Deep Learning-Based Transfer Learning for Classification of Cassava Disease

TL;DR

This study compares four CNN architectures for cassava disease classification, finding EfficientNet-B3 to be the most accurate, demonstrating the potential of deep learning in digital agriculture applications.

Contribution

It provides a performance comparison of CNN models on cassava disease images, highlighting EfficientNet-B3's effectiveness in this domain.

Findings

EfficientNet-B3 achieved 87.7% accuracy.

The study addresses class imbalance with appropriate metrics.

EfficientNet-B3 outperformed other architectures in this task.

Abstract

This paper presents a performance comparison among four Convolutional Neural Network architectures (EfficientNet-B3, InceptionV3, ResNet50, and VGG16) for classifying cassava disease images. The images were sourced from an imbalanced dataset from a competition. Appropriate metrics were employed to address class imbalance. The results indicate that EfficientNet-B3 achieved on this task accuracy of 87.7%, precision of 87.8%, revocation of 87.8% and F1-Score of 87.7%. These findings suggest that EfficientNet-B3 could be a valuable tool to support Digital Agriculture.