Skip to content

Latest commit

 

History

History
87 lines (65 loc) · 2.84 KB

README.md

File metadata and controls

87 lines (65 loc) · 2.84 KB

Attention Network with Optional Activation Function for Unsupervised Video Anomaly Detection

This repository contains the PyTorch implementation of our research paper published in the ETRI Journal. You can read the full paper here: ETRI Journal.

📌 Overview

This project implements a Attention-based Autoencoder Network with Optional Activation Function designed for unsupervised video anomaly detection. The method leverages PyTorch for training and evaluation, ensuring efficient learning and robust performance.

🚀 Getting Started

1. Prerequisites

Ensure you have the following dependencies installed before running the code:

  • Python 3.8+
  • PyTorch 1.7.1
  • Torchvision 0.8.2

2. Install Dependencies

To install the required packages, run:

pip install -r requirements.txt

3. Prepare Datasets

Download and place the dataset in the datasets/ folder. Ensure the data is structured correctly before running the model.

4. Configure Hyperparameters

Modify hyperparameters such as learning rate, number of epochs, batch size, etc., in the .yaml file inside the configs/ directory.

🔥 Training the Model

To train the model, run:

python train.py

🎯 Testing the Model

To evaluate the model on test data, run:

python test.py

📂 Project Structure

Unsupervised-Video-Anomaly-Detection/
│── config/           # Configuration files (configs.yaml)
│── datasets/          # Folder to store datasets
│── models/            # Model architecture definitions
│── utils/             # Helper functions and utilities
│── train.py           # Script for training the model
│── test.py            # Script for testing the model
│── requirements.txt   # List of dependencies
│── README.md          # Project documentation

🛠 Troubleshooting & Tips

  • Ensure you have the correct Python and PyTorch versions installed.
  • If dependencies are missing, manually install them using pip install <package-name>.
  • Adjust hyperparameters in configs.yaml for improved model performance.

📜 Citation

If you find this work useful, please cite our paper:

@article{rakhmonov2024aonet,
  title={AONet: Attention network with optional activation for unsupervised video anomaly detection},
  author={Rakhmonov, Akhrorjon Akhmadjon Ugli and Subramanian, Barathi and Amirian Varnousefaderani, Bahar and Kim, Jeonghong},
  journal={ETRI Journal},
  volume={46},
  number={5},
  pages={890--903},
  year={2024},
  publisher={Wiley Online Library}
}

📬 Contact

For any questions or contributions, feel free to open an issue or reach out.