Home

Awesome

Point to Mesh distance computation

This package provides a PyTorch module that performs point to surface queries on the GPU

Table of Contents

License

Software Copyright License for non-commercial scientific research purposes. By downloading and/or using the Model & Software (including downloading, cloning, installing, and any other use of this github repository), you acknowledge that you have read these terms and conditions, understand them, and agree to be bound by them. If you do not agree with these terms and conditions, you must not download and/or use the Model & Software. Any infringement of the terms of this agreement will automatically terminate your rights under this License.

Description

This repository provides a PyTorch wrapper around a CUDA kernel that implements the method described in Maximizing parallelism in the construction of BVHs, octrees, and k-d trees. More specifically, given a batch of meshes it builds a BVH tree for each one, which can then be used for distance quries.

Installation

Before installing anything please make sure to set the environment variable $CUDA_SAMPLES_INC to the path that contains the header helper_math.h , which can be found in the CUDA Samples repository. To install the module run the following commands:

1. Install the dependencies

pip install -r requirements.txt 

2. Run the setup.py script

python setup.py install

If you want to modify any part of the code then use the following command:

python setup.py build develop

Examples

Dependencies

  1. PyTorch

Example dependencies

  1. open3d
  2. mesh

Running on Cluster

If you want to run this on the cluster you need to build it using the GPU availabe on the cluster. If you use the local build there might be GPU architecture compatibility issue and you can encounter following error message

RuntimeError: parallel_for failed: unrecognized error code: unrecognized error code

Citation

If you find this code useful in your research please cite the relevant work(s) of the following list:

@inproceedings{Karras:2012:MPC:2383795.2383801,
    author = {Karras, Tero},
    title = {Maximizing Parallelism in the Construction of BVHs, Octrees, and K-d Trees},
    booktitle = {Proceedings of the Fourth ACM SIGGRAPH / Eurographics Conference on High-Performance Graphics},
    year = {2012},
    pages = {33--37},
    numpages = {5},
    url = {https://doi.org/10.2312/EGGH/HPG12/033-037}, 
    doi = {10.2312/EGGH/HPG12/033-037},
    publisher = {Eurographics Association}
}

Contact

The code of this repository was implemented by Vassilis Choutas. For commercial licensing, please contact ps-licensing@tue.mpg.de.