Table 2.

Architecture details of the VGG19+CNN model. The maxpooling layer is applied after each convolution layer.

VGG19 + CNN Model	Output Shapes
[Conv (3 × 3) − 64] × 2	224 × 224 × 64
[Conv (3 × 3) − 128] × 2	112 × 112 × 128
[Conv (3 × 3) − 256] × 4	56 × 56 × 256
[Conv (3 × 3) − 512] × 4	28 × 28 × 512
[Conv (3 × 3) − 512] × 4	14 × 14 × 512
Max pool	7 × 7 × 512
Flatten	25,088
Dense (relu)	4096
Dropout (0.5)	4096
Dense (relu)	4096
Dropout (0.5)	4096
Dense (SoftMax)	3