A Study on Speech Enhancement Based on Diffusion Probabilistic Model

Yen Ju Lu, Yu Tsao, Shinji Watanabe

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Diffusion probabilistic models have demonstrated an outstanding capability to model natural images and raw audio waveforms through a paired diffusion and reverse processes. The unique property of the reverse process (namely, eliminating non-target signals from the Gaussian noise and noisy signals) could be utilized to restore clean signals. Based on this prop-erty, we propose a diffusion probabilistic model-based speech enhancement (DiffuSE) model that aims to recover clean speech signals from noisy signals. The fundamental architecture of the proposed DiffuSE model is similar to that of DiffWave-a high-quality audio waveform generation model that has a relatively low computational cost and footprint. To attain better enhancement performance, we designed an advanced reverse process, termed the supportive reverse process, which adds noisy speech in each time-step to the predicted speech. The experimental results show that DiffuSE yields performance that is comparable to related audio generative models on the standardized Voice Bank corpus SE task. Moreover, relative to the generally suggested full sam-pling schedule, the proposed supportive reverse process especially improved the fast sampling, taking few steps to yield better enhancement results over the conventional full step inference process.

Original languageEnglish
Title of host publication2021 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2021 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages659-666
Number of pages8
ISBN (Electronic)9789881476890
Publication statusPublished - 2021
Externally publishedYes
Event2021 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2021 - Tokyo, Japan
Duration: 2021 Dec 142021 Dec 17

Publication series

Name2021 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2021 - Proceedings

Conference

Conference2021 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2021
Country/TerritoryJapan
CityTokyo
Period21/12/1421/12/17

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Signal Processing
  • Instrumentation

Fingerprint

Dive into the research topics of 'A Study on Speech Enhancement Based on Diffusion Probabilistic Model'. Together they form a unique fingerprint.

Cite this