Facial expression recognition with enhanced feature extraction using PSO & EBPNN

Human face-to-face communication plays an important role in human communication and interaction. In recent years, several different approaches have been proposed for developing methods of automatic facial expression analysis. In this paper we have proposed a novel facial expression recognition system which chooses the optimized features using particle swarm optimization (PSO) from the features calculated from principle component analysis (PCA) of input face images. These optimized features are then used to train the emotional backpropagation neural network (EBPNN). Using this neural network classifier, emotions are classified. The proposed architecture yields good results when PSO features compared with normal PCA features.