Deep convolutional neural networks for face and iris presentation attack detection: survey and case study

Abstract

Biometric presentation attack detection (PAD) is gaining increasing attention. Users of mobile devices find it more convenient to unlock their smart applications with finger, face, or iris recognition instead of passwords. In this study, the authors survey the approaches presented in the recent literature to detect face and iris presentation attacks. Specifically, they investigate the effectiveness of fine-tuning very deep convolutional neural networks to the task of face and iris antispoofing. They compare two different fine-tuning approaches on six publicly available benchmark datasets. Results show the effectiveness of these deep models in learning discriminative features that can tell apart real from fake biometric images with a very low error rate. Cross-dataset evaluation on face PAD showed better generalisation than state-of-the-art. They also performed cross-dataset testing on iris PAD datasets in terms of equal error rate, which was not reported in the literature before. Additionally, they propose the use of a single deep network trained to detect both face and iris attacks. They have not noticed accuracy degradation compared to networks trained for only one biometric separately. Finally, they analysed the learned features by the network, in correlation with the image frequency components, to justify its prediction decision.

EnhanceDeepIris Model for Iris Recognition Applications

“…ignoring the parts beyond the boundary, the corresponding backpropagation calculation is obtained as shown in equation (7).…”

“…The results show that the model is able to accurately recognize minute patterns in the iris region [6].Y Safaa El-Din and other scholars summarized the methods for detecting face and iris and compared two different fine-tuning methods on six different datasets. The results showed that the existing DL models were able to discriminate different face and iris features with low error rates [7]. E Garea-Llano and other researchers proposed a YOLO network with image segmentation IR technique in order to effectively minimize the direct contact between human body and objects.…”

Fingerprint Recognition Using a Transfer Learning Method

“…The With the evolution of technology and the increasing use of computers in all areas of life, the issue of securing computer systems has become of paramount importance for scientists. Traditional authentication methods such as passwords, or access badges are no longer able to guarantee secured-access to systems that must meet a high level of security [1]. Verifying an individual's identity in order to avoid the risk of fraudulent access to an IT system requires, the use of secure authentication methods.…”

Performance of Fine-Tuning Convolutional Neural Networks for HEp-2 Image Classification

“…In recent years there have been several fields of use of CNNs that, nowadays, allow the development of important applications. Some examples are: the antispoofing of the face and iris [11], recognition of highway traffic congestion [12], the image steganography and steganalysis [13], the galaxy morphology classification [14], drone detection, and classification [15,16]. However, the analysis of IIF images, as a whole, and in particular, in regards to the analysis of intensity, is extremely complex and linked to the experience of the immunologist [4].…”