SwipeFormer: transformers for mobile touchscreen biometrics
SwipeFormer: transformers for mobile touchscreen biometrics
The growing number of mobile devices over the past few years brings a large amount of personal information, which needs to be properly protected. As a result, several mobile authentication methods have been developed. In particular, behavioural biometrics has become one of the most relevant methods due to its ability to extract the uniqueness of each subject in a secure, non-intrusive, and continuous way. This article presents SwipeFormer, a novel Transformer-based system for mobile subject authentication by means of swipe gestures in an unconstrained scenario (i.e., subjects could use their personal devices freely, without restrictions on the direction of swipe gestures or the position of the device). Our proposed system contains two modules: (i) a Transformer-based feature extractor, and (ii) a similarity computation module. Mobile data from the touchscreen and different background sensors (accelerometer and gyroscope) have been studied, including in the analysis both Android and iOS operating systems. A complete analysis of SwipeFormer is carried out using an in-house large-scale database acquired in unconstrained scenarios. In these operational conditions, SwipeFormer achieves Equal Error Rate (EER) values of 6.6.6 outperforming the state of the art. In addition, we evaluate SwipeFormer on the popular publicly available databases Frank DB and HuMIdb, achieving EER values of 11.0.0 outperforming previous approaches under the same experimental setup.
Behavioural biometrics, Touchscreen, Swipe verification, Transformers, Deep learning, Mobile devices
Delgado-Santos, Paula
61d96aa4-4228-4b7d-9e55-45737560512e
Tolosana, Ruben
93125127-5ac2-4e76-94aa-4d09f28a3e51
Guest, Richard
93533dbd-b101-491b-83cc-39ccfdc18165
Lamb, Parker
104d78a3-5cfe-4800-9ebd-8471b31a36b8
Khmelnitsky, Andrei
002dbee2-0623-46de-ac59-e3d60a37dac1
Coughlan, Colm
9dc7c9ea-bf3d-4674-b93d-1a32307727ed
Fierrez, Julian
c93da818-9beb-4e74-981b-e7a8d394e719
25 September 2023
Delgado-Santos, Paula
61d96aa4-4228-4b7d-9e55-45737560512e
Tolosana, Ruben
93125127-5ac2-4e76-94aa-4d09f28a3e51
Guest, Richard
93533dbd-b101-491b-83cc-39ccfdc18165
Lamb, Parker
104d78a3-5cfe-4800-9ebd-8471b31a36b8
Khmelnitsky, Andrei
002dbee2-0623-46de-ac59-e3d60a37dac1
Coughlan, Colm
9dc7c9ea-bf3d-4674-b93d-1a32307727ed
Fierrez, Julian
c93da818-9beb-4e74-981b-e7a8d394e719
Delgado-Santos, Paula, Tolosana, Ruben, Guest, Richard, Lamb, Parker, Khmelnitsky, Andrei, Coughlan, Colm and Fierrez, Julian
(2023)
SwipeFormer: transformers for mobile touchscreen biometrics.
Expert Systems with Applications, 237 (Part C), [121537].
(doi:10.1016/j.eswa.2023.121537).
Abstract
The growing number of mobile devices over the past few years brings a large amount of personal information, which needs to be properly protected. As a result, several mobile authentication methods have been developed. In particular, behavioural biometrics has become one of the most relevant methods due to its ability to extract the uniqueness of each subject in a secure, non-intrusive, and continuous way. This article presents SwipeFormer, a novel Transformer-based system for mobile subject authentication by means of swipe gestures in an unconstrained scenario (i.e., subjects could use their personal devices freely, without restrictions on the direction of swipe gestures or the position of the device). Our proposed system contains two modules: (i) a Transformer-based feature extractor, and (ii) a similarity computation module. Mobile data from the touchscreen and different background sensors (accelerometer and gyroscope) have been studied, including in the analysis both Android and iOS operating systems. A complete analysis of SwipeFormer is carried out using an in-house large-scale database acquired in unconstrained scenarios. In these operational conditions, SwipeFormer achieves Equal Error Rate (EER) values of 6.6.6 outperforming the state of the art. In addition, we evaluate SwipeFormer on the popular publicly available databases Frank DB and HuMIdb, achieving EER values of 11.0.0 outperforming previous approaches under the same experimental setup.
Text
1-s2.0-S0957417423020390-main
- Version of Record
More information
Accepted/In Press date: 8 September 2023
e-pub ahead of print date: 16 September 2023
Published date: 25 September 2023
Keywords:
Behavioural biometrics, Touchscreen, Swipe verification, Transformers, Deep learning, Mobile devices
Identifiers
Local EPrints ID: 489419
URI: http://eprints.soton.ac.uk/id/eprint/489419
ISSN: 0957-4174
PURE UUID: eff77729-9f9d-46e4-8a4e-61d1eaf6b2b9
Catalogue record
Date deposited: 24 Apr 2024 16:30
Last modified: 25 Apr 2024 02:09
Export record
Altmetrics
Contributors
Author:
Paula Delgado-Santos
Author:
Ruben Tolosana
Author:
Richard Guest
Author:
Parker Lamb
Author:
Andrei Khmelnitsky
Author:
Colm Coughlan
Author:
Julian Fierrez
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics