Vietnamese Automatic Speech Recognition: A Revisit

TL;DR

This paper introduces a new data aggregation and preprocessing pipeline that creates high-quality, large-scale Vietnamese ASR datasets from noisy open-source sources, improving the foundation for robust low-resource language ASR systems.

Contribution

It presents a novel, generalizable pipeline for constructing high-quality ASR datasets from diverse sources, specifically applied to Vietnamese, resulting in a 500-hour dataset.

Findings

Created a 500-hour Vietnamese ASR dataset

Demonstrated improved ASR performance using the dataset

Provided a scalable pipeline for low-resource language data collection

Abstract

Automatic Speech Recognition (ASR) performance is heavily dependent on the availability of large-scale, high-quality datasets. For low-resource languages, existing open-source ASR datasets often suffer from insufficient quality and inconsistent annotation, hindering the development of robust models. To address these challenges, we propose a novel and generalizable data aggregation and preprocessing pipeline designed to construct high-quality ASR datasets from diverse, potentially noisy, open-source sources. Our pipeline incorporates rigorous processing steps to ensure data diversity, balance, and the inclusion of crucial features like word-level timestamps. We demonstrate the effectiveness of our methodology by applying it to Vietnamese, resulting in a unified, high-quality 500-hour dataset that provides a foundation for training and evaluating state-of-the-art Vietnamese ASR systems. Our project page is available at https://github.com/qualcomm-ai-research/PhoASR.