MosbehBarhoumiRAI
MosbehBarhoumiRAI
And if you are asking about the key points, they used Openpose as they mentioned.
Hello, @zero-kappa, could you please review my repository regarding all the pre-processing steps? You can find it at the following link: [VITON-PRE-PROCESSING](https://github.com/MosbehBarhoumi/VITON-PRE-PROCESSING). It is the official one on the VITON...
@Abdullahkhan5 To replicate the results, simply follow the inference steps by downloading the check points for the trained models and placing them in the correct location. Additionally, preprocess your own...
We are waiting (also trying) 👀
Please review the official repository of VITON by visiting this link: https://github.com/xthan/VITON. I have created a Colab Notebook to aid in the pre-processing of VITON, which can be accessed here:...
This is a common problem with diffusion-based virtual try-on models. One approach to fix it (partially) is to postprocess the output image. For example, you can use a segmentation model...
same, it works better with higher resolution images but still not that good results, I believe it's not an issue because this is like what the model(s) can do!
Upgrading to 1.13.1 solved the problem !