Abstract:We propose a nonlinear acoustic echo cancellation system, which aims to model the echo path from the far-end signal to the near-end microphone in two parts. Inspired by the physical behavior of modern hands-free devices, we first introduce a novel neural network architecture that is specifically designed to model the nonlinear distortions these devices induce between receiving and playing the far-end signal. To account for variations between devices, we construct this network with trainable memory length and nonlinear activation functions that are not parameterized in advance, but are rather optimized during the training stage using the training data. Second, the network is succeeded by a standard adaptive linear filter that constantly tracks the echo path between the loudspeaker output and the microphone. During training, the network and filter are jointly optimized to learn the network parameters. This system requires 17 thousand parameters that consume 500 Million floating-point operations per second and 40 Kilo-bytes of memory. It also satisfies hands-free communication timing requirements on a standard neural processor, which renders it adequate for embedding on hands-free communication devices. Using 280 hours of real and synthetic data, experiments show advantageous performance compared to competing methods.

Acoustic Echo Cancellation Based on Nonlinear Pre-processing

Nonlinear Acoustic Echo Cancellation with Deep Learning

Adaptive Sub-Band Acoustic Echo Cancellation Algorithms for Echo Paths Involving Nonlinear Amplifiers and Transducers

Improved Acoustic Echo Cancellation Algorithm

A Residual Echo Suppression Technique for Systems with Nonlinear Acoustic Echo Paths

An efficient acoustic echo cancellation design for systems with long room impulses and nonlinear loudspeakers

An Acoustic Echo Canceller Combined with VoIP

A Channel Shortening Approach for Nonlinear Acoustic Echo Cancellation

Dynamic Neural Network-based Solutions for Acoustic Echo Suppression

Simulation and Research on Adaptive Echo Cancellation Algorithm

Acoustic Echo Cancellation Using A Pseudocoherence Function in the Presence of Memoryless Nonlinearity

A Robust and Cascaded Acoustic Echo Cancellation Based on Deep Learning

A Nonparametric Variable Step-Size Subband Adaptive Filtering Algorithm for Acoustic Echo Cancellation

Study on echo cancellation based on NLMS improved algorithm

Adaptive Acoustic Echo Cancellation in the Presence of Multiple Nonlinearities

An Improved Nlms Algorithm Based on Speech Enhancement

A Two-stage Progressive Neural Network for Acoustic Echo Cancellation.

Adaptive Acoustic Echo Cancellation System Without Double-Talk Detection Using VSS-NLMS Algorithm

A Double-talk Robust and Fast Converging Echo Canceller

A fast converging algorithm for network echo cancellation and its DSP implementation

Sub-band Implementation of Adaptive Nonlinear Filter for Adaptive Nonlinear Echo Cancellation