A greedy inverse kinematics algorithm for animating 3D signing avatars

The Inverse Kinematicss (IK) is a method for calculating the postures, in order to satisfy a given task by the user, based on degree of freedom of each joint. This method plays an important role in the animation and simulation of virtual characters to improve and facilitate the use of interfaces. In this paper, we present the description of an animation solver that allows the automatic moving of the different parts of a signing avatar body based on a greedy algorithm. The objective is to obtain a realistic animation of the virtual character that enables deaf person to visualize realistic gestures. Our approach is based on a thorough study of sign language and gestures classification. The challenge of this project is to find a good compromise between computational time and realistic representation that should be closer to real-time generation signs.