Multiresolution Modeling for Real-Time Facial Animation

This paper proposes the generation scheme of multiresolution mesh for real-time facial animation. Since previous works consider only the geometry of facial model, during generating a multiresolution facial model they may lose the feature points of facial animation. But the proposed scheme can automatically preserve those features, so that the quality of facial animation with a lower-level model is as good as that of the original. In this paper we prove that the multiresolution facial model generated by our scheme is effective in facial animation by applying an expression-retargeting technique to it