XAI-MaizeNet: An Explainable Deep Learning Framework for Transparent and Interpretable Maize Leaf Disease Detection and Classification

Abstract

Early and precise diagnosis of maize leaf diseases is vital to reduce crop losses and to ensure sustainable farming production. Nevertheless, traditional Deep Learning models are usually “black boxes,” making them unreliable and unpopular in real-world agricultural decision-making processes. To provide transparent and interpretable maize leaf disease detection and classification, this paper introduces XAI-MaizeNet, an explainable deep learning framework. The framework encompasses image pre-processing, data augmentation, deep convolutional feature extraction, disease classification, and explainable artificial intelligence techniques, creating a comprehensive diagnostic pathway. The deep CNN learns discriminative spatial and textural features of the disease symptoms, and explainability mechanisms produce visual attribution maps and feature-level interpretation maps to highlight the image regions that are associated with the classification. Common performance metrics such as accuracy, precision, recall, F1 score, specificity and area under the receiver operating characteristic curve are applied in evaluating the proposed framework. Experimental evaluations show that XAI-MaizeNet demonstrates accurate classification accuracy and gives human-readable explanations of disease prediction. This framework enhances model transparency, allows for expert validation, and enables the creation of trustworthy, AI-driven disease diagnosis, providing a practical decision support tool for precision agriculture and intelligent crop health management.