Spatiotemporal Bilateral Gradient Filtering for Inverse Rendering

This repository contains the implementation for the SIGGRAPH Asia 2024 paper Spatiotemporal Bilateral Gradient Filtering for Inverse Rendering .

Running

The code requires Python 3.10 and above and installing required dependencies by:

pip install -r requirements.txt

Then you should be able to run each notebook in the notebooks directory (either as jupyter notebooks or as python scripts).

Important Files

• notebooks/texture_run.py: End-to-end texture optimization.
• notebooks/volume_run.py: End-to-end volume optimization.
• optimizers/filter_*.slang: Slang filter implementations.
• optimizers/spatiotemporal_adam*.py: Our spatiotemporal optimizers.

Citation

@inproceedings{WXY2024Stadam,
  title = {Spatiotemporal Bilateral Gradient Filtering for Inverse Rendering},
  author = {Chang, Wesley and Yang, Xuanda and Belhe, Yash and Ramamoorthi, Ravi and Li, Tzu-Mao},
  booktitle = {ACM SIGGRAPH Asia 2024 Conference Proceedings},
  numpages = {11},
  year = {2024},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  location = {Tokyo, Japan},
  series = {SIGGRAPH Asia '24},
  url = {https://doi.org/10.1145/3680528.3687606},
  doi = {10.1145/3680528.3687606}
}

Acknowledgement

The code examples adapted the following scenes:

• At the Window, from Bernhard Vogl.
• Autumn Field, from Jarod Guest and Sergej Majboroda.
• High-Res Smoke Plume, from JangaFX.