Matches in SemOpenAlex for { <https://semopenalex.org/work/W2012174101> ?p ?o ?g. }
Showing items 1 to 99 of
99
with 100 items per page.
- W2012174101 abstract "In this paper, we propose the use of deep neural networks to expand conventional methods of statistical feature enhancement based on piecewise linear transformation. Stereo-based piecewise linear compensation for environments (SPLICE), which is a powerful statistical approach for feature enhancement, models the probabilistic distribution of input noisy features as a mixture of Gaussians. However, soft assignment of an input vector to divided regions is sometimes done inadequately and the vector comes to go through inadequate conversion. Especially when conversion has to be linear, the conversion performance will be easily degraded. Feature enhancement using neural networks is another powerful approach which can directly model a non-linear relationship between noisy and clean feature spaces. In this case, however, it tends to suffer from over-fitting problems. In this paper, we attempt to mitigate this problem by reducing the number of model parameters to estimate. Our neural network is trained whose output layer is associated with the states in the clean feature space, not in the noisy feature space. This strategy makes the size of the output layer independent of the kind of a given noisy environment. Firstly, we characterize the distribution of clean features as a Gaussian mixture model and then, by using deep neural networks, estimate discriminatively the state in the clean space that an input noisy feature corresponds to. Experimental evaluations using the Aurora 2 dataset demonstrate that our proposed method has the best performance compared to conventional methods." @default.
- W2012174101 created "2016-06-24" @default.
- W2012174101 creator A5010841595 @default.
- W2012174101 creator A5041213266 @default.
- W2012174101 creator A5055097130 @default.
- W2012174101 creator A5078386192 @default.
- W2012174101 date "2013-12-01" @default.
- W2012174101 modified "2023-10-02" @default.
- W2012174101 title "Discriminative piecewise linear transformation based on deep learning for noise robust automatic speech recognition" @default.
- W2012174101 cites W1524333225 @default.
- W2012174101 cites W1553154706 @default.
- W2012174101 cites W1997063331 @default.
- W2012174101 cites W2025768430 @default.
- W2012174101 cites W2119809277 @default.
- W2012174101 cites W2132695714 @default.
- W2012174101 cites W2136922672 @default.
- W2012174101 cites W2141520175 @default.
- W2012174101 cites W2144538561 @default.
- W2012174101 cites W2163648514 @default.
- W2012174101 cites W2290318471 @default.
- W2012174101 cites W58750608 @default.
- W2012174101 cites W96504731 @default.
- W2012174101 doi "https://doi.org/10.1109/asru.2013.6707755" @default.
- W2012174101 hasPublicationYear "2013" @default.
- W2012174101 type Work @default.
- W2012174101 sameAs 2012174101 @default.
- W2012174101 citedByCount "3" @default.
- W2012174101 countsByYear W20121741012014 @default.
- W2012174101 countsByYear W20121741012015 @default.
- W2012174101 crossrefType "proceedings-article" @default.
- W2012174101 hasAuthorship W2012174101A5010841595 @default.
- W2012174101 hasAuthorship W2012174101A5041213266 @default.
- W2012174101 hasAuthorship W2012174101A5055097130 @default.
- W2012174101 hasAuthorship W2012174101A5078386192 @default.
- W2012174101 hasBestOaLocation W20121741012 @default.
- W2012174101 hasConcept C104317684 @default.
- W2012174101 hasConcept C115961682 @default.
- W2012174101 hasConcept C138885662 @default.
- W2012174101 hasConcept C153180895 @default.
- W2012174101 hasConcept C154945302 @default.
- W2012174101 hasConcept C17095337 @default.
- W2012174101 hasConcept C185592680 @default.
- W2012174101 hasConcept C204241405 @default.
- W2012174101 hasConcept C2524010 @default.
- W2012174101 hasConcept C2776401178 @default.
- W2012174101 hasConcept C33923547 @default.
- W2012174101 hasConcept C41008148 @default.
- W2012174101 hasConcept C41895202 @default.
- W2012174101 hasConcept C50644808 @default.
- W2012174101 hasConcept C55493867 @default.
- W2012174101 hasConcept C83665646 @default.
- W2012174101 hasConcept C97931131 @default.
- W2012174101 hasConcept C99498987 @default.
- W2012174101 hasConceptScore W2012174101C104317684 @default.
- W2012174101 hasConceptScore W2012174101C115961682 @default.
- W2012174101 hasConceptScore W2012174101C138885662 @default.
- W2012174101 hasConceptScore W2012174101C153180895 @default.
- W2012174101 hasConceptScore W2012174101C154945302 @default.
- W2012174101 hasConceptScore W2012174101C17095337 @default.
- W2012174101 hasConceptScore W2012174101C185592680 @default.
- W2012174101 hasConceptScore W2012174101C204241405 @default.
- W2012174101 hasConceptScore W2012174101C2524010 @default.
- W2012174101 hasConceptScore W2012174101C2776401178 @default.
- W2012174101 hasConceptScore W2012174101C33923547 @default.
- W2012174101 hasConceptScore W2012174101C41008148 @default.
- W2012174101 hasConceptScore W2012174101C41895202 @default.
- W2012174101 hasConceptScore W2012174101C50644808 @default.
- W2012174101 hasConceptScore W2012174101C55493867 @default.
- W2012174101 hasConceptScore W2012174101C83665646 @default.
- W2012174101 hasConceptScore W2012174101C97931131 @default.
- W2012174101 hasConceptScore W2012174101C99498987 @default.
- W2012174101 hasLocation W20121741011 @default.
- W2012174101 hasLocation W20121741012 @default.
- W2012174101 hasOpenAccess W2012174101 @default.
- W2012174101 hasPrimaryLocation W20121741011 @default.
- W2012174101 hasRelatedWork W1972326848 @default.
- W2012174101 hasRelatedWork W1984205139 @default.
- W2012174101 hasRelatedWork W2047588738 @default.
- W2012174101 hasRelatedWork W2080834119 @default.
- W2012174101 hasRelatedWork W2109239522 @default.
- W2012174101 hasRelatedWork W2117857642 @default.
- W2012174101 hasRelatedWork W2119809277 @default.
- W2012174101 hasRelatedWork W2124479510 @default.
- W2012174101 hasRelatedWork W2126199866 @default.
- W2012174101 hasRelatedWork W2129967803 @default.
- W2012174101 hasRelatedWork W2140898856 @default.
- W2012174101 hasRelatedWork W2153513468 @default.
- W2012174101 hasRelatedWork W2161693361 @default.
- W2012174101 hasRelatedWork W2166114080 @default.
- W2012174101 hasRelatedWork W2171070969 @default.
- W2012174101 hasRelatedWork W2233657726 @default.
- W2012174101 hasRelatedWork W2320912073 @default.
- W2012174101 hasRelatedWork W2547013604 @default.
- W2012174101 hasRelatedWork W2766672686 @default.
- W2012174101 hasRelatedWork W2900715170 @default.
- W2012174101 isParatext "false" @default.
- W2012174101 isRetracted "false" @default.
- W2012174101 magId "2012174101" @default.
- W2012174101 workType "article" @default.