Related Experiment Video
Updated: Jan 8, 2026
13:19
Deep Neural Networks for Image-Based Dietary Assessment
Published on: March 13, 2021
9.9K
Deep learning-based disease detection in potato and mango leaves: a comparative study of CNN, AlexNet, ResNet, and
Utkarsh Mishra1, Ansh Pandey1, Logeswari G2
1Vellore Institute of Technology, Chennai, Tamil Nadu, India.
Scientific Reports
|December 23, 2025
Summary
Deep learning models accurately identify potato and mango plant diseases using leaf images. EfficientNet achieved the highest accuracy, showing promise for smart farming and early disease detection.
Area of Science:
- Agricultural Science
- Computer Science
- Artificial Intelligence
Background:
- Accurate plant disease detection is vital for global food security.
- Deep learning (DL) offers potential for automated plant health monitoring.
Purpose of the Study:
- To develop and evaluate DL models for automatic disease identification in potato and mango leaves.
- To compare the performance of four DL architectures: CNN, AlexNet, ResNet, and EfficientNet.
Main Methods:
- Utilized two public datasets: PlantVillage Potato Leaf Disease and Kaggle Mango Leaf Disease.
- Pre-processed and augmented images, splitting into 80:20 train-test sets.
- Evaluated four DL architectures for multi-class disease classification.
Main Results:
- Baseline CNN achieved 92.61% test accuracy.
- ResNet demonstrated efficient convergence with 96.7% validation accuracy.
- EfficientNet outperformed others, achieving 97.8% validation accuracy with no overfitting.
Conclusions:
- DL models, particularly EfficientNet, show high accuracy and generalization for plant disease diagnosis.
- This research supports early remediation and precision agriculture through scalable smart farming solutions.
