This semester I have an Independent Learning module and Research methods, then follows the Dissertation. Thankfully my Pathway Leader and Supervisors have been very helpful in enabling me to plan my work such that I can study the history and theory of 3D motion sensing for the ILM and work up a prototype implementation for the Dissertation.
To this end, my research starts now. I’m investigating the capabilities (and potential) of consumer-level solutions, naturally leading me to Microsoft’s novel new Xbox 360 peripheral, Kinect. Today is the first time in two weeks that I’ve had the opportunity to get back to hacking Kinect. My goal is to find all the hacked and official drivers out there and get them working in as many environments as possible, i.e. at least Windows, Mac and Linux.
After this I’ll be looking at the application-layer libraries that are available. Thankfully there’s a burgeoning open-source community writing open-standards-based gesture recognition libraries. But perhaps even better than this, Microsoft have announced that they will be releasing official drivers and libraries for their XNA Game Studio, although a launch date has not yet been given. And best of all, PrimeSense, the company Microsoft commissioned to write the Kinect sensor algorithms, has release much of its Kinect code to the open-source community. They are long-standing experts in gesture recognition and were wise enough, many years ago, to recognise the potential of such hardware-software tie-ups. They architected their software such that it is hardware agnostic. This is great news for us as it means that those big-budget industrial implementations they’ve done over the years will work beautifully with the Kinect. Therefore this immature consumer technology is actually rather mature, well-developed and heavily tested.
So…watch this space. I have various implementations working in Windows and Linux, and will probably attack OSX tomorrow.