search

Zhangjinso / PISE

123 stars 28 forks source link

About the extraction of "Fp" #23

Closed happyday521 closed 3 years ago

happyday521 commented 3 years ago

Hi, In your paper, you "concatenate the source image Is, the source parsing map Ss, the generated parsing map Sg and the target pose Pt in depth (channel) dimension and extract its feature Fp ", as shown in the Fig. image

However, in my opinion, Fp should aims to provide the target pose information. Why do you additionally use the source image Is and the source parsing map Ss as input? Do you try to only use Sg and Pt to extract Fp ? Thanks!

Zhangjinso commented 3 years ago

Hi, We found this pipeline can disentangle the shape and style information in the texture transfer model. I believe it is a good try to only take Sg and Pt as the inputs of the image generator, and I have tried to do it. Besides, if you would like to do it, more normalization blocks are recommended to adopt in the image generator.

happyday521 commented 3 years ago

Thanks！