[robotics-worldwide] Invitation to attend Workshop "Inside Data Association" at RSS08

José Neira jneira at unizar.es
Mon Jun 23 08:42:55 PDT 2008

***                WORKSHOP: "Inside Data Association"               ***
***                ===================================               ***
***                              at the                              ***
***              ROBOTICS: SCIENCE AND SYSTEMS (RSS) 2008            ***
***                                                                  ***
***             ETH Zurich, Switzerland,       28. June 2008         ***
***                                                                  ***


Workshop homepage:

Short Description

Data association is an important problem for many fields. In this
workshop we want to take a deeper look at fundamental methods and
problem analysis of the data association problem. 

Final Program

 9:00 -  9:10  	Opening remarks

 9:10 - 10:00 	Invited talk: Appearance-Only SLAM and the FabMap Algorithm
			Mark Cummins, University of Oxford

10:00 - 10:30 	Bayesian Formulation of Data Association and Markov Chain Monte Carlo Data Association
			Songhwai Oh, University of California

10:30 - 11:00 	coffee break

11:00 - 11:30 	Implicit Data Association from Spectrally Clustered Local Matches
			Edwin Olson, MIT

11:30 - 12:00 	Active Matching
			Margarita Chli and Andrew J. Davison, Imperial College London

12:00 - 12:30 	Learning to Associate with CRF-Matching
			F. Ramos, D. Fox, University of Sydney

12:30 - 13:30 	lunch break

13:30 - 14:00 	An overview of the benchmark data set

14:00 - 14:30 	Vision SLAM needs appearance based methods
			O. Booij, et al., University of Amsterdam

14:30 - 15:00 	Keypoint Design and Evaluation for Global Localization in 2D Lidar Maps
			Michael Bosse and Robert Zlot, CSIRO ICT Centre, Australia

15:00 - 15:30 	A comparison of loop closing techniques in monocular SLAM
			Brian Williams, et al., University of Oxford

15:30 - 16:00 	coffee break

16:00 - 17:00 	Plenary discussion: The data-association problem -- Can we solve it? Can we circumvent it?

17:00 - 17:15 	Closing remarks

Inside Data Association

Correctly associating sets of features is a core problem in many
applications including robot self-localization and mapping (SLAM),
visual tracking, and object recognition. Features usually need to be
extracted from sensor data and extraction itself can be a difficult
application-specific task in its own right. However, there are many
common characteristics of data association problems which are worth

In this workshop we want to take a deeper look at the data-association
problem and we explicitly encourage insightful contributions focussing
on theoretical or empirical analysis, providing novel views on the
problem, or even proposing a candidate for a gold-standard algorithm.

To facilitate evaluation of data association methods, we particularly
encourage using a  SLAM data set which is available from the workshop
homepage (see below). The data set is preprocessed to provide geometric
features and ground-truth is also available. Please take a look at the
video on the workshop homepage.

Best Regards,
  the organizers:

    José Neira
    Udo Frese
    Diedrich Wolter
    Jörg Kurlbaum

Jose Neira

Departamento de Informática e Ingeniería de Sistemas
Centro Politécnico Superior
Universidad de Zaragoza
Edificio Ada Byron
María de Luna 1
E-50018 Zaragoza, Spain
Fax: (34) 976 761 914; Voice: (34) 976 761 947; Office D.1.19
Email: jneira at unizar.es; URL: http://webdiis.unizar.es/~neira

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: message-footer.txt
Url: http://duerer.usc.edu/pipermail/robotics-worldwide/attachments/20080623/0d59a69e/attachment.txt 

More information about the robotics-worldwide mailing list