The Self-Detection Method of the Puppet Attack in Biometric Fingerprinting

Li, Guyue ORCID: 0000-0003-1145-1168, Ma, Yiyun ORCID: 0009-0008-5455-1314, Wang, Wenhao, Zhang, Junqing ORCID: 0000-0002-3502-2926 and Luo, Hongyi ORCID: 0009-0006-4750-4451 (2024) The Self-Detection Method of the Puppet Attack in Biometric Fingerprinting. IEEE Internet of Things Journal, PP (99). p. 1.

Text
IoTJ 2024 Fingerprint.pdf - Author Accepted Manuscript
Available under License Creative Commons Attribution.
Download (4MB) | Preview

Official URL: http://dx.doi.org/10.1109/jiot.2024.3365714

Abstract

Fingerprint authentication has become a staple in securing access to personal devices and sensitive information in our daily lives, with the security level of such systems being paramount. Recent attention has been drawn to the puppet attack, a forced fingerprint unlocking scenario that exploits legitimate user fingerprints for unauthorized access. Traditional authentication methods are constrained by their reliance on additional sensors and are typically limited to static authentication scenarios, lacking versatility in dynamic or mobile contexts. In this study, we employ physical modeling to elucidate puppet attack, unraveling the distinctive stress patterns and points of application associated with forced interactions. By scrutinizing the physical alterations induced during such attacks, our investigation unveils discernible changes in the texture of fingerprints, specifically reflecting variations linked to different force patterns. Consequently, we introduce a detection system that operates without the need for external sensors, solely utilizing fingerprint images to extract texture features, thereby offering a broadly applicable solution. To address the challenge posed by the absence of puppet attack samples in existing datasets, we constructed a comprehensive database, incorporating a substantial number of puppet attack fingerprints collected from 70 volunteers aged between 20 and 75. This database facilitates a more robust detection of puppet attack. Our system demonstrates accuracy rates of 85.5%, 97.2%, 86.5%, and 78.1% across four distinct scenarios within our puppet attack database.

Item Type:	Article
Divisions:	Faculty of Science and Engineering > School of Electrical Engineering, Electronics and Computer Science
Depositing User:	Symplectic Admin
Date Deposited:	12 Feb 2024 09:19
Last Modified:	02 Apr 2024 15:46
DOI:	10.1109/jiot.2024.3365714
Related URLs:	Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3178601