专利名称：PARAMETER CLUSTERING AND SHARING
FOR VARIABLE-PARAMETER HIDDENMARKOV MODELS
发明人：Dong Yu,Li Deng,Yifan Gong,Alejandro Acero申请号：US12211115申请日：20080916
公开号：US20100070280A1公开日：20100318
专利附图：
摘要：A speech recognition system uses Gaussian mixture variable-parameter hiddenMarkov models (VPHMMs) to recognize speech. The VPHMMs include Gaussian
parameters that vary as a function of at least one environmental conditioning parameter.The relationship of each Gaussian parameter to the environmental conditioningparameter(s) is modeled using a piecewise fitting approach, such as by using splinefunctions. In a training phase, the recognition system can use clustering to identify classesof spline functions, each class grouping together spline functions which are similar toeach other based on some distance measure. The recognition system can then store setsof spline parameters that represent respective classes of spline functions. An instance ofa spline function that belongs to a class can make reference to an associated shared setof spline parameters. The Gaussian parameters can be represented in an efficient formthat accommodates the use of sharing in the above-summarized manner.
申请人：Dong Yu,Li Deng,Yifan Gong,Alejandro Acero
地址：Kirkland WA US,Redmond WA US,Sammamish WA US,Bellevue WA US
国籍：US,US,US,US
更多信息请下载全文后查看
因篇幅问题不能全部显示，请点此查看更多更全内容