Table 3. Components of the proposed CNN model applied on the VMD data.

Layer	No. of Layers	Layer Size	No. of Kernels	Kernel Features
Input Layer	1	512 × 512 × 3	-	-
Convolution/ReLU	3	3 × 3	32	Same Padding
Pooling Layer	1	2 × 2	32	No Padding
Convolution/ReLU	3	3 × 3	64	Same Padding
Pooling Layer	1	2 × 2	64	No Padding
Convolution/ReLU	3	3 × 3	128	Same Padding
Pooling Layer	1	2 × 2	128	No Padding
Convolution/ReLU	3	3 × 3	256	Same Padding
Pooling Layer	1	2 × 2	256	No Padding
Convolution/ReLU	2	3 × 3	512	Same Padding
Pooling Layer	1	2 × 2	512	No Padding
Convolution/ReLU	1	3 × 3	1024	Same Padding
Pooling Layer	1	2 × 2	1024	No Padding
Fully Connected/ ReLU	3	128, 64, 32,16	1	-
Fully Connected	1	2	1	-
Softmax/Classification	1		1	-