Table 1.

Detailed configuration of the proposed deep convolutional neural network architecture.

#	Type	Input	Kernel	Output
1	Convolutional	64 × 64 × 1	5 × 5	64 × 64 × 64
2	Convolutional	64 × 64 × 64	5 × 5	60 × 60 × 64
3	Convolutional	60 × 60 × 64	5 × 5	56 × 56 × 64
4	Max pooling	56 × 56 × 64	2 × 2	28 × 28 × 64
5	Convolutional	28 × 28 × 64	3 × 3	26 × 26 × 128
6	Convolutional	26 × 26 × 128	3 × 3	24 × 24 × 128
7	Convolutional	24 × 24 × 128	3 × 3	22 × 22 × 128
8	Max pooling	22 × 22 × 128	2 × 2	11 × 11 × 128
9	Convolutional	11 × 11 × 128	3 × 3	11 × 11 × 256
10	Convolutional	11 × 11 × 256	3 × 3	11 × 11 × 256
11	Convolutional	11 × 11 × 256	3 × 3	11 × 11 × 256
12	Max pooling	11 × 11 × 128	2 × 2	5 × 5 × 256
13	Fully connected	5 × 5 × 256	N/A	512
14	Dropout	512	N/A	512
15	Fully connected	512	N/A	2