Active Morphable Model: an efficient method for face analysis

Multidimensional Morphable Model is a powerful model to analyze and synthesize human faces. However, the stochastic gradient descent algorithm adopted to match the Morphable Model to a novel face image is not efficient enough. In this paper, a very efficient optimization method devised for Morphable Model matching is proposed, called Active Morphable Model (AMM). The kernel of AMM is an iterative algorithm directly utilizing the heuristic information provided by the novel image, and updating the model parameters in a computationally economic fashion. AMM is more efficient than general optimization methods in matching a Morphable Model, it has much higher convergent rate and matching speed. Furthermore, it is insensitive to the initial estimation of the face pose, and is robust when used to match novel faces with large variations in translation, rotation and scaling. Experimental results are given to validate the efficiency and robustness of the proposed method.

[1]  David Beymer,et al.  Vectorizing Face Images by Interleaving Shape and Texture Computations , 1995 .

[2]  Thomas Vetter,et al.  A morphable model for the synthesis of 3D faces , 1999, SIGGRAPH.

[3]  Michael Jones,et al.  Multidimensional Morphable Models: A Framework for Representing and Matching Object Classes , 2004, International Journal of Computer Vision.

[4]  Tomaso A. Poggio,et al.  Multidimensional morphable models , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[5]  Tomaso A. Poggio,et al.  A bootstrapping algorithm for learning linear models of object classes , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[6]  Stan Z. Li,et al.  Direct appearance models , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[7]  Timothy F. Cootes,et al.  Statistical models of appearance for medical image analysis and computer vision , 2001, SPIE Medical Imaging.

[8]  Timothy F. Cootes,et al.  Active Shape Models-Their Training and Application , 1995, Comput. Vis. Image Underst..

[9]  P. Anandan,et al.  Hierarchical Model-Based Motion Estimation , 1992, ECCV.

[10]  Michael J. Jones,et al.  Gaze estimation using morphable models , 1998, Proceedings Third IEEE International Conference on Automatic Face and Gesture Recognition.

[11]  Tomaso Poggio,et al.  Trainable Videorealistic Speech Animation , 2004, FGR.

[12]  Timothy F. Cootes,et al.  Statistical models of appearance for computer vision , 1999 .

[13]  Sami Romdhani,et al.  Face Identification by Fitting a 3D Morphable Model Using Linear Shape and Texture Error Functions , 2002, ECCV.

[14]  Takeo Kanade,et al.  An Iterative Image Registration Technique with an Application to Stereo Vision , 1981, IJCAI.

[15]  Timothy F. Cootes,et al.  Active Appearance Models , 2001, IEEE Trans. Pattern Anal. Mach. Intell..