DeepFake Detection with 3D Face Perception

With the development of deep learning technology, the emergence of DeepFake technology makes fake video and image become more real, it brought many safe hidden trouble to the society. In this paper, we propose a DeepFake detection method based on 3D face perception, which aims to effectively identify fake face images. By generating and detecting facial features, we use a binary classifier to discriminate the generated weights. In addition, combining Generative adversarial Network (GAN) and Convolutional Neural Network (CNN) technology, we designed a feature detection and extraction mechanism to improve the accuracy and robustness of detection. The experimental results show that the proposed method has significant performance advantages in DeepFake detection task.