caa a22 4500 467914354 CHVBK 20180406152857.0 cr unu---uuuuu 170328e20060801xx s 000 0 eng 10.1007/s11263-006-6993-y doi (NATIONALLICENCE)springer-10.1007/s11263-006-6993-y Object Recognition as Many-to-Many Feature Matching [Elektronische Daten] [M. Demirci, Ali Shokoufandeh, Yakov Keselman, Lars Bretzner, Sven Dickinson] Object recognition can be formulated as matching image features to model features. When recognition is exemplar-based, feature correspondence is one-to-one. However, segmentation errors, articulation, scale difference, and within-class deformation can yield image and model features which don't match one-to-one but rather many-to-many. Adopting a graph-based representation of a set of features, we present a matching algorithm that establishes many-to-many correspondences between the nodes of two noisy, vertex-labeled weighted graphs. Our approach reduces the problem of many-to-many matching of weighted graphs to that of many-to-many matching of weighted point sets in a normed vector space. This is accomplished by embedding the initial weighted graphs into a normed vector space with low distortion using a novel embedding technique based on a spherical encoding of graph structure. Many-to-many vector correspondences established by the Earth Mover's Distance framework are mapped back into many-to-many correspondences between graph nodes. Empirical evaluation of the algorithm on an extensive set of recognition trials, including a comparison with two competing graph matching approaches, demonstrates both the robustness and efficacy of the overall approach. Springer Science + Business Media, LLC, 2006 graph matching nationallicence graph embedding nationallicence Earth Mover's Distance (EMD) nationallicence object recognition nationallicence Demirci M. Department of Computer Science, Drexel University, 19104, Philadelphia, PA, USA aut Shokoufandeh Ali Department of Computer Science, Drexel University, 19104, Philadelphia, PA, USA aut Keselman Yakov School of Computer Science, Telecommunications and Information Systems, DePaul University, 60604, Chicago, IL, USA aut Bretzner Lars Department of Numerical Analysis and Computer Science, KTH, Computational Vision and Active Perception Laboratory, Stockholm, Sweden aut Dickinson Sven Department of Computer Science, University of Toronto, M5S 3G4, Toronto, Ontario, Canada aut International Journal of Computer Vision Kluwer Academic Publishers 69/2(2006-08-01), 203-222 0920-5691 69:2<203 2006 69 11263 https://doi.org/10.1007/s11263-006-6993-y text/html Onlinezugriff via DOI 1 research-article jats NATIONALLICENCE 856 40 https://doi.org/10.1007/s11263-006-6993-y text/html Onlinezugriff via DOI NATIONALLICENCE 700 1- Demirci M. Department of Computer Science, Drexel University, 19104, Philadelphia, PA, USA aut NATIONALLICENCE 700 1- Shokoufandeh Ali Department of Computer Science, Drexel University, 19104, Philadelphia, PA, USA aut NATIONALLICENCE 700 1- Keselman Yakov School of Computer Science, Telecommunications and Information Systems, DePaul University, 60604, Chicago, IL, USA aut NATIONALLICENCE 700 1- Bretzner Lars Department of Numerical Analysis and Computer Science, KTH, Computational Vision and Active Perception Laboratory, Stockholm, Sweden aut NATIONALLICENCE 700 1- Dickinson Sven Department of Computer Science, University of Toronto, M5S 3G4, Toronto, Ontario, Canada aut NATIONALLICENCE 773 0- International Journal of Computer Vision Kluwer Academic Publishers 69/2(2006-08-01), 203-222 0920-5691 69:2<203 2006 69 11263 Metadata rights reserved Springer special CC-BY-NC licence nationallicence BK010053 XK010053 XK010000 NATIONALLICENCE NATIONALLICENCE NL-springer