Unsupervised Domain Adaptation Schemes for Building ASR in Low-resource Languages

TL;DR

This paper demonstrates that unsupervised domain adaptation techniques, specifically adversarial training and domain separation networks, can significantly improve low-resource language ASR performance using high-resource language data.

Contribution

It introduces and evaluates two UDA architectures for building ASR in low-resource languages, showing notable improvements over baseline models.

Findings

GRL and DSN architectures improve WER by over 6% and 7%.

Proper source language selection enhances adaptation performance.

UDA schemes reduce the need for large annotated datasets in low-resource languages.

Abstract

Building an automatic speech recognition (ASR) system from scratch requires a large amount of annotated speech data, which is difficult to collect in many languages. However, there are cases where the low-resource language shares a common acoustic space with a high-resource language having enough annotated data to build an ASR. In such cases, we show that the domain-independent acoustic models learned from the high-resource language through unsupervised domain adaptation (UDA) schemes can enhance the performance of the ASR in the low-resource language. We use the specific example of Hindi in the source domain and Sanskrit in the target domain. We explore two architectures: i) domain adversarial training using gradient reversal layer (GRL) and ii) domain separation networks (DSN). The GRL and DSN architectures give absolute improvements of 6.71% and 7.32%, respectively, in word error rate over the baseline deep neural network model when trained on just 5.5 hours of data in the target domain. We also show that choosing a proper language (Telugu) in the source domain can bring further improvement. The results suggest that UDA schemes can be helpful in the development of ASR systems for low-resource languages, mitigating the hassle of collecting large amounts of annotated speech data.