4/22/2013
Written by
Two computer science professors have been selected to receive 2013 Sloan Research Fellowships from the Alfred P. Sloan Foundation. Professors Derek Hoiem and Svetlana Lazebnik are among 126 early career scientists and researchers from 50 colleges and universities who have been honored with a two-year fellowship. The $50,000 fellowships recognize researchers for distinguished performance and their unique potential to make substantial scientific contributions.
Hoiem and Lazebnik both work in computer vision, an area where researchers explore methods to acquire, process, analyze, and understand images. “Derek and Svetlana are pioneering incredible new methods and technologies in computer vision that are laying foundations for the future of the field,” said Rob A. Rutenbar, Head of the Department of Computer Science. “We are extremely proud that their work has been honored with this prestigious award.”
Hoiem is working on general visual scene understanding—interpreting images in terms of real-world space, surfaces, objects, and relationships. Rather than breaking vision into a set of narrowly defined tasks and tackling them as independent pattern matching or image processing problems, Hoiem seeks to build widely-applicable frameworks that can infer general properties from images and describe the unfamiliar in familiar terms.
He recently co-wrote a book, Representations and Techniques for 3-D Object Recognition and Scene Interpretation, which organizes and makes accessible recent advances in the field. Hoiem is also a recipient of a Best Paper Award at CVPR 2006, a 2009 Google Research Award, a 2011 NSF CAREER Award, and a 2012 Intel Early Career Faculty Award. In addition, he received an Honorable Mention for the 2008 ACM Doctoral Dissertation Award. Before joining the faculty in 2009, Hoiem was a postdoctoral fellow at the Beckman Institute for Advanced Science and Technology. He earned a doctorate from Carnegie Mellon University in 2007.
Lazebnik seeks to create a new generation of recognition systems that handle the large-scale, constantly evolving, heterogeneous photo and video collections that are now found online. Her research is aimed at discovering the collective structure of these collections to create compact and scalable representations for accessing their content. Her work has yielded advances in several areas, from fast techniques for fundamental operations such as similarity search; to efficient methods for organizing photo collections based on perceptual and geometric constraints; to higher-level systems capable of interpreting images in terms of their constituent objects, parts and materials.
She is also the recipient of a 2008 NSF CAREER Award, a 2009 Microsoft Faculty Fellowship, a 2011 DARPA Computer Science Study Group Award, and was recently program chair for the European Conference on Computer Vision. Lazebnik earned a doctoral degree from the University of Illinois at Urbana-Champaign in 2006. After several years as Assistant Professor for the University of North Carolina, she joined the Illinois faculty in 2012.
Hoiem and Lazebnik join previous Sloan Research Fellowship recipients in the Department of Computer Science: Professors Karrie Karahalios (2010), ChengXiang Zhai (2008), Jeff Erickson (1999), Sarita Adve (1998), and Adjunct Professor Yuanyuan Zhou (2007). Fellow Illinois researcher Taylor Hughes, a professor of physics, is also a recipient of a 2013 fellowship.