I would be interested in how he plans on dealing with the high CPU percentage of using skeleton streams while watching High Definition Media. I am also interested in how he plans on utilizing the kinect microphone array to pinpoint users for voice commands while the environment is saturated in either DTS, or any type of HD audio being produced by media browser??
Finally, and most importantly I would be extremely interested in how much he will charge for his kinect media browser integration application? Will it be free because he will utilize the Microsoft SDK, or will he decide to utilize some of the other SDKs in the internet to build his app?
The bottom line is this, kinect is a process hog, it wants to use every single bit of memory and CPU percentage your PC has to offer up. In order to wield most its features you have to be smart about turning off and on certain features, building speech grammars for EXACTLY your position in media browser/ home automation commands.
The final thing I will say about kinect media browser and home automation is this:
If you use the kinect SDK they recommend that you use their "Microsoft.speech" namespace. The problem is that, although this particular set of speech recognition objects don't need any voice training, it only allows for a maximum of 300 possible grammar objects to be added to the recognizer. That my friends is an extremely small library of media, and forget about home automation commands on top of it.
So, how do you over come this obstacle? You use a completely different speech namespace with the ability to handle large sets of commands, and also means a whole bunch of extra work to get it working with kinect(two years worth of work to be exact).
This is not a venting session, not to be taken the wrong way, I will use his applications for sure. I have five kinects running in my home. All running media browser and some even communicating with Eventghost and a USBUIrT to turn off and on lights and TV by voice command.
Edited by chef, 18 April 2014 - 09:18 PM.