Robust Face Liveness Detection for Biometric Authentication using Single Image
Poulami Raha, Yeongnam Chae
TL;DR
This work tackles the vulnerability of Face Recognition Systems to 2D presentation attacks by introducing a lightweight, end-to-end CNN for robust face liveness detection. The model leverages background cues via cropped-face padding to improve discrimination between live faces and spoof artifacts, enabling 1-2 second CPU inference. A new 2D spoof dataset is released, comprising over 500 videos from 60 subjects across five attack types, and a web-based demonstrator showcases real-time liveness detection. The approach achieves strong practical impact for secure biometric authentication, including an ACER of $2.7\%$ on the authors' dataset and deployment-ready web demonstration.
Abstract
Biometric technologies are widely adopted in security, legal, and financial systems. Face recognition can authenticate a person based on the unique facial features such as shape and texture. However, recent works have demonstrated the vulnerability of Face Recognition Systems (FRS) towards presentation attacks. Using spoofing (aka.,presentation attacks), a malicious actor can get illegitimate access to secure systems. This paper proposes a novel light-weight CNN framework to identify print/display, video and wrap attacks. The proposed robust architecture provides seamless liveness detection ensuring faster biometric authentication (1-2 seconds on CPU). Further, this also presents a newly created 2D spoof attack dataset consisting of more than 500 videos collected from 60 subjects. To validate the effectiveness of this architecture, we provide a demonstration video depicting print/display, video and wrap attack detection approaches. The demo can be viewed in the following link: https://rak.box.com/s/m1uf31fn5amtjp4mkgf1huh4ykfeibaa