RemixIT: Continual self-training of speech enhancement models via bootstrapped remixing

TL;DR

RemixIT is a self-supervised speech enhancement method that uses bootstrapped remixing and iterative self-training to improve performance without relying on clean in-domain signals, effectively handling domain mismatch.

Contribution

It introduces a novel self-training scheme with remixing for speech enhancement that does not require clean target signals and can adapt across domains.

Findings

Outperforms prior speech enhancement methods in various datasets

Compatible with any separation model and domain adaptation tasks

Student models improve despite degraded pseudo-targets

Abstract

We present RemixIT, a simple yet effective self-supervised method for training speech enhancement without the need of a single isolated in-domain speech nor a noise waveform. Our approach overcomes limitations of previous methods which make them dependent on clean in-domain target signals and thus, sensitive to any domain mismatch between train and test samples. RemixIT is based on a continuous self-training scheme in which a pre-trained teacher model on out-of-domain data infers estimated pseudo-target signals for in-domain mixtures. Then, by permuting the estimated clean and noise signals and remixing them together, we generate a new set of bootstrapped mixtures and corresponding pseudo-targets which are used to train the student network. Vice-versa, the teacher periodically refines its estimates using the updated parameters of the latest student models. Experimental results on multiple speech enhancement datasets and tasks not only show the superiority of our method over prior approaches but also showcase that RemixIT can be combined with any separation model as well as be applied towards any semi-supervised and unsupervised domain adaptation task. Our analysis, paired with empirical evidence, sheds light on the inside functioning of our self-training scheme wherein the student model keeps obtaining better performance while observing severely degraded pseudo-targets.