JunlinHan
One sided inference
Hey @JunlinHan,
Just wanted to say thanks for all the hard work you've put into this project, it is impressive
I was wondering if you might be able to help out with something that would really make things easier for everyone working on the project, a script for inference that loads images from a folder, runs the Generator A to B using any saved .pth model checkpoint and save the results in a folder. I know that myself (Would be difficult as I am new to PyTorch, mostly worked on TensorFlow) or others might be able to write this script, but it would likely take a significant amount of effort from their end. Your familiarity with your own code and project would make it really simple for you.
Other have asked you for it as well and any help in this regard would be greatly appreciated and I understand that you are likely busy, but any help you can provide would be greatly appreciated.
Best regards,
Hello @Sir-Zeeshan , Many thanks for your nice words!
I'm currently quite busy with CVPR rebuttal and PhD application. I will provide a simple solution instead of writing a script, please try!
Simply change line 56 of dcl_model.py into visual_names_B = ['real_B'] (remove "fake_A") and comment line 165 will enable one-side inference ( domain A ->B). Change line 55 and comment line 164 will perform domain B->A inference.
Hope it helps!
