Matches in SemOpenAlex for { <https://semopenalex.org/work/W2066378046> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W2066378046 abstract "This paper proposes an efficient one-pass decoding method for realtime speech recognition employing a recurrent neural network language model (RNNLM). An RNNLM is an effective language model that yields a large gain in recognition accuracy when it is combined with a standard n-gram model. However, since every word probability distribution based on an RNNLM is dependent on the entire history from the beginning of the speech, the search space in Viterbi decoding grows exponentially with the length of the recognition hypotheses and makes computation prohibitively expensive. Therefore, an RNNLM is usually used by N-best rescoring or by approximating it to a back-off n-gram model. In this paper, we present another approach that enables one-pass Viterbi decoding with an RNNLM without approximation, where the RNNLM is represented as a prefix tree of possible word sequences, and only the part needed for decoding is generated on-the-fly and used to rescore each hypothesis using an on-the-fly composition technique we previously proposed. Experimental results on the MIT lecture transcription task show that our proposed method enables one-pass decoding with small overhead for the RNNLM and achieves a slightly higher accuracy than 1000-best rescoring. Furthermore, it reduces the latency from the end of each utterance in two-pass decoding by a factor of 10." @default.
- W2066378046 created "2016-06-24" @default.
- W2066378046 creator A5010061679 @default.
- W2066378046 creator A5018620798 @default.
- W2066378046 creator A5087554069 @default.
- W2066378046 date "2014-05-01" @default.
- W2066378046 modified "2023-09-23" @default.
- W2066378046 title "Real-time one-pass decoding with recurrent neural network language model for speech recognition" @default.
- W2066378046 cites W143425341 @default.
- W2066378046 cites W1508752711 @default.
- W2066378046 cites W1531348272 @default.
- W2066378046 cites W1585876329 @default.
- W2066378046 cites W179875071 @default.
- W2066378046 cites W1934041838 @default.
- W2066378046 cites W1970689298 @default.
- W2066378046 cites W1984196678 @default.
- W2066378046 cites W2005733316 @default.
- W2066378046 cites W2046932483 @default.
- W2066378046 cites W2138889249 @default.
- W2066378046 cites W2171928131 @default.
- W2066378046 cites W2185726469 @default.
- W2066378046 cites W2293185259 @default.
- W2066378046 cites W2129057864 @default.
- W2066378046 doi "https://doi.org/10.1109/icassp.2014.6854829" @default.
- W2066378046 hasPublicationYear "2014" @default.
- W2066378046 type Work @default.
- W2066378046 sameAs 2066378046 @default.
- W2066378046 citedByCount "29" @default.
- W2066378046 countsByYear W20663780462014 @default.
- W2066378046 countsByYear W20663780462015 @default.
- W2066378046 countsByYear W20663780462016 @default.
- W2066378046 countsByYear W20663780462017 @default.
- W2066378046 countsByYear W20663780462018 @default.
- W2066378046 countsByYear W20663780462019 @default.
- W2066378046 countsByYear W20663780462020 @default.
- W2066378046 countsByYear W20663780462021 @default.
- W2066378046 countsByYear W20663780462022 @default.
- W2066378046 countsByYear W20663780462023 @default.
- W2066378046 crossrefType "proceedings-article" @default.
- W2066378046 hasAuthorship W2066378046A5010061679 @default.
- W2066378046 hasAuthorship W2066378046A5018620798 @default.
- W2066378046 hasAuthorship W2066378046A5087554069 @default.
- W2066378046 hasConcept C11413529 @default.
- W2066378046 hasConcept C117379686 @default.
- W2066378046 hasConcept C137293760 @default.
- W2066378046 hasConcept C147168706 @default.
- W2066378046 hasConcept C154945302 @default.
- W2066378046 hasConcept C28490314 @default.
- W2066378046 hasConcept C41008148 @default.
- W2066378046 hasConcept C50644808 @default.
- W2066378046 hasConcept C57273362 @default.
- W2066378046 hasConcept C60582962 @default.
- W2066378046 hasConceptScore W2066378046C11413529 @default.
- W2066378046 hasConceptScore W2066378046C117379686 @default.
- W2066378046 hasConceptScore W2066378046C137293760 @default.
- W2066378046 hasConceptScore W2066378046C147168706 @default.
- W2066378046 hasConceptScore W2066378046C154945302 @default.
- W2066378046 hasConceptScore W2066378046C28490314 @default.
- W2066378046 hasConceptScore W2066378046C41008148 @default.
- W2066378046 hasConceptScore W2066378046C50644808 @default.
- W2066378046 hasConceptScore W2066378046C57273362 @default.
- W2066378046 hasConceptScore W2066378046C60582962 @default.
- W2066378046 hasLocation W20663780461 @default.
- W2066378046 hasOpenAccess W2066378046 @default.
- W2066378046 hasPrimaryLocation W20663780461 @default.
- W2066378046 hasRelatedWork W1517386993 @default.
- W2066378046 hasRelatedWork W1520465330 @default.
- W2066378046 hasRelatedWork W179875071 @default.
- W2066378046 hasRelatedWork W1849603325 @default.
- W2066378046 hasRelatedWork W1964742985 @default.
- W2066378046 hasRelatedWork W1970689298 @default.
- W2066378046 hasRelatedWork W1994536225 @default.
- W2066378046 hasRelatedWork W2026149468 @default.
- W2066378046 hasRelatedWork W2034785571 @default.
- W2066378046 hasRelatedWork W2064675550 @default.
- W2066378046 hasRelatedWork W2091981305 @default.
- W2066378046 hasRelatedWork W2132339004 @default.
- W2066378046 hasRelatedWork W2171928131 @default.
- W2066378046 hasRelatedWork W2290469434 @default.
- W2066378046 hasRelatedWork W2396464458 @default.
- W2066378046 hasRelatedWork W2402268235 @default.
- W2066378046 hasRelatedWork W2404974730 @default.
- W2066378046 hasRelatedWork W2407031592 @default.
- W2066378046 hasRelatedWork W3184026069 @default.
- W2066378046 hasRelatedWork W36903255 @default.
- W2066378046 isParatext "false" @default.
- W2066378046 isRetracted "false" @default.
- W2066378046 magId "2066378046" @default.
- W2066378046 workType "article" @default.