Optical Flow-Based Anticipatory Audio Cues for Cybersickness Mitigation in Passive Navigation VR Environments

Virtual reality (VR) applications face a critical trade-off: while visual immersion is the core experience, it often induces cybersickness (CS), leading to user attrition. Existing visual interventions (e.g., field-of-view restriction) alleviate discomfort at the cost of compromising immersion. Auditory interventions have been widely applied in mitigating motion sickness (MS), such as through music interventions and synchronized sound effects. Among all MS mitigation approaches, anticipatory audio cues have demonstrated effectiveness in vehicular MS contexts, reducing symptoms by approximately 17%. However, questions remain regarding VR cybersickness (CS): Do anticipatory audio cues remain effective in highly immersive virtual environments, where the dominance of visual interaction in VR may alter cue perception compared to that in MS contexts? How can VR motion intensity information be encoded, given that such information is largely absent in current mitigation strategies? How can anticipatory audio cues be systematically designed for CS mitigation through explicit mapping frameworks that link visual motion parameters to audio features? To address these gaps, this study establishes an optical flow mapping framework that encodes motion direction through stereo positioning and intensity through temporal patterns, delivering the results 2 seconds in advance. A within-subjects experiment (N = 24) was conducted in VR roller coaster environments, in which four conditions were compared. Results demonstrated a significant reduction in cybersickness across all audio conditions (direction-only: d = 0.862; intensity-only: d = 0.670; combined: d = 1.497, p