TSTNN: Two-Stage Transformer Based Neural Network for Speech Enhancement in the Time Domain.

Kai Wang, Bengbeng He, Wei-Ping Zhu

Browse the full ICASSP paper archive.