The team shot outdoor scenes in Kiev, Ukraine, before recreating the entire town on a set inside the country’s largest airplane hangar. The “ground,” however, was built six feet off the floor, to allow space for trampolines built into the sidewalks. […]
For a scene where he falls sideways beside a woman on a bench, two practical shots were merged into a single one. The actor bounces off a specially-crafted surface, and the camera was turned 90 degrees to film the woman, who was strapped into a bench built into a wall. The entire production was shot in just 12 days, a feat that required 200 artists and technicians.
Jeff’s mane is a little thin on top, and Gregg is more folliclularly challenged. So, when Jeff returned from vacation to Taiwan, he was rather unhappy to find that Quick Selection was selecting only his head, missing the wispy bits of hair on top. As he proclaimed while making a quick whiteboard self portrait, “I need to keep all the hair I’ve got!”
Smash-cut forward 13 years (cripes…), and researchers are developing a way to use multiple cameras to capture one’s hair, then reconstruct it in 3D (!). Check it out:
Here’s an example of what happens when a team leverages deep learning to make light fields practical. It’s gonna be really fun to try enabling these capture & display capabilities to mass scale.
More than 10,000 artworks from 208 partners worldwide have been captured with Art Camera and digitized in ultra-high resolution, from the fluffy fabric from which Vivienne Westwood tailored the Keith Haring “Witches” dress, to the almost photographic View of Delft by Vermeer. You can see these works in intricate detail simply by browsing on the Google Arts & Culture app. Explore Art Zoom online at g.co/ArtZoom, or download our free app for iOS or Android.
It’s crazy to think that this stuff works in realtime on a telephone, when just 7 years ago here’s how Content-Aware Fill looked when applied to video:
Do I know what the hell is going on here? No, of course not! (Have you ever met me? 😌) But thankfully my colleagues Noah, Richard, and co. do, and it promises a way to capture & display rich, dimensional photos (see interactive example that lets you play with parallax & see depth; more are on the site). Check it out:
The glue my team developed to connect & coordinate machine learning, computer vision, and other processes is now available for developers:
The main use case for MediaPipe is rapid prototyping of applied machine learning pipelines with inference models and other reusable components. MediaPipe also facilitates the deployment of machine learning technology into demos and applications on a wide variety of different hardware platforms (e.g., Android, iOS, workstations).
If you’ve tried any of the Google AR examples I’ve posted in the last year+ (Playground, Motion Stills, YouTube Stories or ads, etc.), you’ve already used MediaPipe, and now you can use it to remove some drudgery when creating your own apps.
People have been trying to combine the power of vector & raster drawing/editing for decades. (Anybody else remember Creature House Expression, published by Fractal & then acquired by Microsoft? Congrats on also being old! 🙃) It’s a tough line to walk, and the forthcoming Adobe Fresco app is far from Adobe’s first bite at the apple (I remember you, Fireworks).
Back in 2010, I transitioned off of Photoshop proper & laid out a plan by which different mobile apps/modules (painting, drawing, photo library) would come together to populate a share, object-centric canvas. Rather than build the monolithic (and now forgotten) Photoshop Touch that we eventually shipped, I’d advocated for letting Adobe Ideas form the drawing module, Lightroom Mobile form the library, and a new Photoshop-derived painting/bitmap editor form the imaging module. We could do the whole thing on a new imaging stack optimized around mobile GPUs.
Obviously that went about as well as conceptually related 90’s-era attempts at OpenDoc et al.—not because it’s hard to combine disparate code modules (though it is!), but because it’s really hard to herd cats across teams, and I am not Steve Fucking Jobs.
Sadly, I’ve learned, org charts do matter, insofar as they represent alignment of incentives & rewards—or lack thereof. “If you want to walk fast, walk alone; if you want to walk far, walk together.” And everyone prefers “innovate” vs. “integrate,” and then for bonus points they can stay busy for years paying down the resulting technical debt. “…Profit!”
But who knows—maybe this time crossing the streams will work. Or, see you again in 5-10 years the next time I write this post. 😌
