bkbrowncom

And here is yet some more fun with Kinect-based face tracking, only this time I'm not doing any sort of device control, but rather combining the data stream with a NeuroSky MindWave brainwave monitor: http://bkbrown.com/projects-KinectMindwave.html

This video shows my latest work with the servo-based webcam:

http://youtu.be/xb65iuOj8VE

In this video I'm using the latest SDK v1.5.1, and I also modified the control electronics and software to use a Pololu Micro Maestro servo controller. I wrapped the Maestro DLL in a C# class library, and this time instead of making method calls directly from within the OnAllFramesReady event handler, I spun a worker thread and did some running averaging on the face coordinates to help smooth the the webcam's motion.

Another interesting thing shown in this latest video is Kinect Studio v1.5.1 running with the 3D Viewer and Depth Viewer activated. In a couple of instances I move in and out of the Kinect's field of view to demonstrate how quickly the face tracking picks up. This is a big improvement over my last experiment, where there was some delay experienced before tracking was accomplished.

I thought I'd take the recently released Kinect SDK v1.5 out for a drive and experiment with the facial tracking features using my servo-controlled webcam. Here's my first video clip:

http://www.youtube.com/watch?v=EPWxGrKn-r0

On a similar topic, here's a link to my latest video showing a Kinect servo-controlled camera:

http://www.youtube.com/watch?v=HyRy_04b0dY

This is the first stage of a larger project that will also incorporate speech recognition.

This wrapper is working great in a project I'm working on (http://www.youtube.com/watch?v=xln6hLe_gF4&feature=youtu.be).This project blends Kinect gesture control in Unity along with physical hardware control using a Programmable Automation Controller. Some of my other work with machine control and the Kinect SDKs (both the MS Research beta and v1.0) are described on my website: http://bkbrown.com/ . 

My compliments to the author (who is credited in the YouTube video). I'd be happy to assist anyone who has questions on using this very nice piece of work in an actual application.