search

sangyun884 / HR-VITON

Official PyTorch implementation for the paper High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled Conditions (ECCV 2022).
831 stars 171 forks source link

Help wanted : Unable to preprocess data . #1

Closed Arnavgoyal00 closed 2 years ago

Arnavgoyal00 commented 2 years ago

I am unable the find the correct scripts to preprocess the data.

sangyun884 commented 2 years ago

Currently, we do not support preprocessing code for datasets other than those we utilized. Please refer to the paper for detailed preprocessing procedures.

Arnavgoyal00 commented 2 years ago

Can you please share the code that you have utilized to preprocess the data if possible? And also can you please share the paper .

koo616 commented 2 years ago

@Arnavgoyal00 We used the pre-trained model of https://github.com/Engineering-Course/CIHP_PGN for human segmentation, and obtained an agnostic image and an agnostic segmentation map by the method in VITON-HD. For Densepose, please check detectron2.

Please refer to our paper(https://arxiv.org/pdf/2206.14180.pdf) for more information on the pre-processing :)

halcyon370 commented 2 years ago

@Arnavgoyal00 We used the pre-trained model of https://github.com/Engineering-Course/CIHP_PGN for human segmentation, and obtained an agnostic image and an agnostic segmentation map by the method in VITON-HD. For Densepose, please check detectron2.

Please refer to our paper(https://arxiv.org/pdf/2206.14180.pdf) for more information on the pre-processing :)

Hello, could you simply explain what the data format in 'image-densepose' looks like? Is it range from 0-255 in IUV?

MosbehBarhoumi commented 1 year ago

@Arnavgoyal00 If you are still looking for a solution for the pre-processing you can check this issue (it's a tutorial actually): https://github.com/sangyun884/HR-VITON/issues/45