Non-Autoregressive Neural Text-to-Speech

Kainan Peng · Wei Ping · Zhao Song · Kexin Zhao

Keywords: [ Speech Processing ] [ Applications - Language, Speech and Dialog ]

[ Abstract ]
Tue 14 Jul 9 a.m. PDT — 9:45 a.m. PDT
Tue 14 Jul 10 p.m. PDT — 10:45 p.m. PDT


In this work, we propose ParaNet, a non-autoregressive seq2seq model that converts text to spectrogram. It is fully convolutional and brings 46.7 times speed-up over the lightweight Deep Voice 3 at synthesis, while obtaining reasonably good speech quality. ParaNet also produces stable alignment between text and speech on the challenging test sentences by iteratively improving the attention in a layer-by-layer manner. Furthermore, we build the parallel text-to-speech system by applying various parallel neural vocoders, which can synthesize speech from text through a single feed-forward pass. We also explore a novel VAE-based approach to train the inverse autoregressive flow~(IAF) based parallel vocoder from scratch, which avoids the need for distillation from a separately trained WaveNet as previous work.

Chat is not available.