Face Recognition Based on Hierarchical Model and X (HMAX)

The Face Detection System is a biometric system which applies smart automatic methods to detect and/or verify a person’s identity based on physiological features. The current study aims to use the improved HMAX model for face recognition. HMAX is a biological model inspired by the human vision system. Hence, to improve the function of HMAX model we used learning automata as it has free parameters of Alpha and Beta. Learning automata is able to predict in uncertain environments and is applied to increase the rate of human face recognition.  In this study used the standard FEI dataset as the input of the proposed model which incorporates 200 photos of Brazilian people. When the photos are read by the MATLAB software commands, they enter the phase of feature extract which is done through HMAX model filters. To measure the rate of face detection, all the extracted characteristics are categorized. The HMAX model parameters are determined through learning automata. HMAX is a hierarchical model with a four-layered system of C2,S2,C1,S1 for recognizing the fine features of photos. Moreover, we compared the improved HMAX model with the Genetic algorithm to demonstrate the efficiency of the proposed model. The results of dataset analyses show a 94.08 percent of face detection. So, we conclude that the face detection rate in the improved HMAX is more than the Genetic algorithm.