Repository for our ICIP 2021 paper, titled Cyclic Diffeomorphic Transformer Nets for Contour Alignment co-authored by: Ilya Kaufman, Ron Shapira Weber and Oren Freifeld.
licpab [2] is a python package supporting the CPAB transformations [1] in Numpy, Tensorflow and Pytorch.
Our code uses a modified version of libcpab which allows for circular warps.
Ilya Kafuman (email: ilyakau@post.bgu.ac.il)
- Standard Python(>=3.6) packages: numpy, matplotlib, tqdm
- PyTorch >= 1.4
- For Nvidia GPU iimplementation: CUDA==11.0 + appropriate cuDNN as well. You can follow the instructions here.
The code uses Python 3.6 and it was tested on Pytorch 1.9.0
Install pip and virtualenv
sudo apt-get install python-pip python-virtualenv
Clone the git project:
$ git clone git@github.com:BGU-CS-VIL/CDTNCA.git
Set up virtual environment:
$ mkdir <your_home_dir>/.virtualenvs
$ virtualenv -p python3 <your_home_dir>/.virtualenvs/CDTNCA
Activate virtual environment:
$ cd CDTNCA
$ source <your_home_dir>/.virtualenvs/CDTNCA/bin/activate
The requirements can be installed using:
pip install -r requirements.txt
Alignment of a subset of shapes generated from the 2D Shape Structure dataset [3]. For the entire archive, please visit: http://2dshapesstructure.github.io/
python alignment.py [args]
- tess_size: list, with the number of cells in each dimension
- zero_boundary: bool, determines if the velocity at the boundary is zero
- circularity: bool, allows for circular warps
[1] @article{freifeld2017transformations,
title={Transformations Based on Continuous Piecewise-Affine Velocity Fields},
author={Freifeld, Oren and Hauberg, Soren and Batmanghelich, Kayhan and Fisher, John W},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
year={2017},
publisher={IEEE}
}
[2] @misc{detlefsen2018,
author = {Detlefsen, Nicki S.},
title = {libcpab},
year = {2018},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/SkafteNicki/libcpab}},
}
[3] @article{Carlier:CG:20162:2DShapes,
title={The 2d shape structure dataset: A user annotated open access database},
author={Carlier, Axel and Leonard, Kathryn and Hahmann, Stefanie and Morin, Geraldine and Collins, Misha},
journal={Computers \& Graphics},
volume={58},
pages={23--30},
year={2016},
publisher={Elsevier}
}
This software is released under the MIT License (included with the software). Note, however, that if you are using this code (and/or the results of running it) to support any form of publication (e.g., a book, a journal paper, a conference paper, a patent application, etc.) then we request you will cite our paper:
@INPROCEEDINGS{kaufman2021contouralign,
author={Kaufman, Ilya and Weber, Ron Shapira and Freifeld, Oren},
booktitle={2021 IEEE International Conference on Image Processing (ICIP)},
title={Cyclic Diffeomorphic Transformer Nets For Contour Alignment},
year={2021},
pages={349-353}
}