PCE-FL: A Personalized, Clustered, and Communication-Efficient Federated Learning Framework for Robust Tomato Leaf Disease Detection
Open Access
- 6 May 2026
- journal article
- research article
- Published by MDPI AG in Agriengineering
- Vol. 8 (5) , 182
Abstract
Tomato leaf diseases represent a persistent threat to global food security, causing annual crop losses of 20% to 40%. Although deep learning models achieve accuracies exceeding 95% in centralized settings, their deployment across distributed farms is constrained by data privacy concerns, communication bottlenecks, and heterogeneous data quality. This paper proposes Personalized, Clustered, and Communication-Efficient Federated Learning (PCE-FL), a framework that integrates three synergistic components: (1) server-side client clustering to group farms with similar data distributions for personalized model training; (2) federated knowledge distillation to reduce communication overhead by over 91%; and (3) reputation-based aggregation to ensure robustness against unreliable contributions. Extensive experiments on realistic non-IID simulations of the PlantVillage tomato dataset Dirichlet() demonstrate that PCE-FL achieves 89.1% accuracy under extreme heterogeneity (), surpassing FedAvg by 10.9 and IFCA by 4.8 percentage points, while maintaining a 91% reduction in communication cost. All improvements are statistically significant (). These results advance the practical deployment of privacy-preserving collaborative AI in resource-constrained agricultural environments.
Keywords
This publication has 40 references indexed in Scilit:
- AGRIFOLD: AGRIculture Federated learning for Optimized Leaf disease DetectionExpert Systems with Applications, 2025
- A clustering-based federated deep learning approach for enhancing diabetes management with privacy-preserving edge artificial intelligenceHealthcare Analytics, 2025
- Issues in federated learning: some experiments and preliminary resultsScientific Reports, 2024
- Adaptive federated learning for resource-constrained IoT devices through edge intelligence and multi-edge clusteringScientific Reports, 2024
- A state-of-the-art survey on solving non-IID data in Federated LearningFuture Generation Computer Systems, 2022
- Identification of rice plant diseases using lightweight attention networksExpert Systems with Applications, 2021
- Factors influencing the use of deep learning for plant disease recognitionBiosystems Engineering, 2018
- Deep learning models for plant disease detection and diagnosisComputers and Electronics in Agriculture, 2018
- Advanced methods of plant disease detection. A reviewAgronomy for Sustainable Development, 2014
- A systematic analysis of performance measures for classification tasksInformation Processing & Management, 2009