Deep Learning-Integrated Crime Detection System Using Convolutional Neural Networks

Abstract

The work delivered here provides a new deep learning-based approach for crime detection, using Convolutional Neural Networks (CNNs) as a tool. Given the abundance of surveillance data which is becoming more available, the provision of automated systems to the law enforcement agencies ought to help in the identification of crimes seems to be the urgent need of the hour. We aim by this study to build a CNN model that could identify scenes in video simplifying them as criminal or non-criminal. The model is trained on the large dataset of the video clips which are labeled and it can promise the accuracy and efficiency attracted by this model. We test the proposed system on various baseline datasets and then its performance compared with state-of-the-art methods. The experimental results thereby show the effectiveness of our method in the detection of criminal activities, which highlights its potentiality for real world implementation in preventive measures and crime enforcement.