Default behavior for determining “Trained” Faces

[mlavicka]

I have recently imported some photos with existing embedded face annotations previously created by iMatch.  The person already exists in iMatch and iMatch correctly adds them as confirmed faces to that persons face manager collection.  My question is: How does iMatch determine which of the newly imported faces will be used as a trained face?  Does it just go in sequential order and stops adding them as trained faces after reaching 10 trained faces? I make use of DOB/DOD, can it not use a face as trained if the date of the photo is outside that range?  Any other factors that affect how iMatch will determine which faces it uses as trained?  I know I can always modify whcih faces are used as "trained, I was just curious what the default behavior was.  Thanks.

[Mario]

Does it just go in sequential order and stops adding them as trained faces after reaching 10 trained faces? I

Yes. I tried not to make this overly complex and prone to errors. You can use the Face Manager to see and change trained faces. 

In most cases, trained faces are created while users assign persons to faces and confirm. If a imported face region has a tag matching the tag of an existing person, it is considered as confirmed, and may make it into the trained faces. The age range affects only suggestions made by IMatch during face matching, not confirmed faces.