Crowds (and research in animation and games) CSE 3541 Matt Boggus

Foundation of Digital Games See site for paper topics

CASA – computer animation and social agents Social Agents and Avatars Emotion and Personality Virtual Humans Autonomous Actors AI-based Animation Social and Conversational Agents Inter-Agent Communication Social Behavior Crowd Simulation Understanding Human Activity Memory and Long-term Interaction, etc.

SCA – symposium on computer animation 2D, 3D, and N-D animation systems autonomous characters clothing animation and simulation expressive motion / communication facial animation group and crowd behavior intuitive interfaces for creating and editing animations mathematical foundations of animation methods of control and artistic direction of simulations nature in motion (natural phenomena, plants, clouds,...) new time-based art forms on the computer novel time-varying phenomena perceptual metrics for animation perceptual foundations of animation physical realism / measuring the real world for animation physical simulation fluid animation planning / learning / optimization for animation real-time and interactive methods camera control methods for computer animation sound and speech for animation

I3D – symposium on interactive 3D graphics and games Interaction devices and techniques 3D game techniques Interactive modeling Level-of-detail approaches Pre-computed lighting Visibility computation Real-time surface shading Fast shadows, caustics and reflections Imposters and image-based techniques Animated models GPU techniques Navigation methods Interactive visualization Virtual and augmented reality User studies of interactive techniques and applications Sketch based 3D interaction

Siggraph See

Crowd modeling applications Entertainment: Games Computer animation Art Evaluation: Architecture Robotics Training: Virtual reality simulation

Individuals and crowds Individuals are agents – Reactive vs. planning – Goal vs. need driven Groups – set of similar agents – Spatially close – Like minded (butter-side up or butter-side down) Crowds – many individuals, with or without groups – Emergent behavior – similar to flocking, flocking system – Uniformity – sameness of members – Quantity & density – average distance between members

Execution environment Real-time / Interactive – Simple computations – Avoid n 2 or higher algorithms – Limit size Off-line – Can use complex models for behavior – Can allow interaction between all agents (n 2 ) – Size limited only by hardware memory and storage

Example 1 Autonomous Pedestrians Emulating real pedestrians in urban environments Motions controlled at different levels – Reactive behaviors – Navigational and motivational behaviors – Cognitive behaviors Information stored in mental states

Example 2 Autonomous Pedestrians Controlling Agents in High-Density Crowd Simulation fA&feature=related

Spatial organization Cellular decomposition: Regular 2D grid Adjacency accessible Density limited Cells define obstructions Continuous space: Step in any direction Need to decipher obstructions Perception needed

Navigation Rule-based Fluid flow: density fields, potential functions Cognitive modeling Flocking systems: individual perception, navigation Particle systems: Individual navigation Cellular automata

Panic & Congestion handling Personal space Packing people during evacuation Exit awareness

Motion & Navigation Potential fields Path planning Passing on pathways Roadmaps Forming & maintaining subgroups

Recent papers on crowd simulation Simulating Heterogeneous Crowd Behaviors Using Personality Trait Theory Environment-aware Real-time Crowd Control ontrol.html ontrol.html A Synthetic-Vision-Based Steering Approach for Crowd Simulation (scroll down to find paper and video)

Commercial crowd simulation MASSIVE What is Massive? elated elated LQ1 LQ2 elmfu elmfu