Complex-Valued AutoEncoders for Object Discovery

We present the Complex AutoEncoder – an object discovery approach that takes inspiration from neuroscience to implement distributed object-centric representations. After introducing complex-valued activations into a convolutional autoencoder, it learns to encode feature information in the activations’ magnitudes and object affiliation in their phase values.

This repo provides a reference implementation for the Complex AutoEncoder (CAE) as introduced in our paper "Complex-Valued AutoEncoders for Object Discovery" (https://arxiv.org/abs/2204.02075) by Sindy Löwe, Phillip Lippe, Maja Rudolph and Max Welling.

Setup

Make sure you have conda installed (you can find instructions here).

Then, run bash setup.sh to download the 2Shapes, 3Shapes and MNIST&Shapes datasets and to create a conda environment with all required packages.

The script installs PyTorch with CUDA 11.3, which is the version we used for our experiments. If you want to use a different version, you can change the version number in the setup.sh script.

Run Experiments

To train and test the CAE, run one of the following commands, depending on the dataset you want to use:

python -m codebase.main +experiment=CAE_2Shapes

python -m codebase.main +experiment=CAE_3Shapes

python -m codebase.main +experiment=CAE_MNISTShapes

Citation

When using this code, please cite our paper:

@article{lowe2022complex,
  title={Complex-Valued Autoencoders for Object Discovery},
  author={L{\"o}we, Sindy and Lippe, Phillip and Rudolph, Maja and Welling, Max},
  journal={Transactions on Machine Learning Research (TMLR)},
  year={2022}
}

Contact

For questions and suggestions, feel free to open an issue on GitHub or send an email to loewe.sindy@gmail.com.