Paper
19 October 2023 Tomato leaf disease detection based on peer-to-peer federated learning in wireless networks
Ke Liu
Author Affiliations +
Proceedings Volume 12709, Fourth International Conference on Artificial Intelligence and Electromechanical Automation (AIEA 2023); 127094Q (2023) https://doi.org/10.1117/12.2685057
Event: Fourth International Conference on Artificial Intelligence and Electromechanical Automation (AIEA 2023), 2023, Nanjing, China
Abstract
As a research highlight of plant disease control, the tomato leaf disease is the one of the most important reasons resulting in tomato yield decline. Since the relevant studies not considering transmission errors in wireless environment and the privacy of datasets, a method based on P2P (Peer-to-Peer) federated learning in wireless networks to detect tomato leaf disease is proposed, which applies ARQ (Automatic Repeat-reQuest) protocols for reliability and ensures the privacy of datasets by transmitting model parameters. In this paper, the training/validation loss, precision and false alarm are analyzed for five different kinds of tomato leaf diseases images. Results show that accuracy and precision reach 90 percent and the average of false alarm is three percent for all kinds, where gap is below five percent compared to centralized learning method, thus proving the effectiveness and high performance of the proposed method and providing a reference to tomato leaf disease control in reality with machine learning.
© (2023) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Ke Liu "Tomato leaf disease detection based on peer-to-peer federated learning in wireless networks", Proc. SPIE 12709, Fourth International Conference on Artificial Intelligence and Electromechanical Automation (AIEA 2023), 127094Q (19 October 2023); https://doi.org/10.1117/12.2685057
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Machine learning

Diseases and disorders

Education and training

Automatic repeat request

Data modeling

Data privacy

Wireless communications

Back to Top