Matches in SemOpenAlex for { <https://semopenalex.org/work/W3094189801> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W3094189801 abstract "Robust voice activity detection (VAD) is a challenging task in low signal-to-noise (SNR) environments. Recent studies show that speech enhancement is helpful to VAD, but the performance improvement is limited. To address this issue, here we propose a speech enhancement aided end-to-end multi-task model for VAD. The model has two decoders, one for speech enhancement and the other for VAD. The two decoders share the same encoder and speech separation network. Unlike the direct thought that takes two separated objectives for VAD and speech enhancement respectively, here we propose a new joint optimization objective -- VAD-masked scale-invariant source-to-distortion ratio (mSI-SDR). mSI-SDR uses VAD information to mask the output of the speech enhancement decoder in the training process. It makes the VAD and speech enhancement tasks jointly optimized not only at the shared encoder and separation network, but also at the objective level. It also satisfies real-time working requirement theoretically. Experimental results show that the multi-task method significantly outperforms its single-task VAD counterpart. Moreover, mSI-SDR outperforms SI-SDR in the same multi-task setting." @default.
- W3094189801 created "2020-10-29" @default.
- W3094189801 creator A5018286848 @default.
- W3094189801 creator A5082846858 @default.
- W3094189801 date "2020-10-23" @default.
- W3094189801 modified "2023-10-16" @default.
- W3094189801 title "Speech enhancement aided end-to-end multi-task learning for voice activity detection" @default.
- W3094189801 cites W1522301498 @default.
- W3094189801 cites W1974387177 @default.
- W3094189801 cites W1985242443 @default.
- W3094189801 cites W1989364685 @default.
- W3094189801 cites W2024490156 @default.
- W3094189801 cites W2048497537 @default.
- W3094189801 cites W2059203007 @default.
- W3094189801 cites W2067295501 @default.
- W3094189801 cites W2098265087 @default.
- W3094189801 cites W2109000787 @default.
- W3094189801 cites W2197404611 @default.
- W3094189801 cites W2289394825 @default.
- W3094189801 cites W2396495723 @default.
- W3094189801 cites W2513345070 @default.
- W3094189801 cites W2612770767 @default.
- W3094189801 cites W2791616807 @default.
- W3094189801 cites W2797121075 @default.
- W3094189801 cites W2889224635 @default.
- W3094189801 cites W2892300106 @default.
- W3094189801 cites W2917987043 @default.
- W3094189801 cites W2952218014 @default.
- W3094189801 cites W2972861996 @default.
- W3094189801 cites W3011405319 @default.
- W3094189801 cites W3022404557 @default.
- W3094189801 cites W3036462239 @default.
- W3094189801 cites W3099628266 @default.
- W3094189801 doi "https://doi.org/10.48550/arxiv.2010.12484" @default.
- W3094189801 hasPublicationYear "2020" @default.
- W3094189801 type Work @default.
- W3094189801 sameAs 3094189801 @default.
- W3094189801 citedByCount "1" @default.
- W3094189801 countsByYear W30941898012021 @default.
- W3094189801 crossrefType "posted-content" @default.
- W3094189801 hasAuthorship W3094189801A5018286848 @default.
- W3094189801 hasAuthorship W3094189801A5082846858 @default.
- W3094189801 hasBestOaLocation W30941898011 @default.
- W3094189801 hasConcept C111919701 @default.
- W3094189801 hasConcept C118505674 @default.
- W3094189801 hasConcept C127413603 @default.
- W3094189801 hasConcept C154945302 @default.
- W3094189801 hasConcept C163294075 @default.
- W3094189801 hasConcept C201995342 @default.
- W3094189801 hasConcept C204201278 @default.
- W3094189801 hasConcept C2776182073 @default.
- W3094189801 hasConcept C2780451532 @default.
- W3094189801 hasConcept C28490314 @default.
- W3094189801 hasConcept C41008148 @default.
- W3094189801 hasConcept C61328038 @default.
- W3094189801 hasConcept C74296488 @default.
- W3094189801 hasConceptScore W3094189801C111919701 @default.
- W3094189801 hasConceptScore W3094189801C118505674 @default.
- W3094189801 hasConceptScore W3094189801C127413603 @default.
- W3094189801 hasConceptScore W3094189801C154945302 @default.
- W3094189801 hasConceptScore W3094189801C163294075 @default.
- W3094189801 hasConceptScore W3094189801C201995342 @default.
- W3094189801 hasConceptScore W3094189801C204201278 @default.
- W3094189801 hasConceptScore W3094189801C2776182073 @default.
- W3094189801 hasConceptScore W3094189801C2780451532 @default.
- W3094189801 hasConceptScore W3094189801C28490314 @default.
- W3094189801 hasConceptScore W3094189801C41008148 @default.
- W3094189801 hasConceptScore W3094189801C61328038 @default.
- W3094189801 hasConceptScore W3094189801C74296488 @default.
- W3094189801 hasLocation W30941898011 @default.
- W3094189801 hasOpenAccess W3094189801 @default.
- W3094189801 hasPrimaryLocation W30941898011 @default.
- W3094189801 hasRelatedWork W2151333624 @default.
- W3094189801 hasRelatedWork W2884250895 @default.
- W3094189801 hasRelatedWork W2950108968 @default.
- W3094189801 hasRelatedWork W2963453742 @default.
- W3094189801 hasRelatedWork W3094189801 @default.
- W3094189801 hasRelatedWork W3115290769 @default.
- W3094189801 hasRelatedWork W3160071434 @default.
- W3094189801 hasRelatedWork W3163142165 @default.
- W3094189801 hasRelatedWork W4225287045 @default.
- W3094189801 hasRelatedWork W4287329374 @default.
- W3094189801 isParatext "false" @default.
- W3094189801 isRetracted "false" @default.
- W3094189801 magId "3094189801" @default.
- W3094189801 workType "article" @default.