Self-training And Pre-training Are Complementary For Speech Recognition

Abstract

Self-training and unsupervised pre-training have emerged as effective approaches to improve speech recognition systems using unlabeled data. However, it is not clear whether they learn similar patterns or if they can be effectively combined. In this paper, we show that pseudo-labeling and pre-training with wav2vec 2.0 are complementary in a variety of labeled data setups. Using just 10 minutes of labeled data from Libri-light as well as 53k hours of unlabeled data from LibriVox achieves WERs of 3.0%/5.2% on the clean and other test sets of Librispeech - rivaling the best published systems trained on 960 hours of labeled data only a year ago. Training on all labeled data of Librispeech achieves WERs of 1.5%/3.1%.

Stats

• citations76
• S2 citations—
• github stars0
• HF likes0
• heat score14.15
• arxiv keyxu2020self

Related papers

• Self-training For End-to-end Speech Recognition (2019)15.48
• Wav2vec 2.0: A Framework For Self-supervised Learning Of Speech Representations (2020)0.00
• Censer: Curriculum Semi-supervised Learning For Speech Recognition Based On Self-supervised Pre-training (2022)4.52
• Weakly-supervised Speech Pre-training: A Case Study On Target Speech Recognition (2023)8.09
• Wav2vec: Unsupervised Pre-training For Speech Recognition (2019)0.00
• Unsupervised Speech Recognition (2021)0.00
• Joint Speech Transcription And Translation: Pseudo-labeling With Out-of-distribution Data (2022)0.00
• Robust Wav2vec 2.0: Analyzing Domain Shift In Self-supervised Pre-training (2021)25.07