Adapting Self-Supervised Speech Representations for Cross-lingual Dysarthria Detection in Parkinson's Disease
This work addresses the challenge of limited dysarthric speech data for cross-lingual detection, which is important for healthcare applications but incremental in method.
The paper tackled the problem of cross-lingual dysarthria detection in Parkinson's disease by proposing a representation-level language shift method, which improved sensitivity and F1 scores in cross-lingual settings and showed consistent gains in multilingual settings.
The limited availability of dysarthric speech data makes cross-lingual detection an important but challenging problem. A key difficulty is that speech representations often encode language-dependent structure that can confound dysarthria detection. We propose a representation-level language shift (LS) that aligns source-language self-supervised speech representations with the target-language distribution using centroid-based vector adaptation estimated from healthy-control speech. We evaluate the approach on oral DDK recordings from Parkinson's disease speech datasets in Czech, German, and Spanish under both cross-lingual and multilingual settings. LS substantially improves sensitivity and F1 in cross-lingual settings, while yielding smaller but consistent gains in multilingual settings. Representation analysis further shows that LS reduces language identity in the embedding space, supporting the interpretation that LS removes language-dependent structure.