search

hustvl / TopFormer

TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation, CVPR2022
Other
373 stars 42 forks source link

size of output #8

Closed javadmozaffari closed 2 years ago

javadmozaffari commented 2 years ago

Hello, As shown in figure 2, 1/4 scale of input is present in the output. so for 448448 input, there is a 5656 image in the output!!! what is the point that I can't see?

mulinmeng commented 2 years ago

Hello, As shown in figure 2, 1/4 scale of input is present in the output. so for 448_448 input, there is a 56_56 image in the output!!! what is the point that I can't see?

Thanks for your ask! We have provided the detailed implementation in Sec.3.4 and conducted related experiments in Tab. 3. You can decide whether to add 1/4 features according to your hardware device.