Anomaly-Detection-in-Surveillance-Videos

It contains the code for the following paper Anomaly detection in surveillance videos using transformer based attention model

Dataset

ShanghaiTech Campus Dataset (https://svip-lab.github.io/dataset/campus_dataset.html)

Setup

For feature extraction we've followed this repo on default settings.

Extracted videoswin features link

Specify the features path in Args class and change the rgb_list_file path in the Dataset class.

To run the code execute the provided notebook cell by cell.

Thanks repo for the starter code.