Awesome
<h1 align="center">Deep Graph Matching Consensus</h1> <img width="100%" src="https://raw.githubusercontent.com/rusty1s/deep-graph-matching-consensus/master/figures/overview.png?token=ABU7ZATGTDHG6DPI3ANYBUS5TXPOQ" />
This is a PyTorch implementation of Deep Graph Matching Consensus, as described in our paper:
Matthias Fey, Jan E. Lenssen, Christopher Morris, Jonathan Masci, Nils M. Kriege: Deep Graph Matching Consensus (ICLR 2020)
Requirements
- PyTorch (>=1.2.0)
- PyTorch Geometric (>=1.5.0)
- KeOps (>=1.1.0)
Installation
$ python setup.py install
Head over to our documentation for a detailed overview of the DGMC module.
Running examples
We provide training and evaluation procedures for the PascalVOC with Berkely annotations dataset, the WILLOW-ObjectClass dataset, the PascalPF dataset, and the DBP15K dataset. Experiments can be run via:
$ cd examples/
$ python pascal.py
$ python willow.py
$ python pascal_pf.py
$ python dbp15k.py --category=zh_en
Cite
Please cite our paper if you use this code in your own work:
@inproceedings{Fey/etal/2020,
title={Deep Graph Matching Consensus},
author={Fey, M. and Lenssen, J. E. and Morris, C. and Masci, J. and Kriege, N. M.},
booktitle={International Conference on Learning Representations (ICLR)},
year={2020},
}
Running tests
$ python setup.py test