FusionNet-ATT: Hybrid Transfer Learning Model with Attention Mechanism for Potato Leaf Disease Detection

Both datasets used in this study, namely Dataset 1 ((D_1)) and Dataset 2 ((D_2)), were obtained from Kaggle to ensure accessibility, reproducibility, and ease of use for future research. The (D_1) dataset represents potato leaf images captured under relatively controlled conditions, while the (D_2) dataset consists of images collected in uncontrolled, real-world environments, introducing variations such as lighting, background noise, and camera angles. These datasets were selected deliberately to evaluate the robustness and generalization capability of the proposed FusionNet-ATT model across both ideal and challenging scenarios. For transparency and to support further validation, both datasets have also been uploaded alongside this work.