Matches in SemOpenAlex for { <https://semopenalex.org/work/W2967913447> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W2967913447 abstract "In this paper, we propose personal VAD, a system to detect the voice activity of a target speaker at the frame level. This system is useful for gating the inputs to a streaming on-device speech recognition system, such that it only triggers for the target user, which helps reduce the computational cost and battery consumption, especially in scenarios where a keyword detector is unpreferable. We achieve this by training a VAD-alike neural network that is conditioned on the target speaker embedding or the speaker verification score. For each frame, personal VAD outputs the probabilities for three classes: non-speech, target speaker speech, and non-target speaker speech. Under our optimal setup, we are able to train a model with only 130K parameters that outperforms a baseline system where individually trained standard VAD and speaker recognition networks are combined to perform the same task." @default.
- W2967913447 created "2019-08-22" @default.
- W2967913447 creator A5001306222 @default.
- W2967913447 creator A5050898122 @default.
- W2967913447 creator A5058886181 @default.
- W2967913447 creator A5063131315 @default.
- W2967913447 creator A5091363775 @default.
- W2967913447 date "2019-08-12" @default.
- W2967913447 modified "2023-10-16" @default.
- W2967913447 title "Personal VAD: Speaker-Conditioned Voice Activity Detection" @default.
- W2967913447 cites W1494198834 @default.
- W2967913447 cites W1536583098 @default.
- W2967913447 cites W1821462560 @default.
- W2967913447 cites W1999454387 @default.
- W2967913447 cites W2023582935 @default.
- W2967913447 cites W2034940213 @default.
- W2967913447 cites W2062335243 @default.
- W2967913447 cites W2240641835 @default.
- W2967913447 cites W2295098554 @default.
- W2967913447 cites W2497335382 @default.
- W2967913447 cites W2612434969 @default.
- W2967913447 cites W2617258110 @default.
- W2967913447 cites W2625979394 @default.
- W2967913447 cites W2696967604 @default.
- W2967913447 cites W2726515241 @default.
- W2967913447 cites W2742061524 @default.
- W2967913447 cites W2775336875 @default.
- W2967913447 cites W2787752687 @default.
- W2967913447 cites W2890964092 @default.
- W2967913447 cites W2892300106 @default.
- W2967913447 cites W2896538040 @default.
- W2967913447 cites W2902094805 @default.
- W2967913447 cites W2962760690 @default.
- W2967913447 cites W2962788625 @default.
- W2967913447 cites W2963432880 @default.
- W2967913447 cites W2963470929 @default.
- W2967913447 cites W2963912924 @default.
- W2967913447 cites W2964121744 @default.
- W2967913447 cites W2972495969 @default.
- W2967913447 cites W2973062255 @default.
- W2967913447 doi "https://doi.org/10.48550/arxiv.1908.04284" @default.
- W2967913447 hasPublicationYear "2019" @default.
- W2967913447 type Work @default.
- W2967913447 sameAs 2967913447 @default.
- W2967913447 citedByCount "0" @default.
- W2967913447 crossrefType "posted-content" @default.
- W2967913447 hasAuthorship W2967913447A5001306222 @default.
- W2967913447 hasAuthorship W2967913447A5050898122 @default.
- W2967913447 hasAuthorship W2967913447A5058886181 @default.
- W2967913447 hasAuthorship W2967913447A5063131315 @default.
- W2967913447 hasAuthorship W2967913447A5091363775 @default.
- W2967913447 hasBestOaLocation W29679134471 @default.
- W2967913447 hasConcept C126042441 @default.
- W2967913447 hasConcept C127413603 @default.
- W2967913447 hasConcept C133892786 @default.
- W2967913447 hasConcept C149838564 @default.
- W2967913447 hasConcept C154945302 @default.
- W2967913447 hasConcept C201995342 @default.
- W2967913447 hasConcept C204201278 @default.
- W2967913447 hasConcept C2780451532 @default.
- W2967913447 hasConcept C28490314 @default.
- W2967913447 hasConcept C31258907 @default.
- W2967913447 hasConcept C41008148 @default.
- W2967913447 hasConcept C41608201 @default.
- W2967913447 hasConcept C61328038 @default.
- W2967913447 hasConceptScore W2967913447C126042441 @default.
- W2967913447 hasConceptScore W2967913447C127413603 @default.
- W2967913447 hasConceptScore W2967913447C133892786 @default.
- W2967913447 hasConceptScore W2967913447C149838564 @default.
- W2967913447 hasConceptScore W2967913447C154945302 @default.
- W2967913447 hasConceptScore W2967913447C201995342 @default.
- W2967913447 hasConceptScore W2967913447C204201278 @default.
- W2967913447 hasConceptScore W2967913447C2780451532 @default.
- W2967913447 hasConceptScore W2967913447C28490314 @default.
- W2967913447 hasConceptScore W2967913447C31258907 @default.
- W2967913447 hasConceptScore W2967913447C41008148 @default.
- W2967913447 hasConceptScore W2967913447C41608201 @default.
- W2967913447 hasConceptScore W2967913447C61328038 @default.
- W2967913447 hasLocation W29679134471 @default.
- W2967913447 hasOpenAccess W2967913447 @default.
- W2967913447 hasPrimaryLocation W29679134471 @default.
- W2967913447 hasRelatedWork W2020970176 @default.
- W2967913447 hasRelatedWork W2059891707 @default.
- W2967913447 hasRelatedWork W3025260599 @default.
- W2967913447 hasRelatedWork W3087422378 @default.
- W2967913447 hasRelatedWork W3163933965 @default.
- W2967913447 hasRelatedWork W3179294409 @default.
- W2967913447 hasRelatedWork W3198543387 @default.
- W2967913447 hasRelatedWork W4226428303 @default.
- W2967913447 hasRelatedWork W4287102200 @default.
- W2967913447 hasRelatedWork W4292862526 @default.
- W2967913447 isParatext "false" @default.
- W2967913447 isRetracted "false" @default.
- W2967913447 magId "2967913447" @default.
- W2967913447 workType "article" @default.