Improved Kernel Common Vector Method for Face Recognition

The common vector (CV) method is a linear subspace classifier for datasets, such as those arising in image and word recognition. In this approach, a class subspace is modeled from the common features of all samples in the corresponding class. Since the class subspace are modeled as a separate subspace for each class in feature domain, there is overlapping between these subspaces and there is loss of information in the common vector of a class which in turn reduces the recognition performance. In CV method the followed criterion considers only the class scatter matrices. Thus the neglecting of the influence of neighboring classes in CV method also reduces the recognition performance. Generally a linear subspace classifier fails to extract the non-linear features of samples which describe the complexity of face image due to illumination, facial expressions and pose variations. In this paper, we propose a new method called ¿improved kernel common vector method¿ which solves the above problems by means of its appealing properties. First, the introduced between-class and within-class scatter matrices consider the neighboring classes and covariance of a class and this makes the obtained common vector has more discriminant information which increases the recognition performance. Second like all kernel methods, it handles non-linearity in a disciplined manner which extracts the non-linear features of samples representing the complexity of face images. Experimental results on real time face database demonstrate the promising performance of the proposed methodology.