Enhanced public safety
The system empowers controlling authorities to take proactive measures, ensuring the safety and security of individuals by autonomously detecting anomalous events in real time.
Revolutionising public safety
Advanced video analytics for anomaly detection in surveillance
Our research group has developed a video surveillance system that autonomously detects and monitors anomalous events, so emergencies get a swift response.
VADE: Video Anomaly Detection Engine
What is a video anomaly?
Video anomalies are unusual appearances or motion patterns that differ from the normal stream. Examples include:
- Abuse
- Theft
- Arson
- Assault
- Fights
- Crime scene
Benefits of the system
Why autonomous anomaly detection matters for public safety.
Increased efficiency
It automates the detection and monitoring of anomalous activities, reducing reliance on manual surveillance and enabling 24/7 autonomous surveillance without human limitations.
Accuracy and reliability
Using machine learning, computer vision and AI, the system identifies normal and anomalous behaviour in video streams with high accuracy, minimising false alarms.
Alert generation
It generates alerts and notifications when anomalies are detected, so security personnel can respond promptly to potential threats.
Scalability and adaptability
It supports multiple CCTV cameras, making it suitable for deployment in diverse environments.
Cost-effectiveness
By automating surveillance tasks and optimising resource allocation, it costs less than traditional manual surveillance.
Real-time monitoring and response
Real-time inference enables immediate detection of anomalous events as they occur, so incidents can be handled quickly.
Results on a real-world camera stream
Detections on a real surveillance camera video stream. Select the image to enlarge it.
