objectrecognition
Latest
Google lands patent for automatic object recognition in videos, leaves no stone untagged
Google has already been working on patents that could pick out faces and song melodies in our YouTube clips. Now, it might just have the ultimate tool: the technique in a just-granted patent could pick out objects in a video, whether they're living or not. Instead of asking the creator to label objects every time, Google proposes using a database of "feature vectors" such as color, movement, shape and texture to automatically identify subjects in the frame through their common traits -- a cat's ears and fast movement would separate it from the ball of yarn it's attacking, for example. Movie makers themselves could provide a lot of the underlying material just by naming and tagging enough of their clips, with the more accurate labels helping to separate the wheat from the chaff if an automated visual ranking system falls short. The one mystery is what Google plans to do with its newfound observational skills, if anything, although the most logical step would be to fill in YouTube keywords without any user intervention -- a potential time-saver when we're uploading that twelfth consecutive pet video.
Sony makes a patent move for Kinect style controller
We're not suggesting Sony has a Kinect style controller coming soon, but it certainly filed a patent for one. The application for a "User-Driven Three-Dimensional Interactive Gaming Environment" was initially logged in October last year by Richard Marks of PlayStation Eye fame. The controller can "determine when to interact with the system by allowing part of the user's body, or an object, to cross the maximum depth range plane" meaning input could be limited to just the hands etc. This wouldn't be the first time Sony has created controller technology similar to its competitors, but the company is careful not to outright state that's what it's for. That said, amongst the legalese it does go as far too say the system can be implemented by an entertainment system "such as" its flagship console. So, sadly there's no way of knowing if we'll be getting a PlayStation Motion any time soon, but feel free to wave over the source link to see the patent for yourself.
How I Met Your Robot Mother: a Qbo 'First' (video)
Is consciousness programmable? Do robots feel? Would Kubrick have approved of Spielberg's handling of AI? While you wrap your mind around those conundrums, set aside a bit of free grey matter to soak in another Qbo milestone. When last we left the little bot, it was just coming to terms with its place in the grand scheme of things and recognizing its reflection. Since, robotic hangers-on and the geeks that love'em have been pelting The Corpora (the brains behind the operation) with questions pertaining to the self-awareness of this artificially intelligent machine. The outfit's reply? A Qbo meet-cute with a "female" member of the automated species that highlights the droid's understanding of individuality. It's a far cry from a Lady and the Tramp-style nose nuzzle, but these bots don't have to touch -- a random series of nasal flashes serve to indicate their distinctiveness and, once that's done, polite flirtation ensues. Don't believe us? You can see the sparks of android love for yourself in the video after the break.
Google+ and Goggles apps for Android updated with new features
It's time to stop ignoring that "updates available" notification and start enjoying some refreshed Google goodies on Android. You'll now be able to search posts and people from within the Google+ app, as well as add your +1 to photos and comments, see the 'online' and 'typing' status of your Messenger buddies, and upload videos and full-res photos. Meanwhile, Goggles now has a 'continuous mode' that can log multiple items without you having to fiddle with the shutter button, the ability to link straight to any online versions of hard-copy text, and also better crowdsourcing so users can add less well-known objects like toothpaste squeezers and Off the Hook shower heads to Google's ever-expanding database.
Qbo robot looks in a mirror, learns who he is (video)
A robot equipped with stereoscopic vision and some object and facial recognition programs opens up a whole number of different possibilities, but few quite as interesting as what happens when you place said robot in front of a mirror. It took a bit of guidance (and no doubt a bit of prep), but the open source Qbo robot was quickly able to both recognize and learn to identify himself when he was recently placed in just such a situation. See what happened after the break. Update: The Corpora, the company behind Qbo, has now added a more detailed explanation about exactly what is happening here on its blog.
Google acquires PittPatt, wants to know you on a face-to-face basis
Google's quietly pitter-pattering its acquisitive ways back into the controversial realm of facial recognition technology. To do that, the company busted out its oversized wallet to fold Pittsburgh-based PittPatt into the Mountain View borg. Founded by a trio of PhD's from Carnegie Mellon University, this three-man strong outfit specializes in the sort of object recognition software you've come to know as "tagging." Is this a reversal of the Do No Evil tech giant's prior waffling on the dubious visioning tech, or just another massive weapon in its social networking crusade against Facebook? We'd err on the side of both, although the company's new employees aren't exactly playing their cards for us to see. A brief statement on the triumvirate's site makes vague mention of "computer vision technology" being core to Google's products and points to the tech's planned integration in photo, video and mobile applications. So, basically, expect to see Picasa, Goggles, YouTube and Google+ watch you as you flaunt your internet celebrity ways to that front-facing camera.
Microsoft licenses GeoVector's augmented reality search for local guidance (video)
After the ho-hum AR demonstration of Windows Phone Mango, Microsoft appears to be stepping up its game by licensing a mature set of technologies from GeoVector, (a company previously known for its defunct World Surfer application). While the details remain elusive, Ballmer's crew was granted a multi-year, non-exclusive right to use and abuse the pointing-based local search and augmented reality elements of GeoVector's portfolio -- surely capable of bringing Local Scout to the next level. While much of the technology relies on GPS and a compass for directional-based discovery, the licensor also holds intellectual property for object recognition (à la Google Goggles), although it's unclear whether this element falls within the agreement. Of course, Microsoft could have turned to Nokia's Live View AR for many of the same tools, but that would have been far too obvious. Just beyond the break, you'll find the full PR along with an (admittedly dated) video of GeoVector's technology.
Sony's SmartAR demoed live, raises the bar for augmented reality (video)
Remember Sony's SmartAR? The markerless AR technology that promises reality augmentation without the need for unsightly tattoos? It's back again, showing itself once more after an all-too-brief 48 hour layoff. A new live-demo shows Sony's markerless object recognition system focusing on posters, tables, books, and coffee cups in lieu of the traditional AR card -- allowing it recognize multiple objects at once. Focusing on objects rather than markers allow augmented entities to interact more naturally with their environment. For instance, bouncing AR balls plummet off the edge of a table, and realistically ricochet off of a book placed in their path. Objects don't even need to remain on screen, as demonstrated by an AR pop-up menu that remained viewable even after the object-marker that spawned it left the viewer's field of vision. Sony seems to have built the groundwork of an augmented reality system that might actually be useful -- pair this up with a set of swank AR glasses (or better yet, holographic AR glasses), and we'll have a vision of the future we can really look forward to.
Aurasma's AR iPhone app to turn everyday objects into multimedia triggers (video)
We've seen augmented reality done what seems like a million different ways, but we've never seen it quite like this. The New York Times reported Wednesday on a forthcoming iPhone app called Aurasma that has the power to turn ink-and-paper publications into interactive mine fields. Aurasma, conceived by enterprise software firm Autonomy, uses a scaled down version of the outfit's IDOL pattern recognizer to identify images stored in a vast database, and then converts those images into related video. Unfortunately, the first release of the app, scheduled for sometime next month, comes in the form of an AR advertisement / game for an unidentified upcoming movie, and the company's founder seems solidly focused on the technology's marketing potential. No word yet on when or if we can expect to see our New York Times come to life, as seen in the video at the source link below, but if this is the future of augmented reality, count us in.
Harvard and MIT researchers working to simulate the visual cortex to give computers true sight
It sounds like a daunting task, but some researchers at Harvard and MIT have banded together to basically "reverse engineer" the human brain's ability to process visual data into usable information. However, instead of testing one processing model at a time, they're using a screening technique borrowed from molecular biology to test a range of thousands of models up against particular object recognition tasks. To get the computational juice to accomplish this feat, they've been relying heavily on GPUs, saying the off-the-shelf parallel computing setup they've got gives them hundred-fold speed improvements over conventional methods. So far they claim their results are besting "state-of-the-art computer vision systems" (which, if iPhoto's skills are any indication, wouldn't take much), and they hope to not only improve tasks such as face recognition, object recognition and gesture tracking, but also to apply their knowledge back into a better understanding of the brain's mysterious machinations. A delicious cycle! There's a video overview of their approach after the break. [Thanks, David]
Sony's PlayStation Eye to gain facial recognition capabilities
It was almost two weeks ago that we first heard about the patent for Sony's PlayStation Eye-powered object motion detection, and now the company's announcing facial recognition for the console. According to Gamasutra, Kish Hirani, SCEE's Head of Developer Services, said that the facial recognition software would "detect gender and even the age of the face, separate facial features such as the nose, eyes and ears, and even detect whether you're smiling or not." This new gear will also work with the Sony Motion Controller, with the camera tracking X, Y and Z motion by using the glowing ball. Speaking at the Develop Conference in Brighton, Hirani stressed that companies won't need to develop their own tech to work with the new controllers. "If you are working with the PlayStation Eye and think there is some new tech you're going to have to develop for the motion controllers, just get in touch with us. We have a wealth of libraries available, and the chances are you won't have to develop any technology yourself." You hear that, guys? get to work![Via SlashGear]
Taito introduces Surface-like arcade game
Well, it didn't take long for Microsoft's Surface to trickle down into the fast-paced world of card gaming (ala Magic). Feast your eyes on the next device your children will be shoveling money into like it's going out of style, a new Square Enix (makers of Final Fantasy) and Taito collaboration called The Wheel of Eternity. The card-based RPG appears to function much in the same way that Bill Gates described Surface's object recognition properties as working, "sensing" cards which are placed on the game area, and reacting to their status, movement and arrangement. There's no word on when or if this game is coming to the states, but with kids' unnatural obsession with card games (particularly those of a Japanese ilk), it seems likely this will be popping up on these shores before long.
