Face Reconstruction with Low Resolution Facial Images by Feature Vector Projection in Kernel Space

In spite of increasing interest in person identification based on biometrics, face recognition technology has not been applied into real world. It is caused by appearance changes such as illumination, noise, degradation, and occlusion. Among these problems, we focus on the low resolution problem and propose a new face recognition method of extending the SVDD (support vector data description). In the proposed method, we first solve the SVDD problem for the data belonging to the given prototype facial images, and model the data region for the normal faces as the ball resulting from the SVDD problem. Next, for each input facial image in low resolution, we project its feature vector onto the decision boundary of the SVDD ball so that it can be tailored enough to belong to the normal region. Finally, we synthesize facial images which are obtained from the pre-image of the projection, and then perform the face recognition. The applicability of the proposed method is illustrated via some experiments using general recognition algorithm