Matches in SemOpenAlex for { <https://semopenalex.org/work/W4322717169> ?p ?o ?g. }
Showing items 1 to 91 of
91
with 100 items per page.
- W4322717169 abstract "The performance of the keyword spotting (KWS) system based on audio modality, commonly measured in false alarms and false rejects, degrades significantly under the far field and noisy conditions. Therefore, audio-visual keyword spotting, which leverages complementary relationships over multiple modalities, has recently gained much attention. However, current studies mainly focus on combining the exclusively learned representations of different modalities, instead of exploring the modal relationships during each respective modeling. In this paper, we propose a novel visual modality enhanced end-to-end KWS framework (VE-KWS), which fuses audio and visual modalities from two aspects. The first one is utilizing the speaker location information obtained from the lip region in videos to assist the training of multi-channel audio beamformer. By involving the beamformer as an audio enhancement module, the acoustic distortions, caused by the far field or noisy environments, could be significantly suppressed. The other one is conducting cross-attention between different modalities to capture the inter-modal relationships and help the representation learning of each modality. Experiments on the MSIP challenge corpus show that our proposed model achieves 2.79% false rejection rate and 2.95% false alarm rate on the Eval set, resulting in a new SOTA performance compared with the top-ranking systems in the ICASSP2022 MISP challenge." @default.
- W4322717169 created "2023-03-03" @default.
- W4322717169 creator A5016304163 @default.
- W4322717169 creator A5021471823 @default.
- W4322717169 creator A5029797051 @default.
- W4322717169 creator A5029817092 @default.
- W4322717169 creator A5032367559 @default.
- W4322717169 creator A5034776912 @default.
- W4322717169 creator A5056265693 @default.
- W4322717169 creator A5073965149 @default.
- W4322717169 date "2023-02-27" @default.
- W4322717169 modified "2023-09-30" @default.
- W4322717169 title "VE-KWS: Visual Modality Enhanced End-to-End Keyword Spotting" @default.
- W4322717169 doi "https://doi.org/10.48550/arxiv.2302.13523" @default.
- W4322717169 hasPublicationYear "2023" @default.
- W4322717169 type Work @default.
- W4322717169 citedByCount "0" @default.
- W4322717169 crossrefType "posted-content" @default.
- W4322717169 hasAuthorship W4322717169A5016304163 @default.
- W4322717169 hasAuthorship W4322717169A5021471823 @default.
- W4322717169 hasAuthorship W4322717169A5029797051 @default.
- W4322717169 hasAuthorship W4322717169A5029817092 @default.
- W4322717169 hasAuthorship W4322717169A5032367559 @default.
- W4322717169 hasAuthorship W4322717169A5034776912 @default.
- W4322717169 hasAuthorship W4322717169A5056265693 @default.
- W4322717169 hasAuthorship W4322717169A5073965149 @default.
- W4322717169 hasBestOaLocation W43227171691 @default.
- W4322717169 hasConcept C120665830 @default.
- W4322717169 hasConcept C121332964 @default.
- W4322717169 hasConcept C144024400 @default.
- W4322717169 hasConcept C154945302 @default.
- W4322717169 hasConcept C177264268 @default.
- W4322717169 hasConcept C17744445 @default.
- W4322717169 hasConcept C185592680 @default.
- W4322717169 hasConcept C188027245 @default.
- W4322717169 hasConcept C192209626 @default.
- W4322717169 hasConcept C199360897 @default.
- W4322717169 hasConcept C199539241 @default.
- W4322717169 hasConcept C202444582 @default.
- W4322717169 hasConcept C2776359362 @default.
- W4322717169 hasConcept C2779506182 @default.
- W4322717169 hasConcept C2779903281 @default.
- W4322717169 hasConcept C2780226545 @default.
- W4322717169 hasConcept C2781213101 @default.
- W4322717169 hasConcept C28490314 @default.
- W4322717169 hasConcept C33923547 @default.
- W4322717169 hasConcept C36289849 @default.
- W4322717169 hasConcept C41008148 @default.
- W4322717169 hasConcept C71139939 @default.
- W4322717169 hasConcept C94625758 @default.
- W4322717169 hasConcept C9652623 @default.
- W4322717169 hasConceptScore W4322717169C120665830 @default.
- W4322717169 hasConceptScore W4322717169C121332964 @default.
- W4322717169 hasConceptScore W4322717169C144024400 @default.
- W4322717169 hasConceptScore W4322717169C154945302 @default.
- W4322717169 hasConceptScore W4322717169C177264268 @default.
- W4322717169 hasConceptScore W4322717169C17744445 @default.
- W4322717169 hasConceptScore W4322717169C185592680 @default.
- W4322717169 hasConceptScore W4322717169C188027245 @default.
- W4322717169 hasConceptScore W4322717169C192209626 @default.
- W4322717169 hasConceptScore W4322717169C199360897 @default.
- W4322717169 hasConceptScore W4322717169C199539241 @default.
- W4322717169 hasConceptScore W4322717169C202444582 @default.
- W4322717169 hasConceptScore W4322717169C2776359362 @default.
- W4322717169 hasConceptScore W4322717169C2779506182 @default.
- W4322717169 hasConceptScore W4322717169C2779903281 @default.
- W4322717169 hasConceptScore W4322717169C2780226545 @default.
- W4322717169 hasConceptScore W4322717169C2781213101 @default.
- W4322717169 hasConceptScore W4322717169C28490314 @default.
- W4322717169 hasConceptScore W4322717169C33923547 @default.
- W4322717169 hasConceptScore W4322717169C36289849 @default.
- W4322717169 hasConceptScore W4322717169C41008148 @default.
- W4322717169 hasConceptScore W4322717169C71139939 @default.
- W4322717169 hasConceptScore W4322717169C94625758 @default.
- W4322717169 hasConceptScore W4322717169C9652623 @default.
- W4322717169 hasLocation W43227171691 @default.
- W4322717169 hasOpenAccess W4322717169 @default.
- W4322717169 hasPrimaryLocation W43227171691 @default.
- W4322717169 hasRelatedWork W2323788589 @default.
- W4322717169 hasRelatedWork W2474574787 @default.
- W4322717169 hasRelatedWork W2546190447 @default.
- W4322717169 hasRelatedWork W2559372442 @default.
- W4322717169 hasRelatedWork W2742860341 @default.
- W4322717169 hasRelatedWork W2949074159 @default.
- W4322717169 hasRelatedWork W2952745240 @default.
- W4322717169 hasRelatedWork W3211385060 @default.
- W4322717169 hasRelatedWork W4298715519 @default.
- W4322717169 hasRelatedWork W4301143707 @default.
- W4322717169 isParatext "false" @default.
- W4322717169 isRetracted "false" @default.
- W4322717169 workType "article" @default.