Method and system for matching speech data

  • US 7,707,032 B2
  • Filed: 10/20/2005
  • Issued: 04/27/2010
  • Est. Priority Date: 10/20/2005
  • Status: Active Grant
  • ×
    • Pin Icon | RPX Insight
    • Pin
First Claim
Patent Images

1. A method for matching speech data used to determine the similarity between an input speech data and a sample speech data, the method comprising:

  • segmenting the input speech data into a plurality of input speech frames;

    segmenting the sample speech data into a plurality of sample speech frames;

    building a matching matrix, wherein each element of the matching matrix corresponds to one of the input speech frames and one of the sample speech frames and indicates a distance value between the corresponding input speech frame and the corresponding sample speech frame;

    determining a minimum value of the distance values indicated in each row of elements of the matching matrix, thereby obtaining a plurality of minimum distance values of the respective rows of elements of the matching matrix, determining a second least value of the distance values indicated in each row of elements of the matching matrix, thereby obtaining a plurality of second least distance values of the respective rows of elements of the matching matrix;

    summing up the minimum distance values and the second least distance value of the distance values indicated in each row of elements of the matching matrix, thereby obtaining a row score, determined by;

    row



    score
    =

    j = 1 r




    min r

    C


    [ MM

    ( i , j )
    ]
    +

    j = 1 r


    min r

    C
    - i c


    [ MM

    ( i , j )
    ]
    ;

    determining a minimum value of the distance values indicated in each column of elements of the matching matrix, thereby obtaining a plurality of another minimum distance values of the respective columns of elements of the matching matrix, determining a second least value of the distance values indicated in each column of elements of the matching matrix, thereby obtaining a plurality of second least distance values of the respective columns of elements of the matching matrix;

    summing up the another minimum distance values and the second least value of the indicated distance values in each column of elements of the matching matrix distance values, thereby obtaining a column score, wherein;

    column



    score
    =

    i = 1 e




    min j

    R


    [ MM

    ( i , j )
    ]
    +

    i = 1 e


    min j

    R - j k


    [ MM

    ( i , j )
    ]
    ;

    calculating a matching score obtained by combining the distance row score and the column score; and

    determining whether the input speech data and the sample speech data are similar according to the matching score.

View all claims
    ×
    ×

    Thank you for your feedback

    ×
    ×