FlagThis

Abstract

Biometric systems are increasingly deployed in security applications; however, they remain vulnerable to spoofing attacks, wherein attackers exploit counterfeit biometric data to gain unauthorized access. This research evaluates the effectiveness of state-of-the-art machine learning modelsMobileNetV2, DenseNet-121, Inception-v3, and Spoof Trace Disentanglement (STD)in detecting spoofing attacks within facial recognition systems. Using the CelebA-Spoof dataset, the study evaluates model performance using metrics such as accuracy, precision, recall, and F1 Score. Cross-dataset validation is performed on the MSU-MFSD dataset to assess generalizability. The results reveal MobileNetV2 as the most efficient model, achieving 92% accuracy while balancing computational efficiency, making it suitable for real-world applications. Inception-v3 shows moderate robustness, while DenseNet121 and STD experience challenges with generalization. The findings emphasize the need for advancements in domain adaptation and hybrid architectures to enhance biometric security systems.