Media & Enterprise
Most sites do not need more cameras. They need the cameras they have to search recorded footage, watch a restricted area, notice a crowd building and raise something while nobody happens to be looking at that screen. FaceOff adds all of that to a camera network you already own, with an operator deciding on every alert
Overview
Add search, restricted area alerts, crowd monitoring and weapon detection to camera networks that are already installed.
A camera estate is usually finished long before it becomes useful. Screens get watched when somebody is free, footage gets searched when somebody has hours to spare, and the value of the whole installation quietly comes to depend on whether a person happened to be looking in the right direction at the right moment.
The gap is not coverage, it is attention. Nobody can watch forty feeds at once, and after an incident the recording that holds the answer is reviewed by scrubbing through hours of it. By the time the footage has given up what it knows, the moment when somebody could have acted on it has usually passed.
Everything below works with an installed estate rather than replacing it. The cameras carry on doing what they already do, and what changes is that a question can be put to them, and that something can be raised while there is still time to do something about it.
What you deploy
These are the products integrators and camera manufacturers add most often to an estate that is already installed, covering watch list matching, restricted areas, crowd safety, weapon detection and searching recorded footage.
Live face recognition across your camera network, matched against your watch-lists.
ExploreKey capabilities
What this sector asks for most often. Each one runs at the moment a decision is being made, and each returns a result a reviewer can explain to somebody else.
Specific areas can be marked, and entry by somebody who should not be there raises an alert straight away. The area stays watched whether or not anybody is looking at that particular screen, which is the entire point of an alert.
Each area can be given its own limit and monitored across several cameras at once, and the area is named when numbers pass that limit. Crowd safety depends on acting early, and early means before the problem is obvious from the control room.
Faces on live cameras are matched against the lists you build and maintain, and every match arrives as a graded alert for an operator to accept or reject. It runs on your own hardware inside your own network, and the operator's verdict is stored with the alert.
These run on an estate that already exists rather than requiring a new one, and more than one check can share the same coverage, so a site can watch for crowding and look for a threat on the same cameras at the same time.
Recorded video can be searched for a person by the clothing they were wearing and the bags or belongings they were carrying, returning a snapshot and a record for each person found. A search can run when no name is known, which is almost always the situation right after an incident.
Crowded areas are monitored continuously for a visible weapon and an alert is raised when one is identified, so staff hear about it from the system rather than from the reaction of the people standing nearest to it.
Where it is used
Every entry names a product on this site and describes it the way its own page does.
CrossCam PersonTrack matches faces against the watch lists you build, with an operator reviewing each match.
SecureZone Monitor raises an alert when someone enters an area they should not, or when it holds more people than allowed.
CrowdSentinel AI gives every area its own limit and names the area when numbers pass it.
WeaponGuard AI looks for a visible weapon among a crowd, and AnonymousPersonID makes recorded footage searchable by appearance.
Common questions
In most cases yes, and it is checked before anything is bought rather than after installation. CrossCam PersonTrack includes a survey of camera models and mounting distances that flags any position which will not deliver a good enough view.
No. These add search, alerting and monitoring to the camera estate and the recorder you already run, rather than asking you to change either of them.
Yes. The same coverage can carry more than one kind of check, which is how a site watches crowding and looks for a threat at once without installing a second set of cameras.
Yes. AnonymousPersonID searches recorded footage by clothing colour and by the bags and belongings a person was carrying, and returns a snapshot and a record for each person found, so a search can run when no name or personal detail is known.
No. A match arrives as a graded alert that an operator accepts or rejects, and that verdict is stored with the alert. The decision belongs to your staff, and the record shows that it did.
CrossCam PersonTrack runs on your own hardware and makes no call outside your network in normal running, so an estate can be covered without footage being sent anywhere else.
Recognition keeps running. CrossCam PersonTrack can run with no internet at all, which matters for a remote site or a secure area where connectivity is not something anybody should be depending on.
Yes. Areas can be marked individually and each given its own crowd limit, so a loading bay, a public concourse and a secure store can all be treated differently on the same estate.
Yes. Any method of detection can be mistaken in either direction, which is exactly why a match is presented as a graded alert for a person to adjudicate rather than as an instruction. What gets recorded and acted on is the operator's decision, not the system's.
Bring your own footage. We will run it through the Adaptive Cognito Engine and walk you through the findings.
