Exploring the Interactions between Target Positive and Negative Information for Acoustic Echo Cancellation

TL;DR

This paper introduces CMNet, a novel acoustic echo cancellation model that leverages both positive and negative target information through a collaboration module to improve discrimination between speech and interference signals.

Contribution

The paper proposes a new encoder-decoder architecture with a collaboration module that models interactions between positive and negative information for better echo cancellation.

Findings

CMNet outperforms recent methods in experiments.

Incorporating negative information improves discrimination.

The collaboration module effectively models positive-negative interactions.

Abstract

Acoustic echo cancellation (AEC) aims to remove interference signals while leaving near-end speech least distorted. As the indistinguishable patterns between near-end speech and interference signals, near-end speech can't be separated completely, causing speech distortion and interference signals residual. We observe that besides target positive information, e.g., ground-truth speech and features, the target negative information, such as interference signals and features, helps make pattern of target speech and interference signals more discriminative. Therefore, we present a novel AEC model encoder-decoder architecture with the guidance of negative information termed as CMNet. A collaboration module (CM) is designed to establish the correlation between the target positive and negative information in a learnable manner via three blocks: target positive, target negative, and interactive block. Experimental results demonstrate our CMNet achieves superior performance than recent methods.