Relative Transfer Matrix Estimator using Covariance Subtraction

TL;DR

This paper introduces a covariance subtraction method for estimating the Relative Transfer Matrix (ReTM) in multichannel sound source separation, demonstrating improved performance in noisy and reverberant environments.

Contribution

The paper presents a novel covariance subtraction approach for practical ReTM estimation applicable to multiple independent sound sources.

Findings

Effective separation at low SNR levels

Validated in simulated and real environments

Outperforms existing ReTM-based estimators

Abstract

The Relative Transfer Matrix (ReTM), recently introduced as a generalization of the relative transfer function for multiple receivers and sources, shows promising performance when applied to speech enhancement and speaker separation in noisy environments. Blindly estimating the ReTM of sound sources by exploiting the covariance matrices of multichannel recordings is highly beneficial for practical applications. In this paper, we use covariance subtraction to present a flexible and practically viable method for estimating the ReTM for a select set of independent sound sources. To show the versatility of the method, we validated it through a speaker separation application under reverberant conditions. Separation performance is evaluated at low signal-to-noise ratio levels in comparison with existing ReTM-based and relative transfer function-based estimators, in both simulated and real-life environments.