Ranking pre-trained speech embeddings in Parkinson's disease detection: Does Wav2Vec 2.0 outperform its 1.0 version across speech modes and languages?
Authors:
Open Access
- 1 January 2025
- journal article
- Published by American Association for the Advancement of Science (AAAS) in Computational and Structural Biotechnology Journal
- Vol. 27, 2584-2601
Abstract
Speech and language technologies are effective tools for identifying the distinct speech changes associated with Parkinson's disease (PD), enabling earlier and more accurate diagnosis. Models leveraging recent advancements in self-supervised speech pretraining, such as Wav2Vec, have demonstrated superior performance over traditional feature extraction methods. While Wav2Vec 2.0 has been successfully utilized for PD detection, a rigorous quantitative comparison with Wav2Vec 1.0 is needed to comprehensively evaluate its advantages, limitations, and applicability across different speech modes in PD. This study presents a systematic comparison of Wav2Vec 1.0 and Wav2Vec 2.0 embeddings across three multilingual datasets using various classification approaches to classify normal (healthy controls; HC) and PD-affected speech. Additionally, both Wav2Vec 1.0 and 2.0 were benchmarked against traditional baseline features across diverse linguistic contexts, including spontaneous speech, non-spontaneous speech, and isolated vowels. A multicriteria TOPSIS approach was employed to rank feature extraction methods, revealing that Wav2Vec 2.0 excelled across speech modes, with its first transformer layer demonstrating the best performance for classifying read text and monologue, and its feature extractor performing best in vowel-based classification. In contrast, Wav2Vec 1.0, while generally outperformed by Wav2Vec 2.0, still provided a more efficient alternative with competitive performance. Finally, we combined selected layers from both architectures and have demonstrated improved diagnostic accuracy in vowel-based classification. This comparative analysis underscores the strengths of both Wav2Vec architectures and informs their optimal use in PD detection.Keywords
Funding Information
- Ministerstvo Školství, Mládeže a Tělovýchovy
This publication has 34 references indexed in Scilit:
- A Web Application for exploratory data analysis and classification of Parkinson’s Disease patients using machine learning models on different datasetsSoftware Impacts, 2024
- From prodromal stages to clinical trials: The promise of digital speech biomarkers in Parkinson's diseaseNeuroscience & Biobehavioral Reviews, 2024
- Comparison of wav2vec 2.0 models on three speech processing tasksInternational Journal of Speech Technology, 2024
- Voice Disorder Classification Using Wav2vec 2.0 Feature ExtractionJournal of Voice, 2024
- Exploring the potential of Wav2vec 2.0 for speech emotion recognition using classifier combination and attention-based feature fusionThe Journal of Supercomputing, 2024
- Pre-trained models for detection and severity level classification of dysarthria from speechSpeech Communication, 2024
- Interpretable speech features vs. DNN embeddings: What to use in the automatic assessment of Parkinson’s disease in multi-lingual scenariosComputers in Biology and Medicine, 2023
- On the inter-dataset generalization of machine learning approaches to Parkinson's disease detection from voiceInternational Journal of Medical Informatics, 2023
- Analysis of spontaneous speech in Parkinson's disease by natural language processingParkinsonism & Related Disorders, 2023
- An improved framework for Parkinson’s disease prediction using Variational Mode Decomposition-Hilbert spectrum of speech signalBiocybernetics and Biomedical Engineering, 2021