# Non-Autoregressive Neural Text-to-Speech

Source: [Non-Autoregressive Neural Text-to-Speech](<https://www.sota2.com/research/paper/68fa67d51e209131cea18cc1>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T22:24:01.852Z

Authors: Kainan Peng, Wei Ping, Zhao Song, Kexin Zhao

Published: 2019-05-21T06:36:15.000Z

## Abstract

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 and test 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.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Text-to-Speech on ParaNet 100 sentences (test)](<https://www.sota2.com/research/sota/text-to-speech-on-paranet-100-sentences-test.md>): Text-to-Speech; ParaNet 100 sentences (test); Repeat Errors: 1.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1905.08459>)

[PDF](<https://arxiv.org/pdf/1905.08459.pdf>)