Awesome
Connecting Consistency Distillation to Score Distillation for Text-to-3D Generation
Zongrui Li*, Minghui Hu*, Qian ZhengβοΈ, Xudong Jiang
*: Equal Contribution. βοΈ: Corresponding author.
<div align=center> <img src="assets/teaser.gif" width="90%"/> </div>π© Brief
We present an improved 2D-to-3D distillation method using theories from consistency distillation.
π» Installation
conda create -n gcs python=3.9.16 cudatoolkit=11.8
conda activate gcs
pip install -r requirements.txt
pip install submodules/diff-gaussian-rasterization/
pip install submodules/simple-knn/
pip install submodules/point-e/
pip install tensorboard
π₯ Training
python train.py --opt configs/full_model/cat_armor.yaml
# train script
sh scripts/train_0.sh
π§Ύ Todo List
- Release the basic training codes
π Citation
@article{li2024gcs,
title={Connecting Consistency Distillation to Score Distillation for Text-to-3D Generation},
author={Li, Zongrui and Hu, Minghui and Zheng, Qian and Jiang, Xudong},
journal={arXiv preprint arXiv:2407.13584},
year={2024}
}
π Acknowledgement
Our work is developed on LucidDreamer. Thanks for their contribution to this task!