Closed cslvjt closed 1 year ago
You have a better reconstruction than mine, for sure! Its possible my simpler implementation isn't working as well as the original code, or perhaps batch processing is changing the optimization a bit.
You should see these images saved under results/deform_XXX.png
when you call run.py
Thank you for your reply. I think the problem of artifacts is caused by the lack of optical flow loss function during training.
Thank you for introduction about CoDeF. I have a question about your blog.
how to generate this image and Why do you generate images with so many artifacts? but I didn't generate it follow the original author code. This is my generation. Thank you.