PR1030: Machine Learning VocalSensePro

Asst. Prof. Dr. Kirandeep Kaur Universiti Tunku Abdul Rahman (UTAR)

VIC26 | Professional

CR: 0.0000 | 0 Likes | 4 Views | 2 times | LS: 2.0
Like it? | Support them now!

Machine Learning VocalSensePro is an intelligent voice analysis and dysphonia detection system developed to support the early identification and monitoring of vocal disorders using artificial intelligence and machine learning technologies. Dysphonia is acondition that affects voice quality, pitch, loudness, and speech clarity, often caused by excessive vocal strain, improper voice usage, or underlying medical conditions. Although vocal disorders are increasingly common, especially among professional voice users, there are still limited accessible and affordable tools available for early screening and continuous monitoring. Conventional diagnostic methodstypically require specialized clinical equipment, expert evaluation, and hospital-based assessments, making early detection difficult for many individuals, particularly in schools, workplaces, rural healthcare settings, and remote areas. This limitation has created the need for a practical, portable, and technology-driven solution, leading to the development of Machine Learning VocalSensePro. The system utilizes speech signal processing and supervised machine learning algorithms to analyze voice recordings and identify patterns associated with dysphonia. Important acoustic features such as pitch, jitter, shimmer, harmonics-to-noise ratio, spectral parameters, and Mel-frequency cepstral coefficients (MFCCs) are extracted and processed to differentiate between healthy and dysphonic voices with high accuracy and consistency. By reducing dependence on subjective interpretation, the system improves reliability and efficiency in preliminary voice disorder screening. Machine Learning VocalSensePro is designed as a non-invasive, user-friendly, and portable platform that can operate through computers or mobile devices, enabling real-time voice assessment and automated feedback without requiring specialized clinical tools. The innovation is highly suitable for telehealth applications, speech therapy support, occupational 
health monitoring, and preventive healthcare services. The targeted commercialization industries for this innovation include educational institutions, healthcare providers, speech therapy centers, broadcasting and media industries, music and entertainment sectors, and occupational health organizations. The system is especially beneficial for teachers, lecturers, singers, vocalists, broadcasters, public speakers, and call center operators who heavily depend on vocal performance in their professions. Future improvements of the system include expanding multilingual voice datasets, integrating deep learning techniques for enhanced prediction accuracy, developing cloud-based monitoring systems, incorporating mobile application support, and adding personalized vocal health recommendations and real-time voice coaching features. Overall, Machine Learning VocalSensePro offers a scalable, cost-effective, and innovative solution for modern vocal healthcare and early dysphonia detection.