Logical Anomaly Detection

This is the official Logicial Anomaly Detection Algorithm developed by Jin Er*, Qihui Feng, Yongli Mou

We recommend to use virtual environment for setting environment. This package have tested with python==3.10.13 under ubuntu 22.04.4 LTS

Please check the modify .env file and change it to the location where you save your package

Please make sure you have miniconda or anaconda installed in your system

Below is an example of how you can install all the relevant packages

1. Create virual environment

    # create env with miniconda/anaconda
    yes | conda create -n logic python=3.10
    pip install torchvision==0.12.0+cu113 torch==1.11.0+cu113 -i https://download.pytorch.org/whl/cu113
    pip install -e .
    pip install requirements.txt

Logical Anomaly Detection

The config file is saved in src/anomalib/models/logicad/config.yaml

python tools/train.py --config ./src/anomalib/models/logicad/config.yaml

or

python tools/train.py --model logicad

This package is built based on anomalib, openclip, lighting and hydra

@misc{falcon2019pytorch,
  title={PyTorch Lightning The lightweight PyTorch wrapper for high-performance AI research. Scale your models, not the boilerplate},
  author={Falcon, W and Team, TPL},
  year={2019}
}

@Misc{Yadan2019Hydra,
  author =       {Omry Yadan},
  title =        {Hydra - A framework for elegantly configuring complex applications},
  howpublished = {Github},
  year =         {2019},
  url =          {https://github.com/facebookresearch/hydra}
}
@misc{anomalib,
      title={Anomalib: A Deep Learning Library for Anomaly Detection},
      author={Samet Akcay and
              Dick Ameln and
              Ashwin Vaidya and
              Barath Lakshmanan and
              Nilesh Ahuja and
              Utku Genc},
      year={2022},
      eprint={2202.08341},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}