Matches in SemOpenAlex for { <https://semopenalex.org/work/W2619810135> ?p ?o ?g. }
- W2619810135 abstract "We address the problem of reconstructing sparse signals from noisy and compressive measurements using a feed-forward deep neural network (DNN) with an architecture motivated by the iterative shrinkage-thresholding algorithm (ISTA). We maintain the weights and biases of the network links as prescribed by ISTA and model the nonlinear activation function using a linear expansion of thresholds (LET), which has been very successful in image denoising and deconvolution. The optimal set of coefficients of the parametrized activation is learned over a training dataset containing measurement-sparse signal pairs, corresponding to a fixed sensing matrix. For training, we develop an efficient second-order algorithm, which requires only matrix-vector product computations in every training epoch (Hessian-free optimization) and offers superior convergence performance than gradient-descent optimization. Subsequently, we derive an improved network architecture inspired by FISTA, a faster version of ISTA, to achieve similar signal estimation performance with about 50% of the number of layers. The resulting architecture turns out to be a deep residual network, which has recently been shown to exhibit superior performance in several visual recognition tasks. Numerical experiments demonstrate that the proposed DNN architectures lead to 3 to 4 dB improvement in the reconstruction signal-to-noise ratio (SNR), compared with the state-of-the-art sparse coding algorithms." @default.
- W2619810135 created "2017-06-05" @default.
- W2619810135 creator A5005652970 @default.
- W2619810135 creator A5024424541 @default.
- W2619810135 creator A5035377071 @default.
- W2619810135 date "2017-05-20" @default.
- W2619810135 modified "2023-09-27" @default.
- W2619810135 title "Deep Sparse Coding Using Optimized Linear Expansion of Thresholds." @default.
- W2619810135 cites W171902450 @default.
- W2619810135 cites W1900669773 @default.
- W2619810135 cites W1926981477 @default.
- W2619810135 cites W1999726054 @default.
- W2619810135 cites W2007203285 @default.
- W2619810135 cites W2017812570 @default.
- W2619810135 cites W2026114584 @default.
- W2619810135 cites W2043003570 @default.
- W2619810135 cites W2063978378 @default.
- W2619810135 cites W2100556411 @default.
- W2619810135 cites W2101675075 @default.
- W2619810135 cites W2109952872 @default.
- W2619810135 cites W2115706991 @default.
- W2619810135 cites W2118103795 @default.
- W2619810135 cites W2119667497 @default.
- W2619810135 cites W2119883478 @default.
- W2619810135 cites W2121058967 @default.
- W2619810135 cites W2127271355 @default.
- W2619810135 cites W2129812935 @default.
- W2619810135 cites W2135046866 @default.
- W2619810135 cites W2140742952 @default.
- W2619810135 cites W2145094598 @default.
- W2619810135 cites W2153663612 @default.
- W2619810135 cites W2160979406 @default.
- W2619810135 cites W2162718566 @default.
- W2619810135 cites W2164452299 @default.
- W2619810135 cites W2167219413 @default.
- W2619810135 cites W2194775991 @default.
- W2619810135 cites W2252143850 @default.
- W2619810135 cites W2289917018 @default.
- W2619810135 cites W2346728112 @default.
- W2619810135 cites W2950320139 @default.
- W2619810135 cites W2950942950 @default.
- W2619810135 cites W2951287198 @default.
- W2619810135 cites W2963542991 @default.
- W2619810135 cites W2964288706 @default.
- W2619810135 cites W3100774664 @default.
- W2619810135 hasPublicationYear "2017" @default.
- W2619810135 type Work @default.
- W2619810135 sameAs 2619810135 @default.
- W2619810135 citedByCount "4" @default.
- W2619810135 countsByYear W26198101352019 @default.
- W2619810135 countsByYear W26198101352020 @default.
- W2619810135 countsByYear W26198101352021 @default.
- W2619810135 crossrefType "posted-content" @default.
- W2619810135 hasAuthorship W2619810135A5005652970 @default.
- W2619810135 hasAuthorship W2619810135A5024424541 @default.
- W2619810135 hasAuthorship W2619810135A5035377071 @default.
- W2619810135 hasConcept C11413529 @default.
- W2619810135 hasConcept C153180895 @default.
- W2619810135 hasConcept C153258448 @default.
- W2619810135 hasConcept C154945302 @default.
- W2619810135 hasConcept C155512373 @default.
- W2619810135 hasConcept C157553263 @default.
- W2619810135 hasConcept C163294075 @default.
- W2619810135 hasConcept C174576160 @default.
- W2619810135 hasConcept C193415008 @default.
- W2619810135 hasConcept C203616005 @default.
- W2619810135 hasConcept C28826006 @default.
- W2619810135 hasConcept C33923547 @default.
- W2619810135 hasConcept C38652104 @default.
- W2619810135 hasConcept C41008148 @default.
- W2619810135 hasConcept C50644808 @default.
- W2619810135 hasConcept C77637269 @default.
- W2619810135 hasConceptScore W2619810135C11413529 @default.
- W2619810135 hasConceptScore W2619810135C153180895 @default.
- W2619810135 hasConceptScore W2619810135C153258448 @default.
- W2619810135 hasConceptScore W2619810135C154945302 @default.
- W2619810135 hasConceptScore W2619810135C155512373 @default.
- W2619810135 hasConceptScore W2619810135C157553263 @default.
- W2619810135 hasConceptScore W2619810135C163294075 @default.
- W2619810135 hasConceptScore W2619810135C174576160 @default.
- W2619810135 hasConceptScore W2619810135C193415008 @default.
- W2619810135 hasConceptScore W2619810135C203616005 @default.
- W2619810135 hasConceptScore W2619810135C28826006 @default.
- W2619810135 hasConceptScore W2619810135C33923547 @default.
- W2619810135 hasConceptScore W2619810135C38652104 @default.
- W2619810135 hasConceptScore W2619810135C41008148 @default.
- W2619810135 hasConceptScore W2619810135C50644808 @default.
- W2619810135 hasConceptScore W2619810135C77637269 @default.
- W2619810135 hasLocation W26198101351 @default.
- W2619810135 hasOpenAccess W2619810135 @default.
- W2619810135 hasPrimaryLocation W26198101351 @default.
- W2619810135 hasRelatedWork W2078204800 @default.
- W2619810135 hasRelatedWork W2118103795 @default.
- W2619810135 hasRelatedWork W2529783450 @default.
- W2619810135 hasRelatedWork W2591984662 @default.
- W2619810135 hasRelatedWork W2765969184 @default.
- W2619810135 hasRelatedWork W2788847598 @default.
- W2619810135 hasRelatedWork W2793080158 @default.
- W2619810135 hasRelatedWork W2805747803 @default.
- W2619810135 hasRelatedWork W2883220346 @default.