garychan22

@jinmaodaomaye2021 @sanbuddhacharyas Have you ever solved this?

I think the problem is how to get the exact training data of VITON-HD according to the documentation > I saved a parse map image using PIL.Image with P mode....

but the results produced by this https://colab.research.google.com/drive/1JOwOPaChoc9GzyBi5FUEYTSaP2qxJl10?usp=sharing#scrollTo=VtZDKLL8jWLr are far from the ones in viton regardless of the color example here the one in viton  the one produced by the...

@FengLi-ust the output result in opencv window is distored with input 512 * 512 by running demo.py

@bluehyena have you resolved this?

看起来这个重要的bug并没有被修复

@JincanDeng i have met the same error, have you solved this?

> Do you run it with your own prompt? > > I think it's mostly because you have an empty geometry, so in validation, you get None from the geometry...

my solution is change the model_type param in the customized model config before calling any "xxx.register(...)", then the conflict disappears

oh, i have just found the projects mmpose4aigc and will give it a try