Abstract
Collaborative Virtual Reality (VR) systems increasingly rely on shared control mechanisms to enhance user coordination, but existing solutions often face trade-offs between individual autonomy and joint efficiency. This study builds on the foundational work of Zhou et al. (2024) on CoplayVR and PairPlayVR, aiming to optimize shared control by integrating adaptive authority allocation and real-time intent prediction. We conducted a user experiment with 32 participants, comparing the optimized mechanism against the baseline CoplayVR system across three collaborative tasks (object assembly, path navigation, and target selection). Results show that the optimized system reduces task completion time by 18.7% (p<0.05) and increases user satisfaction (SUS score: 82.3 vs. 71.5 for CoplayVR). Qualitative feedback further highlights improved perceived control and reduced coordination friction. This work advances shared control design for collaborative VR, complementing prior research on shared hand control and user experience in VR collaboration.
