FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2890886265> ?p ?o ?g. }

• W2890886265 abstract "This paper addresses the problem of localizing facial landmarks with deformable face models using cascaded regression strategies. Recently, these methods have become quite popular, standing out as simple and efficient approaches to optimize nonlinear objective functions. In this paper, we target the well-known Lucas and Kanade (LK) image alignment formulation and introduce the Simultaneous Cascaded Regression (SCR) technique, which can be considered as a cascaded regression extension of the Simultaneous Forwards Additive / Inverse Composition approaches. In contrast to previous LK techniques (Newton based optimizations) which require to recompute Jacobian and Hessians matrices at each iteration, our approach learns (offline) a sequence of descent directions, effectively behaving as averaged steepest descent matrices. Under this revised technique, we propose a part-based generative model (with a linear warp function), that accounts with the underlying shape and appearance structure embedded into regression process itself. Our method is validated on a number of experiments on several datasets (LFPW, LFW, HELEN, 300W), demonstrating a noticeable gain in accuracy/fitting performance when compared with other face alignment solutions." @default.
• W2890886265 created "2018-09-27" @default.
• W2890886265 creator A5024233672 @default.
• W2890886265 creator A5045422006 @default.
• W2890886265 date "2018-10-01" @default.
• W2890886265 modified "2023-09-26" @default.
• W2890886265 title "Simultaneous Cascaded Regression" @default.
• W2890886265 cites W1527240141 @default.
• W2890886265 cites W1604643028 @default.
• W2890886265 cites W1682276745 @default.
• W2890886265 cites W1782590233 @default.
• W2890886265 cites W1796263212 @default.
• W2890886265 cites W1915668717 @default.
• W2890886265 cites W1925384340 @default.
• W2890886265 cites W1964846093 @default.
• W2890886265 cites W1998294030 @default.
• W2890886265 cites W2018718299 @default.
• W2890886265 cites W2032558548 @default.
• W2890886265 cites W2033961931 @default.
• W2890886265 cites W2035379092 @default.
• W2890886265 cites W2038044681 @default.
• W2890886265 cites W2045965808 @default.
• W2890886265 cites W2047508432 @default.
• W2890886265 cites W2058961190 @default.
• W2890886265 cites W2068709348 @default.
• W2890886265 cites W2076017598 @default.
• W2890886265 cites W2082308025 @default.
• W2890886265 cites W2087681821 @default.
• W2890886265 cites W2096503284 @default.
• W2890886265 cites W2109434346 @default.
• W2890886265 cites W2111372597 @default.
• W2890886265 cites W2117955708 @default.
• W2890886265 cites W2118877769 @default.
• W2890886265 cites W2128958079 @default.
• W2890886265 cites W2130563197 @default.
• W2890886265 cites W2133401711 @default.
• W2890886265 cites W2138406903 @default.
• W2890886265 cites W2152826865 @default.
• W2890886265 cites W2156584657 @default.
• W2890886265 cites W2161969291 @default.
• W2890886265 cites W2163998463 @default.
• W2890886265 cites W2171262059 @default.
• W2890886265 cites W2284800790 @default.
• W2890886265 cites W2294304715 @default.
• W2890886265 cites W2396190289 @default.
• W2890886265 cites W2523270796 @default.
• W2890886265 cites W3097096317 @default.
• W2890886265 doi "https://doi.org/10.1109/icip.2018.8451313" @default.
• W2890886265 hasPublicationYear "2018" @default.
• W2890886265 type Work @default.
• W2890886265 sameAs 2890886265 @default.
• W2890886265 citedByCount "1" @default.
• W2890886265 countsByYear W28908862652020 @default.
• W2890886265 crossrefType "proceedings-article" @default.
• W2890886265 hasAuthorship W2890886265A5024233672 @default.
• W2890886265 hasAuthorship W2890886265A5045422006 @default.
• W2890886265 hasConcept C105795698 @default.
• W2890886265 hasConcept C111472728 @default.
• W2890886265 hasConcept C11413529 @default.
• W2890886265 hasConcept C115961682 @default.
• W2890886265 hasConcept C119857082 @default.
• W2890886265 hasConcept C121332964 @default.
• W2890886265 hasConcept C138885662 @default.
• W2890886265 hasConcept C14036430 @default.
• W2890886265 hasConcept C144024400 @default.
• W2890886265 hasConcept C152877465 @default.
• W2890886265 hasConcept C153180895 @default.
• W2890886265 hasConcept C153258448 @default.
• W2890886265 hasConcept C154945302 @default.
• W2890886265 hasConcept C158622935 @default.
• W2890886265 hasConcept C200331156 @default.
• W2890886265 hasConcept C207467116 @default.
• W2890886265 hasConcept C2524010 @default.
• W2890886265 hasConcept C2779304628 @default.
• W2890886265 hasConcept C2780586882 @default.
• W2890886265 hasConcept C28826006 @default.
• W2890886265 hasConcept C33923547 @default.
• W2890886265 hasConcept C36289849 @default.
• W2890886265 hasConcept C41008148 @default.
• W2890886265 hasConcept C46889948 @default.
• W2890886265 hasConcept C50644808 @default.
• W2890886265 hasConcept C62520636 @default.
• W2890886265 hasConcept C78458016 @default.
• W2890886265 hasConcept C83546350 @default.
• W2890886265 hasConcept C86803240 @default.
• W2890886265 hasConceptScore W2890886265C105795698 @default.
• W2890886265 hasConceptScore W2890886265C111472728 @default.
• W2890886265 hasConceptScore W2890886265C11413529 @default.
• W2890886265 hasConceptScore W2890886265C115961682 @default.
• W2890886265 hasConceptScore W2890886265C119857082 @default.
• W2890886265 hasConceptScore W2890886265C121332964 @default.
• W2890886265 hasConceptScore W2890886265C138885662 @default.
• W2890886265 hasConceptScore W2890886265C14036430 @default.
• W2890886265 hasConceptScore W2890886265C144024400 @default.
• W2890886265 hasConceptScore W2890886265C152877465 @default.
• W2890886265 hasConceptScore W2890886265C153180895 @default.
• W2890886265 hasConceptScore W2890886265C153258448 @default.
• W2890886265 hasConceptScore W2890886265C154945302 @default.
• W2890886265 hasConceptScore W2890886265C158622935 @default.
• W2890886265 hasConceptScore W2890886265C200331156 @default.

• next