Gaze on the Go: Effect of Spatial Reference Frame on Visual Target Acquisition During Physical Locomotion in Extended Reality

Title of the Paper

Gaze on the Go: Effect of Spatial Reference Frame on Visual Target Acquisition During Physical Locomotion in Extended Reality

Paper Information

• Research Area: Human-Computer Interaction (HCI), Virtual Reality (VR), and Augmented Reality (AR)
• Keywords: Spatial User Interface (Spatial UI), Reference Frame, UI Placement, Physical Locomotion, Extended Reality, Gaze Interaction, Eye Tracking

Research Background and Problem

• What issues or challenges did the authors identify?
  Spatial interaction using head-mounted displays (HMDs) requires users to quickly and accurately acquire visual targets. However, when users are in motion, such as walking or jogging, target acquisition time and tracking accuracy are significantly affected by the movement.

• Why is this issue important?
  As head-mounted displays are increasingly used in dynamic environments (e.g., AR technician operations or target selection in VR games), understanding gaze interaction performance during motion is crucial for enhancing user experience and improving UI design.

• Research Motivation and Related Work
  While there has been research on target acquisition and gaze performance for fixed or moving targets, most studies focus on static conditions or simple head-fixed reference frames (Head Frame) and world-fixed reference frames (World Frame). This study is the first to experimentally compare visual target acquisition performance across different reference frames, including a newly proposed Path reference frame.

Solution

• What methods or solutions did the authors propose?
  This study designed four spatial reference frames:

  1. Head: Targets are fixed relative to the head and move with its translation and rotation.
  2. HeadDelay: Similar to Head but with added delay to simulate inertia.
  3. Path (newly proposed frame): Targets remain stable relative to the path and are unaffected by head rotation.
  4. World: Targets are fixed in the environment and do not move with the user’s position.

• What is innovative about this solution?
  The innovation lies in the proposal of the Path reference frame and the first-time investigation of the feasibility of the HeadDelay frame in dynamic environments. The study compares the time and accuracy performance of these four reference frames through VR experiments.

• What are the implementation steps and key technologies used?

  1. Using the HTC Vive Pro Eye headset and Tobii eye-tracking equipment, virtual test scenarios were built on the Unity platform.
  2. The experiment involved user walking and jogging states, with targets fixed in different reference frames.
  3. Measurements included target acquisition time, gaze tracking precision, and gaze tracking trueness, alongside subjective user evaluations.
  4. Data analysis methods included four-factor repeated measures ANOVA and non-parametric tests.

Research Findings

• What specific results were achieved?

  1. Acquisition Time: Path and World reference frames performed better during motion, with significantly shorter acquisition times compared to the Head reference frame.
  2. Gaze Tracking Precision: Path and World frames maintained higher precision across different distances and motion speeds, while Head and HeadDelay frames showed lower precision.
  3. Gaze Tracking Trueness: The impact of reference frames on trueness was more complex, but the Head frame generally performed poorly.
  4. Subjective Evaluation: Users found the Path and World reference frames easier to operate during gaze and motion tasks.

• What advantages does it have compared to existing solutions?
  The Path frame combines the mobility of the Head frame with the environmental stability of the World frame, making it particularly effective in dynamic scenarios. The HeadDelay frame partially compensates for the shortcomings of direct positioning in the Head frame.

• What were the experimental or evaluation results?

  • At higher speeds, target acquisition became more challenging, but the Path frame maintained stable acquisition times.
  • During motion, the gaze precision of the Head reference frame decreased significantly.
  • Distance had minimal impact on gaze tracking precision, but nearby targets were easier to maintain using the Path frame.

• Limitations and Future Directions

  • Limitations:
    • The study was primarily conducted in virtual reality environments, without covering the complexity of real-world augmented reality (AR) scenarios.
    • The linear and flat motion design may not fully reflect complex use cases.
  • Future Directions:
    • Investigate the effects of reference frames in scenarios with obstacles and free movement paths.
    • Explore the applicability of reference frames in complex environments (e.g., navigation and multi-user collaboration applications).
    • Develop more real-time optimization solutions to address increased visual noise.

Conclusion

This study systematically compared users' gaze target acquisition and tracking abilities across different reference frames during motion, demonstrating the potential of the Path and World frames to enhance user experience. It provides important design guidelines for HMD development.