Vision Transformers, Foundation Models, and Agentic AI for Plant Disease and Pest Recognition

Vision transformers and multimodal foundation models are rapidly transforming automated plant disease and pest recognition, moving beyond traditional CNN-based image classification toward richer, context-aware diagnostic systems. At the same time, agentic AI is emerging as a natural extension of this progress: AI agents built on foundation models can integrate visual perception, domain knowledge, external tools, and multi-step reasoning to support more comprehensive disease and pest diagnosis.This Research Topic invites contributions that advance vision transformer and foundation-model architectures for plant disease and pest recognition, and extend them toward agentic, reasoning-capable AI systems for plant health diagnostics. The focus is on rigorous method development, benchmarking, and validation against real plant pathology and entomology data — not on field deployment, robotic hardware, or farm-management systems, which fall under the scope of our Sustainable and Intelligent Phytoprotection section.We welcome submissions on themes including, but not limited to:- Vision transformer architectures for plant disease and pest image recognition- Foundation models and large-scale pretraining strategies applied to plant pathology and entomology imagery- Agentic AI and AI agents for multimodal perception, reasoning, and decision support in plant disease and pest diagnosis- Retrieval-augmented and knowledge-augmented diagnostic reasoning grounded in plant pathology data- Tool-augmented AI agents for autonomous, multi-stage disease and pest analysis- Benchmarking and comparative evaluation of vision transformers, foundation models, and agentic pipelines against conventional CNN-based methods- Explainability, interpretability, and biological validation of AI-driven plant disease and pest diagnostic systems- Multimodal data integration (imagery, textual, and domain knowledge) for plant disease and pest recognitionSubmissions should provide genuine methodological advances, full experimental detail, and comparison with existing approaches, in line with the Technical Advances in Plant Science section’s requirements. Topic editors: Guoxiong Zhou, Liujun Li, Weiwei Cai Source: https://www.frontiersin.org/research-topics/85385/vision-transformers-foundation-models-and-agentic-ai-for-plant-disease-and-pest-recognition