Documentation for Face Intellect 7.4. Documentation for other versions of Intellect is available too.
By default, when viewing the analytical data, the similar faces are sorted at the moment of data construction, which causes a system load and can lead to an error if there is a large number of unique faces within a specified time period.
To increase the speed of working with analytical data, it is recommended to enable the grouping of similar faces.
Note
If the grouping of similar faces is enabled, then during the face recognition server operation, each new captured face will be compared with a previously captured face that has the maximum level of similarity, but no less than specified in the settings (see Configuring the captured faces recognition parameters). If the detected face is not similar to any previous ones, then this face will become a parent for all subsequent faces similar to it. If the parent face is excluded from the collection, for example, when cleaning the archive on a loop or narrowing the data request period, the place of the common parent will be taken by the face which first referred to the common parent in an accessible time period.
If the grouping of similar faces is enabled, then the Similarity field on the Analytics tab of the Face recognition and search interface object will be hidden (see Viewing the analytics).
The grouping of similar persons is configured as follows:
- Go to the settings panel of the Face Recognition Server object. Go to the Analytics tab.
- Set the Indicate similar faces checkbox (1).
- Click the Apply button (2) to save the changes.
The grouping of similar faces is now configured.