Spatial Gaze Markers: Supporting Effective Task Switching in Augmented Reality

Document Title

Spatial Gaze Markers: Supporting Effective Task Switching in Augmented Reality

Document Information

• Subject Area: Human-computer interaction and task switching in augmented reality
• Keywords: augmented reality, eye tracking, gaze interaction, task switching, attention shift

Research Background and Problem

• Problem and Challenges:

  • Humans may face memory limitations when switching tasks in complex environments, such as remembering the target location of previous work in multitasking scenarios.
  • Spatial memory can be affected by visual distractions, task complexity, or interference, leading to reduced efficiency and task continuity interruptions.

• Significance:

  • Task switching is a common behavior in daily life and work. Reducing spatial memory confusion caused by attention shifts can significantly improve operational efficiency and reduce cognitive load.

• Research Motivation and Related Work:

  • Previous studies have shown that gaze cues on screens can help users switch tasks more quickly, such as technologies like Gazemarks and EyeBookmark. These studies focus on task assistance in digital spaces, leaving room for exploration in physical environments.
  • Visual guidance in augmented reality is often based on deep knowledge of the environment and tasks. Developing task-agnostic tools can enhance generalizability.

Solution

• Method and Solution:

  • Proposes an augmented reality system—Spatial Gaze Markers (SGM)—that automatically detects users' attention shifts through eye tracking and leaves markers in the visual space to assist task resumption.
  • SGM operates based on eye tracking and scene depth without requiring knowledge of specific tasks or activities, making it a "task-agnostic" solution.

• Innovativeness:

  • Unlike existing screen-based task cues, SGM extends to real physical environments, allowing users to be guided by visual markers when returning to a previously attended area.
  • It features low intrusiveness and the capability for large-scale deployment without pre-configuring the workspace.

• Implementation Steps and Key Technologies:

  • Gaze Detection: Uses a real-time I-DT algorithm to detect fixation points and generate 3D gaze positions.
  • Attention Shift Detection: Determines whether the user has shifted attention away from the task area. When the gaze point moves out of the augmented reality display's field of view, the system records the shift location.
  • Marker Visualization: Displays a visual marker of the last fixation position when the user returns attention to the task area, using a transparent circular design to avoid task interference.
  • Marker Removal: Automatically removes the marker when the user gazes at its location, minimizing impact on the user interface.

Research Outcomes

• Specific Results:

  • SGM significantly reduced task resumption time and alleviated visual search burden in simulated maintenance and inspection task experiments.
  • Experiments showed that under conditions with distractions, SGM was more effective than baseline conditions without markers in reducing task resumption time (relative reduction of 41-52%).
  • User feedback indicated that SGM performed well in aiding memory of target locations and reducing task load.

• Advantages:

  • Reduced cognitive load: NASA TLX scale scores showed that using SGM significantly lowered mental demand, effort, and frustration levels.
  • Task-agnostic: SGM can be quickly deployed across various scenarios without pre-configuration or additional markers.

• Experimental and Evaluation Results:

  • Under both distraction and non-distraction conditions, SGM consistently outperformed the baseline in task resumption time and reduced the number of re-checks.
  • In "inspection" tasks, users rated SGM as having lower mental load and requiring fewer re-checks.

• Limitations and Future Directions:

  • The abstraction of task design may limit the external validity of the results. Future research should investigate the effectiveness of SGM in more realistic and long-term use cases.
  • The limited field of view of HoloLens 2 constrains SGM's practicality. Future work could explore performance with wider-angle display devices.
  • Current user feedback on the disappearance speed of visual markers suggests optimizing visual design, such as using gestures or actions to avoid accidental deletions.

Conclusion

This study introduces Spatial Gaze Markers, an eye-tracking-based task-switching assistance system. By reducing visual search difficulty and improving task resumption efficiency, it provides a task-agnostic, low-intrusiveness solution for the augmented reality field. With low deployment costs and broad application potential, this system holds promise for both everyday and industrial environments. However, future research should further optimize visual design, expand device applicability, and validate its performance in more complex scenarios.