A Hybrid Approach Based on ViT and Capsule Networks for Classification of Agricultural Plant Diseases

Key Points

• The classification approach shows superior accuracy compared to traditional methods in agricultural plant diseases.
• Using hybrid techniques like ViT and capsule networks enhances performance, achieving notable improvements in reliability.
• The method involves machine learning models that combine visual information and spatial hierarchies for better results.
• This research highlights the potential for improved agricultural practices, indicating a need for wider applicability in various crops.