AI Meets Climate: Advanced Transformer Multimodal Model for Early and Accurate Plant Disease Detection

The world's food security and biodiversity are seriously threatened by the rising frequency and severity of plant diseases, which are made worse by climate change. Plant diseases must be accurately and promptly detected to reduce these hazards, especially when environmental changes are likely to occur. This paper presents a unique method to enhance plant disease classification across 22 classes by integrating structured climatic data with image data. This paper provides a multimodal fusion model with improved prediction accuracy by fusing Swin Transformer models that support image data with an RNN model that promotes climatic data. Our proposed hybrid model integrates environmental parameters into the disease prediction framework, addressing the issues posed by climate change on the evolution of plant diseases. In addition, we propose a stacking transformer model that combines the advantages of the ViT and Swin Transformer as base models with a Random Forest (RF) as a meta-learner to enhance the classification of plant diseases using image data. The findings show that the multimodal and stacking transformer model works better than conventional models, attaining higher recall, accuracy, precision, and F1-score. We emphasize how crucial it is to use multimodal data integration and sophisticated hybrid models to create a reliable plant disease detection system.s that can adjust to the effects of climate change