[ACM MM 2024] Shape-Guided Clothing Warping for Virtual Try-On

This repository is the official implementation of SCW-VTON

Shape-Guided Clothing Warping for Virtual Try-On
Xiaoyu Han, Shunyuan Zheng, Zonglin Li, Chenyang Wang, Xin Sun, Quanling Meng

[Paper]  [Checkpoints] 

 

Environments

1. Clone the repository

git clone https://github.com/xyhanHIT/SCW-VTON.git
cd SCW-VTON

2. Install Python dependencies

conda create -n scw-vton python=3.8
conda activate scw-vton
bash environment.sh

3. Download the pretrained vgg checkpoint and put it in models/vgg/

Inference

1. Download the VITON-HD datasets from VITON-HD or GP-VTON.

2. Download the checkpoints from Baidu Netdisk or Google Drive.

3. Modify the "dataroot" and "ckpts" parameters in "test.sh" according to the downloaded files above.

4. Run the "test.sh".

bash test.sh

5. Evaluation

python metrics.py

Results

Our Team's Researches

• [ACM MM'22] PL-VTON - Progressive Limb-Aware Virtual Try-On
• [IEEE TMM'23] PL-VTONv2 - Limb-Aware Virtual Try-On Network With Progressive Clothing Warping
• [ACM MM'24] SCW-VTON - Shape-Guided Clothing Warping for Virtual Try-On

Acknowledgements

Our code references the implementation of DCI-VTON. Thanks for their awesome works.

Citation

If you use our code or models, please cite with:

@inproceedings{han2024shape,
  title={Shape-Guided Clothing Warping for Virtual Try-On},
  author={Han, Xiaoyu and Zheng, Shunyuan and Li, Zonglin and Wang, Chenyang and Sun, Xin and Meng, Quanling},
  booktitle={Proceedings of the 32nd ACM International Conference on Multimedia},
  pages={2593--2602},
  year={2024}
}