Plant Disease Identification using Deep Learning Approach

This study explores the performance of SOTA models for plant disease identification, extending beyond accuracy to include in-depth explainability through visualization. Additionally, three novel models (PlantAIM, FF-ViT, and CL-ViT) are proposed to address existing challenges in the field. For the research community, this study provides a comprehensive performance benchmark and introduces new methodologies to tackle current limitations. For the broader society, the findings enhance public trust in automated plant disease identification models and contribute to more robust models adaptable to various tasks and environments.