Motion capture-driven simulations that hit and react

Abstract

Controllable, reactive human motion is essential in many video games and training environments. Characters in these applications often perform tasks based on modified motion data, but response to unpredicted events is also important in order to maintain realism. We approach the problem of motion synthesis for interactive, humanlike characters by combining dynamic simulation and human motion capture data. Our control systems use trajectory tracking to follow motion capture data and a balance controller to keep the character upright while modifying sequences from a small motion library to accomplish specified tasks, such as throwing punches or swinging a racket. The system reacts to forces computed from a physical collision model by changing stiffness and damping terms. The freestanding, simulated humans respond automatically to impacts and smoothly return to tracking. We compare the resulting motion with video and recorded human data.

Simulating balance recovery responses to trips based on biomechanical principles

“…One approach is to change the gains of PD controllers to deal with perturbations [ZH02,ACSF07]. For example, Zordan and Hodgins proposed a method to simulate a reactive motion from motion capture data by lowering the gains of the simulation during the perturbation, allowing the simulation to be pushed off the desired trajectory by the forces of the interaction, and then returning slowly to an appropriate motion capture trajectory [ZH02]. They also implemented a balancing controller that could maintain balance during the disturbances caused by the motion of the upper body.…”

Fluid Simulation with Articulated Bodies

“…Zordan and Hodgins [20] create simulations from motion capture data that are able to hit and react to collisions. This work is closest to ours in spirit.…”

Pose Control in Dynamic Conditions

“…The problem here is that stiff tracking eliminates the natural response to forces and perturbations, rendering the system effectively equivalent to computationally expensive kinematic animation. Methods have been proposed to change the control parameters when a collision occurs [18], but these approaches require specific ad hoc parameters and it is unclear how such systems can deal with frequent contacts.…”

“…The application of precise PD control to articulated characters has been considered previously [1,8,18], and is not addressed here since the proposed method is fully compatible with any existing use of PD control where the moment of inertia can be computed.…”