Facial modeling can be used for a variety of applications, such as face recognition, face tracking, age detection, expression recognition or facial animation. In this thesis, a 3D facial model, which is also able to synthesize facial expressions, is constructed with faces from a 3D face database. Therefore, dense point-to-point correspondences between the 3D faces are computed and then a Principal Component Analysis is applied to a set of faces of the database. The functionality of the facial model is demonstrated by means of two applications. The facial movements and the micro-movements within the faces of probands are tracked in multi-camera video sequences. Furthermore, the facial model is employed for animation purposes. Faces are replaced in video sequences, in which probands are talking and carrying out facial expressions.
«
Facial modeling can be used for a variety of applications, such as face recognition, face tracking, age detection, expression recognition or facial animation. In this thesis, a 3D facial model, which is also able to synthesize facial expressions, is constructed with faces from a 3D face database. Therefore, dense point-to-point correspondences between the 3D faces are computed and then a Principal Component Analysis is applied to a set of faces of the database. The functionality of the facial mo...
»