This is the repo for DIG: Draping Implicit Garment over the Human Body.
Here we provide the code and the pretrained models to generate garments and drape them on the posed human body.
The code is implemented with python 3.6 and torch 1.9.0+cu102 (other versions may also work).
Download the female SMPL model from http://smplify.is.tue.mpg.de/ and place basicModel_f_lbs_10_207_0_v1.0.0.pkl in the folder of ./smpl_pytorch.
Other dependencies include trimesh, torchgeometry, scikit-image.
In infer.py you can find the script for inference. It loads the pretrained SDF models and skinning models (which are placed in ./extra-data/pretrained), reconstructs garments using the learned latent codes and deforms them with the given SMPL parameters /extra-data/pose-beta-sample.pt. You can simply run the following command and the output mesh will be saved at ./output.
python infer.py
To generate results for other garments and bodies, you can replace the latent code and SMPL parameters with other values.
Check here for the instruction of training.
If you find this work helpful for your research, please cite
@inproceedings{ren2022dig,
author = {Ren, Li and Guillard, Benoit and Remelli, Edoardo and Fua, Pascal},
title = {{DIG: Draping Implicit Garment over the Human Body}},
booktitle = {Asian Conference on Computer Vision},
year = {2022}
}