Speech Dereverberation with a Reverberation Time Shortening Target

TL;DR

This paper introduces a reverberation time shortening (RTS) target for speech dereverberation, improving training stability and performance by better modeling reverberation decay, and adapts the FullSubNet denoising network for this task.

Contribution

It proposes a novel RTS learning target that preserves reverberation decay properties, enhancing dereverberation training and performance, and demonstrates the effectiveness of adapting FullSubNet for dereverberation.

Findings

RTS outperforms direct-path and early reflection targets in suppressing reverberation.

FullSubNet achieves outstanding dereverberation performance with RTS.

RTS reduces signal distortion caused by prediction errors.

Abstract

This work proposes a new learning target based on reverberation time shortening (RTS) for speech dereverberation. The learning target for dereverberation is usually set as the direct-path speech or optionally with some early reflections. This type of target suddenly truncates the reverberation, and thus it may not be suitable for network training. The proposed RTS target suppresses reverberation and meanwhile maintains the exponential decaying property of reverberation, which will ease the network training, and thus reduce signal distortion caused by the prediction error. Moreover, this work experimentally study to adapt our previously proposed FullSubNet speech denoising network to speech dereverberation. Experiments show that RTS is a more suitable learning target than direct-path speech and early reflections, in terms of better suppressing reverberation and signal distortion. FullSubNet is able to achieve outstanding dereverberation performance.