GGA-AAM: Novel heuristic method of gradient driven Genetic Algorithm for Active Appearance Models

This paper proposes an optimization technique of gradient based genetic algorithm (GGA) to make a robust, efficient and real time face alignment application for embedded systems. It uses 2.5D Active Appearance Model (AAM) for the face search, the model is generated by taking 3D landmarks and 2D texture of the face image. 3D face alignment requires to optimize 6DOF (Degrees of Freedom) pose and appearance parameters of AAM. These parameters span in a huge face search space. Genetic Algorithm (GA) searches face globally whereas gradient descent helps GA to search face locally. In other words exploitation property of gradient descent and exploration property of GA are combined to make an efficient optimization system. For this combination we propose a new gradient operator in GA, which functions in conjunction with the existing operator of mutation in GA. Thus it does not increase the computational cost of the system. We compare it with classical search algorithm by gradient descent. Face poses of SUPELECpsila08 and Pointingpsila04 databases are estimated and results validate the efficiency, accuracy and robustness achieved.