Enhanced Face Anti-Spoofing Using a Hybrid Method: Combining Extendedlbp and Mobilenet

With the current era of the digital world, antispoofing technology plays a critical role in ensuring the integrity, security, and reliability of facial recognition systems. This paper introduces an advanced anti-spoofing system leveraging an innovative combination of Extended Local Binary Patterns (ExLBP) and the MobileNetV2 architecture to robustly detect and mitigate spoofing attempts. Texture features extracted using ExtendedLBP were further analysed in-depth by the lightweight MobileNetV2 CNN model, achieving exceptional performance in detecting spoofed faces. The experiment and evaluation carried on LCC FASD database reveals that the proposed approach attains an improved accuracy of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{93.59 \%}$</tex>, significantly outperforming other existing methods. The outcomes highlight how this hybrid model can strengthen facial recognition systems against the spoofing attempts, improving the security and dependability of biometric authentication procedures.