Physically-based Animations of 3D Biped Characters Using Genetic Algorithms
Maurizio Conventi and Marco Fratarcangeli
The 10th Congress of the Italian Association for Artificial Intelligence - AI for Expressive Media (AIIA-EM 2007)
Roma, Italy, September 10-13, 2007
Abstract
Synthesizing the realistic motion of a humanoid is a challenging task, studied in different research areas. This paper addresses the problem of synthesizing realistic animations of 3D biped characters in a physics simulated environment, using real coded genetic algorithms. Characters are represented by a structure of rigid bodies linked each other with 1DOF joints. Such joints are controlled by sinusoidal functions whose parameters are calculated by the genetic algorithm. Our system allows for automatic synthesis of physically plausible animations by setting only few key-poses of the characters, thus it is suitable for not skilled users. We present results obtained by testing and comparing several different genetic operators.