Table 10.

Input size and preprocessing input function for each CNN

CNN	Input Shape	Preprocess Input Function
VGG16 [36]	224x224	It converts RGB to BGR, The images are converted from RGB to BGR, then each color channel is zero-centered with respect to the ImageNet dataset, without scaling.
InceptionV3 [38]	299x299	The inputs pixel values are scaled between -1 and 1, sample-wise.
ResNet152V2 [40]	224x224	The inputs pixel values are scaled between -1 and 1, sample-wise.
InceptionResNetV2 [41]	299x299	The inputs pixel values are scaled between -1 and 1, sample-wise.
MobileNetV2 [46]	224x224	The inputs pixel values are scaled between -1 and 1, sample-wise.
DenseNet201 [43]	224x224	The input pixels values are scaled between 0 and 1 and each channel is normalized with respect to the ImageNet dataset.
Xception [42]	299x299	The inputs pixel values are scaled between -1 and 1, sample-wise.
NasNetLarge [44]	331x331	The inputs pixel values are scaled between -1 and 1, sample-wise.
EfficientNetV2L [49]	480x480	Nothing