Weighted network gesture recognition

Gesture detection using weighted networks is one of the advanced approaches in gesture detection. In this section we will have a quick walkthrough about this approach and we will see how this works, rather than doing an actual implementation. We have seen most of the measurements of gestures are done on the basis of skeleton joint movements, and for all the earlier cases, joints were clearly visible to the sensors (which we have checked with the joints-tracking state). On the other hand, can we just say that gesture detection of joints can be done only for the tracked joints?

Think about some body movements that are very dynamic in nature. For example, bending exercises (forward, sideways) where we initially start with a straight position, then slowly raise our hands forward, and try to touch our toes (forward bending) or standing upright with legs apart, keeping one hand on the waist and raising the other hand while trying to bend sideways (sideways bending...