Emerging Paradigms: A Federated Learning CNN Approach for Tomato Leaf Disease Detection

Abstract

This study examines how a Convolutional Neural Network (CNN) model based on federated learning may be used to categorize tomato leaf diseases. Four customers from each of the five classes-four representing the four tomato leaf illnesses and one representing the healthy class-are included in the research. Precision, Recall, F1-Score, and Accuracy are the metrics used to assess the model's performance. With overall accuracies ranging between 98% and 99%, the data show consistently strong performance across all customers and courses. Most Precision, Recall, and F1-Score values are over 94%, demonstrating the model's ability to accurately identify healthy and diseased tomato leaf pictures with few misclassifications. In all three averaging techniques (Macro average, Weighted average, and Micro average), Client 2 regularly surpasses the other clients, demonstrating its better performance in the categorization job. This study emphasizes how federated learning may be used to build reliable and precise models for categorizing tomato leaf diseases while maintaining data privacy. It is clear from the consistent results across all customers and classes that this strategy may be used in the agricultural industry and other privacy-sensitive industries. The findings provide insightful information for further study and practical federated learning applications in comparable settings. This study offers a CNN model for categorising tomato leaf diseases based on federated learning. This study highlights the potential of federated learning in developing reliable and accurate models for categorising tomato leaf disease while maintaining data privacy.

Topics & Concepts