Transformer-Based Text-to-Speech with Weighted Forced Attention.

Takuma Okamoto, Tomoki Toda, Yoshinori Shiga, Hisashi Kawai

Browse the full ICASSP paper archive.