Ensemble of Multi-View Learning Classifiers for Cross-Domain Iris Presentation Attack Detection

Pinto

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

et al. 2020

Self Cite

This paper addresses the pixelwise classification problem based on temporal profiles, which are encoded in twodimensional representations based on Recurrence Plots, Gramian Angular/Difference Fields, and Markov Transition Field. We propose a multi-representational fusion scheme that exploits the complementary view provided by those time series representations, and different data-driven feature extractors and classifiers. We validate our ensemble scheme in the problem related to the classification of eucalyptus plantations in remote sensing images. Achieved results demonstrate that our proposal overcomes recently proposed baselines, and now represents the new state-of-the-art classification solution for the target dataset.

Section: B Fusion Approachmentioning

confidence: 99%

A Multirepresentational Fusion of Time Series for Pixelwise Classification

Pinto

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

et al. 2020

Self Cite

“…The use of CNN by itself for the sake of classifying attack versus bona-fide iris images has been investigated in recent years. Kuehlkamp et al [96] proposed to use a lightweight CNN on binarised statistical image features extracted from the iris image; they evaluated their proposed algorithm on LivDet-Warsaw 2017 [110] and other three benchmark datasets. In [97], Hoffman et al proposed to use 25 patches of the iris image as input to the CNN instead of the full iris image.…”

Section: Learned-featuresmentioning

confidence: 99%

“…Kuehlkamp et al . [96] proposed to use a lightweight CNN on binarised statistical image features extracted from the iris image; they evaluated their proposed algorithm on LivDet‐Warsaw 2017 [110] and other three benchmark datasets. In [97], Hoffman et al .…”

Section: Passive Software‐based Padmentioning

confidence: 99%

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

2020

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.

“…However, in the past few years, researchers encounter the challenge of cross-domain iris recognition, where heterogeneous cameras/sensors are used for iris image acquisition [21]. Since the vendors use distinct technologies to manufacture such sensors, they might vary in the illumination, wavelength range, and the core hardware.…”

Section: Introductionmentioning

confidence: 99%

CCRNet: a novel data-driven approach to improve cross-domain Iris recognition

Choudhary

Tiwari

Udutalapally

2020

Multimed Tools Appl

In spite of the prominence and robustness of iris recognition systems, iris images acquisition using heterogeneous cameras/sensors, is the prime concern in deploying them for wide-scale applications. The textural qualities of iris samples (images) captured through distinct sensors substantially differ due to the differences in illumination and the underlying hardware that yields intra-class variation within the iris dataset. This paper examines three miscellaneous configurations of convolution and residual blocks to improve cross-domain iris recognition. Further, the finest architecture amongst three is identified by the Friedman test, where the statistical differences in proposed architectures are identified based on the outcomes of Nemeny and Bonferroni-Dunn tests. The quantitative performances of these architectures are perceived on several experiments simulated on two iris datasets; ND-CrossSensor-Iris-2013 and ND-iris-0405. The finest model is referred to as "Collaborative Convolutional Residual Network (CCRNet)" and is further examined on several experiments prepared in similar and cross-domains. Results depict that least two error rates reported by CCRNet are 1.06% and 1.21% that enhances the benchmark for the state of the arts. This is due to fast convergence and rapid weights updation achieved from convolution and residual connections, respectively. It helps in recognizing the micro-patterns existing within the iris region and results in better feature discrimination among large numbers of iris subjects.