Skip to content

huanngzh/threestudio-dreamwaltz

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

threestudio-dreamwaltz

DreamWaltz extension of threestudio. To use it, please install threestudio first and then install this extension in threestudio custom directory.

Installation

cd custom
git clone https://github.com/huanngzh/threestudio-dreamwaltz.git
cd threestudio-dreamwaltz

pip install -r requirements.txt

If installing the pytorch3d package fails, please see the detailed instructions at pytorch3d/INSTALL.md.

Prepare SMPL Weights

We use smpl and vposer models for avatar creation and animation learning, please follow the instructions in smplx and human_body_prior to download the model weights, and build a directory with the following structure:

smpl_models
├── smpl
│   ├── SMPL_FEMALE.pkl
│   └── SMPL_MALE.pkl
│   └── SMPL_NEUTRAL.pkl
└── vposer
    └── v2.0
        ├── snapshots
        ├── V02_05.yaml
        └── V02_05.log

Then, update the model paths SMPL_ROOT and VPOSER_ROOT in utils/smpl/smpl_prompt.py.

Quick Start

Static Avatar Creation

All in one (SMPL Initializaion + Canonical Avatar Creation):

python launch.py --config custom/threestudio-dreamwaltz/configs/dreamwaltz-static.yaml --train --gpu 0 system.prompt_processor.prompt="Naruto"

Divided into multiple stages:

# SMPL Initializaion
python launch.py --config custom/threestudio-dreamwaltz/configs/experimental/dreamwaltz-1-warmup.yaml --train --gpu 0 system.prompt_processor.prompt="Naruto"
# Canonical Avatar Creation
python launch.py --config custom/threestudio-dreamwaltz/configs/experimental/dreamwaltz-2-nerf.yaml --train --gpu 0 system.prompt_processor.prompt="Naruto" resume=path/to/trial/dir/ckpts/last.ckpt

Animatable Avatar Learning

Not yet implemented!

Citing

If you find DreamWaltz helpful, please consider citing:

@article{huang2023dreamwaltz,
    title={DreamWaltz: Make a Scene with Complex 3D Animatable Avatars},
    author={Yukun Huang and Jianan Wang and Ailing Zeng and He Cao and Xianbiao Qi and Yukai Shi and Zheng-Jun Zha and Lei Zhang},
    journal = {arXiv:2305.12529},
    year={2023},
}

About

DreamWaltz extension of threestudio.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages