AI

As photographers, we’re all familiar with needing to pull the most detail we can out of an image. Now, Google’s Brain team has come up with a way not just to extract detail, but to add it based on comparing the blurry version with a reference library of potentially-similar images. Since they’ve trained the system with a library of popular media stars, you might like the results even more than the original. Of course, it can’t really be called a photograph when it’s finished. More like an impressionist painting. You can read more about how it works in this article I wrote for ExtremeTech.