Finger Millet Leaf Disease Detection and Prediction with Machine Learning

Abstract

The increasing threat of plant diseases, especially in staple crops like Finger Millet, is posing significant challenges to global food security and agricultural sustainability sets. The traditionally relied-upon methods for disease detection, such as visual inspection and manual sampling, are not only labor-intensive but also prone to errors, often leading to delayed interventions that result in significant crop yield losses. With the advent of modern technology, there has been an urgent need to develop more accurate, efficient, and automated methods for early disease detection and prediction. The current methodologies fail to capture the complex spatial and temporal dynamics of plant diseases, leading to suboptimal disease detection and prediction accuracy sets. In light of these challenges, this work explores and proposes several advanced methodologies that integrate cutting-edge machine learning techniques to predict and detect diseases in Finger Millet leaves. Specifically, this study presents four distinct frameworks that leverage the power of graph networks, dynamic temporal models, multimodal fusion, and advanced neural networks for precise and timely disease management. newline