It's been three months since my last entry. I wonder what that means. I suspect it means that I've been expressing myself enough in private conversations that I haven't felt compelled to write anything here.
Finally, something worth shouting about. NYU's Image Analogies project is extremely cool. Given three images, A, B, and C, you tell it that B is some sort of transform of A, and it will tell you what you'd get if you applied the same transformation to C. What makes it so cool is the complexity of the transformations their software is able to handle in a reasonable way. Be sure to check out the example applications, especially "Texture-by-numbers".
It's not entirely automated; there are a few parameters to tweak. But the software is available (license not specified) so we can all take it for a spin.