A Deep Learning Attention Approach for Diagnosis of Rice Diseases

Rice diseases pose a major threat to the volume of crops and food security. It is especially vital to identify them at the early stage to apply effective management and prevention tactics. In this respect, convolutional neural networks (CNN), have been increasingly used due to their ability to automate the process of disease detection. In this paper, we combine attention mechanism to CNN to further improve the disease prediction. Experiment on rice disease dataset suggests the best results in terms of performance (93% accuracy) compare to prior methods.