A Real-Time Approach to Facial Anti-Spoofing using Liveness Detection on CPU Bound Devices
Rajni Jindal, Tanuj Chandolia, Tanmaya Yadav · 2022 6th International Conference on Intelligent Computing and Control Systems (ICICCS) · 2022
With the fast pacing and advancements in technology, especially in the fields relating to computer science and engineering as well as electronic engineering, facial recognition has become one of the most widely deployed authentication and authorization method in the field of biometric security and identity, placing at second with fingerprint biometric authentication security coming at first. Most of the research in facial recognition has had the main aim of improving the performance of the algorithm or the systems for faster and efficient completion of identification and verification tasks. Much less light was thrown in the direction of security vulnerabilities of these systems. Even though many techniques have been incorporated to solve this problem, they require high memory usage as well as processing power to work, which makes them less useful for real life scenarios where there may not be enough resources. FaceBoxes and 3DDFA have been incorporated to immensely increase processing speeds on CPU Bound Devices.