A Small Object Detection Network Based on Multiple Feature Enhancement and Feature Fusion

Table 1

MFEFNet vs. other two-stage and one-stage detectors on COCO test-dev.

Method	Backbone	AP	AP50	AP75	APS	APM	APL
Two-stage methods
DeNet [33]	ResNet-101	33.8	53.4	36.1	12.3	36.1	50.8
CoupleNet [28]	ResNet-101	34.4	54.8	37.2	13.4	38.1	50.8
Faster R-CNN +++ [4]	ResNet-101	34.9	55.7	37.4	15.6	38.7	50.9
Faster R-CNN FPN [15]	ResNet-101	36.2	59.1	39.0	18.2	39.0	48.2
Mask R-CNN [5]	ResNet-101	38.2	60.3	41.7	20.1	41.1	50.2
Cascade R-CNN [23]	ResNet-101	42.8	62.1	46.3	23.7	45.5	55.2
One-stage methods
YOLOv2 [7]	DarkNet-19	21.6	44.0	19.2	5.0	22.4	35.5
SSD513 [10]	ResNet-101	31.2	50.4	33.3	10.2	34.5	49.8
RetinaNet [13]	ResNet-50	32.5	50.9	34.8	13.9	35.8	46.7
YOLOv3 [8]	DarkNet-53	33.0	57.9	34.4	18.3	35.4	51.1
DSSD513 [11]	ResNet-101	33.2	53.3	35.2	13.0	35.4	51.1
EfficientDet-D0 [14]	EfficientNet	34.6	53.0	37.1	—	—	—
RetinaNet [13]	ResNet-101	34.4	53.1	36.8	14.7	38.5	49.1
RefineDet512 [12]	ResNet-101	36.4	57.5	39.5	16.6	39.9	51.4
MFEFNet (ours)	ResNet-50	34.8	52.8	37.4	16.8	37.3	47.9
MFEFNet (ours)	ResNet-101	36.2	54.2	38.3	17.6	39.5	50.1