Jonah Mubuuke Kyagaba’s research while affiliated with Makerere University and other places

Banana plants are susceptible to the dangerous fungal disease known as Black Sigatoka, which has a negative impact on global economies. Early detection and timely intervention are crucial for preventing the spread of the disease. In recent years, machine learning (ML) has shown great potential for detecting and diagnosing plant diseases, including Black Sigatoka. However, the lack of transparency and interpretability of ML models raises concerns about their use. In this paper, we propose explainable AI approaches for Black Sigatoka detection using Local Interpretable Model-Agnostic Explanations (LIME) and Integrated Gradients. Our methodology involves the utilization of Mobilenet V2 and AlexNet models which are trained on an extensive dataset of banana leaf images and generating explanations to provide a better understanding of the CNN’s decision-making process. We demonstrate the effectiveness of our approach through extensive experiments and show that it outperforms existing state-of-the-art methods for Black Sigatoka detection. Our approach not only provides accurate and interpretable results but also promotes responsible AI practices for plant disease diagnosis.