Biometric Liveness Detection Guarding Against Spoofing Attacks using Machine Learning Techniques
Meghana Lokhande, Mrudula Khedkar, Nikita Desale, Mrunal Chopade, Pratiksha Ghodke · 2024
Biometric authentication is widely used for securing digital identities, but it faces vulnerabilities to various spoofing attacks. Iris recognition, in particular, requires robust liveness detection systems to ensure its security and reliability. This paper introduces an advanced iris-liveness detection system to differentiate between genuine iris traits and spoofing attempts. Existing liveness detection methods, often heuristic or rule-based, have limitations such as limited detection accuracy, susceptibility to evolving spoofing techniques, and operational complexity. Our proposed system leverages state-of-the-art deep neural networks and machine learning techniques to process iris scans in real time, enhancing the detection of live signals over static spoofing attempts. By integrating these advanced AI techniques, our system will significantly improve flexibility and resilience against sophisticated spoofing methods, ensuring higher accuracy and robustness in iris-based biometric authentication.