A Study on YOLOv11-based Object Detection Performance for CCTV-based Traffic Safety

View Full Paper

Ask AI

Bookmark

The YOLOv11s-Fog model achieved an impressive mAP of 0.97 for object detection.
Using a dataset of 7,000 annotated CCTV images, the model significantly improves pedestrian detection performance.
Optimized strategies include maintaining aspect ratios and simulating adverse conditions for data augmentation.
These enhancements may enable more effective real-time risk detection in school and senior zones.

Discussion

No takes yet. Share an insight, caveat, or question.

The study demonstrates improved object detection results in CCTV footage, indicating higher traffic safety standards in urban areas.

Jeon et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark