Tracing Stress and Arousal in Virtual Reality Games Using Players’ Motor and Vocal Behaviour

In this study, we tackle the integration of stressors and voice interaction in a Virtual Reality game to assess players’ arousal and stress levels. The selected game genre and its characteristic components are used as a basis to create stress-inducing elements. Additionally, a voice interaction module has been created using a voice assistant called Minerva. The module allows for real-time detection and recording of players’ emotional responses based on variations in pitch and intensity of speech. The game consists of a single level divided into four areas with increasing levels of stress. The experiment involved 16 volunteer students who played the game while their prosodic and behavioral movement data were collected. Participants also completed questionnaires and produced ratings to assess their perceived stress and arousal levels. The collected data were analyzed to evaluate the effectiveness of the real-time estimation of arousal and stress.