pith. machine review for the scientific record. sign in

A wavenet for speech denoising

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

eess.AS 1

years

2020 1

verdicts

UNVERDICTED 1

representative citing papers

DiffWave: A Versatile Diffusion Model for Audio Synthesis

eess.AS · 2020-09-21 · unverdicted · novelty 8.0

DiffWave is a non-autoregressive diffusion model that generates high-fidelity audio waveforms from noise in constant steps, matching WaveNet vocoder quality while being orders of magnitude faster and outperforming prior models in unconditional generation.

citing papers explorer

Showing 1 of 1 citing paper.

  • DiffWave: A Versatile Diffusion Model for Audio Synthesis eess.AS · 2020-09-21 · unverdicted · none · ref 11

    DiffWave is a non-autoregressive diffusion model that generates high-fidelity audio waveforms from noise in constant steps, matching WaveNet vocoder quality while being orders of magnitude faster and outperforming prior models in unconditional generation.