Skip to main content

Table 1 MobileNetV2 network structure

From: Ancient mural segmentation based on a deep separable convolution network

Input

Operator

t

c

n

s

2242 × 3

conv2d

32

1

2

1122 × 32

bottleneck

1

16

1

1

1122 × 16

bottleneck

6

24

2

2

562 × 24

bottleneck

6

32

3

2

282 × 32

bottleneck

6

64

4

2

142 × 64

bottleneck

6

96

3

1

142 × 96

bottleneck

6

160

3

2

72 × 160

bottleneck

6

320

1

1

72 × 320

conv2d 1 × 1

1 280

1

1

72 × 1 280

Avgpool 7 × 7

1

1 × 1 × 1 280

conv2d 1 × 1

k

 
  1. t is the expansion factor, c represents the number of output channels or the number of convolution kernels of the concerned layer, n is the number of repetitions of the convolution layer, and s is the stride, which represents the moving length of the convolution kernel