A Survey on Video Anomaly Detection in Surveillance System

Surveillance video anomaly detection is a challenging and important task that aims to identify unusual or suspicious events in video streams without human intervention. Anomaly detection can help enhance the security and safety of various critical infrastructure systems, such as football stadiums, a...

Full description

Saved in:
Bibliographic Details
Published in:IEEE Recent Advances in Intelligent Computational Systems pp. 1 - 5
Main Authors: T, Manesh, Nataraj, Neetha K, Jayaraj, Akshaya, Joby, Alen, Ananthakrishnan, P V, Thankachan, Basil K
Format: Conference Proceeding
Language:English
Published: IEEE 16.05.2024
Subjects:
CNN
Mask R-CNN
One class detection
R-CNN
Safety
Security
Spatio Temporal autoencoders
Streaming media
Surveillance
Surveys
Training
Weapons
ISSN:2769-5565
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Surveillance video anomaly detection is a challenging and important task that aims to identify unusual or suspicious events in video streams without human intervention. Anomaly detection can help enhance the security and safety of various critical infrastructure systems, such as football stadiums, and cricket stadiums. However, anomaly detection also faces many difficulties, such as the diversity and complexity of normal and abnormal behaviors, the lack of sufficient and consistent annotations, and the high computational cost of processing large-scale video data. In this survey, we review the recent advances and challenges of artificial intelligence techniques in surveillance video anomaly detection. We highlight the strengths and limitations of different techniques, such as deep learning, generative models, graph neural networks, and multiple instance learning. We conclude by providing some future research directions and open issues in this field.
ISSN:2769-5565
DOI:10.1109/RAICS61201.2024.10690095