Divide-and-Conquer

This is the repository for the paper Divide and Conquer in Video Anomaly Detection: A Comprehensive Review and New Approach.

The experimental code in this paper is based on the code from two papers, STG-NF and Jigsaw. Following the instructions provided in the papers, the replication of the research can be accomplished.

We replicated these two papers on different datasets and attempted the divide-and-conquer approach. The results are as follows:

We achieved the highest AUC score on the ShanghaiTech dataset, as of August 20, 2023. Please refer to the Paper with Code for more details.

As you can see, our results reported a negative outcome, aligning with the spirit of open science.