Very small image face recognition using deep convolution neural networks

Abstract

The very small size of face image recognition is one of a significant case in computer vision because it can be applied in many applications in the real world. The main problem of face recognition with a small size is that the area of the acquired face is very small so that many facial features cannot be obtained. In this paper, we propose a novel method using deep convolution neural networks (DCNNs). First the image resolution is enhanced using CNN and then it is classified using the backpropagation algorithm. To verify the proposed method, we use AR Database. The results show that the identification accuracy is very good but still need a larger database for further verification.