- 1、本文档共5页,可阅读全部内容。
- 2、原创力文档(book118)网站文档一经付费(服务费),不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
- 3、本站所有内容均由合作方或网友上传,本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺!文档内容仅供研究参考,付费前请自行鉴别。如您付费,意味着您自己接受本站规则且自行承担风险,本站不退款、不进行额外附加服务;查看《如何避免下载的几个坑》。如果您已付费下载过本站文档,您可以点击 这里二次下载。
- 4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“版权申诉”(推荐),也可以打举报电话:400-050-0827(电话支持时间:9:00-18:30)。
查看更多
MP稀疏分解快速算法及其在语音识别中的应用
122201 and
0,46(1) ComputerEngineeringApplications计算机工程与应用
MP稀疏分解快速算法及其在语音识别中的应用
李雨昕,尹忠科,王建英
LIYu-xin,YIN
Zhong-ke,WANGJian-ying
1
西南交通大学信息科学与技术学院,成都61003
SchoolofInformation 610031,China
ScienceTechnology,SouthwestJiaotongUniversity,Chengdu
E-mail:yuxinlee@163.corn
LIYu-xin.YIN forMP andits in
Zhong-ke。WANGJian-ying.Fastalgorithm sparsedecompositionapplicationspeech
and
recognition.ComputerEngineeringApplications,2010。46(1):122—124.
Abstract:After Pursuit(MP)based new
studying signal decomposition。asparsedecompositionalgorithm
Matching speech sparse
is tothecharactersofthe andthefactthatthe andatomsareall onthe
proposed.According speechsignal signal real,based
of the atom view convertsinner calculationsinto
partition over-completedictionary,thisalgorithm product
calculationsthatarefastdone Fast withMPbased decomposition
by HartleyTransform(FHT).Therefore,comparedsignalsparse
with cannot reducethe also the ofthe
F兀’,this memoryconsumptionhalf,butheighten
algorithm only by speed decomposition
24.8%.The is realized Vector uses
by recognition Machine(SVM)thatsignals’Mel-scaled
speech bySupport speech Frequency
Coefficient(MFCC)andthefeaturesextractedthis11CMR the theeffectivenessof
Cepstrum by algorithm.Finallyexperimentsverify
the
proposedalgorithm.
Keywo
文档评论(0)