Face Liveness and Anti-Spoofing Detection using CNN

Face recognition is a very popular biometric method that has developed quickly lately. It is much easier, more user-friendly and convenient than most of the methods used. However, these systems can be spoofed using fabricated faces to the system. To increase security measures, liveness detection needs to be integrated to prevent the spoofing attempts. Our work researches applying Convolutional Neural Networks (CNNs) and deep learning techniques to enhance face liveness detection, enhancing the robustness of biometric authentication systems. A comprehensive review of existing methods is presented, a novel architecture particularly designed for liveness detection is proposed, and its performance is evaluated on a diverse dataset. Our findings suggest that this proposed machine-learning system was in addition highly accurate and exhibited low false acceptance rates compared to traditional approaches for fighting the problem of spoofing.