Table 6. Selected values of parameters for benchmark CNNs.

CNNs	Epochs, pixel size, dropout, batch size, dense nodes,	Average Running Time for 200 epochs
Best CNN	Batch-size = 16, dropout = 0.7, nodes [3 conv layers) = 8,16,32, Dense Nodes: 64, Epochs:200, Pixel = [256x256x3)	418 sec
VGG16	Pixel size (256x256x3), batch size = 64, dropout = 0.2, dense nodes 2x512, epochs = 200	1005 sec
ResNet50	Pixel size (300x300x3), batch size = 8, dropout = 0.2, Global Average Pooling, dense nodes 512x512, epochs = 200	2434 sec
MobileNet	Pixel size (300x300x3), batch size = 16, dropout = 0.2, global average pooling, epochs = 200	1380 sec
Inception V3	Pixel size (250x250x3), batch size = 16, dropout = 0.7, dense nodes = 1500x1500, epochs = 200	1317 sec
Xception	Pixel size (300x300x3), batch size = 8, dropout = 0.7, Global Average Pooling, dense nodes = 512x512, epochs = 200	3532 sec