Learns what normal looks like for each individual camera and alerts on activity that deviates from it, with no rules, zones, or labels to configure. A short learning period runs when you enable it, after which the camera continuously watches for the unexpected.
Visylix Anomaly Detection is self-learning surveillance for the things you did not think to write a rule for. Instead of asking you to define zones, lines, or labels, it learns what normal activity looks like for each individual camera during a short learning period when you first enable it. From then on it continuously watches that scene and alerts on activity that deviates from the pattern it learned, catching unforeseen events that rule-based analytics would miss. Because the baseline is per camera, it adapts to each scene on its own and works best as a complement to the rule-based models, adding a safety net for the unexpected at perimeters, remote sites, and unmanned facilities.
Core capabilities of the Anomaly Detection model.
Learns what normal looks like for each individual camera without manual setup.
Needs no zones, lines, or labels to configure before it starts working.
Alerts on the unexpected that rule-based analytics were never told to look for.
Runs a short learning period when you enable it, then starts watching on its own.
Builds a per-camera baseline so every scene is judged on its own normal.
Adds a safety net alongside the rule-based models rather than replacing them.
Real-world applications for Anomaly Detection.
Watches fence lines and remote locations for activity that does not belong.
Flags unusual movement in spaces that should be quiet outside working hours.
Adds an extra layer of awareness at high-value and sensitive sites.
Keeps watch over lights-out sites where no operator is present to notice.
Performance and deployment details.
Add Anomaly Detection to your video pipeline in minutes.
Assign the model to specific cameras with zone definitions and sensitivity settings through the web UI or API.
The model processes video frames in real time, generating structured detection events with bounding boxes and metadata.
Receive instant alerts via webhooks, trigger automated workflows, or query detections through the REST API.
Explore other computer vision capabilities.
Talk to our team to see this model in action on your video feeds.