(published in Computer Animation '95 Proceedings, IEEE, 1995)

Disney Meets Darwin
The Evolution of Funny Animated Figures


JJ Ventrella
Visible Language Workshop
MIT Media Laboratory
20 Ames Street, Cambridge, MA 02139
ventrell@media.mit.edu


Abstract

In this paper I discuss an approach to the generation of entertaining animated motion, using a genetic algorithm. Genetic algorithms have been used successfully to optimize the physical behaviors of articulated stick figures and other animated creatures for goal-directed behavior. But an emphasis has not been put on the "optimization of expressivity", whereby an animator inserts him/herself into the optimizing loop to aesthetically influence the evolution of motion behavior in such figures. A technique for bringing both automatic and interactive evolution together into one tool is discussed as a means to bring evolutionary tools closer to the concerns of the character animator, who may be just as interested in developing amusing behaviors in a world of "Cartoon Laws", as in simulating realistic animals in a world of Newtonian physics.


1. Introduction

2. The Art of Motion

3. Background

4. The Figures

5. Physics

6. Muscles

7. Genotype/Phenotype

8. Evolution

9. The Biological Clock

10. Results

11. Discussion

12. Improvements

13. Summary

14. Acknowledgments/References