Matches in SemOpenAlex for { <https://semopenalex.org/work/W3197343310> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W3197343310 abstract "This paper proposes a serialized multi-layer multi-head attention for neural speaker embedding in text-independent speaker verification. In prior works, frame-level features from one layer are aggregated to form an utterance-level representation. Inspired by the Transformer network, our proposed method utilizes the hierarchical architecture of stacked self-attention mechanisms to derive refined features that are more correlated with speakers. Serialized attention mechanism contains a stack of self-attention modules to create fixed-dimensional representations of speakers. Instead of utilizing multi-head attention in parallel, the proposed serialized multi-layer multi-head attention is designed to aggregate and propagate attentive statistics from one layer to the next in a serialized manner. In addition, we employ an input-aware query for each utterance with the statistics pooling. With more layers stacked, the neural network can learn more discriminative speaker embeddings. Experiment results on VoxCeleb1 dataset and SITW dataset show that our proposed method outperforms other baseline methods, including x-vectors and other x-vectors + conventional attentive pooling approaches by 9.7% in EER and 8.1% in DCF0.01." @default.
- W3197343310 created "2021-09-13" @default.
- W3197343310 creator A5004287909 @default.
- W3197343310 creator A5032690182 @default.
- W3197343310 creator A5052393310 @default.
- W3197343310 date "2021-08-30" @default.
- W3197343310 modified "2023-10-16" @default.
- W3197343310 title "Serialized Multi-Layer Multi-Head Attention for Neural Speaker Embedding" @default.
- W3197343310 doi "https://doi.org/10.21437/interspeech.2021-2210" @default.
- W3197343310 hasPublicationYear "2021" @default.
- W3197343310 type Work @default.
- W3197343310 sameAs 3197343310 @default.
- W3197343310 citedByCount "6" @default.
- W3197343310 countsByYear W31973433102021 @default.
- W3197343310 countsByYear W31973433102022 @default.
- W3197343310 countsByYear W31973433102023 @default.
- W3197343310 crossrefType "proceedings-article" @default.
- W3197343310 hasAuthorship W3197343310A5004287909 @default.
- W3197343310 hasAuthorship W3197343310A5032690182 @default.
- W3197343310 hasAuthorship W3197343310A5052393310 @default.
- W3197343310 hasBestOaLocation W31973433102 @default.
- W3197343310 hasConcept C114793014 @default.
- W3197343310 hasConcept C121332964 @default.
- W3197343310 hasConcept C126042441 @default.
- W3197343310 hasConcept C127313418 @default.
- W3197343310 hasConcept C153180895 @default.
- W3197343310 hasConcept C154945302 @default.
- W3197343310 hasConcept C165801399 @default.
- W3197343310 hasConcept C178790620 @default.
- W3197343310 hasConcept C185592680 @default.
- W3197343310 hasConcept C2775852435 @default.
- W3197343310 hasConcept C2779227376 @default.
- W3197343310 hasConcept C2780312720 @default.
- W3197343310 hasConcept C28490314 @default.
- W3197343310 hasConcept C41008148 @default.
- W3197343310 hasConcept C41608201 @default.
- W3197343310 hasConcept C50644808 @default.
- W3197343310 hasConcept C62520636 @default.
- W3197343310 hasConcept C66322947 @default.
- W3197343310 hasConcept C70437156 @default.
- W3197343310 hasConcept C76155785 @default.
- W3197343310 hasConcept C97931131 @default.
- W3197343310 hasConceptScore W3197343310C114793014 @default.
- W3197343310 hasConceptScore W3197343310C121332964 @default.
- W3197343310 hasConceptScore W3197343310C126042441 @default.
- W3197343310 hasConceptScore W3197343310C127313418 @default.
- W3197343310 hasConceptScore W3197343310C153180895 @default.
- W3197343310 hasConceptScore W3197343310C154945302 @default.
- W3197343310 hasConceptScore W3197343310C165801399 @default.
- W3197343310 hasConceptScore W3197343310C178790620 @default.
- W3197343310 hasConceptScore W3197343310C185592680 @default.
- W3197343310 hasConceptScore W3197343310C2775852435 @default.
- W3197343310 hasConceptScore W3197343310C2779227376 @default.
- W3197343310 hasConceptScore W3197343310C2780312720 @default.
- W3197343310 hasConceptScore W3197343310C28490314 @default.
- W3197343310 hasConceptScore W3197343310C41008148 @default.
- W3197343310 hasConceptScore W3197343310C41608201 @default.
- W3197343310 hasConceptScore W3197343310C50644808 @default.
- W3197343310 hasConceptScore W3197343310C62520636 @default.
- W3197343310 hasConceptScore W3197343310C66322947 @default.
- W3197343310 hasConceptScore W3197343310C70437156 @default.
- W3197343310 hasConceptScore W3197343310C76155785 @default.
- W3197343310 hasConceptScore W3197343310C97931131 @default.
- W3197343310 hasLocation W31973433101 @default.
- W3197343310 hasLocation W31973433102 @default.
- W3197343310 hasOpenAccess W3197343310 @default.
- W3197343310 hasPrimaryLocation W31973433101 @default.
- W3197343310 hasRelatedWork W11553578 @default.
- W3197343310 hasRelatedWork W12014881 @default.
- W3197343310 hasRelatedWork W13074359 @default.
- W3197343310 hasRelatedWork W13158784 @default.
- W3197343310 hasRelatedWork W14719770 @default.
- W3197343310 hasRelatedWork W1569375 @default.
- W3197343310 hasRelatedWork W5289323 @default.
- W3197343310 hasRelatedWork W7082836 @default.
- W3197343310 hasRelatedWork W9190101 @default.
- W3197343310 hasRelatedWork W9547545 @default.
- W3197343310 isParatext "false" @default.
- W3197343310 isRetracted "false" @default.
- W3197343310 magId "3197343310" @default.
- W3197343310 workType "article" @default.