Federated Learning for Crop Disease Detection Without Sharing Farm Data

Rina Kusuma; Hendra Wijaya

doi:10.5555/fji-demo-005

Open Access DOI: 10.5555/fji-demo-005 How to use a DOI?

Federated Learning for Crop Disease Detection Without Sharing Farm Data

Rina Kusuma^* iD¹, Hendra Wijaya²

^*Corresponding author: Rina Kusuma

¹Universitas Cakrawala
²Institute of Technology Nusantara

Case Study · Pages 9–16 · Sustainable Technology Proceedings, Proceedings of the 2nd International Conference on Sustainable Technology (ICST 2026) · Published 2026-06-15 · 242 views · 98 downloads

Article history: Received 2026-04-10 · Accepted 2026-05-22 · Available online 2026-06-15

Abstract

Farmers are reluctant to share crop imagery due to privacy concerns. We train a federated disease-detection model across 40 farms without centralizing images. The federated model reaches 91.4% accuracy, within 1.2 points of a centralized baseline, demonstrating that privacy-preserving training is viable for agricultural AI.

Keywords: federated learning; agriculture; privacy; computer vision

⇩ Download PDF

Cite this article

Rina Kusuma, Hendra Wijaya (2026). Federated Learning for Crop Disease Detection Without Sharing Farm Data. Sustainable Technology Proceedings, 2, 9–16. https://doi.org/10.5555/fji-demo-005

Download the citation in your preferred format:

BibTeX RIS

This is an open access article distributed under the CC BY-NC 4.0 license. CC BY-NC 4.0

DOI: 10.5555/fji-demo-005
Article type: Case Study
Pages: 9–16
Published: 2026-06-15
Metrics: 242 views · 98 downloads