Spatial Processing Front-End For Distant ASR Exploiting Self-Attention Channel Combinator

TL;DR

This paper introduces a novel multi-channel front-end for distant ASR that combines channel shortening, self-attention-based channel combination, and beamforming, significantly improving recognition accuracy and dereverberation.

Contribution

It proposes a new combination of WPE, self-attention-based channel combination, and beamforming for distant ASR, demonstrating superior performance over existing systems.

Findings

21.6% relative WER reduction on LibriSpeech dataset

13.6 dB dereverberation improvement with WPE

SACC weights enable accurate spatial information extraction

Abstract

We present a novel multi-channel front-end based on channel shortening with theWeighted Prediction Error (WPE) method followed by a fixed MVDR beamformer used in combination with a recently proposed self-attention-based channel combination (SACC) scheme, for tackling the distant ASR problem. We show that the proposed system used as part of a ContextNet based end-to-end (E2E) ASR system outperforms leading ASR systems as demonstrated by a 21.6% reduction in relative WER on a multi-channel LibriSpeech playback dataset. We also show how dereverberation prior to beamforming is beneficial and compare the WPE method with a modified neural channel shortening approach. An analysis of the non-intrusive estimate of the signal C50 confirms that the 8 channel WPE method provides significant dereverberation of the signals (13.6 dB improvement). We also show how the weights of the SACC system allow the extraction of accurate spatial information which can be beneficial for other speech processing applications like diarization.