Mixing in Reverse Optical Flow to Mitigate Vection and Simulation Sickness in Virtual Reality

Scientific concepts / nature & physiology phenomena

• Virtual reality (VR) vection and simulation sickness

  • Motion symptoms (e.g., nausea, disorientation, headaches) are attributed to sensory mismatch between:
    • Visual motion cues (what the eyes/visual system detect)
    • Vestibular signals (what the inner ear/vestibular system detects)

• Sensory conflict mechanism

  • The brain interprets the environment as moving when the vestibular system indicates otherwise (or vice versa), producing sickness.

• Optical flow as a visual-motion representation

  • Optical flow describes the apparent motion of image patterns.

• Reverse optical flow mitigation idea

  • The method adds/artificially injects optical flow in the reverse direction of the user’s perceived motion to cancel or reduce the visual motion component that contributes to sensory mismatch.

Method / system (as described)

• Online reverse optical-flow injection

  • At fixed time windows, the algorithm:
    • Extracts visual pictures once at the start of the time window (locations are fixed).
    • Over the window duration, it completes the injected motion by leveraging those initial extracted frames at predetermined positions.

• Pre-processing

  • Rendered VR frames are downsampled and converted to a contrast/gray “clear level” image (as described).

• Line-segment processing and overlay

  • The algorithm uses line segments and overlays processed results onto the image plane to create the reverse optical flow visualization.

• Evaluation measurements

  • Compares conditions with and without the proposed “reverse optical flow” approach.
  • Tracks:
    • Self-reported sickness
    • Presence/immersion (emergence/presence measures)
    • Motion profile data and relationships to motion type
    • An implicit/observed “impact” across different categories/content types

Main findings (reported)

• Sickness reduction

  • The proposed approach shows significant reduction of sickness in many conditions/categories.

• Presence effects

  • Presence/immersion was significantly reduced in one experiment (e.g., first experiment), but not in another, suggesting the effect may depend on scenario/content or tuning.

• Limitations / content dependence

  • Shakiness reduced but not eliminated.
  • The technique appears only partially effective:
    • Not all visual information/pictures are extracted or used.
    • Performance was less effective for “shimmer light” / certain light content (auto-caption wording).
  • The injected reverse-flow visualization can be intrusive to the original content (“still somewhat invasive”).

• Visual saliency dependence

  • In some cases (e.g., “space exploration content” described), reverse-flow line segments were more distinct and eye-catching, correlating with stronger mitigation.

Proposed future directions

• Use actual motion tracking (or source-predicted optical flow) instead of relying only on the current approximation.
• Combine with other mitigation techniques for practical VR deployment.
• Adaptive application based on camera motion (e.g., “sudo reverse flow from camera motion adaptively apply”).
• Support different visualization messaging strategies.

Researchers / sources featured

• No specific individual researchers or external sources are named in the provided subtitles (aside from “Franta” and “Korea University” as an affiliation reference).