Speech Dereverberation in the STFT Domain

TL;DR

This paper introduces a new single-channel speech dereverberation method in the STFT domain that uses a linear filter based on the channel impulse response to improve speech quality and intelligibility.

Contribution

A novel dereverberation technique in the STFT domain utilizing a linear filter derived from the channel impulse response, enhancing speech clarity.

Findings

Effective dereverberation with known impulse response

Improves speech quality and intelligibility

Resolves reverberant signals successfully

Abstract

Reverberation is damaging to both the quality and the intelligibility of a speech signal. We propose a novel single-channel method of dereverberation based on a linear filter in the Short Time Fourier Transform domain. Each enhanced frame is constructed from a linear sum of nearby frames based on the channel impulse response. The results show that the method can resolve any reverberant signal with knowledge of the impulse response to a non-reverberant signal.