Back to Catalog

EWA-DB

Dataset

EWA-DB is a clinical speech database featuring audio recordings and ASR transcripts from 1,649 speakers, including healthy controls and patients with Alzheimer's, Parkinson's, or mild cognitive impairment. The dataset provides WAV files, manual annotations, and metadata across four distinct language tasks to support medical speech analysis and diagnostic research.

Gated access
Contributors
Language
sk
Modality
Speech/audio