LUMS Faculty, PhD Student Collaborate with CMU & Disney
Dr. Sohaib Khan, Department Chair and Associate Professor, Department of Computer Science at LUMS and PhD student, Ijaz Akther recently collaborated with Carnegie Mellon University (CMU) and Disney Research on graphics research.
Their paper titled “Bilinear Spatiotemporal Basis Models” was presented at SIGGRAPH 12, the International Conference on Computer Graphics and Interactive Techniques, at the Los Angeles Convention Centre on August 6, 2012, which is the top forum for Graphics research. Other members on this research team included Yaser Sheikh, Assistant Research Professor in Carnegie Mellon's Robotics Institute; Tomas Simon, a PhD student at CMU and Iain Matthews, Senior Research Scientist at Disney Research.
The LUMS team along with researchers at CMU and Disney Research, Pittsburgh, have developed a graphics software which could greatly simplify the editing process of producing computer-animated movies and games. This could benefit computer graphic artists who spend a lot of time creating subtle movements such as expressions on faces, gestures on bodies and the draping of clothes.
Graphics software usually represents dynamic objects, such as an expressive face, as a sequence of shapes, with each shape composed of a set of points in space. Another way to model an expressive face is to chart each point on the face as it shifts location over time. Each method has its advantages, but the sheer number of possible variations is tremendous, which results in models that are large and difficult to manage.
The researchers discovered that they could create a model that simultaneously takes into account both space and time - a bilinear spatiotemporal basis model. This approach enabled the researchers to create a much more compact, powerful and easy-to-manage model. They were able to demonstrate that they could reproduce a dynamic sequence, with millimeter precision, after discarding 99 percent of the original data points.
The research team explained that the natural limitations on spatial movements, such as the characteristic ways that the face changes shape as someone is talking or expressing an emotion, combine with the natural constraints on how much movement can occur over a given stretch of time. This enables the models to be very compact and efficient. The new software will enable graphics artists to do things more sensibly with less work.
The related paper and videos can be viewed here.