Ensemble transfer learning meets explainable AI: A deep learning approach for leaf disease detection

Global food security is threatened by plant diseases and manual detection methods are often labor-intensive and time-consuming. Deep learning offers a promising solution by enabling early and accurate detection of leaf diseases. This study presents a novel deep-learning model designed to address the challenges of real-world leaf disease identification. To enhance the model's robustness, we incorporated six datasets (LD, LD1, LD2, LD3, LD4, LD5) which include image augmentation techniques, like flipped versions (LD1) and controlled noise (LD2, LD3). Additionally, we introduced new datasets with additional noise types (LD4) and real-world scenarios (LD5). To further improve accuracy, we employed an ensemble approach, combining MobileNetV3_Small and EfficientNetV2B3 with weighted voting. Our model achieved exceptional performance, surpassing 94 % accuracy on imbalanced data (LD) and exceeding 99 % on balanced, high-quality data (LD1). Even in noisy environments (LD2, LD3, LD4, LD5), our model consistently outperformed other approaches, maintaining an accuracy rate above 90 %. To ensure transparency and interpretability, we utilized Explainable AI (LIME) to visualize the model's decision-making process. These results demonstrate the potential of our model as a reliable and accurate tool for leaf disease detection in practical agricultural settings. • Ensemble Learning: Combining multiple pre-trained models (MobileNetV3_Small, EfficientNetV2B3) significantly improved accuracy and robustness. • Data Augmentation: Flipping and noise addition improved model generalization and performance. • Model Interpretability: LIME (Explainable AI technique) enhanced model understanding and trust. • Robustness to Noise: Model performed well in noisy environments (varied brightness, contrast, blur, clutter, noise). • Real-world Applicability: Model's accuracy on LD5 dataset shows its potential for practical agriculture.