Face Recognition Using Cortex Mechanism and SVM

This paper introduces a novel classified method for face recognition, which is based on cortex mechanism and support vector machine (SVM). It describes a feature extracting system that closely follows the organization of visual cortex and builds a complex and invariant feature representation by alternating between it template matching and a maximum pooling operation, and then the classifier adopts the binary tree structure in the training stage and the test stage. The experiment results demonstrate the capabilities of the approach and insensitivity to the light illumination and the scale of the face image, so it has good recognition rate for face recognition in real-time scenery.