Federated Multi-Camera Person Re-Identification with Privacy-Preserving Feature Decoupling Mechanisms
Keywords:
federated learning, person re-identification, privacy preservation, feature decoupling, multi-camera systems, decentralized AIAbstract
The growing deployment of multi-camera surveillance and analytics systems across urban, commercial, and institutional environments has amplified both the utility and the risks of person re-identification. Centralizing visual data from distributed camera networks poses severe privacy threats, regulatory burdens, and single points of failure, while purely localized models suffer from degraded performance due to limited data diversity. Federated learning offers a decentralized paradigm that enables collaborative model training without raw data exchange, yet conventional federated re-identification still transmits deep features that may inadvertently encode sensitive personal attributes. This paper presents a system-level framework for federated multi-camera person re-identification that integrates privacy-preserving feature decoupling mechanisms. Within each camera edge node, a deep representation is decomposed into identity-discriminative components that are invariant to transient appearance, clothing, and camera-specific biases, and into a complementary set of non-identity attributes that remain local. Only the identity-specific embeddings are encrypted and aggregated across the federated network, while the sensitive part is discarded or stored solely for local inference. The paper examines the architecture, communication topology, and adversarial disentanglement strategies that underpin this decoupling, and it critically analyzes structural trade-offs among accuracy, communication efficiency, fairness, and robustness. Further, it addresses deployment considerations at scale, governance frameworks that align with data protection regulations such as GDPR, and long-term sustainability aspects including energy consumption, continual adaptation, and resilience against distributional shifts. The analysis highlights how feature decoupling reconfigures the privacy-utility frontier and establishes a principled foundation for accountable, transparent, and auditable person re-identification infrastructures. The discussion also identifies open challenges related to heterogeneous camera capabilities, adversarial evasion, and the need for standardized benchmarks that evaluate both re-identification accuracy and privacy leakage under federated constraints. By weaving together architectural design, policy implications, and infrastructure sustainability, this work provides a comprehensive systems perspective on the next generation of privacy-conscious visual intelligence networks.
References
1. McMahan, B., Moore, E., Ramage, D., Hampson, S., & y Arcas, B. A. (2017). Communication-efficient learning of deep networks from decentralized data. Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS), 1273–1282.
2. Zheng, L., Yang, Y., & Hauptmann, A. G. (2016). Person re-identification: Past, present and future. arXiv preprint arXiv:1610.02984.
3. Liu, B., Ding, M., Shaham, S., Rahwan, T., & Duel-Hallen, A. (2021). Privacy-preserving deep learning: A comprehensive survey. ACM Computing Surveys, 54(6), 1–41.
4. Higgins, I., Matthey, L., Pal, A., Burgess, C., Glorot, X., Botvinick, M., … & Lerchner, A. (2017). Beta-Vae: Learning basic visual concepts with a constrained variational framework. In International Conference on Learning Representations (ICLR).
5. Ristani, E., & Tomasi, C. (2018). Features for multi-target multi-camera tracking and re-identification. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 6036–6046.
6. Kairouz, P., McMahan, H. B., Avent, B., Bellet, A., Bennis, M., Bhagoji, A. N., … & Zhao, S. (2021). Advances and open problems in federated learning. Foundations and Trends in Machine Learning, 14(1–2), 1–210.
7. Yang, Q., Wu, A., & Zheng, W. S. (2021). Person re-identification by contour sketch under moderate clothing change. IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(9), 3160–3173.
8. Ding, Y., Wang, X., Yuan, H., Qu, M., & Jian, X. (2025). Decoupling feature-driven and multimodal fusion attention for clothing-changing person re-identification. Artificial Intelligence Review, 58(8), 241.
9. Abadi, M., Chu, A., Goodfellow, I., McMahan, H. B., Mironov, I., Talwar, K., & Zhang, L. (2016). Deep learning with differential privacy. Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, 308–318.
10. Bonawitz, K., Ivanov, V., Kreuter, B., Marcedone, A., McMahan, H. B., Patel, S., … & Seth, K. (2017). Practical secure aggregation for privacy-preserving machine learning. Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security, 1175–1191.
11. Wang, S., Tuor, T., Salonidis, T., Leung, K. K., Makaya, C., He, T., & Chan, K. (2019). Adaptive federated learning in resource constrained edge computing systems. IEEE Journal on Selected Areas in Communications, 37(6), 1205–1221.
12. Buolamwini, J., & Gebru, T. (2018). Gender shades: Intersectional accuracy disparities in commercial gender classification. Proceedings of the 1st Conference on Fairness, Accountability and Transparency, 77–91.
13. Hinton, G., Vinyals, O., & Dean, J. (2015). Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531.
14. Kirkpatrick, J., Pascanu, R., Rabinowitz, N., Veness, J., Desjardins, G., Rusu, A. A., … & Hadsell, R. (2017). Overcoming catastrophic forgetting in neural networks. Proceedings of the National Academy of Sciences, 114(13), 3521–3526.
15. Luo, H., Gu, Y., Liao, X., Lai, S., & Jiang, W. (2020). Bag of tricks and a strong baseline for image-based person re-identification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (pp. 420–421).
16. Voigt, P., & von dem Bussche, A. (2017). The EU General Data Protection Regulation (GDPR): A practical guide. Springer.
17. Shokri, R., & Shmatikov, V. (2015). Privacy-preserving deep learning. Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security, 1310–1321.
18. Goodfellow, I. J., Shlens, J., & Szegedy, C. (2015). Explaining and harnessing adversarial examples. In International Conference on Learning Representations (ICLR).
19. Strubell, E., Ganesh, A., & McCallum, A. (2019). Energy and policy considerations for deep learning in NLP. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (pp. 3645–3650).
20. Zanella, A., Bui, N., Castellani, A., Vangelista, L., & Zorzi, M. (2014). Internet of things for smart cities. IEEE Internet of Things Journal, 1(1), 22–32.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 International Journal of Artificial Intelligence Engineering and Systems
This work is licensed under a Creative Commons Attribution 4.0 International License.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
