Matches in SemOpenAlex for { <https://semopenalex.org/work/W3049198593> ?p ?o ?g. }
Showing items 1 to 91 of
91
with 100 items per page.
- W3049198593 abstract "Generalized end-to-end (GE2E) model is widely used in speaker verification (SV) fields due to its expandability and generality regardless of specific languages. However, the long-short term memory (LSTM) based on GE2E has two limitations: First, the embedding of GE2E suffers from vanishing gradient, which leads to performance degradation for very long input sequences. Secondly, utterances are not represented as a properly fixed dimensional vector. In this paper, to overcome issues mentioned above, we propose a novel framework for SV, end-to-end trainable self-attentive shallow network (SASN), incorporating a time-delay neural network (TDNN) and a self-attentive pooling mechanism based on the self-attentive x-vector system during an utterance embedding phase. We demonstrate that the proposed model is highly efficient, and provides more accurate speaker verification than GE2E. For VCTK dataset, with just less than half the size of GE2E, the proposed model showed significant performance improvement over GE2E of about 63%, 67%, and 85% in EER (Equal error rate), DCF (Detection cost function), and AUC (Area under the curve), respectively. Notably, when the input length becomes longer, the DCF score improvement of the proposed model is about 17 times greater than that of GE2E." @default.
- W3049198593 created "2020-08-21" @default.
- W3049198593 creator A5046603228 @default.
- W3049198593 creator A5075631319 @default.
- W3049198593 creator A5089889189 @default.
- W3049198593 date "2020-08-14" @default.
- W3049198593 modified "2023-09-26" @default.
- W3049198593 title "End-to-End Trainable Self-Attentive Shallow Network for Text-Independent Speaker Verification." @default.
- W3049198593 cites W1589137271 @default.
- W3049198593 cites W1635512741 @default.
- W3049198593 cites W2046056978 @default.
- W3049198593 cites W2064675550 @default.
- W3049198593 cites W2150769028 @default.
- W3049198593 cites W2281877999 @default.
- W3049198593 cites W2402146185 @default.
- W3049198593 cites W2584329820 @default.
- W3049198593 cites W2597655663 @default.
- W3049198593 cites W2612434969 @default.
- W3049198593 cites W2890964092 @default.
- W3049198593 cites W2896538040 @default.
- W3049198593 cites W2962739339 @default.
- W3049198593 cites W2963341956 @default.
- W3049198593 cites W2963432880 @default.
- W3049198593 cites W2963470929 @default.
- W3049198593 cites W2972495969 @default.
- W3049198593 cites W2973062255 @default.
- W3049198593 hasPublicationYear "2020" @default.
- W3049198593 type Work @default.
- W3049198593 sameAs 3049198593 @default.
- W3049198593 citedByCount "0" @default.
- W3049198593 crossrefType "posted-content" @default.
- W3049198593 hasAuthorship W3049198593A5046603228 @default.
- W3049198593 hasAuthorship W3049198593A5075631319 @default.
- W3049198593 hasAuthorship W3049198593A5089889189 @default.
- W3049198593 hasConcept C14036430 @default.
- W3049198593 hasConcept C153180895 @default.
- W3049198593 hasConcept C154945302 @default.
- W3049198593 hasConcept C15744967 @default.
- W3049198593 hasConcept C2775852435 @default.
- W3049198593 hasConcept C2780767217 @default.
- W3049198593 hasConcept C28490314 @default.
- W3049198593 hasConcept C41008148 @default.
- W3049198593 hasConcept C41608201 @default.
- W3049198593 hasConcept C50644808 @default.
- W3049198593 hasConcept C542102704 @default.
- W3049198593 hasConcept C70437156 @default.
- W3049198593 hasConcept C74296488 @default.
- W3049198593 hasConcept C78458016 @default.
- W3049198593 hasConcept C86803240 @default.
- W3049198593 hasConceptScore W3049198593C14036430 @default.
- W3049198593 hasConceptScore W3049198593C153180895 @default.
- W3049198593 hasConceptScore W3049198593C154945302 @default.
- W3049198593 hasConceptScore W3049198593C15744967 @default.
- W3049198593 hasConceptScore W3049198593C2775852435 @default.
- W3049198593 hasConceptScore W3049198593C2780767217 @default.
- W3049198593 hasConceptScore W3049198593C28490314 @default.
- W3049198593 hasConceptScore W3049198593C41008148 @default.
- W3049198593 hasConceptScore W3049198593C41608201 @default.
- W3049198593 hasConceptScore W3049198593C50644808 @default.
- W3049198593 hasConceptScore W3049198593C542102704 @default.
- W3049198593 hasConceptScore W3049198593C70437156 @default.
- W3049198593 hasConceptScore W3049198593C74296488 @default.
- W3049198593 hasConceptScore W3049198593C78458016 @default.
- W3049198593 hasConceptScore W3049198593C86803240 @default.
- W3049198593 hasLocation W30491985931 @default.
- W3049198593 hasOpenAccess W3049198593 @default.
- W3049198593 hasPrimaryLocation W30491985931 @default.
- W3049198593 hasRelatedWork W115323076 @default.
- W3049198593 hasRelatedWork W1539377945 @default.
- W3049198593 hasRelatedWork W171514242 @default.
- W3049198593 hasRelatedWork W2352468006 @default.
- W3049198593 hasRelatedWork W2378656144 @default.
- W3049198593 hasRelatedWork W2587150483 @default.
- W3049198593 hasRelatedWork W2787051072 @default.
- W3049198593 hasRelatedWork W2898678194 @default.
- W3049198593 hasRelatedWork W2902572630 @default.
- W3049198593 hasRelatedWork W2920731986 @default.
- W3049198593 hasRelatedWork W2951343470 @default.
- W3049198593 hasRelatedWork W2952783970 @default.
- W3049198593 hasRelatedWork W2996819981 @default.
- W3049198593 hasRelatedWork W3010590357 @default.
- W3049198593 hasRelatedWork W3012459487 @default.
- W3049198593 hasRelatedWork W3013246586 @default.
- W3049198593 hasRelatedWork W3024231486 @default.
- W3049198593 hasRelatedWork W3035714152 @default.
- W3049198593 hasRelatedWork W3171638953 @default.
- W3049198593 hasRelatedWork W3205788551 @default.
- W3049198593 isParatext "false" @default.
- W3049198593 isRetracted "false" @default.
- W3049198593 magId "3049198593" @default.
- W3049198593 workType "article" @default.