A Survey on the Latest Development and Typical Methods of Face Anti-Spoofing

Identity authentication based on face recognition is widely applied because of its simplicity, convenience and non-contact nature. However, face recognition systems in real-world scenarios are subject to malicious attacks, which severely threatens national security and people’s daily life. It is therefore urgent to explore counter measures of face spoofing. This survey collects 68 classical and typical research articles among which over 50 ones were published in the major international academic journals and conferences in the past two years. We organize those articles according to their technical motivation into 7 classes. More information is given to some typical works so that readers may gain some insight into the latest trends in this area. We further summarize each class. After reviewing the methods, we make a brief introduction to 6 face spoofing detection databases released since 2019, which contributes to a better understanding of the development and validation platforms of face spoofing detection. In the last section, we conclude the relations among the 7 classes, describe their general status in theoretical research and industrial applications, exhibit the major challenges and point out the research directions.