Matches in SemOpenAlex for { <https://semopenalex.org/work/W3178113344> ?p ?o ?g. }
- W3178113344 endingPage "1281" @default.
- W3178113344 startingPage "1253" @default.
- W3178113344 abstract "Dual decomposition approaches in nonconvex optimization may suffer from a duality gap. This poses a challenge when applying them directly to nonconvex problems such as MAP-inference in a Markov random field with continuous state spaces. To eliminate such gaps, this paper considers a reformulation of the original nonconvex task in the space of measures. This infinite-dimensional reformulation is then approximated by a semi-infinite one, which is obtained via a piecewise polynomial discretization in the dual. We provide a geometric intuition behind the primal problem induced by the dual discretization and draw connections to optimization over moment spaces. In contrast to existing discretizations which suffer from a grid bias, we show that a piecewise polynomial discretization better preserves the continuous nature of our problem. Invoking results from optimal transport theory and convex algebraic geometry we reduce the semi-infinite program to a finite one and provide a practical implementation based on semidefinite programming. We show, experimentally and in theory, that the approach successfully reduces the duality gap. To showcase the scalability of our approach, we apply it to the stereo matching problem between two images." @default.
- W3178113344 created "2021-07-19" @default.
- W3178113344 creator A5006959568 @default.
- W3178113344 creator A5008471432 @default.
- W3178113344 creator A5051884518 @default.
- W3178113344 creator A5077620562 @default.
- W3178113344 creator A5087710605 @default.
- W3178113344 date "2022-08-11" @default.
- W3178113344 modified "2023-10-16" @default.
- W3178113344 title "Lifting the Convex Conjugate in Lagrangian Relaxations: A Tractable Approach for Continuous Markov Random Fields" @default.
- W3178113344 cites W1003865189 @default.
- W3178113344 cites W1535868076 @default.
- W3178113344 cites W1585160083 @default.
- W3178113344 cites W1597928342 @default.
- W3178113344 cites W173642128 @default.
- W3178113344 cites W1967344706 @default.
- W3178113344 cites W1983127118 @default.
- W3178113344 cites W1993336109 @default.
- W3178113344 cites W2032938251 @default.
- W3178113344 cites W2037091587 @default.
- W3178113344 cites W2038887391 @default.
- W3178113344 cites W2077011847 @default.
- W3178113344 cites W2078762555 @default.
- W3178113344 cites W2085352155 @default.
- W3178113344 cites W2092663520 @default.
- W3178113344 cites W2108457948 @default.
- W3178113344 cites W2123457441 @default.
- W3178113344 cites W2125668872 @default.
- W3178113344 cites W2135414191 @default.
- W3178113344 cites W2138322123 @default.
- W3178113344 cites W2146807782 @default.
- W3178113344 cites W2166798101 @default.
- W3178113344 cites W2550994434 @default.
- W3178113344 cites W2593062745 @default.
- W3178113344 cites W2750185643 @default.
- W3178113344 cites W2895609902 @default.
- W3178113344 cites W2963305757 @default.
- W3178113344 cites W2963601861 @default.
- W3178113344 cites W2963719189 @default.
- W3178113344 cites W2964142180 @default.
- W3178113344 cites W2967043085 @default.
- W3178113344 cites W3100717425 @default.
- W3178113344 cites W3125749597 @default.
- W3178113344 cites W3156884384 @default.
- W3178113344 cites W4205353090 @default.
- W3178113344 cites W4240422548 @default.
- W3178113344 cites W4249513058 @default.
- W3178113344 cites W4249667877 @default.
- W3178113344 cites W4254742099 @default.
- W3178113344 cites W4255442211 @default.
- W3178113344 cites W4255839052 @default.
- W3178113344 cites W4293052541 @default.
- W3178113344 cites W63091017 @default.
- W3178113344 cites W653291882 @default.
- W3178113344 doi "https://doi.org/10.1137/21m1433241" @default.
- W3178113344 hasPublicationYear "2022" @default.
- W3178113344 type Work @default.
- W3178113344 sameAs 3178113344 @default.
- W3178113344 citedByCount "3" @default.
- W3178113344 countsByYear W31781133442021 @default.
- W3178113344 countsByYear W31781133442022 @default.
- W3178113344 crossrefType "journal-article" @default.
- W3178113344 hasAuthorship W3178113344A5006959568 @default.
- W3178113344 hasAuthorship W3178113344A5008471432 @default.
- W3178113344 hasAuthorship W3178113344A5051884518 @default.
- W3178113344 hasAuthorship W3178113344A5077620562 @default.
- W3178113344 hasAuthorship W3178113344A5087710605 @default.
- W3178113344 hasBestOaLocation W31781133442 @default.
- W3178113344 hasConcept C101901036 @default.
- W3178113344 hasConcept C112680207 @default.
- W3178113344 hasConcept C118615104 @default.
- W3178113344 hasConcept C126255220 @default.
- W3178113344 hasConcept C134306372 @default.
- W3178113344 hasConcept C137836250 @default.
- W3178113344 hasConcept C157972887 @default.
- W3178113344 hasConcept C163863214 @default.
- W3178113344 hasConcept C164660894 @default.
- W3178113344 hasConcept C2524010 @default.
- W3178113344 hasConcept C2778023678 @default.
- W3178113344 hasConcept C28826006 @default.
- W3178113344 hasConcept C33923547 @default.
- W3178113344 hasConcept C5274546 @default.
- W3178113344 hasConcept C73000952 @default.
- W3178113344 hasConceptScore W3178113344C101901036 @default.
- W3178113344 hasConceptScore W3178113344C112680207 @default.
- W3178113344 hasConceptScore W3178113344C118615104 @default.
- W3178113344 hasConceptScore W3178113344C126255220 @default.
- W3178113344 hasConceptScore W3178113344C134306372 @default.
- W3178113344 hasConceptScore W3178113344C137836250 @default.
- W3178113344 hasConceptScore W3178113344C157972887 @default.
- W3178113344 hasConceptScore W3178113344C163863214 @default.
- W3178113344 hasConceptScore W3178113344C164660894 @default.
- W3178113344 hasConceptScore W3178113344C2524010 @default.
- W3178113344 hasConceptScore W3178113344C2778023678 @default.
- W3178113344 hasConceptScore W3178113344C28826006 @default.
- W3178113344 hasConceptScore W3178113344C33923547 @default.
- W3178113344 hasConceptScore W3178113344C5274546 @default.
- W3178113344 hasConceptScore W3178113344C73000952 @default.