Research Article
Fault Line Selection Method Based on Transfer Learning Depthwise Separable Convolutional Neural Network
Table 3
The specific structural parameters of DSCNN.
| Layer | Convolution kernel/sampling window size | Output feature map size |
| Input | —— | 150 150 | SeparableConv2D_1 | 3 3 | 148 148 32 | MaxPooling2D_1 | 2 2 | 74 74 32 | SeparableConv2D_2 | 3 3 | 72 72 64 | MaxPooling2D_2 | 2 2 | 36 36 64 | SeparableConv2D_3 | 3 3 | 34 34 128 | MaxPooling2D_3 | 2 2 | 17 17 128 | SeparableConv2D_4 | 3 3 | 15 15 128 | MaxPooling2D_4 | 2 2 | 7 7 128 | Flatten | —— | 6272 1 | Dense | —— | 512 1 | Output | —— | 6 1 |
|
|