Google AI can rate photos based on aesthetic appeal
Photo apps can help you find objects in your pictures, but they don’t tell you whether or not those images are worth sharing. For now, that’s still up to you. If Google has its way, though, AI may serve as an art critic. It just detailed work on a Neural Image Assessment (NIMA) system that uses a deep convolutional neural network to rate photos based on what it believes you’d like, both technically and aesthetically. It trains on a set of images based on a histogram of ratings (such as from photo contests) that give a sense of the overall quality of a picture in different areas, not just a mean score or a simple high/low rating.
The approach can use reference photos if they’re available, but it can turn to statistical models if there’s no ideal target photo.
The result is a neural network system that “closely” replicates the mean scores of humans when judging photos. That, in turn, has all kinds of implications for photography apps. To begin with, it could help you quickly find your best photos while avoiding blurry or poorly composed shots. Google adds that it’d be helpful for editing, too, as you could use it to tune automatic editing tools. Your favorite editing app could tweak exposure, brightness and other details based on artistic appeal rather than arbitrary values. While there’s a lot of work to be done, this hints at a day when your phone could have as discerning a taste in photos as you do.
(37)
