Thunder : Unified Regression-Diffusion Speech Enhancement with a Single Reverse Step using Brownian Bridge

TL;DR

Thunder introduces a unified regression-diffusion speech enhancement model using Brownian bridge, enabling efficient inference with fewer steps by predicting clean speech directly, thus reducing complexity and improving performance.

Contribution

The paper presents a novel unified model that combines regression and diffusion modes with a single reverse step, simplifying speech enhancement and improving efficiency.

Findings

Competitive speech enhancement performance with fewer reverse steps

Reduced model complexity and size compared to traditional diffusion models

Effective mitigation of gradient instability by predicting clean speech directly

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.