Video Anomaly Detection (VAD)
Automatic anomaly detection without rules set-up
Large camera networks are becoming an integral part of many smart cities, and as part of solutions in many other domains (e.g., traffic, intelligent transportation systems (ITS), parking, buildings, retails and security). These sensor networks collect significant volumes of video data; however, this data must be filtered so that users are not overloaded with irrelevant information. . The challenge is to be able to learn, automatically, what is relevant in a camera’s field-of-view.
Efficient filtering can be achieved with AGT’s algorithms for anomaly detection. With this machine learning-based approach, our system automatically learns what is normal and reports deviations to system users in real time. In contrast to traditional methods, no rules set-up is required.
This is a game-changer in the industry.