Michael Kaess @ Georgia Tech

I am a Ph.D. student in the College of Computing at the Georgia Institute of Technology. I am interested in 3D perception with a focus on mobile robotics applications. My advisor is Frank Dellaert.

My research interests span the areas of computer vision, probabilistic methods, large-scale optimization and autonomous robots. My thesis work focuses on simultaneous localization and mapping (SLAM), which is the problem of mapping a previously unknown environment while at the same time using this map for localization. In addition to exploring efficient solutions to the continuous estimation problem, I am working on probabilistic solutions for the discrete data association problem.

Selected Publications:

“iSAM: Fast Incremental Smoothing and Mapping with Efficient Data Association” by M. Kaess, A. Ranganathan, and F. Dellaert. In IEEE Intl. Conf. on Robotics and Automation, ICRA, (Rome, Italy), Apr. 2007, pp. 1670-1677. Details. Download: PDF.

“Fast Incremental Square Root Information Smoothing” by M. Kaess, A. Ranganathan, and F. Dellaert. In Intl. Joint Conf. on Artificial Intelligence, IJCAI, (Hyderabad, India), Jan. 2007, pp. 2129-2134. Oral presentation acceptance ratio 15.7% (212 of 1353). Details. Download: PDF.

“Square Root SAM: Simultaneous Localization and Mapping via Square Root Information Smoothing” by F. Dellaert and M. Kaess. Intl. J. of Robotics Research, vol. 25, no. 12, Dec. 2006, pp. 1181-1204. Details. Download: PDF.

“MCMC-based Multiview Reconstruction of Piecewise Smooth Subdivision Curves with a Variable Number of Control Points” by M. Kaess, R. Zboinski, and F. Dellaert. In Eur. Conf. on Computer Vision, ECCV, (Prague, Czech Republic), May 2004, pp. 329-341. Acceptance ratio 34.2% (190 of 555). Details. Download: PDF.

all publications...

Current Research:

	iSAM: Incremental Smoothing and Mapping My thesis work: A fast incremental solution to the full SLAM problem, including access to marginal covariances for real-time data association.
	Visual Odometry Visual odometry provides both improved localization and useful input for visual SLAM.
	Visual SLAM Localization and mapping using stereo cameras as well as a multi-camera rig.

Previous Work:

	Loop Closing A probabilistic approach to loop closing, with some results on laser data.
	3D Boundary Reconstruction from Multiple Views Multiple View Rao-Blackwellized Reversible-Jump Markov Chain Monte Carlo Tagged Subdivision Curve Fitting - the paper in one sentence ;)
	AAAI Search and Rescue Competition 2002 We got third place and received the technical award for the best mapping!
	3D Models from Laser Range Data Cool pictures! And software to download...
	Global Robot Localization Monte Carlo Localization based on omnidirectional images with PCA-based dimensionality reduction.
	Learning for Behavior-based Robots Case Based Reasoning for Behavioral Parameter Selection, as part of the DARPA MARS program.

Michael Kaess
Center for Robotics and Intelligent Machines, Georgia Tech	GT CoC IC GVU RIM@GT BORG