Table 2.

Six types of feature extraction networks composed of different network components.

Model		Backbone			Neck		Head	Number of Convolutional Layers
	Focus	ResBlock	CSPNet	SPP block	FPN	PANet	YOLO
YOLOV3-tiny					√		√	13
YOLOV3		√			√		√	52
YOLO-RP		√				√	√	30
YOLO-RCP		√	√			√	√	30
YOLO-A30	√	√	√	√		√	√	30
YOLO-A78	√	√	√	√		√	√	78