Web Concepts

Posts Tagged ‘augmented reality’

Unlike existing applications that overlay virtual features onto real-world video, this program doesn't add any synthetic elements to the video feed. It creates a 3D model of the user's face, tracks their features, and then subtly warps the video. The user can then see how they would look with a smaller nose, wider mouth, or Powerpuff Girl eyes.

The software could be a valuable visualization tool for plastic surgeons. The creators are also considering adding in texture-based modifications that could allow the user to apply virtual makeup to their face. Check out the mirror at work in the video below:

[New Scientist]

Swedish startup 13th Lab is implementing a computer vision technique called Simultaneous Localization and Mapping (SLAM), which constructs a 3-D map of a local environment in real time and calculates the current position within it. The result is a new iPad app called Ball Invasion (get it here), wherein the camera’s view becomes a playing field. Instead of advanced robotic sensors and controls, the app just needs the camera and other sensors native to mobile devices.

This is quite a feat, and a potential new avenue for augmented reality. Unlike most AR systems, it doesn’t require previously known markers to trigger the virtual display — it can augment any environment, previously seen or unseen, simply by using the iPad2’s camera.

The goal of the game is to shoot malicious balls hiding in the real world, which becomes part of the playing field — you can bounce virtual items off the actual walls, for instance. The video below shows it in action.

SLAM was developed to help robots determine where they are, by looking around and building a 3-D model of their environment and then determining their place in it. It’s a tough chicken-or-egg problem and one of the most complicated topics in robotics sensing, but, as GigaOM points out, 13th Lab has figured out how to compress this complex capability into a consumer device. So far, it’s only possible with the iPad2’s powerful dual-core A5 CPU, 13th Lab says (though they probably haven't tried using the next-gen quad-core Kal-El yet).

13th Lab’s overall goal is far broader than 3-D games: they want to build a 3-D toolkit for other app developers, according to GigaOM. Ball Invasion is simply the first example.

This type of technology could conceivably be used for many other things, from architecture design to augmented-reality tours. For now, though, this game seems pretty fun:

[via Slashdot]

Get ready for the first complete synthetic human brain, moon mining, and much more

Click to launch the photo gallery

2020, of course, is just a convenient target date for roughly-ten-years-off predictions. "It's not any more particularly interesting, in my opinion, than 2019 or 2021," says Mike Liebhold, a Distinguished Fellow at the Institute for the Future, and an all-around technology expert with a resume that includes stints with Intel, Apple, and even Netscape. "There's a continuum of technological development, and that's just an easy date for an editor or a writer to get a handle on.

After spending decades helping various top-tier tech companies develop and deploy their cutting edge technologies around the world, Liebhold now helps clients take a long view of their businesses so they can make better decisions in the short term. He and his colleagues at the Institute for the Future don't help clients read tea leaves (predictions are for soothsayers and crystal ball gazers) but they do help them read what he calls the signals -- those things you can see in the world today that allow you to make reasonable forecasts about what the future holds.

"We help people think systematically about the future," Liebhold says. "We don't give them answers, we give them foresight."

In other words, the year 2020 (and 2019, and 2021) is Liebhold's business. And he forecasts a pretty interesting world a decade from now. For instance, given the current forward momentum of mobile technology and the ever-present forces of economies of scale, Liebhold says it's conceivable that most of the world's population will be able to afford a Web-enabled smartphone or tablet device by 2020, offering everyone on the planet geo-location services and access to global information and communication (the forces working against this, he notes, are political rather than technological).

Facial recognition and other biometrics will be commonplace, he says. High-performance data visualizations that currently require supercomputing power will become commonplace as well, driving technological and scientific innovation at even faster rates. We'll see wider distribution of things like AI and immersive media experiences like viewpoint-independent 3-D. We'll finally have some decent augmented reality glasses.

And what won't happen? We won't be uploading the human mind to a machine by 2020, a la Ray Kurzweil. We won't be cruising the streets in self-driving vehicles, and while robots may be rolling around on the moon, we won't be mining minerals from extraterrestrial sources.

So what will the world look like in 2020? With Liebhold riding shotgun, we took a quick spin through 2020 to see what the future might hold. Click through the gallery to see some of the bolder 2020 forecasts we've seen--and why some of them don't stand a chance.

ShelvAR consists of an Android app and a set of coded tags, representing call numbers, that are placed on books' spines. When a librarian holds a smartphone or tablet camera up to a shelf, the app reads all the tags at once, thanks to a new algorithm that can decipher multiple patterns even though they're small when viewed at a distance. Then the app uses a simple sorting method—at least for computers, which aren't fazed by complex letter-digit combos like Q164 .G72 2009--to figure out the correct order and the shortest number of moves needed to achieve it. The phone's screen displays red X's over any misfiled books, along with arrows that show where they really belong.

The prototype app, built by computer science professor Bo Brinkman and research assistant Matt Hodges, has successfully analyzed a dozen books with half-inch tags. The team is now working on scaling up to 75 to 150 quarter-inch-thin books, so that they can scan a full shelf in one shot, and in December, they'll test the app in part of the university library. Adding ShelvAR tags could save libraries time and money in the long run, since workers now do frequent shelf checks by hand.

If all goes well, a beta version of ShelvAR will be released next spring. Librarians are already envisioning other uses for the technology, Brinkmann tells us, such as displaying a star rating over recommended books or helping lost students find the book they're looking for.

ShelvAR consists of an Android app and a set of coded tags, representing call numbers, that are placed on books' spines. When a librarian holds a smartphone or tablet camera up to a shelf, the app reads all the tags at once, thanks to a new algorithm that can decipher multiple patterns even though they're small when viewed at a distance. Then the app uses a simple sorting method—at least for computers, which aren't fazed by complex letter-digit combos like Q164 .G72 2009--to figure out the correct order and the shortest number of moves needed to achieve it. The phone's screen displays red X's over any misfiled books, along with arrows that show where they really belong.

The prototype app, built by computer science professor Bo Brinkman and research assistant Matt Hodges, has successfully analyzed a dozen books with half-inch tags. The team is now working on scaling up to 75 to 150 quarter-inch-thin books, so that they can scan a full shelf in one shot, and in December, they'll test the app in part of the university library. Adding ShelvAR tags could save libraries time and money in the long run, since workers now do frequent shelf checks by hand.

If all goes well, a beta version of ShelvAR will be released next spring. Librarians are already envisioning other uses for the technology, Brinkmann tells us, such as displaying a star rating over recommended books or helping lost students find the book they're looking for.

Forrester Research just released a new report on mobile augmented reality (AR) that says few companies are delivering “real” augmented reality (AR) today.

Augmented reality overlays digital information over your view of the physical world in a mobile device’s camera view, as pictured left. It currently works by using a combination of the mobile phone’s camera, compass, GPS data and other sensors to identify objects in the field of view, retrieve relevant data and overlay that data over the camera view. For example, you could point your camera at a building to get information about apartments for sale or the history of the building.

The Forrester report says that AR requires object recognition on the mobile device itself as well as 3D rendering to superimpose images on the real world and that only a few companies like Metaio and Total Immersion provide this. Visual search like Google Goggles and GPS browser-based experiences such as that provided by Layar don’t meet that definition of AR.

Forrester also concludes that AR is not yet scalable or mature. VentureBeat highlighted some of the problems (and possible solutions) with mobile AR technology at the start of 2010. Thus far, few consumers are using the technology. Layar is ahead of the pack with 1 million active users, while Metaio’s Junaio mobile AR solution has been downloaded 500,000 times.

Mobile AR will merge with other technologies like location-based services, visual search, image and facial recognition and barcodes, predicts Forrester. In fact this already started to happen in 2010. In February we saw an application called Recognizr which combined facial recognition with AR to recognize people and pull up profile information. Both companies involved in Recognizr were acquired in 2010, facial recognition company Polar Rose by Apple and interface specialists TAT (The Astonishing Tribe) by Blackberry-makers RIM. Metaio combined visual object search with AR so you can point your camera at everyday objects and get information about them. More recently, we saw the wonderfully useful WordLens which translates Spanish text to English (and vice versa) in the camera view in real time.

In spite of these provisos, Forrester is optimistic on the future of mobile AR and expects more blue chip companies to join the AR fray in 2010. Currently the space is dominated by startups. The report points out that Qualcomm opened an AR research center. And eBay classifieds recently partnered with Metaio.

Funding for mobile AR companies in 2010 went mainly to Layar (which did two rounds, 3.4 million in February and 14 million in December) and gaming companies like Tonchidot which received 12 million.

Tags: augmented reality, visual search

Companies: Forrester, layar, Metaio, Qualcomm, Total Immersion

Just when you thought you’ve seen everything mobile apps have to offer, along comes an entry like Word Lens that makes you feel like you’re in the future. The app instantly translates Spanish into English (and vice versa) whenever you point your iPhone’s camera on text.

Translation apps aren’t anything new, but what’s amazing about Word Lens is that it translates text in real-time — there’s no need to take a photo and wait for the translation to process. Instead, the app offers an augmented reality (AR) experience by laying its translated text over a video feed from your iPhone’s camera.

This means that you can point your phone at a menu or sign in Spanish, and the app will instantly translate it into English text on your iPhone’s screen. The app also doesn’t require an internet connection to work, so you can use it even when you don’t have reception.

Word Lens is developed by Quest Visual, and the app itself is free. Inside the app, you pay $4.99 for translation dictionaries (only Spanish to English, and English to Spanish are available right now). The app doesn’t come bundled with any dictionaries, so it will effectively cost you $4.99 to start using it. With its in-app purchasing model, Word Lens could easily add support for other translation dictionaries.

Judging from the amount of buzz Word Lens is generating online this morning, I think Quest Visual has a hit on its hands. It’s also notable as a genuinely useful implementation of augmented reality, even more so than Yelp’s Monocle feature. Instead of being flooded with AR data, as many AR apps tend to do, Word Lens simply offers another way to view what’s directly in front of you.

You can view a demonstration of the app below to see just how innovative it is.

Tags: apps, augmented reality, iPhone, translation, Word Lens

Companies: Quest Visual, Yelp