Table 1 Network structure of BiUnet-Dense.
From: Urban change detection of remote sensing images via deep-feature extraction
Time | Block | Conv layer | Filter | Outputsize | |
|---|---|---|---|---|---|
Encoder | T1 | block1 | con11–con13 | [3 × 3,16]×3 | [16,35,3,3] |
block2 | con21–con23 | [3 × 3,32]×3 | [32,115,3,3] | ||
block3 | con31–con33 | [3 × 3,64]×3 | [64,275,3,3] | ||
block4 | con41–con43 | [3 × 3,128]×3 | [128,595,3,3] | ||
T2 | block1 | con11–con13 | [3 × 3,16]×3 | [16,35,3,3] | |
block2 | con21–con23 | [3 × 3,32]×3 | [32,115,3,3] | ||
block3 | con31–con33 | [3 × 3,64]×3 | [64,275,3,3] | ||
block4 | con41–con43 | [3 × 3,128]×3 | [128,595,3,3] | ||
Decoder | block1 | con43d–con41d | [3 × 3,128]×2 [3 × 3,64]×1 | [512,64,3,3] | |
block2 | con33d–con31d | [3 × 3,64]×2 [3 × 3,32]×1 | [256,32,3,3] | ||
block3 | con23d–con21d | [3 × 3,32]×2 [3 × 3,16]×1 | [128,16,3,3] | ||
block4 | con13d–con11d | [3 × 3,16]×2 [3 × 3,2]×1 | [16,2,3,3] |
