Update README.md

1. Add OTC paper link
2. Add OTC bibtex
3. Add pre-trained model link
This commit is contained in:
Dongji Gao 2023-09-28 15:06:15 -04:00 committed by GitHub
parent 8a3051bea7
commit 949d32e33a
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -1,7 +1,7 @@
# Introduction # Introduction
This is a weakly supervised ASR recipe for the LibriSpeech (clean 100 hours) dataset. We train a This is a weakly supervised ASR recipe for the LibriSpeech (clean 100 hours) dataset. We train a
conformer model using [Bypass Temporal Classification](https://arxiv.org/pdf/2306.01031.pdf) (BTC)/Omni-temporal Classification (OTC) with transcripts with synthetic errors. In this README, we will describe conformer model using [Bypass Temporal Classification](https://arxiv.org/pdf/2306.01031.pdf) (BTC)/[Omni-temporal Classification](https://arxiv.org/pdf/2309.15796.pdf) (OTC) with transcripts with synthetic errors. In this README, we will describe
the task and the BTC/OTC training process. the task and the BTC/OTC training process.
Note that OTC is an extension of BTC and supports all BTC functions. Therefore, in the following, we only describe OTC. Note that OTC is an extension of BTC and supports all BTC functions. Therefore, in the following, we only describe OTC.
@ -203,6 +203,9 @@ export CUDA_VISIBLE_DEVICES="0"
</tr> </tr>
</table> </table>
## Pre-trained Model
Pre-trained model: <https://huggingface.co/dgao/icefall-otc-librispeech-conformer-ctc2>
## Citations ## Citations
``` ```
@inproceedings{gao2023bypass, @inproceedings{gao2023bypass,
@ -211,4 +214,11 @@ export CUDA_VISIBLE_DEVICES="0"
booktitle={INTERSPEECH}, booktitle={INTERSPEECH},
year={2023} year={2023}
} }
@inproceedings{gao2023learning,
title={Learning from Flawed Data: Weakly Supervised Automatic Speech Recognition},
author={Gao, Dongji and Xu, Hainan and Raj, Desh and Garcia, Leibny Paola and Povey, Daniel and Khudanpur, Sanjeev},
booktitle={IEEE ASRU},
year={2023}
}
``` ```