I saw a similar request in the video editor and thought I’d request the same for Photo AI: the ability to generate face models for known faces to aid Photo AI in face recovery.
I’m an amateur who purchased Photo AI to enhance old 640x480 digital photos and scans of old film photos. I have a large number of photos containing my face and/or my wife’s face. Each time face recovery runs, it draws a face, but not my face, and each time the face is different. Overall the photo is miles better looking than the original, but it’s not me you’re seeing in the final product. I have hundreds if not thousands of higher quality photos of our faces that could be used to establish a training model for Photo AI. Then I could have the option to tag a specific face as me or her and let the AI use that for the basis of reconstructing our faces. I hope this makes sense.