Sampling-based contact-rich motion control

Liu, Libin; Yin, KangKang; Panne, Michiel van de; Shao, Tianjia; Xu, Weiwei

doi:10.1145/1778765.1778865

Cited by 127 publications

(81 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently Liu et al [2010] Control Composition: Research on the composition of control policies for heterogenous motions is relatively sparse. Direct parameter interpolation is sufficient for the transitions between various quadruped gaits [Coros et al 2011].…”

Section: Physics-based Methodsmentioning

confidence: 99%

“…We use the sampling-based method of Liu et al [2010] to construct open-loop controls from a single example of each of these actions. These controls are PD (Proportional Derivative) target-angle trajectories that are computed using an offline search procedure and therefore cannot adapt to external perturbations.…”

Section: Feedback Control From Single Examplementioning

confidence: 99%

“…We thus manually choose low-dimensional action vector a and sensory vector s for each motor skill. We note that the remaining degrees of freedom are simply controlled using the open-loop control provided by the method of Liu et al [2010].…”

Section: Feedback Control From Single Examplementioning

confidence: 99%

“…We can also choose to blendã as well. However, the reconstructed reference controls from Liu et al [2010] are PD target-angle trajectories that offset the reference motions. Hence as long as we re-calculate the reference actionsã from blended reference statess, we in effect have blended the reference control as well.…”

Section: Reference Control Blendingmentioning

confidence: 99%

See 3 more Smart Citations

Terrain runner

et al. 2012

Self Cite

View full text Add to dashboard Cite

Figure 1: An animated character runs, vaults, jumps, and drop-rolls across a parkour terrain during a real-time physics-based simulation. Given a single motion capture clip of each of these four skills as input, our method uses an offline process to develop robust control policies for parameterized versions of these skills, as well as robust transition motions. AbstractIn this paper we learn the skills required by real-time physics-based avatars to perform parkour-style fast terrain crossing using a mix of running, jumping, speed-vaulting, and drop-rolling. We begin with a single motion capture example of each skill and then learn reduced-order linear feedback control laws that provide robust execution of the motions during forward dynamic simulation. We then parameterize each skill with respect to the environment, such as the height of obstacles, or with respect to the task parameters, such as running speed and direction. We employ a continuation process to achieve the required parameterization of the motions and their affine feedback laws. The continuation method uses a predictor-corrector method based on radial basis functions. Lastly, we build control laws specific to the sequential composition of different skills, so that the simulated character can robustly transition to obstacle clearing maneuvers from running whenever obstacles are encountered. The learned transition skills work in tandem with a simple online step-based planning algorithm, and together they robustly guide the character to achieve a state that is well-suited for the chosen obstacle-clearing motion.

show abstract

Section: Physics-based Methodsmentioning

confidence: 99%

Section: Feedback Control From Single Examplementioning

confidence: 99%

Section: Feedback Control From Single Examplementioning

confidence: 99%

Section: Reference Control Blendingmentioning

confidence: 99%

See 2 more Smart Citations

Terrain runner

et al. 2012

Self Cite

View full text Add to dashboard Cite

show abstract

“…To make optimization easier, [5] learns from low dimensional motion data of stylistic motion. In order to adapt input motions, [6][7] use randomized search. Such strategies have produced many perfect lifelike results.…”

Section: Related Workmentioning

confidence: 99%

Optimization control for biped motion trajectory

Wang

Bao

et al. 2014

2014 International Conference on Audio, Language and Image Processing

View full text Add to dashboard Cite

Character animation is important in many applications, such as video games, movie production, and somatosensory interaction. However, it is difficult to control the locomotion of a virtual character in a dynamic environment, because bipeds are under-actuated and unstable. In this paper, we present a novel optimization strategy to improve the realism and stability of virtual character animation. Our trajectory optimization method produces lifelike motions by using the motion capture data. The method describes systematic computation of controllers that can optimize locomotion. We use center of mass (COM) control and foot control to enhance the sense of reality in physical movement, and use tracking control to maintain similarity with the reference motion. Experiment results demonstrate our approach is applicable by simulating walking motion and the optimized motion is both natural and recognizable.

show abstract

Designing controllers for physics‐based characters with motion networks

Backman

Kallmann

2013

Computer Animation & Virtual

View full text Add to dashboard Cite

We present a system that allows non-programmers to create generic controllers for physically simulated characters. The core of our system is based on a directed acyclic graph of trajectory transformations, which can be modified by feedback terms and serve as reference motions tracked by the physically simulated character. We then introduce tools to enable the automatic creation of robust and parameterized controllers suitable for running in real-time applications, such as in computer games. The entire process is accomplished by means of a graphical user interface and we demonstrate how our system can be intuitively used to design a simbicon-like walking controller and a parameterized jump controller to be used in real-time simulations.

show abstract

Sampling-based contact-rich motion control

Cited by 127 publications

References 32 publications

Terrain runner

Terrain runner

Optimization control for biped motion trajectory

Designing controllers for physics‐based characters with motion networks

Contact Info

Product

Resources

About