Workshop Program
Workshop Program
Program at a Glance
The workshop will be held on May 10th 2013 in conjunction with ICRA in Karlsruhe, Germany.
The workshop features two tutorials in the morning, two invited talks and two sessions for the six contributed papers. An open discussion concludes the workshop.
The workshop takes place in the room „Kleiner Saal“ in the „Konzerthaus“.
Session 1
08:50 - 09:00
09:00 - 10:00
10:00 - 10:30
Session 2
10:30 - 11:30
11:30 - 11:50
11:50 - 12:10
12:10 - 12:50
12:30 - 12:50
12:50 - 14:00
Session 3
14:00 - 14:20
14:20 - 14:40
14:45 - 15:30
15:30 - 16:00
Session 4
16:00 - 16:45
16:45 - 17:15
Welcome & Introduction
Tutorial 1 - Inference in Factor Graphs (Frank Dellaert)
Coffee Break
Tutorial 2 - Incremental Inference and Applications (Michael Kaess)
Detecting the Correct Graph Structure in Pose Graph SLAM
(Yasir Latif, César Cadena, and José Neira) [ Slides ]
A Robust Graph-based Framework for Building Precise Maps from Laser Range Scans (Marian Himstedt, Sabrina Keil, Sven Hellbach and Hans-Joachim Böhme) [ Slides ]
(Max Pfingsthorn and Andreas Birk)
Dynamic Covariance Scaling for Robust Map Optimization
(Pratik Agarwal, Gian Diego Tipaldi, Luciano Spinello, Cyrill Stachniss, and Wolfram Burgard) [ Slides ]
Lunch Break
L1 Factor Graph SLAM: Going Beyond the L2 Norm
(J.J Casafranca, L.M Paz, and P. Piniés)
Nonparametric Density Estimation for Learning Noise Distributions in Mobile Robotics (David M. Rosen and John J. Leonard)
Invited Talk - 3D SLAM at the Level of Objects (Andrew Davison)
Coffee Break
Invited Talk - Learning Max Mixtures (Edwin Olson)
Workshop conclusions & open discussion
Tutorials
•Tutorial 1 - Inference in Factor Graphs (Frank Dellaert)
•Tutorial 2 - Incremental Inference and Applications (Michael Kaess)
Invited Speakers
•Andrew Davison - 3D SLAM at the Level of Objects
•Edwin Olson - Learning Max Mixtures
In this talk, I'll briefly review the Max Mixtures approach for multi-modal inference and describe our ongoing work in learning how to represent real-world sensor models using max mixtures. In particular, we explore error models for GPS, and show how a simple
learning framework can be used to compute error models that are more robust to the noise of GPS measurements.
Technische Universität Chemnitz, 09107 Chemnitz - Impressum - Copyright © 2013 by TU Chemnitz. Alle Rechte vorbehalten.