Skip to main content

Table 1 The details of the proposed initial CNN model

From: Breast cancer histopathology image classification through assembling multiple compact CNNs

Type

Patch size/Stride

Output

Depth

Params

Convolution

7 ×7/2

112 ×112×64

1

2.7K

Max pool

3 ×3/2

56 ×56×64

0

 

Convolution

1 ×1/1

56 ×56×64

1

0.8K

Convolution

3 ×3/1

56 ×56×192

1

112K

Max pool

3 ×3/2

28 ×28×192

0

 

Inception(3a)

 

28 ×28×256

2

159K

SEP block

1 ×1

28 ×28×256

2

32K

Inception(3b)

 

28 ×28×480

2

380K

SEP block

1 ×1

28 ×28×480

2

32K

Max pool

3 ×3/2

14 ×14×480

0

 

Inception(4a)

 

14 ×14×512

2

364K

SEP block

1 ×1

14 ×14×512

2

32K

Inception(4b)

 

14 ×14×512

2

437K

SEP block

1 ×1

14 ×14×512

2

32K

Inception(4c)

 

14 ×14×512

2

840K

SEP block

1 ×1

14 ×14×512

2

32K

Inception(4d)

 

14 ×14×528

2

580K

SEP block

1 ×1

14 ×14×528

2

32K

Inception(4e)

 

14 ×14×1856

2

840K

SEP block

1 ×1

14 ×14×1856

2

32K

Max pool

3 ×3/2

7 ×7×1856

0

 

Ave pool

7 ×7/1

1 ×1×1856

0

 

Linear

 

1 ×1×2

1

2K

Softmax

 

1 ×1×2

0

 
  1. The output of the convolution layer and SEP block may change after the channel pruning stage in every model compression loop