Scholarlink

Motorcycle riders'safety risks greatly increase when they ride without helmet. Despite the increasing number of accidents and fatalities, there is a worrying lack of understanding and practice about the usage of helmet by two-wheeler riders. To recognize helmet in real time for all weather and daylight conditions is a major task. The study suggests real-time helmet violation detecting system. We used a Generative Artificial Intelligence model with an object detection model,i.e.,YOLOv11.Stablediffusionmodel used as a Generative AI model to generate the data for all weather conditionsandYOLOv11is the YOLO(You Only Look Once) series model, which is identify the helmet. Our system produces the mAP.05 score of 0.955, mAP.05–0.95 score of 0.866 in validation and mAP.05 score of 0.822, mAP .05–0.95 score of 0.757 in testing, which is the best result in comparison with other YOLO series models.

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

Real-Time Helmet Detection Using Synthetic Dataset and Yolov11

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Kapil Kumar Pandey,
  • S. K. Singh,
  • J. N. Singh

摘要

Motorcycle riders'safety risks greatly increase when they ride without helmet. Despite the increasing number of accidents and fatalities, there is a worrying lack of understanding and practice about the usage of helmet by two-wheeler riders. To recognize helmet in real time for all weather and daylight conditions is a major task. The study suggests real-time helmet violation detecting system. We used a Generative Artificial Intelligence model with an object detection model,i.e.,YOLOv11.Stablediffusionmodel used as a Generative AI model to generate the data for all weather conditionsandYOLOv11is the YOLO(You Only Look Once) series model, which is identify the helmet. Our system produces the mAP.05 score of 0.955, mAP.05–0.95 score of 0.866 in validation and mAP.05 score of 0.822, mAP .05–0.95 score of 0.757 in testing, which is the best result in comparison with other YOLO series models.