Matches in SemOpenAlex for { <https://semopenalex.org/work/W2511423218> ?p ?o ?g. }
- W2511423218 endingPage "115" @default.
- W2511423218 startingPage "108" @default.
- W2511423218 abstract "The adaptation of context-dependent deep neural network acoustic models is particularly challenging, because most of the context-dependent targets will have no occurrences in a small adaptation data set. Recently, a multi-task training technique has been proposed that trains the network with context-dependent and context-independent targets in parallel. This network structure offers a straightforward way for network adaptation by training only the context-independent part during the adaptation process. Here, we combine this simple adaptation technique with the KL-divergence regularization method also proposed recently. Employing multi-task training we attain a relative word error rate reduction of about 3 % on a broadcast news recognition task. Then, by using the combined adaptation technique we report a further error rate reduction of 2 % to 5 %, depending on the duration of the adaptation data, which ranged from 20 to 100 s." @default.
- W2511423218 created "2016-09-16" @default.
- W2511423218 creator A5020608163 @default.
- W2511423218 creator A5088559776 @default.
- W2511423218 date "2016-01-01" @default.
- W2511423218 modified "2023-09-26" @default.
- W2511423218 title "Adaptation of DNN Acoustic Models Using KL-divergence Regularization and Multi-task Training" @default.
- W2511423218 cites W118756463 @default.
- W2511423218 cites W1545083717 @default.
- W2511423218 cites W1891007208 @default.
- W2511423218 cites W1989549063 @default.
- W2511423218 cites W2010362084 @default.
- W2511423218 cites W2019973879 @default.
- W2511423218 cites W2056445351 @default.
- W2511423218 cites W2056825827 @default.
- W2511423218 cites W2080005694 @default.
- W2511423218 cites W2087006792 @default.
- W2511423218 cites W2094035326 @default.
- W2511423218 cites W2094147890 @default.
- W2511423218 cites W2160306971 @default.
- W2511423218 cites W2169840288 @default.
- W2511423218 cites W2296748324 @default.
- W2511423218 cites W2407793339 @default.
- W2511423218 doi "https://doi.org/10.1007/978-3-319-43958-7_12" @default.
- W2511423218 hasPublicationYear "2016" @default.
- W2511423218 type Work @default.
- W2511423218 sameAs 2511423218 @default.
- W2511423218 citedByCount "4" @default.
- W2511423218 countsByYear W25114232182017 @default.
- W2511423218 countsByYear W25114232182019 @default.
- W2511423218 countsByYear W25114232182020 @default.
- W2511423218 countsByYear W25114232182022 @default.
- W2511423218 crossrefType "book-chapter" @default.
- W2511423218 hasAuthorship W2511423218A5020608163 @default.
- W2511423218 hasAuthorship W2511423218A5088559776 @default.
- W2511423218 hasConcept C119857082 @default.
- W2511423218 hasConcept C120665830 @default.
- W2511423218 hasConcept C121332964 @default.
- W2511423218 hasConcept C138885662 @default.
- W2511423218 hasConcept C139807058 @default.
- W2511423218 hasConcept C151730666 @default.
- W2511423218 hasConcept C153180895 @default.
- W2511423218 hasConcept C154945302 @default.
- W2511423218 hasConcept C162324750 @default.
- W2511423218 hasConcept C183322885 @default.
- W2511423218 hasConcept C187736073 @default.
- W2511423218 hasConcept C207390915 @default.
- W2511423218 hasConcept C2776135515 @default.
- W2511423218 hasConcept C2779343474 @default.
- W2511423218 hasConcept C2780451532 @default.
- W2511423218 hasConcept C2781238097 @default.
- W2511423218 hasConcept C28490314 @default.
- W2511423218 hasConcept C40969351 @default.
- W2511423218 hasConcept C41008148 @default.
- W2511423218 hasConcept C41895202 @default.
- W2511423218 hasConcept C50644808 @default.
- W2511423218 hasConcept C86803240 @default.
- W2511423218 hasConceptScore W2511423218C119857082 @default.
- W2511423218 hasConceptScore W2511423218C120665830 @default.
- W2511423218 hasConceptScore W2511423218C121332964 @default.
- W2511423218 hasConceptScore W2511423218C138885662 @default.
- W2511423218 hasConceptScore W2511423218C139807058 @default.
- W2511423218 hasConceptScore W2511423218C151730666 @default.
- W2511423218 hasConceptScore W2511423218C153180895 @default.
- W2511423218 hasConceptScore W2511423218C154945302 @default.
- W2511423218 hasConceptScore W2511423218C162324750 @default.
- W2511423218 hasConceptScore W2511423218C183322885 @default.
- W2511423218 hasConceptScore W2511423218C187736073 @default.
- W2511423218 hasConceptScore W2511423218C207390915 @default.
- W2511423218 hasConceptScore W2511423218C2776135515 @default.
- W2511423218 hasConceptScore W2511423218C2779343474 @default.
- W2511423218 hasConceptScore W2511423218C2780451532 @default.
- W2511423218 hasConceptScore W2511423218C2781238097 @default.
- W2511423218 hasConceptScore W2511423218C28490314 @default.
- W2511423218 hasConceptScore W2511423218C40969351 @default.
- W2511423218 hasConceptScore W2511423218C41008148 @default.
- W2511423218 hasConceptScore W2511423218C41895202 @default.
- W2511423218 hasConceptScore W2511423218C50644808 @default.
- W2511423218 hasConceptScore W2511423218C86803240 @default.
- W2511423218 hasLocation W25114232181 @default.
- W2511423218 hasOpenAccess W2511423218 @default.
- W2511423218 hasPrimaryLocation W25114232181 @default.
- W2511423218 hasRelatedWork W143957211 @default.
- W2511423218 hasRelatedWork W1513820424 @default.
- W2511423218 hasRelatedWork W1545083717 @default.
- W2511423218 hasRelatedWork W1571447047 @default.
- W2511423218 hasRelatedWork W2394882406 @default.
- W2511423218 hasRelatedWork W2397381476 @default.
- W2511423218 hasRelatedWork W2397725648 @default.
- W2511423218 hasRelatedWork W2407793339 @default.
- W2511423218 hasRelatedWork W2508220198 @default.
- W2511423218 hasRelatedWork W2587088898 @default.
- W2511423218 hasRelatedWork W2589099502 @default.
- W2511423218 hasRelatedWork W2769025471 @default.
- W2511423218 hasRelatedWork W2791636785 @default.
- W2511423218 hasRelatedWork W2921591317 @default.
- W2511423218 hasRelatedWork W2922158904 @default.
- W2511423218 hasRelatedWork W2922215817 @default.