×

Speech processing system

  • US 6,801,891 B2
  • Filed: 11/13/2001
  • Issued: 10/05/2004
  • Est. Priority Date: 11/20/2000
  • Status: Expired due to Fees
First Claim
Patent Images

1. An apparatus for identifying one or more words corresponding to a sequence of sub-word units output by a recognition system in response to a rendition of the one or more words, the apparatus comprising:

  • first receiving means for receiving the recognised sequence of sub-word units representative of the one or more words to be identified;

    second receiving means for receiving a plurality of dictionary sub-word sequences, each representative of one or more known words;

    means for comparing sub-word units of the recognised sequence with sub-word units of each dictionary sequence to provide a set of comparison scores;

    means for combining the comparison scores obtained by comparing the sub-word units of the recognised sequence with the sub-word units of the same dictionary sequence to provide, for each dictionary word, a measure of the similarity between the recognised sequence and the dictionary sequence; and

    means for identifying said one or more words using the similarity measures provided by the combining means;

    wherein said comparing means comprises;

    means for aligning sub-word units of the recognised sequence with sub-word units of the same dictionary sequence to form, for each dictionary sequence, a number of aligned pairs of sub-word units;

    first sub-comparing means for comparing, for each aligned pair, the recognised sequence sub-word unit in the aligned pair with each of a plurality of sub-word units taken from a set of predetermined sub-word units, to generate a corresponding plurality of intermediate comparison scores representative of the similarities between the recognised sequence sub-word unit and the respective sub-word units of the set;

    second sub-comparing means for comparing, for each aligned pair, the dictionary sequence sub-word unit in the aligned pair with each of said plurality of sub-word units from the set to separate a further corresponding plurality of intermediate comparison scores representative of the similarities between said dictionary sequence sub-word unit and the respective sub-word units of the set; and

    means for calculating, for each aligned pair, a comparison score representative of the similarity between the sub-word units of the aligned pair by combining the pluralities of intermediate comparison scores generated by said first and second sub-comparing means, to provide said set of comparison scores.

View all claims
  • 1 Assignment
Timeline View
Assignment View
    ×
    ×