Thunder : Unified Regression-diffusion Speech Enhancement With A Single Reverse Step Using Brownian Bridge

Abstract

Diffusion-based speech enhancement has shown promising results, but can suffer from a slower inference time. Initializing the diffusion process with the enhanced audio generated by a regression-based model can be used to reduce the computational steps required. However, these approaches often necessitate a regression model, further increasing the system's complexity. We propose Thunder, a unified regression-diffusion model that utilizes the Brownian bridge process which can allow the model to act in both modes. The regression mode can be accessed by setting the diffusion time step closed to 1. However, the standard score-based diffusion modeling does not perform well in this setup due to gradient instability. To mitigate this problem, we modify the diffusion model to predict the clean speech instead of the score function, achieving competitive performance with a more compact model size and fewer reverse steps.

Stats

• citations6
• S2 citations—
• github stars0
• HF likes0
• heat score6.34
• arxiv keytrachu2024thunder

Related papers

• Storm: A Diffusion-based Stochastic Regeneration Model For Speech Enhancement And Dereverberation (2022)15.43
• Single And Few-step Diffusion For Generative Speech Enhancement (2023)10.21
• Cold Diffusion For Speech Enhancement (2022)11.85
• Speech Enhancement And Dereverberation With Diffusion-based Generative Models (2022)23.51
• Diffusion-based Speech Enhancement With Schr\"odinger Bridge And Symmetric Noise Schedule (2024)0.00
• Investigating The Design Space Of Diffusion Models For Speech Enhancement (2023)10.07
• Undiff: Unsupervised Voice Restoration With Unconditional Diffusion Model (2023)5.24
• DDTSE: Discriminative Diffusion Model For Target Speech Extraction (2023)5.84