Table 1 The structure of IMV2.

From: MFCA-Net: a deep learning method for semantic segmentation of remote sensing images

Input

Operator

t

c

n

s

5122 × 3

conv2d

–

32

1

2

2562 × 32

Bottleneck1 + CA

1

16

1

1

2562 × 16

Bottleneck2

6

24

2

2

1282 × 24

Bottleneck3

6

32

3

2

642 × 32

Bottleneck4

6

64

4

2

322 × 64

Bottleneck5

6

96

3

1

322 × 96

Bottleneck6 + CA

6

160

3

2

162 × 160

Bottleneck7

6

320

1

1

162 × 320

conv2d

–

1280

1

1

162 × 1280

avgpool

–

–

1

–

1 × 1 × 1280

conv2d

–

K

–

 
Back to article page