ICON: Learning Regular Maps through Inverse Consistency

ICON: Learning Regular Maps through Inverse Consistency. Hastings Greer, Roland Kwitt, Francois-Xavier Vialard, Marc Niethammer. ICCV 2021 https://arxiv.org/abs/2105.04459

Cite this work

@InProceedings{Greer_2021_ICCV,
    author    = {Greer, Hastings and Kwitt, Roland and Vialard, Francois-Xavier and Niethammer, Marc},
    title     = {ICON: Learning Regular Maps Through Inverse Consistency},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
    month     = {October},
    year      = {2021},
    pages     = {3396-3405}
}

Video Presentation

Running our code

We are available on PyPI!

pip install icon-registration

To run our pretrained model in the cloud on 4 sample image pairs from OAI knees (as above), visit our google colab notebook

To train from scratch on the synthetic triangles and circles dataset:

git clone https://github.com/uncbiag/ICON
cd ICON

pip install -e .

python training_scripts/2d_triangles_example.py