Table 2. Comparison of detection performance of YOLOv4-GNet and other models.

Model	F1-score	Precision (%)	Sensitivity (%)	Specificity (%)
R-CNN	0.39	56.25	30.25	33.63
SPP-Net	0.26	68.95	16.03	20.75
Fast R-CNN	0.45	58.31	36.83	39.98
Faster R-CNN	0.56	70.00	47.06	46.88
FPN	0.57	83.00	43.70	46.87
Yolov4-GNet	0.87	86.34	86.69	90.63

R-CNN, region-convolutional neural network; SPP, spatial pyramid pooling; CNN, convolutional neural network; FPN, feature pyramid networks.