Continual self-training with bootstrapped remixing for speech enhancement

Abstract

We propose RemixIT, a simple and novel self-supervised training method for speech enhancement. The proposed method is based on a continuously self-training scheme that overcomes limitations from previous studies including assumptions for the in-domain noise distribution and having access to clean target signals. Specifically, a separation teacher model is pre-trained on an out-of-domain dataset and is used to infer estimated target signals for a batch of in-domain mixtures. Next, we bootstrap the mixing process by generating artificial mixtures using permuted estimated clean and noise signals. Finally, the student model is trained using the permuted estimated sources as targets while we periodically update teacher's weights using the latest student model. Our experiments show that RemixIT outperforms several previous state-of-the-art self-supervised methods under multiple speech enhancement tasks. Additionally, RemixIT provides a seamless alternative for semi-supervised and unsupervised domain adaptation for speech enhancement tasks, while being general enough to be applied to any separation task and paired with any separation model.

To appear in Proc. ICASSP 2022, May 22-27, 2022, Singapore

References in corpus (4)

• Training Speech Enhancement Systems with Noisy Speech Datasets
• Momentum Pseudo-Labeling for Semi-Supervised Speech Recognition
• Teacher-Student MixIT for Unsupervised and Semi-supervised Speech Separation
• Incorporating Real-world Noisy Speech in Neural-network-based Speech Enhancement Systems

Cited by in corpus (5)

• Objective and subjective evaluation of speech enhancement methods in the UDASE task of the 7th CHiME challenge
• The CHiME-7 UDASE task: Unsupervised domain adaptation for conversational speech enhancement
• MixCycle: Unsupervised Speech Separation via Cyclic Mixture Permutation Invariant Training
• On monoaural speech enhancement for automatic recognition of real noisy speech using mixture invariant training
• Reinforced Domain Selection for Continuous Domain Adaptation