A few days ago the image above started going around the social networks, attributed to “a friend working on AI”. Apparently a deliberate leak, now we know where it came from: a research group at Google working on neural networks for image recognition. Their research is worth reading about, and the images are fascinating: Inceptionism.
I have to admit, machine learning makes me jealous. Why does the machine get to do the learning? Why can’t I do the learning? But projects like this make me feel a little better. When the black box opens and we see the writhing mass inside, we get to learn how the machine does the learning. Everyone wins.
And the machines still have some catching up to do. As soon as I saw the amazing gallery of inceptionist images, I recognized the “inceptionist style” from the mysterious grey squirrel. Could a neural network do that?
Here’s something different: I have a new job! Today was my first day at Google Spotlight Stories. I’ll be working with some amazing filmmakers and technologists who are busy inventing a new kind of narrative visual storytelling uniquely suited to handheld mobile devices. If that sounds crazy, that’s because it is. It’s my kind of crazy. It’s exactly the kind of wild, inventive, “let’s try this and see what happens” attitude that got me interested in computer graphics in the first place, all those years ago. I couldn’t be more excited.
The video above really does a great job of capturing the delight of experiencing one of these stories for the first time. It’s almost impossible not to grin like a ninny. There’s not much more I can say about it right now, but there’s been some terrificpress about the projects they’ve created so far. I’ll share more when I can!
Inspired by this brilliant interactive demo of the perceptually uniform CIE L*a*b* color space, I decided to try a L*a*b* version of my pseudocolor scheme. I don’t find this version as pretty to look at, but it has the advantage that higher values are always mapped to colors that are perceptually brighter than lower values. In other words, if you squint at the image above, the bright and dark regions correspond pretty much exactly to what you’d see if it were greyscale. (For the L*a*b* to RGB conversion, I grabbed pseudocode from this handy page.)
I was tinkering with some GLSL shaders in Processing, and I needed a way to visualize a value that smoothly changes from 0 to 1, showing a lot more than the 256 levels of gray that you’d normally see. So I wrote a little pseudocolor function that spirals through colorspace from black to white, hitting various hues along the way. It’s fun, and pretty, and very rungy-chungy, so I thought I’d post it here.
In the 1990’s, it was typical for production companies to start their demo reels with a clock-wipe countdown. Some companies would use this as an opportunity to say something about their style, and do custom animation. (Will Vinton Studios, I remember, had a particularly cool stop-motion countdown at that time.) I thought it would be cool for Xaos to have a countdown that reflected our particular style of work, so I created this drippy ink effect counting down from 10 to 1, and a drippy version of the Xaos logo, which we put at the beginning and end of our demo reel respectively.
Some time after this, our producer got a call from MTV. They had this show called “Top 20 Video Countdown”. They wanted to use this effect for the bumpers and interstitials. Could we add the numbers 11 through 20? Our producer said “of course”, and charged them a lot of money. But it didn’t really take me that long.
Here is my first short film. I made this at PDI in 1995, during a gap between commercials. I modeled and rigged the characters, did most of the animation, and developed the wobbly ink-line look.
The pigeons’ torso was a metaball surface driven by a series of spheres along a spline between the head and the body, which were both separate IK joints (so I could easily get that pigeon-head movement style without counteranimating.) The eyes, beak, legs and wings were separate objects, each of which got rendered in its own separate pass. Each layer had its vector outline traced (using a tool originally written for scanning corporate photostats for flying logos!) I processed the curves using a procedural scripting language to give them some physics and personality, and then rendered them as black ink lines with varying thickness (using a tool written by Drew Olbrich). Finally, I ran the rendered lines through some image processing filters to get the edge darkening effect, and did some iterated stochastic silhouette dilation to add random ink blotches where the lines were thickest. Simple, really! ;-)
Cleary Connolly, an artist in Ireland, has built a lovely series of perception-altering helmets, including the “Hammerhead” (above), which is effectively a head-mounted, lens-based version of the Telestereoscope. The helmets look beautifully crafted and durable. The site has some great photos of the work in progress, and don’t miss the drawings of what they’ve got planned for the next helmets in the series. (I especially love the Siamese Helmet concept– great fun!)