On Streams and Incentives: A Synthesis of Individual and Collective Crowd Motion

Abstract

We present a novel crowd simulation model that combines the advantages of agent-based and ow-based paradigms while only relying on local information. Our model can handle arbitrary and dynamically changing crowd densities, and it enables agents to gradually interpolate between individual and coordinated behavior. This interpolation is based on a dynamically changing incentive value for each agent. The incentive value reflects an agent's willingness to coordinate with the crowd. A central new concept in our model is the notion of a stream of agents. Our model lets agents automatically form multiple streams with nearby agents as the density of the crowd increases. Our model can be used with any existing global path planning and local collision-avoidance method. Experiments show that our model yields energy-efficient and visually convincing crowd behavior for high-density scenarios while maintaining individual agent behavior at lower densities.