September 2, 2012
I use depth cameras (along with 2D IR cameras) for gesture/body motion input, as well as head tracking for VR.
I use Kinects and Xtions (preferred for its 60fps mode), in "vvvv" using OpenNI-based plugins and custom C# code. I do not use the built-in skeleton tracking (too much latency), instead doing direct processing on the point-cloud data; so I'm currently interested in learning more about various methods of focused data extraction, and arbitrary camera orientation calibration for combining point-clouds from multiple depth cameras.
Yes, I have a number of experiments where a person directly interacts with abstract sound and light in an immersive 3D VR environment that I think people would find tons of fun, and I would love to get feedback on.
I create immersive interactive visual and sound environments, that provide a space for exploration, creation, and play via direct interaction with sound and light.