TU Darmstadt / ULB / TUbiblio

Demographic Bias in Presentation Attack Detection of Iris Recognition Systems

Fang, Meiling and Damer, Naser and Kirchbuchner, Florian and Kuijper, Arjan (2021):
Demographic Bias in Presentation Attack Detection of Iris Recognition Systems.
In: 28th European Signal Processing Conference (EUSIPCO 2020) : Proceedings, pp. 835-839,
IEEE, 28th European Signal Processing Conference, virtual Conference, 18.-21.01.2021, ISBN 978-9-0827-9705-3,
DOI: 10.23919/Eusipco47968.2020.9287321,
[Conference or Workshop Item]

Abstract

With the widespread use of biometric systems, the demographic bias problem raises more attention. Although many studies addressed bias issues in biometric verification, there are no works that analyze the bias in presentation attack detection (PAD) decisions. Hence, we investigate and analyze the demographic bias in iris PAD algorithms in this paper. To enable a clear discussion, we adapt the notions of differential performance and differential outcome to the PAD problem. We study the bias in iris PAD using three baselines (hand-crafted, transfer-learning, and training from scratch) using the NDCLD- 2013 [18] database. The experimental results point out that female users will be significantly less protected by the PAD, in comparison to males.

Item Type: Conference or Workshop Item
Erschienen: 2021
Creators: Fang, Meiling and Damer, Naser and Kirchbuchner, Florian and Kuijper, Arjan
Title: Demographic Bias in Presentation Attack Detection of Iris Recognition Systems
Language: English
Abstract:

With the widespread use of biometric systems, the demographic bias problem raises more attention. Although many studies addressed bias issues in biometric verification, there are no works that analyze the bias in presentation attack detection (PAD) decisions. Hence, we investigate and analyze the demographic bias in iris PAD algorithms in this paper. To enable a clear discussion, we adapt the notions of differential performance and differential outcome to the PAD problem. We study the bias in iris PAD using three baselines (hand-crafted, transfer-learning, and training from scratch) using the NDCLD- 2013 [18] database. The experimental results point out that female users will be significantly less protected by the PAD, in comparison to males.

Title of Book: 28th European Signal Processing Conference (EUSIPCO 2020) : Proceedings
Publisher: IEEE
ISBN: 978-9-0827-9705-3
Uncontrolled Keywords: Biometrics, Computer vision, Machine learning
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Interactive Graphics Systems
20 Department of Computer Science > Mathematical and Applied Visual Computing
Event Title: 28th European Signal Processing Conference
Event Location: virtual Conference
Event Dates: 18.-21.01.2021
Date Deposited: 25 Jan 2021 11:45
DOI: 10.23919/Eusipco47968.2020.9287321
Additional Information:

28th European Signal Processing Conference, EUSIPCO 2020, which will be held in the Beurs van Berlage (Amsterdam Conference Centre) from January 18 to January 22, 2021. The EURASIP board of directors have decided to reschedule the physical conference from August 24 - 28, 2020, to January 18 - 22, 2021.

Corresponding Links:
Export:
Suche nach Titel in: TUfind oder in Google
Send an inquiry Send an inquiry

Options (only for editors)
Show editorial Details Show editorial Details