Very interesting technology and looks almost too good to be true, which is a shame as with the content of many of your video's it led me to be highly sceptical at first. I guess part of the problem here is that many of the video's are targeted towards gaining interest from the public and gamers as opposed to hard facts that developers would want to see.
For example in the promo video you posted are claims of full 3D skeleton and 3D depth whilst only mentioning a 2D webcam which seems unlikely? Further more many of the video's on your youTube channel have a fake look to them, mainly due to the editing and poor reporting. For example in this video at 55 seconds in the reporter says the Panda is copying the moves the dancer makes. But its clear that the panda is moving before the dancer does. However, as it appears to be a classic 'dance game' this is to be expected as the player is meant to copy the virtual dancer not the other way around. Like I said poor reporting on behalf of the guy in front of the camera.
Sorry if that seems harsh, but it was my first impression, which is a shame as it looks like you've got some very impressive technology on your hands. Obviously much of what is shown in the actual demonstrations is clearly possible, such as full body tracking from a 2D webcam (presumably through segmented filtering and machine learning) and its not like you've sprung up over-night.
So as I've worked with Kinect over the last couple of years for various projects this makes your Extreme Motion technology very intriguing and appealing. Removing the need for specific hardware would be hugely beneficial so it was a no-brainer to sign up and request SDK access, especially as there is Unity plugin support. Hopefully my application will be approved as i'm keen to put this tech through its paces and see just what it is capable of achieving as well as how it could free us from the Kinect.
Edited by noisecrime, 20 May 2013 - 10:03 PM.