Deep Learning For Face Spoofing and Liveness Detection

Abstract

Pandemic situation turns every ones lives in online mode even classes were conducting through virtual meeting sessions. But for most of the meetings duplication or spoofing of faces was traced even for some secured meetings if they got meeting link they used get connected. But our system helps, authenticated users only can login and the rest couldn't make it possible. In this paper, face recognition based identification approach used to provide authentication for the user there is possibility of face spoofing therefore the early stage of the process is to verify the face is real or fake then adjust the database consists of the registered one or not. To handle this problem, face cascaded feature were calculated and using deep learning approach a solution was finalized.