Investigating the VR rollercoaster experience by applying AI models to biometric data

Detecting emotional components within a Virtual Reality is crucial for evaluating the user experience. However, it’s not always easy to detect changes in emotional states during immersive scenarios. This paper investigates the use of an IoT stress detection device aimed at collecting heartbeat and galvanic skin data during the VR rollercoaster simulation, consisting of a moving chair and a synchronized visor, providing a 3D scenario with adjustable speed. The biometric data was processed by a machine learning model, accurately selected after a comparative analysis between three Clustering models, to identify the different stress levels felt by the participants during the ride. Their stress levels were then cross-referenced with their perceptions on the ride to demonstrate whether VR can have an impact on their emotions. The results show how the VR can affect the perceived levels of stress and how a VR simulation could be comparable to a real roller coaster.

This version of the article has been accepted for publication, after peer review (when applicable) and is subject to Springer Nature’s AM terms of use (https://www.springernature.com/gp/open-research/policies/accepted-ma...), but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: http://dx.doi.org/10.1007/978-3-032-11355-9_30