search

bcmi / DCI-VTON-Virtual-Try-On

[ACM Multimedia 2023] Taming the Power of Diffusion Models for High-Quality Virtual Try-On with Appearance Flow.
https://arxiv.org/abs/2308.06101
MIT License
387 stars 56 forks source link

warp model #7

Closed hamaadtahiir closed 11 months ago

hamaadtahiir commented 1 year ago

Will you release the warp model as well? Also does your experiment include results for 1024x768 resolution?

Limbor commented 1 year ago

Yes, we will update the warping model soon. About other resolution model, we do not plan to release. You can train it by modifying the model config. Thank you for your interest in our work.

garychan22 commented 1 year ago

@Limbor Hi, should we load the warp model using https://github.com/geyuying/PF-AFN/tree/main since the params' dimensions mismatch the default network setting in PF-AFN

Limbor commented 1 year ago

Hi @garychan22 We have released the inference and training code for warping module

FYtrace commented 1 year ago

do we need pose, densepose, cloth_mask, and others? It seems cp_datasets.py read these messages. warp/test/data/cp_dataset.py#L203

Limbor commented 1 year ago

@FYtrace Yes, the VITON-HD dataset should include these data. If you want to test on your own data, you should do some preprocessing to get these data.