Bilingual Dual-Head Deep Model for Parkinson's Disease Detection from Speech
This addresses Parkinson's disease detection from speech for bilingual populations, offering an incremental improvement over existing methods by handling language variations.
The paper tackles Parkinson's disease detection from speech in bilingual settings by proposing a dual-head deep neural architecture that processes diadochokinetic and natural speech patterns separately, using self-supervised learning and wavelet transforms with adaptive layers to reduce language variations. Results show the model improves cross-linguistic generalization on Slovak and Spanish datasets, outperforming conventional single-language models and naive dataset combinations.
This work aims to tackle the Parkinson's disease (PD) detection problem from the speech signal in a bilingual setting by proposing an ad-hoc dual-head deep neural architecture for type-based binary classification. One head is specialized for diadochokinetic patterns. The other head looks for natural speech patterns present in continuous spoken utterances. Only one of the two heads is operative accordingly to the nature of the input. Speech representations are extracted from self-supervised learning (SSL) models and wavelet transforms. Adaptive layers, convolutional bottlenecks, and contrastive learning are exploited to reduce variations across languages. Our solution is assessed against two distinct datasets, EWA-DB, and PC-GITA, which cover Slovak and Spanish languages, respectively. Results indicate that conventional models trained on a single language dataset struggle with cross-linguistic generalization, and naive combinations of datasets are suboptimal. In contrast, our model improves generalization on both languages, simultaneously.