5 ms·

I don’t mean to be a downer but I’m not sure how productive this is. I applaud his wanting to jump start something but this is a huge project. Think about it.

I don’t mean to be a downer but I’m not sure how productive this is. I applaud his wanting to jump start something but this is a huge project. Think about it. Just because the drivers let your computer see a hand doesn’t mean your computer can recognize it is a hand. That has to be done in software and doing that will be a massive amount of work. And that's just figuring out it's a hand. Not telling the program what a hand does. Anyone who has used a Kinect on the Xbox knows the software plays a huge part in its execution. I’d wager the configuration program alone took Microsoft far longer than the 6 allotted weeks this contest gives. I’d much rather see a high profile person like Matt Cutts start a fund for open source development of a Kinect library for Linux and then giving people the time to accomplish something great (it's certainly a project I'd be willing to donate to)

olalonde 16y ago
Right. At first sight, the video sent in by the Kinect looks like it could have been generated by a good computer vision algorithm.
donohoe 16y ago
At the very least it will spark the imagination. The beauty of this is that it doesn't have to have any winner before the end of the year to be a huge success. All Matt has to do is keep the conversation going, and get people thinking about what is possible. People will build great things, regardless of prize money, if they are inspired - that might be by Dec 31st or it could be 6 months from now... To this end, its a success.
bigiain 16y ago
A lot of the heavy lifting has been done. OpenCv and HandVu ( http://www.movesinstitute.org/~kolsch/HandVu/HandVu.html http://www.movesinstitute.org/~kolsch/HandVu/HandVu.html ) have got a lot of the "Huge project" bit done and packaged into easy to use libraries already.
Matt_Cutts 16y ago
Half of the prize money is set aside for people who make it easier to work with the Kinect, e.g. working on a Linux library. Mainly I want to get people excited about doing neat things with very cool device.
orangecat 16y ago
That has to be done in software and doing that will be a massive amount of work. But it's fun and challenging work, exactly the sort of problem hackers are willing to devote lots of effort to. I expect progress to be surprisingly rapid.
greendestiny 16y ago
6 weeks isn't a long amount time, but I don't think the point of the competition is replicating Kinect with open source. Surely there is more to open source than the copying of proprietary technology.

olalonde 16y ago

Right. At first sight, the video sent in by the Kinect looks like it could have been generated by a good computer vision algorithm.

donohoe 16y ago

At the very least it will spark the imagination. The beauty of this is that it doesn't have to have any winner before the end of the year to be a huge success. All Matt has to do is keep the conversation going, and get people thinking about what is possible. People will build great things, regardless of prize money, if they are inspired - that might be by Dec 31st or it could be 6 months from now... To this end, its a success.

bigiain 16y ago

A lot of the heavy lifting has been done. OpenCv and HandVu ( http://www.movesinstitute.org/~kolsch/HandVu/HandVu.html http://www.movesinstitute.org/~kolsch/HandVu/HandVu.html ) have got a lot of the "Huge project" bit done and packaged into easy to use libraries already.

Matt_Cutts 16y ago

Half of the prize money is set aside for people who make it easier to work with the Kinect, e.g. working on a Linux library. Mainly I want to get people excited about doing neat things with very cool device.

orangecat 16y ago

That has to be done in software and doing that will be a massive amount of work. But it's fun and challenging work, exactly the sort of problem hackers are willing to devote lots of effort to. I expect progress to be surprisingly rapid.

greendestiny 16y ago

6 weeks isn't a long amount time, but I don't think the point of the competition is replicating Kinect with open source. Surely there is more to open source than the copying of proprietary technology.