https://github.com/erikwijmans/Pointnet2_PyTorch

PyTorch implementation of Pointnet2/Pointnet++
https://github.com/erikwijmans/Pointnet2_PyTorch

hydra point-cloud pointnet2 pytorch pytorch-lightning

Last synced: about 2 months ago
JSON representation

PyTorch implementation of Pointnet2/Pointnet++

• Host: GitHub
• URL: https://github.com/erikwijmans/Pointnet2_PyTorch
• Owner: erikwijmans
• License: unlicense
• Created: 2017-12-26T23:20:11.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2024-02-28T14:32:35.000Z (7 months ago)
• Last Synced: 2024-07-21T22:36:19.596Z (2 months ago)
• Topics: hydra, point-cloud, pointnet2, pytorch, pytorch-lightning
• Language: Python
• Homepage:
• Size: 953 KB
• Stars: 1,454
• Watchers: 17
• Forks: 332
• Open Issues: 67
• Metadata Files:
  • Readme: README.rst

Awesome Lists containing this project

README

        
Pointnet2/Pointnet++ PyTorch

============================

**Project Status**: Unmaintained.  Due to finite time, I have no plans to update this code and I will not be responding to issues.

* Implemention of Pointnet2/Pointnet++ written in `PyTorch `_.

* Supports Multi-GPU via `nn.DataParallel `_.

* Supports PyTorch version >= 1.0.0.  Use `v1.0 `_

  for support of older versions of PyTorch.

See the official code release for the paper (in tensorflow), `charlesq34/pointnet2 `_,

for official model definitions and hyper-parameters.

The custom ops used by Pointnet++ are currently **ONLY** supported on the GPU using CUDA.

Setup

-----

* Install ``python`` -- This repo is tested with ``{3.6, 3.7}``

* Install ``pytorch`` with CUDA -- This repo is tested with ``{1.4, 1.5}``.

  It may work with versions newer than ``1.5``, but this is not guaranteed.

* Install dependencies

  ::

    pip install -r requirements.txt

Example training

----------------

Install with: ``pip install -e .``

There example training script can be found in ``pointnet2/train.py``.  The training examples are built

using `PyTorch Lightning `_ and `Hydra `_.

A classifion pointnet can be trained as

::

  python pointnet2/train.py task=cls

  # Or with model=msg for multi-scale grouping

  python pointnet2/train.py task=cls model=msg

Similarly, semantic segmentation can be trained by changing the task to ``semseg``

::

  python pointnet2/train.py task=semseg

Multi-GPU training can be enabled by passing a list of GPU ids to use, for instance

::

  python pointnet2/train.py task=cls gpus=[0,1,2,3]

Building only the CUDA kernels

----------------------------------

::

  pip install pointnet2_ops_lib/.

  # Or if you would like to install them directly (this can also be used in a requirements.txt)

  pip install "git+git://github.com/erikwijmans/Pointnet2_PyTorch.git#egg=pointnet2_ops&subdirectory=pointnet2_ops_lib"

Contributing

------------

This repository uses `black `_ for linting and style enforcement on python code.

For c++/cuda code,

`clang-format `_ is used for style.  The simplest way to

comply with style is via `pre-commit `_

::

  pip install pre-commit

  pre-commit install

Citation

--------

::

  @article{pytorchpointnet++,

        Author = {Erik Wijmans},

        Title = {Pointnet++ Pytorch},

        Journal = {https://github.com/erikwijmans/Pointnet2_PyTorch},

        Year = {2018}

  }

  @inproceedings{qi2017pointnet++,

      title={Pointnet++: Deep hierarchical feature learning on point sets in a metric space},

      author={Qi, Charles Ruizhongtai and Yi, Li and Su, Hao and Guibas, Leonidas J},

      booktitle={Advances in Neural Information Processing Systems},

      pages={5099--5108},

      year={2017}

  }