Table 8. Training parameters and accuracy of models in comparison with YO-CNN.

Training parameters and accuracy
Model	Labels	Image size	Batch size	Weight size	Recall	Precision	mAP
YO-CNN	2	640 ×640	16	14.5 MB	0.976	0.97	0.98
Vgg16				528	0.88	0.901	0.923
Vgg19				549	0.89	0.900	0.921
ResNet50				98	0.879	0.88	0.890
ResNet50v2				100	0.889	0.899	0.90
ResNet101v2				109	0.89	0.901	0.91
YOLOv5				14MB	0.90	0.91	0.91