Paper Title
Particle Swam Optimization Based Identification of Face Images From Video

Abstract
To advance the performance of the existing face identification techniques, a new method is projected ,which is based on particle swarm optimization (PSO) with Support Vector Machine (SVM). The proposed video face recognition system performance is analyzed by taking number videos and compared with the existing face recognition techniques. PSO is utilized to optimize the parameters of the SVM. To accomplish the video face recognition with PSO-SVM, initially preprocessing process is performed on the video database by the filtering method and then the feature extraction process is carried on these preprocessed video face images by the AAM (Active Appearance Modeling) method. The extracted AAM based features from the feature extraction are stored in the feature database. These stored features are given to the SVM training and testing process. During the training process, the SVM parameters are simultaneously optimized by the PSO technique. The optimized parameters in SVM by PSO efficiently perform the video face recognition process. The proposed video face recognition system performance is analyzed by taking number videos and also the proposed video face recognition system is compared with the existing AAM based face recognition technique. Index Terms- Face Recognition, PSO, SVM, Adaptive Median filtering, AAM