Towards Inclusive ASR: Investigating Voice Conversion for Dysarthric Speech Recognition in Low-Resource Languages
Abstract
Automatic speech recognition (ASR) for dysarthric speech remains challenging due to data scarcity, particularly in non-English languages. To address this, we fine-tune a voice conversion model on English dysarthric speech (UASpeech) to encode both speaker characteristics and prosodic distortions, then apply it to convert healthy non-English speech (FLEURS) into non-English dysarthric-like speech. The generated data is then used to fine-tune a multilingual ASR model, Massively Multilingual Speech (MMS), for improved dysarthric speech recognition. Evaluation on PC-GITA (Spanish), EasyCall (Italian), and SSNCE (Tamil) demonstrates that VC with both speaker and prosody conversion significantly outperforms the off-the-shelf MMS performance and conventional augmentation techniques such as speed and tempo perturbation. Objective and subjective analyses of the generated data further confirm that the generated speech simulates dysarthric characteristics.
Keywords
Cite
@article{arxiv.2505.14874,
title = {Towards Inclusive ASR: Investigating Voice Conversion for Dysarthric Speech Recognition in Low-Resource Languages},
author = {Chin-Jou Li and Eunjung Yeo and Kwanghee Choi and Paula Andrea Pérez-Toro and Masao Someki and Rohan Kumar Das and Zhengjun Yue and Juan Rafael Orozco-Arroyave and Elmar Nöth and David R. Mortensen},
journal= {arXiv preprint arXiv:2505.14874},
year = {2025}
}
Comments
5 pages, 1 figure, Proceedings of Interspeech