About the 'Figure 11: Visual comparison of super resolution results on CelebA-HQ.'

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

About the 'Figure 11: Visual comparison of super resolution results on CelebA-HQ.' #237

Closed gzhhhere closed 2 years ago

gzhhhere commented 2 years ago

Hi! Could you please tell me how do you train the pix2pixHD model in super resolution tasks? Do you perform random bi-cubic down-sampling or fixed a down sampling factor?
Thank you !

yuval-alaluf commented 2 years ago

We perform a random bi-cubic down-sampling. We provide a command for reproducing the results: https://github.com/eladrich/pixel2style2pixel#super-resolution-1