Face recognition

Developmenthistory

Theresearchoffacerecognitionsystembeganinthe1960s,andafterthe1980s,withthedevelopmentofcomputertechnologyandopticalimagingtechnology,ithasenteredtheprimaryapplicationThestageisinthelate1990s,andismainlybasedonthetechnologyimplementationoftheUnitedStates,GermanyandJapan;thekeytothesuccessofthefacerecognitionsystemiswhetherithasacutting-edgecorealgorithm,andtherecognitionresulthasapracticalrecognitionrateandrecognitionspeed;"Facerecognitionsystem"integratesmanyprofessionaltechnologiessuchasartificialintelligence,machinerecognition,machinelearning,modeltheory,expertsystem,videoimageprocessing,etc.Italsoneedstocombinethetheoryandimplementationofintermediatevalueprocessing.Itisthelatestapplicationofbiometricrecognition.Therealizationofitscoretechnologyshowsthetransformationfromweakartificialintelligencetostrongartificialintelligence.

Technicalfeatures

Traditionalfacerecognitiontechnologyismainlybasedonvisiblelightimagefacerecognition,whichisalsoafamiliarrecognitionmethodthathasbeendevelopedformorethan30years.However,thismethodhasinsurmountabledefects,especiallywhentheambientlightchanges,therecognitioneffectwilldropsharply,whichcannotmeettheneedsoftheactualsystem.Solutionstothelightingproblemincludethree-dimensionalimagefacerecognitionandthermalimagingfacerecognition.Butthesetwotechnologiesarefarfrommature,andtherecognitioneffectisnotsatisfactory.

Arapidlydevelopingsolutionisthemulti-lightsourcefacerecognitiontechnologybasedonactivenear-infraredimages.Itcanovercometheinfluenceoflightchangesandhasachievedexcellentrecognitionperformance.Theoverallsystemperformanceintermsofaccuracy,stabilityandspeedexceedsthatof3Dimagefacerecognition.Thistechnologyhasdevelopedrapidlyinthepasttwotothreeyears,makingfacerecognitiontechnologygraduallypractical.

Humanfaceisinnatelikeotherbiologicalcharacteristicsofthehumanbody(fingerprints,iris,etc.).Itsuniquenessandgoodcharacteristicsthatarenoteasytobecopiedprovidethenecessaryprerequisitesforidentityauthentication.Comparedwithbiometricrecognition,facerecognitionhasthefollowingcharacteristics:

Optional:Theuserdoesnotneedtospecificallycooperatewithfacecollectionequipment,andcanobtainfaceimagesalmostunconsciously.ThissamplingmethodThereisno“mandatory”;

Non-contact:userscanobtainfaceimageswithoutdirectcontactwiththedevice;

Concurrency:multiplepeoplecanbeusedinactualapplicationscenariosThesorting,judgmentandrecognitionoffaces;

Inaddition,italsoconformstothevisualcharacteristics:thecharacteristicsof"recognizingpeoplebyappearance",andthecharacteristicsofsimpleoperation,intuitiveresults,andgoodconcealment.

Technicalprocess

Thefacerecognitionsystemmainlyincludesfourcomponents,namely:faceimageacquisitionanddetection,faceimagepreprocessing,faceimagefeatureextractionandmatchingAndrecognition.

Faceimagecollectionanddetection

Faceimagecollection:Differentfaceimagescanbecollectedthroughthecameralens,suchasstaticimages,dynamicimages,differentpositions,differentExpressionsandotheraspectscanbewellcollected.Whentheuseriswithintheshootingrangeofthecapturedevice,thecapturedevicewillautomaticallysearchforandcapturetheuser'sfaceimage.

Facedetection:Inpractice,facedetectionismainlyusedforpreprocessingoffacerecognition,thatis,toaccuratelycalibratethepositionandsizeofthefaceintheimage.Thepatternfeaturescontainedinfaceimagesareveryrich,suchashistogramfeatures,colorfeatures,templatefeatures,structuralfeatures,andHaarfeatures.Facedetectionistopickouttheusefulinformationandusethesefeaturestorealizefacedetection.

ThemainstreamfacedetectionmethodusestheAdaboostlearningalgorithmbasedontheabovefeatures.TheAdaboostalgorithmisamethodforclassification.ItcombinessomeweakerclassificationmethodstoformanewstrongTheclassificationmethod.

Inthefacedetectionprocess,theAdaboostalgorithmisusedtoselectsomerectangularfeatures(weakclassifiers)thatbestrepresenttheface,andtheweakclassifierisconstructedintoastrongclassifieraccordingtotheweightedvotingmethod,andthentheSeveralstrongclassifiersobtainedbytrainingareconnectedinseriestoformacascadestructuredcascadedclassifier,whicheffectivelyimprovesthedetectionspeedoftheclassifier.

Faceimagepreprocessing

Faceimagepreprocessing:Theimagepreprocessingofthefaceisbasedontheresultoffacedetection,theimageisprocessedandfinallyservedforfeatureextractionprocess.Theoriginalimageacquiredbythesystemisoftennotdirectlyusedduetovariousconditionsandrandominterference.Itmustbepre-processedintheearlystagesofimageprocessingsuchasgray-scalecorrectionandnoisefiltering.Forfaceimages,thepreprocessingprocessmainlyincludeslightcompensation,grayscaletransformation,histogramequalization,normalization,geometriccorrection,filteringandsharpeningofthefaceimage.

Faceimagefeatureextraction

Faceimagefeatureextraction:Thefeaturesthatcanbeusedinfacerecognitionsystemsareusuallydividedintovisualfeatures,pixelstatisticalfeatures,faceimagetransformationcoefficientfeatures,Thealgebraicfeaturesoffaceimages,etc.Facialfeatureextractionisbasedoncertainfeaturesofthehumanface.Facefeatureextraction,alsoknownasfacerepresentation,istheprocessoffeaturemodelingofhumanfaces.Themethodsoffacefeatureextractioncanbesummarizedintotwocategories:oneistherepresentationmethodbasedonknowledge;theotheristherepresentationmethodbasedonalgebraicfeaturesorstatisticallearning.

Theknowledge-basedrepresentationmethodismainlybasedonthedescriptionoftheshapeofthefaceorgansandthedistancecharacteristicsbetweenthemtoobtainfeaturedatathatishelpfulforfaceclassification.ThefeaturecomponentsusuallyincludetheOuDistance,curvatureandangle,etc.Thehumanfaceiscomposedofpartssuchaseyes,nose,mouth,andchin.Thegeometricdescriptionofthesepartsandthestructuralrelationshipbetweenthemcanbeusedasimportantfeaturestorecognizetheface.Thesefeaturesarecalledgeometricfeatures.Knowledge-basedfacerepresentationmainlyincludesgeometricfeature-basedmethodsandtemplatematchingmethods.

Faceimagematchingandrecognition

Faceimagematchingandrecognition:Thefeaturedataoftheextractedfaceimageissearchedandmatchedwiththefeaturetemplatestoredinthedatabase.BysettingaThreshold,whenthesimilarityexceedsthisthreshold,thematchingresultisoutput.Facerecognitionistocomparethefacefeaturestoberecognizedwiththeobtainedfacefeaturetemplates,andjudgetheidentityinformationofthefaceaccordingtothedegreeofsimilarity.Thisprocessisdividedintotwocategories:oneisconfirmation,whichisaone-to-oneimagecomparisonprocess,andtheotherisidentification,whichisaone-to-manyimagematchingandcomparisonprocess.

Recognitionalgorithm

Generallyspeaking,thefacerecognitionsystemincludesimagecapture,facepositioning,imagepreprocessing,andfacerecognition(identityconfirmationoridentitysearch).Theinputofthesystemisgenerallyoneoraseriesoffaceimageswithundeterminedidentities,andseveralfaceimageswithknownidentitiesinthefacedatabaseorcorrespondingcodes,andtheoutputisaseriesofsimilarityscores,indicatingTheidentityofthefacetoberecognized.

Facerecognitionalgorithmclassification

Feature-basedrecognitionalgorithms(Feature-basedrecognitionalgorithms).

Appearance-basedrecognitionalgorithms(Appearance-basedrecognitionalgorithms).

Template-basedrecognitionalgorithms.

Recognitionalgorithmsusingneuralnetwork.

Basedonthelightestimationmodeltheory

ProposesalightpreprocessingmethodbasedonGammagraycorrection,andbasedonthelightestimationmodel,correspondingLightcompensationandlightbalancestrategy.

Optimizeddeformationstatisticalcorrectiontheory

Basedonstatisticaldeformationcorrectiontheory,optimizefacepose;Strengtheniterativetheory

TheenhancediterativetheoryisaneffectiveextensionoftheDLFAfacedetectionalgorithm;

Theoriginalreal-timefeaturerecognitiontheory

ThetheoryfocusesonBasedontheintermediatevalueprocessingoffacereal-timedata,itcanachievethebestmatchingeffectbetweentherecognitionrateandtherecognitionperformance.

Recognitiondata

FacerecognitionneedstoaccumulateandcollectAlargenumberoffaceimage-relateddataareusedtoverifyalgorithmsandcontinuouslyimproverecognitionaccuracy.ThesedatasuchasANeuralNetworkFaceRecognitionAssignment(neuralnetworkfacerecognitiondata),orlfacedatabase,MITbiologyandcomputingThefacerecognitiondatabaseoftheLearningCenter,thefacerecognitiondataoftheSchoolofComputerandElectronicEngineering,UniversityofEssex,etc.

Degreeofcooperation

Existingfacerecognitionsystemscanachievesatisfactoryresultswhentheusercooperatesandtheacquisitionconditionsarerelativelyideal.However,iftheuserdoesnotcooperateandtheacquisitionconditionsarenotideal,therecognitionrateoftheexistingsystemwilldropsharply.Forexample,whenafaceiscompared,thereisadiscrepancywiththefacestoredinthesystem.Forexample,ashavedbeard,achangeofhairstyle,moreglasses,andachangeofexpressionmaycausethecomparisontofail.

Advantagesanddifficulties

Advantages

Theadvantagesoffacerecognitionareitsnaturalnessandthecharacteristicsofbeingundetectedbytheindividualbeingtested.

Theso-callednaturalnessmeansthattheidentificationmethodisthesameasthebiologicalcharacteristicsusedbyhumans(orevenothercreatures)forindividualidentification.Forexample,facerecognition,humansalsodistinguishandconfirmidentitiesthroughobservationandcomparisonoffaces.Inaddition,naturalrecognitionincludesvoicerecognition,bodyshaperecognition,etc.,whilefingerprintrecognitionandirisrecognitionarenotnatural,becausehumansorotherBiologydoesnotdistinguishindividualsbysuchbiologicalcharacteristics.

Thecharacteristicofbeingimperceptibleisalsoveryimportantforanidentificationmethod,whichmakestheidentificationmethodnotobjectionable,andbecauseitisnoteasytoattractpeople'sattention,itisnoteasytobedeceived.Facerecognitionhasthisfeature.Itcompletelyusesvisiblelighttoobtainfacialimageinformation.Unlikefingerprintrecognitionoririsrecognition,itneedstouseanelectronicpressuresensortocollectfingerprints,oruseinfraredtocollectirisimages.ThesespecialcollectionmethodsareeasyBeingnoticedbyothersmakesitmorelikelytobedeceivedbydisguise.

Difficult

Facerecognitionisconsideredtobeoneofthemostdifficultresearchtopicsinthefieldofbiometricsandeveninthefieldofartificialintelligence.Thedifficultyoffacerecognitionismainlycausedbythecharacteristicsofthefaceasabiologicalfeature.

Similarity

Thereislittledifferencebetweendifferentindividuals.Thestructureofallhumanfacesissimilar,andeventhestructureandappearanceoffaceorgansareverysimilar.resemblance.Suchafeatureisadvantageousforusinghumanfacesforpositioning,butitisdisadvantageousforusinghumanfacestodistinguishhumanindividuals.

Variability

Theshapeofthehumanfaceisveryunstable.Peoplecanproducemanyexpressionsthroughthechangesoftheface,andatdifferentviewingangles,thefaceThevisualimagesarealsoverydifferent.Inaddition,facerecognitionisalsoaffectedbylightingconditions(suchasdayandnight,indoorandoutdoor,etc.),manycoveringsoftheface(suchasmasks,sunglasses,hair,beards,etc.),age,etc.Theinfluenceofvariousfactors.

Infacerecognition,thefirsttypeofchangeshouldbeamplifiedandusedasacriterionfordistinguishingindividuals,whilethesecondtypeofchangeshouldbeeliminatedbecausetheycanrepresentthesameindividual.Thefirsttypeofchangeisusuallycalledinter-classdifference,andthesecondtypeofchangeiscalledintra-classdifference.Forhumanfaces,intra-classchangesareoftengreaterthaninter-classchanges,whichmakesitextremelydifficulttodistinguishindividualsbyusinginter-classchangesundertheinterferenceofintra-classchanges.

Mainpurpose

Facerecognitionismainlyusedforidentityrecognition.Duetotherapidspreadofvideosurveillance,manyvideosurveillanceapplicationsurgentlyneedafastidentificationtechnologyinalong-distance,user-uncooperativestate,inordertoquicklyconfirmtheidentityofapersonfromalongdistanceandrealizeintelligentearlywarning.Facerecognitiontechnologyisundoubtedlythebestchoice.Usingfastfacedetectiontechnologycanfindfacesfromsurveillancevideoimagesinrealtime,andcomparethemwithfacedatabasesinrealtimetoachieverapididentityrecognition.

Applicationprospects

Biometrictechnologyhasbeenwidelyusedingovernment,military,banking,socialwelfare,e-commerce,securityanddefenseandotherfields.Forexample,adepositorwalksintoabank,andhedoesnotbringabankcard,nordoesherecallthepasswordtowithdrawmoney.WhenhewithdrawsmoneyattheATM,acamerascanstheuser’seyesandthenquicklyandaccuratelyTheuseridentificationiscompletedlocallyandthebusinessiscompleted.ThisisarealscenethattookplaceinasalesdepartmentofUnionBankofTexas,USA.Andwhatthebusinessdepartmentusesisthe"irisrecognitionsystem"inmodernbiometrictechnology.Inaddition,afterthe"9.11"incidentintheUnitedStates,anti-terrorismactivitieshavebecometheconsensusofgovernmentsofallcountries,anditisveryimportanttostrengthenthesecurityanddefenseofairports.ThefacialrecognitiontechnologyoftheUSVisagcompanyhasshownitsmagicintwoairportsintheUnitedStates.Itcanpickoutacertainfaceinacrowdedcrowdtodetermineifheisawantedcriminal.

Thefrequentoccurrenceofburglary,robbery,woundingandothercasesinthecurrentsocietycontinuestooccur.Forthisreason,securitydoorshavebeguntoenterthousandsofhouseholds,bringingpeacetofamilies;however,withWiththedevelopmentofsociety,theadvancementoftechnology,theaccelerationofthepaceoflife,andtheimprovementofconsumptionlevels,people’sexpectationsforhomefurnishingsarealsoincreasing,andtherequirementsforconveniencearebecomingmoreandmoreurgent.Theanti-theftdoorsbasedontraditionalpurelymechanicaldesign,inadditiontoInadditiontobeingruggedanddurable,itisdifficulttoquicklymeettheseemergingneeds:convenience,dooropeningrecordsandotherfunctions.Facerecognitiontechnologyhasbeenwidelyrecognized,butitsapplicationthresholdisstillhigh:hightechnicalthreshold(longdevelopmentcycle),higheconomicthreshold(highprice).

Facerecognitionproductshavebeenwidelyusedinfinance,justice,military,publicsecurity,bordercontrol,government,aerospace,electricpower,factories,education,medicaltreatmentandmanyenterprisesandinstitutions.Withthefurthermaturityoftechnologyandtheimprovementofsocialrecognition,facerecognitiontechnologywillbeappliedinmorefields.

1.Enterpriseandresidentialsecurityandmanagement.Suchasfacerecognitionaccesscontrolandattendancesystems,facerecognitionanti-theftdoors,etc.

2.ElectronicpassportandIDcard.China'se-passportprogram,theMinistryofPublicSecurity,issteppingupitsplanningandimplementation.

3.Publicsecurity,justiceandcriminalinvestigation.Suchastheuseoffacialrecognitionsystemsandnetworkstosearchforfugitivesacrossthecountry.

4.Self-service.

5.Informationsecurity.Suchascomputerlogin,e-governmentande-commerce.Ine-commerce,alltransactionsarecompletedonline,andmanyapprovalproceduresine-governmenthavealsobeenmovedonline.Atpresent,theauthorizationoftransactionsorapprovalsisachievedbypasswords.Ifthepasswordisstolen,securitycannotbeguaranteed.However,theuseofbiologicalcharacteristicscanachievetheunificationoftheparties’digitalandrealidentitiesontheInternet,therebygreatlyincreasingthereliabilityofe-commerceande-governmentsystems.

Mainproducts

Digitalcameras

Faceautofocusandsmileshuttertechnology:Thefirstisfacialcapture.Itjudgesaccordingtothepositionoftheperson'shead.First,theheadisdetermined,andthentheheadfeaturessuchaseyesandmoutharejudged.Throughthecomparisonofthefeaturelibrary,itisconfirmedthatitisahumanface,andthefacecaptureiscompleted.Thenautofocuswiththefaceasthefocus,whichcangreatlyimprovetheclarityofthephotostaken.Smileshuttertechnologyisbasedonfacerecognition,completefacialcapture,andthenstarttojudgethedegreeofmouthbendandthedegreeofeyebend,tojudgewhetheritisasmile.Alltheabovecapturesandcomparisonsaredoneinthecaseofcomparingthefeaturelibrary,sothefeaturelibraryisthebasis,whichcontainsvarioustypicalfacialandsmilingfeaturedata.

Accesscontrolsystem

Asecurity-protectedareacanidentifytheidentityofthepersontryingtoenterthroughfacialrecognition.Thefacerecognitionsystemcanbeusedforenterpriseandresidentialsecurityandmanagement.Suchasfacerecognitionaccesscontrolandattendancesystems,facerecognitionanti-theftdoors,etc.

Facerecognitionaccesscontrol

Facerecognitionaccesscontrolisbasedonadvancedfacerecognitiontechnology,combinedwithmatureIDcardandfingerprintrecognitiontechnology.Safeandpracticalaccesscontrolproducts.Theproductadoptsasplitdesign,thecollectionofface,fingerprintandIDcardinformationandtheseparationofbiologicalinformationrecognitionandaccesscontrolcontrolarehighinpracticality,safeandreliable.Thesystemadoptsnetworkinformationencryptiontransmission,supportsremotecontrolandmanagement,andcanbewidelyusedinaccesscontrolsecuritycontrolinkeyareassuchasbanks,themilitary,publicsecuritybureaus,andintelligentbuildings.

Identification

Suchase-passportandIDcard.Thismaybeafuturescaleapplication.TheInternationalCivilAviationOrganizationhasdeterminedthatfromApril1,2010,its118membercountriesandregionsmustusemachine-readablepassports.Facerecognitiontechnologyisthefirstrecognitionmode.Thisregulationhasbecomeaninternationalstandard.TheUnitedStateshasrequiredcountriesthathavevisa-freeaccessagreementswithittouseanelectronicpassportsystemthatincorporatesfacialfingerprintsandotherbiometricsbeforeOctober26,2006.Bytheendof2006,morethan50countrieshadimplementedsuchasystem.TheTransportationSecurityAdministrationplanstopromoteabiometric-baseddomestictraveldocumentthroughouttheUnitedStates.ManycountriesinEuropearealsoplanningorimplementingsimilarprogramstoidentifyandmanagepassengerswithbiometricdocuments.China'se-passportprogram,theMinistryofPublicSecurity,issteppingupitsplanningandimplementation.

Peoplecanbemonitoredinpublicplacessuchasairports,stadiums,andsupermarkets.Forexample,asurveillancesystemcanbeinstalledatairportstopreventterroristsfromboarding.Forexample,iftheuser'scardandpasswordarestolenfromthebank'sATM,itwillbetakenbyothersascash.Atthesametime,theapplicationoffacerecognitionwillavoidthissituation.Findoutwhetherthereisbasicinformationofkeypopulationsinthedatabasebyqueryingthetargetportraitdata.Forexample,installsystemsatairportsorstationstocatchfugitives.

Webapplication

Facerecognitionprocess(2photos)

Entertainmentapplications

Facerecognitiontechnologyiswidelyusedindailylife,suchascamerashooting,picturecomparison,etc.,especiallyinthepasttwoyears,blinddateprogramshavebeeninfullswing,amongwhichZhejiangTV’sLoveLinkThebestcoupleportraitsessionusesfacecomparisontechnologytotestthesimilarityofthefacesofthemaleandfemaleprotagonists.

WiththeriseofthemobileInternet,somefacerecognitiontechnologydevelopershaveappliedthistechnologytotheentertainmentfield,suchasapplyinghappycelebrityfaces,etc.,accordingtothecontour,skincolor,texture,andtextureoftheface.,Color,lightingandothercharacteristicstocalculatethesimilaritybetweentheprotagonistandthestarinthephoto.

Applicationexample

OnApril13,2012,theBeijing-Shanghaihigh-speedrailwaysecuritycheckareafacerecognitionsystemprojectbeganbidding,andthethreestationsofShanghaiHongqiaoStation,TianjinWestStationandJinanWestStationweresecuritycheckedTheareawillinstallahigh-techsecuritychecksystemforidentityrecognition-thefacerecognitionsystemtoassistthepublicsecuritydepartmentincatchingcriminalsatlarge,andhigh-techinnovativecompanieswithfacialrecognitionproductsandsystemsolutions.ThecoretechnologyR&Dteamiscomposedofexpertsinthisfield,focusingonproductdesignandR&Dprojectsinvariousfieldssuchasattendance,accesscontrolandsecuritywithfacerecognitiontechnologyasthecore.Nowadays,facerecognitionproductshavebeenwidelyusedinthefieldsoffinance,justice,military,publicsecurity,borderinspection,government,aerospace,electricpower,factories,education,medicaltreatmentandmanyenterprisesandinstitutions.

OnSeptember5,2013,thefacialpaymentsystemwasunveiledattheChinaInternationalFinanceExhibition.ThefacialrecognitionpaymentsystemisbasedonthebiometriccloudfinancialplatformindependentlydevelopedbyTianchengShengye.Itintegratesthemilitary-levelfacialrecognitionalgorithmwithindependentintellectualpropertyrightsandtheexistingpaymentsystemtoconnectthepayment,transfer,andsettlementinourlives.Andtradinglinks.Whenpaying,peoplenolongerneedbankcards,passbooksandpasswords,orevenmobilephones.Theyonlyneedtonodtothecameraandshowasmile.Theface-swipingpaymentsystemwillcompleteidentityverification,accountreading,andtransferpaymentwithinafewseconds.,Transactionconfirmationandotherone-stoppaymentlinks,creatingabetterpaymentexperienceforusers.

StartingfromAugust2014,Japanwillrestarttheexperimentofthefacerecognitionsystemattheimmigrationinspection(frontierinspection)atsomeairports.Thefirstexperimentcarriedoutin2012wastemporarilysuspendedduetofrequenterrors,buttheMinistryofJusticebelievedthatthespeedofborderinspectionsneededtobeincreasedtomeetthe2020TokyoOlympics,soitdecidedtorestarttheexperiment.Theexperimentwillbeconductedforabout5weeksstartinginAugust2014,andthesubjectswillbeJapaneseflyingatHanedaAirportandNaritaAirport.Thecompanyresponsiblefortheexperimentwillbefinalizedinthenearfuture.TheJapanesegovernmenthassetupautomaticbordergatesthatcanbepassedthroughonlybyfingerprintrecognitionatvariousairports.However,duetotheneedtoregisterfingerprintsinadvance,theutilizationrateofpassengersisnothigh.Thereisnoneedtoregisterinadvanceforfacerecognition.

OnMarch15,2015,theHannoverITExpo(CeBIT)openedinGermany.AlibabafounderJackMa,astheonlyinvitedentrepreneurrepresentative,deliveredakeynotespeechattheopeningceremony.Afterthespeech,JackMaalsodemonstratedAntFinancial'sSmiletoPayfacescanningtechnologyforGermanChancellorMerkelandChineseDeputyPrimeMinisterMaKai,andbrushedhisfaceonthespottobuygiftsfortheguests.ThegiftMaYunchosewasa1948HanovercommemorativestamponTaobao.com.HeuseshismobilephonetologintoTaobaoandfirstselecttheproduct;thesecondstepistoenterthepaymentsystem,andafterconfirmingthepayment,apagetoscanhisfaceappears;thenhescanshisface(aftertakingaphoto)forbackgroundauthentication;thenitshowsthatthepaymentissuccessful.JackMapresentedaspecialgifttoGermanChancellorMerkelonthespot:acommemorativeversionoftheGermancalendarpage,whichhappenedtobethebirthdateofthefemaleChancellor.

Lawsandregulations

OnJuly28,2021,withtherapiddevelopmentofinformationtechnology,facerecognitionhasgraduallypenetratedintoallaspectsofpeople’slives.Whilefacerecognitiontechnologyplaysahugeroleinmanyfields,itisalsoabused.TheSupremePeople'sCourthasissuedjudicialinterpretationstoregulatefacerecognition.

OnJuly28,2021,theProvisionsoftheSupremePeople'sCourtonSeveralIssuesConcerningtheApplicationofLawintheTrialofCivilCasesRelatedtotheUseofFaceRecognitionTechnologytoProcessPersonalInformationwereofficiallyreleased.The"Regulations"clarify:"Realestateserviceenterprisesorotherbuildingmanagersusefacerecognitionastheonlyverificationmethodforownersorpropertyuserstoenterandexitthepropertyservicearea,andtheownersorpropertyuserswhodisagreerequestotherreasonableverificationmethods.Thepeople’scourtsupportsitinaccordancewiththelaw.”Accordingtothisprovision,thecommunitypropertyshallobtaintheconsentoftheownerorpropertyuserwhenusingthefacerecognitionaccesscontrolsystemtoenterfacialinformation.Forthosewhodonotagree,thecommunityshallThepropertyshallprovidealternativeverificationmethods,andshallnotinfringethepersonalityrightsandotherlegalrightsoftheownerortheuseroftheproperty.

InAugust2021,asfacerecognitiontechnologyiswidelyusedinreallife,thepublic'sconcernaboutitsabuseisalsoincreasing,andthecallforstrengtheningtheprotectionoffaceinformationisincreasing.TheSupremeLawpromulgatedthe"RegulationsonSeveralIssuesConcerningtheApplicationofLawintheTrialofCivilCasesRelatedtotheUseofFaceRecognitionTechnologytoProcessPersonalInformation",whichclearlydefinesthecollectionoffacialinformationwithouttheindividualconsentofnaturalpersonsortheirguardiansas"infringement",whichisnotonlyaresponseThishashelpedthepeopletocallfortheprotectionoffaceinformation,andprovidedalegalbasisforthejudicialdepartmenttodealwithsuchdisputes,andstrengthenedcitizens'legalconfidenceintheprotectionoftheirpersonalrightsandinterests.

OnAugust20,2021,the30thmeetingoftheStandingCommitteeofthe13thNationalPeople'sCongressvotedtopassthe"PersonalInformationProtectionLawofthePeople'sRepublicofChina",whichwillcomeintoforceonNovember1,2021.Inresponsetotheabuseoffacerecognitiontechnology,thislawrequiresthattheinstallationofimagecollectionandpersonalidentificationequipmentinpublicplacesshouldbesetupwithprominentreminders;thecollectedpersonalimagesandidentificationinformationcanonlybeusedforthepurposeofmaintainingpublicsafety.

ThecontentofBaiduBaikeisco-editedbynetizens.Ifyoufindthatthecontentofyourentryisinaccurateorimperfect,pleaseuseyourentryeditingservice(freeofcharge)toparticipateinthecorrection.Gonow>>