Matches in SemOpenAlex for { <https://semopenalex.org/work/W2892300106> ?p ?o ?g. }
Showing items 1 to 74 of
74
with 100 items per page.
- W2892300106 abstract "Voice activity detection (VAD) is the task of predicting which parts of an utterance contains speech versus background noise. It is an important first step to determine which samples to send to the decoder and when to close the microphone. The long short-term memory neural network (LSTM) is a popular architecture for sequential modeling of acoustic signals, and has been successfully used in several VAD applications. However, it has been observed that LSTMs suffer from state saturation problems when the utterance is long (i.e., for voice dictation tasks), and thus requires the LSTM state to be periodically reset. In this paper, we propose an alternative architecture that does not suffer from saturation problems by modeling temporal variations through a stateless dilated convolution neural network (CNN). The proposed architecture differs from conventional CNNs in three respects: it uses dilated causal convolution, gated activations and residual connections. Results on a Google Voice Typing task shows that the proposed architecture achieves 14% relative FA improvement at a FR of 1% over state-of-the-art LSTMs for VAD task. We also include detailed experiments investigating the factors that distinguish the proposed architecture from conventional convolution." @default.
- W2892300106 created "2018-09-27" @default.
- W2892300106 creator A5001053064 @default.
- W2892300106 creator A5001306222 @default.
- W2892300106 creator A5003562101 @default.
- W2892300106 creator A5039776447 @default.
- W2892300106 creator A5045069209 @default.
- W2892300106 creator A5050938071 @default.
- W2892300106 creator A5070513394 @default.
- W2892300106 date "2018-04-01" @default.
- W2892300106 modified "2023-09-25" @default.
- W2892300106 title "Temporal Modeling Using Dilated Convolution and Gating for Voice-Activity-Detection" @default.
- W2892300106 cites W1594494252 @default.
- W2892300106 cites W1689711448 @default.
- W2892300106 cites W2048060899 @default.
- W2892300106 cites W2064675550 @default.
- W2892300106 cites W2117671523 @default.
- W2892300106 cites W2194775991 @default.
- W2892300106 cites W2513345070 @default.
- W2892300106 cites W2617258110 @default.
- W2892300106 cites W2734724284 @default.
- W2892300106 cites W2742061524 @default.
- W2892300106 doi "https://doi.org/10.1109/icassp.2018.8461921" @default.
- W2892300106 hasPublicationYear "2018" @default.
- W2892300106 type Work @default.
- W2892300106 sameAs 2892300106 @default.
- W2892300106 citedByCount "54" @default.
- W2892300106 countsByYear W28923001062018 @default.
- W2892300106 countsByYear W28923001062019 @default.
- W2892300106 countsByYear W28923001062020 @default.
- W2892300106 countsByYear W28923001062021 @default.
- W2892300106 countsByYear W28923001062022 @default.
- W2892300106 countsByYear W28923001062023 @default.
- W2892300106 crossrefType "proceedings-article" @default.
- W2892300106 hasAuthorship W2892300106A5001053064 @default.
- W2892300106 hasAuthorship W2892300106A5001306222 @default.
- W2892300106 hasAuthorship W2892300106A5003562101 @default.
- W2892300106 hasAuthorship W2892300106A5039776447 @default.
- W2892300106 hasAuthorship W2892300106A5045069209 @default.
- W2892300106 hasAuthorship W2892300106A5050938071 @default.
- W2892300106 hasAuthorship W2892300106A5070513394 @default.
- W2892300106 hasConcept C154945302 @default.
- W2892300106 hasConcept C194544171 @default.
- W2892300106 hasConcept C28490314 @default.
- W2892300106 hasConcept C41008148 @default.
- W2892300106 hasConcept C42407357 @default.
- W2892300106 hasConcept C45347329 @default.
- W2892300106 hasConcept C50644808 @default.
- W2892300106 hasConcept C86803240 @default.
- W2892300106 hasConceptScore W2892300106C154945302 @default.
- W2892300106 hasConceptScore W2892300106C194544171 @default.
- W2892300106 hasConceptScore W2892300106C28490314 @default.
- W2892300106 hasConceptScore W2892300106C41008148 @default.
- W2892300106 hasConceptScore W2892300106C42407357 @default.
- W2892300106 hasConceptScore W2892300106C45347329 @default.
- W2892300106 hasConceptScore W2892300106C50644808 @default.
- W2892300106 hasConceptScore W2892300106C86803240 @default.
- W2892300106 hasLocation W28923001061 @default.
- W2892300106 hasOpenAccess W2892300106 @default.
- W2892300106 hasPrimaryLocation W28923001061 @default.
- W2892300106 hasRelatedWork W2312116756 @default.
- W2892300106 hasRelatedWork W2368779261 @default.
- W2892300106 hasRelatedWork W2778699561 @default.
- W2892300106 hasRelatedWork W2794438528 @default.
- W2892300106 hasRelatedWork W2893763841 @default.
- W2892300106 hasRelatedWork W2995996972 @default.
- W2892300106 hasRelatedWork W3034099097 @default.
- W2892300106 hasRelatedWork W3107474891 @default.
- W2892300106 hasRelatedWork W3128571556 @default.
- W2892300106 hasRelatedWork W4304891817 @default.
- W2892300106 isParatext "false" @default.
- W2892300106 isRetracted "false" @default.
- W2892300106 magId "2892300106" @default.
- W2892300106 workType "article" @default.