Matches in SemOpenAlex for { <https://semopenalex.org/work/W2401396087> ?p ?o ?g. }
Showing items 1 to 82 of
82
with 100 items per page.
- W2401396087 abstract "Copyright © 2015 ISCA. The form of hidden activation functions has been always an important issue in deep neural network (DNN) design. The most common choices for acoustic modelling are the standard Sigmoid and rectified linear unit (ReLU), which are normally used with fixed function shapes and no adaptive parameters. Recently, there have been several papers that have studied the use of parameterised activation functions for both computer vision and speaker adaptation tasks. In this paper, we investigate generalised forms of both Sigmoid and ReLU with learnable parameters, as well as their integration with the standard DNN acoustic model training process. Experiments using conversational telephone speech (CTS) Mandarin data, result in an average of 3.4% and 2.0% relative word error rate (WER) reduction with Sigmoid and ReLU parameterisations." @default.
- W2401396087 created "2016-06-24" @default.
- W2401396087 creator A5002191410 @default.
- W2401396087 creator A5089966579 @default.
- W2401396087 date "2015-09-06" @default.
- W2401396087 modified "2023-10-15" @default.
- W2401396087 title "Parameterised sigmoid and reLU hidden activation functions for DNN acoustic modelling" @default.
- W2401396087 cites W1524333225 @default.
- W2401396087 cites W1533861849 @default.
- W2401396087 cites W1553004968 @default.
- W2401396087 cites W1554663460 @default.
- W2401396087 cites W1665214252 @default.
- W2401396087 cites W1677182931 @default.
- W2401396087 cites W2029947361 @default.
- W2401396087 cites W2033256038 @default.
- W2401396087 cites W2035424729 @default.
- W2401396087 cites W2036351241 @default.
- W2401396087 cites W2062227835 @default.
- W2401396087 cites W2085628288 @default.
- W2401396087 cites W2094147890 @default.
- W2401396087 cites W2100495367 @default.
- W2401396087 cites W2107878631 @default.
- W2401396087 cites W2154423186 @default.
- W2401396087 cites W2156387975 @default.
- W2401396087 cites W2158069733 @default.
- W2401396087 cites W2158289097 @default.
- W2401396087 cites W2160306971 @default.
- W2401396087 cites W2184045248 @default.
- W2401396087 cites W2296705629 @default.
- W2401396087 cites W2411442511 @default.
- W2401396087 cites W3207342693 @default.
- W2401396087 cites W75153181 @default.
- W2401396087 cites W2407698769 @default.
- W2401396087 doi "https://doi.org/10.21437/interspeech.2015-649" @default.
- W2401396087 hasPublicationYear "2015" @default.
- W2401396087 type Work @default.
- W2401396087 sameAs 2401396087 @default.
- W2401396087 citedByCount "30" @default.
- W2401396087 countsByYear W24013960872015 @default.
- W2401396087 countsByYear W24013960872016 @default.
- W2401396087 countsByYear W24013960872018 @default.
- W2401396087 countsByYear W24013960872019 @default.
- W2401396087 countsByYear W24013960872020 @default.
- W2401396087 countsByYear W24013960872021 @default.
- W2401396087 countsByYear W24013960872022 @default.
- W2401396087 countsByYear W24013960872023 @default.
- W2401396087 crossrefType "proceedings-article" @default.
- W2401396087 hasAuthorship W2401396087A5002191410 @default.
- W2401396087 hasAuthorship W2401396087A5089966579 @default.
- W2401396087 hasConcept C121332964 @default.
- W2401396087 hasConcept C154945302 @default.
- W2401396087 hasConcept C24890656 @default.
- W2401396087 hasConcept C28490314 @default.
- W2401396087 hasConcept C38365724 @default.
- W2401396087 hasConcept C41008148 @default.
- W2401396087 hasConcept C50644808 @default.
- W2401396087 hasConcept C81388566 @default.
- W2401396087 hasConceptScore W2401396087C121332964 @default.
- W2401396087 hasConceptScore W2401396087C154945302 @default.
- W2401396087 hasConceptScore W2401396087C24890656 @default.
- W2401396087 hasConceptScore W2401396087C28490314 @default.
- W2401396087 hasConceptScore W2401396087C38365724 @default.
- W2401396087 hasConceptScore W2401396087C41008148 @default.
- W2401396087 hasConceptScore W2401396087C50644808 @default.
- W2401396087 hasConceptScore W2401396087C81388566 @default.
- W2401396087 hasLocation W24013960871 @default.
- W2401396087 hasOpenAccess W2401396087 @default.
- W2401396087 hasPrimaryLocation W24013960871 @default.
- W2401396087 hasRelatedWork W1504105233 @default.
- W2401396087 hasRelatedWork W2041004593 @default.
- W2401396087 hasRelatedWork W2811324119 @default.
- W2401396087 hasRelatedWork W2895192346 @default.
- W2401396087 hasRelatedWork W3024979424 @default.
- W2401396087 hasRelatedWork W3032499992 @default.
- W2401396087 hasRelatedWork W3110577345 @default.
- W2401396087 hasRelatedWork W3134817226 @default.
- W2401396087 hasRelatedWork W3136579697 @default.
- W2401396087 hasRelatedWork W4283785902 @default.
- W2401396087 isParatext "false" @default.
- W2401396087 isRetracted "false" @default.
- W2401396087 magId "2401396087" @default.
- W2401396087 workType "article" @default.