Hacking the Kinect Leads to Very Cool Things

Recently, Microsoft launched the Kinect for Xbox. A stereographic camera set capable of detecting and tracking movement in the room—something to help combat the extreme popularity of the Wii as a motion-sensing controller device. The Kinect has exploded onto the scene with a giant number of titles, but really where it’s shining today is how it’s caught the attention of the tinkering developer community.

Shortly after its release, hackers took to their workstations and started looking at how it works and what they could do with it. One intrepid tinkerer created drivers to capture Kinect feeds on Windows; and another shortly thereafter released a set of open-source drivers to do the same.

Now, still others have developed even more interesting applications for this highly versatile device. A report over at EDGE about the sort of things people have been doing with the Kinect. And they’ve got some amazing videos!

Kinect’s standard calibration routine offers a glimpse under the bonnet of the hardware, showing the skeleton it maps onto players in order to make sense of their movements. McDonald, along with Theo Watson and Joshua Noble, have explored an alternative in ‘blob detection’, a technique which can recognise distinctive elements such as hands in real time.

Memo Akten, meanwhile, has written a program which allows users to draw in 3D. Kinect automatically switches from one-hand drawing to manipulation once a second hand is raised, allowing the user to rotate the newly created image, and continue drawing just by dropping the second hand again.

And if Kinect can recognise hands, then why not a book, or even a giraffe? On his Youtube channel, Yankeyan demonstrates his software which allows you to teach Kinect to recognise objects. By holding up a variety of things, from a doll to the aforementioned giraffe, and then saying the name of that object, he shows how Kinect can amasses a database of associations that allow it to identify the same object the next time it is shown.

The stereoscopic effect of the Kinect lends it to a great deal of interesting applications. Especially identifying people, objects, and other things. It could be combined readily with Augmented Reality to produce entirely new types of games and interactions with the system—for example the make-any-surface an interactive virtual keyboard.

Microsoft has responded to people purchasing its new device and looking at how it works with extreme hostility—claiming that people doing so would be breaking the law. Today, it’s easy to see why the giant corporation is acting like such a bully over its technology: it hasn’t even exploited every facet it’s capable of in its initial release. Not quite the brightest move on Microsoft’s part in this case as people like McDonald, Watson, and Noble are just students and technological explorers who fill a role in the community for the curious. As a video game technology vendor, Microsoft has much more to fear from competing corporations who can throw hundreds of thousands of dollars at dissecting and reverse-engineering their device that snapping at these activities deals them only bad PR.

Perhaps we can use the exploration of those curious as a roadmap for what we can expect Microsoft to eventually do with the Kinect.

Who knows, we just might see a game from Microsoft that involves children bringing their stuffed animals into the room and then using them to play out their fantasy on screen.

This post was originally written by Kit Dotson for SiliconAngle and is reposted here with permission.

One thought on “Hacking the Kinect Leads to Very Cool Things”

Leave a Reply