Improvement of the Face Recognition Systems Security Against Morph Attacks using the Developed Siamese Neural Network
Article Sidebar
Main Article Content
Abstract
Face Recognition Systems (FRS) are increasingly targeted by morphing attacks, where facial features of multiple individuals are blended into a synthetic image to deceive biometric verification. This paper proposes an enhanced Siamese Neural Network (SNN)-based system for robust morph detection. The methodology involves four stages. First, a dataset of real and morphed images is generated using StyleGAN, producing high-quality facial images. Second, facial regions are extracted using Faster Region-based Convolutional Neural Networks (R-CNN) to isolate relevant features and eliminate background noise. Third, a Local Binary Pattern-Convolutional Neural Network (LBP-CNN) is used to build a baseline FRS and assess its susceptibility to deception by morphed images. Finally, morph detection and classification are conducted using the proposed SNN framework, which incorporates a novel feature fusion strategy based on Canonical Correlation Analysis (CCA) to enhance discriminative power. The model is trained and evaluated using publicly available Face Recognition Technology (FERET) and Face Recognition Grand Challenge (FRGC) datasets, comprising 1,030 real and 2,000 morphed images. Experimental results demonstrate that the proposed method significantly strengthens the resilience of FRS to morphing attacks, achieving a high detection accuracy of 99.9%. This confirms the model’s effectiveness in distinguishing between real and manipulated images with minimal errors.
Article Details
Issue
Section
This work is licensed under a Creative Commons Attribution 4.0 International License.
How to Cite
References
[1] K. A. Ibrahim, B. N. Al-Din Abed, and S. A. S. Hussien, “A novel diffusion-based cryptographic method for cyber security,” Mesopotamian J. Cybersecurity, vol. 5, no. 2, pp. 842–862, 2025, doi: 10.58496/MJCS/2025/048 .
[2] W. A. H. Salman and C. Huah Yong, “An efficient distributed intrusion detection system that combines traditional machine learning techniques with advanced deep learning,” Mesopotamian J. Cybersecurity, vol. 5, no. 2, pp. 721–734, 2025, doi: 10.58496/MJCS/2025/043 .
[3] I. Saleem and B. K. Shukr, “Techniques and challenges for generation and detection of face morphing attacks: A survey,” Iraqi J. Sci., pp. 385–404, 2023, doi: 10.24996/ijs.2023.64.1.36 .
[4] S. V., R. Raghavendra, R. Kiran, and C. Busch, “Face morphing attack generation and detection: A comprehensive survey,” IEEE Trans. Technol. Soc., 2021, doi: 10.1109/TTS.2021.3066254 .
[5] R. Ramachandra and C. Busch, “Presentation attack detection methods for face recognition systems: A comprehensive survey,” ACM Comput. Surv., vol. 50, no. 1, pp. 1–37, 2017, doi: 10.1145/3038924 .
[6] S. A. S. Hussien, B. N. A. Din Abed, and K. A. Ibrahim, “Encrypting text messages via iris recognition and gaze tracking technology,” Mesopotamian J. Cybersecurity, vol. 5, no. 1, pp. 90–103, 2025, doi: 10.58496/MJCS/2025/007 .
[7] O. O. Petrova and K. B. Bulatov, “Methods of machine-readable zone recognition results post-processing,” in Proc. Int. Conf. Microelectron. Comput. (ICMV), vol. 11041, 2019, pp. 387–393, doi: 10.1117/12.2522792 .
[8] B. S. Mahmmed and A. I. Majeed, “Face detection and recognition using Google-Net architecture,” Iraqi J. Inf. Commun. Technol., vol. 6, no. 1, pp. 66–79, 2023.
[9] Y. Zhang et al., “Evaluation of StyleGAN3 morphing in biometric systems,” in Proc. Int. Joint Conf. Biometrics (IJCB), 2024, pp. 1–8, doi: 10.1109/IJCB.2024.10230219 .
[10] M. Ivanovska and V. Štruc, “Face morphing attack detection with denoising diffusion probabilistic models,” in Proc. Int. Workshop Biometrics Forensics (IWBF), 2023, pp. 1–6, doi: 10.1109/IWBF57495.2023.10156877 .
[11] J. Tapia and C. Busch, “Impact of synthetic images on morphing attack detection using a Siamese network,” arXiv preprint arXiv:2403.09380 [cs.CV], 2024.
[12] C.-K. Jia, Y.-C. Liu, and Y.-L. Chen, “Face morphing attack detection based on high-frequency features and progressive enhancement learning,” Front. Neurorobot., vol. 17, p. 1182375, 2023, doi: 10.3389/fnbot.2023.1182375 .
[13] R. Dwivedi et al., “An efficient ensemble explainable AI (XAI) approach for morphed face detection,” arXiv preprint arXiv:2304.14509 [cs.CV], 2023.
[14] R. Ramachandra and S. Venkatesh, “Time-frequency based convolution neural network for differential morphing attack detection,” in Computer Vision and Image Processing (CVIP 2024), J. Kakarla et al., Eds. Cham, Switzerland: Springer, 2024, vol. 2473, pp. 345–358, doi: 10.1007/978-3-031-93688-3_29 .
[15] M. Jafari and R. Singh, “Towards explainable morphing attack detection: A visual saliency approach,” Comput. Vis. Image Underst., vol. 245, p. 103491, 2024, doi: 10.1016/j.cviu.2024.103491 .
[16] H. Chen and R. Zhao, “Adaptive feature fusion for robust Siamese networks in morph detection,” Pattern Recognit. Lett., vol. 166, pp. 1–7, 2023, doi: 10.1016/j.patrec.2023.01.012 .
[17] J. M. Singh, S. Venkatesh, and R. Ramachandra, “Robust face morphing attack detection using fusion of multiple features and classification techniques,” arXiv preprint arXiv:2305.03264 [cs.CV], 2023.
[18] A. Ali et al., “MorphAttackNet: A lightweight CNN for real-time face morphing attack detection,” IEEE Trans. Biometrics, Behav., Identity Sci., vol. 5, no. 3, pp. 412–423, 2023, doi: 10.1109/TBIOM.2023.3251234 .
[19] L. DeBruine and B. Jones, “Face Research Lab London Set,” figshare, 2017. [Online]. Available: https://figshare.com/articles/dataset/Face_Research_Lab_London_Set/5047666
[20] J. Bird, “Football players and staff faces,” Kaggle Dataset, 2020. [Online]. Available: https://www.kaggle.com/datasets/birdy654/football-players-and-staff-faces/
[21] L. DeBruine and B. Jones, “Young adult white faces with manipulated versions,” figshare, 2022. [Online]. Available: https://figshare.com/articles/dataset/Young_Adult_White_Faces_with_Manipulated_Versions/4220517
[22] J. M. Chen, J. B. Norman, and Y. Nam, “Broadening the stimulus set: Introducing the American Multiracial Faces Database,” Behav. Res. Methods, vol. 53, pp. 371–389, 2021, doi: 10.3758/s13428-020-01447-8 .
[23] M. Walker, S. Schönborn, R. Greifeneder, and T. Vetter, “The Basel Face Database: A validated set of photographs reflecting systematic differences in Big Two and Big Five personality dimensions,” PLoS ONE, vol. 13, no. 2, p. e0193190, 2018, doi: 10.1371/journal.pone.0193190 .
[24] S. A. Saribay, A. F. Biten, E. O. Meral, P. Aldan, V. Trebicky, and K. Kleisner, “The Bogazici Face Database: Standardized photographs of Turkish faces with supporting materials,” PLoS ONE, vol. 13, no. 2, p. e0192018, 2018, doi: 10.1371/journal.pone.0192018 .
[25] D. S. Ma, J. Kantner, and B. Wittenbrink, “Chicago Face Database: Multiracial expansion,” Behav. Res. Methods, vol. 53, pp. 1289–1300, 2021, doi: 10.3758/s13428-020-01482-5 .
[26] N. Strohminger et al., “MR2 Face Database,” Behav. Res. Methods, vol. 48, pp. 1197–1204, 2016, doi: 10.3758/s13428-016-0775-2.
(Note: Corrected DOI based on standard format; original had typo)
[27] S. Price, S. Soleymani, and N. M. Nasrabadi, “Landmark enforcement and style manipulation for generative morphing,” in Proc. Int. Joint Conf. Biometrics (IJCB), 2022, pp. 1–10, doi: 10.1109/IJCB54206.2022.10008001 .
[28] C. Cao et al., “An improved Faster R-CNN for small object detection,” IEEE Access, vol. 7, pp. 106838–106846, 2019, doi: 10.1109/ACCESS.2019.2932731 .
[29] M. Karthikeyan and T. Subashini, “Automated object detection of mechanical fasteners using faster region-based convolutional neural networks,” Int. J. Electr. Comput. Eng., vol. 11, no. 6, pp. 5430–5437, 2021, doi: 10.11591/ijece.v11i6.pp5430-5437 .
[30] C. Zhang, W. Liu, H. Ma, and H. Fu, “Siamese neural network based gait recognition for human identification,” in Proc. IEEE Int. Conf. Acoust., Speech Signal Process. (ICASSP), 2016, pp. 2832–2836, doi: 10.1109/ICASSP.2016.7472194 .
[31] N. Serrano and A. Bellogin, “Siamese neural networks in recommendation,” Neural Comput. Appl., vol. 35, pp. 13941–13953, 2023, doi: 10.1007/s00521-023-08610-0 .