With or Without You: Effect of Contextual and Responsive Crowds on VR-based Crowd Motion Capture

Area of Science:

• Human-computer interaction
• Computer graphics
• Social simulation

Background:

• Capturing real human crowd motion data is challenging.
• Virtual Reality (VR) offers potential for crowd data acquisition through simulated or motion-capture-based crowds.
• Existing methods like Record-and-Replay (2R) have limitations impacting data quality.

Purpose of the Study:

• To introduce the concept of contextual crowds for more consistent crowd data.
• To evaluate two novel paradigms: Replace-Record-Replay (3R) and Replace-Record-Replay-Responsive (4R).
• To enhance the naturalness and consistency of captured crowd datasets in VR.

Main Methods:

• Development and implementation of 3R and 4R paradigms within a VR environment.
• Immersing users into simulated crowds that are progressively replaced by user-captured motion (3R).
• Endowing pre-recorded agents with responsive capabilities in the 4R paradigm.
• Evaluation using two real-world-based scenarios replicated in VR.

Main Results:

• The 3R and 4R paradigms significantly improve the consistency of captured crowd datasets.
• User behaviors observed in VR were more natural when surrounded by agents from the start.
• Contextual crowds effectively leverage both simulation and motion capture for better data.

Conclusions:

• The proposed 3R and 4R paradigms offer a more effective approach to capturing consistent crowd motion data.
• Integrating simulated agents and user motion capture in VR enhances data realism.
• Contextual crowds represent a promising advancement for crowd behavior research and simulation.