Face Recognition Algorithm Based on Convolutional Neural Network

Abstract

In response to the needs of modern enterprises for intelligent security of face authentication, a face recognition algorithm based on deep convolutional neural network is introduced in this paper. In order to improve the accuracy of face recognition, the design of 11 convolutional layers and 4 pooling layers. The network uses the standard face data set CASIA-WebFace for training, and the face recognition accuracy on the LTW database can reach more than 97.8%. Based on this deep learning network, a face recognition management system was designed. The system realized 1: 1 face authentication and 1: N face recognition. After applying the convolutional neural network algorithm, the face search time is significantly reduced, indicating that the algorithm has higher efficiency and practicability.