Low-Complexity, Real-Time Joint Neural Echo Control and Speech Enhancement Based On PercepNet

TL;DR

This paper introduces a low-complexity, real-time joint neural system for echo cancellation and speech enhancement that outperforms traditional methods while using minimal computational resources.

Contribution

It presents a hybrid DSP/DNN approach combining traditional echo cancellation with a neural residual suppressor, achieving high performance with very low CPU usage.

Findings

Outperforms traditional and neural methods in echo suppression

Operates in real-time using only 5.5% CPU

Can be scaled to lower complexity levels

Abstract

Speech enhancement algorithms based on deep learning have greatly surpassed their traditional counterparts and are now being considered for the task of removing acoustic echo from hands-free communication systems. This is a challenging problem due to both real-world constraints like loudspeaker non-linearities, and to limited compute capabilities in some communication systems. In this work, we propose a system combining a traditional acoustic echo canceller, and a low-complexity joint residual echo and noise suppressor based on a hybrid signal processing/deep neural network (DSP/DNN) approach. We show that the proposed system outperforms both traditional and other neural approaches, while requiring only 5.5% CPU for real-time operation. We further show that the system can scale to even lower complexity levels.