search

tamarott / SinGAN

Official pytorch implementation of the paper: "SinGAN: Learning a Generative Model from a Single Natural Image"
https://tamarott.github.io/SinGAN.htm
Other
3.31k stars 611 forks source link

i want to know the usage of the mask #120

Open Alphver opened 4 years ago

Alphver commented 4 years ago

I originally thought the trained model only need a naive paste image to inference, until I found my result tend to be weird. Then I found in harmonization part and image editing part, you proposed a mask function. What is the usage of the mask function? I think this might be the key.

Alphver commented 4 years ago

And by the way, neither your paper nor SM talk about the mask

tamarott commented 4 years ago
  1. as we explain in the paper we combine the generated image with the original background. We do this using a binary mask.
  2. the way to use the mask is described in the readme