Adaptively weighted discrete Laplacian for inverse rendering
Hyeonjang An, Wonjun Lee, Bochang Moon
Overview
This code is the official implementation of The Visual Computer paper, Adaptively weighted discrete Laplacian for inverse rendering. For the further informations, please refer to the project page.
Requirements
We recommend running the code under conda environment.
conda create -f environment.yml
conda activate awl
Usage
We have tested our discrete Laplacian in existing differentiable rendering frameworks, Large Steps in Inverse Rendering of Geometry, and Continuous Remeshing For Inverse Rendering. Please replace the existing Laplacian in the frameworks with ours for the test.
Example code
from lap import laplacian_cotangent, laplacian_adaptive
L_c = laplacian_cotangent(mesh.verts, mesh.faces) # cotangent Laplacian
L_a = laplacian_adaptive(mesh.verts, mesh.faces, LAMBDA, SCALE) # adaptively weighted Laplacian
Here, LAMBDA,
Issue
If there is an issue, please send an email to this address, hyeonjang2021@gmail.com
License
All source codes are released under a BSD License
Citation
@article{An2023,
title={Adaptively weighted discrete Laplacian for inverse rendering},
author={An, Hyeonjang and Lee, Wonjun and Moon, Bochang},
journal={The Visual Computer},
year={2023},
issn={1432-2315},
doi={10.1007/s00371-023-02955-2},
url={https://doi.org/10.1007/s00371-023-02955-2}
}