Keyword recognition system and method
AssigneeIntel Corporation
InventorADORAM ERELL
Life of this patent
Estimating the term…
Abstract
A keyword recognition system for speaker dependent, dynamic time warping (DTW) recognition systems uses all of the trained word templates in the system, (keyword and vocabulary), to determine if an utterance is a keyword utterance or not. The utterance is selected as the keyword if a keyword score indicates a significant match to the keyword template and if the keyword score indicates a better match than do the entirety of scores to the vocabulary word templates.
Claims24
A method for recognizing an utterance as a keyword which activates a speaker dependent recognition system on a plurality of vocabulary words, the method comprising the steps of:
- matching said utterance to a keyword template and producing a corresponding keyword score indicating the quality of the match between said utterance and said keyword template;
- matching said utterance to a plurality of vocabulary templates of said vocabulary words and producing a corresponding plurality of vocabulary scores each indicating the quality of the match between said utterance and one of said vocabulary templates;
- selecting said utterance as said keyword if said keyword score indicates a significant match to said keyword template and if said keyword score indicates a better match than do the entirety of said vocabulary scores;
- and
- activating said speaker dependent recognition system to match at least a second utterance to said vocabulary words.
A system for recognizing an utterance as a keyword which activates a speaker dependent recognition system on a plurality of vocabulary words, the system comprising:
- a pattern matcher for matching an input utterance to a keyword template and to a plurality of vocabulary templates of said vocabulary words and for producing a keyword score indicating the quality of the match between said utterance and said keyword template and a corresponding plurality of vocabulary scores each indicating the quality of the match between said utterance and one of said vocabulary templates;
- a criterion determiner for selecting said utterance as said keyword if said keyword score indicates a significant match to said keyword template and if said keyword score indicates a better match than do the entirety of said vocabulary scores and for activating said speaker dependent recognition system to match at least a second utterance to said vocabulary words.
A method for recognizing an utterance as a keyword which activates a speaker dependent recognition system on a plurality of vocabulary words, the method comprising the steps of:
- matching said utterance to at least two keyword templates and producing at least two corresponding keyword scores indicating the quality of the match between said utterance and one of said at least two keyword templates;
- matching said utterance to a plurality of vocabulary templates of said vocabulary words and producing a corresponding plurality of vocabulary scores each indicating the quality of the match between said utterance and one of said vocabulary templates, said vocabulary words being divided into subsets wherein each subset is associated with a different one of said at least two keyword template;
- selecting said utterance as a keyword if the kth one of said at least two keyword scores indicates a significant match to said its corresponding keyword template and if said kth keyword score indicates a better match than do the entirety of said vocabulary scores;
- and
- activating said speaker dependent recognition system to match at least a second utterance to a subset of said vocabulary words corresponding to said kth keyword.
A system for recognizing an utterance as a keyword which activates a speaker dependent recognition system on a plurality of vocabulary words, the system comprising:
- a pattern matcher for matching an input utterance to at least two keyword templates and to a plurality of vocabulary templates of said vocabulary words and for producing corresponding keyword scores indicating the quality of the match between said utterance and said keyword templates and a corresponding plurality of vocabulary scores each indicating the quality of the match between said utterance and one of said vocabulary templates;
- a criterion determiner for selecting said utterance as a keyword if the kth one of said at least two keyword scores indicates a significant match to said its corresponding keyword template and if said kth keyword score indicates a better match than do the entirety of said vocabulary scores and activating said speaker dependent recognition system to match at least a second utterance to a subset of said vocabulary words corresponding to said kth keyword.
Dependencies are inferred from the claim text.