SAFR can recognize individuals based on stored identities differing in age, makeup, facial hair, glasses, hoodies, or changes in facial expression. This makes the identity database resilient over time. This is accomplished by merging new face images with existing identities. SAFR automatically gives users this option when importing new facial images. You can also manually merge two stored identities via the People Window in the SAFR Desktop Client.
Furthermore, SAFR can automatically add new facial images to existing identities when it finds a significantly different but matching facial image. This is controlled via by selecting one of the video processing modes and the merge settings defined in the Recognition Preferences menu for that mode.