Scholarlink

As the population of the earth grows, the demand for food grows proportionally. Early and cost-effective detection of plant diseases can result in less food loss. The current methods for image-based plant disease detection tend to fail in field conditions. The proposed pipeline uses an ensemble of an YOLOv8 model trained for disease detection and a disease detection module made of YOLOv8 (as the localizer) and Vision Transformer (as the classifier). The ensembling is performed with a method called Soft-NMS. Our pipeline performs disease detection with 46.12% mAP, beating the YOLOv8 by 14.72%, which detects with 31.4% mAP in the Plant-Doc dataset.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Ensembling YOLO and ViT for Plant Disease Detection

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Debojyoti Misra,
  • Suryansh Goel,
  • Tushar Sandhan

摘要

As the population of the earth grows, the demand for food grows proportionally. Early and cost-effective detection of plant diseases can result in less food loss. The current methods for image-based plant disease detection tend to fail in field conditions. The proposed pipeline uses an ensemble of an YOLOv8 model trained for disease detection and a disease detection module made of YOLOv8 (as the localizer) and Vision Transformer (as the classifier). The ensembling is performed with a method called Soft-NMS. Our pipeline performs disease detection with 46.12% mAP, beating the YOLOv8 by 14.72%, which detects with 31.4% mAP in the Plant-Doc dataset.