Spot The Odd One Out: Regularized Complete Cycle Consistent Anomaly Detector GAN

RCALAD official code

The code for the paper "Spot The Odd One Out: Regularized Complete Cycle Consistent Anomaly Detector GAN" (authors: Zahra Dehghanian, Saeed Saravani, Maryam Amirmazlaghani and Mohamad Rahmati) is now open source!

Please reach us via emails or via github issues for any enquiries!

Prerequisites.

This code package was developed and tested with Python 3.7.6. Make sure all dependencies specified in the requirements.txt file are satisfied before running the model. This can be achieved by

conda create --name tf1 python=3.7
conda activate tf1
pip3 install -r requirements.txt

Usage.

Running the code with different options

python main.py <model>  <dataset> --nb_epochs=<number_epochs> --label=<0, 1, 2, 3, 4, 5, 6, 7, 8, 9> --sn=<bool> --enable_dzz=<bool> --rd=<int> --d-<int> etc. 

The default option will run the RCALAD model on Arrhythmia dataset with 1000 epoches.

Please refer to the argument parser in main.py for more details.

Cite.

Please cite our work if you find it useful for your research and work.

@article{Dehghanian2023Spot,
  title={Spot The Odd One Out: Regularized Complete Cycle Consistent Anomaly Detector GAN},
  author={Zahra Dehghanian, Saeed Saravani, Maryam Amirmazlaghani and Mohamad Rahmati},
  year={2023}
}