Microsoft unveils even more tempting Kinect offering: Open source First it was developer tools, then Kinect for the PC, now Microsoft's given hackers a shot at the Kinect code under an open-source license. Specifically, Redmond has now released samples of the Kinect code under an Apache license to serve as a template for hackers building apps for the hands-free motion controller that's been …
Bit of both. You can't just slap a wide angle adapter on it and carry on using it, because lots of bits of the software make assumptions about the camera's field of view and lens distortion characteristics.
The bits of software that do make these assumptions are either a) quite low level or b) of sufficient value to MS that they haven't (and indeed probably won't) release them with an open source license. The harder it is for you to run, say, the skeleton tracking system on another OS the better, as far as they're concerned. Face tracking is far less interesting by comparison, and much easier to cobble together with existing open source stuff like OpenCV et al, so giving it away doesn't actually cost them anything.
I think it depends upon the version - there is an XBOX version, and a PC version. There isn't much difference in hardware, but I believe the PC version is more accurate at nearer distances. The PC version is also more expensive but works with the official SDK (there is a 3rd party open source alternative) whereas last I looked the XBOX version didn't. I don't know if this announcement changes that.
Which it transpired actually meant.
"Guinness World Record for fastest shipping consumer electronics device"
As for the following 8 months, Microsoft's sales of Kinect fell off a cliff because of their channel stuffing they they did to achieve this bullshit headline..
Or... The following 8 months the sales fell off a cliff because everyone who wanted/could afford one had one already. It's a much more simple and plausible answer than "big faceless megacorp evilly attempted to make it look like their fabulously popular product was more popular than it actually was."
Why does it always have to be a conspiracy?
Will you be accusing them of burning Kinnects next, so that they can stuff the channel with more of them, as there as been space filled up?
How could Microsoft 'Stuff the channel' when they don't control it? Microsoft can only ship goods to fill orders.
You seem to live in some alternative reality where you think Microsoft can just dump a container load at the local Shopping Mall...
Every second hand shop is full of them, as as a gaming device, it's failed spectacularly, and massively flopped despite the initial hype of it being a game changer.
(Still it achieved Microsoft's goal, they fooled plenty of press and gamers into believing it was amazing).
Second-hand hardware and open source code? Can only be a good thing. Look at what happened with the CueCats etc. - they went mad once people weren't using them and code appeared for them.
I would quite happily buy a Kinect if it can do the same things on the open source code AND a few relevant games pop up to support just that. But, to be honest, I can't think of a game I play that would benefit from it. Hell, I am one button away from supporting Wiimotes on my main machine - but I can't think of anything that I'd actually use it for past a very cumbersome remote pointer (and a £5 wireless presenter thing I have does a better job of that).
At the end of the day, it's a games gadget that requires hardware, calibration, software, a controlled environment, and a need to recognise bodily input. There's not much non-gimmicky that you can actually use it for.
There's been millions of wiimotes sold as well. Doesn't mean it's a killer device. All it proves is the marketing worked.
If *you* enjoy flailing around in front of a camera just for the sake of it then more power to you I guess. I'll stick with a control scheme that's not likely to knock over my beer and go to the gym when I want exercise.
> People are lazy, simple as that. Moving your butt instead of your thumb is a lot more difficult.
Not sure about kinect, but once you find that you can sit down with the wiimote and get the same, if not better results with a flick of the wrist (in say, bowling or golf), it rather takes the fun out of things. On the other hand, having a wheel to turn is a lot more fun than a more accurate mouse or keyboard for driving games.
I suspect Valve will say thank-you very much for the FLOSS, you will now be assimilated.
The downside is that the PC will get more crummy xbox ports.
What might be interesting is to have wearable sensors. How about an update to Karateka or Donkey Kong where you have sensors on your arms and legs and the whole thing is projected onto walls and/or floor rather than a screen?
Anyone that says kinect changes games is not much of a gamer. The general consensus these days is that motions controls are a novelty that quickly wears out it's welcome. Kinect Starwars is a great example of the gimmick that is motion controlled games. That's not to say that there's an occasional game the concept works well with. It's just few and far between and many developers were quick to buy in to the hype by slapping motion controls onto games that were not designed with them in mind.
In other fields it might hold promise.
Kinect wasn't the first attempt at camera based motion controls so it is hardly revolutionary. Evolutionary perhaps. Still it is cool that they released some of the code. Devils in the details. Doesn't seem like a microsoft move to do something that won't make them money or help with consumer/developer lock-in.
im not sure motion gaming is the way forward myself either. i am more interested with the other stuff like facial recognition and speech recognition in gaming.
i play a fair bit of fifa offline yet with mates in the house it is always more interesting, why? because with 2 people v the cpu you can actually form an attack. so, why not have things like shouting 'left wing' or 'far post' etc will make players listen to your commands.
even stuff like changing ammo types in guns - like judge dredd - would be cool.
i say all this as someone who owns PS Move and rarely used it, apart from killzone3 which was kinda fun with the rifle attachment.
Agree with you that V1 isn't quite there yet.
Although Kinect is actually more accurate and less laggy than PS Move, it doesn't work well as it cant focus on a single point like the light on the Move - but must track whole bodies.
Sounds like this might be fixed soon: http://www.engadget.com/2013/03/06/microsoft-research-mid-air-multitouch-kinect/
Motion controls for games probably is a novelty, or at least more of a very kid-centric sort of toy. But Kinect development has gone way beyond games at this point. There are business and medical apps that use that motion sensor package and MS is hoping to build on that.
Of COURSE they did this to ultimately increase their bottom line. Helloooo...corporation? You know, just like Apple, IBM, all the rest?
"Doesn't seem like a microsoft move to do something that won't make them money or help with consumer/developer lock-in"
Please. People always say that. Yet Microsoft are involved in stacks of open source projects both in benevolence and making it work better on windows fashion (the latter isn't evil, it's just what any sane company would do, but Apple don't - I know people who work on Open Source projects that Apple have said "sure" and then said it'll cost $XX,XXX for their time).
There's a lot of old gear out there. All of the various iThings have already been through several iterations of updated product. You can't say the same for Microsoft's flavor of the month.
The Apple based comparison would be more like pawn shops being full of iPhone 5's and iPad 3's.
Otherwise you're trying to pretend that an Apples an oranges comparison is relevant.
This to me looks like an attempt to at least partially head off the Leap Motion. The Kinect SDK is a huge monolith as stands and this is MS trying to get the community to help out make it a bit smaller and more (consumer oriented) developer friendly.
The Leap Motion SDK is already very developer friendly and the hardware is cheaper, smaller and generally sexier than the Kinect. Will be good to see what happens between these two devices in the ' gesture space' over the next six months.
This post has been deleted by its author
Leap is quite a way behind Kinect - and Kinect 2 is out soon. Leap also sucks quite badly at present: http://9to5mac.com/2013/01/25/first-hands-on-with-the-leap-motion-developer-edition-the-69-gesture-controller-for-mac-and-pcs/
Maybe that will all be fixed by launch, but I wouldn't count on it....
The guy writing that Leap article doesn't have a clue what he's talking about. He's gauging everything off the stock visualiser app's behaviour, but that's just one crude interpretation of the information from the sensors. It's up to Devs to make the best use of the device. It's akin to saying you can never do proper photo-editing on Windows, and basing that opinion on MSPaint, whilst making no attempt to install Photoshop.
Most views from people who have used the Leap I've read are amazed by the accuracy of the technology, but question whether they could use it for long periods of time. I look forward appropriate software and plugins being developed for it.
Anyway, it is a very different beast to the Kinnect- the Kinnect can make real-time scans of rooms, for example, the Leap is better for finger-scale input.
So he is providing an accurate review on what is available now, and what currently ships with the device. Seems quite rational to me - can't see where the 'doesn't have a clue' comes into it.
A lot of the issues he describes sound like underlying driver issues to me - nothing to do with the layered application.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
