Cross-dataset evaluation of deep learning models for plant pest and disease diagnosis

Abstract

Deep learning models have shown significant potential for plant pest and disease (PPD) diagnosis; however, their real-world effectiveness is often limited by variability between datasets, where models trained on one dataset perform poorly on others collected under different conditions. In this study, I evaluated the cross-dataset generalization of widely used deep learning architectures, including ResNet, EfficientNet, Inception, and MobileNet, across multiple tomato pest and disease datasets. As expected, models trained and tested on the same dataset achieved high performance. However, substantial performance degradation occurred when these models were tested on different datasets, highlighting the challenges posed by dataset variability. This trend was consistent across all evaluated architectures, indicating that changing the model architecture alone is insufficient to address these issues. The findings emphasize the need for more diverse and representative datasets to better capture variability in agricultural data and enhance the practical deployment of deep learning models for PPD diagnosis.

Code and data availability

The paper's cross-dataset evaluation uses three public tomato pest/disease image datasets (PlantVillage, Tomato-Village, Tomato Leaf Disease), each with an explicit public URL in the data availability statement. No author analysis code or trained models are deposited.