在聯邦學習架構中，當多個用戶端的資料分佈存在非獨立同分佈(Non-IID)時，模型的語意一致性與整體泛化能力將面臨嚴峻挑戰。為改善此問題，本論文提出一套整合特徵語意對齊與正規化調控機制的聯邦學習架構，應用於跨域交通資料的分類與物件偵測。系統透過實例層級的特徵提取與對抗式語意對齊策略，引導模型學習具有遷移性的共享語意表示；同時導入特徵正規化項控制模型特徵空間的多樣性與穩定性。
本論文進一步設計一套以群內特徵分散度（Within-Cluster Dispersion, WCD）為核心之自適應調控機制，根據各用戶端模型在目標域資料上的語意分佈表現，動態調整正規化強度，使學習過程能依據資料差異進行彈性適配。實驗涵蓋日夜轉換、天候變異與極端 domain 差異等多種跨域非IID情境，並以準確率與 mAP 作為評估指標。結果顯示，於具顯著資料分佈偏移之場景下，本研究方法相較於傳統聯邦平均與單一策略方法展現出更穩定之表現與泛化能力；而在資料差異較小之基準設定下，各方法表現差距有限，顯示所提出機制主要在異質環境中發揮其優勢，具備實務應用於跨域交通資料之潛力。

In federated learning (FL) systems, the presence of non-independent and identically distributed (Non-IID) data across clients poses significant challenges to semantic consistency and model generalization. To address this issue, this study proposes a federated training framework that integrates instance-level semantic alignment and feature regularization control, and applies it to cross-domain traffic data for both classification and object detection tasks. The framework employs instance-level feature extraction combined with adversarial semantic alignment to encourage the learning of transferable shared representations. In addition, feature-space regularization is introduced to enhance representation stability while preserving structural diversity.
Furthermore, we design an adaptive regulation mechanism centered on Within-Cluster Dispersion (WCD), which dynamically adjusts regularization strength according to each client model’s semantic distribution performance on the target domain. This allows the training process to flexibly adapt to heterogeneous data characteristics across clients. Experiments are conducted under multiple cross-domain Non-IID scenarios, including day-to-night transitions, weather variations, and extreme domain discrepancies. Evaluation results based on classification accuracy and mean Average Precision (mAP) indicate that the proposed method achieves more stable performance and improved generalization compared to conventional federated averaging and single-strategy baselines in settings with significant distribution shifts. Under near-IID baseline configurations, performance differences among methods remain limited, suggesting that the proposed mechanism primarily demonstrates its advantages in heterogeneous environments. These findings highlight the practical potential of the proposed framework for cross-domain traffic applications.

[1] B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y Arcas, "Communication-efficient learning of deep networks from decentralized data," in Artificial intelligence and statistics, 2017: PMLR, pp. 1273-1282.
[2] Y. Zhao, M. Li, L. Lai, N. Suda, D. Civin, and V. Chandra, "Federated learning with non-iid data," arXiv preprint arXiv:1806.00582, 2018.
[3] T. Li, A. K. Sahu, M. Zaheer, M. Sanjabi, A. Talwalkar, and V. Smith, "Federated optimization in heterogeneous networks," Proceedings of Machine learning and systems, vol. 2, pp. 429-450, 2020.
[4] J. Wang, Q. Liu, H. Liang, G. Joshi, and H. V. Poor, "Tackling the objective inconsistency problem in heterogeneous federated optimization," Advances in neural information processing systems, vol. 33, pp. 7611-7623, 2020.
[5] H. Wang, M. Yurochkin, Y. Sun, D. Papailiopoulos, and Y. Khazaeni, "Federated learning with matched averaging," arXiv preprint arXiv:2002.06440, 2020.
[6] D. A. E. Acar, Y. Zhao, R. M. Navarro, M. Mattina, P. N. Whatmough, and V. Saligrama, "Federated learning based on dynamic regularization," arXiv preprint arXiv:2111.04263, 2021.
[7] H. Wu and P. Wang, "Fast-convergent federated learning with adaptive weighting," IEEE Transactions on Cognitive Communications and Networking, vol. 7, no. 4, pp. 1078-1088, 2021.
[8] M. Asad, A. Moustafa, and T. Ito, "Fedopt: Towards communication efficiency and privacy preservation in federated learning," Applied Sciences, vol. 10, no. 8, p. 2864, 2020.
[9] N. Shoham et al., "Overcoming forgetting in federated learning on non-iid data," arXiv preprint arXiv:1910.07796, 2019.
[10] Y. Chen, W. Li, C. Sakaridis, D. Dai, and L. Van Gool, "Domain adaptive faster r-cnn for object detection in the wild," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 3339-3348.
[11] K. Saito, Y. Ushiku, T. Harada, and K. Saenko, "Strong-weak distribution alignment for adaptive object detection," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 6956-6965.
[12] Q. Cai, Y. Pan, C.-W. Ngo, X. Tian, L. Duan, and T. Yao, "Exploring object relation in mean teacher for cross-domain detection," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 11457-11466.
[13] X. Peng, Z. Huang, Y. Zhu, and K. Saenko, "Federated adversarial domain adaptation," arXiv preprint arXiv:1911.02054, 2019.
[14] D. Li and J. Wang, "Fedmd: Heterogenous federated learning via model distillation," arXiv preprint arXiv:1910.03581, 2019.
[15] P. J. Lu, C.-Y. Jui, and J.-H. Chuang, "FedDAD: Federated domain adaptation for object detection," IEEE Access, vol. 11, pp. 51320-51330, 2023.
[16] A. T. Nguyen, P. Torr, and S. N. Lim, "Fedsr: A simple and effective domain generalization method for federated learning," Advances in Neural Information Processing Systems, vol. 35, pp. 38831-38843, 2022.
[17] S. Gupta, V. Sutar, V. Singh, and A. Sethi, "FedAlign: Federated Domain Generalization with Cross-Client Feature Alignment," arXiv preprint arXiv:2501.15486, 2025.
[18] S. Dai et al., "Fedgp: Buffer-based gradient projection for continual federated learning," 2023.
[19] Y. Huang et al., "FedDA-TSformer: Federated Domain Adaptation with Vision TimeSformer for Left Ventricle Segmentation on Gated Myocardial Perfusion SPECT Image," arXiv preprint arXiv:2502.16709, 2025.
[20] P. Voigt and A. Von dem Bussche, "The eu general data protection regulation (gdpr)," A practical guide, 1st ed., Cham: Springer International Publishing, vol. 10, no. 3152676, pp. 10-5555, 2017.
[21] Z. Lu, H. Pan, Y. Dai, X. Si, and Y. Zhang, "Federated learning with non-iid data: A survey," IEEE Internet of Things Journal, 2024.
[22] C. T Dinh, N. Tran, and J. Nguyen, "Personalized federated learning with moreau envelopes," Advances in neural information processing systems, vol. 33, pp. 21394-21405, 2020.
[23] M. G. Arivazhagan, V. Aggarwal, A. K. Singh, and S. Choudhary, "Federated learning with personalization layers," arXiv preprint arXiv:1912.00818, 2019.
[24] E. Seo, D. Niyato, and E. Elmroth, "Resource-efficient federated learning with non-iid data: An auction theoretic approach," IEEE Internet of Things Journal, vol. 9, no. 24, pp. 25506-25524, 2022.
[25] Y. Cheng, L. Zhang, and A. Li, "GFL: Federated learning on non-IID data via privacy-preserving synthetic data," in 2023 IEEE International Conference on Pervasive Computing and Communications (PerCom), 2023: IEEE, pp. 61-70.
[26] I. Goodfellow et al., "Generative adversarial networks," Communications of the ACM, vol. 63, no. 11, pp. 139-144, 2020.
[27] L. Pinheiro Cinelli, M. Araújo Marins, E. A. Barros da Silva, and S. Lima Netto, "Variational autoencoder," in Variational methods for machine learning with applications to deep networks: Springer, 2021, pp. 111-149.
[28] R. Ye, M. Xu, J. Wang, C. Xu, S. Chen, and Y. Wang, "Feddisco: Federated learning with discrepancy-aware collaboration," in International Conference on Machine Learning, 2023: PMLR, pp. 39879-39902.
[29] J. Lu, H. Zhang, P. Zhou, X. Wang, C. Wang, and D. O. Wu, "FedLaw: Value-Aware Federated Learning With Individual Fairness and Coalition Stability," IEEE Transactions on Emerging Topics in Computational Intelligence, 2024.
[30] Y. Mi, Y. Mu, S. Zhou, and J. Guan, "Fedmdr: Federated model distillation with robust aggregation," in Asia-Pacific Web (APWeb) and Web-Age Information Management (WAIM) Joint International Conference on Web and Big Data, 2021: Springer, pp. 18-32.
[31] J. Liu, J. H. Wang, C. Rong, Y. Xu, T. Yu, and J. Wang, "Fedpa: An adaptively partial model aggregation strategy in federated learning," Computer Networks, vol. 199, p. 108468, 2021.
[32] C. Xu, Y. Qu, Y. Xiang, and L. Gao, "Asynchronous federated learning on heterogeneous devices: A survey," Computer Science Review, vol. 50, p. 100595, 2023.
[33] W. Wu, L. He, W. Lin, R. Mao, C. Maple, and S. Jarvis, "SAFA: A semi-asynchronous protocol for fast federated learning with low overhead," IEEE Transactions on Computers, vol. 70, no. 5, pp. 655-668, 2020.
[34] M. Fang, X. Cao, J. Jia, and N. Gong, "Local model poisoning attacks to {Byzantine-Robust} federated learning," in 29th USENIX security symposium (USENIX Security 20), 2020, pp. 1605-1622.
[35] R. Shokri, M. Stronati, C. Song, and V. Shmatikov, "Membership inference attacks against machine learning models," in 2017 IEEE symposium on security and privacy (SP), 2017: IEEE, pp. 3-18.
[36] A. El Ouadrhiri and A. Abdelhadi, "Differential privacy for deep and federated learning: A survey," IEEE access, vol. 10, pp. 22359-22380, 2022.
[37] H. Fang and Q. Qian, "Privacy preserving machine learning with homomorphic encryption and federated learning," Future Internet, vol. 13, no. 4, p. 94, 2021.
[38] V. Mugunthan, A. Polychroniadou, D. Byrd, and T. H. Balch, "Smpai: Secure multi-party computation for federated learning," in Proceedings of the NeurIPS 2019 Workshop on Robust AI in Financial Services, 2019, vol. 21: MIT Press Cambridge, MA, USA.
[39] F. Mo, H. Haddadi, K. Katevas, E. Marin, D. Perino, and N. Kourtellis, "PPFL: Privacy-preserving federated learning with trusted execution environments," in Proceedings of the 19th annual international conference on mobile systems, applications, and services, 2021, pp. 94-108.
[40] P. Ramanan and K. Nakayama, "Baffle: Blockchain based aggregator free federated learning," in 2020 IEEE international conference on blockchain (Blockchain), 2020: IEEE, pp. 72-81.
[41] Z. Wang, Z. Dai, B. Póczos, and J. Carbonell, "Characterizing and avoiding negative transfer," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 11293-11302.
[42] D.-H. Lee, "Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks," in Workshop on challenges in representation learning, ICML, 2013, vol. 3, no. 2: Atlanta, p. 896.
[43] Y. Grandvalet and Y. Bengio, "Semi-supervised learning by entropy minimization," Advances in neural information processing systems, vol. 17, 2004.
[44] J. Tian, J. Zhang, W. Li, and D. Xu, "VDM-DA: Virtual domain modeling for source data-free domain adaptation," IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 6, pp. 3749-3760, 2021.
[45] H. Xia, H. Zhao, and Z. Ding, "Adaptive adversarial network for source-free domain adaptation," in Proceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 9010-9019.
[46] S. Yang, Y. Wang, J. Van De Weijer, L. Herranz, and S. Jui, "Generalized source-free domain adaptation," in Proceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 8978-8987.
[47] A. Vaswani et al., "Attention is all you need," Advances in neural information processing systems, vol. 30, 2017.
[48] R. Girshick, "Fast r-cnn," in Proceedings of the IEEE international conference on computer vision, 2015, pp. 1440-1448.
[49] D. P. Kingma and M. Welling, "Auto-encoding variational bayes," ed: Banff, Canada, 2013.
[50] S. Ren, K. He, R. Girshick, and J. Sun, "Faster R-CNN: Towards real-time object detection with region proposal networks," IEEE transactions on pattern analysis and machine intelligence, vol. 39, no. 6, pp. 1137-1149, 2016.
[51] K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 770-778.
[52] F.-H. Chan, Y.-T. Chen, Y. Xiang, and M. Sun, "Anticipating accidents in dashcam videos," in Computer Vision–ACCV 2016: 13th Asian Conference on Computer Vision, Taipei, Taiwan, November 20-24, 2016, Revised Selected Papers, Part IV 13, 2017: Springer, pp. 136-153.
[53] W. Bao, Q. Yu, and Y. Kong, "Uncertainty-based traffic accident anticipation with spatio-temporal relational learning," in Proceedings of the 28th ACM International Conference on Multimedia, 2020, pp. 2682-2690.