MVP: Multi-source Voice Pathology detection

2025/05/26 by Alkis Koudounas, Moreno La Quatra, Koudounas, Alkis +15 · 1 citation
Computer Science · Medicine · #Audio and Speech Processing (eess.AS) #Computation and Language (cs.CL) #FOS: Computer and information sciences #FOS: Electrical engineering #Speech Recognition and Synthesis #Voice and Speech Disorders #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2505.20050

openalex publication_date 2025/05/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Voice disorders significantly impact patient quality of life, yet non-invasive automated diagnosis remains under-explored due to both the scarcity of pathological voice data, and the variability in recording sources. This work introduces MVP (Multi-source Voice Pathology detection), a novel approach that leverages transformers operating directly on raw voice signals. We explore three fusion strategies to combine sentence reading and sustained vowel recordings: waveform concatenation, intermediate feature fusion, and decision-level combination. Empirical validation across the German, Portuguese, and Italian languages shows that intermediate feature fusion using transformers best captures the complementary characteristics of both recording types. Our approach achieves up to +13% AUC improvement over single-source methods.

Citations

• Bilingual Dual-Head Deep Model for Parkinson's Disease Detection from Speech
• voc2vec: A Foundation Model for Non-Verbal Vocalization
• Exploring ASR-Based Wav2Vec2 for Automated Speech Disorder Assessment: Insights and Analysis
• Exploiting Foundation Models and Speech Enhancement for Parkinson's Disease Detection from Speech in Real-World Operative Conditions
• A Voice Disease Detection Method Based on MFCCs and Shallow CNN
• Detecting dementia from speech and transcripts using transformers
• HuBERT: Self-Supervised Speech Representation Learning by Masked\n Prediction of Hidden Units
• wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
• FiLM: Visual Reasoning with a General Conditioning Layer
• Gated Multimodal Units for Information Fusion

Cited by

• A Concept-based approach to Voice Disorder Detection

Related

• Attentive-based Multi-level Feature Fusion for Voice Disorder Diagnosis
• Voice Pathology Detection Using Phonation
• Pathological Voice Classification Using Mel-Cepstrum Vectors and Support Vector Machine
• Parameterization of Sequence of MFCCs for DNN-based voice disorder detection
• VocalAgent: Large Language Models for Vocal Health Diagnostics with Safety-Aware Evaluation