Assessing the quality of swipe interactions for mobile biometric systems
Entidad
UAM. Departamento de Tecnología Electrónica y de las ComunicacionesEditor
Institute of Electrical and Electronics Engineers Inc. (IEEE)Fecha de edición
2020-10-01Cita
10.1109/IJCB48548.2020.9304858
2020 IEEE International Joint Conference on Biometrics (IJCB). IEEE, 2020. 1-8
ISBN
9781728191867DOI
10.1109/IJCB48548.2020.9304858Financiado por
Thanks to support by projects: PRIMA (H2020-MSCAITN-2019-860315), TRESPASS-ETN (H2020-MSCAITN-2019-860813), BIBECA (RTI2018-101248-B-I00 MINECO/FEDER), and BioGuard (Ayudas Fundacion BBVA a Equipos de Investigacion Científica 2017)Proyecto
info:eu-repo/grantAgreement/EC/H2020/860315/EU/PriMa-ITN; info:eu-repo/grantAgreement/EC/H2020/860813/EU/TReSPAsS-ETN; Gobierno de España. RTI2018-101248-B-I00Versión del editor
http://doi.org/10.1109/IJCB48548.2020.9304858Materias
Behavioural biometrics; Mobile biometrics; Sample quality; Swipe gestures; User quality; TelecomunicacionesNota
© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other worksDerechos
© Institute of Electrical and Electronics EngineersResumen
Quality estimation is a key study in biometrics, allowing optimisation and improvement of existing authentication systems by giving a prediction on the model performance based on the goodness of the sample or the user. In this paper, we propose a quality metric for swipe gestures on mobile devices. We evaluate a quality score for subjects on enrollment and for swipe samples, we estimate three quality groups and explore the correlation between our quality score and a state-of-art biometric authentication classifier performance. A further analysis based on the combined effects of subject quality and the amount of enrollment samples is conducted, investigating if increasing or decreasing enrollment size affects the authentication performance for different quality groups. Results are shown for three different public datasets, highlighting how higher quality users score a lower equal error rate compared to medium and low quality users, while high quality samples get a higher similarity score from the classifier
Lista de ficheros
Google Scholar:Santopietro, Marco
-
Vera Rodríguez, Rubén
-
Guest, Richard
-
Morales Moreno, Aythami
-
Acién Ayala, Alejandro
Lista de colecciones del ítem
Registros relacionados
Mostrando ítems relacionados por título, autor, creador y materia.