Authors:
Guilherme Silva
1
;
Pedro Silva
1
;
Mariana Mota
2
;
Eduardo Luz
1
and
Gladston Moreira
1
Affiliations:
1
Computing Department, Federal University of Ouro Preto (UFOP), MG, Brazil
;
2
Department of Control and Automation Engineering, UFOP, MG, Brazil
Keyword(s):
Spoofing, CNN, EfficientNets.
Abstract:
Spoofing detection, when differentiating illegitimate users from genuine ones, is a major problem for biometric systems and these techniques could be an enhancement in the industry. Nowadays iris recognition systems are very popular, once it is more precise for person authentication when compared to fingerprints and other biometric modalities. Nevertheless, iris recognition systems are vulnerable to spoofing via textured cosmetic contact lenses and techniques to avoid those attacks are imperative for a well system behavior and could be embedded. In this work, attention is centered on a three-class iris spoofing detection problem: textured/colored contact lenses, soft contact lenses, and no lenses. Our approach adapts the Inverted Bottleneck Convolution blocks from the EfficientNets to build deep image representation. Experiments are conducted in comparison with the literature on two public iris image databases for contact lens detection: Notre Dame and IIIT-Delhi. With transfer learn
ing, we surpass previous approaches in most of the cases for both databases with very promising results.
(More)