -94,6 +94,7 @@ verbatim: -have- a [] [nice] day
``` ```
### Training ### Training
The training uses synthetic data based on the train-clean-100 subset.
``` ```
otc_lang_dir=data/lang_bpe_200 otc_lang_dir=data/lang_bpe_200
@ -195,10 +196,10 @@ export CUDA_VISIBLE_DEVICES="0"
## Citations ## Citations
``` ```
@article{gao2023bypass, @inproceedings{gao2023bypass,
title={Bypass Temporal Classification: Weakly Supervised Automatic Speech Recognition with Imperfect Transcripts}, title={Bypass Temporal Classification: Weakly Supervised Automatic Speech Recognition with Imperfect Transcripts},
author={Gao, Dongji and Wiesner, Matthew and Xu, Hainan and Garcia, Leibny Paola and Povey, Daniel and Khudanpur, Sanjeev}, author={Gao, Dongji and Wiesner, Matthew and Xu, Hainan and Garcia, Leibny Paola and Povey, Daniel and Khudanpur, Sanjeev},
journal={arXiv preprint arXiv:2306.01031}, booktitle={INTERSPEECH},
year={2023} year={2023}
} }
``` ```