Facial animation is one of the key points in the realism of 3D scenes featuring humans. This is due to several reasons. First, face and the many muscles that compose it can generate a multitude of expressions. Then, our faculty of perception provides us a great ability to detect and analyze its smallest variations. This complexity is reflected in existing approaches by the fact that it is very difficult to create an animation without a long and tedious manual work.
This thesis aims to develop different techniques contributing to the process of creating facial animation:
- automatic configuration of a face for the 3D animation
- monocular capture without marker of face's motion
- animation transfer ( from 3D face to capture or from 3D face to 3D face)
- generation of skin's deformation during motion
The goal is to propose methods for realistic facial animations, esay to use, real-time oriented and automated maximum while leaving, if necessary, the possibility of a human/graphic designer to improve the outcome.