Research Article
Deep Learning Joint Inversion of Electrical Data for Ahead-Prospecting in Tunneling
Table 1
Setting for each convolutional layer module.
| | Encoder | Type | Filter | Stride | Decoder | Type | Filter | Stride |
| | conv_down1 | Conv | 3 × 3 × 3 | 1 | conv_up1 | Conv | 3 × 3 × 3 | 1 | | Conv | 3 × 3 × 3 | 1 | Conv | 3 × 3 × 3 | 1 | | conv_down2 | Max pool | 2 × 2 × 2 | 2 | conv_up2 | Up-conv | 3 × 3 × 3 | 1 | | Conv | 3 × 3 × 3 | 1 | Conv | 3 × 3 × 3 | 1 | | Conv | 3 × 3 × 3 | 1 | Conv | 3 × 3 × 3 | 1 |
| | conv_down3 | Max pool | 2 × 2 × 2 | 2 | conv_up3 | Up-conv | 3 × 3 × 3 | 1 | | Conv | 3 × 3 × 3 | 1 | Conv | 3 × 3 × 3 | 1 | | Conv | 3 × 3 × 3 | 1 | Conv | 3 × 3 × 3 | 1 |
| | conv_down4 | Max pool | 2 × 2 × 2 | 2 | conv_up4 | Up-conv | 3 × 3 × 3 | 1 | | Conv | 3 × 3 × 3 | 1 | Conv | 3 × 3 × 3 | 1 | | Conv | 3 × 3 × 3 | 1 | Conv | 3 × 3 × 3 | 1 | | Up-conv | 3 × 3 × 3 | 1 | Conv | 3 × 3 × 3 | 1 |
|
|
