FFA-YOLOv7: Improved YOLOv7 Based on Feature Fusion and Attention Mechanism for Wearing Violation Detection in Substation Construction Safety

Table 3

Comparison of performance among state-of-the-art object detection models.

Model	P (%)	R (%)	mAP (%)	F1 (%)	Speed (ms)
YOLOv5-s	94.62	95.32	96.45	94.93	8.6
YOLOv5-m	95.14	95.33	96.67	95.22	12.0
YOLOv5-l	95.18	96.02	96.71	95.53	14.0
YOLOv7	95.76	95.92	97.64	95.84	8.8
YOLOv7-x	95.55	96.45	97.82	96.00	11.5
FFA-YOLOv7	95.92	97.13	98.16	96.50	9.6