Torch PBR

Torch PBR is a light-weight library for differentiable PBR written purely in Python/PyTorch.

For example usage, please check IntrinsicAvatar.

If you find our code useful, please cite:

@inproceedings{WangCVPR2024,
  title   = {IntrinsicAvatar: Physically Based Inverse Rendering of Dynamic Humans from Monocular Videos via Explicit Ray Tracing},
  author  = {Shaofei Wang and Bo\v{z}idar Anti\'{c} and Andreas Geiger and Siyu Tang},
  booktitle = {IEEE Conf. on Computer Vision and Pattern Recognition (CVPR)},
  year    = {2024}
}

TODO

• Implement importance sampling and PDF evaluation for EnvironmentLightSG (https://arxiv.org/abs/2303.16617)

Acknowledgement

The basic utility functions in utils/nvdiffrecmc_util.py are borrowed from nvdiffrecmc. We thank authors of the paper for their wonderful works which greatly facilitates the development of this project.