Skip to main content
Visual Computing for Industry, Biomedicine, and Art

Table 1 Description of the architecture of Focus-RCNet

From: Focus-RCNet: a lightweight recyclable waste classification algorithm based on focus and knowledge distillation

Layer

Output_size

Ksize

Stride

Repeat

Output_channel

Image

380 × 380

   

3

Focus

190 × 190

1 × 1

1

1

24

Stage1

95 × 95

 

2

1

48

95 × 95

 

1

3

Stage2

48 × 48

 

2

1

96

48 × 48

 

1

2

Stage3

24 × 24

 

2

1

192

24 × 24

 

1

1

Stage4

12 × 12

 

2

1

384

12 × 12

 

1

1

Conv5

12 × 12

1 × 1

1

1

512

GAP

1 × 1

7 × 7

   

FC

    

1000

Flops

    

418.8 M

Back to article page