Mockingjay: Unsupervised Speech Representation Learning with Deep Bidirectional Transformer Encoders

TL;DR

Mockingjay introduces a bidirectional Transformer-based speech representation learning method that pre-trains on unlabeled speech, significantly improving downstream task performance and outperforming traditional features even with minimal labeled data.

Contribution

It is the first to use bidirectional Transformer encoders for unsupervised speech representation learning, enhancing downstream task accuracy with minimal labeled data.

Findings

Outperforms previous methods on phoneme, speaker, and sentiment classification.

Achieves significant performance gains with only 2 training epochs.

Outperforms Mel-features with all labeled data in low-resource scenarios.

Abstract

We present Mockingjay as a new speech representation learning approach, where bidirectional Transformer encoders are pre-trained on a large amount of unlabeled speech. Previous speech representation methods learn through conditioning on past frames and predicting information about future frames. Whereas Mockingjay is designed to predict the current frame through jointly conditioning on both past and future contexts. The Mockingjay representation improves performance for a wide range of downstream tasks, including phoneme classification, speaker recognition, and sentiment classification on spoken content, while outperforming other approaches. Mockingjay is empirically powerful and can be fine-tuned with downstream models, with only 2 epochs we further improve performance dramatically. In a low resource setting with only 0.1% of labeled data, we outperform the result of Mel-features that uses all 100% labeled data.