Matches in SemOpenAlex for { <https://semopenalex.org/work/W2180339972> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W2180339972 endingPage "142" @default.
- W2180339972 startingPage "130" @default.
- W2180339972 abstract "In real-world environments, noisy utterances with variable noise levels are recorded and then converted to i-vectors for cosine distance or PLDA scoring. This paper investigates the effect of noise-level variability on i-vectors. It demonstrates that noise-level variability causes the i-vectors to shift, causing the noise contaminated i-vectors to form clusters in the i-vector space. It also demonstrates that optimal subspaces for discriminating speakers are noise-level dependent. Based on these observations, this paper proposes using signal-to-noise ratio (SNR) of utterances as guidance for training mixture of PLDA models. To maximize the coordination among the PLDA models, mixtures of PLDA models are trained simultaneously via an EM algorithm using the utterances contaminated with noise at various levels. For scoring, given a test i-vector, the marginal likelihoods from individual PLDA models are linearly combined by the posterior probabilities of the test utterance’s SNR. Verification scores are the ratio of the marginal likelihoods. Results based on NIST 2012 SRE suggest that the SNR-dependent mixture of PLDA is not only suitable for the situations where the test utterances exhibit a wide range of SNR, but also beneficial for the test utterances with unknown SNR distribution. Supplementary materials containing full derivations of the EM algorithms and scoring functions can be found in http://bioinfo.eie.polyu.edu.hk/mPLDA/SuppMaterials.pdf ." @default.
- W2180339972 created "2016-06-24" @default.
- W2180339972 creator A5031169351 @default.
- W2180339972 creator A5061908942 @default.
- W2180339972 creator A5068768998 @default.
- W2180339972 date "2016-01-01" @default.
- W2180339972 modified "2023-09-24" @default.
- W2180339972 title "Mixture of PLDA for Noise Robust I-Vector Speaker Verification" @default.
- W2180339972 cites W1968709559 @default.
- W2180339972 cites W1970872899 @default.
- W2180339972 cites W1985043720 @default.
- W2180339972 cites W1994244061 @default.
- W2180339972 cites W2002854845 @default.
- W2180339972 cites W2015549073 @default.
- W2180339972 cites W2037786411 @default.
- W2180339972 cites W2041823554 @default.
- W2180339972 cites W2074262827 @default.
- W2180339972 cites W2076818396 @default.
- W2180339972 cites W2090716422 @default.
- W2180339972 cites W2112856142 @default.
- W2180339972 cites W2121812409 @default.
- W2180339972 cites W2129066450 @default.
- W2180339972 cites W2129066856 @default.
- W2180339972 cites W2129379984 @default.
- W2180339972 cites W2136879537 @default.
- W2180339972 cites W2143054781 @default.
- W2180339972 cites W2145846419 @default.
- W2180339972 cites W2150769028 @default.
- W2180339972 cites W2395750323 @default.
- W2180339972 cites W2400085354 @default.
- W2180339972 cites W2400622930 @default.
- W2180339972 doi "https://doi.org/10.1109/taslp.2015.2499038" @default.
- W2180339972 hasPublicationYear "2016" @default.
- W2180339972 type Work @default.
- W2180339972 sameAs 2180339972 @default.
- W2180339972 citedByCount "59" @default.
- W2180339972 countsByYear W21803399722014 @default.
- W2180339972 countsByYear W21803399722016 @default.
- W2180339972 countsByYear W21803399722017 @default.
- W2180339972 countsByYear W21803399722018 @default.
- W2180339972 countsByYear W21803399722019 @default.
- W2180339972 countsByYear W21803399722020 @default.
- W2180339972 countsByYear W21803399722021 @default.
- W2180339972 countsByYear W21803399722022 @default.
- W2180339972 countsByYear W21803399722023 @default.
- W2180339972 crossrefType "journal-article" @default.
- W2180339972 hasAuthorship W2180339972A5031169351 @default.
- W2180339972 hasAuthorship W2180339972A5061908942 @default.
- W2180339972 hasAuthorship W2180339972A5068768998 @default.
- W2180339972 hasBestOaLocation W21803399722 @default.
- W2180339972 hasConcept C104317684 @default.
- W2180339972 hasConcept C115961682 @default.
- W2180339972 hasConcept C133892786 @default.
- W2180339972 hasConcept C154945302 @default.
- W2180339972 hasConcept C185592680 @default.
- W2180339972 hasConcept C28490314 @default.
- W2180339972 hasConcept C2982762665 @default.
- W2180339972 hasConcept C41008148 @default.
- W2180339972 hasConcept C55493867 @default.
- W2180339972 hasConcept C63479239 @default.
- W2180339972 hasConcept C99498987 @default.
- W2180339972 hasConceptScore W2180339972C104317684 @default.
- W2180339972 hasConceptScore W2180339972C115961682 @default.
- W2180339972 hasConceptScore W2180339972C133892786 @default.
- W2180339972 hasConceptScore W2180339972C154945302 @default.
- W2180339972 hasConceptScore W2180339972C185592680 @default.
- W2180339972 hasConceptScore W2180339972C28490314 @default.
- W2180339972 hasConceptScore W2180339972C2982762665 @default.
- W2180339972 hasConceptScore W2180339972C41008148 @default.
- W2180339972 hasConceptScore W2180339972C55493867 @default.
- W2180339972 hasConceptScore W2180339972C63479239 @default.
- W2180339972 hasConceptScore W2180339972C99498987 @default.
- W2180339972 hasIssue "1" @default.
- W2180339972 hasLocation W21803399721 @default.
- W2180339972 hasLocation W21803399722 @default.
- W2180339972 hasOpenAccess W2180339972 @default.
- W2180339972 hasPrimaryLocation W21803399721 @default.
- W2180339972 hasRelatedWork W1519782245 @default.
- W2180339972 hasRelatedWork W2121417792 @default.
- W2180339972 hasRelatedWork W2336323613 @default.
- W2180339972 hasRelatedWork W2388258002 @default.
- W2180339972 hasRelatedWork W2588576637 @default.
- W2180339972 hasRelatedWork W2950380002 @default.
- W2180339972 hasRelatedWork W3047097207 @default.
- W2180339972 hasRelatedWork W3114320604 @default.
- W2180339972 hasRelatedWork W4287692034 @default.
- W2180339972 hasRelatedWork W4288323920 @default.
- W2180339972 hasVolume "24" @default.
- W2180339972 isParatext "false" @default.
- W2180339972 isRetracted "false" @default.
- W2180339972 magId "2180339972" @default.
- W2180339972 workType "article" @default.