In a former life, I did a lot of work with image and video compression—in fact, I spent a good part of grad school studying it. Not that I’ve really touched the stuff in the last few years… but I digress…
Somewhat related to that is my current budding Photoshop habit, which (along with that former life) keeps me attached to what’s going on nowadays in computer graphics. Sort of. Kind of. (Well, not really at all… but again, I digress…)
WARNING: Large, slow links ahead. Take necessary precautions (read: don’t use dial-up) before following them.
All of which makes this SIGGRAPH04 paper really cool. (SIGGRAPH is the annual computer graphics super-conference.) You know those old black-and-white movies that have been recolored (badly, in many cases)? Well, this paper describes a new algorithm for accomplishing the same thing, and the results are pretty jaw-dropping in my opinion. Basically, you take a B&W image or video, scribble in some colored lines, and the algorithm takes care of coloring the rest of the frame(s) using the colors you’ve prescribed—for example, taking a B&W image, scribbling in the colors you’d like, and ending up with a fully-colored image. I’d love to see a comparison of the results of this method against one of the commercially recolored B&W movies.
Granted, this has nothing at all to do with compression (it’s purely graphics manipulation), but still, really freakin’ cool. Or maybe I just miss my former life a little.