Frame-Synchronous Hand Gesture Detection by Projected Winding Order
Gesture recognition on video is normally posed as classification: assign a label to each frame, then act on the label. That formulation is adequate for control, where a command may be obeyed several frames late without a user noticing, and inadequate for synchronisation, where an output must be aligned to the frame on which the gesture physically occurred. We take the synchronisation problem for a specific and common movement, the rotation of an open hand about its own long axis, and show that it admits an exact solution requiring no classifier, no training data and no calibration. Let s denote the normalised two-dimensional cross product of the two palm edges, taken at the wrist and the two outer knuckles, under the projection the camera already performs.