Towards Inclusive ASR: Investigating Voice Conversion for Dysarthric Speech Recognition in Low-Resource Languages

TL;DR

This paper explores using voice conversion to generate dysarthric speech in multiple languages, improving ASR performance for dysarthric speakers in low-resource settings by augmenting training data.

Contribution

It introduces a novel voice conversion approach that encodes dysarthric speech characteristics and applies it to enhance multilingual ASR models for low-resource languages.

Findings

Voice conversion significantly improves dysarthric speech recognition accuracy.

Generated speech effectively simulates dysarthric characteristics.

Proposed method outperforms traditional data augmentation techniques.

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.