FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386075809> ?p ?o ?g. }

• W4386075809 abstract "Outlier-robust estimation involves estimating some parameters (e.g., 3D rotations) from data samples in the presence of outliers, and is typically formulated as a non-convex and non-smooth problem. For this problem, the classical method called iteratively reweighted least-squares (IRLS) and its variants have shown impressive performance. This paper makes several contributions towards understanding why these algorithms work so well. First, we incorporate majorization and graduated non-convexity (GNC) into the IRLS framework and prove that the resulting IRLS variant is a convergent method for outlier-robust estimation. Moreover, in the robust regression context with a constant fraction of outliers, we prove this IRLS variant converges to the ground truth at a global linear and local quadratic rate for a random Gaussian feature matrix with high probability. Experiments corroborate our theory and show that the proposed IRLS variant converges within 5–10 iterations for typical problem instances of outlier-robust estimation, while state-of-the-art methods need at least 30 iterations. A basic implementation of our method is provided: https://github.com/liangzu/IRLS-CVPR2023" @default.
• W4386075809 created "2023-08-23" @default.
• W4386075809 creator A5011256828 @default.
• W4386075809 creator A5042095059 @default.
• W4386075809 creator A5072149973 @default.
• W4386075809 date "2023-06-01" @default.
• W4386075809 modified "2023-10-09" @default.
• W4386075809 title "On the Convergence of IRLS and Its Variants in Outlier-Robust Estimation" @default.
• W4386075809 cites W1966661320 @default.
• W4386075809 cites W1967023569 @default.
• W4386075809 cites W1977899734 @default.
• W4386075809 cites W1982029035 @default.
• W4386075809 cites W1988874269 @default.
• W4386075809 cites W1993803475 @default.
• W4386075809 cites W2004544971 @default.
• W4386075809 cites W2018816711 @default.
• W4386075809 cites W2026155476 @default.
• W4386075809 cites W2046033161 @default.
• W4386075809 cites W2050968963 @default.
• W4386075809 cites W2053608531 @default.
• W4386075809 cites W2085261163 @default.
• W4386075809 cites W2087714402 @default.
• W4386075809 cites W2088616581 @default.
• W4386075809 cites W2094694151 @default.
• W4386075809 cites W2094900096 @default.
• W4386075809 cites W2096758544 @default.
• W4386075809 cites W2107105904 @default.
• W4386075809 cites W2109181422 @default.
• W4386075809 cites W2119883478 @default.
• W4386075809 cites W2159719921 @default.
• W4386075809 cites W2161765392 @default.
• W4386075809 cites W2168745297 @default.
• W4386075809 cites W2409932664 @default.
• W4386075809 cites W2508393166 @default.
• W4386075809 cites W2519911873 @default.
• W4386075809 cites W2547811925 @default.
• W4386075809 cites W2605778869 @default.
• W4386075809 cites W2749287221 @default.
• W4386075809 cites W2769591697 @default.
• W4386075809 cites W2941355095 @default.
• W4386075809 cites W2973258311 @default.
• W4386075809 cites W3034373437 @default.
• W4386075809 cites W3035025229 @default.
• W4386075809 cites W3111598592 @default.
• W4386075809 cites W3124680869 @default.
• W4386075809 cites W3126158429 @default.
• W4386075809 cites W3131624727 @default.
• W4386075809 cites W3214667715 @default.
• W4386075809 cites W4206103517 @default.
• W4386075809 cites W4206335957 @default.
• W4386075809 cites W4220695485 @default.
• W4386075809 cites W4230266908 @default.
• W4386075809 cites W4244769743 @default.
• W4386075809 cites W4250589301 @default.
• W4386075809 cites W4282913827 @default.
• W4386075809 cites W4292589475 @default.
• W4386075809 cites W4312934854 @default.
• W4386075809 cites W4386071591 @default.
• W4386075809 doi "https://doi.org/10.1109/cvpr52729.2023.01708" @default.
• W4386075809 hasPublicationYear "2023" @default.
• W4386075809 type Work @default.
• W4386075809 citedByCount "1" @default.
• W4386075809 countsByYear W43860758092023 @default.
• W4386075809 crossrefType "proceedings-article" @default.
• W4386075809 hasAuthorship W4386075809A5011256828 @default.
• W4386075809 hasAuthorship W4386075809A5042095059 @default.
• W4386075809 hasAuthorship W4386075809A5072149973 @default.
• W4386075809 hasConcept C106159729 @default.
• W4386075809 hasConcept C11413529 @default.
• W4386075809 hasConcept C126090379 @default.
• W4386075809 hasConcept C126255220 @default.
• W4386075809 hasConcept C151730666 @default.
• W4386075809 hasConcept C153180895 @default.
• W4386075809 hasConcept C154945302 @default.
• W4386075809 hasConcept C162324750 @default.
• W4386075809 hasConcept C167928553 @default.
• W4386075809 hasConcept C2777303404 @default.
• W4386075809 hasConcept C2779343474 @default.
• W4386075809 hasConcept C33923547 @default.
• W4386075809 hasConcept C41008148 @default.
• W4386075809 hasConcept C45923927 @default.
• W4386075809 hasConcept C50522688 @default.
• W4386075809 hasConcept C70259352 @default.
• W4386075809 hasConcept C72134830 @default.
• W4386075809 hasConcept C739882 @default.
• W4386075809 hasConcept C79337645 @default.
• W4386075809 hasConcept C86803240 @default.
• W4386075809 hasConceptScore W4386075809C106159729 @default.
• W4386075809 hasConceptScore W4386075809C11413529 @default.
• W4386075809 hasConceptScore W4386075809C126090379 @default.
• W4386075809 hasConceptScore W4386075809C126255220 @default.
• W4386075809 hasConceptScore W4386075809C151730666 @default.
• W4386075809 hasConceptScore W4386075809C153180895 @default.
• W4386075809 hasConceptScore W4386075809C154945302 @default.
• W4386075809 hasConceptScore W4386075809C162324750 @default.
• W4386075809 hasConceptScore W4386075809C167928553 @default.
• W4386075809 hasConceptScore W4386075809C2777303404 @default.
• W4386075809 hasConceptScore W4386075809C2779343474 @default.
• W4386075809 hasConceptScore W4386075809C33923547 @default.
• W4386075809 hasConceptScore W4386075809C41008148 @default.

• next