The Hierarchical Perception Model for Crowd Simulation

This paper proposes a hierarchical perception model for simulation of crowd perception in different density. In this model, there are two tiers: the low density and high density perception modules. They are both composed of perceptual filters, short-term and long-term memory. However, considering the changes of the perception as the density increases, we adopt different methods to model them. We also take into account the impact of the human subjective factors on the selection of the object the virtual agent focuses. Our hierarchical perception model can well reflect the characteristics of the human perception; moreover, it can provide more plausible information for the virtual human among the crowd to respond properly to certain event.