FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4372266801> ?p ?o ?g. }

• W4372266801 abstract "Disfluency detection has mainly been solved in a pipeline approach, as post-processing of speech recognition. In this study, we propose Transformer-based encoder-decoder models that jointly solve speech recognition and disfluency detection, which work in a streaming manner. Compared to pipeline approaches, the joint models can leverage acoustic information that makes disfluency detection robust to recognition errors and provide non-verbal clues. Moreover, joint modeling results in low-latency and lightweight inference. We investigate two joint model variants for streaming disfluency detection: a transcript-enriched model and a multi-task model. The transcript- enriched model is trained on text with special tags indicating the starting and ending points of the disfluent part. However, it has problems with latency and standard language model adaptation, which arise from the additional disfluency tags. We propose a multi-task model to solve such problems, which has two output layers at the Transformer decoder; one for speech recognition and the other for disfluency detection. It is modeled to be conditioned on the currently recognized token with an additional token-dependency mechanism. We show that the proposed joint models outperformed a BERT-based pipeline approach in both accuracy and latency, on both the Switch- board and the corpus of spontaneous Japanese." @default.
• W4372266801 created "2023-05-07" @default.
• W4372266801 creator A5001291873 @default.
• W4372266801 creator A5011937407 @default.
• W4372266801 creator A5034216052 @default.
• W4372266801 creator A5047892839 @default.
• W4372266801 creator A5055097130 @default.
• W4372266801 creator A5081903232 @default.
• W4372266801 creator A5089876843 @default.
• W4372266801 date "2023-06-04" @default.
• W4372266801 modified "2023-09-27" @default.
• W4372266801 title "Streaming Joint Speech Recognition and Disfluency Detection" @default.
• W4372266801 cites W2756923881 @default.
• W4372266801 cites W2899773543 @default.
• W4372266801 cites W2904571617 @default.
• W4372266801 cites W2936123380 @default.
• W4372266801 cites W2962760690 @default.
• W4372266801 cites W2962780374 @default.
• W4372266801 cites W2964172015 @default.
• W4372266801 cites W2972328063 @default.
• W4372266801 cites W3007433671 @default.
• W4372266801 cites W3148654612 @default.
• W4372266801 cites W3162665866 @default.
• W4372266801 cites W3171851140 @default.
• W4372266801 cites W3173625927 @default.
• W4372266801 cites W3175745257 @default.
• W4372266801 cites W3197433369 @default.
• W4372266801 cites W3198653233 @default.
• W4372266801 cites W3217767527 @default.
• W4372266801 cites W4224917162 @default.
• W4372266801 cites W4225369246 @default.
• W4372266801 cites W4296068599 @default.
• W4372266801 cites W4296069267 @default.
• W4372266801 doi "https://doi.org/10.1109/icassp49357.2023.10094620" @default.
• W4372266801 hasPublicationYear "2023" @default.
• W4372266801 type Work @default.
• W4372266801 citedByCount "0" @default.
• W4372266801 crossrefType "proceedings-article" @default.
• W4372266801 hasAuthorship W4372266801A5001291873 @default.
• W4372266801 hasAuthorship W4372266801A5011937407 @default.
• W4372266801 hasAuthorship W4372266801A5034216052 @default.
• W4372266801 hasAuthorship W4372266801A5047892839 @default.
• W4372266801 hasAuthorship W4372266801A5055097130 @default.
• W4372266801 hasAuthorship W4372266801A5081903232 @default.
• W4372266801 hasAuthorship W4372266801A5089876843 @default.
• W4372266801 hasBestOaLocation W43722668011 @default.
• W4372266801 hasConcept C111919701 @default.
• W4372266801 hasConcept C118505674 @default.
• W4372266801 hasConcept C121332964 @default.
• W4372266801 hasConcept C127413603 @default.
• W4372266801 hasConcept C137293760 @default.
• W4372266801 hasConcept C153083717 @default.
• W4372266801 hasConcept C154945302 @default.
• W4372266801 hasConcept C165801399 @default.
• W4372266801 hasConcept C170154142 @default.
• W4372266801 hasConcept C18555067 @default.
• W4372266801 hasConcept C199360897 @default.
• W4372266801 hasConcept C204201278 @default.
• W4372266801 hasConcept C23224414 @default.
• W4372266801 hasConcept C2776214188 @default.
• W4372266801 hasConcept C28490314 @default.
• W4372266801 hasConcept C38652104 @default.
• W4372266801 hasConcept C41008148 @default.
• W4372266801 hasConcept C43521106 @default.
• W4372266801 hasConcept C48145219 @default.
• W4372266801 hasConcept C61328038 @default.
• W4372266801 hasConcept C62520636 @default.
• W4372266801 hasConcept C66322947 @default.
• W4372266801 hasConcept C76155785 @default.
• W4372266801 hasConcept C82876162 @default.
• W4372266801 hasConceptScore W4372266801C111919701 @default.
• W4372266801 hasConceptScore W4372266801C118505674 @default.
• W4372266801 hasConceptScore W4372266801C121332964 @default.
• W4372266801 hasConceptScore W4372266801C127413603 @default.
• W4372266801 hasConceptScore W4372266801C137293760 @default.
• W4372266801 hasConceptScore W4372266801C153083717 @default.
• W4372266801 hasConceptScore W4372266801C154945302 @default.
• W4372266801 hasConceptScore W4372266801C165801399 @default.
• W4372266801 hasConceptScore W4372266801C170154142 @default.
• W4372266801 hasConceptScore W4372266801C18555067 @default.
• W4372266801 hasConceptScore W4372266801C199360897 @default.
• W4372266801 hasConceptScore W4372266801C204201278 @default.
• W4372266801 hasConceptScore W4372266801C23224414 @default.
• W4372266801 hasConceptScore W4372266801C2776214188 @default.
• W4372266801 hasConceptScore W4372266801C28490314 @default.
• W4372266801 hasConceptScore W4372266801C38652104 @default.
• W4372266801 hasConceptScore W4372266801C41008148 @default.
• W4372266801 hasConceptScore W4372266801C43521106 @default.
• W4372266801 hasConceptScore W4372266801C48145219 @default.
• W4372266801 hasConceptScore W4372266801C61328038 @default.
• W4372266801 hasConceptScore W4372266801C62520636 @default.
• W4372266801 hasConceptScore W4372266801C66322947 @default.
• W4372266801 hasConceptScore W4372266801C76155785 @default.
• W4372266801 hasConceptScore W4372266801C82876162 @default.
• W4372266801 hasLocation W43722668011 @default.
• W4372266801 hasLocation W43722668012 @default.
• W4372266801 hasOpenAccess W4372266801 @default.
• W4372266801 hasPrimaryLocation W43722668011 @default.
• W4372266801 hasRelatedWork W148740283 @default.
• W4372266801 hasRelatedWork W1542012215 @default.

• next