Convolutive Transfer Function Invariant SDR training criteria for Multi-Channel Reverberant Speech Separation

TL;DR

This paper introduces a novel training criterion for multi-channel reverberant speech separation using a convolutive transfer function invariant SDR loss, significantly improving performance over traditional methods.

Contribution

It proposes the first use of CI-SDR as a training objective for neural network-based multi-channel reverberant speech separation.

Findings

Approaches single-source non-reverberant error rates

Outperforms permutation invariant training methods

Achieves large margin improvements over alternative objectives

Abstract

Time-domain training criteria have proven to be very effective for the separation of single-channel non-reverberant speech mixtures. Likewise, mask-based beamforming has shown impressive performance in multi-channel reverberant speech enhancement and source separation. Here, we propose to combine neural network supported multi-channel source separation with a time-domain training objective function. For the objective we propose to use a convolutive transfer function invariant Signal-to-Distortion Ratio (CI-SDR) based loss. While this is a well-known evaluation metric (BSS Eval), it has not been used as a training objective before. To show the effectiveness, we demonstrate the performance on LibriSpeech based reverberant mixtures. On this task, the proposed system approaches the error rate obtained on single-source non-reverberant input, i.e., LibriSpeech test_clean, with a difference of only 1.2 percentage points, thus outperforming a conventional permutation invariant training based system and alternative objectives like Scale Invariant Signal-to-Distortion Ratio by a large margin.