Meeting Date: 10.05.1999
What I did last week: I had Colin, Dan and Shaun compile a training set and have their gestures recognized. It works quite well apart from the limitations mentioned earlier (depends on the choice of gestures). I made a statistical evaluation of a tyical training set for the existing gestures. Details can be found here.
What I'll do this week: I'll write some more stuff on what I think the gesture recognizer can do and what not and where I see it going. Then I will have to focus a bit more on getting my Gesture Recognizer actually built into CoRgi. To do that will require some major work seeing as I have not the faintest idea of how to go about that!!! I will also start looking into selecting vertices, since I will ultimately have to deal with that problem in my main application. I decided to break the problem down into a simpler problem (selecting vertices in 2D) and then abstracting it to 3D.