Waveglow: A Flow-based Generative Network for Speech Synthesis

TLDR

The paper proposes WaveGlow, a flow‑based network that generates high‑quality speech directly from mel‑spectrograms. WaveGlow merges Glow and WaveNet ideas into a single network trained with a single likelihood‑maximizing cost, enabling fast, efficient, non‑autoregressive audio synthesis. In PyTorch, WaveGlow produces audio at over 500 kHz on an NVIDIA V100 and achieves Mean Opinion Scores comparable to the best publicly available WaveNet. All code will be made publicly available online.

Abstract

In this paper we propose WaveGlow: a flow-based network capable of generating high quality speech from mel-spectrograms. WaveGlow combines insights from Glow [1] and WaveNet [2] in order to provide fast, efficient and high-quality audio synthesis, without the need for auto-regression. WaveGlow is implemented using only a single network, trained using only a single cost function: maximizing the likelihood of the training data, which makes the training procedure simple and stable. Our PyTorch implementation produces audio samples at a rate of more than 500 kHz on an NVIDIA V100 GPU. Mean Opinion Scores show that it delivers audio quality as good as the best publicly available WaveNet implementation. All code will be made publicly available online [3].