简介:为随机的珍视的噪音贿赂的deblurring图象,二阶段的方法首先仅仅与选择数据选择可能未来的reliables(没被随机的珍视的噪音贿赂的数据)然后deblur图象。然而,二阶段的方法经常引起有缺点的数据人工制品,它被混合数据的缺乏引起的失踪的数据人工制品和主要虚伪地选择的孤立点(被随机的珍视的噪音贿赂的数据)引起的吵闹的数据人工制品的结果。在这份报纸,压制这些有缺点的数据人工制品,我们建议一个变模糊的模型基于可靠选择的技术选择to-be-recovered象素值做所有贡献的尽可能许多选择了的reliables数据,当尽可能精确地排除孤立点时。我们也建议一种正规化技术在恢复象素价值补偿不一致的率。我们在Gaussian和斜deblurring上进行了模拟研究评估建议技术的表演。模拟结果证明那建议技术改进了二阶段的方法的性能,由有效地压制有缺点的数据人工制品。[从作者抽象]
简介:TheeffectsofmagnituderoundingandofthepresenceofnoiseintheroundedmagnitudesontheestimationoftheGutenberg–Richterb-valueareexplored,andthewaystocorrectfortheseeffectsareproposed.Fortypicalvalues,b=1androundingintervalDM=0.1,theroundingerrorisapproximately-10-3anditcanbecorrectedtoanegligibleapproximately-10-5.Forthesametypicalvalues,theeffectofnoisecanbelarger,dependingonthecharacteristicsofthenoisedistribution;fornormallydistributednoisewithstandarddeviationr=0.1,thecorrectb-valuemaybeunderestimatedbyafactor*0.97.
简介:Anideaofestimatingthedirectsequencespreadspectrum(DSSS)signalpseudo-noise(PN)sequenceispresented.WithouttheapriorityknowledgeabouttheDSSSsignalinthenon-cooperationcondition,weproposeaself-organizingfeaturemap(SOFM)neuralnetworkalgorithmtodetectandidentifythePNsequence.Anon-supervisedlearningalgorithmisproposedaccordingtheKohonenruleinSOFM.TheblindalgorithmcanalsoestimatethePNsequenceinalowsignal-to-noise(SNR)andcomputersimulationdemonstratesthatthealgorithmiseffective.Comparedwiththetraditionalcorrelationalgorithmbasedonslip-correlation,theproposedalgorithm’sbiterrorrate(BER)andcomplexityarelower.
简介:Objective:TodemonstratetheperformancebenefitoftheAutomaticSceneClassifier(SCAN)algorithmavailableintheNucleus6(CP900series)soundprocessoroverthedefaultprocessingalgorithmsofthepreviousgenerationNucleus5(CP810)andFreedomHybridTMsoundprocessors.Methods:Eighty-twocochlearimplantrecipients(40Nucleus5processorusersand42FreedomHybridprocessorusers)listenedtoandrepeatedAzBiosentencesinnoisewiththeircurrentprocessorandwiththeNucleus6processor.Results:TheSCANalgorithmwhenenabledyieldedstatisticallysignificantnon-inferiorandsuperiorperformancewhencomparedtotheNucleus5andFreedomHybridsoundprocessorsprogrammedwithASCtADRO.Conclusion:TheresultsofthesestudiesdemonstratethesuperiorperformanceandclinicalutilityoftheSCANalgorithmintheNucleus6processorovertheNucleus5andFreedomHybridprocessors.
简介:Noisepollutionisamajorhazardousfactortohumanhealthandislikelyharmfulforvulnerablegroupssuchaspre-terminfantsunderlifesupportsysteminanintensivecareunit.Previousstudieshavesuggestedthatnoiseexposureimpairschildren’slearningabilityandcognitiveperformanceandcognitivefunctionsinanimalmodelsinwhichtheeffectismainlyattributedtotheoxidantstressofnoiseonthecognitivebrain.Thepotentialroleofnoiseinducedhearingloss(NIHL),ratherthantheoxidantstress,hasalsobeenindicatedbyadepressionofneurogenesisinthehippocampuslongafterabriefnoiseexposure,whichproducesonlyatentativeoxidantstress.ItisnotclearifnoiseexposureandNIHLduringearlydevelopmentexertsalongtermimpactoncognitivefunctionandneurogenesistowardsadulthood.Inthepresentstudy,abriefnoiseexposureathighsoundlevelwasperformedinneonatalC57BL/6Jmice(15daysafterbirth)toproduceasignificantamountofpermanenthearinglossasproved2monthsafterthenoise.Atthisage,thenoise-exposedanimalsshoweddeterioratedspatiallearningandmemoryabilitiesandareductionofhippocampalneurogenesisascomparedwiththecontrol.Theaveragedhearingthresholdwasfoundtobestronglycorrelatedwiththescoresforspatiallearningandmemory.Weconsidertheeffectsobservedarelargelyduetothelossofhearingsensitivity,ratherthantheoxidantstress,duetothelongintervalbetweennoiseexposureandtheobservations.
简介:Olderadultsoftenfinditdifficulttoperceivespeech,especiallyinnoisyconditions.Thoughhearingaidisoneoftherehabilitativedevicesavailabletoolderadultstoalleviatehearingloss,someofthemmayexperienceannoyancethroughhearingaidandhencerejectit.maybeduetocircuitrynoiseand/orbackgroundnoise.Acceptablenoiselevelisadirectbehaviouralmeasuretoestimatetheextentofhowmuchapersonisabletoputupwithnoisewhilesimultaneouslylisteningtospeech.Acceptablenoiselevelisacentralauditorymeasureanditisnotinfluencedbyage.gender,presentationlevelorspeaker.Usingthismeasure,wecanquantifytheannoyancelevelexperiencedbyanindividual.Thisinformationisofutmostimportanceandcautionshouldbepaidbeforesettingtheparametersinhearingaid,especiallyforthosewhoareunabletoacceptnoise.Inthisreviewarticle,anattempthasbeenmadetodocumenthowtooptimizethehearingaidprogrambysettingparameterssuchasnoisereductioncircuit,microphonesensitivityandgain.Theseadjustmentsofparametersmighthelptoreducerejectionrateofhearingaids,especiallyinthoseindividualswhoareannoyedbybackgroundnoise.