Matches in SemOpenAlex for { <https://semopenalex.org/work/W1982172017> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W1982172017 endingPage "106" @default.
- W1982172017 startingPage "97" @default.
- W1982172017 abstract "A neural network for time-frequency mask prediction is proposed.The network is trained using simulated speech to produce naturally occurring masks.After a post-processing stage, the mask is used as a post-filter of a beamformer.Speech mixtures recorded with a circular array in two rooms are separated.The method shows best intelligibility and SNR values compared to contrast methods. Speech separation algorithms are faced with a difficult task of producing high degree of separation without containing unwanted artifacts. The time-frequency (T-F) masking technique applies a real-valued (or binary) mask on top of the signal's spectrum to filter out unwanted components. The practical difficulty lies in the mask estimation. Often, using efficient masks engineered for separation performance leads to presence of unwanted musical noise artifacts in the separated signal. This lowers the perceptual quality and intelligibility of the output.Microphone arrays have been long studied for processing of distant speech. This work uses a feed-forward neural network for mapping microphone array's spatial features into a T-F mask. Wiener filter is used as a desired mask for training the neural network using speech examples in simulated setting. The T-F masks predicted by the neural network are combined to obtain an enhanced separation mask that exploits the information regarding interference between all sources. The final mask is applied to the delay-and-sum beamformer (DSB) output.The algorithm's objective separation capability in conjunction with the separated speech intelligibility is tested with recorded speech from distant talkers in two rooms from two distances. The results show improvement in instrumental measure for intelligibility and frequency-weighted SNR over complex-valued non-negative matrix factorization (CNMF) source separation approach, spatial sound source separation, and conventional beamforming methods such as the DSB and minimum variance distortionless response (MVDR)." @default.
- W1982172017 created "2016-06-24" @default.
- W1982172017 creator A5057962401 @default.
- W1982172017 creator A5075555953 @default.
- W1982172017 date "2015-04-01" @default.
- W1982172017 modified "2023-09-26" @default.
- W1982172017 title "Distant speech separation using predicted time–frequency masks from spatial features" @default.
- W1982172017 cites W146976060 @default.
- W1982172017 cites W1963970749 @default.
- W1982172017 cites W1987906574 @default.
- W1982172017 cites W2004222308 @default.
- W1982172017 cites W2013608223 @default.
- W1982172017 cites W2021196544 @default.
- W1982172017 cites W2027884847 @default.
- W1982172017 cites W2034040413 @default.
- W1982172017 cites W2054221677 @default.
- W1982172017 cites W2057200980 @default.
- W1982172017 cites W2060741034 @default.
- W1982172017 cites W2064949872 @default.
- W1982172017 cites W2065115983 @default.
- W1982172017 cites W2065660120 @default.
- W1982172017 cites W2068175049 @default.
- W1982172017 cites W2073006643 @default.
- W1982172017 cites W2086139506 @default.
- W1982172017 cites W2093010905 @default.
- W1982172017 cites W2094461119 @default.
- W1982172017 cites W2096855653 @default.
- W1982172017 cites W2099655464 @default.
- W1982172017 cites W2111732517 @default.
- W1982172017 cites W2113990625 @default.
- W1982172017 cites W2117678320 @default.
- W1982172017 cites W2123649031 @default.
- W1982172017 cites W2127851351 @default.
- W1982172017 cites W2135158232 @default.
- W1982172017 cites W2135823751 @default.
- W1982172017 cites W2141188634 @default.
- W1982172017 cites W2141411743 @default.
- W1982172017 cites W2141520175 @default.
- W1982172017 cites W2141892084 @default.
- W1982172017 cites W2141998673 @default.
- W1982172017 cites W2144404214 @default.
- W1982172017 cites W2149693148 @default.
- W1982172017 cites W2152919025 @default.
- W1982172017 cites W2168273590 @default.
- W1982172017 cites W2168379380 @default.
- W1982172017 cites W2290318471 @default.
- W1982172017 cites W2401485387 @default.
- W1982172017 cites W2403228223 @default.
- W1982172017 cites W4231807801 @default.
- W1982172017 doi "https://doi.org/10.1016/j.specom.2015.01.006" @default.
- W1982172017 hasPublicationYear "2015" @default.
- W1982172017 type Work @default.
- W1982172017 sameAs 1982172017 @default.
- W1982172017 citedByCount "34" @default.
- W1982172017 countsByYear W19821720172015 @default.
- W1982172017 countsByYear W19821720172016 @default.
- W1982172017 countsByYear W19821720172017 @default.
- W1982172017 countsByYear W19821720172018 @default.
- W1982172017 countsByYear W19821720172019 @default.
- W1982172017 countsByYear W19821720172020 @default.
- W1982172017 countsByYear W19821720172021 @default.
- W1982172017 countsByYear W19821720172022 @default.
- W1982172017 crossrefType "journal-article" @default.
- W1982172017 hasAuthorship W1982172017A5057962401 @default.
- W1982172017 hasAuthorship W1982172017A5075555953 @default.
- W1982172017 hasBestOaLocation W19821720172 @default.
- W1982172017 hasConcept C119857082 @default.
- W1982172017 hasConcept C154945302 @default.
- W1982172017 hasConcept C2776061190 @default.
- W1982172017 hasConcept C28490314 @default.
- W1982172017 hasConcept C41008148 @default.
- W1982172017 hasConceptScore W1982172017C119857082 @default.
- W1982172017 hasConceptScore W1982172017C154945302 @default.
- W1982172017 hasConceptScore W1982172017C2776061190 @default.
- W1982172017 hasConceptScore W1982172017C28490314 @default.
- W1982172017 hasConceptScore W1982172017C41008148 @default.
- W1982172017 hasLocation W19821720171 @default.
- W1982172017 hasLocation W19821720172 @default.
- W1982172017 hasOpenAccess W1982172017 @default.
- W1982172017 hasPrimaryLocation W19821720171 @default.
- W1982172017 hasRelatedWork W2066106687 @default.
- W1982172017 hasRelatedWork W2312116756 @default.
- W1982172017 hasRelatedWork W2368779261 @default.
- W1982172017 hasRelatedWork W2374918184 @default.
- W1982172017 hasRelatedWork W2778699561 @default.
- W1982172017 hasRelatedWork W2794438528 @default.
- W1982172017 hasRelatedWork W2893763841 @default.
- W1982172017 hasRelatedWork W2995996972 @default.
- W1982172017 hasRelatedWork W3128571556 @default.
- W1982172017 hasRelatedWork W4304891817 @default.
- W1982172017 hasVolume "68" @default.
- W1982172017 isParatext "false" @default.
- W1982172017 isRetracted "false" @default.
- W1982172017 magId "1982172017" @default.
- W1982172017 workType "article" @default.