Hardware Design on Face Recognition System by Deep Neural Network for Access Control System

Based on the advance in GPU and big data, deep learning has brought progress, especially in computer vision. But the acceleration by GPU is not cost-effective on the dedicated design. This paper implements a face recognition system with the deep learning technique as an access control system. We accelerate the algorithm and design it with the combination of FPGA and ARM to realize an access control system. The recognition system can only be executed when there is an object near the camera, and only compute the detected face.