WaveGrad: Estimating Gradients for Waveform Generation

TL;DR

WaveGrad is a novel waveform generation model that estimates data density gradients, enabling high-quality audio synthesis with adjustable inference speed and bridging the gap between non-autoregressive and autoregressive models.

Contribution

It introduces WaveGrad, a gradient-based waveform generation method that allows flexible trade-offs between speed and quality, outperforming existing non-autoregressive models.

Findings

High fidelity audio generated with as few as six iterations

Outperforms adversarial non-autoregressive baselines

Matches autoregressive baseline with fewer sequential steps

Abstract

This paper introduces WaveGrad, a conditional model for waveform generation which estimates gradients of the data density. The model is built on prior work on score matching and diffusion probabilistic models. It starts from a Gaussian white noise signal and iteratively refines the signal via a gradient-based sampler conditioned on the mel-spectrogram. WaveGrad offers a natural way to trade inference speed for sample quality by adjusting the number of refinement steps, and bridges the gap between non-autoregressive and autoregressive models in terms of audio quality. We find that it can generate high fidelity audio samples using as few as six iterations. Experiments reveal WaveGrad to generate high fidelity audio, outperforming adversarial non-autoregressive baselines and matching a strong likelihood-based autoregressive baseline using fewer sequential operations. Audio samples are available at https://wavegrad.github.io/.