In this paper, we provide a new perspective on self-supervised speech models from how the training targets are obtained. We generalize the targets extractor into Offline Targets Extractor (Off-TE) and Online Targets Extractor (On-TE). Based on this, we propose a new multi-tasking learning framework for self-supervised learning, MT4SSL, which stands for Boosting Self-Supervised Speech Representation Learning by Integrating Multiple Targets. MT4SSL uses the K-means algorithm as an Off-TE and a teacher network without gradients as an On-TE, respectively. Our model outperforms previous SSL methods by nontrivial margins on the LibriSpeech benchmark, and is comparable to or even better than the best-performing models with fewer data. Furthermore, we find that using both Off-TE and On-TE results in better convergence in the pre-training phase. With both effectiveness and efficiency, we think doing multi-task learning on self-supervised speech models from our perspective is a promising trend.
@article{arxiv.2211.07321,
title = {MT4SSL: Boosting Self-Supervised Speech Representation Learning by Integrating Multiple Targets},
author = {Ziyang Ma and Zhisheng Zheng and Changli Tang and Yujin Wang and Xie Chen},
journal= {arXiv preprint arXiv:2211.07321},
year = {2023}
}
Comments
Accepted to Interspeech 2023. Code available at: https://github.com/ddlBoJack/MT4SSL