Methods to Increase the Amount of Data for Speech Recognition for Low Resource Languages

TL;DR

This paper tackles data scarcity in ASR for two low-resource languages, Armenian and Georgian, by systematically evaluating diverse data-extension strategies. It combines expanding Common Voice, creating open audiobook and text resources, crowd-sourcing speech, and leveraging pseudo-labeling within robust data-processing pipelines, all evaluated on a compact FastConformer model. The results show substantial improvements in WER, including a Georgian WER of $5.73\%$ and Armenian improvements reaching near Georgian performance, underscoring the value of multi-source data augmentation. The work provides practical guidance on cost-effective data collection for low-resource languages and publicly releases models and datasets to support continued research and real-world deployment.

Abstract

This study explores methods to increase data volume for low-resource languages using techniques such as crowdsourcing, pseudo-labeling, advanced data preprocessing and various permissive data sources such as audiobooks, Common Voice, YouTube. While these methods are well-explored for highresource languages, their application for low-resource languages remains underexplored. Using Armenian and Georgian as case studies, we demonstrate how linguistic and resource-specific characteristics influence the success of these methods. This work provides practical guidance for researchers to choose cost-effective and quality-driven dataset extension strategies for low-resource languages. The key takeaway from various data extension approaches is that paid crowd-sourcing offers the best balance between cost and quality, outperforming volunteer crowd-sourcing, open-source audiobooks, and unlabeled data usage. Ablation study shows that models trained on the expanded datasets outperform existing baselines and achieve 5.73% for Gergian and 9.9% for Armenian ASR word error rate using a relatively small FastConformer architecture. We open-sourced both the Armenian and Georgian models to allow further research and practical applications.