Unsupervised Pre-Training for Vietnamese Automatic Speech Recognition in the HYKIST Project

TL;DR

This paper explores unsupervised pre-training methods for Vietnamese ASR in the medical domain, focusing on low-resource scenarios and comparing publicly available models with custom pre-trained models.

Contribution

It investigates the effectiveness of unsupervised pre-training and data strategies for Vietnamese medical speech recognition, a low-resource language task.

Findings

Unsupervised pre-training improves ASR performance in low-resource settings.

Customized pre-trained models outperform publicly available models.

Data combination strategies enhance recognition accuracy.

Abstract

In today's interconnected globe, moving abroad is more and more prevalent, whether it's for employment, refugee resettlement, or other causes. Language difficulties between natives and immigrants present a common issue on a daily basis, especially in medical domain. This can make it difficult for patients and doctors to communicate during anamnesis or in the emergency room, which compromises patient care. The goal of the HYKIST Project is to develop a speech translation system to support patient-doctor communication with ASR and MT. ASR systems have recently displayed astounding performance on particular tasks for which enough quantities of training data are available, such as LibriSpeech. Building a good model is still difficult due to a variety of speaking styles, acoustic and recording settings, and a lack of in-domain training data. In this thesis, we describe our efforts to construct ASR systems for a conversational telephone speech recognition task in the medical domain for Vietnamese language to assist emergency room contact between doctors and patients across linguistic barriers. In order to enhance the system's performance, we investigate various training schedules and data combining strategies. We also examine how best to make use of the little data that is available. The use of publicly accessible models like XLSR-53 is compared to the use of customized pre-trained models, and both supervised and unsupervised approaches are utilized using wav2vec 2.0 as architecture.