A Conformer-based Waveform-domain Neural Acoustic Echo Canceller Optimized for ASR Accuracy

TL;DR

This paper introduces a conformer-based waveform-domain neural acoustic echo canceller optimized for ASR accuracy, demonstrating significant WER reduction and efficiency improvements over spectral-domain models.

Contribution

It presents a novel waveform-domain neural AEC model based on conformer architecture, trained jointly with ASR loss, outperforming spectral-domain models in accuracy and size.

Findings

56-59% WER reduction over linear AEC

20-29% improvement over spectral neural AEC

Effective in resource-limited applications

Abstract

Acoustic Echo Cancellation (AEC) is essential for accurate recognition of queries spoken to a smart speaker that is playing out audio. Previous work has shown that a neural AEC model operating on log-mel spectral features (denoted "logmel" hereafter) can greatly improve Automatic Speech Recognition (ASR) accuracy when optimized with an auxiliary loss utilizing a pre-trained ASR model encoder. In this paper, we develop a conformer-based waveform-domain neural AEC model inspired by the "TasNet" architecture. The model is trained by jointly optimizing Negative Scale-Invariant SNR (SISNR) and ASR losses on a large speech dataset. On a realistic rerecorded test set, we find that cascading a linear adaptive AEC and a waveform-domain neural AEC is very effective, giving 56-59% word error rate (WER) reduction over the linear AEC alone. On this test set, the 1.6M parameter waveform-domain neural AEC also improves over a larger 6.5M parameter logmel-domain neural AEC model by 20-29% in easy to moderate conditions. By operating on smaller frames, the waveform neural model is able to perform better at smaller sizes and is better suited for applications where memory is limited.