Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
Microsoft Input Devices Windows Technology

Microsoft Releases Kinect For Windows 130

nk497 writes "Microsoft has released Kinect for Windows, featuring a new "near mode" that lets the gesture control tech be used as close as 40cm. The Kinect for Windows hardware will retail at $249 — well above the price of the version for Xbox 360 consoles. Microsoft defended the price difference, saying sales of games and Xbox Live subscriptions help subsidize the console version. The new version will support Windows 7 and the Windows 8 developer preview, as well as Windows Embedded 7 devices."
This discussion has been archived. No new comments can be posted.

Microsoft Releases Kinect For Windows

Comments Filter:
  • WoR k z !!! (Score:5, Funny)

    by Anonymous Coward on Wednesday February 01, 2012 @03:19PM (#38895379)
    i JU-st P000000steD tHIz using mi NEw keNe C t!
  • by Deathnerd ( 1734374 ) on Wednesday February 01, 2012 @03:19PM (#38895381)
    Considering the tech they're supplying. If you bought similar tech for a hobby robot or something, I'd bet the price would be even higher. (I don't know if that's the case as I haven't done hobby robotics since the RCX Lego Programmable Brick. Ah, memories)
    • by Tsingi ( 870990 )

      It's a good deal.

      I don't have a windows os on anything at home, but I have a Kinect.

    • by Missing.Matter ( 1845576 ) on Wednesday February 01, 2012 @03:24PM (#38895431)
      The cheapest 2D LIDAR you're going to find is about $2000 (Hokuyo URG [acroname.com]). It has pretty terrible range for a LIDAR, but it's still a good sensor. For 3D ranging you're going to spend at least a grand. The IFM O3D 2XX is the cheapest 3D Flash LIDAR I know of, and you're only getting 64x48 pixels of resolution, so essentially 1% of the pixels you're getting back from the Kinect for 10x the price. Given this, the Kinect is truely an amazing sensor.
      • I'd love to see something comparable with a passive sensor. You'd need a pair of cameras and some serious algorithms though, wouldn't you?

        • by Missing.Matter ( 1845576 ) on Wednesday February 01, 2012 @03:34PM (#38895537)
          That's the main problem facing perception systems today. Humans have these two simple exteroceptive eyeballs and yet we can do incredible things. That's thanks to the amazing computational power of our brain, which we hardly understand. Thus, when we try to replicate our cognitive abilities we end up with algorithms that are completely intractable. I think this is in a large part due to computer scientists tendency to approach things with an engineering perspective instead of a biological perspective.
          • by tlhIngan ( 30335 ) <slashdot&worf,net> on Wednesday February 01, 2012 @05:23PM (#38896953)

            That's the main problem facing perception systems today. Humans have these two simple exteroceptive eyeballs and yet we can do incredible things. That's thanks to the amazing computational power of our brain, which we hardly understand. Thus, when we try to replicate our cognitive abilities we end up with algorithms that are completely intractable. I think this is in a large part due to computer scientists tendency to approach things with an engineering perspective instead of a biological perspective.

            Well, stuff like that isn't "wow".

            The problem is well, we treat computers as automation. We let computers do stuff we find hard or boring. The stuff we find easy, it turns out, is very hard to do on computers - natural language processing (face it - a lot of people went "so what?" when they saw Watson last year), vision processing (object recognition, character/word recognition), and hearing.

            It's stuff we don't think about - and it's boring to most people who can't comprehend how we can do stuff like read printed text, but the computer can't do a reliable job of it.

            It's probably one of the ironies in life. We have computers doing stuff easily that we find hard, and stuff we do easily computers find hard.

          • Thus, when we try to replicate our cognitive abilities we end up with algorithms that are completely intractable. I think this is in a large part due to computer scientists tendency to approach things with an engineering perspective instead of a biological perspective.

            It's because we try to replicate those cognitive abilities on systems that are fundamentally different. Computers aren't subjective, they don't make assumptions, they don't estimate, etc... and even if they do those things they do them in the exact way humans have told them to do it.

            • "Computers aren't subjective, they don't make assumptions, they don't estimate, etc..."

              I don't know why you say that? Any decent heuristic algorithm is all about building in a good bias (or prior), and solving simplified models of the real world, which is the essence of those supposedly human-only behaviors. Computers also do lots of things that are not foreseeable outcomes of the input (i.e. there's no way to get the answer other than to perform the full calculation). They also have entropy sources w

              • I don't know why you say that? Any decent heuristic algorithm is all about building in a good bias (or prior), and solving simplified models of the real world, which is the essence of those supposedly human-only behaviors.

                I don't think you've taken the comment in context, the fundamental difference is that we have to build algorithms to introduce 'human error' or cognitive bias artificially, a computer cannot reach a conclusion until you tell it how to reach that conclusion.

                Computers also do lots of things that are not foreseeable outcomes of the input (i.e. there's no way to get the answer other than to perform the full calculation).

                And they do them based on a defined set of rules.

        • by fuzzyfuzzyfungus ( 1223518 ) on Wednesday February 01, 2012 @03:35PM (#38895561) Journal
          That, and a rifle to use on all the assholes who keep hanging Escher prints in the cameras' field of view and laughing hysterically at your algorithm's attempt to cope...
      • Kinetic depth resolution is only 320x240.
    • Re: (Score:3, Informative)

      Also, from the article: "Although you will be able to download the SDK and use it with an existing Kinect for Xbox 360 for your own, personal development purposes <snip>"

      Sounds good enough for hobbyists.
  • Too late (Score:5, Funny)

    by jdastrup ( 1075795 ) on Wednesday February 01, 2012 @03:22PM (#38895417)
    Looks like Windows 8 touch interface is already outdated. Don't bother with the Kinect, mind control will be available soon.
    • But the question is, will mind control support Windows!!????
      • by Anonymous Coward

        But the question is, will mind control support Windows!!????

        Are you kidding? Who else do you think will be controlling your mind???

  • by getto man d ( 619850 ) on Wednesday February 01, 2012 @03:27PM (#38895477)
    I thought OpenNI's implementation was supposed to be pretty darn good (compared to say the freenect drivers) as far as offering more options / tools. Just looking at their APIs shows how much you can really do. What I don't understand is this 'Kinect for Windows': is the difference just updated code..?
    • Re:New Drivers? (Score:5, Informative)

      by Missing.Matter ( 1845576 ) on Wednesday February 01, 2012 @03:36PM (#38895565)
      According to Microsoft [msdn.com]:
      • Support for up to four Kinect sensors plugged into the same computer
      • Significantly improved skeletal tracking, including the ability for developers to control which user is being tracked by the sensor
      • Near Mode for the new Kinect for Windows hardware, which enables the depth camera to see objects as close as 40 centimeters in front of the device
      • Many API updates and enhancements in the managed and unmanaged runtimes
      • The latest Microsoft Speech components (V11) are now included as part of the SDK and runtime installer
      • Improved “far-talk” acoustic model that increases speech recognition accuracy
      • New and updated samples, such as Kinect Explorer, which enables developers to explore the full capabilities of the sensor and SDK, including audio beam and sound source angles, color modes, depth modes, skeletal tracking, and motor controls
      • A commercial-ready installer which can be included in an application’s set-up program, making it easy to install the Kinect for Windows runtime and driver components for end-user deployments.
      • Robustness improvements including driver stability, runtime fixes, and audio fixes
      • If the hardware is the same, and it sounds like it is, then it should be a simple matter to modify the driver inf to support the XBox Kinect vendor/hardware IDs.

        • Apparently the hardware is modestly different(supports closer-in operation, probably costs peanuts at scale, might be a pain to find the right IR-band lenses to mod one yourself); but you might run into some trouble in that MS has been pushing toward signed drivers only for a while now. XP and 32-bit versions of the later stuff can be told to ignore it; but 64-bit 7 does some serious whining if the signatures don't check out...
          • by Hadlock ( 143607 )

            You can buy a slip-over "wide angle lens" for about $20 on ebay that's designed to work specifically with the existing 360 Kinect. I don't think the existing lenses handle 16" but it might be able to handle 36 or 40". A 16" desktop model slip over lens isn't too far off though, I suspect.

          • I don't think the signing applies to the .inf file, just the executable code.
  • Nothing To Do Yet (Score:5, Informative)

    by Anonymous Coward on Wednesday February 01, 2012 @03:28PM (#38895487)

    "... Microsoft has already confirmed that the Kinect will ... not even run on Windows PCs that aren't also running the developer's kit and using the device otherwise may actually void the warranty."

    http://news.consumerreports.org/electronics/2012/01/ces-2012-kinect-for-windows-doesnt-mean-youll-be-playing-games-on-your-pc.html [consumerreports.org]

    So if you are not a developer, save your money.

    • Re: (Score:3, Informative)

      by GodInHell ( 258915 )
      Since the developer's kit is free . . . what's the issue? [read TFA]

      -GiH
    • by jader3rd ( 2222716 ) on Wednesday February 01, 2012 @04:14PM (#38896123)
      Prerequisites have existed for software installers for a few years now. I don't know how easy it is to add this SDK to an installer (haven't had to do it), but eventually (if not already) there'll be a template/plugin for Visual Studio and all a programmer will need to do when creating their installer, is check the box for the Kinect SDK prerequisite. Maybe they won't even have to do that, the current VS 2010 installer project will autodect a bunch of prerequisites for your project automatically and just do the right thing.
      • ...Which would be a copyright violation as Microsoft haven't (as far as I'm aware) provided a license to redistribute the SDK in such a manner.
        • If it behaves in the same way as the other prerequsites, the installer doesn't include a copy of the SDK. It contains some smarts about checking to see if it's installed on the computer, and if not knows where to go get it (ie the Microsoft download servers). So I don't think that would break copyright, given that it's the Microsoft servers making the copy. According to the Kinect SDK FAQ [microsoft.com] "The commercial license authorizes development and distribution of commercial applications." I'm no lawyer, but I read
          • If it behaves in the same way as the other prerequsites, the installer doesn't include a copy of the SDK. It contains some smarts about checking to see if it's installed on the computer, and if not knows where to go get it (ie the Microsoft download servers).

            If I were distributing software on a CD, I wouldn't want to require downloading additional components from the internet. It's an acceptable workaround for online distribution however.

            According to the Kinect SDK FAQ [microsoft.com] "The commercial license authorizes development and distribution of commercial applications." I'm no lawyer, but I read that as allowing for the deployment of applications to systems which aren't used for development.

            That merely describes that the application build using the SDK is distributable for commercial use, it says nothing for distributing Microsoft prerequisite components. However, the EULA for the SDK which I've just had a quick read through does state that some (not all) components are redistributable (specified in REDIST.TXT, of

    • by LocalH ( 28506 )

      Wow, why leave out the "at least initially"? Doesn't say that it never will, only that it won't work at first.

    • Re: (Score:3, Informative)

      by Xest ( 935314 )

      Yet from a post that listed Microsoft's official blurb a bit further up:

      "A commercial-ready installer which can be included in an applicationâ(TM)s set-up program, making it easy to install the Kinect for Windows runtime and driver components for end-user deployments."

      I think I'll trust the official release notes thanks.

  • Since I'm too lazy to RTFA, is the embedded device SDK meant to bring Kinect functionality to embedded devices/phones using your phone's built-in camera and mic? Or do you get to plug your Kinect into your phone and bring it with you on the bus?
  • User: Fuuuus-DOH-sah! Healing. (Raise left hand)

    Seems pretty straightforward to me.

  • (I just measured it in my office in plain sight of my colleagues, with a ruler sticking out of my head). My hands would be mere inches away from the Kinect. How's that going to work?

    For one thing, I'd have to lean back to bitchslap in Duke Nukem Forever, which frankly is a bit girly.

  • by twmcneil ( 942300 ) on Wednesday February 01, 2012 @05:08PM (#38896781)
    I bash MS quite often so in all fairness...

    IMHO, this is an awesome piece of kit at an incredibly affordable price. Good going Microsoft, you've done well.

    Damn, saying that has made my face twitch.
  • Does this mean I'll have to dance in front of my computer to upgrade to ubuntu 12.0?
  • Have you seen the design of a kinect? An up-down motor that runs once per power cycle and really only needs to be realigned when it is moved to a different position. Sensors 7cm apart that are housed in a 30cm frame of cheap tacky plastic, leading to a much bigger shipping box than necessary.. The kinect should be cheaper and smaller than it is now, what on earth is in all the empty space anyway?

    • by cbhacking ( 979169 ) <been_out_cruisin ... nospAM.yahoo.com> on Wednesday February 01, 2012 @07:12PM (#38897985) Homepage Journal

      The up/down is for people who are different hights, or for games which are better played sitting/kneeling vs. standing. I don't know, maybe everybody who uses your kinect are all the same hight and always use it from the same position, but for the rest of us that motor is pretty important.

      The spacing on the visual sensors doesn't require such a wide sensor bar, but the spacing on the microphones (for effective direction-sensing and noise-cancelation) does. People always focus on the optical portion of the sensor, and ignore the highly-focused microphones (possible because they're harder to see).

    • by mapuche ( 41699 )

      what on earth is in all the empty space anyway?

      Electronic circuits? the kinect is not just a couple of cheap webcams tied together.

  • Too busy to fully search, but what is the difference between Kinect on Windows and on 360. I know it says it has near mode, but is that only supported on Windows Kinects? Do I need to buy ANOTHER kinect (one for my 360, and one for my pc) or can I just use the Kinect I have for my X-Box 360 on my PC?
  • ...you'd think they'd want to help with getting some Open Source drivers developed for Kinect.

    I personally cannot see a use for Kinect but I recognise the algorithms that make it work are pretty cool and Microsoft is rightfully quite proud of it. Hardware hackers are definitely going to be interested in it but if the majority of those are Open Source freaks, I doubt many of them will end up paying for a Windows 7 license just to use it on a desktop PC.

    Common sense would have been for Microsoft to show off s

  • else the screen turning blue will be the last thing you see
  • Saw the following from a semi-famous developer in my twitter feed today:

    Microsoft Store in Santa Clara apparently not selling Kinect for Windows unless customer can "prove it will be used for commercial purposes"

    (later...)

    Wow. You actually have to BRING BUSINESS DOCUMENTATION to the MS Santa Clara store to get a Kinect for Windows. Also sign licenses.

The 11 is for people with the pride of a 10 and the pocketbook of an 8. -- R.B. Greenberg [referring to PDPs?]

Working...