- 最后登录
- 2018-6-29
- 注册时间
- 2011-7-1
- 阅读权限
- 20
- 积分
- 359

- 纳金币
- 335582
- 精华
- 0
|
Controlling Physics-Based Characters Using Soft Contacts
Sumit Jain C. Karen Liu
Georgia Institute of Technology
Abstract
In this paper, we investigate the impact of the deformable bodies on
the control algorithms for physically simulated characters. We hy-
pothesize that ignoring the effect of deformable bodies at the site of
contact negatively affects the control algorithms, leading to less ro-
bust and unnatural character motions. To verify the hypothesis, we
introduce a compact representation for an articulated character with
deformable soft tissue and develop a practical system to simulate
two-way coupling between rigid and deformable bodies in a robust
and efficient manner. We then apply a few simple and widely used
control algorithms, such as pose-space tracking control, Cartesian-
space tracking control, and a biped controller (SIMBICON), to sim-
ulate a variety of behaviors for both full-body locomotion and hand
manipulation. We conduct a series of experiments to compare our
results with the motion generated by these algorithms on a char-
acter comprising only rigid bodies. The evaluation shows that the
character with soft contact can withstand larger perturbations in a
more noisy environment, as well as produce more realistic motion.
CR Categories: I.3.7 [Computer Graphics]: Three-Dimensional
Graphics and Realism—Animation;
Keywords: character animation, articulated rigid body, de-
formable body, linear complementarity problem
1 Introduction
One of the fundamental simplifications that researchers in physics-
based human motion synthesis make, is that motion is the product
of an articulated rigid body system with actuated joints representing
bones and active skeletal muscles. On the surface this abstraction
does capture the most fundamental aspects of the human muscu-
loskeletal system. Utilizing this assumption, researchers have de-
veloped several control algorithms that can synthesize movement
for various tasks like balance and walking. Although these con-
trollers work well in their specific problem domain, they still cannot
achieve the same level of agility the human body displays.
In this paper we revisit the fundamental assumption that an articu-
lated rigid body system, by itself, captures the fundamental prop-
erties that enable human-like motion. We focus on one aspect of
the motion that is not captured by this simplified model: the contact
with the environment primarily occurs through the soft tissue. This
factor comes into play in any situation where there is a collision be-
tween the character and the environment. Collisions between rigid
bodies usually result in sporadic contact points and highly discon-
tinuous pressure distribution. Although a character consist of only
rigid bodies is ideal for efficiently simulating human movement,
we postulate that the simplified rigid contact model inadvertently
increases the difficulty in controller design and results in unrealis-
tic motion.
![]()
The primary contribution of this paper is to demonstrate that simple
control strategies coupled with the simulation of soft tissue defor-
mation at the site of contact can achieve very robust and realistic
motion. We develop a practical system that allows us to simulate
two-way coupling between rigid and deformable bodies in a robust
and efficient manner. We then apply a few simple and widely used
control algorithms, such as pose-space tracking control, Cartesian-
space tracking control, and SIMBICON, to simulate a variety of
both full-body locomotion and hand manipulation. The resulting
motions are compared with the motion generated by these algo-
rithms on a character comprising only of rigid bodies. These sim-
ple controllers demonstrate that the character with soft contacts can
withstand larger perturbations in a more noisy environment, with-
out the need of designing more sophisticated control algorithms.
Simulating deformable bodies can be achieved in a few different
ways and the design choice often has to balance the required accu-
racy and performance. We hypothesize that the accuracy offered
by sophisticated but expensive methods, such as Finite Element
Method (FEM) is unnecessary for our application for two reasons.
First, unlikemost previous work that simulates deformation of com-
plex volumetric meshes for aesthetic purpose, the primary goal of
our work is to produce deformation for more physically correct con-
tacts. Second, average human body deforms marginally due to the
support of bones. In particular, the deformation due to contacts is
typically small and localized. We take advantage of these proper-
ties to design a simple and accurate model that only computes the
surface of deformable bodies, rather than the entire volume.
全文请下载附件: |
|