metadata

title: LucidDreamer
emoji: 👁
colorFrom: pink
colorTo: yellow
sdk: gradio
sdk_version: 4.7.1
python_version: 3.9
app_file: gradio_demo.py
pinned: false
license: mit
tags:
  - arXiv:2311.11284
  - Diffusion
  - 3D Generation
  - Gaussian Splatting

LucidDreamer: Towards High-Fidelity Text-to-3D Generation via Interval Score Matching

Yixun Liang* Xin Yang*, Jiantao Lin, Haodong Li, Xiaogang Xu, Yingcong Chen**

*: Equal contribution. **: Corresponding author.

Paper PDF (Arxiv) | Project Page (Coming Soon)

Note: we compress these motion pictures for faster previewing.

Examples of text-to-3D content creations with our framework, the LucidDreamer, within ~35mins on A100.

🎏 Abstract

We present a text-to-3D generation framework, named the LucidDreamer, to distill high-fidelity textures and shapes from pretrained 2D diffusion models.

CLICK for the full abstract

The recent advancements in text-to-3D generation mark a significant milestone in generative models, unlocking new possibilities for creating imaginative 3D assets across various real-world scenarios. While recent advancements in text-to-3D generation have shown promise, they often fall short in rendering detailed and high-quality 3D models. This problem is especially prevalent as many methods base themselves on Score Distillation Sampling (SDS). This paper identifies a notable deficiency in SDS, that it brings inconsistent and low-quality updating direction for the 3D model, causing the over-smoothing effect. To address this, we propose a novel approach called Interval Score Matching (ISM). ISM employs deterministic diffusing trajectories and utilizes interval-based score matching to counteract over-smoothing. Furthermore, we incorporate 3D Gaussian Splatting into our text-to-3D generation pipeline. Extensive experiments show that our model largely outperforms the state-of-the-art in quality and training efficiency.

🔧 Training Instructions

Our code is now released! Please refer to this link for detailed training instructions.

🚧 Todo

Release the basic training codes
Release the guidance documents
Release the training codes for more applications

📍 Citation

@misc{EnVision2023luciddreamer,
      title={LucidDreamer: Towards High-Fidelity Text-to-3D Generation via Interval Score Matching}, 
      author={Yixun Liang and Xin Yang and Jiantao Lin and Haodong Li and Xiaogang Xu and Yingcong Chen},
      year={2023},
      eprint={2311.11284},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

Acknowledgement

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.