Related Experiment Video
Updated: Mar 3, 2026
08:20
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
2.7K
Lightweight deep learning for tomato disease detection: trends, challenges, and edge AI perspectives
Harshinisree Gunasekaran1, Sujatha Rajkumar2, Lincy Kirubhadharsini B3
1School of Biosciences and Technology (SBST), Vellore Institute of Technology (VIT), Vellore, India.
Frontiers in Plant Science
|March 2, 2026
Summary
This study explores lightweight deep learning and edge AI for accurate tomato disease detection, achieving 99.9% accuracy. It proposes an AI framework combined with microbial biocontrol for sustainable, region-specific crop management.
Area of Science:
- Agricultural Science
- Computer Science
- Artificial Intelligence
Background:
- Tomato production is significantly impacted by diseases, necessitating efficient detection and sustainable management strategies.
- Precision agriculture requires advanced tools for early disease diagnosis to mitigate crop quality and yield losses.
Purpose of the Study:
- To review and evaluate lightweight deep learning models and edge AI for tomato disease detection.
- To propose an integrated framework combining AI-driven diagnosis with microbial biocontrol for sustainable agriculture.
Main Methods:
- Comprehensive literature survey on deep learning models (CNNs, transformers), optimization techniques (pruning, quantization, distillation), and explainable AI.
- Experimental validation using MobileNetV2 and EfficientNetB0 on prevalent tomato diseases in Tamil Nadu.
Main Results:
- MobileNetV2 and EfficientNetB0 achieved 99.9% accuracy and a macro-F1 score of nearly 0.99 in detecting tomato diseases.
- A novel framework integrating AI diagnosis with microbial biocontrol recommendations was developed.
Conclusions:
- Lightweight deep learning and edge AI show significant promise for practical, real-time tomato disease detection in precision agriculture.
- The proposed integrated framework offers an eco-friendly, region-specific solution for resilient and farmer-friendly agricultural systems.
