differential geometric computer vision for deep learning
Project description
The PyTorch Geometry package is a geometric computer vision library for PyTorch.
It consists of a set of routines and differentiable modules to solve generic geometry computer vision problems. At its core, the package uses PyTorch as its main backend both for efficiency and to take advantage of the reverse-mode auto-differentiation to define and compute the gradient of complex functions.
Development Setup
Assuming that you are on ubuntu 16.04, with nvidia-drivers installed.
In bash, source the path.bash.inc script. This will install a local conda environment under ./.dev_env, which includes pytorch and some dependencies (no root required).
source ./path.bash.inc
python -c "import torchgeometry; print(torchgeometry.__version__)"
To install, or update the conda environment run setup_dev_env.sh
./setup_dev_env.sh
Quick Usage
import torch
import torchgeometry as tgm
x_rad = tgm.pi * torch.rand(1, 3, 3)
x_deg = tgm.rad2deg(x_rad)
torch.allclose(x_rad, tgm.deg2rad(x_deg)) # True
Examples
Run our Jupyter notebooks examples to learn to use the library.
Installation
From source:
python setup.py install
Testing
python setup.py test
Cite
If you are using torchgeometry in your research-related documents, it is recommended that you cite the poster.
@misc{Arraiy2018,
author = {E. Riba, M. Fathollahi, W. Chaney, E. Rublee and G. Bradski}
title = {torchgeometry: when PyTorch meets geometry},
booktitle = {PyTorch Developer Conference},
year = {2018},
url = {https://drive.google.com/file/d/1xiao1Xj9WzjJ08YY_nYwsthE-wxfyfhG/view?usp=sharing}
}
Future work
The roadmap will add more functions to allow developers to solve geometric problems.
Contributing
We appreciate all contributions. If you are planning to contribute back bug-fixes, please do so without any further discussion. If you plan to contribute new features, utility functions or extensions, please first open an issue and discuss the feature with us.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for torchgeometry-0.1.1-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 4f61d35cb2827ff2bc83ca54af0f6c3108d9bf18fb17e2ac4f9766edbc285ee8 |
|
MD5 | 3d9ff4bc097ddb1470af5d3ce910892d |
|
BLAKE2b-256 | 36cf12f033742afceeb27833bc346d42697f13788f3441631672d8e9bbbe73d5 |