Table 2.

The proposed VGG19+CNN architecture.

Layer (type)	Output Shape	Parameters
vgg19 (Functional)	(None, 7, 7, 512)	20024384
reshape (Reshape)	(None, 7, 7, 512)	0
conv2d (Conv2D)	(None, 7, 7, 128)	1638528
activation (Activation)	(None, 7, 7, 128)	0
conv2d_1(Conv2D)	(None, 7, 7, 128)	409728
activation_1	(None, 7, 7, 128)	0
batch_normalization
(BatchNormalization)	(None, 7, 7, 128)	512
max_pooling2d
(MaxPooling2d)	(None, 2, 2, 128)	0
dropout (Dropout)	(None, 2, 2, 128)	0
flatten (Flatten)	(None, 512)	0
dense (Dense)	(None, 512)	262656
dropout_1 (Dropout)	(None, 512)	0
dense_1 (Dense)	(None, 4)	2052
Total parameters: 22,337,860
Trainable parameters: 22,337,604
Non-trainable parameters: 256