Consistency-aware multi-channel speech enhancement using deep neural networks

TL;DR

This paper introduces a consistency-aware multi-channel speech enhancement system using deep neural networks that optimizes the quality of the reconstructed time-domain signal, addressing issues with spectrogram inconsistency.

Contribution

It proposes a novel objective function based on reconstructed time-domain signals for training DNN-based multi-channel speech enhancement systems.

Findings

Improved speech quality over traditional T-F masking methods

Effective enhancement with reconstructed time-domain objective functions

Demonstrated superiority in experimental comparisons

Abstract

This paper proposes a deep neural network (DNN)-based multi-channel speech enhancement system in which a DNN is trained to maximize the quality of the enhanced time-domain signal. DNN-based multi-channel speech enhancement is often conducted in the time-frequency (T-F) domain because spatial filtering can be efficiently implemented in the T-F domain. In such a case, ordinary objective functions are computed on the estimated T-F mask or spectrogram. However, the estimated spectrogram is often inconsistent, and its amplitude and phase may change when the spectrogram is converted back to the time-domain. That is, the objective function does not evaluate the enhanced time-domain signal properly. To address this problem, we propose to use an objective function defined on the reconstructed time-domain signal. Specifically, speech enhancement is conducted by multi-channel Wiener filtering in the T-F domain, and its result is converted back to the time-domain. We propose two objective functions computed on the reconstructed signal where the first one is defined in the time-domain, and the other one is defined in the T-F domain. Our experiment demonstrates the effectiveness of the proposed system comparing to T-F masking and mask-based beamforming.