FADS: Few-shot Anomaly Detection and Classification Through Reinforced Data Selection

A Pytorch implementation of FADS.

Configuration

• Ubuntu 20.04
• NVIDIA driver 460.73.01
• CUDA 11.2
• Python 3.9
• PyTorch 1.9.0

Installation

This code requires the packages listed in requirements.txt. A virtual environment is recommended to run this code

On macOS and Linux:

python3 -m pip install --user virtualenv
python3 -m venv env
source env/bin/activate
pip install -r requirements.txt
deactivate

Reference: https://packaging.python.org/guides/installing-using-pip-and-virtual-environments/

Instructions

Please download the datasets from Google Drive: https://drive.google.com/drive/folders/1B3Y8oIvr4bS4IBO-3YzyXw1W6uWfZAxh?usp=sharing

Clone the template project, replacing my-project with the name of the project you are creating:

    git clone https://github.com/hanxiao0607/FADS.git my-project
    cd my-project

Run and test:

    python3 main_CERT.py
    or
    python3 main_IDS.py
    or
    python3 main_UNSW.py

Citation

@inproceedings{han2022few,
  title={Few-shot Anomaly Detection and Classification Through Reinforced Data Selection},
  author={Han, Xiao and Xu, Depeng and Yuan, Shuhan and Wu, Xintao},
  booktitle={2022 IEEE International Conference on Data Mining (ICDM)},
  pages={963--968},
  year={2022},
  organization={IEEE}
}