Matches in SemOpenAlex for { <https://semopenalex.org/work/W2776923314> ?p ?o ?g. }
Showing items 1 to 77 of
77
with 100 items per page.
- W2776923314 abstract "English. This paper introduces a novel application of the hybrid deep neural network (DNN)-hidden Markov model (HMM) approach for automatic speech recognition (ASR) to target groups of speakers of a specific age/gender. The group-specific training of DNN is investigated and shown to be inefficient when the amount of training data is limited. To overcome this problem, the recent approach that consists in adapting a general DNN to domain/language specific data is extended to target age/gender groups in the context of hybrid DNN-HMM systems, reducing consistently the phone error rate by 15-20% relative for the three different speaker groups. Italiano. Questo articolo propone l'applicazione del modello ibrido rete neurale artificiale multistrato-modelli di Markov nascosti al riconoscimento automatico del parlato per gruppi di parlanti di una specifica fascia di eta o genere che in questo caso sono costituiti da: bambini, maschi adulti e femmine adulte. L'addestramente della rete neurale multistrato si e dimostrato poco efficace quando i dati di addestra-mento erano disponibili solo in piccola quantita per uno specifico gruppo di par-lanti. Per migliorare le prestazioni, un re-cente approccio proposto per adattare una rete neurale multistrato pre-addestrata ad un nuovo domino o ad una nuova lingualingua e stato esteso al caso di gruppi di par-lanti di diverse eta e genere. L'adozione di di una rete multistrato adattata per cias-cun gruppo di parlanti ha consentito di ottenere una riduzione dell'errore nel ri-conoscimento di fonemi del 15-20% rela-tivo per ciascuno dei tre gruppi di parlanti considerati." @default.
- W2776923314 created "2018-01-05" @default.
- W2776923314 creator A5057130762 @default.
- W2776923314 creator A5064988262 @default.
- W2776923314 date "2014-12-01" @default.
- W2776923314 modified "2023-09-24" @default.
- W2776923314 title "Deep neural network adaptation for children's and adults' speech recognition" @default.
- W2776923314 cites W1559401186 @default.
- W2776923314 cites W1603759003 @default.
- W2776923314 cites W1932968309 @default.
- W2776923314 cites W1993660824 @default.
- W2776923314 cites W1993882792 @default.
- W2776923314 cites W2056914870 @default.
- W2776923314 cites W2100118853 @default.
- W2776923314 cites W2117620049 @default.
- W2776923314 cites W2120137673 @default.
- W2776923314 cites W2120209245 @default.
- W2776923314 cites W2124812654 @default.
- W2776923314 cites W2130827473 @default.
- W2776923314 cites W2131042651 @default.
- W2776923314 cites W2136922672 @default.
- W2776923314 cites W2137325773 @default.
- W2776923314 cites W2138857742 @default.
- W2776923314 cites W2147768505 @default.
- W2776923314 cites W2150919661 @default.
- W2776923314 cites W2153107398 @default.
- W2776923314 cites W2164317961 @default.
- W2776923314 cites W2184045248 @default.
- W2776923314 cites W2606321545 @default.
- W2776923314 cites W97192492 @default.
- W2776923314 hasPublicationYear "2014" @default.
- W2776923314 type Work @default.
- W2776923314 sameAs 2776923314 @default.
- W2776923314 citedByCount "6" @default.
- W2776923314 countsByYear W27769233142016 @default.
- W2776923314 countsByYear W27769233142018 @default.
- W2776923314 countsByYear W27769233142019 @default.
- W2776923314 countsByYear W27769233142020 @default.
- W2776923314 countsByYear W27769233142021 @default.
- W2776923314 crossrefType "proceedings-article" @default.
- W2776923314 hasAuthorship W2776923314A5057130762 @default.
- W2776923314 hasAuthorship W2776923314A5064988262 @default.
- W2776923314 hasBestOaLocation W27769233141 @default.
- W2776923314 hasConcept C139807058 @default.
- W2776923314 hasConcept C154945302 @default.
- W2776923314 hasConcept C15744967 @default.
- W2776923314 hasConcept C169760540 @default.
- W2776923314 hasConcept C175202392 @default.
- W2776923314 hasConcept C28490314 @default.
- W2776923314 hasConcept C41008148 @default.
- W2776923314 hasConcept C50644808 @default.
- W2776923314 hasConceptScore W2776923314C139807058 @default.
- W2776923314 hasConceptScore W2776923314C154945302 @default.
- W2776923314 hasConceptScore W2776923314C15744967 @default.
- W2776923314 hasConceptScore W2776923314C169760540 @default.
- W2776923314 hasConceptScore W2776923314C175202392 @default.
- W2776923314 hasConceptScore W2776923314C28490314 @default.
- W2776923314 hasConceptScore W2776923314C41008148 @default.
- W2776923314 hasConceptScore W2776923314C50644808 @default.
- W2776923314 hasLocation W27769233141 @default.
- W2776923314 hasLocation W27769233142 @default.
- W2776923314 hasOpenAccess W2776923314 @default.
- W2776923314 hasPrimaryLocation W27769233141 @default.
- W2776923314 hasRelatedWork W1679636228 @default.
- W2776923314 hasRelatedWork W2156694948 @default.
- W2776923314 hasRelatedWork W2329734087 @default.
- W2776923314 hasRelatedWork W2386387936 @default.
- W2776923314 hasRelatedWork W2396375857 @default.
- W2776923314 hasRelatedWork W2794558692 @default.
- W2776923314 hasRelatedWork W2883762569 @default.
- W2776923314 hasRelatedWork W3142820373 @default.
- W2776923314 hasRelatedWork W3175075966 @default.
- W2776923314 hasRelatedWork W1629725936 @default.
- W2776923314 isParatext "false" @default.
- W2776923314 isRetracted "false" @default.
- W2776923314 magId "2776923314" @default.
- W2776923314 workType "article" @default.