You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, we do not need to train on the old photo dataset. You can directly try to restore the images in the folder DDNM/exp/datasets/ by running: python main.py --ni --simplified --config celeba_hq.yml --path_y solvay --eta 0.85 --deg "sr_averagepooling" --deg_scale 4.0 --sigma_y 0.1 -i demo , as you can see in Demo: Real-World Super-Resolution.
it seems that only celeba and imagenet dataset used in the paper are provided in the drives
网盘里似乎只有imagenet和celeba的数据集,没有老照片修复的,请问在哪里可以找到呢
The text was updated successfully, but these errors were encountered: