Pre-trained models for detection and severity level classification of dysarthria from speech

Farhad Javanmardi, Sudarsana Reddy Kadiri, Paavo Alku

Abstract: Highlights•Layer-wise analysis of pre-trained models for detection of dysarthria.•Layer-wise analysis of pre-trained models for severity level classification.•Systematic comparison of conventional features and pre-trained model embeddings.•The pre-trained model embeddings outperformed the conventional features.•The HuBERT features performed better than the wav2vec2 features.