Matches in SemOpenAlex for { <https://semopenalex.org/work/W2004940028> ?p ?o ?g. }
- W2004940028 endingPage "2352" @default.
- W2004940028 startingPage "2301" @default.
- W2004940028 abstract "We present a hierarchical architecture and learning algorithm for visual recognition and other visual inference tasks such as imagination, reconstruction of occluded images, and expectation-driven segmentation. Using properties of biological vision for guidance, we posit a stochastic generative world model and from it develop a simplified world model (SWM) based on a tractable variational approximation that is designed to enforce sparse coding. Recent developments in computational methods for learning overcomplete representations (Lewicki & Sejnowski, 2000; Teh, Welling, Osindero, & Hinton, 2003) suggest that overcompleteness can be useful for visual tasks, and we use an overcomplete dictionary learning algorithm (Kreutz-Delgado, et al., 2003) as a preprocessing stage to produce accurate, sparse codings of images. Inference is performed by constructing a dynamic multilayer network with feedforward, feedback, and lateral connections, which is trained to approximate the SWM. Learning is done with a variant of the back-propagation-through-time algorithm, which encourages convergence to desired states within a fixed number of iterations. Vision tasks require large networks, and to make learning efficient, we take advantage of the sparsity of each layer to update only a small subset of elements in a large weight matrix at each iteration. Experiments on a set of rotated objects demonstrate various types of visual inference and show that increasing the degree of overcompleteness improves recognition performance in difficult scenes with occluded objects in clutter." @default.
- W2004940028 created "2016-06-24" @default.
- W2004940028 creator A5013678553 @default.
- W2004940028 creator A5042298001 @default.
- W2004940028 date "2007-09-01" @default.
- W2004940028 modified "2023-09-23" @default.
- W2004940028 title "Visual Recognition and Inference Using Dynamic Overcomplete Sparse Learning" @default.
- W2004940028 cites W1488411625 @default.
- W2004940028 cites W1802356529 @default.
- W2004940028 cites W1967011375 @default.
- W2004940028 cites W1968098270 @default.
- W2004940028 cites W1981814724 @default.
- W2004940028 cites W1985813166 @default.
- W2004940028 cites W2000308859 @default.
- W2004940028 cites W2001951139 @default.
- W2004940028 cites W2003766373 @default.
- W2004940028 cites W2020999234 @default.
- W2004940028 cites W2032677875 @default.
- W2004940028 cites W2041828808 @default.
- W2004940028 cites W2042492924 @default.
- W2004940028 cites W2048489451 @default.
- W2004940028 cites W2052916917 @default.
- W2004940028 cites W2057653135 @default.
- W2004940028 cites W2058079620 @default.
- W2004940028 cites W2058670155 @default.
- W2004940028 cites W2067126098 @default.
- W2004940028 cites W2077910828 @default.
- W2004940028 cites W2085927826 @default.
- W2004940028 cites W2086556872 @default.
- W2004940028 cites W2091845343 @default.
- W2004940028 cites W2091987367 @default.
- W2004940028 cites W2093793583 @default.
- W2004940028 cites W2096388912 @default.
- W2004940028 cites W2098580305 @default.
- W2004940028 cites W2100495367 @default.
- W2004940028 cites W2101295242 @default.
- W2004940028 cites W2103212315 @default.
- W2004940028 cites W2105464873 @default.
- W2004940028 cites W2122741244 @default.
- W2004940028 cites W2125663122 @default.
- W2004940028 cites W2128084896 @default.
- W2004940028 cites W2128465753 @default.
- W2004940028 cites W2131689618 @default.
- W2004940028 cites W2136922672 @default.
- W2004940028 cites W2137105523 @default.
- W2004940028 cites W2140499889 @default.
- W2004940028 cites W2142940228 @default.
- W2004940028 cites W2145889472 @default.
- W2004940028 cites W2148848065 @default.
- W2004940028 cites W2149194912 @default.
- W2004940028 cites W2165409036 @default.
- W2004940028 cites W4214649671 @default.
- W2004940028 cites W4249612153 @default.
- W2004940028 cites W4253572625 @default.
- W2004940028 doi "https://doi.org/10.1162/neco.2007.19.9.2301" @default.
- W2004940028 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/17650062" @default.
- W2004940028 hasPublicationYear "2007" @default.
- W2004940028 type Work @default.
- W2004940028 sameAs 2004940028 @default.
- W2004940028 citedByCount "29" @default.
- W2004940028 countsByYear W20049400282012 @default.
- W2004940028 countsByYear W20049400282013 @default.
- W2004940028 countsByYear W20049400282014 @default.
- W2004940028 countsByYear W20049400282015 @default.
- W2004940028 countsByYear W20049400282016 @default.
- W2004940028 countsByYear W20049400282017 @default.
- W2004940028 countsByYear W20049400282018 @default.
- W2004940028 crossrefType "journal-article" @default.
- W2004940028 hasAuthorship W2004940028A5013678553 @default.
- W2004940028 hasAuthorship W2004940028A5042298001 @default.
- W2004940028 hasConcept C119857082 @default.
- W2004940028 hasConcept C153180895 @default.
- W2004940028 hasConcept C154945302 @default.
- W2004940028 hasConcept C177264268 @default.
- W2004940028 hasConcept C199360897 @default.
- W2004940028 hasConcept C2776214188 @default.
- W2004940028 hasConcept C41008148 @default.
- W2004940028 hasConcept C77637269 @default.
- W2004940028 hasConcept C89600930 @default.
- W2004940028 hasConceptScore W2004940028C119857082 @default.
- W2004940028 hasConceptScore W2004940028C153180895 @default.
- W2004940028 hasConceptScore W2004940028C154945302 @default.
- W2004940028 hasConceptScore W2004940028C177264268 @default.
- W2004940028 hasConceptScore W2004940028C199360897 @default.
- W2004940028 hasConceptScore W2004940028C2776214188 @default.
- W2004940028 hasConceptScore W2004940028C41008148 @default.
- W2004940028 hasConceptScore W2004940028C77637269 @default.
- W2004940028 hasConceptScore W2004940028C89600930 @default.
- W2004940028 hasIssue "9" @default.
- W2004940028 hasLocation W20049400281 @default.
- W2004940028 hasLocation W20049400282 @default.
- W2004940028 hasOpenAccess W2004940028 @default.
- W2004940028 hasPrimaryLocation W20049400281 @default.
- W2004940028 hasRelatedWork W2014105136 @default.
- W2004940028 hasRelatedWork W2080679404 @default.
- W2004940028 hasRelatedWork W2510758617 @default.
- W2004940028 hasRelatedWork W2773500201 @default.
- W2004940028 hasRelatedWork W2897195263 @default.