Smart City Gnosys
Smart city article details
| Title | Communication Efficient Model-Aware Federated Learning For Visual Crowd Counting And Density Estimation In Smart Cities |
|---|---|
| ID_Doc | 14905 |
| Authors | Armacki A.; Milosevic N.; Bajovic D.; Kar S.; Jakovetic D.; Bakhtiarnia A.; Esterle L.; Muscat A.; Festi T. |
| Year | 2023 |
| Published | European Signal Processing Conference |
| DOI | http://dx.doi.org/10.23919/EUSIPCO58844.2023.10289938 |
| Abstract | Federated learning (FL) is an attractive paradigm where a number of users can improve their local models via sharing trained models or model increments with a central server, while the users' data is kept private. However, when model sizes are huge, FL incurs a significant communication overhead. In such scenarios, strategies that perform user sampling, so that only informative users communicate their models to the server, are desired. In this paper, we make several contributions on user sampling in FL. On the theoretical side, we consider a general framework that exhibits user heterogeneity across several dimensions: activation probabilities, gradient noise variance, number of updates per epoch, and communication channel quality. In this setting, we derive convergence rate of the FedAvg method. The rate explicitly characterizes the effects of heterogeneity and enables us to derive optimal user sampling probabilities in an offline setting, when the sampling probabilities are pre-computed. We then show how these derived probabilities naturally connect with existing optimized sampling strategies in an adaptive-online setting. On the practical side, we study visual crowd counting (VCC) as a representative deep learning application with huge-sized models. We provide an implementation of the FL system over real-world data across three pilot sites - Valetta, Trento and Novi Sad. The evaluation results demonstrate significant model accuracy benefits through employing FL over the multiple cities, and significant communication savings via non-uniform user sampling strategies. © 2023 European Signal Processing Conference, EUSIPCO. All rights reserved. |
| Author Keywords | Communication efficient protocol; Federated learning; Optimized client selection; Visual crowd counting |
Similar Articles
| Id | Similarity | Authors | Title | Published |
|---|---|---|---|---|
26337  | 0.904 | Pang Y.; Ni Z.; Zhong X. | Federated Learning For Crowd Counting In Smart Surveillance Systems | IEEE Internet of Things Journal, 11, 3 (2024) |
| 26332 | 0.881 | Jiang Y.; Cong R.; Shu C.; Yang A.; Zhao Z.; Min G. | Federated Learning Based Mobile Crowd Sensing With Unreliable User Data | Proceedings - 2020 IEEE 22nd International Conference on High Performance Computing and Communications, IEEE 18th International Conference on Smart City and IEEE 6th International Conference on Data Science and Systems, HPCC-SmartCity-DSS 2020 (2020) |
| 16111 | 0.868 | Ilyas N.; Shahzad A.; Kim K. | Convolutional-Neural Network-Based Image Crowd Counting: Review, Categorization, Analysis, And Performance Evaluation | Sensors (Switzerland), 20, 1 (2020) |
| 16357 | 0.867 | Solmaz G.; Baranwal P.; Cirillo F. | Countmein: Adaptive Crowd Estimation With Wi-Fi In Smart Cities | 2022 IEEE International Conference on Pervasive Computing and Communications, PerCom 2022 (2022) |
| 16685 | 0.86 | Hu Y.; Liu Y.; Cao G.; Wang J. | Crowdcl: Unsupervised Crowd Counting Network Via Contrastive Learning | IEEE Internet of Things Journal, 12, 12 (2025) |
| 26359 | 0.859 | Gandhi M.; Singh S.K.; Ravikumar R.N.; Vaghela K. | Federated Learning In Secure Smart City Sensing: Challenges And Opportunities | Edge of Intelligence: Exploring the Frontiers of AI at the Edge (2025) |
| 21817 | 0.857 | Siddiqa A.; Khan W.Z.; Alkinani M.H.; Aldhahri E.A.; Khan M.K. | Edge-Assisted Federated Learning Framework For Smart Crowd Management | Internet of Things (Netherlands), 27 (2024) |
| 50724 | 0.855 | De Cock L.; Verstockt S.; Vandeviver C.; Van de Weghe N. | Smart Crowd Management: The Data, The Users And The Solution | Leibniz International Proceedings in Informatics, LIPIcs, 240 (2022) |
| 21192 | 0.854 | Song S.; Du S.; Song Y.; Zhu Y. | Dualpfl: A Dual Sparse Pruning Method With Efficient Federated Learning For Edge-Based Object Detection | Applied Sciences (Switzerland), 14, 22 (2024) |
| 11921 | 0.853 | Xie Z.; Shao F.; Mu B.; Chen H.; Jiang Q.; Lu C.; Ho Y.-S. | Bgdfnet: Bidirectional Gated And Dynamic Fusion Network For Rgb-T Crowd Counting In Smart City System | IEEE Transactions on Instrumentation and Measurement, 73 (2024) |