Research on the Detection Technique of Situation Elements in Obscure Overlapping Scenes

Table 1

Comparison of the average accuracy rates of YOLOv3, SE-YOLO, and SEDNet.

	YOLOv3(%)	SE-YOLO(%)	SEDNet(%)
Submarine	89.09	94.10	99.12
UAV	77.20	80.00	92.93
Dog	90.03	90.86	90.65
Tank	65.71	79.87	89.20
Warship	72.96	83.10	87.92
Fighter	72.52	76.58	82.46
Soldier	61.34	73.63	80.38
Helicopter	55.98	75.33	78.49
Truck	48.87	67.41	72.55
Gun	38.53	48.90	64.64
mAP	67.22	76.98	83.83