Convolutive Transfer Function Invariant SDR Training Criteria for Multi-Channel Reverberant Speech Separation
Christoph Böddeker, Wangyou Zhang, Tomohiro Nakatani, Keisuke Kinoshita, Tsubasa Ochiai, Marc Delcroix, Naoyuki Kamo, Yanmin Qian
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.
BibTeX
@inproceedings{icassp2021_convolutivetrans,
title = {Convolutive Transfer Function Invariant SDR Training Criteria for Multi-Channel Reverberant Speech Separation},
author = {Christoph Böddeker and Wangyou Zhang and Tomohiro Nakatani and Keisuke Kinoshita and Tsubasa Ochiai and Marc Delcroix and Naoyuki Kamo and Yanmin Qian and Reinhold Haeb-Umbach},
booktitle = {ICASSP 2021},
year = {2021}
}