CUE-Net: Violence Detection Video Analytics with Spatial Cropping, Enhanced UniformerV2 and Modified Efficient Additive Attention

Introduction

This repository contains the code and resources for the "CUE-Net: Violence Detection Video Analytics with Spatial Cropping, Enhanced UniformerV2 and Modified Efficient Additive Attention" paper accepted for 6th Workshop on Affective Behavior Analysis in-the-wild (ABAW) in CVPR Conference 2024 (https://openaccess.thecvf.com/content/CVPR2024W/ABAW/papers/Senadeera_CUE-Net_Violence_Detection_Video_Analytics_with_Spatial_Cropping_Enhanced_UniformerV2_CVPRW_2024_paper.pdf).

Abstract

In this paper we introduce CUE-Net, a novel architecture designed for automated violence detection in video surveillance. As surveillance systems become more prevalent due to technological advances and decreasing costs, the challenge of efficiently monitoring vast amounts of video data has intensified. CUE-Net addresses this by combining spatial Cropping with an enhanced version of the UniformerV2 architecture, integrating convolutional and self-attention mechanisms alongside a novel Modified Efficient Additive Attention mechanism (which reduces the quadratic time complexity of self-attention) so as to effectively and efficiently identify violent activities. This approach aims to overcome traditional challenges such as capturing distant or partially obscured subjects within video frames. By focusing on both local and global spatio-temporal features, CUE-Net achieves state-of-the-art performance on the RWF-2000 and RLVS datasets, surpassing existing methods.

Codes

Spatial Cropping Module

You will find the Jupyter Notebook which explains the Spatial Cropping Module in our Algorithm in the folder "Spatial_Cropping_Module".

Enhanced Uniformer V2 Architecture

You will find instructions on how to install Uniforemr V2 architecture based on the original GitHUb repository and do the modifications as instructed in the "Enhanced_Uniformer_V2" folder.