rmurai0610/diff-gaussian-rasterization-w-pose

Differential Gaussian Rasterization with Camera Pose Jacobians

This software is used as the rasterization engine in the paper "Gaussian Splatting SLAM", and supports:

• Analytical gradient for SE(3) camera poses.
• Analytical gradient for rendered depth.

The code is built on top of the original Differential Gaussian Rasterization used in "3D Gaussian Splatting for Real-Time Rendering of Radiance Fields".

If you can make use of it in your own research, please be so kind to cite both papers.

BibTeX

@Article{kerbl3Dgaussians,
      author       = {Kerbl, Bernhard and Kopanas, Georgios and Leimk{\"u}hler, Thomas and Drettakis, George},
      title        = {3D Gaussian Splatting for Real-Time Radiance Field Rendering},
      journal      = {ACM Transactions on Graphics},
      number       = {4},
      volume       = {42},
      month        = {July},
      year         = {2023},
      url          = {https://repo-sam.inria.fr/fungraph/3d-gaussian-splatting/}
}

@inproceedings{Matsuki:Murai:etal:CVPR2024,
  title={{G}aussian {S}platting {SLAM}},
  author={Hidenobu Matsuki and Riku Murai and Paul H. J. Kelly and Andrew J. Davison},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year={2024}
}

diff-gaussian-rasterization-w-pose's Issues

Compared to Original Implementaiton

@muskie82 @rmurai0610
Thank you for your excellent work. I want to ask, compared with the original rendering code, will it take up much more GPU memory? If so, approximately how much?