Binary Face Templates with Mobile-Class CNNs: A Reproducible Benchmark for Smart-Card-Constrained Authentication
DOI:
https://doi.org/10.56294/dm20251223Keywords:
Biometrics, Face recognition, Deep learning, Binary templates, Mobile neural networks, Smart cardsAbstract
Facial recognition systems are increasingly deployed in privacy-sensitive and resource-constrained environments such as smart cards. However, traditional face verification relies on high-dimensional floating-point embeddings, which are unsuitable for compact and efficient matching on such platforms. To address this challenge, this work investigates the generation of binary face templates that retain identity information while reducing storage and computational cost.
The objective of this study is to benchmark binary biometric representations derived from mobile-class convolutional neural networks (CNNs), aiming to support reproducible, lightweight face verification pipelines.
We evaluate four lightweight CNNs—EfficientNet-B0, MobileNetV2, ShuffleNetV2, and SqueezeNet1_1—trained on the MORPH dataset. Binary templates are generated via Principal Component Analysis followed by Iterative Quantization (PCA–ITQ) at 32, 64, and 128 bits. Models are tested cross-dataset on the Georgia Tech Face Database (GT Face) to assess generalization.
At 128 bits, EfficientNet-B0 and MobileNetV2 achieve strong verification performance, with area under the curve (AUC) ≈ 0.895–0.899 and equal error rate (EER) ≈ 0.182–0.185. A Hamming-distance analysis confirms clear separation between genuine and impostor pairs, and the bit-flip rate (~17%) indicates intra-subject consistency. Bit-length scaling further reveals monotonic improvements in AUC from 32 to 128 bits, highlighting a trade-off between accuracy and compactness.
These results demonstrate that binary templates from lightweight CNNs can deliver efficient, privacy-preserving authentication with limited performance degradation. The proposed pipeline supports reproducibility and aligns with FAIR data principles, making it suitable for secure biometric deployments on constrained hardware.
References
1. Taigman Y, Yang M, Ranzato M, Wolf L. DeepFace: closing the gap to human-level performance in face verification. Proc. CVPR. 2014. DOI: 10.1109/CVPR.2014.220
2. Schroff F, Kalenichenko D, Philbin J. FaceNet: a unified embedding for face recognition and clustering. Proc. CVPR. 2015. DOI: 10.1109/CVPR.2015.7298682
3. Sun Y, Liang D, Wang X, Tang X. DeepID3: face recognition with very deep neural networks. arXiv preprint arXiv:1502.00873; 2015.
4. Drozdowski P, Struck F, Rathgeb C, Busch C. Benchmarking binarisation schemes for deep face templates. Proc. IEEE Int. Conf. on Image Processing (ICIP). 2018. DOI: 10.1109/ICIP.2018.8451291
5. Das A, Sengupta A, Saqib M, Pal U, Blumenstein M. More realistic and efficient face-based mobile authentication using CNNs. Proc. IJCNN. 2018. DOI: 10.1109/IJCNN.2018.8489070
6. Adebayo O, Boulgouris NV. Binary face representation for privacy-aware authentication. IEEE Trans. Inf. Forensics Security. 2012;7(6):1780–1785. DOI: 10.1109/TIFS.2012.2207726
7. Pandey R, Zhou Y, Urala Kota B, Govindaraju V. Deep secure encoding for face template protection. Proc. CVPR Workshops. 2016. DOI: 10.1109/CVPRW.2016.17
8. Mai G, Cao K, Lan X, Yuen PC. SecureFace: face template protection. IEEE Trans. Inf. Forensics Security. 2021;16:223–238. DOI: 10.1109/TIFS.2020.3009590
9. Parkhi OM, Vedaldi A, Zisserman A. Deep face recognition. Proc. BMVC. 2015. DOI: 10.5244/C.29.41
10. Tan M, Le Q. EfficientNet: Rethinking model scaling for convolutional neural networks. Proc. ICML. 2019. DOI: 10.48550/arXiv.1905.11946
11. Sandler M, Howard A, Zhu M, Zhmoginov A, Chen L-C. MobileNetV2: Inverted residuals and linear bottlenecks. CVPR. 2018. DOI: 10.48550/arXiv.1801.04381
12. Ma N, Zhang X, Zheng H-T, Sun J. ShuffleNet V2: Practical guidelines for efficient CNN architecture design. ECCV. 2018. DOI: 10.1007/978-3-030-01264-9_8
13. Iandola FN, Han S, Moskewicz MW, Ashraf K, Dally WJ, Keutzer K. SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size. arXiv preprint arXiv:1602.07360; 2016. DOI: 10.48550/arXiv.1602.07360
14. Gong Y, Lazebnik S, Gordo A, Perronnin F. Iterative Quantization: A Procrustean approach to learning binary codes. CVPR. 2013. DOI: 10.1109/CVPR.2013.214
15. Ricanek K, Tesafaye T. MORPH: A Longitudinal Image Database of Normal Adult Age‑Progression. FGR. IEEE; 2006. DOI: 10.1109/FGR.2006.78
16. Liu W, Wen Y, Yu Z, Li M, Raj B, Song L. SphereFace: Deep hypersphere embedding for face recognition. CVPR. 2017. DOI: 10.1109/CVPR.2017.713
17. Deng J, Guo J, Xue N, Zafeiriou S. ArcFace: Additive angular margin loss for deep face recognition. IEEE TPAMI. 2022;44(10):5962–5979. DOI: 10.1109/TPAMI.2021.3084828
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Abdelilah Ganmati, Karim Afdel, Lahcen Koutti (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.
The article is distributed under the Creative Commons Attribution 4.0 License. Unless otherwise stated, associated published material is distributed under the same licence.
