The One-Man-Crowd: Single User Generation of Crowd Motions Using Virtual Reality

Area of Science:

• Computer Science
• Human-Computer Interaction
• Simulation

Background:

• Realistic crowd motion data is crucial for understanding and simulating crowd behaviors.
• Acquiring such data from real-world experiments is challenging due to ethical concerns, logistical difficulties, and limited scenario coverage.

Purpose of the Study:

• To propose and evaluate a novel Virtual Reality (VR) based approach for acquiring crowd motion data.
• To overcome the limitations of real-world data collection methods.

Main Methods:

• A single user is immersed in virtual scenarios, successively embodying each crowd member.
• User's past trajectories and body movements are recorded and displayed on virtual characters.
• Crowd behavior is progressively built by the user acting as individual members.

Main Results:

• The feasibility of the VR approach was validated by replicating three real-world experiments.
• Emergent phenomena and individual interactions in the generated virtual crowd data were compared to real datasets.
• Realistic collective behaviors emerged from the virtual crowd data, though with less behavioral variety than real situations.

Conclusions:

• The proposed VR approach offers a viable method for generating realistic crowd motion data.
• The approach provides valuable insights for building virtual crowd experiences.
• Further improvements can focus on increasing the variety of behaviors in virtual crowd simulations.