Durand’s research addresses all aspects of image synthesis and capture, and this integration enables him to address transversal issues such as 3-D modeling from 2-D images, relighting of photographs, real-time photorealistic effects and material appearance capture. His research combines computer science, mathematics, physics, visual perception and the visual arts.

Subhash Khot
Georgia Institute of Technology
Assistant professor
College of Computing

 

Khot works in the area of theoretical computer science, with an emphasis on complexity theory. He tackles problems that are among the most difficult and long-standing in computer-science theory, using novel techniques that draw on fields such as coding theory, linear algebra and Fourier analysis. He has provided specific leadership in the use of Probabilistically Checkable Proof Systems to prove many inapproximability results, an approach that has been proven powerful.

Dan Klein
University of California at Berkeley
Assistant professor
Computer Science Division

 

Klein’s research demonstrates the feasibility of unsupervised methods of learning to natural language processing problems such as grammar induction and machine learning. His efforts to enable computers to learn important language information, such as grammar, from abundantly occurring data, as opposed to hand-labeled data, could have an enormous impact.

Radhika Nagpal
Harvard University
Assistant professor of Computer Science
School of Engineering and Applied Sciences

 

Nagpal is interested in robust programming paradigms for systems composed of large numbers of embedded, locally interacting, identically programmed nodes, such as sensor-actuator networks, smart materials, and self-assembling and swarm robotics. Her research draws on concepts from embryo development suggested by biologists to explain how globally robust behavior can emerge from the decentralized interactions of less reliable cells.

Wei Wang
University of North Carolina at Chapel Hill
Assistant professor
Department of Computer Science

 

Wei proposes to use novel techniques in data mining, automatic classification and natural language text retrieval to address a central challenge of molecular biology: linking proteins to their function. She has developed algorithms to find recurring amino acid packing patterns in protein structures and to select those patters whose occurrences are highly associated with known functionalities.