Analysis of the synthetic periocular iris images for robust Presentation Attacks Detection algorithms

Abstract The LivDet‐2020 competition focuses on Presentation Attacks Detection (PAD) algorithms, has still open problems, mainly unknown attack scenarios. It is crucial to enhance PAD methods. This can be achieved by augmenting the number of Presentation Attack Instruments (PAI) and Bona fide (genui...

Full description

Saved in:
Bibliographic Details
Main Authors: Jose Maureira, Juan E. Tapia, Claudia Arellano, Christoph Busch
Format: Article
Language:English
Published: Wiley 2022-07-01
Series:IET Biometrics
Online Access:https://doi.org/10.1049/bme2.12084
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850213508652728320
author Jose Maureira
Juan E. Tapia
Claudia Arellano
Christoph Busch
author_facet Jose Maureira
Juan E. Tapia
Claudia Arellano
Christoph Busch
author_sort Jose Maureira
collection DOAJ
description Abstract The LivDet‐2020 competition focuses on Presentation Attacks Detection (PAD) algorithms, has still open problems, mainly unknown attack scenarios. It is crucial to enhance PAD methods. This can be achieved by augmenting the number of Presentation Attack Instruments (PAI) and Bona fide (genuine) images used to train such algorithms. Unfortunately, the capture and creation of PAI and even the capture of Bona fide images are sometimes complex to achieve. The generation of synthetic images with Generative Adversarial Networks (GAN) algorithms may help and has shown significant improvements in recent years. This paper presents a benchmark of GAN methods to achieve a novel synthetic PAI from a small set of periocular near‐infrared images. The best PAI was obtained using StyleGAN2, and it was tested using the best PAD algorithm from the LivDet‐2020. The synthetic PAI was able to fool such an algorithm. As a result, all images were classified as Bona fide. A MobileNetV2 was trained using the synthetic PAI as a new class to achieve a more robust PAD. The resulting PAD was able to classify 96.7% of synthetic images as attacks. BPCER10 was 0.24%. Such results demonstrated the need for PAD algorithms to be constantly updated and trained with synthetic images.
format Article
id doaj-art-83fbc7ab65bd4f358a0e06e5011a23b9
institution OA Journals
issn 2047-4938
2047-4946
language English
publishDate 2022-07-01
publisher Wiley
record_format Article
series IET Biometrics
spelling doaj-art-83fbc7ab65bd4f358a0e06e5011a23b92025-08-20T02:09:07ZengWileyIET Biometrics2047-49382047-49462022-07-0111434335410.1049/bme2.12084Analysis of the synthetic periocular iris images for robust Presentation Attacks Detection algorithmsJose Maureira0Juan E. Tapia1Claudia Arellano2Christoph Busch3TOC Biometrics, R&D Center SR‐226 Santiago Chileda/sec‐Biometrics and Internet Security Research Group Hochschule Darmstadt Darmstadt GermanyUniversidad Adolfo Ibañez Santiago Chileda/sec‐Biometrics and Internet Security Research Group Hochschule Darmstadt Darmstadt GermanyAbstract The LivDet‐2020 competition focuses on Presentation Attacks Detection (PAD) algorithms, has still open problems, mainly unknown attack scenarios. It is crucial to enhance PAD methods. This can be achieved by augmenting the number of Presentation Attack Instruments (PAI) and Bona fide (genuine) images used to train such algorithms. Unfortunately, the capture and creation of PAI and even the capture of Bona fide images are sometimes complex to achieve. The generation of synthetic images with Generative Adversarial Networks (GAN) algorithms may help and has shown significant improvements in recent years. This paper presents a benchmark of GAN methods to achieve a novel synthetic PAI from a small set of periocular near‐infrared images. The best PAI was obtained using StyleGAN2, and it was tested using the best PAD algorithm from the LivDet‐2020. The synthetic PAI was able to fool such an algorithm. As a result, all images were classified as Bona fide. A MobileNetV2 was trained using the synthetic PAI as a new class to achieve a more robust PAD. The resulting PAD was able to classify 96.7% of synthetic images as attacks. BPCER10 was 0.24%. Such results demonstrated the need for PAD algorithms to be constantly updated and trained with synthetic images.https://doi.org/10.1049/bme2.12084
spellingShingle Jose Maureira
Juan E. Tapia
Claudia Arellano
Christoph Busch
Analysis of the synthetic periocular iris images for robust Presentation Attacks Detection algorithms
IET Biometrics
title Analysis of the synthetic periocular iris images for robust Presentation Attacks Detection algorithms
title_full Analysis of the synthetic periocular iris images for robust Presentation Attacks Detection algorithms
title_fullStr Analysis of the synthetic periocular iris images for robust Presentation Attacks Detection algorithms
title_full_unstemmed Analysis of the synthetic periocular iris images for robust Presentation Attacks Detection algorithms
title_short Analysis of the synthetic periocular iris images for robust Presentation Attacks Detection algorithms
title_sort analysis of the synthetic periocular iris images for robust presentation attacks detection algorithms
url https://doi.org/10.1049/bme2.12084
work_keys_str_mv AT josemaureira analysisofthesyntheticperiocularirisimagesforrobustpresentationattacksdetectionalgorithms
AT juanetapia analysisofthesyntheticperiocularirisimagesforrobustpresentationattacksdetectionalgorithms
AT claudiaarellano analysisofthesyntheticperiocularirisimagesforrobustpresentationattacksdetectionalgorithms
AT christophbusch analysisofthesyntheticperiocularirisimagesforrobustpresentationattacksdetectionalgorithms