speech and noise dual-stream spectrogram refine network with speech distortion loss for robust speech recognition
There will be a delay. Please wait one minute. The simulated data link is: https://tjueducn-my.sharepoint.com/:u:/g/personal/luhaoyu_831_tju_edu_cn/EUZwebz6WANDllOtd_9DI20Bg6Pr98aBA0C_Ni68yY9P-g?e=MChca1
The noise link is: http://web.cse.ohio-state.edu/pnl/corpus/HuNonspeech/
Sorry, there's been a slight error in the equation (9) in our paper, and the link to the latest version of the paper is: https://arxiv.org/abs/2305.17860