Virtual reality (VR) sickness remains a significant challenge in the widespread adoption of immersive VR experiences. VR sickness is typically measured by collecting self-reported data from the user post-exposure through various questionnaires. Traditional self-reported assessments of VR sickness suffer from subjectivity and non-real-time data collection, leading to unreliable results. Moreover, gender differences in reporting symptoms further complicate the evaluation process.
To address these limitations, these researchers propose a novel approach using computational models to objectively measure VR sickness. The project aims to combine physiological sensor data, scene statistics, subjective self-reports, and historical user data to develop accurate and robust models for predicting VR sickness in real time. By collecting real-time performance data, alongside subjective questionnaire responses, the researchers will build a comprehensive dataset of VR sickness data. This dataset will then be used to train reliable and robust models for measuring predicting VR sickness in real time. This dataset will also serve as a valuable resource for the VR community, enhancing the safety and usability of VR experiences. The computational demands of processing and analyzing this extensive dataset require advanced computing resources. This project holds the potential to significantly advance the current methods of VR sickness measurement and improve the overall VR user experience.