Show simple item record

dc.contributor.authorDellaert, Frank
dc.contributor.authorSeitz, Steven M.
dc.contributor.authorThorpe, Charles E.
dc.contributor.authorThrun, Sebastian
dc.date.accessioned2011-04-22T01:51:56Z
dc.date.available2011-04-22T01:51:56Z
dc.date.issued2003
dc.identifier.citationDellaert, F., Seitz, S. M., Thorpe, C. E., & Thrun, S. (2003). “EM, MCMC, and Chain Flipping for Structure from Motion with Unknown Correspondence”. Machine Learning, Vol. 50, no. 1-2, (January 01, 2003), pp. 45-71.en_US
dc.identifier.issn0885-6125
dc.identifier.urihttp://hdl.handle.net/1853/38687
dc.description©2002 Kluwer Academic Publishers. ©2004 Springer-Verlag Berlin Heidelberg. The original publication is available at www.springerlink.comen_US
dc.descriptionBertelsmannSpringer and Kluwer Academic Publishers merged in 2004.
dc.descriptionDOI: 10.1023/A:1020245811187
dc.description.abstractLearning spatial models from sensor data raises the challenging data association problem of relating model parameters to individual measurements. This paper proposes an EM-based algorithm, which solves the model learning and the data association problem in parallel. The algorithm is developed in the context of the the structure from motion problem, which is the problem of estimating a 3D scene model from a collection of image data. To accommodate the spatial constraints in this domain, we compute virtual measurements as sufficient statistics to be used in the M-step. We develop an efficient Markov chain Monte Carlo sampling method called chain flipping, to calculate these statistics in the E-step. Experimental results show that we can solve hard data association problems when learning models of 3D scenes, and that we can do so efficiently. We conjecture that this approach can be applied to a broad range of model learning problems from sensor data, such as the robot mapping problem.en_US
dc.language.isoen_USen_US
dc.publisherGeorgia Institute of Technologyen_US
dc.subjectComputer visionen_US
dc.subjectCorrespondence Problemen_US
dc.subjectData associationen_US
dc.subjectEfficient samplingen_US
dc.subjectExpectation-maximizationen_US
dc.subjectMarkov chain Monte Carloen_US
dc.subjectStructure from motionen_US
dc.titleEM, MCMC, and Chain Flipping for Structure from Motion with Unknown Correspondenceen_US
dc.typeArticleen_US
dc.contributor.corporatenameGeorgia Institute of Technology. Center for Robotics and Intelligent Machines
dc.contributor.corporatenameGeorgia Institute of Technology. College of Computing
dc.contributor.corporatenameCarnegie-Mellon University. School of Computer Science
dc.contributor.corporatenameUniversity of Washington. Dept. of Computer Science and Engineering
dc.publisher.originalKluwer Academic Publishers


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record