Wednesday, 3 August

Wednesday, 3 August

3:45 - 5:30 pm

Hall B

Session Chair: Mathieu Desbrun, California Institute of Technology

Animating Sand as a Fluid

Yongning Zhu
Robert Bridson

The University of British Columbia

Coupling Water and Smoke to Thin Deformable and Rigid Shells

Eran Guendelman

Stanford University and Industrial Light & Magic

Andrew Selle

Stanford University and Intel Corporation

Frank Losasso
Ronald Fedkiw

Stanford University and Industrial Light & Magic

Real-Time Subspace Integration for St. Venant-Kirchhoff Deformable Models

Jernej Barbic
Doug L. James

Carnegie Mellon University

Interactive Collision Detection Between Deformable Models Using Chromatic Decomposition

Naga Govindaraju
David Knott
Nitin Jain
Ilknurk Kabal

University of North Carolina at Chapel Hill

Rasmus Tamstorf

Walt Disney Feature Animation

Russel Gayle
Ming C. Lin
Dinesh Manocha

University of North Carolina at Chapel Hill

Wednesday, 3 August

3:45 - 5:30 pm

Petree Hall C

How and when do academic research ideas make their way into feature animation and visual effects production facilities? What kinds of graphics research ideas have made good production tools, and how are they transformed by practical experience and needs? What pressing production issues should be considered in academic circles? To what extent is the industry using standardized tools, which might be slowing adoption of new techniques? How are intellectual property issues resolved? How could academia and industry work together more closely to bridge the gaps? Panelists from academia and production explore these and other issues in this vital relationship.

Moderator

Daniel Goldman

University of Washington

Panelists

Tony DeRose

Pixar Animation Studios

Andrew Hendrickson

PDI/DreamWorks

Barbara Mones

University of Washington

Paul Salvini

Side Effects Software Inc.

Steve Sullivan

Industrial Light & Magic

Wednesday, 3 August

3:45 - 5:30 pm

Room 515B

Session Chair: Eric Haines, Autodesk, Inc.

Time-Varying Textures: Definition, Acquisition, and Synthesis

A new class of textures with simple algorithms for synthesis and controllability: Time-varying textures, in which appearance evolves over time. A database of these textures is being created.

Sebastian Enrique

Columbia University
senrique (at) cs.columbia.edu

Melissa Koudelka

Yale University

Peter Belhumeur

Columbia University

Julie Dorsey

Yale University

Shree Nayar
Ravi Ramamoorthi

Columbia University

Efficient Magnification of Bi-Level Textures

Standard tri-linear filtering blurs road signs and similar items excessively. We use a single-channel MIP map of small size and an efficient pixel shader to significantly improve bi-level textures.

Jörn Loviscach

Hochschule Bremen
jlovisca (at) informatik.hs-bremen.de

Real-Time Image-Based Control of Skin Melanin Texture

A novel process to control skin melanin texture over a continuous range. The process is implemented on graphics hardware and can achieve real-time processing for live video streams.

Norimichi Tsumura

Chiba University, PRESTO Japan Science and Technology Corporation
tsumura (at) faculty.chiba-u.jp

Toshiya Nakaguchi

Chiba University

Nobutoshi Ojima

Kao Corporation

Koichi Takase
Saya Okaguchi
Ryoko Usuba

Chiba University

NRT-Based Texture Replacement in Real Videos

A near-regular-texture-based approach for dynamic texture replacement in video that preserves geometric and lighting effects on the original texture.

Wen-Chieh Lin

Carnegie Mellon University
wclin (at) cs.cmu.edu

Yanxi Liu

Carnegie Mellon University

Wednesday, 3 August

3:45 - 5:30 pm

Room 511 AB

Session Chair: Hanspeter Pfister, Mistubishi Electric Research Laboratories (MERL) .

Boneless Motion Reconstruction

A new automated technique for character animation from motion-capture data that does not require global information (skeleton) and can be applied directly to animating mesh models.

Vladislav Kraevoy

The University of British Columbia
vlady (at) cs.ubc.ca

Alla Sheffer

The University of British Columbia

Surfacing By Numbers

Combining graph-cut-based methods for surface selection and detail transfer to provide a fast and flexible system for reusing geometry and texture across 3D surfaces.

Steve Zelinka

University of Illinois at Urbana-Champaign
zelinka (at) uiuc.edu

Michael Garland

University of Illinois at Urbana-Champaign

Knowledge-Based Modeling of Laser-Scanned Trees

A new method that uses knowledge about tree structures and allometric theory to produce full polygonal models of range-scanned trees, without suffering from the inherent incompleteness of such data.

Hui Xu

University of Minnesota
hxu (at) cs.umn.edu

Nathan Gossett
Baoquan Chen

University of Minnesota

Interaction Capture and Synthesis

Capture and reuse of interaction is improved by measuring both forces and motion. Assuming that captured motion emulated impedance-based control, the technique computes compliances from capture data and retargets via simulation.

Paul G. Kry

The University of British Columbia
kry (at) cs.rutgers.edu

Dinesh K. Pai

Rutgers University

Wednesday, 3 August

3:45 - 5:30 pm

Room 407

Moderator

Roy Ascott

Planetary Collegium, University of Plymouth

Featured Panelists

Steve Dietz

Director, ZeroOne: The Network
Director, ISEA2006 Symposium +
ZeroOne San Jose: A Global Festival of Art on the Edge

Carl Goodman

Curator of Digital Media/Director of New Media Projects
The American Museum of the Moving Image

Tomoe Moriyama

Curator of Media Art in the Images and Technology Gallery of The Tokyo Museum of Photography

Gerfried Stocker

Artistic Director, Ars Electronica

Christine Schopf

Artistic Director, Ars Electronica