Logo image
Back
Machine Learning Classifiers for Voice Health Assessment Under Simulated Room Acoustics
Conference proceeding   Open access

Machine Learning Classifiers for Voice Health Assessment Under Simulated Room Acoustics

Ahmed Yousef and Eric J Hunter
Engineering Proceedings, Vol.81(1), 16
The 1st International Online Conference on Bioengineering
05/06/2025
DOI: 10.3390/engproc2024081016
PMCID: PMC12862568
PMID: 41635464
url
https://doi.org/10.3390/engproc2024081016View
Published (Version of record) Open Access

Abstract

Machine learning (ML) robustness for voice disorder detection was evaluated using reverberation-augmented recordings. Common vocal health assessment voice features from steady vowel samples (135 pathological, 49 controls) were used to train/test six ML classifiers. Detection performance was evaluated under low-reverb and simulated medium (med = 0.48 s) and high-reverb times (high = 1.82 s). All models’ performance declined with longer reverberation. Support Vector Machine exhibited slight robustness but faced performance challenges. Random Forest and Gradient Boosting, though strong under low reverb, lacked generalizability in med/high reverb. Training/testing ML on augmented data is essential to enhance their reliability in real-world voice assessments.
 Machine Learning data augmentation overfitting random forest reverberation room acoustics speech acoustics support vector machine voice assessment voice disorders

Details

Metrics

22 Record Views
Logo image