README.md

Layered-Garment Net

This repository contains a pytorch implementation (training and testing) of

Layered-Garment Net: Generating Multiple Implicit Garment Layers from a Single Image

Alakh Aggarwal, Jikai Wang, Steven Hogue, Saifeng Ni, Madhukar Budagavi, Xiaohu Guo

ACCV 2022

[Paper]

Demo

This demo runs for sample dataset given in sample/data/ folder. Refer the folder and source code to run on custom data.

1. Download trained model from checkpoints and save in home folder of the project by name checkpoints.
2. run the following script:

python -m apps.update

The output will be saved in sample/data/ folder. The resulting mesh are watertight and can be visualized using meshlab. 3. To trim the generated mesh, run the following script:

python -m apps.trim

Citation

@inproceedings{aggarwal2022layered,
  title={Layered-Garment Net: Generating Multiple Implicit Garment Layers from a Single Image},
  author={Aggarwal, Alakh and Wang, Jikai and Hogue, Steven and Ni, Saifeng and Budagavi, Madhukar and Guo, Xiaohu},
  booktitle={Proceedings of the Asian Conference on Computer Vision (ACCV)},
  year={2022}
}