You can use Apple's Vision framework to - among other things - detect faces in photos and get back locations. One usage could be let users easily add emoji or meme faces over faces found in photographs without the need to manually drag them in place and scale appropriately.
Not sure if this would in practice save a lot of time.