Detecting Hostile Behaviour Through Real-time Voice Analysis

About the Project

This PhD study will explore the automation of hostility assessment from the human voice in real-time. The project will leverage recent advances in machine learning to train a hostility predictor model to identify and classify hostile behaviours.

Demonstrating persistent high levels of hostility in are associated with several morbidities including coronary artery disease. As such, a real-time, non-invasive approach to passive detection of hostile behaviour may have wide application in early detection and treatment of significant healthcare issues.

The PhD may explore:

• Gathering, processing and labelling of training data for an automated hostility predictor.
• Development of a novel approach to automated hostility prediction.
• Consideration of deployment (such as mobile applications) and/or healthcare applications (such as detection of coronary artery disease.

Supervisory team

• Dr Chris Bowers, Principal Lecturer, Digital Innovation and Intelligent Systems Research Group, University of Worcester
• Dr Felix Why, Senior Lecturer, School of Health and Wellbeing, Department of Psychology and Mental Health

Research Group: Digital Innovation and Intelligent Systems Research Group

Application Process:

To begin the application process please go to: https://www.worc.ac.uk/research/research-degrees/applying-for-a-phd/.