DoubleMe: Local Blending in Multi-Display Environments with Augmented Reality to Facilitate Co-Located Collaboration

Paper Title

DoubleMe: Local Blending in Multi-Display Environments with Augmented Reality to Facilitate Co-Located Collaboration

Publication Info

• Topic area: Augmented Reality (AR) for co-located collaboration in multi-display environments.
• Keywords: Augmented Reality, co-located collaboration, multi-display environments, workspace duplication, avatars, spatial constraints, blended interaction spaces, proxemics, deictic gestures, user interfaces.

Background and Problem

• Problem / challenge: Co-located collaboration in multi-display environments (MDEs) is hindered by spatial constraints, such as stationary displays and limited physical mobility, making cross-referencing between displays challenging.
• Significance: Addressing these spatial constraints is critical for improving collaborative efficiency in settings like crisis management rooms, offices, and educational environments.
• Motivation and related work: Prior research on blended interaction spaces and AR systems has focused on remote collaboration, with limited exploration of spatial challenges in co-located contexts. Existing systems often require dedicated software or fail to address the dual presence of physical and virtual elements in co-located settings. This paper builds on these insights to explore local blending for co-located collaboration.

Solution

• Proposed approach: DoubleMe, an AR system that creates virtual duplicates of collaborators’ workspaces, including displays and avatars, allowing users to spatially rearrange these duplicates within their own workspace.
• Novelty:
  1. Introduction of locally blended interaction spaces for co-located collaboration.
  2. Development of the DoubleMe prototype to implement this concept.
  3. Empirical evaluation of how local blending influences collaboration strategies and user interaction.
• Procedure and key techniques:
  • Duplicates consist of virtual copies of displays and avatars, maintaining internal alignment and reciprocal spatial consistency.
  • Users can translate and rotate duplicates to optimize workspace layout.
  • Minimalist avatars are used to support deictic gestures while minimizing visual intrusion.
  • Two studies were conducted: one to assess avatar design and interaction with duplicates, and another to evaluate DoubleMe’s impact on collaborative tasks.

Results

• Concrete findings:
  • Participants relied on duplicates rather than physical displays, with frequent gaze and gesture switches between their own workspace and the duplicate.
  • Minimalist avatars were preferred for tasks involving deictic gestures, especially when duplicates violated physical or social norms.
  • DoubleMe reduced mental, physical, and temporal demands, and improved perceived performance and collaboration efficiency.
• Advantage over baselines:
  • DoubleMe outperformed non-augmented setups by enabling seamless cross-referencing and reducing the need for physical movement between displays.
  • The UnRealistic duplicate orientation (aligned with the user’s display) facilitated cross-referencing better than the Realistic orientation.
• Experiments / evaluation:
  • Study 1: Tested avatar types (Minimal vs. Volumetric) and orientations (Realistic vs. UnRealistic), with 12 participants performing tasks involving deictic gestures. Results showed higher comfort and efficiency with Minimal avatars.
  • Study 2: Compared three conditions (None, Realistic, UnRealistic) in a collaborative game inspired by the Pandemic board game, with 12 participants in dyads. DoubleMe enabled diverse collaborative strategies and improved task performance.
• Limitations and future work:
  • Small sample size (N=12) limits generalizability.
  • Current design supports only two users; extending to larger groups poses challenges in maintaining spatial consistency.
  • Future work could explore alternative avatar designs, support for heterogeneous tasks, and adaptation to remote collaboration.

Summary

DoubleMe introduces locally blended interaction spaces to address spatial constraints in co-located multi-display environments. By creating virtual duplicates of collaborators’ workspaces, DoubleMe enables efficient cross-referencing and diverse collaborative strategies. Empirical studies demonstrate that participants rely on duplicates for interaction, with minimalist avatars preferred for tasks involving deictic gestures. While the system shows promise in enhancing collaboration, future work is needed to scale to larger groups and explore broader applications.