Out of context augmented navfields: Designing crowd choreographies

Abstract

This paper presents a way to dynamically influence the shape and movements of a simulated crowd. We propose a tool and system that allows to modify a crowd's dynamics in an intuitive, semantically rich and out of context fashion, while being independent from the global path finding architecture and having a low computational cost. We follow a mixed approach where user-specified navigation fields are combined with steering and global A*pathfinding. © 2012 Springer-Verlag Berlin Heidelberg.