Appeal No. 1999-2294
Application No. 08/807,430
Appellants argue that Journal teaches and suggests the use of a two-pass parametric
analysis which uses two statistical models or templates rather than the individual
utterances by single speakers or a nonparametric vocabulary. (See brief at page 6.) We
agree with appellants. The templates taught by Journal are composites of multiple
individual utterances rather than a single utterance by a single speaker as in a
nonparametric vocabulary as required by the language of claims 1, 12, 19, and 21.
Therefore, we will not sustain the rejection of claim 1 and its dependent claims 2-11.
Appellants argue that claim 12 requires a two-pass approach wherein the first pass
compares the speech sample to a model to identify a portion of the model and then in a
second pass compares the speech sample to the training observations in a subset of the
identified portion of the model. (See brief at page 8.) We agree with appellants that
Journal does not teach this identification of a portion and then using the training
observations in a subset of the identified portion of the model as required by claim 12.
Therefore, we will not sustain the rejection of claim 12 and its dependent claims 13-18.
With respect to claim 19, Journal does not teach or suggest the use of a
nonparametric vocabulary as discussed above. Therefore, we will not sustain the rejection
of claim 19 and its dependent claim 20.
6
Page: Previous 1 2 3 4 5 6 7 8 9 Next
Last modified: November 3, 2007