search

raven38 / GroupDNet

Unofficial Pytorch implementation of "Semantically Multi-modal Image Synthesis"
3 stars 1 forks source link

deepfasion dataset #3

Open amandazw opened 3 years ago

amandazw commented 3 years ago

Hi I got the deepfasion dataset, but the structure of the image folder is wrong. where can I get the image which have 'lbl' or 'val' image folder name

amandazw commented 3 years ago

Dataset not found or incomplete. Please make sure all required folders for the specified "split" and "mode" are inside the "root" directory

raven38 commented 3 years ago

Hi, thank you for finding this repository.

I prepared the data for the DeepFashion dataset following the procedure in section 4.3 of the SMIS paper. https://openaccess.thecvf.com/content_CVPR_2020/papers/Zhu_Semantically_Multi-Modal_Image_Synthesis_CVPR_2020_paper.pdf

We use the official off-the-self implementation for the CIHP dataset. https://github.com/Engineering-Course/CIHP_PGN

However, I think we need some modification for preparing the DeepFashion dataset. Then, you can use generated dataset I made. https://drive.google.com/file/d/1f_PvqnsewaKrPeNiSA3DAtc5aNJMJM-Q/view?usp=sharing

Please feel free to ask me if you have any questions.

amandazw commented 3 years ago

Ok,thank a lot