Select an emotional quadrant, then play a gesture from the library that the Random Forest model predicts falls in that region of the valence/arousal plane. Each click picks a different random gesture.
Predicted V/A values will appear here after a gesture plays.