Face recognition
Face recognition on FortiRecorder uses artificial intelligence (AI) to identify unique faces and enact policies based on configured information. For example, you can identify faces as "known" and enter information about them in the face recognition database, such as their job role or their department. The information can be used to track when the person usually appears and how often so that you can create notification policies about unexpected occurrences or unauthorized locations.
Alternatively, you can process face modeling and face recognition in a distributed way, with FortiCentral installations. Both methods store their analysis results on FortiRecorder. However, the results are stored in separate databases. On FortiRecorder:
|
Face recognition requires Internet access and is supported on FortiRecorder-200D, 400D, 400F and VM. For VM users, a minimum of 4 GB of memory is required, but 8 GB or more is recommended. |
To enable face recognition in the GUI
-
Go to Dashboard > Status. In the License Information widget, verify that you have a valid license for face recognition.
If the license is expired or invalid, then only the previous 7 days of face recognition data (face clusters, timeline, events, and activity) will display.
- Go to Camera > Configuration > Camera Profile.
- Select the camera profile to enable face recognition on and then select Edit.
- Expand the Recording section and enable Motion detection in the Recording.
- Enable FortiRecorder in the Store.
- Click OK.
- Go to Face Recognition > User Asset > AI Cameras.
-
In the AI Status column, enable AI for one or more cameras.
Face recognition AI only analyzes video from the selected cameras. Analysis results in face clusters, which you must review and confirm in order to ensure accurate face recognition. Continue with Identifying faces.
To enable face recognition in the CLI
Enter these commands:
config system global
set face-recognition enable
end