Relational Fusion Networks

This library contains a reference implementation of the Relational Fusion Network (RFN). The RFN first appeared in a paper presented at ACM SIGSPATIAL 2019 [1] which is available through the ACM Digital Library. An extended version of this paper has since appeared in IEEE Transactions on Intelligent Transportation Systems [2] and is available through IEEE Xplore. A preprint of the extended paper is also available on arXiv.

You can install the required dependencies using pipenv or pip by executing pipenv install and pip install -r requirements.txt, respectively, in the root directory. See Tutorial.ipynb for instructions on how to use the library.

Citation

If you find RFNs relevant to your research, please cite the following paper:

@article{rfn,
  author={T. S. {Jepsen} and C. S. {Jensen} and T. D. {Nielsen}},
  journal={IEEE Transactions on Intelligent Transportation Systems}, 
  title={Relational Fusion Networks: Graph Convolutional Networks for Road Networks}, 
  year={2020},
  pages={1-12},}

References

[1] Jepsen, T.S., Jensen, C.S. and Nielsen, T.D., 2019. Graph Convolutional Networks for Road Networks. In Proceedings of the 27th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (pp. 460-463).
[2] Jepsen, T.S., Jensen, C.S. and Nielsen, T.D., 2020. Relational Fusion Networks: Graph Convolutional Networks for Road Networks. IEEE Transactions on Intelligent Transportation Systems (pp. 1-12).