FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2891283181> ?p ?o ?g. }

• W2891283181 abstract "Speech separation is a fundamental problem in speech and signal processing. A particular challenge is monaural separation of cochannel speech, or a two-talker mixture, in a reverberant environment. In this paper, we study recurrent neural networks (RNNs) with long short-term memory (LSTM) in separating and enhancing speech signals in reverberant cochannel mixtures. Our investigation shows that RNNs are effective in separating reverberant speech signals. In addition, RNNs significantly outperform deep feedforward networks based on objective speech intelligibility and quality measures. We also find that the best performance is achieved when the ideal ratio mask (IRM) is used as the training target in comparison with alternative training targets. While trained using reverberant signals generated by simulated room impulse responses (RIRs), our model generalizes well to conditions where the signals are generated by recorded RIRs." @default.
• W2891283181 created "2018-09-27" @default.
• W2891283181 creator A5012541772 @default.
• W2891283181 creator A5051837453 @default.
• W2891283181 date "2018-04-01" @default.
• W2891283181 modified "2023-09-26" @default.
• W2891283181 title "Recurrent Neural Networks for Cochannel Speech Separation in Reverberant Environments" @default.
• W2891283181 cites W1482149378 @default.
• W2891283181 cites W1897240248 @default.
• W2891283181 cites W2006129368 @default.
• W2891283181 cites W2031647436 @default.
• W2891283181 cites W2069681747 @default.
• W2891283181 cites W2087368178 @default.
• W2891283181 cites W2114218864 @default.
• W2891283181 cites W2114719288 @default.
• W2891283181 cites W2125562558 @default.
• W2891283181 cites W2136922672 @default.
• W2891283181 cites W2144786298 @default.
• W2891283181 cites W2152037357 @default.
• W2891283181 cites W2221409856 @default.
• W2891283181 cites W2304609584 @default.
• W2891283181 cites W2404892923 @default.
• W2891283181 cites W2558649592 @default.
• W2891283181 cites W2598036111 @default.
• W2891283181 cites W2734774145 @default.
• W2891283181 cites W2964121744 @default.
• W2891283181 doi "https://doi.org/10.1109/icassp.2018.8462014" @default.
• W2891283181 hasPublicationYear "2018" @default.
• W2891283181 type Work @default.
• W2891283181 sameAs 2891283181 @default.
• W2891283181 citedByCount "6" @default.
• W2891283181 countsByYear W28912831812019 @default.
• W2891283181 countsByYear W28912831812020 @default.
• W2891283181 countsByYear W28912831812021 @default.
• W2891283181 crossrefType "proceedings-article" @default.
• W2891283181 hasAuthorship W2891283181A5012541772 @default.
• W2891283181 hasAuthorship W2891283181A5051837453 @default.
• W2891283181 hasConcept C102894143 @default.
• W2891283181 hasConcept C111472728 @default.
• W2891283181 hasConcept C121332964 @default.
• W2891283181 hasConcept C127413603 @default.
• W2891283181 hasConcept C133731056 @default.
• W2891283181 hasConcept C134306372 @default.
• W2891283181 hasConcept C138885662 @default.
• W2891283181 hasConcept C147168706 @default.
• W2891283181 hasConcept C153180895 @default.
• W2891283181 hasConcept C154945302 @default.
• W2891283181 hasConcept C163294075 @default.
• W2891283181 hasConcept C24890656 @default.
• W2891283181 hasConcept C2776182073 @default.
• W2891283181 hasConcept C2776864781 @default.
• W2891283181 hasConcept C28490314 @default.
• W2891283181 hasConcept C33923547 @default.
• W2891283181 hasConcept C38858127 @default.
• W2891283181 hasConcept C41008148 @default.
• W2891283181 hasConcept C50644808 @default.
• W2891283181 hasConcept C60048801 @default.
• W2891283181 hasConcept C61328038 @default.
• W2891283181 hasConcept C62520636 @default.
• W2891283181 hasConcept C70836080 @default.
• W2891283181 hasConcept C72279823 @default.
• W2891283181 hasConcept C95851461 @default.
• W2891283181 hasConceptScore W2891283181C102894143 @default.
• W2891283181 hasConceptScore W2891283181C111472728 @default.
• W2891283181 hasConceptScore W2891283181C121332964 @default.
• W2891283181 hasConceptScore W2891283181C127413603 @default.
• W2891283181 hasConceptScore W2891283181C133731056 @default.
• W2891283181 hasConceptScore W2891283181C134306372 @default.
• W2891283181 hasConceptScore W2891283181C138885662 @default.
• W2891283181 hasConceptScore W2891283181C147168706 @default.
• W2891283181 hasConceptScore W2891283181C153180895 @default.
• W2891283181 hasConceptScore W2891283181C154945302 @default.
• W2891283181 hasConceptScore W2891283181C163294075 @default.
• W2891283181 hasConceptScore W2891283181C24890656 @default.
• W2891283181 hasConceptScore W2891283181C2776182073 @default.
• W2891283181 hasConceptScore W2891283181C2776864781 @default.
• W2891283181 hasConceptScore W2891283181C28490314 @default.
• W2891283181 hasConceptScore W2891283181C33923547 @default.
• W2891283181 hasConceptScore W2891283181C38858127 @default.
• W2891283181 hasConceptScore W2891283181C41008148 @default.
• W2891283181 hasConceptScore W2891283181C50644808 @default.
• W2891283181 hasConceptScore W2891283181C60048801 @default.
• W2891283181 hasConceptScore W2891283181C61328038 @default.
• W2891283181 hasConceptScore W2891283181C62520636 @default.
• W2891283181 hasConceptScore W2891283181C70836080 @default.
• W2891283181 hasConceptScore W2891283181C72279823 @default.
• W2891283181 hasConceptScore W2891283181C95851461 @default.
• W2891283181 hasLocation W28912831811 @default.
• W2891283181 hasOpenAccess W2891283181 @default.
• W2891283181 hasPrimaryLocation W28912831811 @default.
• W2891283181 hasRelatedWork W1992879732 @default.
• W2891283181 hasRelatedWork W2006129368 @default.
• W2891283181 hasRelatedWork W2027804983 @default.
• W2891283181 hasRelatedWork W2094461119 @default.
• W2891283181 hasRelatedWork W2145037686 @default.
• W2891283181 hasRelatedWork W2587989377 @default.
• W2891283181 hasRelatedWork W2598036111 @default.
• W2891283181 hasRelatedWork W2609317876 @default.
• W2891283181 hasRelatedWork W2635471793 @default.
• W2891283181 hasRelatedWork W2711335087 @default.

• next