Haoran Li, Yuli Tian, Yonghui Wang, Yong Liao, Lin Wang, Yuyang Wang, Peng Yuan Zhou
This repository contains the official implementation for Text-to-3D Generation by 2D Editing.
Note: We compress these motion pictures for faster previewing. More high-quality 3D results can be found on our project homepage GE3D.
| A wooden rocking chair, rustic, comfortable, 8K. | A fluffy squirrel wearing a tiny wizard hat. | Black Widow in Marvel, head, photorealistic, 8K, HDR. | Ninja in black outfit, photorealistic, 8K, HDR. |
 |
 |
 |
 |
git clone https://github.com/Jahnsonblack/GE3D.git
cd GE3D
conda create -n ge3d python=3.10
conda activate ge3d
pip install -r requirements.txt -f https://download.pytorch.org/whl/cu118/torch_stable.html
git clone --recursive https://github.com/Jahnsonblack/diff-gaussian-rasterization.git
git clone https://github.com/YixunLiang/simple-knn.git
pip install diff-gaussian-rasterization/
pip install simple-knn/
# Install point-e
git clone https://github.com/crockwell/Cap3D.git
cd Cap3D/text-to-3D/point-e/
pip install -e .cd GE3D
mkdir point_e_model_cache
# Optional: Initialize with better point-e
wget https://huggingface.co/datasets/tiange/Cap3D/resolve/main/misc/our_finetuned_models/pointE_finetuned_with_825kdata.pth
mv pointE_finetuned_with_825kdata.pth point_e_model_cache/
# Modify the parameter init_guided in the configuration file to pointe_825k
# or
wget https://huggingface.co/datasets/tiange/Cap3D/resolve/main/misc/our_finetuned_models/pointE_finetuned_with_330kdata.pth
mv pointE_finetuned_with_330kdata.pth point_e_model_cache/
# Modify the parameter init_guided in the configuration file to pointe_330kpython train.py --opt configs/sample.yamlTo achieve higher quality 3D generation results, you can increase the number of iterations and use smaller step sizes in the later stages of optimization.
This work is built on many amazing research works and open-source projects :
Thanks for their excellent work and great contribution to 3D generation area.
If you find it useful in your research, please consider citing our paper as follows:
@article{li2024enhanced,
title={Enhanced 3D Generation by 2D Editing},
author={Li, Haoran and Tian, Yuli and Liao, Yong and Wang, Lin and Wang, Yuyang and Zhou, Peng Yuan},
journal={arXiv preprint arXiv:2412.05929},
year={2024}
}