Show simple item record

dc.contributor.authorDellaert, Frank
dc.contributor.authorCarlson, Justin
dc.contributor.authorIla, Viorela
dc.contributor.authorNi, Kai
dc.contributor.authorThorpe, Charles E.
dc.date.accessioned2011-03-29T21:20:36Z
dc.date.available2011-03-29T21:20:36Z
dc.date.issued2010
dc.identifier.citationDellaert, F., Carlson, J., Ila, V., Ni, K., & Thorpe, C.E. (2010). "Subgraph-Preconditioned Conjugate Gradients for Large Scale SLAM.” Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2010), 18-22 October 2010, 2566-2571.en_US
dc.identifier.issn2153-0858
dc.identifier.urihttp://hdl.handle.net/1853/38327
dc.description©2010 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works.en_US
dc.descriptionPresented at the 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 18-22 October 2010, Taipei, Taiwan.
dc.descriptionDOI: 10.1109/IROS.2010.5650422
dc.description.abstractIn this paper we propose an efficient preconditioned conjugate gradients (PCG) approach to solving large-scale SLAM problems. While direct methods, popular in the literature, exhibit quadratic convergence and can be quite efficient for sparse problems, they typically require a lot of storage as well as efficient elimination orderings to be found. In contrast, iterative optimization methods only require access to the gradient and have a small memory footprint, but can suffer from poor convergence. Our new method, subgraph preconditioning, is obtained by re-interpreting the method of conjugate gradients in terms of the graphical model representation of the SLAM problem. The main idea is to combine the advantages of direct and iterative methods, by identifying a sub-problem that can be easily solved using direct methods, and solving for the remaining part using PCG. The easy sub-problems correspond to a spanning tree, a planar subgraph, or any other substructure that can be efficiently solved. As such, our approach provides new insights into the performance of state of the art iterative SLAM methods based on re-parameterized stochastic gradient descent. The efficiency of our new algorithm is illustrated on large datasets, both simulated and real.en_US
dc.language.isoen_USen_US
dc.publisherGeorgia Institute of Technologyen_US
dc.subjectPreconditioned conjugate gradientsen_US
dc.subjectSimultaneous localization and mappingen_US
dc.subjectSubgraph preconditioningen_US
dc.titleSubgraph-preconditioned Conjugate Gradients for Large Scale SLAMen_US
dc.typePost-printen_US
dc.typeProceedings
dc.contributor.corporatenameGeorgia Institute of Technology. Center for Robotics and Intelligent Machines
dc.contributor.corporatenameGeorgia Institute of Technology. College of Computing
dc.contributor.corporatenameInstitut de Robòtica i Informàtica Industrial
dc.publisher.originalInstitute of Electrical and Electronics Engineers


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record