Boosting Cross-Domain Speech Recognition with Self-Supervision

This repository is the official implementation of the TASLP 2023 paper: Boosting Cross-Domain Speech Recognition with Self-Supervision

Example

• We provide an example with the SwitchBoard dataset in run.sh.
• The implementation is based on fairseq. Our modification is here.

Citation

@article{zhu2023boosting,
  title={Boosting Cross-Domain Speech Recognition with Self-Supervision},
  author={Zhu, Han and Cheng, Gaofeng and Wang, Jindong and How, Wenxin and Zhang, Pengyuan and Yan, Yonghong},
  journal={IEEE/ACM Transactions on Audio, Speech, and Language Processing},
  year={2023},
  publisher={IEEE}
}

Contact

• zhuhan@hccl.ioa.ac.cn