RWF-2000: An Open Large Scale Video Database for Violence Detection

TL;DR

This paper introduces RWF-2000, a large-scale real-world surveillance video dataset for violence detection, and proposes a novel 3D-CNN based method achieving high accuracy, with open access to data and code.

Contribution

It provides a new extensive violence detection dataset and a novel Flow Gated Network method utilizing 3D-CNNs and optical flow for improved accuracy.

Findings

Achieved 87.25% accuracy on RWF-2000 test set.

Provided an open dataset and source code for violence detection.

Compared favorably with existing datasets and methods.

Abstract

In recent years, surveillance cameras are widely deployed in public places, and the general crime rate has been reduced significantly due to these ubiquitous devices. Usually, these cameras provide cues and evidence after crimes are conducted, while they are rarely used to prevent or stop criminal activities in time. It is both time and labor consuming to manually monitor a large amount of video data from surveillance cameras. Therefore, automatically recognizing violent behaviors from video signals becomes essential. This paper summarizes several existing video datasets for violence detection and proposes the RWF-2000 database with 2,000 videos captured by surveillance cameras in real-world scenes. Also, we present a new method that utilizes both the merits of 3D-CNNs and optical flow, namely Flow Gated Network. The proposed approach obtains an accuracy of 87.25% on the test set of our proposed database. The database and source codes are currently open to access.