Aletheia: A Hybrid AI Framework for Fake Biometric Fingerprint Detection

View Sample PDF

Author(s): Claudia Mattar (Issam Fares Faculty of Technology, University of Balamand, Lebanon), Claritta AlSaneh (Issam Fares Faculty of Technology, University of Balamand, Lebanon)and Soraia Oueida (College of Engineering and Technology, American University of the Middle East, Kuwait)
Copyright: 2026
Pages: 28
Source title: Exploring the Intersection of Forensics and Biometrics
Source Author(s)/Editor(s): Sarah Benziane (University of Science and Technology in Oran, Algeria)
DOI: 10.4018/979-8-3373-4972-5.ch010

Keywords: Biometrics / Criminal Science and Forensics / Information Science Reference / Medicine & Healthcare

Purchase

View Aletheia: A Hybrid AI Framework for Fake Biometric Fingerprint Detection on the publisher's website for pricing and purchasing information.

Abstract

The purpose of this work is to create a fake biometric detection system that detects fake biometric images. The system utilizes generative artificial intelligence, particularly a Deep Convolutional Generative Adversarial Network, to generate fake fingerprints in the best accuracy possible. Tested on fingerprint images, extracted from SOCOFing database, the system resulted with an accuracy of 99.17%. Two different datasets: the real dataset and the fake dataset are then passed through a Support Vector Machine model with linear Kernel to test and calculate the accuracy of detection of both systems. A Sanity Check Accuracy test is performed to check the sanity of the system, offering promising results after label shuffling and retraining of the SVM. The proposed system also uses image quality analysis like Laplacian Blur estimation, Local Binary Pattern features, Edge density, and Histogram Entropy to extract features for testing and training the SVM.

The IRMA Community

Research IRM

Aletheia: A Hybrid AI Framework for Fake Biometric Fingerprint Detection

Purchase

Abstract

Related Content

IRMA Sponsors