A Data-Driven Method for Crowd Simulation in Urban Scenes

This paper proposes a new data-driven method for crowd simulation,which can be used for adding scalable and complicated crowd animation into a virtual urban scene.By the method firstly,we capture several typical pedestrian movement videos,from which state-action data sets are extracted.The data sets are then integrated into an informed virtual urban model.And finally,a crowd animation synthesis system of pedestrian in the virtual urban is established based on the informed virtual urban model.