Research And Design On Multi-Factor Enabled Facial Liveness Detection Systems

Face authentication becomes popular to protect personal information systems and electronic payments,which is a promising replacement of legacy password based authentication method due to its memory-less authentication process.However,face authentication is vulnerable to spoofing attacks.Spoofing attack means that an adversary deceives the authentication system by using the victim's photos or videos.There are lots of facial recognition applications widely used in real world,such as Construction Bank's face registration,Alipay's face payment,face check-in and so on.Well-known domestic and oversea companies like Google,Microsoft,Ali,Keda Xunfei,and Face++,have already made significant achievements in facial recognition by improving the recognition rate and robustness,which requires additional powerful hardware support.The face recognition system provided plenty of methods earlier,but it is vulnerable to the attacks based on photos,videos,and 3D virtual face.In this article,we propose a multi-factor enabled liveness detection system based on human face and human-computer interaction to enhance the security of face authentication.There are many detection methods,such as opening mouth,blinking and shaking to the real-time detection based on eye gaze,the analysis of facial expressions and the latest changes in key points to defend against the attacks based on photos,videos and3 D virtual face.Thus,we establish a corresponding method set and apply the state-of-the-art method to generate a secure random number,which can be called in real time for the method set so as to give the user a challenge,and detect the response.The response is analyzed to achieve the purpose of liveness detection.There are numbers of methods for random number generation,from random sequences,pseudo-random algorithms,to the latest quantum random numbers,which enhance the robustness of the system.As the new methods emerge,we have more candidate liveness detection factors to be used as challenges.The more randomized the system is,the more resistant to photos,videos,and 3D face attacks the liveness detection is.In this paper,we design a secure,robust,and practical multi-factor enabled liveness detection system to resist spoofing attacks and meet the needs of multiple platforms,especially mobile terminal platforms.Through the challenge-response liveness detection method,it realizes a portable detection system that is easy to transplant and it can resist attacks based on video and 3D virtual face.On the terminal equipmentwith superior performance,powerful computing capability and fast feedback make human-computer interaction more convenient and make the system more complete,which is more suitable for the development of terminal equipment.