A Sample of Monte Carlo Methods in Robotics and Vision

Frank Dellaert 
College of Computing, Georgia Institute of Technology 

Approximate inference by sampling from an appropriately constructed posterior 
has recently seen a dramatic increase in popularity in both the robotics and 
computer vision community. I will describe a number of approaches in which we 
have used Sequential Monte Carlo methods and Markov chain Monte Carlo sampling 
to solve a variety of difficult and challenging inference problems. Very 
recently, we have also used sampling over variable dimension state spaces to 
perform automatic model selection. I will present two examples of this, one in 
the domain of computer vision, the other in mobile robotics. In both cases Rao-
Blackwellization was used to integrate out the variable dimension-part of the 
state space, and hence the sampling was done purely over the (combinatorially 
large) space of different models.

Most of the work I will describe was presented at ECCV or will be presented at 
CVPR and IROS later this year, and was done in collaboration with my students 
Zia Khan, Michael Kaess, and Ananth Ranganathan. The electronic versions of the 
papers can be found at 
http://www.cc.gatech.edu/~dellaert/html/2004_highlights.html

Frank Dellaert is an Assistant Professor at the College of Computing, Georgia 
Institute of Technology. He graduated in 2001 with a Ph.D. from Carnegie Mellon 
University. His research focuses on probabilistic methods in Robotics and 
Computer Vision: he has applied Markov chain Monte Carlo sampling methodologies 
in a variety of novel settings, most notably to address the correspondence 
problem in computer vision. Before that, with Dieter Fox and Sebastian Thrun, he 
has introduced the Monte Carlo localization method for estimating and tracking 
the pose of robots, which is now a standard and popular tool in mobile robotics. 
Since coming to Georgia Tech, he explored the theme of probabilistic, model-
based reasoning paired with randomized approximation methods in three main 
research areas: Advanced sequential Monte Carlo methods, Spatio-Temporal 
Reconstruction from Images, and Simultaneous Localization and Mapping. His 
homepage can be found at http://www.cc.gatech.edu/~dellaert