A Novel Self-training Approach for Low-resource Speech Recognition

TL;DR

This paper introduces a self-training method that enhances low-resource speech recognition by generating accurate pseudo-labels, significantly reducing word error rates and outperforming existing models on Punjabi datasets.

Contribution

The paper presents a novel self-training approach tailored for low-resource languages, improving speech recognition accuracy with pseudo-labels and achieving state-of-the-art results.

Findings

14.94% relative WER improvement over baseline

Significant performance gains on Punjabi datasets

Effective pseudo-label generation for low-resource speech

Abstract

In this paper, we propose a self-training approach for automatic speech recognition (ASR) for low-resource settings. While self-training approaches have been extensively developed and evaluated for high-resource languages such as English, their applications to low-resource languages like Punjabi have been limited, despite the language being spoken by millions globally. The scarcity of annotated data has hindered the development of accurate ASR systems, especially for low-resource languages (e.g., Punjabi and M\=aori languages). To address this issue, we propose an effective self-training approach that generates highly accurate pseudo-labels for unlabeled low-resource speech. Our experimental analysis demonstrates that our approach significantly improves word error rate, achieving a relative improvement of 14.94% compared to a baseline model across four real speech datasets. Further, our proposed approach reports the best results on the Common Voice Punjabi dataset.