|Logo|

.. image:: https://github.com/vislearn/FrEIA/workflows/CI/badge.svg :alt: Build Status

This is the Fr\ amework for E\ asily I\ nvertible A\ rchitectures (FrEIA).

• Construct Invertible Neural Networks (INNs) from simple invertible building blocks.
• Quickly construct complex invertible computation graphs and INN topologies.
• Forward and inverse computation guaranteed to work automatically.
• Most common invertible transforms and operations are provided.
• Easily add your own invertible transforms.

.. contents:: Table of contents :backlinks: top :local:

Papers

Our following papers use FrEIA, with links to code given below.

"Generative Classifiers as a Basis for Trustworthy Image Classification" (CVPR 2021)

• Paper: https://arxiv.org/abs/2007.15036
• Code: https://github.com/RayDeeA/ibinn_imagenet

"Training Normalizing Flows with the Information Bottleneck for Competitive Generative Classification" (Neurips 2020)

• Paper: arxiv.org/abs/2001.06448 <https://arxiv.org/abs/2001.06448>_
• Code: github.com/vislearn/IB-INN <https://github.com/vislearn/IB-INN>_

"Disentanglement by Nonlinear ICA with General Incompressible-flow Networks (GIN)" (ICLR 2020)

• Paper: arxiv.org/abs/2001.04872 <https://arxiv.org/abs/2001.04872>_
• Code: github.com/vislearn/GIN <https://github.com/vislearn/GIN>_

"Guided Image Generation with Conditional Invertible Neural Networks" (2019)

• Paper: arxiv.org/abs/1907.02392 <https://arxiv.org/abs/1907.02392>_
• Supplement: drive.google.com/file/d/1_OoiIGhLeVJGaZFeBt0OWOq8ZCtiI7li <https://drive.google.com/file/d/1_OoiIGhLeVJGaZFeBt0OWOq8ZCtiI7li>_
• Code: github.com/vislearn/conditional_INNs <https://github.com/vislearn/conditional_INNs>_

"Analyzing inverse problems with invertible neural networks." (ICLR 2019)

• Paper: arxiv.org/abs/1808.04730 <https://arxiv.org/abs/1808.04730>_
• Code: github.com/vislearn/analyzing_inverse_problems <https://github.com/vislearn/analyzing_inverse_problems>_

Installation

FrEIA has the following dependencies:

+---------------------------+-------------------------------+ | Package | Version | +---------------------------+-------------------------------+ | Python | >= 3.7 | +---------------------------+-------------------------------+ | Pytorch | >= 1.0.0 | +---------------------------+-------------------------------+ | Numpy | >= 1.15.0 | +---------------------------+-------------------------------+ | Scipy | >= 1.5 | +---------------------------+-------------------------------+

Through pip ^^^^^^^^^^^^^^^^^^^^^^^^^^^

.. code:: sh

pip install FrEIA

Manually ^^^^^^^^^^^^^^^^^^^^^^^^^^^

For development:

.. code:: sh

first clone the repository

git clone https://github.com/vislearn/FrEIA.git cd FrEIA

install the dependencies

pip install -r requirements.txt

install in development mode, so that changes don't require a reinstall

python setup.py develop

Documentation

The full manual can be found at https://vislearn.github.io/FrEIA including

• Quickstart guide <https://vislearn.github.io/FrEIA/_build/html/tutorial/quickstart.html>_
• Tutorial <https://vislearn.github.io/FrEIA/_build/html/tutorial/tutorial.html>_
• Examples <https://vislearn.github.io/FrEIA/_build/html/tutorial/examples.html>_
• API documentation <https://vislearn.github.io/FrEIA/_build/html/index.html#package-documentation>_

How to cite this repository

If you used this repository in your work, please cite it as below:

.. code-block::

@software{freia, author = {Ardizzone, Lynton and Bungert, Till and Draxler, Felix and Köthe, Ullrich and Kruse, Jakob and Schmier, Robert and Sorrenson, Peter}, title = {{Framework for Easily Invertible Architectures (FrEIA)}}, year = {2018-2022}, url = {https://github.com/vislearn/FrEIA} }

.. |Logo| image:: docs/freia_logo_invertible.svg