Behavioral Geography for Crowd Animation and Computational Social Science

Monday, October 26, 2009

Abstract

The film and television industry has, for some time, made use of computer animation to script and render synthetic characters as stand-ins for actors on movie sets. Traditionally, development of these characters has focused on visual acuity, with their behavior limited to quick bursts of choreographed movements and gestures. More recently, there has been a move toward replacing whole crowds of actors and stand-ins digitally and the number of characters in effects sequences, as well as audience and directors’ expectations for their realism, has led animators to automatic generation of actor proxies to replace manual posing and scripting in production pipelines. The special effects industry has turned to agent-based modeling in response to these growing demands for added functionality and the technology is now used for automatically generating behaviorally-plausible synthetic characters. The fidelity of animated crowds is most acute when the models driving them reproduce the geography of their behaviors appropriately by placing people in the right places, at the right times, doing the right things. Much of the built-in realism ascribed to special effects agents relates to their behavioral geography, particularly in representing the intricacies of how agents think about their surroundings, how they interact physically with their environment, how they plan their activity, and how they choreograph movement through built and social infrastructure based on that information. This associates animation with computational social science. The heuristics classically used to achieve agent behaviors in special effects sequences are quite simplistic and abstract and special effects technology could benefit from the infusion of realistic behavioral geography into agent design. Similarly, academic experimentation with agent-based modeling in computational social science could profit from the infusion of technologies from the special effects industry, particularly in adopting schemes for representing agent characters in greater detail, in distributing the computation of large volumes of interactive agents, and in animating agent dynamics in simulation. In this talk, I will introduce the work that I have been pursuing in developing realistic behavioral geography for agents and in connecting those agents to technologies used in special effects pipelines. I will demonstrate the usefulness of this approach in simulating academic scenarios related to human movement and crowd formation in dense urban environments.

Bio

Dr. Paul M. Torrens is an Associate Professor in the School of Geographical Sciences and Urban Planning at Arizona State University and Director of its Geosimulation Research Laboratory. He is also affiliated with the University’s Center for Social Dynamics and Complexity and the GeoDa Center for Geospatial Analysis and Computation. Paul’s work has featured in a diverse array of outlets, from traditional academic publications to popular media such as Vanity Fair, Popular Mechanics, Forbes, and Discover Magazine. His projects have been supported by the U.K. Economic and Social Research Council, the U.S. National Science Foundation, the Herberger Foundation, Science Foundation Arizona, Autodesk, Inc., and Alias Research. His work earned him a Faculty Early Career Development Award from the U.S. National Science Foundation in 2007 and he was awarded the Presidential Early Career Award for Scientists and Engineers by President George W. Bush in 2008. The Presidential Early Career Award is the highest honor that the U.S. government bestows upon young scientists; Paul is the first geographer to receive the Award.