Matches in SemOpenAlex for { <https://semopenalex.org/work/W2513592917> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W2513592917 abstract "Substantial amounts of resources are usually required to robustly develop a language model for an open vocabulary speech recognition system as out-of-vocabulary (OOV) words can hurt recognition accuracy. In this work, we applied a hybrid lexicon of word and sub-word units to resolve the problem of OOV words in a resource-efficient way. As sub-lexical units can be combined to form new words, a compact set of hybrid vocabulary can be used while still maintaining a low OOV rate. For Thai, a syllable-based unit called pseudo-morpheme (PM) was chosen as a sub-word unit. To also benefit from different levels of linguistic information embedded in different input types, a hybrid recurrent neural network language model (RNNLM) framework is proposed. An RNNLM can model not only information from multiple-type input units through a hybrid input vector of words and PMs, but can also capture long context history through recurrent connections. Several hybrid input representations were also explored to optimize both recognition accuracy and computational time. The hybrid LM has shown to be both resource-efficient and well-performed on two Thai LVCSR tasks: broadcast news transcription and speech-to-speech translation. The proposed hybrid lexicon can constitute an open vocabulary for Thai LVCSR as it can greatly reduce the OOV rate to less than 1 % while using only 42 % of the vocabulary size of the word-based lexicon. In terms of recognition performance, the best proposed hybrid RNNLM, which uses a mixed word-PM input, obtained 1.54 % relative WER reduction when compared with a conventional word-based RNNLM. In terms of computational time, the best hybrid RNNLM has the lowest training and decoding time among all RNNLMs including the word-based RNNLM. The overall relative reduction on WER of the proposed hybrid RNNLM over a traditional n-gram model is 6.91 %." @default.
- W2513592917 created "2016-09-16" @default.
- W2513592917 creator A5002206098 @default.
- W2513592917 creator A5061820788 @default.
- W2513592917 creator A5065708730 @default.
- W2513592917 date "2016-08-08" @default.
- W2513592917 modified "2023-10-18" @default.
- W2513592917 title "A hybrid input-type recurrent neural network for LVCSR language modeling" @default.
- W2513592917 cites W2001331936 @default.
- W2513592917 cites W2009840120 @default.
- W2513592917 cites W2023289124 @default.
- W2513592917 cites W2026149468 @default.
- W2513592917 cites W2112844297 @default.
- W2513592917 cites W2150907703 @default.
- W2513592917 cites W2160807445 @default.
- W2513592917 cites W2160914720 @default.
- W2513592917 cites W2171928131 @default.
- W2513592917 doi "https://doi.org/10.1186/s13636-016-0093-x" @default.
- W2513592917 hasPublicationYear "2016" @default.
- W2513592917 type Work @default.
- W2513592917 sameAs 2513592917 @default.
- W2513592917 citedByCount "7" @default.
- W2513592917 countsByYear W25135929172017 @default.
- W2513592917 countsByYear W25135929172018 @default.
- W2513592917 countsByYear W25135929172020 @default.
- W2513592917 countsByYear W25135929172021 @default.
- W2513592917 countsByYear W25135929172022 @default.
- W2513592917 crossrefType "journal-article" @default.
- W2513592917 hasAuthorship W2513592917A5002206098 @default.
- W2513592917 hasAuthorship W2513592917A5061820788 @default.
- W2513592917 hasAuthorship W2513592917A5065708730 @default.
- W2513592917 hasBestOaLocation W25135929171 @default.
- W2513592917 hasConcept C109089402 @default.
- W2513592917 hasConcept C137293760 @default.
- W2513592917 hasConcept C138885662 @default.
- W2513592917 hasConcept C147168706 @default.
- W2513592917 hasConcept C151730666 @default.
- W2513592917 hasConcept C154945302 @default.
- W2513592917 hasConcept C165297611 @default.
- W2513592917 hasConcept C204321447 @default.
- W2513592917 hasConcept C2777601683 @default.
- W2513592917 hasConcept C2778121359 @default.
- W2513592917 hasConcept C2779343474 @default.
- W2513592917 hasConcept C28490314 @default.
- W2513592917 hasConcept C40969351 @default.
- W2513592917 hasConcept C41008148 @default.
- W2513592917 hasConcept C41895202 @default.
- W2513592917 hasConcept C50644808 @default.
- W2513592917 hasConcept C86803240 @default.
- W2513592917 hasConcept C90805587 @default.
- W2513592917 hasConceptScore W2513592917C109089402 @default.
- W2513592917 hasConceptScore W2513592917C137293760 @default.
- W2513592917 hasConceptScore W2513592917C138885662 @default.
- W2513592917 hasConceptScore W2513592917C147168706 @default.
- W2513592917 hasConceptScore W2513592917C151730666 @default.
- W2513592917 hasConceptScore W2513592917C154945302 @default.
- W2513592917 hasConceptScore W2513592917C165297611 @default.
- W2513592917 hasConceptScore W2513592917C204321447 @default.
- W2513592917 hasConceptScore W2513592917C2777601683 @default.
- W2513592917 hasConceptScore W2513592917C2778121359 @default.
- W2513592917 hasConceptScore W2513592917C2779343474 @default.
- W2513592917 hasConceptScore W2513592917C28490314 @default.
- W2513592917 hasConceptScore W2513592917C40969351 @default.
- W2513592917 hasConceptScore W2513592917C41008148 @default.
- W2513592917 hasConceptScore W2513592917C41895202 @default.
- W2513592917 hasConceptScore W2513592917C50644808 @default.
- W2513592917 hasConceptScore W2513592917C86803240 @default.
- W2513592917 hasConceptScore W2513592917C90805587 @default.
- W2513592917 hasLocation W25135929171 @default.
- W2513592917 hasOpenAccess W2513592917 @default.
- W2513592917 hasPrimaryLocation W25135929171 @default.
- W2513592917 hasRelatedWork W1488211819 @default.
- W2513592917 hasRelatedWork W1501139663 @default.
- W2513592917 hasRelatedWork W1512898506 @default.
- W2513592917 hasRelatedWork W2029461472 @default.
- W2513592917 hasRelatedWork W2050469586 @default.
- W2513592917 hasRelatedWork W2091981305 @default.
- W2513592917 hasRelatedWork W2101295087 @default.
- W2513592917 hasRelatedWork W2345190899 @default.
- W2513592917 hasRelatedWork W2586419604 @default.
- W2513592917 hasRelatedWork W2735949438 @default.
- W2513592917 hasRelatedWork W2762466393 @default.
- W2513592917 hasRelatedWork W2892115357 @default.
- W2513592917 hasRelatedWork W2896411932 @default.
- W2513592917 hasRelatedWork W2905678190 @default.
- W2513592917 hasRelatedWork W2912418351 @default.
- W2513592917 hasRelatedWork W2914603327 @default.
- W2513592917 hasRelatedWork W2944691285 @default.
- W2513592917 hasRelatedWork W2951327905 @default.
- W2513592917 hasRelatedWork W3034311908 @default.
- W2513592917 hasRelatedWork W3109672743 @default.
- W2513592917 isParatext "false" @default.
- W2513592917 isRetracted "false" @default.
- W2513592917 magId "2513592917" @default.
- W2513592917 workType "article" @default.