Progressive Limb-Aware Virtual Try-On, ACM MM'22.

Official code for ACM MM 2022 paper 'Progressive Limb-Aware Virtual Try-On'

We propose a novel progressive limb-aware virtual try-on framework named PL-VTON. PL-VTON consists of Multi-attribute Clothing Warping (MCW), Human Parsing Estimator (HPE), and Limb-aware Texture Fusion (LTF), which produces stable clothing deformation and handles the texture retention well in the final try-on result.

[Paper]

[Checkpoints]

Pipeline

Environment

python 3.7

torch 1.9.0+cu111

torchvision 0.10.0+cu111

Dataset

For the dataset, please refer to VITON.

Inference

1. Download the checkpoints from here.

2. Get VITON dataset.

3. Run the "test.py".

python test.py

Note that the results of our pretrained model are guaranteed in VITON dataset only.

License

The use of this code is restricted to non-commercial research and educational purposes.

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

Citation

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

@inproceedings{han2022progressive,
  title={Progressive Limb-Aware Virtual Try-On},
  author={Han, Xiaoyu and Zhang, Shengping and Liu, Qinglin and Li, Zonglin and Wang, Chenyang},
  booktitle={Proceedings of the 30th ACM International Conference on Multimedia},
  pages={2420--2429},
  year={2022}
}