Skip to content
All projects

2023

Parkinson's Detection from Voice

Vocal Tab Transformer that classifies Parkinson's disease from dysphonia features, beating gradient-boosted trees by 1%+ AUC.

Parkinson's disease often shows up early in a person's voice. We extracted dysphonia measures from voice recordings and treated them as a large, complex tabular feature set.

We proposed the Vocal Tab Transformer, a transformer for tabular vocal features, which outperformed the previous state of the art - Gradient-Boosted Decision Trees - by at least 1% AUC with better precision and recall.

Highlights

  • Published in MDPI Biomimetics (2023)
  • Outperforms GBDT by ≥1% AUC