Using the correct keywords is a far more efficient system IMHO
But what does that have to do with anything? Apart from the fact that photographers have more than just photos of family and friends in their collections, automatic face recognition (like in Google Photos) could help describe photos – without the need to manually click through a list of keywords, i.e. automatically fill in the “persons shown” field. I photograph many music, business, social, and sports events, as well as private parties… Finding all the photos of John Doe in a few moments is probably the best role for a photo database! Having 10 people in one photo, 20 in the next, 3 in another, then 5 or 7 again – you have to click through keywords for EVERY PHOTO that many times. A face recognition machine, after a few parties and assigning names to found faces – could do it automatically in every subsequent photo in which it finds that face! Apart from the time-consuming manual tagging, a person will not always notice the one person in a group of people that they know. Often, he also forgets the name and wastes time trying to remember it again (or most likely, he will simply skip it, not wanting to waste time). The machine would do it without human intervention. If we also added a database of all recognized faces, but not described – we could describe them at our leisure, and the program would automatically fill in the appropriate field.
Honestly… As for me, at the moment, in the matter of cataloging photos, such a possibility simply imposes itself, as a natural way of developing this program.
