Self-training and Pre-training are Complementary for Speech Recognition

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-traini...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2020-10
Main Authors: Xu, Qiantong, Baevski, Alexei, Likhomanenko, Tatiana, Paden Tomasello, Conneau, Alexis, Collobert, Ronan, Synnaeve, Gabriel, Auli, Michael
Format: Article
Language:English
Subjects:
Speech recognition
System effectiveness
Test sets
Training
Voice recognition
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary: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%.
ISSN:2331-8422