August 23, 2025ForestsOpen Access

F3-YOLO: A Robust and Fast Forest Fire Detection Model

View Full Paper

Ask AI

Bookmark

F3-YOLO achieves superior accuracy with a highest mAP@50 of 68.5% on the dataset while needing only 5.4 GFLOPs of computational cost.
The model incorporates conditionally parameterized convolution and frequency domain-based self-attention to enhance detection capabilities.
Structured pruning is applied in F3-YOLO to minimize computational overhead, making it accessible for edge devices.
Overall, this model supports real-time forest fire early warning systems by ensuring low latency and reliability.

Discussion

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

F3-YOLO demonstrates superior accuracy and efficiency for forest fire detection using advanced algorithms and techniques.

Zhang et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark