Abstract:
With the rapid spread of deep network generated fake face technology, criminals fake face images and videos to implement telecommunications fraud, manipulate public opinion, disseminate obscenity and other purposes. How to efficiently and accurately detect fake faces from massive data has become a research focus. In this paper, we systematically summarize, analyze and compare the current deep network generative forgery face detection methods from two fields: generative forgery face image and generative forgery face video. For forged face images, this paper introduces the detection methods in detail from five categories: based on digital image processing foundation, deep feature extraction, spatial domain feature analysis, multi-feature fusion analysis, and fingerprint detection. In addition, the detection methods of fake face videos are discussed from four categories: physiological signals, identity information, multi-modal and spatio-temporal inconsistency. The research summary of this paper is of great significance for preventing the spread of false information, protecting personal privacy and maintaining social justice.