Cybersickness assessment for frame rate of ultra-wide display with deep model of visual perception mismatch

Abstract

This study investigates the method of assessing the cybersickness according to the frame rate in the ultra-wide display, which is one of the immersive virtual environments. Cybersickness is a similar symptom to motion sickness while experiencing the virtual environments, causing serious concern about the viewing safety of the immersive virtual environments. In particular, frame rate is one of the most important and common causes of cybersickness. In this study, we propose an objective cybersickness assessment method due to frame rate by modeling the visual perception mismatch using deep learning. To do this, we devise the visual expectation network through unsupervised learning and the cybersickness prediction network through supervised learning. We also conducted an extensive subject measurement experiments on the cybersickness score at various frame rates to evaluate performance of the proposed network.