CMAD: A CNN-Based model for Morphing Attack Detection

DOI

10.2991/978-94-6463-858-5_16How to use a DOI?

Keywords

morphing; morphing attack detection; Siamese model; CNN

Abstract

Authentication systems play a crucial role in verifying the identity of individuals, whether they are accessing physical spaces, digital platforms, or services. Traditional methods of authentication, such as passwords, PINs, and security questions, have limitations, including vulnerability to theft, guessing, or brute force attacks. Biometric authentication has emerged as a more secure and convenient alternative. Biometrics refers to the use of unique physical characteristics such as fingerprints, facial features, iris patterns, or voice recognition to identify individuals. For the purpose of security, these biometric systems are being employed in the banking sector, healthcare, offices, airports etc. At the airports, Automatic Border Control (ABC) gates are being used which capture a live photo and compare it with the photo in e-MRTD. The traveller is only allowed to pass the gate when both of these photos match. Today, these ABC systems are exposed to face morphing attacks in which a person tries to pass the gates by using a forged photo in his passport. This photo is created by combining two images together which resembles both the participating images and is used to fool the security systems. This problem is alarming as it is a matter of national security. Therefore, the proposed model detects whether a photo is morphed or not. The dataset contains both genuine and morphed images. Various pre-processing techniques is applied on the images to make them ready for the training. Three blocks of convolutional layers followed by a fully connected layer is used in the architecture. Adam optimizer is used for optimization and binary cross entropy is employed as a loss function. The model outperforms in comparison to existing techniques when various performance metrics are applied on the test dataset.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Pooja Arora
AU  - Gurpreet Singh
AU  - Aaisha Makkar
PY  - 2025
DA  - 2025/11/04
TI  - CMAD: A CNN-Based model for Morphing Attack Detection
BT  - Proceedings of International Conference on Computer Science and Communication Engineering (ICCSCE 2025)
PB  - Atlantis Press
SP  - 175
EP  - 186
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-858-5_16
DO  - 10.2991/978-94-6463-858-5_16
ID  - Arora2025
ER  -