Here is an explanation of the face recognition triggering:
The face recognition works hand-in-hand with the Proactive Dialogs. I.e. In the project itself you can set the "retrigger" time of the Proactive mode. So if you have set a value of 30sec there, for example, the Proactive Mode will start every 30sec and Pepper will say a random Proactive sentence to the person standing in front of her. (of course only as long as someone stays in the Proactive area)
If face recognition is now also switched on, then Proactive Mode and face recognition alternate. I.e. Assuming a person steps into Pepper's Proactive area, has been greeted by a proactive dialogue and remains in Pepper's field of vision, the sequence would be as follows:
Proactive Dialog -> 30sec pause (depending on what has been set) -> Face recognition Dialog -> 30sec pause -> Proactive Dialog -> 30sec pause -> Face recognition dialog ...
So if a person has been recognized by Pepper and greeted personally by Pepper, then the minimum time until the next person is recognized by Pepper is:
30sec (depending on the setting) + 1 Proactive Dialog + another 30sec
On the other hand: If Pepper has said a Proactive Dialog to someone and he then walks away and the next person is recognized by Pepper, then they will certainly be greeted directly.