search

greatlog / UnpairedSR

This is an offical implementation of the CVPR2022's paper [Learning the Degradation Distribution for Blind Image Super-Resolution](https://arxiv.org/abs/2203.04962)
162 stars 11 forks source link

Training tips for degradation model? #33

Open agarya opened 1 year ago

agarya commented 1 year ago

I have tried my best to train 2020Track1 by using psnr/2020Track1.yml and pretrained SR weight edsr_baseline_x4-new.pt, but I just can get PSNR 26.2 and cannot get the same PSNR 26.80 as the author mentioned in paper. What can I do to get the same PSNR in paper, or the author can provide checkpoint of degradation model?

Thanks!

mrgreen3325 commented 1 year ago

I have tried my best to train 2020Track1 by using psnr/2020Track1.yml and pretrained SR weight edsr_baseline_x4-new.pt, but I just can get PSNR 26.2 and cannot get the same PSNR 26.80 as the author mentioned in paper. What can I do to get the same PSNR in paper, or the author can provide checkpoint of degradation model?

Thanks!

Hello is that the paper use y channel to measure the psnr? Also, have you try the preper/2020Track2.yml? the author said this is the latest end to end pdm to sr training yml. However, someone said there is some errors.