Aircraft Detection for Remote Sensing Image Based on Bidirectional and Dense Feature Fusion

Table 2

The contrastive results in the RSOD-DataSet.

Method	Backbone	AP (%)					FPS
		Aircraft	Oil tank	Overpass	Playground	mAP
SSD [7]	VGG-16	69.17	71.20	70.23	81.26	72.97	61.5
DSSD [33]	ResNet-101	72.12	72.49	72.10	83.56	75.07	6.1
FFSSD [34]	VGG-16	72.95	73.24	73.17	84.08	75.86	38.2
ESSD [35]	VGG-16	73.08	72.94	73.61	84.27	75.98	37.3
DC-SPP-YOLO [36]	Figure 5 in [35]	73.16	73.52	74.82	84.82	76.58	33.5
UAV-YOLO [37]	Figure 1 in [36]	74.68	74.20	76.32	85.96	77.79	30.12
FRCN [12]	VGG-16	85.85	86.67	88.15	90.35	87.76	6.1
DConvNet [38]	ResNet-101	71.87	90.35	89.59	99.88	87.92	6.7
MRFF-YOLO [39]	Figure 5 in [38]	87.16	86.56	87.56	92.05	88.33	25.1
Improved-YOLOv3 [40]	Figure 3 in [39]	86.42	87.57	89.37	91.56	88.73	25.8
SigNMS [41]	VGG-16	80.60	90.60	87.40	99.10	89.40	6.7
BDFFDN (ours)	Figure 6	90.81	90.73	84.12	100.00	91.41	26