Multi-Spectral Convolutional Neural Networks for Biometric Presentation Attack Detection
Nowadays, reliable and automatic subject authentication has become of the utmost importance in multiple application scenarios. Over the last decades, biometric recognition has shown to be a good alternative to password based systems. In spite of their numerous advantages, biometric systems are vulnerable to presentation attacks (PAs), i.e., attempts to log into the system with a fake biometric characteristic or presentation attack instrument (PAI). These attacks pose a severe threat to the security of the authentication system: any person could eventually fabricate or order a gummy nger to impersonate someone else. Therefore, the development of accurate presentation attack detection (PAD) schemes is key to the wider deployment of secure biometric systems. In this paper, we present a novel approach for ngerprint PAD based on short wave infrared (SWIR) images and multi-spectral convolutional neural networks (CNNs). In particular, four samples are acquired at dierent SWIR wavelengths, which are subsequently fed to ve different CNN models. These networks rst pre-process the multi-spectral information to obtain images
with three channels, and then apply regular CNN models. The approach is evaluated on a database comprising over 4700 samples, stemming from 562 different subjects and35 different PAI species. The results show the soundness of the proposed approach with a detection equal error rate (D-EER) as low as 0.5%, outperforming the state-of-the-art D-EER of 1.4%. In addition, fusing the SWIR information with laser speckle contrast
imaging (LSCI) sequences leads to an even lower D-EER of 0.2%.