Wayne H. Enright, Department of Computer
Science, University of Toronto
Fast Visualization/Animation of Approximate Solutions of PDEs
Visualizing and displaying approximate solutions of PDEs on high resolution
screens and other devices has become a significant part of scientific
computation. On the other hand the prime focus of numerical analysts
has been and still is on the development of effective numerical methods
for generating accurate approximate solutions on relatively coarse meshes
(covering the domain of interest). In order to display or visualize
the results of such methods at the resolution required by the current
display devices significant computational effort is often required and
we are investigating efficient algorithms for carrying out these tasks.
In this presentation we will consider the efficient generation of contour
plots and suface plots and the use of animation. We have developed and
implemented (in MATLAB) an approach that directly generates surface
plots and contour lines associated with a function $u(x,y)$ that does
not have to solve a PDE on a fine mesh.
Our approach exploits knowledge of an underlying PDE to define a multivariate
piecewise polynomial on a coarse unstructured mesh. This piecewise polynomial
can then be used to generate accurate off-mesh data (in particular data
required for effective visualization).We will show how this allows us
to generate animated contour plots at a much lower cost than that involved
in generating animated surface plots. Examples will be presented to
show the effectiveness of this approach on a variety of realistic problems.
Jessica Hodgins, School of Computer Science,
Carnegie Mellon University
Animating Human Characters
Computer animations and virtual environments both require a controllable
source of motion for their characters. Two possible solutions are simulation
and motion capture and over the past 10 years, we have explored both
techniques separately. For example, we developed control algorithms
that allow rigid body models to run or bicycle, bounce on a trampoline,
and perform a handspring vault. More recently we explored several interfaces
for controlling a avatar animated from a database of motion capture
data. Recently, we have begun to combine simulations with motion capture
data in the hope that these techniques will benefit both from the physical
realism of simulation and from the humanlike motion provided by captured
Jessica Hodgins joined the Robotics Institute and Computer Science Department
at Carnegie Mellon University as a Associate Professor in fall of 2000.
Prior to moving to CMU, she was an an Associate Professor and Assistant
Dean in the College of Computing at Georgia
Institute of Technology. She received her Ph.D. in Computer Science
from Carnegie Mellon University in 1989. Her research focuses on computer
graphics, animation, and robotics. She has received a NSF Young Investigator
Award, a Packard Fellowship, and a Sloan Fellowship. She is editor-in-chief
of ACM Transactions on Graphics and will be SIGGRAPH Papers Chair in
James Kuffner, School of Computer Science,
Carnegie Mellon University
Task-level Character Motion Synthesis as High-dimensional Search
Gross body motions for animated characters can be represented by time-parameterized
function curves in the joint configuration space. The problem of automatically
synthesizing animation for characters can then be formulated as a discrete
search over a continuous, high-dimensional configuration space for a
motion trajectory that satisfies a number of constraints. The constraints
on the motion can arise from joint limits, obstacles in the world, self-collisions,
body physics, or aesthetic considerations such as naturalness and style.
This talk considers the application of robot motion planning techniques
to the problem of automatic motion synthesis for articulated characters.
Key research issues will be discussed along with demonstrations of interactive,
task-based control interfaces for character navigation, object manipulation,
and full-body motions.
Realism and Psychology in Animation
The notion that computer graphics can recreate 'photorealistic' reality
had been an ongoing fetish in the industry since its inception a generation
ago. And in spite of its use, or more likely because of it, it is largely
perceived as a sterile, impersonal, crass, perhaps even menacing medium.
Part of the answer, in my opinion, is that CG artists are not yet
exploring another sort of realism that CGI can show--the realism of
human thoughts, emotions, and psychological nuances. What I'm interested
in is not achieving 'photorealism' per se, but coopting elements of
photorealism to recreate the realism of the complex, chaotic, conflicted,
sometimes mundane and always glorious mess that we call "human
nature". How can one capture that mathematically?
I will show examples of how "Psychorealism" has been explored
in animation, including my short films "the end" and "Bingo".
I will talk a bit about the emerging animation form of the 'animated
documentary', and my current project, an animated documentary "Ryan",
based on the life of the Canadian animator Ryan Larkin.
Raghu Machiraju, The Ohio State University
Synthesis and Animation Through Analysis of Data
Visualization and data analysis can allow for the systematic detection
of features. For example, features like vortices in a flow field can
be detected using efficient and robust methods derived from the physics
of the problem. Similarly, discrete simulation models can be used to
generate phenomena. These methods can generate rough looking surfaces
are amenable to user control.
Animation has relied on the direct numerical simulation of phenomenon.
It is often hard to achieve to allow interactive control. Perhaps, one
could use results from visualization and discrete simulation in an inverse
fashion to generate necessary imagery. I will first describe our recent
work in visualization and modeling and then conclude with some thoughts
for using the same towards animation.
Dinesh K. Pai, Division of Computer and Information
Sciences, Rutgers University
Multisensory Interactive Animation
Humans experience the world with all their senses, including vision,
touch, and hearing. I will describe recent progress in my group towards
constructing multisensory animations with integrated graphics, haptics,
and sounds. The animations are based on interactive simulation of physically
based models, and provide the immediacy of interacting with the real
world. I will describe how we can construct mathematical models suitable
for multisensory simulation, and reality based modeling using the UBC
Active Measurement Facility (ACME) and the new Rutgers Haptic, Auditory,
and Visual Environment.
Dinesh K. Pai is a Professor in the Department of Computer Science at
Rutgers, the State University of NJ. He is also a Professor at the University
of British Columbia and a fellow of the BC Advanced Systems Institute.
He received his Ph.D. from Cornell University, Ithaca, NY. His research
interests span the areas of graphics, robotics, and multisensory human-computer
interaction. One current research focus is reality-based modeling, i.e.,
building multisensory computational models of the physical world from
measurements. This includes a recent thrust in reconstruction from medical
ultrasound images. Another focus is fast simulation with integrated
sound, haptics, and graphics, especially simulation of contact.
Rick Parent, Department of Computer and
Information Science, Ohio State University
Human speech is complex. The main visible articulators, the lips and
tongue, are deformable. The muscles which drive speech produce visible
secondary effects which convey meaning. Speech is context dependent
and is changed by intent and emotion. Incorporating all of this into
a computational model for the visual aspects of speech is a challenging
task and will be discussed.
Nancy Pollard, Brown University
Animating Manipulation Tasks from Human Motion Data
Captured human motion data has been used to create convincing human
motion for a wide variety of tasks including locomotion, sports, and
dance. However, generalizing motion capture data to new situations remains
difficult, and tasks that involve manipulating objects in the virtual
world are rarely addressed. This talk considers the problem of adapting
a motion capture data for a quasistatic manipulation task to new object
geometries and friction conditions. I will argue that correct physics
is critical for this type of task and will present an algorithm that
generates manipulation plans for new objects by using the captured motion
data to constrain the solution space to a set of plans physically similar
to the original. Our planner provides guarantees on maximum task forces
and flexibility in contact placement and works for a wide range of object
geometries and coefficients of friction. I will show results for the
task of tumbling large, heavy objects. We have demonstrated these results
with a humanoid robot as well as with animated characters.
Michiel van de Panne, University of British
Haptic control envelopes and other unfinished projects
Animation tools are perhaps at their best when they neither over-constrain
nor under-constrain the motions to be created. After discussing some
of the types of constraints that of possible use in defining animated
motion, I will present some recent experiments with interactive systems
that enforce different types of computable constraints. I'll describe
other unfinished projects and unsolved problems as time allows.
Back to top