Media Server can run face recognition under a visual license or a surveillance license.

If your configuration includes tasks that require a visual license (such as optical character recognition or object detection), then a visual license must be used and the process action does not start unless a visual channel is available. In this case the MaxFaces parameter is ignored.

If the configuration only includes tasks that require a surveillance license, set the MaxFaces parameter to the maximum number of face identities you intend to use. If you set MaxFaces=Unlimited, a single visual channel is used to run the process action. The number of faces that you can recognize using a visual channel is not limited. Otherwise, the value you set determines the number of surveillance channels that are used for face recognition. If there are insufficient surveillance channels but a visual channel is available, a single visual channel is used instead.


Face detection and recognition with up to 250 faces requires one surveillance channel. Every additional 250 faces requires an additional surveillance channel. For example, if your face database contains 300 faces, set MaxFaces=300 and Media Server will use two surveillance channels for face detection and recognition.

If the value of MaxFaces is less than the number of faces in the database, the process action does not start. If you expect faces to be added to the database while the face recognition task runs, you should set MaxFaces to the expected total number of faces after the additional faces have been added.

Type: Integer
Default: 250
Required: No
Configuration Section: TaskName
Example: MaxFaces=500
See Also: Channels Configuration Parameters