Early and precise diagnosis of corn leaf diseases is essential for maximizing crop yield and ensuring food security. Current progressions in deep learning have significantly boosted the ability to recognize and classify countless crop diseases. In this study, we propose an explainable parallel convolutional neural network (CNN) framework for corn leaf disease detection and classification. The model services a two-stream feature extraction strategy: one stream utilizes Dense Blocks to enrich deep feature representation, while the other incorporates Squeeze-and-Excitation (SE) blocks to emphasize relevant features through adaptive channel recalibration. Experimental evaluations on benchmark corn disease datasets show that the proposed model performs better than current deep learning techniques in terms of trainable parameters, classification accuracy, and precision. Further more, explainability techniques are integrated to provide transparency and foster belief in the model’s estimates. This work pays to the development of intelligent, accurate, and interpretable deep learning solutions for precision agriculture.