Table 3.

Performance metrics of the compared DNN models.

Model	Accuracy	Precision	Recall	AUROC
DenseNet201	0.788	0.751	0.99	0.967
ViT	0.8	0.759	0.995	0.959
MobileNetV2	0.843	0.815	0.969	0.955
NASNetMobile	0.806	0.773	0.977	0.945
ResNet101V2	0.792	0.756	0.985	0.94
Xception	0.793	0.755	0.99	0.936
InceptionV3	0.792	0.756	0.985	0.931
InceptionResNetV2	0.827	0.794	0.977	0.926