Matches in SemOpenAlex for { <https://semopenalex.org/work/W4323664761> ?p ?o ?g. }
- W4323664761 endingPage "119833" @default.
- W4323664761 startingPage "119833" @default.
- W4323664761 abstract "As an attractive research in biometric authentication, Text Independent Speaker Verification (TI-SV) problem aims to specify whether two given unconstrained utterances come from the same speaker or not. As state-of-the-art solutions, end-to-end approaches using deep neural networks seek to learn a highly discriminative speaker embedding space. In this paper, we propose a novel end-to-end approach for speaker embedding learning by focusing on two crucial factors: speaker embedder architecture and objective function. The proposed module in the speaker embedder is composed of an Efficient Multi-resolution feature Representation (EMR) block followed by a Multi-scale Channel Attention Fusion (MCAF) block. The EMR effectively addresses the issue of fixed resolution convolutional kernels which commonly used in most embedder architectures. Moreover, the MCAF significantly improves the simple summation-based feature fusion used in residual embedder networks. Regarding the objective function, we conduct the speaker embedding space towards learning the embedding-to-embedding relations, in addition to only embedding-to-training class relations employed by most previous methods. So, we propose to employ a dynamic graph attention network, on top of the proposed embedder to learn all informative relations between embeddings, and then learn both embedder and graph-based networks in an end-to-end manner. We conduct various experiments on a large-scale benchmark dataset called VoxCeleb1&2. The effectiveness of all proposed components is verified through an ablation study. We show the superior or competitive performances of the proposed approach compared to seven well-known embedding architectures and 32 SV systems, regarding two evaluation metrics, EER and minDCF, as well as the number of embedder parameters." @default.
- W4323664761 created "2023-03-10" @default.
- W4323664761 creator A5017188788 @default.
- W4323664761 creator A5039023891 @default.
- W4323664761 date "2023-07-01" @default.
- W4323664761 modified "2023-10-16" @default.
- W4323664761 title "End-to-end deep speaker embedding learning using multi-scale attentional fusion and graph neural networks" @default.
- W4323664761 cites W2041823554 @default.
- W4323664761 cites W2046056978 @default.
- W4323664761 cites W2097117768 @default.
- W4323664761 cites W2107638917 @default.
- W4323664761 cites W2150769028 @default.
- W4323664761 cites W2183341477 @default.
- W4323664761 cites W2404617565 @default.
- W4323664761 cites W2549139847 @default.
- W4323664761 cites W2565639579 @default.
- W4323664761 cites W2696967604 @default.
- W4323664761 cites W2726515241 @default.
- W4323664761 cites W2747165665 @default.
- W4323664761 cites W2748488820 @default.
- W4323664761 cites W2748501745 @default.
- W4323664761 cites W2802973008 @default.
- W4323664761 cites W2808631503 @default.
- W4323664761 cites W2889385246 @default.
- W4323664761 cites W2889519245 @default.
- W4323664761 cites W2890964092 @default.
- W4323664761 cites W2916104401 @default.
- W4323664761 cites W2922509574 @default.
- W4323664761 cites W2928165649 @default.
- W4323664761 cites W2938358845 @default.
- W4323664761 cites W2940070181 @default.
- W4323664761 cites W2963125010 @default.
- W4323664761 cites W2963173190 @default.
- W4323664761 cites W2963466847 @default.
- W4323664761 cites W2963669405 @default.
- W4323664761 cites W2963988212 @default.
- W4323664761 cites W2964271799 @default.
- W4323664761 cites W2972369255 @default.
- W4323664761 cites W2972633940 @default.
- W4323664761 cites W2972657845 @default.
- W4323664761 cites W2972885011 @default.
- W4323664761 cites W2991234496 @default.
- W4323664761 cites W3013020904 @default.
- W4323664761 cites W3015276760 @default.
- W4323664761 cites W3015302980 @default.
- W4323664761 cites W3015537910 @default.
- W4323664761 cites W3020953549 @default.
- W4323664761 cites W3034202663 @default.
- W4323664761 cites W3034303554 @default.
- W4323664761 cites W3036173644 @default.
- W4323664761 cites W3037854933 @default.
- W4323664761 cites W3096084197 @default.
- W4323664761 cites W3099206234 @default.
- W4323664761 cites W3103152812 @default.
- W4323664761 cites W3118639295 @default.
- W4323664761 cites W3121150787 @default.
- W4323664761 cites W3126757411 @default.
- W4323664761 cites W3142516134 @default.
- W4323664761 cites W3164772801 @default.
- W4323664761 cites W3205878676 @default.
- W4323664761 cites W3206114047 @default.
- W4323664761 cites W4220961503 @default.
- W4323664761 cites W4221010269 @default.
- W4323664761 cites W4224916701 @default.
- W4323664761 cites W4224917447 @default.
- W4323664761 cites W4224931294 @default.
- W4323664761 cites W4224933346 @default.
- W4323664761 cites W4285005022 @default.
- W4323664761 cites W4297841773 @default.
- W4323664761 doi "https://doi.org/10.1016/j.eswa.2023.119833" @default.
- W4323664761 hasPublicationYear "2023" @default.
- W4323664761 type Work @default.
- W4323664761 citedByCount "0" @default.
- W4323664761 crossrefType "journal-article" @default.
- W4323664761 hasAuthorship W4323664761A5017188788 @default.
- W4323664761 hasAuthorship W4323664761A5039023891 @default.
- W4323664761 hasConcept C108583219 @default.
- W4323664761 hasConcept C132525143 @default.
- W4323664761 hasConcept C13280743 @default.
- W4323664761 hasConcept C153180895 @default.
- W4323664761 hasConcept C154945302 @default.
- W4323664761 hasConcept C185798385 @default.
- W4323664761 hasConcept C205649164 @default.
- W4323664761 hasConcept C41008148 @default.
- W4323664761 hasConcept C41608201 @default.
- W4323664761 hasConcept C59404180 @default.
- W4323664761 hasConcept C80444323 @default.
- W4323664761 hasConcept C81363708 @default.
- W4323664761 hasConcept C97931131 @default.
- W4323664761 hasConceptScore W4323664761C108583219 @default.
- W4323664761 hasConceptScore W4323664761C132525143 @default.
- W4323664761 hasConceptScore W4323664761C13280743 @default.
- W4323664761 hasConceptScore W4323664761C153180895 @default.
- W4323664761 hasConceptScore W4323664761C154945302 @default.
- W4323664761 hasConceptScore W4323664761C185798385 @default.
- W4323664761 hasConceptScore W4323664761C205649164 @default.
- W4323664761 hasConceptScore W4323664761C41008148 @default.
- W4323664761 hasConceptScore W4323664761C41608201 @default.