Feedup, feedback, and feedforward in curve mid-air 3D gestures

Issuing a mid-air gesture in a three-dimensional space intrinsically suffers for the lack of explicit direct representation of the gesture with which guidance and feedback can be offered. To address this challenge, we decompose the feedback problem into three components: feedup to constantly represent the goal of the gestural task, feedback to respond to what the end user already did related to the initial goal, and feedforward to modify the representation towards the ultimate goal before terminating the gesture production. We exemplify these three components with case studies representing three levels of complexity of Curve Mid-Air 3D Gestures produced in three environments.