Diffusion-Based Speech Enhancement in Matched and Mismatched Conditions Using a Heun-Based Sampler

Philippe Gonzalez, Zheng Hua Tan, Jan Østergaard, Jesper Jensen, Tommy Sonne Alstrøm, Tobias May

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Diffusion models are a new class of generative models that have recently been applied to speech enhancement successfully. Previous works have demonstrated their superior performance in mismatched conditions compared to state-of-the art discriminative models. However, this was investigated with a single database for training and another one for testing, which makes the results highly dependent on the particular databases. Moreover, recent developments from the image generation literature remain largely unexplored for speech enhancement. These include several design aspects of diffusion models, such as the noise schedule or the reverse sampler. In this work, we systematically assess the generalization performance of a diffusion-based speech enhancement model by using multiple speech, noise and binaural room impulse response (BRIR) databases to simulate mismatched acoustic conditions. We also experiment with a noise schedule and a sampler that have not been applied to speech enhancement before. We show that the proposed system substantially benefits from using multiple databases for training, and achieves superior performance compared to state-of-the-art discriminative models in both matched and mismatched conditions. We also show that a Heun-based sampler achieves superior performance at a smaller computational cost compared to a sampler commonly used for speech enhancement.
Original languageEnglish
Title of host publicationProceedings of the 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
PublisherIEEE
Publication date2024
Pages10431-10435
ISBN (Print)979-8-3503-4486-8
ISBN (Electronic)979-8-3503-4485-1
DOIs
Publication statusPublished - 2024
Event2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) - Seoul, Korea, Republic of
Duration: 14 Apr 202419 Apr 2024

Conference

Conference2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Country/TerritoryKorea, Republic of
CitySeoul
Period14/04/202419/04/2024

Keywords

  • Speech enhancement
  • Diffusion models
  • Generalization

Fingerprint

Dive into the research topics of 'Diffusion-Based Speech Enhancement in Matched and Mismatched Conditions Using a Heun-Based Sampler'. Together they form a unique fingerprint.

Cite this