SPARC: Shared Perspective with Avatar Distortion for Remote Collaboration in VR

TL;DR

SPARC tackles the challenge of preserving deictic references and nonverbal cues in multi-user VR collaboration by introducing a shared-perspective arrangement where avatars are not co-located but positioned around a table. It maps references across different seats through environment rotations (eight 45-degree sectors) and spline-based avatar arm visualizations, enabling simultaneous viewing of the workspace from a common perspective while maintaining face-to-face cues. The authors implemented a Unity PoC with Photon networking and evaluated it in a within-subject study (n = 18) against a veridical perspective, using a Bedlam Cube assembly task with three roles. Results indicate SPARC reduces mental workload and improves movement economy, with mixed but generally favorable effects on task performance, suggesting a scalable approach to reduce occlusions and enhance nonverbal communication in multi-user VR collaboration. The work highlights practical implications for remote teamwork in VR, offering a method to support more complex, multi-user interactions without the clutter and occlusion typical of avatar-overlap approaches.

Abstract

Telepresence VR systems allow for face-to-face communication, promoting the feeling of presence and understanding of nonverbal cues. However, when discussing virtual 3D objects, limitations to presence and communication cause deictic gestures to lose meaning due to disparities in orientation. Current approaches use shared perspective, and avatar overlap to restore these references, which cause occlusions and discomfort that worsen when multiple users participate. We introduce a new approach to shared perspective in multi-user collaboration where the avatars are not co-located. Each person sees the others' avatars at their positions around the workspace while having a first-person view of the workspace. Whenever a user manipulates an object, others will see his/her arms stretching to reach that object in their perspective. SPARC combines a shared orientation and supports nonverbal communication, minimizing occlusions. We conducted a user study (n=18) to understand how the novel approach impacts task performance and workspace awareness. We found evidence that SPARC is more efficient and less mentally demanding than life-like settings.

Figures (7)

• Figure 1: Over-the-shoulder collaboration (top) is replaced by shared perspective with stretched avatars (bottom). While the perspective is shared, the avatar visualization is face-to-face, allowing for non-verbal cues.
• Figure 2: Avatar manipulation: [A] - User on the left points at a piece in his local workspace; [B] - User on the right sees this reference mapped to local own workspace
• Figure 3: Transformation applied to the avatar's arm and hand position: [A] - User A points at a point in his workspace, [B] - User B receives the information of where user A pointed and maps it to his own workspace, [C] - User A's opposing arm is rendered as a spline and their hand points at the mapped point in User B's workspace.
• Figure 4: Different conditions in the same POV from the perspective of Instructor 1: [A] and [B] represent the Veridical condition. [C] and [D] represent SPARC
• Figure 5: Raw data from the experiment depicting the three measures variation by trios for the two conditions. The lower, the better. Arrows indicate the direction of the difference for each group: red when veridical is higher and blue when SPARC is higher.