Traditionally computer graphics has focused on developing algorithms and systems for performing efficient simulations that transform analytic representations into images and animations. The employed analytic models are often simplified and heuristically defined. Thus, they are often incapable of representing complexity of real world. In the recent years, we have also witnessed significant technological developments in the areas of high-quality sensors and measurement devices. However, the data provided from these devices are frequently incompatible with the representations assumed by most computer graphics systems. In my research I am exploring new data-driven approaches to computer graphics. These approaches bridge the gap between parametric and empirical modeling. They differ from the classical "simulation-based" computational model that pervades today's computer graphics, and instead depend more on the tools of interpolation, data analysis, and signal processing. In my talk I will discuss three specific computer graphics applications. First, image-based rendering, modeling, and lighting addresses the problem of rendering novel views under novel lighting based on a collection of photographs. I will present image-based techniques in the context of visual hulls, 3D photography, and 3D TV systems. In a second application, I will discuss the problem of interpolating and extrapolating new reflectance models (specifically isotropic BRDFs) from a collection of acquired samples. Finally, I will discuss a new data-driven approach to modeling the space of natural textures.

Bio: Wojciech Matusik is a post-doctoral associate in the Computer Science and Artificial Intelligence Laboratory at the Massachusetts Institute of Technology. He is also a visiting research scientist at Mitsubishi Electric Research Labs. He received a B.S. in EECS from the University of California at Berkeley in 1997, M.S. in EECS from MIT in 2001, and consequently Ph.D. in 2003. His primary research lies in computer graphics with an emphasis on modeling based on measured data. He also works on image-based rendering, modeling, and lighting where he developed efficient algorithms for computing and rendering visual hulls. He is currently developing end-to-end 3D TV systems.