What is a "vocabulary" for a gesture recognition algorithm. I've looked into a few gesture recognition libraries and they require a vocabulary in order to classify the gestures.

The gestures are a series of 3d points which I understand. But I don't understand what the vocabulary is? What kind of data is it and how is used in the recognition algorithm?

Thanks

Just guessing, but I think those would be the gestures you want to recognize. The vocabulary could be the collection of every gesture you want to detect.