Appearance model based face-to-face transform

In this paper, a novel approach to the face-to-face transform is presented. The face-to-face transform is a technique which transforms one person´s facial actions to the others. In general, 3D models of faces are used for such transformation. Therefore, the facial action parameters must be estimated from the 2D input images, which is not an easy task. On the contrary, our proposed approach is based on the 2D appearance model, instead of the 3D model, so that the model is acquired by learning directly from training images. To achieve this, we investigate making use of the hidden Markov model (HMM) framework, which models the correspondence between an input face and the other one as well as the appearances of both faces. The experimental results show the effectiveness of the proposed method.