About the Virtual Textiles Research Group

Simulated virtual prototypes have revolutionized the design process in most engineering fields, eliminating much trial-and-error and enabling designs to be refined before real prototypes are built. As computational power becomes commonplace, a major emerging application of simulation is to design fields where the aesthetics of appearance, shape, and motion are primary concerns on a par with strength, weight, or efficiency. However, certain classes of materials consistently elude the available tools, and textiles form one of the most important examples.

Merely realistic simulated cloth, as long as it’s artist-controllable, suffices for entertainment, but design and retail are about reality rather than virtual worlds: predictive fidelity to real materials is critical. In design, predictive simulations can greatly improve efficiency: textile designers can design fabrics without producing test runs, and designers of clothing or furniture can evaluate their designs before the fabrics to be used have even been produced. In retail, predictive simulations can let customers select fabrics for interiors, evaluate fit and appearance of clothing, and order products, even ones that are unique and will only be produced upon demand. These practices are already common in settings where materials are simple and exact appearance is not crucial, but currently cannot extend to the appearance-critical area of textile products because of limited visual fidelity. Textiles and garments constitute a trillion-dollar global industry that is rapidly becoming more high-tech and is ready to adopt virtual prototyping as soon as the right technology is available.

Current techniques for simulating these materials have two fundamental limitations. First, the problem has been approached as two separate problems: either appearance is paramount, in which case measured geometry or texture is used to render cloth without regard for deformation; or motion is the focus, and detailed sheet or yarn motion is simulated, but rendering is a post-process disconnected from the real material. But appearance and deformation can’t be separated in predictive applications. Second, even the best results in both areas are only qualitative matches to real materials.

We aim to develop integrated methods for realistic cloth simulation that will allow higher fidelity, improved predictive power, and new fast algorithms that will open up a range of new applications.


Research Directions

Appearance  Rendering

Constructed woven and knit textiles are difficult to render computationally, especially given subtleties of appearance based on light, shape and motion. We seek to create the most accurate and realistic cloth surfaces possible using varied models for tracing, lighting and deformation. 

User Experience

We continue to explore the idea of a user interface that can support appearance based rendering as well as a related database of mechanical attributes and tendencies so as to predict the virtual behavior of fabric in space. 

unnamed.png

Yarn Modeling

Virtually all constructed cloth is created using knitting or weaving. Yarn is the integral building block of this process and we seek to better understand the computational behavior of fabric by focussing on attributes at the fiber level. 


The Team

Principle Investigators

Brooks Hagan
Rhode Island School of Design, Textile Department

Brooks Hagan is a textile designer and researcher. He works with numerous industry and technology leaders. His research areas include industrial processes and 3D weaving. View profile.

Kavita Bala
Cornell University, Department of Computer Science

Kavita's research interests span computer graphics, computer vision, deep learning and human perception including. View profile.

viewImage.jpg

Doug L. James
Stanford University, Department of Computer Science

Doug's research interests include computer graphics, computer sound, physically based modeling and animation, and reduced-order physics models.  View Profile.

steve.jpg

Steve Marschner
Cornell University, Department of Computer Science

Steve's research and teaching are in computer graphics and vision, centered around how optics and mechanics determine the appearance of materials. View Profile.

 

Affiliated Researchers

self_170.jpg

Shuang Zhao
Assistant Professor, UC Irvine

Shuang's research aims to computationally reproduce the appearance of the real world using physically based models, allowing the prediction of object appearance under a wide range of conditions.  View profile.

Joy.jpg

Joy Ko
Critic, Architecture Department, RISD

Joy is an architecture and computation critic whose work reflects her background and lifelong focus on design, mathematics and programming. View profile.

Hewett.jpg

Daniel Hewitt
Executive Director of Research, RISD

Daniel is an architect, researcher and critic. His research explores new models for culturally generative infrastructure in changing cities. View profile.

 

RESEARCH ASSISTANTS

IMG_40013.jpg

Anjuli Bernstein
RISD Textile MFA

 

jcleaf.jpg

Jonathan Leaf                                    Stanford University, PhD Candidate

Screen Shot 2018-02-07 at 7.33.23 AM.jpg

Gabrielle Ferreira
RISD Textile MFA

Carolina.jpg

Carolina Jimenez
RISD Textile MFA

 

treman-hike-port.jpg

Eston Schweickart                         Cornell Computer Science, PhD Candidate

 

face.jpg

Rundong Wu
Cornell Computer Science, PhD Candidate

14680762_10105006776062670_5284911451703931597_n.jpg

Emily Robertson
RISD Textile MFA

 

 not pictured:

Emily Holtzman
RISD Textile BFA

Claire Harvey
RISD Textile BFA