search

xinntao / ESRGAN

ECCV18 Workshops - Enhanced SRGAN. Champion PIRM Challenge on Perceptual Super-Resolution. The training codes are in BasicSR.
https://github.com/xinntao/BasicSR
Apache License 2.0
6.02k stars 1.07k forks source link

something about discriminator #40

Closed liang233 closed 5 years ago

liang233 commented 5 years ago

hello，the discriminator ,in pix2pix uses patchgan but srgan do not use the structure and pix2pix use the connection of fake image and real image while srgan only use fake image as input，how to choose two differnet structure ？ thanks

xinntao commented 5 years ago

Usually determined by the experiments. And the original discriminator seems to work fine.

PatchGAN seems to be suitable for SR because SR focuses more on local textures. But too shallow layers may lead to unsatisfactory results. I have not tried much about this. From my previous exp, concatenating the faking and real images does not improve the results.

liang233 commented 5 years ago

thanks