Broad Area Colloquium for Artificial Intelligence,
Geometry, Graphics, Robotics and Vision
Computational Imaging
Shree K. Nayar
Department of Computer Science
Columbia University, New York
Monday, November 19, 2001, 4:15PM
Gates B01 http://robotics.stanford.edu/ba-colloquium/
Abstract
The light field associated with a scene is complex. Conventional still
and video cameras perform a specific sampling of the light field that
is inadequate for many applications in computer vision and computer
graphics. Computational imaging seeks to sample the light field in
unconventional ways to produce new forms of visual
information. Broadly speaking, computational imaging consists of three
components: novel imaging optics, one or more image detectors, and a
computational module. This combination allows one to sample and
process the light field in powerful ways. It provides a general
framework for developing imaging systems that significantly improve
one or more imaging dimensions, such as, field of view, brightness,
color and depth.
In this talk, we present several examples of computational vision
sensors. The first part of the talk focuses on the use of
catadioptrics (lenses and mirrors) for capturing unusually large
fields of view. We describe several methods for obtaining
single-viewpoint and multi-viewpoint images using this approach. The
second part of the talk addresses the problem of acquiring high
dynamic range images using a low dynamic range detector. We present
two approaches for extracting the desired extra bits at each pixel;
one requires multiple images while the other uses just a single
image. Several interactive demonstrations of our results will be
shown. These results have implications for digital imaging, immersive
imaging, image-based rendering, 3D scene modeling, robotics and
advanced interfaces.
About the Speaker
Shree K. Nayar is a Professor in the Department of Computer Science at
Columbia University. He received his PhD degree in Electrical and
Computer Engineering from the Robotics Institute at Carnegie Mellon
University in 1990. He currently heads the Columbia Automated Vision
Environment (CAVE), which is dedicated to the development of advanced
computer vision systems. His research is focused on three areas,
namely: the creation of novel vision sensors, the design of physics
based models for vision, and the development of algorithms for scene
understanding. His work is motivated by applications in the fields of
digital imaging, computer graphics, human-machine interfaces,
robotics, and image understanding. He has received the David Marr
Prize twice (1990 and 1995), the David and Lucile Packard Fellowship
(1992), the National Young Investigator Award (1993) and the Keck
Foundation Award for Excellence in Teaching (1995).
