Violence Detection using Machine Learning
Gokul · Zenodo (CERN European Organization for Nuclear Research) · 2023
The increase in violent acts in public places has encouraged the development of automated systems capable of detecting and responding to violent events in real time. We propose a violence detection system in this research that uses deep learning techniques to automatically identify violent behaviours in video data. The method extracts information from the input video frames before using a pre-trained convolutional neural network (CNN) to classify each frame as violent or non-violent.