TU Darmstadt / ULB / TUbiblio

P-Score: Performance Aligned Normalization and an Evaluation in Score-Level Multi-Biometric Fusion

Damer, Naser and Boutros, Fadi and Terhorst, Philipp and Braun, Andreas and Kuijper, Arjan (2018):
P-Score: Performance Aligned Normalization and an Evaluation in Score-Level Multi-Biometric Fusion.
In: 2018 26th European Signal Processing Conference (EUSIPCO), Los Alamitos, IEEE, In: European Signal Processing Conference (EUSIPCO), Rome, Italy, ISBN 978-90-827970-1-5,
DOI: 10.23919/EUSIPCO.2018.8553553,
[Online-Edition: https://doi.org/10.23919/EUSIPCO.2018.8553553],
[Conference or Workshop Item]

Abstract

Normalization is an important step for different fusion, classification, and decision making applications. Previous normalization approaches considered bringing values from different sources into a common range or distribution characteristics. In this work we propose a new normalization approach that transfers values into a normalized space where their relative performance in binary decision making is aligned across their whole range. Multi-biometric verification is a typical problem where information from different sources are normalized and fused to make a binary decision and therefore a good platform to evaluate the proposed normalization.We conducted an evaluation on two publicly available databases and showed that the normalization solution we are proposing consistently outperformed state-of-the-art and best practice approaches, e.g. by reducing the false rejection rate at 0.01% false acceptance rate by 60- 75% compared to the widely used z-score normalization under the sum-rule fusion.

Item Type: Conference or Workshop Item
Erschienen: 2018
Creators: Damer, Naser and Boutros, Fadi and Terhorst, Philipp and Braun, Andreas and Kuijper, Arjan
Title: P-Score: Performance Aligned Normalization and an Evaluation in Score-Level Multi-Biometric Fusion
Language: English
Abstract:

Normalization is an important step for different fusion, classification, and decision making applications. Previous normalization approaches considered bringing values from different sources into a common range or distribution characteristics. In this work we propose a new normalization approach that transfers values into a normalized space where their relative performance in binary decision making is aligned across their whole range. Multi-biometric verification is a typical problem where information from different sources are normalized and fused to make a binary decision and therefore a good platform to evaluate the proposed normalization.We conducted an evaluation on two publicly available databases and showed that the normalization solution we are proposing consistently outperformed state-of-the-art and best practice approaches, e.g. by reducing the false rejection rate at 0.01% false acceptance rate by 60- 75% compared to the widely used z-score normalization under the sum-rule fusion.

Title of Book: 2018 26th European Signal Processing Conference (EUSIPCO)
Place of Publication: Los Alamitos
Publisher: IEEE
ISBN: 978-90-827970-1-5
Uncontrolled Keywords: Multibiometrics, Information fusion, Biometric fusion, Biometrics
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Mathematical and Applied Visual Computing
Event Title: European Signal Processing Conference (EUSIPCO)
Event Location: Rome, Italy
Date Deposited: 01 Jul 2019 08:47
DOI: 10.23919/EUSIPCO.2018.8553553
Official URL: https://doi.org/10.23919/EUSIPCO.2018.8553553
Export:
Suche nach Titel in: TUfind oder in Google

Optionen (nur für Redakteure)

View Item View Item