Skip to main content

Table 2 The backbone network of EfficientNet. Each row describes the stage, operation, input resolution, output channel, and the number of layers

From: Annotation-free genetic mutation estimation of thyroid cancer using cytological slides from multi-centers

Stage

Operator

Resolution

Channels

Layers

0

 

512x512

3

0

1

Conv3x3

512x512

32

1

2

MBConv1, k3x3

256x256

16

1

3

MBConv6, k3x3

256x256

24

2

4

MBConv6, k5x5

128x128

40

2

5

MBConv6, k3x3

64x64

80

3

6

MBConv6, k5x5

32x32

112

3

7

MBConv6, k5x5

32x32

192

4

8

MBConv6, k3x3

16x16

320

1

9

Conv1x1 &Pooling

16x16

1280

1