The aim of this work is to create a model generic that is able to improve tracking and synthetising by modeling behaviours of deformable objects such as human faces. The model will be an extension of the active appearance model (AAM) designed by Cootes, Edwards and Taylor [15]. The AAM is a probabilistic model used to describe an object in a still image by learning its shape and appearance from a database of hand labelled images of similar objects. The AAM is described in more details in section 3 of this report.
We would like to apply the model in the study of interaction between persons during conversations in order to synthetise one person as a response to the tracking of the other person. For instance, we would expect a virtual person closing his mouth when the tracked person is speaking or the virtual person may smile when the tracked person is smiling.