FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2486340537> ?p ?o ?g. }

• W2486340537 abstract "Feature transformations are commonly used in speech recognition to account for distribution mismatches between the source and target domains also referred to as covariate shift. Linear affine or piecewise linear transformations are typically considered. In this paper, we present deep neural network DNN based nonlinear feature transformations estimated under the maximum likelihood criterion. We use the hidden Markov model HMM to model speech feature sequences and features in each HMM state assume a Gaussian mixture model GMM distribution. The network is pre-trained close to a linear transformation followed by a fine-tuning using the gradient descent algorithm. Due to the nonlinearity, the gradients and the partition functions of GMM-HMM state distributions are evaluated using the Monte Carlo MC method based on importance sampling. In addition, a deep stacked architecture is proposed to hierarchically build a DNN as a series of sub-networks with each representing a nonlinear transformation itself, which can be learned using a block-wise learning strategy. Applications of the proposed nonlinear transformations in speaker/environment adaptation and acoustic modeling in large vocabulary continuous speech recognition tasks show its superior performance over the widely-used constrained maximum likelihood linear regression CMLLR." @default.
• W2486340537 created "2016-08-23" @default.
• W2486340537 creator A5034451965 @default.
• W2486340537 creator A5089852582 @default.
• W2486340537 date "2016-11-01" @default.
• W2486340537 modified "2023-09-24" @default.
• W2486340537 title "Maximum Likelihood Nonlinear Transformations Based on Deep Neural Networks" @default.
• W2486340537 cites W1506806321 @default.
• W2486340537 cites W1509793305 @default.
• W2486340537 cites W1529602138 @default.
• W2486340537 cites W1537275613 @default.
• W2486340537 cites W1599512239 @default.
• W2486340537 cites W1663973292 @default.
• W2486340537 cites W182014611 @default.
• W2486340537 cites W1979651826 @default.
• W2486340537 cites W1988115241 @default.
• W2486340537 cites W1993882792 @default.
• W2486340537 cites W2002342963 @default.
• W2486340537 cites W2006722592 @default.
• W2486340537 cites W2014208555 @default.
• W2486340537 cites W2034368206 @default.
• W2486340537 cites W2038507080 @default.
• W2486340537 cites W2049633694 @default.
• W2486340537 cites W2062164080 @default.
• W2486340537 cites W2078528584 @default.
• W2486340537 cites W2082474452 @default.
• W2486340537 cites W2086796102 @default.
• W2486340537 cites W2100969003 @default.
• W2486340537 cites W2103496339 @default.
• W2486340537 cites W2106554350 @default.
• W2486340537 cites W2113249522 @default.
• W2486340537 cites W2122272452 @default.
• W2486340537 cites W2125234026 @default.
• W2486340537 cites W2125378570 @default.
• W2486340537 cites W2136439176 @default.
• W2486340537 cites W2137983211 @default.
• W2486340537 cites W2146871184 @default.
• W2486340537 cites W2147768505 @default.
• W2486340537 cites W2150907703 @default.
• W2486340537 cites W2160306971 @default.
• W2486340537 cites W2160815625 @default.
• W2486340537 cites W2164931619 @default.
• W2486340537 cites W2184045248 @default.
• W2486340537 cites W2294059674 @default.
• W2486340537 cites W2308181013 @default.
• W2486340537 cites W2394932179 @default.
• W2486340537 cites W2403195671 @default.
• W2486340537 cites W2404166326 @default.
• W2486340537 cites W3020882730 @default.
• W2486340537 cites W3146803896 @default.
• W2486340537 cites W82936479 @default.
• W2486340537 doi "https://doi.org/10.1109/taslp.2016.2594255" @default.
• W2486340537 hasPublicationYear "2016" @default.
• W2486340537 type Work @default.
• W2486340537 sameAs 2486340537 @default.
• W2486340537 citedByCount "2" @default.
• W2486340537 countsByYear W24863405372018 @default.
• W2486340537 crossrefType "journal-article" @default.
• W2486340537 hasAuthorship W2486340537A5034451965 @default.
• W2486340537 hasAuthorship W2486340537A5089852582 @default.
• W2486340537 hasConcept C104317684 @default.
• W2486340537 hasConcept C11413529 @default.
• W2486340537 hasConcept C121332964 @default.
• W2486340537 hasConcept C138885662 @default.
• W2486340537 hasConcept C153180895 @default.
• W2486340537 hasConcept C154945302 @default.
• W2486340537 hasConcept C158622935 @default.
• W2486340537 hasConcept C163716315 @default.
• W2486340537 hasConcept C185592680 @default.
• W2486340537 hasConcept C204241405 @default.
• W2486340537 hasConcept C23224414 @default.
• W2486340537 hasConcept C2776401178 @default.
• W2486340537 hasConcept C28490314 @default.
• W2486340537 hasConcept C41008148 @default.
• W2486340537 hasConcept C41895202 @default.
• W2486340537 hasConcept C50644808 @default.
• W2486340537 hasConcept C55493867 @default.
• W2486340537 hasConcept C61224824 @default.
• W2486340537 hasConcept C62520636 @default.
• W2486340537 hasConceptScore W2486340537C104317684 @default.
• W2486340537 hasConceptScore W2486340537C11413529 @default.
• W2486340537 hasConceptScore W2486340537C121332964 @default.
• W2486340537 hasConceptScore W2486340537C138885662 @default.
• W2486340537 hasConceptScore W2486340537C153180895 @default.
• W2486340537 hasConceptScore W2486340537C154945302 @default.
• W2486340537 hasConceptScore W2486340537C158622935 @default.
• W2486340537 hasConceptScore W2486340537C163716315 @default.
• W2486340537 hasConceptScore W2486340537C185592680 @default.
• W2486340537 hasConceptScore W2486340537C204241405 @default.
• W2486340537 hasConceptScore W2486340537C23224414 @default.
• W2486340537 hasConceptScore W2486340537C2776401178 @default.
• W2486340537 hasConceptScore W2486340537C28490314 @default.
• W2486340537 hasConceptScore W2486340537C41008148 @default.
• W2486340537 hasConceptScore W2486340537C41895202 @default.
• W2486340537 hasConceptScore W2486340537C50644808 @default.
• W2486340537 hasConceptScore W2486340537C55493867 @default.
• W2486340537 hasConceptScore W2486340537C61224824 @default.
• W2486340537 hasConceptScore W2486340537C62520636 @default.
• W2486340537 hasFunder F4320333051 @default.
• W2486340537 hasLocation W24863405371 @default.

• next