Matches in SemOpenAlex for { <https://semopenalex.org/work/W4375869401> ?p ?o ?g. }
Showing items 1 to 84 of
84
with 100 items per page.
- W4375869401 abstract "Recently, end-to-end neural diarization (EEND) is introduced and achieves promising results in speaker-overlapped scenarios. In EEND, speaker diarization is formulated as a multi-label prediction problem, where speaker activities are estimated independently and their dependency are not well considered. To overcome these disadvantages, we employ the power set encoding to reformulate speaker diarization as a single-label classification problem and propose the overlap-aware EEND (EEND-OLA) model, in which speaker overlaps and dependency can be modeled explicitly. Inspired by the success of two-stage hybrid systems, we further propose a novel Two-stage OverLap-aware Diarization framework (TOLD) by involving a speaker overlap-aware post-processing (SOAP) model to iteratively refine the diarization results of EEND-OLA. Experimental results show that, compared with the original EEND, the proposed EEND-OLA achieves a 14.39% relative improvement in terms of diarization error rates (DER), and utilizing SOAP provides another 19.33% relative improvement. As a result, our method TOLD achieves a DER of 10.14% on the CALLHOME dataset, which is a new state-of-the-art result on this benchmark to the best of our knowledge." @default.
- W4375869401 created "2023-05-10" @default.
- W4375869401 creator A5001133136 @default.
- W4375869401 creator A5015807909 @default.
- W4375869401 creator A5055433405 @default.
- W4375869401 date "2023-06-04" @default.
- W4375869401 modified "2023-09-27" @default.
- W4375869401 title "TOLD: a Novel Two-Stage Overlap-Aware Framework for Speaker Diarization" @default.
- W4375869401 cites W2046056978 @default.
- W4375869401 cites W2081074144 @default.
- W4375869401 cites W2123768812 @default.
- W4375869401 cites W2148613904 @default.
- W4375869401 cites W2168941687 @default.
- W4375869401 cites W2514794854 @default.
- W4375869401 cites W2537749207 @default.
- W4375869401 cites W2696967604 @default.
- W4375869401 cites W2734774145 @default.
- W4375869401 cites W2890964092 @default.
- W4375869401 cites W2963470929 @default.
- W4375869401 cites W2963962398 @default.
- W4375869401 cites W2969985801 @default.
- W4375869401 cites W2972680151 @default.
- W4375869401 cites W2972949456 @default.
- W4375869401 cites W3008357631 @default.
- W4375869401 cites W3025260599 @default.
- W4375869401 cites W3095212884 @default.
- W4375869401 cites W3095822285 @default.
- W4375869401 cites W3145204487 @default.
- W4375869401 cites W3163019736 @default.
- W4375869401 cites W3196857193 @default.
- W4375869401 cites W3197916665 @default.
- W4375869401 cites W4214556932 @default.
- W4375869401 cites W4224925084 @default.
- W4375869401 cites W4225661121 @default.
- W4375869401 doi "https://doi.org/10.1109/icassp49357.2023.10096436" @default.
- W4375869401 hasPublicationYear "2023" @default.
- W4375869401 type Work @default.
- W4375869401 citedByCount "0" @default.
- W4375869401 crossrefType "proceedings-article" @default.
- W4375869401 hasAuthorship W4375869401A5001133136 @default.
- W4375869401 hasAuthorship W4375869401A5015807909 @default.
- W4375869401 hasAuthorship W4375869401A5055433405 @default.
- W4375869401 hasBestOaLocation W43758694011 @default.
- W4375869401 hasConcept C125411270 @default.
- W4375869401 hasConcept C13280743 @default.
- W4375869401 hasConcept C133892786 @default.
- W4375869401 hasConcept C149838564 @default.
- W4375869401 hasConcept C154945302 @default.
- W4375869401 hasConcept C177264268 @default.
- W4375869401 hasConcept C185798385 @default.
- W4375869401 hasConcept C19768560 @default.
- W4375869401 hasConcept C199360897 @default.
- W4375869401 hasConcept C205649164 @default.
- W4375869401 hasConcept C28490314 @default.
- W4375869401 hasConcept C41008148 @default.
- W4375869401 hasConceptScore W4375869401C125411270 @default.
- W4375869401 hasConceptScore W4375869401C13280743 @default.
- W4375869401 hasConceptScore W4375869401C133892786 @default.
- W4375869401 hasConceptScore W4375869401C149838564 @default.
- W4375869401 hasConceptScore W4375869401C154945302 @default.
- W4375869401 hasConceptScore W4375869401C177264268 @default.
- W4375869401 hasConceptScore W4375869401C185798385 @default.
- W4375869401 hasConceptScore W4375869401C19768560 @default.
- W4375869401 hasConceptScore W4375869401C199360897 @default.
- W4375869401 hasConceptScore W4375869401C205649164 @default.
- W4375869401 hasConceptScore W4375869401C28490314 @default.
- W4375869401 hasConceptScore W4375869401C41008148 @default.
- W4375869401 hasLocation W43758694011 @default.
- W4375869401 hasLocation W43758694012 @default.
- W4375869401 hasOpenAccess W4375869401 @default.
- W4375869401 hasPrimaryLocation W43758694011 @default.
- W4375869401 hasRelatedWork W1497807607 @default.
- W4375869401 hasRelatedWork W1509309911 @default.
- W4375869401 hasRelatedWork W1521049138 @default.
- W4375869401 hasRelatedWork W1813780412 @default.
- W4375869401 hasRelatedWork W2128773298 @default.
- W4375869401 hasRelatedWork W2144208207 @default.
- W4375869401 hasRelatedWork W2160753975 @default.
- W4375869401 hasRelatedWork W2162158162 @default.
- W4375869401 hasRelatedWork W2499802997 @default.
- W4375869401 hasRelatedWork W2175373321 @default.
- W4375869401 isParatext "false" @default.
- W4375869401 isRetracted "false" @default.
- W4375869401 workType "article" @default.