Jean Ponce, University of Illinois - Urbana-Champaign, ponce@cs.uiuc.edu
Presented under the auspices of the DIMACS Special Year on Large Scale Discrete Optimization.
This workshop will address the fundamental mathematical and
computational problems underlying computer vision and robotics. A set
of talks from major researchers in the field will help give a
perspective on the leading issues today and the likely directions for
future progress. Points of contact with computer science theory and
algorithms, probabilistic inference, and other areas of applied
mathematics in general will be stressed.
Biological vision is used to guide a number of different aspects of
behavior--guiding locomotion, controlling manipulation and recognition
of objects. Computer vision seeks to develop models and algorithms for
these competences, drawing heavily on geometry, probability and
statistics and computation. Likewise, robotics is concerned with
controlling mechanisms such as manipulator arms, grippers and mobile
platforms using sensory input as well as a priori knowledge about the
geometry and physics of these mechanisms and their environment. This
discipline is at the boundary between fields as diverse as mathematics,
mechanics, control theory and computer science.
The intended audience for the workshop is formed of students and
practitioners of applied mathematics and computer science. The
invited speakers will present various facets of vision and robotics,
including object recognition, image segmentation, motion analysis,
mobile robot navigation, visual servoing, geometric motion planning
and object manipulation.