About this document

Violent Crowd Detection via Deep Learning by darkrider9k is a document available to read on EtoBox.

This research presents a deep learning model for detecting violent crowd flows in Bangladesh using a dataset of 160 videos, equally divided between violent and non-violent scenarios. The study employs convolutional neural networks (CNN) and long short-term memory (LSTM) networks, with a pre-trained model achieving an accuracy of 95.67%, outperforming other methods. The lightweight model is designed for easy deployment in security systems like CCTV and UAVs, eliminating the need for constant human supervisio

Author

darkrider9k

Language

EN