Table 6.

Validation accuracy of CNN models using 3 class 3 folds after segment selection.

Accuracy	Folds
	Fold1	Fold2	Fold3	AVG
VGG16	0.85957	0.8383	0.84483	0.85
VGG19	0.84255	0.89711	0.86207	0.87
Xception	0.65106	0.61277	0.67241	0.64
ResNet152V2	0.72766	0.68287	0.69397	0.70
ResNet152	0.75745	0.74468	0.83621	0.78
ResNet101V2	0.75745	0.69787	0.73276	0.73
ResNet101	0.77447	0.74894	0.77155	0.76
ResNet50v2	0.72766	0.69362	0.73707	0.72
ResNet50	0.75745	0.73191	0.78017	0.76
NasNetMobile	0.69362	0.69787	0.68966	0.69
MobileNetV2	0.69787	0.65957	0.69397	0.68
MobileNet	0.74043	0.69787	0.71552	0.72
Inceptionv3	0.70638	0.68511	0.66379	0.68
InceptionResNetV2	0.69787	0.70213	0.67241	0.69
DenseNet201	0.77447	0.7234	0.76293	0.75
DenseNet169	0.74894	0.68085	0.75431	0.73
DenseNet121	0.7234	0.70213	0.72414	0.71