Table 2 Layer wise description of the CNN trained.
S. no. | Input image | 227 × 227 × 3 images with ‘zero center’ normalization |
|---|---|---|
1. | conv1 layer | 96 11 × 11 × 3 convolutions with stride [4 4] and padding [0 0 0 0] |
2. | Relu1 unit | ReLU unit |
3. | Norm1layer | cross channel normalization with 5 channels per element |
4. | Pool1 layer | cross channel normalization with 5 channels per element |
5. | Conv2 layer | 256 5 × 5 × 48 convolutions with stride [1 1] and padding [2 2 2 2] |
6. | Relu2 unit | ReLU unit |
7. | Norm2 layer | Cross Channel Normalization with 5 channels per element |
8. | Pool2 layer | 3 × 3 max pooling with stride [2 2] and padding [0 0 0 0] |
9. | Conv3 layer | 384 3 × 3 × 256 convolutions with stride [1 1] and padding [1 1 1 1] |
10. | Relu3unit | ReLU unit |
11. | Conv4 layer | 384 3 × 3 × 192 convolutions with stride [1 1] and padding [1 1 1 1] |
12. | Relu4 unit | ReLU unit |
13. | Conv5 layer | 256 3 × 3 × 192 convolutions with stride [1 1] and padding [1 1 1 1] |
14. | Relu5 unit | ReLU |
15. | Pool5 layer | 3 × 3 max pooling with stride [2 2] and padding [0 0 0 0] |
16. | Fc6 layer | 4096 fully connected layer |
17. | Relu6 unit | ReLU unit |
18. | Fc7 layer | 4096 fully connected layer |
19. | Relu7 unit | ReLU unit |
20. | Special_2 | 64 fully connected layer |
21. | Relu8 | ReLU unit |
22. | Fc8_2 | 3 fully connected layer |
23. | Softmax | Softmax |
24. | Classification Output | Cross entropy ex |
