Deep Convolutional Neural Network for Real and Fake Face Discrimination

With the progress of society, and the rapid development of science and technology, information identification security issues have become more important. With the use of technologies such as Generative Adversarial Networks (GAN) to generate the generalization of faces, in order to make the face identification system more secure, it is necessary to detect the fake face. With the maturity of artificial intelligence technology, face identification technology is widely used in various fields of real life, especially in the research of identifying computer generated faces. This paper is aimed at discriminating the generation of face problems. An algorithm based on Deep Convolutional Neural Network (DCNN) for neural network architecture for face image style classification is proposed. And provide experimental basis. The results show that the method can obtain satisfactory results, and the average accuracy is above 99.24%.