Matches in SemOpenAlex for { <https://semopenalex.org/work/W2806018063> ?p ?o ?g. }
Showing items 1 to 67 of
67
with 100 items per page.
- W2806018063 abstract "Parsimonious representations are ubiquitous in modeling and processing information. Motivated by the recent Multi-Layer Convolutional Sparse Coding (ML-CSC) model, we herein generalize the traditional Basis Pursuit problem to a multi-layer setting, introducing similar sparse enforcing penalties at different representation layers in a symbiotic relation between synthesis and analysis sparse priors. We explore different iterative methods to solve this new problem in practice, and we propose a new Multi-Layer Iterative Soft Thresholding Algorithm (ML-ISTA), as well as a fast version (ML-FISTA). We show that these nested first order algorithms converge, in the sense that the function value of near-fixed points can get arbitrarily close to the solution of the original problem. We further show how these algorithms effectively implement particular recurrent convolutional neural networks (CNNs) that generalize feed-forward ones without introducing any parameters. We present and analyze different architectures resulting unfolding the iterations of the proposed pursuit algorithms, including a new Learned ML-ISTA, providing a principled way to construct deep recurrent CNNs. Unlike other similar constructions, these architectures unfold a global pursuit holistically for the entire network. We demonstrate the emerging constructions in a supervised learning setting, consistently improving the performance of classical CNNs while maintaining the number of parameters constant." @default.
- W2806018063 created "2018-06-13" @default.
- W2806018063 creator A5020279598 @default.
- W2806018063 creator A5043744017 @default.
- W2806018063 creator A5069363394 @default.
- W2806018063 creator A5086776097 @default.
- W2806018063 date "2018-06-02" @default.
- W2806018063 modified "2023-10-03" @default.
- W2806018063 title "On Multi-Layer Basis Pursuit, Efficient Algorithms and Convolutional Neural Networks" @default.
- W2806018063 hasPublicationYear "2018" @default.
- W2806018063 type Work @default.
- W2806018063 sameAs 2806018063 @default.
- W2806018063 citedByCount "0" @default.
- W2806018063 crossrefType "posted-content" @default.
- W2806018063 hasAuthorship W2806018063A5020279598 @default.
- W2806018063 hasAuthorship W2806018063A5043744017 @default.
- W2806018063 hasAuthorship W2806018063A5069363394 @default.
- W2806018063 hasAuthorship W2806018063A5086776097 @default.
- W2806018063 hasConcept C108583219 @default.
- W2806018063 hasConcept C11413529 @default.
- W2806018063 hasConcept C154945302 @default.
- W2806018063 hasConcept C17744445 @default.
- W2806018063 hasConcept C199360897 @default.
- W2806018063 hasConcept C199539241 @default.
- W2806018063 hasConcept C2776359362 @default.
- W2806018063 hasConcept C2780801425 @default.
- W2806018063 hasConcept C41008148 @default.
- W2806018063 hasConcept C50644808 @default.
- W2806018063 hasConcept C81363708 @default.
- W2806018063 hasConcept C94625758 @default.
- W2806018063 hasConceptScore W2806018063C108583219 @default.
- W2806018063 hasConceptScore W2806018063C11413529 @default.
- W2806018063 hasConceptScore W2806018063C154945302 @default.
- W2806018063 hasConceptScore W2806018063C17744445 @default.
- W2806018063 hasConceptScore W2806018063C199360897 @default.
- W2806018063 hasConceptScore W2806018063C199539241 @default.
- W2806018063 hasConceptScore W2806018063C2776359362 @default.
- W2806018063 hasConceptScore W2806018063C2780801425 @default.
- W2806018063 hasConceptScore W2806018063C41008148 @default.
- W2806018063 hasConceptScore W2806018063C50644808 @default.
- W2806018063 hasConceptScore W2806018063C81363708 @default.
- W2806018063 hasConceptScore W2806018063C94625758 @default.
- W2806018063 hasOpenAccess W2806018063 @default.
- W2806018063 hasRelatedWork W2552905758 @default.
- W2806018063 hasRelatedWork W2740159961 @default.
- W2806018063 hasRelatedWork W2767880545 @default.
- W2806018063 hasRelatedWork W2781309524 @default.
- W2806018063 hasRelatedWork W2785619321 @default.
- W2806018063 hasRelatedWork W2885059062 @default.
- W2806018063 hasRelatedWork W2950529782 @default.
- W2806018063 hasRelatedWork W2951095867 @default.
- W2806018063 hasRelatedWork W2963217415 @default.
- W2806018063 hasRelatedWork W2979374808 @default.
- W2806018063 hasRelatedWork W2991593608 @default.
- W2806018063 hasRelatedWork W3015670122 @default.
- W2806018063 hasRelatedWork W3016031910 @default.
- W2806018063 hasRelatedWork W3022566213 @default.
- W2806018063 hasRelatedWork W3045430853 @default.
- W2806018063 hasRelatedWork W3108060352 @default.
- W2806018063 hasRelatedWork W3131466454 @default.
- W2806018063 hasRelatedWork W3160822690 @default.
- W2806018063 hasRelatedWork W3161221866 @default.
- W2806018063 hasRelatedWork W3195106575 @default.
- W2806018063 isParatext "false" @default.
- W2806018063 isRetracted "false" @default.
- W2806018063 magId "2806018063" @default.
- W2806018063 workType "article" @default.