Table 1.
Simplified SegNet network.
| S. No. | Layer name | Type | Description |
|---|---|---|---|
| 1 | “Image input” | Image | 208 × 1761 images with “zero center” normalization |
| 2 | “encoder1_conv1” | Convolution | 64 3 × 3 × 1 convolutions with stride [1 1] and padding [1 1 1 1] |
| 3 | “encoder1_bn_1” | Batch normalization | Batch normalization with 64 channels |
| 4 | “encoder1_relu_1” | ReLU | ReLU |
| 5 | “encoder1_conv2” | Convolution | 64 3 × 3 × 64 convolutions with stride [1 1] and padding [1 1 1 1] |
| 6 | “encoder1_bn_2” | Batch normalization | Batch normalization with 64 channels |
| 7 | “encoder1_relu_2” | ReLU | ReLU |
| 8 | “encoder1_maxpool” | Max pooling | 2 × 2 max pooling with stride [2 2] and padding [0 0 0 0] |
| 9 | “encoder2_conv1” | Convolution | 64 3 × 3 × 64 convolutions with stride [1 1] and padding [1 1 1 1] |
| 10 | “encoder2_bn_1” | Batch normalization | Batch normalization with 64 channels |
| 11 | “encoder2_relu_1” | ReLU | ReLU |
| 12 | “encoder2_conv2” | Convolution | 64 3× 3 × 64 convolutions with stride [1 1] and padding [1 1 1 1] |
| 13 | “encoder2_bn_2” | Batch normalization | Batch normalization with 64 channels |
| 14 | “encoder2_relu_2” | ReLU | ReLU |
| 15 | “encoder2_maxpool” | Max pooling | 2 × 2 max pooling with stride [2 2] and padding [0 0 0 0] |
| 16 | “decoder2_unpool” | Max unpooling | Max unpooling |
| 17 | “decoder2_conv2” | Convolution | 64 3× 3 × 64 convolutions with stride [1 1] and padding [1 1 1 1] |
| 18 | “decoder2_bn_2” | Batch normalization | Batch normalization with 64 channels |
| 19 | “decoder2_relu_2” | ReLU | ReLU |
| 20 | “decoder2_conv1” | Convolution | 64 3 × 3 × 64 convolutions with stride [1 1] and padding [1 1 1 1] |
| 21 | “decoder2_bn_1” | Batch normalization | Batch normalization with 64 channels |
| 22 | “decoder2_relu_1” | ReLU | ReLU |
| 23 | “decoder1_unpool” | Max unpooling | Max unpooling |
| 24 | “decoder1_conv2” | Convolution | 64 3 × 3 × 64 convolutions with stride [1 1] and padding [1 1 1 1] |
| 25 | “decoder1_bn_2” | Batch normalization | Batch normalization with 64 channels |
| 26 | “decoder1_relu_2” | ReLU | ReLU |
| 27 | “decoder1_conv1” | Convolution | 4 3 × 3 × 64 convolutions with stride [1 1] and padding [1 1 1 1] |
| 28 | “decoder1_bn_1” | Batch normalization | Batch normalization with 4 channels |
| 29 | “decoder1_relu_1” | ReLU | ReLU |
| 30 | “Softmax” | Softmax | Softmax |
| 31 | “Pixel_classify” | Pixel classification layer | Class weighted cross-entropy loss with “background,” “CSF,” “GM,” and “WM” classes |