Table 3.

The obtained accuracies over five folds in the 19 CNN models.

Model	Fold1	Fold2	Fold3	Fold4	Fold5
AlexNet	78.28	89.9	86.87	90.4	89.9
Vgg16	82.32	86.87	86.87	95.96	96.46
Vgg19	79.8	80.81	90.4	89.39	96.46
Inceptionv1	82.32	84.85	83.84	91.92	95.96
Inceptionv3	79.8	90.91	96.97	99.49	97.47
ResNet18	82.83	85.35	92.93	92.42	96.46
ResNet50	81.31	90.91	92.93	97.98	98.99
ResNet101	81.82	94.44	96.97	98.99	99.49
InceptionResv2	77.27	88.89	93.43	93.94	97.98
Xception	78.28	86.87	90.91	93.43	95.45
DenseNet201	86.36	91.94	96.46	97.98	97.47
MobileNetv2	81.31	86.87	89.9	97.47	98.48
ShuffleNet	77.27	83.84	84.85	88.38	95.96
NasnetMobile	78.28	86.36	85.86	85.86	96.46
NasnetLarge	80.3	88.89	92.93	96.46	97.98
DarkNet19	81.31	85.86	84.85	90.91	90.91
DarkNet53	79.29	87.37	91.41	93.94	93.94
EfficientNetB0	84.34	83.84	85.35	88.89	91.92
SqueezeNet	82.32	84.85	82.32	85.86	87.88