Artificial Animals: Biomechanics, Locomotion, Perception, Behavior, Learning & Cognition in Simulated Physical Worlds

This talk presents research spanning the fields of artificial life, computer graphics, and computer vision. We have created physics-based virtual worlds inhabited by realistic ``artificial animals''. These sophisticated artificial life models possess muscle-actuated bodies, eyes, and brains with motor, perception, and behavior centers. Artificial animals are of interest in computer graphics in part because they are self-animating creatures that dramatically advance the state of the art of graphical character animation. As zoomimetic autonomous agents situated in realistic virtual worlds, artificial animals also foster a deeper understanding of biological information processing. In particular, they enable an advantageously fresh paradigm for computer vision research. Artificial animals are able to navigate complex, highly dynamic 3D environments using active vision systems that continuously analyze retinal streams imaged through mobile, foveated eyes. Artificial animals are furthermore valuable to the study of learning and cognition in living systems. They can learn muscle-actuated locomotion and are able to acquire some astonishing motor skills guided by sensory perception. Finally, knowledge representation formalisms from AI now support basic cognition in artificial animals. The talk will be richly illustrated with CG images and video.

BIO:

Demetri Terzopoulos (PhD '84 MIT) is Professor of Computer Science and Electrical & Computer Engineering at the University of Toronto, where he leads the Visual Modeling Group and is a Canada Council Killam Fellow. He also leads the Computer Graphics Animation research group at Intel Corporation in Santa Clara, CA. Prior to joining the University of Toronto in 1989, he was a program leader at Schlumberger corporate research centers in California and Texas. His published works include more than 200 scientific articles and several volumes, primarily in computer vision and graphics, and also in medical imaging, computer-aided design, artificial intelligence, and artificial life. Professor Terzopoulos' contributions have been recognized with awards from the IEEE, the American Association for Artificial Intelligence, the International Digital Media Foundation, Ars Electronica, NICOGRAPH (Japan), and the Canadian Image Processing and Pattern Recognition Society. He has received three of Canada's most prestigious research fellowships and five University of Toronto Excellence Awards. He has served on DARPA, NIH, and NSF advisory committees and was program co-chair of the 1998 IEEE Conference on Computer Vision and Pattern Recognition (CVPR'98).