Following up on my previous column about the Electronic Drawing Board, I came across this impressive video of Jeff Han unveiling his multi-touch system to the TED conference in 2006.

There is something fascinating watching this presentation. As the applications increase in complexity throughout the demonstration, you can feel the excitement of witnessing the birth of a new era in human-computer interfaces.

In terms of digital arts, the most revealing application is the interactive lava lamp, with the animated line drawing puppets coming in close second. Imagine being able to mold virtual clay with both hands on your screen, pinch or smooth areas with two fingers, warm up an area to make it malleable and let it cool down to solidify it. If this is the future of Zbrush or Mudbox, I want it.

The TED conference is an annual gathering of great minds in the areas of Technology Education and Design. Think of a World Economic Forum for Ideas. Seating is very limited and the conference is booked years in advance but website is a real goldmine of thought provoking, exciting talks.

Another jaw dropping technology unveiled at TED this year was Photosynth, presented by Blaise Aguera y Arcas from Microsoft Labs. If  the multi-touch system evoked the user interface from Minority Report, this technology will bring to life the famous ‘photograph exploration’ scene from Blade Runner. It’s only appropriate since the venerable sci-fi masterpiece is celebrating its 25^th anniversary this year, without showing any sign of aging.

 

 

Photosynth compares the content of hundreds of pictures and puts them in correspondence, trying to calculate the points in space where the pictures were taken. Practically, this allows the system to build a 3D model of the reference points or to browse a large library of images according to their spatial context. You can try it for yourself on their online demo.

For the conspiration theorists, this is the ideal interface to the ultimate surveillance system. All you need is to put together images from surveillance cameras, Google street views, and shots from the Flickrs of the web. 

This is only a start for applications of combing through large libraries of images. Already, one of the papers presented at this year’s Siggraph enables to fill in obscured areas of a picture by looking from a database of similar scenes. The resulting image is not necessarily real, but the blanks are filled in a way that doesn’t degrade the original image.

 

 

Another breathtaking aspect of Photosynth is the fluidity of its visual explorer. Watch it as it zooms from a satellite view of thousands of images down to the level of a single character in a scanned book. This is one technology that can’t come to our desktops fast enough. With thousands of favorite pictures downloaded from the web, digital photos , scans, documents, etc… traditional visual explorers are reaching a breaking point.

What will it look like when high density images such as gigapxel photography or very large genealogy trees become more available?

Enough on this subject. SIGGRAPH 07 just started. Time to dive into amazing images and cool technology… 

Laurent Alquier (agiel) has been moderator of the Vue Forum at Renderosity since 2004. When he is not in the forum or the galleries, he finds himself busy with the balancing act of a day job as a Software Engineer and personal explorations of Information Visualization and Computer Graphics.