Date of Publication :7th March 2016

Abstract: The liability of face recognition systems in biometrics is a growing concern today, as still it remains vulnerable to various sophisticated attacks that undetermined the reliability of biometric systems. In this paper, we present a novel approach to accurately detect and mitigate the Spoofy attacks on the face by introducing light field camera (LFC), also known as plenoptic camera. Since the use of a LFC can record the direction of each incoming ray in addition to the intensity, it also exhibits an unique characteristic of rendering multiple depth (or focus) images in a single capture, also known as refocusing, which provides the high quality artefact face features. Introducing a novel idea of exploring the inherent characteristics of Light Field Camera to detect spoof attacks by estimating the variation of the focus between multiple depth images. To this extent, we first collect a new biometric face artefact database using LFC. We then generate the face artefacts samples by simulating three different kinds of spoof effects including photo print and electronic screen attacks. Extensive experiments carried out on the light field face artifact database have revealed the outstanding performance of the proposed anti spoofing scheme when benchmarked with various well established stateof-the-art schemes.

Reference :