What’s truly amazing is that the AI doesn’t just tweak brightness and contrast—it actually recognizes what’s in the shot and changes the composition accordingly.
Look at the photo at the top of the post. Notice how it’s tilted after being boosted? That’s a technique used to emphasize a sense of scale—by giving the impression of looking down, it makes the subject appear enormous.
In other words, the AI recognizes the subject as a Gundam and, beyond just adjusting lighting, it re-composes the image in a way that suits a colossal humanoid robot. Isn’t that incredible?
This demonstrates that the AI has not only learned to identify “Gundam,” but also to apply compositional and cinematic techniques that make a gigantic robot look appropriately dramatic.
I, on the other hand, had only used AI for mundane tasks—like scanning receipts.
To be honest, Google Photos’ Boost feature still has its flaws—it often produces gaudy, over-the-top photos. But if it can teach us about compositional techniques through these transformations, wouldn’t that alone make it worth using?