MADI: Inter-domain Matching and Intra-domain Discrimination for Cross-domain Speech Recognition

End-to-end automatic speech recognition (ASR) usually suffers from performance degradation when applied to a new domain due to domain shift. Unsupervised domain adaptation (UDA) aims to improve the performance on the unlabeled target domain by transferring knowledge from the source to the target dom...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2023-02
Main Authors: Zhou, Jiaming, Zhao, Shiwan, Jiang, Ning, Zhao, Guoqing, Qin, Yong
Format: Article
Language:English
Subjects:
Automatic speech recognition
Domains
Matching
Performance degradation
Performance enhancement
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:End-to-end automatic speech recognition (ASR) usually suffers from performance degradation when applied to a new domain due to domain shift. Unsupervised domain adaptation (UDA) aims to improve the performance on the unlabeled target domain by transferring knowledge from the source to the target domain. To improve transferability, existing UDA approaches mainly focus on matching the distributions of the source and target domains globally and/or locally, while ignoring the model discriminability. In this paper, we propose a novel UDA approach for ASR via inter-domain MAtching and intra-domain DIscrimination (MADI), which improves the model transferability by fine-grained inter-domain matching and discriminability by intra-domain contrastive discrimination simultaneously. Evaluations on the Libri-Adapt dataset demonstrate the effectiveness of our approach. MADI reduces the relative word error rate (WER) on cross-device and cross-environment ASR by 17.7% and 22.8%, respectively.
ISSN:2331-8422