TalkNet: Non-Autoregressive Depth-Wise Separable Convolutional Model for Speech Synthesis.

Stanislav Beliaev, Boris Ginsburg

Browse the full Interspeech paper archive.