search

eladrich / pixel2style2pixel

Official Implementation for "Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation" (CVPR 2021) presenting the pixel2style2pixel (pSp) framework
https://eladrich.github.io/pixel2style2pixel/
MIT License
3.2k stars 568 forks source link

how to change model #222

Closed Hieuslecong closed 3 years ago

Hieuslecong commented 3 years ago

hi, thanks for your great article. I will need your post to solve my problem. I read that I will have to train a styelgan or stylegan2 model before it can be used in pSp. I am looking for a way to change the default model in your post with my stylegan model. my stylegan model is not about the human face. can you guide me on how to change my existing model ? Thanks for your great contribution.

yuval-alaluf commented 3 years ago

The code will load the pre-trained StyleGAN2 generator that is passed using the argument --stylegan_weights. You can set this command-line argument to the path for your model. In addition, you can set --output_size to the output resolution of your generator. Everything else should remain the same.

Hieuslecong commented 3 years ago

thank you very much. Thank you for your help.