Augmenting Dysarthric Speech Severity Assessment with MOS Supervision

2026-06-16

AI Summary

This study addresses the challenge of automatic dysarthric speech assessment, which is hindered by the scarcity of clinically annotated data. It presents the first evidence that synthetic speech artifacts share perceptual similarities with dysarthric speech, and proposes leveraging a large-scale synthetic speech evaluation corpus—QualiSpeech, which includes Mean Opinion Score (MOS) labels—as a source of cross-domain data augmentation. By fine-tuning or jointly training models on this augmented dataset, the approach substantially reduces reliance on scarce clinical annotations. Experimental results demonstrate consistent performance gains in predicting both intelligibility and naturalness, thereby validating the effectiveness and novelty of the proposed data augmentation strategy.

Abstract

Dysarthria is a speech disorder marked by reduced intelligibility and communicative effectiveness. Automatic utterance-level assessment of dysarthric speech can support scalable speech monitoring and therapy-related analysis. Yet training such systems is bottlenecked by the scarcity of clinically annotated dysarthric speech. This work proposes to augment dysarthric speech assessment using data from speech synthesis evaluations, specifically human-annotated utterances with Mean Opinion Score (MOS) labels from the QualiSpeech corpus. Experiments show that fine-tuning on speech synthesis assessment data consistently improves performance on both intelligibility and naturalness prediction, while joint training yields gains primarily on naturalness. These results suggest that synthesis artifacts and dysarthric speech share perceptual commonalities, and speech synthesis evaluation corpora offer a practical augmentation source that reduces reliance on scarce clinical annotations.

Problem

• dysarthric speech
• severity assessment
• clinical annotation scarcity
• speech intelligibility
• automatic assessment

Innovation

• dysarthric speech assessment
• data augmentation
• Mean Opinion Score (MOS)
• speech synthesis evaluation
• perceptual commonality