[CPL Seminar]
[Schedule]
[Jan 9]
[Jan 16]
[Jan 23]
[Jan 30]
[Feb 6]
[Feb 20]
[Feb 25]
[Mar 7 Shum]
[Mar 7 Szeliski]
[Mar 13]
[Mar 20]
[Mar 27]
[April 3]
[April 10]
[April 17]
[April 24]

Frank Dellaert
College of Computing
Georgia Institute of Technology

Vision and Sonar-based Monte Carlo Localization for Mobile Robots
Online Slides Handout

In this talk I will present material from ICRA and CVPR papers on how the condensation algorithm is ideally suited for mobile robot localization in the case of noisy or poor sensors. It also nicely solves the global localization problem. I'll give an in depth treatment of the Condensation algorithm with lots of graphics, and contrast it with other Bayesian filtering paradigms. I'll show some nice results with both sonar and vision-based sensors, in the context of the museum tourguide robots Rhino (in Bonn) and Minerva (in Washington DC).

The work I'll present was done in collaboration with Dieter Fox, Wolfram Burgard, and Sebastian Thrun, and was done at Carnegie Mellon University.