Matches in SemOpenAlex for { <https://semopenalex.org/work/W3199551491> ?p ?o ?g. }
- W3199551491 abstract "In this paper, we propose a novel algorithm for speaker diarization using metric learning for graph based clustering. The graph clustering algorithms use an adjacency matrix consisting of similarity scores. These scores are computed between speaker embeddings extracted from pairs of audio segments within the given recording. In this paper, we propose an approach that jointly learns the speaker embeddings and the similarity metric using principles of self-supervised learning. The metric learning network implements a neural model of the probabilistic linear discriminant analysis (PLDA). The self-supervision is derived from the pseudo labels obtained from a previous iteration of clustering. The entire model of representation learning and metric learning is trained with a binary cross entropy loss. By combining the self-supervision based metric learning along with the graph-based clustering algorithm, we achieve significant relative improvements of 60% and 7% over the x-vector PLDA agglomerative hierarchical clustering (AHC) approach on AMI and the DIHARD datasets respectively in terms of diarization error rates (DER)." @default.
- W3199551491 created "2021-09-27" @default.
- W3199551491 creator A5002536077 @default.
- W3199551491 creator A5019066037 @default.
- W3199551491 date "2021-09-14" @default.
- W3199551491 modified "2023-09-27" @default.
- W3199551491 title "Self-Supervised Metric Learning With Graph Clustering For Speaker Diarization" @default.
- W3199551491 cites W1569447338 @default.
- W3199551491 cites W1589137271 @default.
- W3199551491 cites W1965819578 @default.
- W3199551491 cites W2035890032 @default.
- W3199551491 cites W2115861008 @default.
- W3199551491 cites W2164136210 @default.
- W3199551491 cites W2395750323 @default.
- W3199551491 cites W2405137844 @default.
- W3199551491 cites W2533545350 @default.
- W3199551491 cites W2726515241 @default.
- W3199551491 cites W2781711557 @default.
- W3199551491 cites W2805869973 @default.
- W3199551491 cites W2808631503 @default.
- W3199551491 cites W2890964092 @default.
- W3199551491 cites W2896538040 @default.
- W3199551491 cites W2938358845 @default.
- W3199551491 cites W2962852342 @default.
- W3199551491 cites W2963470929 @default.
- W3199551491 cites W2963745998 @default.
- W3199551491 cites W2973064719 @default.
- W3199551491 cites W2981461916 @default.
- W3199551491 cites W2989929945 @default.
- W3199551491 cites W3008357631 @default.
- W3199551491 cites W3015544392 @default.
- W3199551491 cites W3035193494 @default.
- W3199551491 cites W3047690012 @default.
- W3199551491 cites W3095212884 @default.
- W3199551491 cites W3104056985 @default.
- W3199551491 cites W3108962419 @default.
- W3199551491 cites W3115133930 @default.
- W3199551491 cites W3143367551 @default.
- W3199551491 cites W3208240190 @default.
- W3199551491 cites W48568691 @default.
- W3199551491 cites W2463481471 @default.
- W3199551491 hasPublicationYear "2021" @default.
- W3199551491 type Work @default.
- W3199551491 sameAs 3199551491 @default.
- W3199551491 citedByCount "0" @default.
- W3199551491 crossrefType "posted-content" @default.
- W3199551491 hasAuthorship W3199551491A5002536077 @default.
- W3199551491 hasAuthorship W3199551491A5019066037 @default.
- W3199551491 hasConcept C119857082 @default.
- W3199551491 hasConcept C132525143 @default.
- W3199551491 hasConcept C133892786 @default.
- W3199551491 hasConcept C149838564 @default.
- W3199551491 hasConcept C153180895 @default.
- W3199551491 hasConcept C154945302 @default.
- W3199551491 hasConcept C162324750 @default.
- W3199551491 hasConcept C167981619 @default.
- W3199551491 hasConcept C176217482 @default.
- W3199551491 hasConcept C180356752 @default.
- W3199551491 hasConcept C21547014 @default.
- W3199551491 hasConcept C41008148 @default.
- W3199551491 hasConcept C73555534 @default.
- W3199551491 hasConcept C80444323 @default.
- W3199551491 hasConcept C92835128 @default.
- W3199551491 hasConceptScore W3199551491C119857082 @default.
- W3199551491 hasConceptScore W3199551491C132525143 @default.
- W3199551491 hasConceptScore W3199551491C133892786 @default.
- W3199551491 hasConceptScore W3199551491C149838564 @default.
- W3199551491 hasConceptScore W3199551491C153180895 @default.
- W3199551491 hasConceptScore W3199551491C154945302 @default.
- W3199551491 hasConceptScore W3199551491C162324750 @default.
- W3199551491 hasConceptScore W3199551491C167981619 @default.
- W3199551491 hasConceptScore W3199551491C176217482 @default.
- W3199551491 hasConceptScore W3199551491C180356752 @default.
- W3199551491 hasConceptScore W3199551491C21547014 @default.
- W3199551491 hasConceptScore W3199551491C41008148 @default.
- W3199551491 hasConceptScore W3199551491C73555534 @default.
- W3199551491 hasConceptScore W3199551491C80444323 @default.
- W3199551491 hasConceptScore W3199551491C92835128 @default.
- W3199551491 hasLocation W31995514911 @default.
- W3199551491 hasOpenAccess W3199551491 @default.
- W3199551491 hasPrimaryLocation W31995514911 @default.
- W3199551491 hasRelatedWork W1517890933 @default.
- W3199551491 hasRelatedWork W1566259545 @default.
- W3199551491 hasRelatedWork W1604115184 @default.
- W3199551491 hasRelatedWork W1635833125 @default.
- W3199551491 hasRelatedWork W1991266022 @default.
- W3199551491 hasRelatedWork W2029614125 @default.
- W3199551491 hasRelatedWork W2033258526 @default.
- W3199551491 hasRelatedWork W2037856097 @default.
- W3199551491 hasRelatedWork W2060484766 @default.
- W3199551491 hasRelatedWork W2126459812 @default.
- W3199551491 hasRelatedWork W2127615881 @default.
- W3199551491 hasRelatedWork W2155995145 @default.
- W3199551491 hasRelatedWork W2157098416 @default.
- W3199551491 hasRelatedWork W2164068772 @default.
- W3199551491 hasRelatedWork W2377256194 @default.
- W3199551491 hasRelatedWork W2401611834 @default.
- W3199551491 hasRelatedWork W2943864189 @default.
- W3199551491 hasRelatedWork W3047690012 @default.
- W3199551491 hasRelatedWork W30751557 @default.