Gfpgan blurred photo repair artifact

Git:

https://github.com/TencentARC/GFPGAN

---

Installation dependency ：requirements.txt

torch Use CUDA:

---

requirements requirement torch>=1.7, Just install the following directly , It works

pip install torch==1.11.0+cu113 torchvision==0.12.0+cu113 torchaudio===0.11.0+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html

Command line start ：

python inference_gfpgan.py -i inputs\whole_imgs -o outputImage -v 1.2 -s 2

-i inputs\whole_imgs  #  Enter the image path 
-o outputImage #  Output picture path 
-v 1.2  # Model version number 
-s 2  #  Final upsampling scale of image .  The default value is :2 （2:256*256,4:1024*1024, 8:2048*2048）