Kumail Nanjiani, actor, writer and cast member of HBO’s Silicon Valley, co-hosted the annual Adobe Sneaks, at Adobe MAX, last month in Las Vegas with Adobe Senior Creative Cloud Evangelist Paul Trani.
At the Sneaks event, held before an audience of 12,000 MAX attendees, Adobe showed off programs and technologies still in the development stage. As I sat waiting for the show to begin, I wondered, with so many innovations to Creative Cloud software this year, if Sneaks would be a let down this time. It was a good show and the technology was amazing.
When Trani introduced Nanjiani, he came on to the stage to the music of PAC-MAN. He admitted to traveling with video games and confessed a deep, dark secret. “When I was younger,” he said, “I thought Sonic was cooler than Mario.”
Trani didn’t kick him off stage for that; instead, he asked Nanjiani why he thought he was the perfect person to be at Adobe Sneaks?
Nanjiani told Trani that he went to school for computer programming and philosophy. “I used Adobe Dreamweaver for six years,” he said. Looking overly serious, he added, “Dreamweaver changed my life. I think I’ll do a one-man show about Dreamweaver now.”
Trani asked how Nanjiani moved from computer science to stand-up comedy.
“I wasn’t good at programming,” he admitted, “but I did like doing it. Unfortunately, I missed class one day and I never caught up. In Chicago, I started doing stand-up at night. Dreamweaver by day, comedy at night. A great scene.”
Having established Nanjiani’s tech credentials, they moved on to the sneaks.
Project Scribbler takes existing black and white photographs or sketches and, using Adobe’s facial recognition software within its artificial intelligence (AI) engine, Sensei, it colors in the black and white art.
The demo finished with a sketch of Nanjiani being colored in. His comment: “Wow, it turned me into a white guy. This is really great. It’s like a coloring program and you’re making children obsolete.”
Making Reality Better
The next two sneaks involved creating new images. Project SCENESTITCH is an enhancement to Content Aware Fill in Adobe Photoshop. The existing technology looks within a photo to copy something over an undesired element. Project SCENESTITCH goes further, allowing searches of similar photos to find content to replace the original.
Nanjiani asked whether this would also work on relationships.
Project PHYSICSPAK allows a user to choose one object and have it fill up another, rather like pouring marbles into a jar. In the example, the outline of a letter was filled with smaller geometric shapes.
Project SIDEWINDER takes creation AI into the VR world. A virtual reality landscape may not know, for instance, what the opposite side of a wall looks like. If a VR user changes a point of view to where there is no information, currently the move is prevented, or blank spots appear. With SIDEWINDER, the AI fills in the missing information on the fly.
See and Hear What isn’t There
If you have used computer based graphics programs, you have been exposed to color palettes, usually a square, containing a grid of different colors. The user clicks on one of the colors and that flows from the software pen or brush. In real life an artist’s palette works differently. Project PLAYFULPALLETTE fixes this. As in the real world, the user can take multiple colors, mix them together on the software palette and come up with an original color. Something PLAYFULPALLETTE does, that no real-world palette can, is letting the user replace that color throughout an artwork with a new color with one click.
Nanjiani shared that as a kid he wanted to be an artist. “The only thing that held me back,” he said, “was talent. But, I can draw a Ninja Turtle.”
Another VR enhancement under development is Project SONICSCAPE. Good sound is critical to any video. In virtual reality, filmmakers are challenged with operating in an environment in which the viewer can choose which direction to turn their head or move. To help manage this, Project SONICSCAPE allows a video editor to see a representation of sound on the video screen, so it can be dragged, dropped and associated to positions within the VR environment.
The Next Step in Puppets
For several years my favorite Adobe Sneak was Character Animator, which allowed you to animate a 2D cartoon by having it mimic your actions in front of a camera. Character Animator is now out in its 1.0 release.
Enter Project PUPPETRON. This app facilitates the creation of cartoon characters by taking a sketch or photograph and changing it to a stylized portrait, based on a sample. For instance, you could use a watercolor portrait as a sample, and PUPPETRON can change the photo to a water color with a similar color palette.
If you use a sketch, you can create a 2D rendering which can then be taken into Character Animator and be brought to life. During the demo, Nanjiani was merged with Abraham Lincoln and a lion.
Adobe doesn’t guarantee that these or several other sneaks they demoed will become commercial products. I do hope PUPPETRON makes it. My face plus Brad Pitt’s. It needs to be done.
For video demos of the sneak peaks check the Adobe Blog.