SOTAVerified

Stabilizing Label Assignment for Speech Separation by Self-supervised Pre-training

2020-10-29Code Available1· sign in to hype

Sung-Feng Huang, Shun-Po Chuang, Da-Rong Liu, Yi-Chen Chen, Gene-Ping Yang, Hung-Yi Lee

Code Available — Be the first to reproduce this paper.

Reproduce

Code

Abstract

Speech separation has been well developed, with the very successful permutation invariant training (PIT) approach, although the frequent label assignment switching happening during PIT training remains to be a problem when better convergence speed and achievable performance are desired. In this paper, we propose to perform self-supervised pre-training to stabilize the label assignment in training the speech separation model. Experiments over several types of self-supervised approaches, several typical speech separation models and two different datasets showed that very good improvements are achievable if a proper self-supervised approach is chosen.

Tasks

Benchmark Results

DatasetModelMetricClaimedVerifiedStatus
Libri2MixConv-Tasnet (Libri1Mix speech enhancement pre-trained)SI-SDRi14.1Unverified
Libri2MixConv-Tasnet (Libri1Mix speech enhancement multi-task)SI-SDRi13.7Unverified
Libri2MixConv-TasnetSI-SDRi13.2Unverified
WSJ0-2mixDPTNet (Libri1Mix speech enhancement pre-trained)SI-SDRi21.3Unverified

Reproductions