Predictive Feedback for Interactive Control of Physics-based Characters

Interactive control of a physically simulated character is a challenging problem, due both to the complexity of controlling multiple degrees of freedom with lower dimensional input and because many interesting motions lie on the fringes of character stability. This paper addresses these problems using a novel technique called predictive feedback, where a glimpse into the near future for a few sample inputs is continuously presented to the animator. We discuss issues related to the spatio-temporal distribution of predictions so that they provide meaningful and timely feedback to an animator interactively controlling a physics-based character with simple input devices, like a mouse or keyboard. We propose a visual presentation of this predictive feedback in which control input samples are chosen in the proximity of the user’s current input and the predicted results are co-located with the position of the input necessary to achieve them. We further show how the predictive samples may be automatically interpolated to control aspects of the character’s motion, such as balance, thereby freeing the animator to focus on other details. The paper thus contributes a technique for physically simulated characters that simplifies interactive character control and increases the range of motion that can be performed by both novices and experts. Many of the presented concepts extend beyond our specific input device and dynamic character control setting to more general input tasks.