This repo is the official implementation of STEAD: Spatio-Temporal Efficient Anomaly Detection for Time and Compute Sensitive Applications
Extracted X3D Features for UCF-Crime dataset
UCF-Crime X3D Features on Google drive
Feature extraction code also available for modification
pip install -r requirements.txt
python test.py
python main.py
@misc{gao2025steadspatiotemporalefficientanomaly,
title={STEAD: Spatio-Temporal Efficient Anomaly Detection for Time and Compute Sensitive Applications},
author={Andrew Gao and Jun Liu},
year={2025},
eprint={2503.07942},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2503.07942},
}