Agricultural AI Alliance: Federated Learning CNNs for Papaya Leaf Disease Detection

However, precision agriculture has successfully used Federated Learning (FL) and Convolutional Neural Networks (CNNs), especially in plant pathology. This paper presents a novel FL-CNN model currently used to classify papaya leaf diseases into four levels. This model enhances early disease detection and facilitates specific agricultural interventions. We employed federated averaging, wherein the local information was consolidated into a single global model whereby client data was secured and confidentiality maintained as collective intelligence prevailed. The result analysis showed that macros, micros, and weighted average precision-recall scores were 0.98621357. The macro, micro, and weighted averages in the model for client vl_1 are 91.95% and 91. Client vl_2 demonstrated considerable improvement with 94.25% and 94.21 % mean scores. Client vl_3 showed further progress, averaging 96.42 %, 96.39%, and 70%. The learner identified by vl_ 4 demonstrated continuous improvement, having mean scores of 96.75%, 96.74, and 96.83%. It brings out the fact that the FL strategy is effective. This pattern was at its peak in client vl_5, which reached the highest mean values of 97.56%, 97.S38% and besides, they confirm that the FL-CNN model adapts to changes and learns slowly from gradually added data. Federated averaging has proven its capability to incorporate local data insights into a global data framework, improving scalability with agricultural district surveillance privacy awareness. The accuracy of the FL-CNN model for classifying papaya leaf disease from all severity levels is a vital factor significantly impacting the treatment of this disorder. It offers a new way of perceiving and combating the threat.