HugoP You clearly noticed English is not my mother language 🙂
It is not mine either, so no worries 🙂
HugoP I should wait until the program has found similar faces. In what time should this proces be done?
Reacting on suggestions is definitely much more relaxed and effortless way of marking people faces. Program constantly calculates suggestions on background, you can see what it is doing in the status string donw bellow. However it is hard to say how fast - that depends on too many factors.
I would suggest the following procedure (starting from very beginning):
1) Mark up manually few high quality faces for people you want in your collection. I would start from just a dozen of people.
2) Program will start coming up with suggestion, so switch to reactive mode and confirm/reject while looking for more people in your collection and adding them, filling up relations information, dealing with duplicates, etc. Basically focus on other things while program thinks and suggests.
3) At some point program will saturate and will stop providing you wil new suggestion, this time you will need to provide it more manual input for people you think you still have lots of photos. But this time select photos from different time period (like 10 years ago) in order to show program more variety. It will start suggesting again.
4) At some point this will stop being effective and there will be very few new suggetions. But you will already have something like 50-70% faces marked up. Now you will need to switch to fully manual mode - go event by event, check each photo, manually assing missed faces, add missed people, ignoring strangers, and so on.
This is the algorithm I would suggest in case you want your photos be 100% marked up, but with minimal effort. However, I am sure different users will have different opinion on what is most efficient process for them.
