Enhancing Explainability in Plant Disease Classification using Score-CAM: Improving Early Diagnosis for Agricultural Productivity

This study deals with the issue of explainability in the classification of plant diseases by deep learning methods. In particular, the explainability of deep learning models is shown using the Score-CAM method. Score-CAM is a method used to identify important regions that affect the classification decision of the model. In this study, the Score-CAM method was applied to deep learning models analyzing images of plant leaves, and it was provided to explain the classification decisions of the model and identify important regions for the diagnosis of the disease. As a result, explainability techniques help to achieve more effective and accurate results in the early diagnosis of plant diseases. This in turn helps to reduce the negative effects of plant diseases on economic and food security by increasing agricultural productivity.