Skip to content

Utilizing the texture and depth priors to initialize the Gaussian points

License

Notifications You must be signed in to change notification settings

littlepure2333/prior-driven-init-gs

 
 

Repository files navigation

Prior-driven initialization of Gaussian Splatting

This is the implementation of prior-driven initialization described in "3.2.1 Single frame initialization" in "GFlow: Recovering 4D World from Monocular Video".

This initialization method, which takes an image and its corresponding depth as input, can achieve faster 3DGS reconstruction compared to random initialization.

Method RGB Depth Centers
Prior-init (500steps)
training_rgb_prior.mp4
training_depth_prior.mp4
training_center_prior.mp4
Random-init (5000 steps)
training_rgb_random.mp4
training_depth_random.mp4
training_center_random.mp4

Usage

  1. Install following msplat's instructions.
  2. run python tutorials/run.py and check the results in the logs folder.

Acknowledgment

We thank the msplat team for re-implementing 3DGS in a more developer-friendly way. We also salute the original 3DGS team for their seminal work.

If you find this initialization method helpful, please consider to cite:

@article{wang2024gflow,
  title={GFlow: Recovering 4D World from Monocular Video},
  author={Wang, Shizun and Yang, Xingyi and Shen, Qiuhong and Jiang, Zhenxiang and Wang, Xinchao},
  journal={arXiv preprint arXiv:2405.18426},
  year={2024}
}

About

Utilizing the texture and depth priors to initialize the Gaussian points

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Cuda 55.3%
  • Python 38.3%
  • C++ 5.9%
  • C 0.5%