Skip to content
forked from zju3dv/LoFTR

"A case for using rotation invariant features in state of the art feature matchers", CVPRW 2022

License

Notifications You must be signed in to change notification settings

georg-bn/se2-loftr

This branch is 4 commits ahead of, 7 commits behind zju3dv/LoFTR:master.

Folders and files

NameName
Last commit message
Last commit date

Latest commit

f518157 · May 10, 2022

History

41 Commits
Jun 17, 2021
May 10, 2022
Jun 17, 2021
Apr 13, 2021
Jun 25, 2021
Aug 17, 2021
Aug 17, 2021
Apr 21, 2022
Jun 16, 2021
Apr 21, 2022
Jun 16, 2021
Apr 1, 2021
Apr 21, 2022
May 10, 2022
Apr 21, 2022
Jun 16, 2021
Apr 7, 2021
Jun 16, 2021

Repository files navigation

SE2-LoFTR

This is the repo for the CVPR Image Matching Workshop paper A case for using rotation invariant features in state of the art feature matchers. We implement a rotation equivariant LoFTR-version by using steerable CNNs.

Please see the LoFTR repo or the file LoFTR_README.md for instructions on how to obtain the data and run the code. We add a single dependency, namely e2cnn.

The new config-files configs/loftr/outdoor/loftr_ds_e2_dense*.py contain the parameters used for our SE2-LoFTR experiments.

Models trained on MegaDepth can be found here.

TODOs

  • Implement the rotation equivariant backbone as an EquivariantModule.

Cite

If you find this code useful in your research, please cite our paper as well as the LoFTR and e2cnn papers:

@inproceedings{bokman2022se2loftr,
    title={A case for using rotation invariant features in state of the art feature matchers},
    author={B\"okman, Georg and Kahl, Fredrik},
    booktitle={CVPRW},
    year={2022}
}

@article{sun2021loftr,
    title={{LoFTR}: Detector-Free Local Feature Matching with Transformers},
    author={Sun, Jiaming and Shen, Zehong and Wang, Yuang and Bao, Hujun and Zhou, Xiaowei},
    journal={CVPR},
    year={2021}
}

@inproceedings{e2cnn,
    title={{General E(2)-Equivariant Steerable CNNs}},
    author={Weiler, Maurice and Cesa, Gabriele},
    booktitle={Conference on Neural Information Processing Systems (NeurIPS)},
    year={2019},
}

About

"A case for using rotation invariant features in state of the art feature matchers", CVPRW 2022

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 92.8%
  • Python 6.9%
  • Shell 0.3%