预览加载中，请您耐心等待几秒...

一种基于SCHMM的手语识别方法.docx

预览
1/3
2/3
3/3

在线预览结束，喜欢就下载吧，查找使用更方便

下载提示 文本预览

如果您无法下载资料，请参考说明：

1、部分资料下载需要金币，请确保您的账户上有足够的金币

2、已购买过的文档，再次下载不重复扣费

3、资料包下载后请先用软件解压，在使用对应软件打开

一种基于SCHMM的手语识别方法 Abstract: Signlanguagerecognitionisanimportantareaofresearchthataimstohelpthedeafcommunityintheircommunicationwiththehearingcommunity.TheuseofHiddenMarkovModels(HMMs)hasbeenwidelyacceptedasasuccessfulmethodforrecognizinghandsigns.However,conventionalHMMsdonotaccountforthetemporaldependenciesinthesignlanguagerecognitionsystem.AplausiblesolutiontothisprobleminvolvesusingStateConditionalHiddenMarkovModels(SCHMMs).Inthispaper,wepresentamethodforsignlanguagerecognitionwhichisbasedonSCHMMs. Introduction: Signlanguageistheprimaryformofcommunicationforpeoplewhoaredeaforhardofhearing.Itisavisual-spatiallanguagethatconsistsofacombinationofhandgestures,facialexpressions,andbodylanguage.Therecognitionofsignlanguageisachallengingtaskduetotheinherentcomplexityofthegesturesandthevariabilityinthewaypeopleexpressthemselvesthroughsignlanguage. Signlanguagerecognitionisoneofthemostimportantapplicationsofgesturerecognition.Thegoalofsignlanguagerecognitionistoconvertsignedlanguageintorecognizedsentencesinanaturallanguage,whichcansupportcommunicationbetweenthedeafandhearingcommunities.TheuseofHiddenMarkovModels(HMMs)hasbeenwidelyacceptedasasuccessfulmethodforrecognizinghandsigns.However,conventionalHMMsdonotaccountforthetemporaldependenciesinthesignlanguagerecognitionsystem. StateConditionalHiddenMarkovModels(SCHMMs)areanextensionofconventionalHMMsthathavebeenproposedtoaddressthisproblem.SCHMMsareatypeofprobabilisticmodelthatcapturethedependenciesbetweenstatesintheHMMs.Byconsideringthetemporaldependenciesbetweenadjacentstates,SCHMMscanimprovetheperformanceofsignlanguagerecognition. ThispaperpresentsamethodforsignlanguagerecognitionbasedonSCHMMs.Theproposedmethodusesasetofhand-craftedfeaturesthatdescribetheshapeandmotionofthehandduringsignlanguagegestures.AsetofSCHMMsisthentrainedtorecognizethehandsigns.ExperimentalresultsonalargedatasetshowthattheproposedmethodoutperformstheconventionalHMM-basedmodelintermsofaccuracyandreliability. Methodology: Theproposedmethodforsignla