miccunifi / ladi-vton

[ACM MM 2023] - LaDI-VTON: Latent Diffusion Textual-Inversion Enhanced Virtual Try-On
Other
427 stars 55 forks source link

could not get good result #24

Open renrenzsbbb opened 1 year ago

renrenzsbbb commented 1 year ago

Thanks for your great work. I want to test your model trained by dresscode. However, I don't have dresscode. I use the vitonhd.py to load vitonhd training data and use the pretrained model by dresscode. Unluckly, the sleeve is over original boundary. Can you give me same advice. image

Another question, when I test your code by vitonhd(vitonhd.py to load data and pretrained by vitonhd),I find that the sleeve is not match my human parsing. Can I give depth/canny control to your diffusion model? image

Thanks for any advices.

Jdachang commented 1 year ago

Thanks for your great work. I want to test your model trained by dresscode. However, I don't have dresscode. I use the vitonhd.py to load vitonhd training data and use the pretrained model by dresscode. Unluckly, the sleeve is over original boundary. Can you give me same advice. image

Another question, when I test your code by vitonhd(vitonhd.py to load data and pretrained by vitonhd),I find that the sleeve is not match my human parsing. Can I give depth/canny control to your diffusion model? image

Thanks for any advices.

Which GitHub project is used for human parsing?

renrenzsbbb commented 1 year ago

Thanks for your great work. I want to test your model trained by dresscode. However, I don't have dresscode. I use the vitonhd.py to load vitonhd training data and use the pretrained model by dresscode. Unluckly, the sleeve is over original boundary. Can you give me same advice. image Another question, when I test your code by vitonhd(vitonhd.py to load data and pretrained by vitonhd),I find that the sleeve is not match my human parsing. Can I give depth/canny control to your diffusion model? image Thanks for any advices.

Which GitHub project is used for human parsing? I parsing the human by labelme. And you can refer https://github.com/sangyun884/HR-VITON/blob/main/Preprocessing.md

momolefe24 commented 1 year ago

use the inf_pgn.py file and ensure that the checkpoint is the folder instead. Ensure that the python version is the correct one since they use tensorflow There is a 2022 paper that performs human parsing. Email me at molefemolefe45@gmail.com to collaborate on these projects thanks

renrenzsbbb commented 1 year ago

use the inf_pgn.py file and ensure that the checkpoint is the folder instead. Ensure that the python version is the correct one since they use tensorflow There is a 2022 paper that performs human parsing. Email me at molefemolefe45@gmail.com to collaborate on these projects thanks

Thanks for your reply. But I don't focus on image parsing because I want to try-on to certain one avator. Now, I can't get good result.