Matches in SemOpenAlex for { <https://semopenalex.org/work/W2891816510> ?p ?o ?g. }
Showing items 1 to 70 of
70
with 100 items per page.
- W2891816510 endingPage "46" @default.
- W2891816510 startingPage "39" @default.
- W2891816510 abstract "Abstract Multilingual models for Automatic Speech Recognition (ASR) are attractive as they have been shown to benefit from more training data, and better lend themselves to adaptation to under-resourced languages. However, initialisation from monolingual context-dependent models leads to an explosion of context-dependent states. Connectionist Temporal Classification (CTC) is a potential solution to this as it performs well with monophone labels. We investigate multilingual CTC training in the context of adaptation and regularisation techniques that have been shown to be beneficial in more conventional contexts. The multilingual model is trained to model a universal International Phonetic Alphabet (IPA)-based phone set using the CTC loss function. Learning Hidden Unit Contribution (LHUC) is investigated to perform language adaptive training. During cross-lingual adaptation, the idea of extending the multilingual output layer to new phonemes is introduced and investigated. In addition, dropout during multilingual training and cross-lingual adaptation is also studied and tested in order to mitigate the overfitting problem. Experiments show that the performance of the universal phoneme-based CTC system can be improved by applying dropout and LHUC and it is extensible to new phonemes during cross-lingual adaptation. Updating all acoustic model parameters shows consistent improvement on limited data. Applying dropout during adaptation can further improve the system and achieve competitive performance with Deep Neural Network / Hidden Markov Model (DNN/HMM) systems on limited data." @default.
- W2891816510 created "2018-09-27" @default.
- W2891816510 creator A5047712456 @default.
- W2891816510 creator A5077304441 @default.
- W2891816510 creator A5081808619 @default.
- W2891816510 date "2018-11-01" @default.
- W2891816510 modified "2023-10-03" @default.
- W2891816510 title "Cross-lingual adaptation of a CTC-based multilingual acoustic model" @default.
- W2891816510 cites W2100772283 @default.
- W2891816510 cites W2160815625 @default.
- W2891816510 cites W2515572005 @default.
- W2891816510 cites W2743818757 @default.
- W2891816510 cites W2891816510 @default.
- W2891816510 cites W2962704885 @default.
- W2891816510 cites W2963292011 @default.
- W2891816510 cites W2963303951 @default.
- W2891816510 cites W2963336460 @default.
- W2891816510 cites W2964309797 @default.
- W2891816510 doi "https://doi.org/10.1016/j.specom.2018.09.001" @default.
- W2891816510 hasPublicationYear "2018" @default.
- W2891816510 type Work @default.
- W2891816510 sameAs 2891816510 @default.
- W2891816510 citedByCount "25" @default.
- W2891816510 countsByYear W28918165102018 @default.
- W2891816510 countsByYear W28918165102019 @default.
- W2891816510 countsByYear W28918165102020 @default.
- W2891816510 countsByYear W28918165102021 @default.
- W2891816510 countsByYear W28918165102022 @default.
- W2891816510 crossrefType "journal-article" @default.
- W2891816510 hasAuthorship W2891816510A5047712456 @default.
- W2891816510 hasAuthorship W2891816510A5077304441 @default.
- W2891816510 hasAuthorship W2891816510A5081808619 @default.
- W2891816510 hasBestOaLocation W28918165102 @default.
- W2891816510 hasConcept C139807058 @default.
- W2891816510 hasConcept C154945302 @default.
- W2891816510 hasConcept C15744967 @default.
- W2891816510 hasConcept C169760540 @default.
- W2891816510 hasConcept C204321447 @default.
- W2891816510 hasConcept C28490314 @default.
- W2891816510 hasConcept C41008148 @default.
- W2891816510 hasConceptScore W2891816510C139807058 @default.
- W2891816510 hasConceptScore W2891816510C154945302 @default.
- W2891816510 hasConceptScore W2891816510C15744967 @default.
- W2891816510 hasConceptScore W2891816510C169760540 @default.
- W2891816510 hasConceptScore W2891816510C204321447 @default.
- W2891816510 hasConceptScore W2891816510C28490314 @default.
- W2891816510 hasConceptScore W2891816510C41008148 @default.
- W2891816510 hasFunder F4320338352 @default.
- W2891816510 hasLocation W28918165101 @default.
- W2891816510 hasLocation W28918165102 @default.
- W2891816510 hasLocation W28918165103 @default.
- W2891816510 hasOpenAccess W2891816510 @default.
- W2891816510 hasPrimaryLocation W28918165101 @default.
- W2891816510 hasRelatedWork W1543657218 @default.
- W2891816510 hasRelatedWork W1571518467 @default.
- W2891816510 hasRelatedWork W189527659 @default.
- W2891816510 hasRelatedWork W2008361377 @default.
- W2891816510 hasRelatedWork W2020291234 @default.
- W2891816510 hasRelatedWork W2135121413 @default.
- W2891816510 hasRelatedWork W2135470563 @default.
- W2891816510 hasRelatedWork W2611614995 @default.
- W2891816510 hasRelatedWork W3107474891 @default.
- W2891816510 hasRelatedWork W87991986 @default.
- W2891816510 hasVolume "104" @default.
- W2891816510 isParatext "false" @default.
- W2891816510 isRetracted "false" @default.
- W2891816510 magId "2891816510" @default.
- W2891816510 workType "article" @default.