Los Angeles Convention Center, Room 408B
In recent years, various methods have been introduced to exploit pre-recorded data to improve the performance and/or realism of dynamic deformations, but their differences and similarities have not been adequately analyzed or discussed. So far, the proposed methods have been explored mainly in the research context. They have not been adopted by the computer graphics industry.
This course bridges the gap between research labs and industry to present a unifying theory and understanding of data-driven methods for dynamic deformations that may inspire development of novel solutions. It focuses on application of data-driven methods to three areas of computer animation: dynamic deformation of faces, soft volumetric tissue, and cloth. And it describes how to approach these challenges in a data-driven manner, classifies the various methods, and demonstrates how data-driven methods can work in other settings.
COURSE SCHEDULE
9 am
Introduction and Overview of Methods
Otaduy
9:20 am
Tissue and Cloth Mechanics
Otaduy
9:50 am
Capturing Deformation Examples
Bradley
10:30 am
Break
10:45 am
Facial Animation
Bickel
11:25 am
Cloth Animation
Wang
12:05 pm
Conclusion/Q & A
All
Level
Intermediate
Prerequisites
Basic knowledge of physically based simulation.
Intended Audience
Researchers and programmers in computer animation.
Instructor(s)
Miguel Otaduy
Universidad Rey Juan Carlos Madrid
Bernd Bickel
Disney Research Zürich
Derek Bradley
Disney Research Zürich
Huamin Wang
The Ohio State University