End-to-End Text-to-Speech Using Latent Duration Based on VQ-VAE.

Yusuke Yasuda, Xin Wang, Junichi Yamagishi

Browse the full ICASSP paper archive.