Dynamic speech sharpening
View Patent ↗An enhanced system for speech interpretation is provided. The system may include receiving a user verbalization and generating one or more preliminary interpretations of the verbalization by identifying one or more phonemes in the verbalization. An acoustic grammar may be used to map the phonemes to syllables or words, and the acoustic grammar may include one or more linking elements to reduce a search space associated with the grammar. The preliminary interpretations may be subject to various post-processing techniques to sharpen accuracy of the preliminary interpretation. A heuristic model may assign weights to various parameters based on a context, a user profile, or other domain knowledge. A probable interpretation may be identified based on a confidence score for each of a set of candidate interpretations generated by the heuristic model. The model may be augmented or updated based on various information associated with the interpretation of the verbalization.
1. A method for interpreting natural language utterances using out-of-vocabulary and noise toleration capabilities, comprising:
recognizing, on an electronic device, a phoneme stream contained in an utterance received at the electronic device;
mapping, on the electronic device, the recognized phoneme stream to a syllable series that includes one or more syllables that an acoustic grammar phonemically represents in accordance with an acoustic speech model; and
generating, on the electronic device, an interpretation of the utterance that includes the one or more syllables in the syllable series mapped to the recognized phoneme stream.
2. The method of claim 1 , wherein the acoustic speech model phonemically represents the one or more syllables with acoustic elements for an onset, a nucleus, and a coda.
3. The method of claim 1 , wherein the acoustic speech model includes an unstressed central vowel that links sequential phonemic elements in the acoustic speech model.
4. A method for interpreting natural language utterances using out-of-vocabulary and noise toleration capabilities, comprising:
recognizing, on an electronic device, a phoneme stream contained in an utterance received at the electronic device;
mapping, on the electronic device, the recognized phoneme stream to a syllable series that includes one or more syllables using an acoustic grammar that constrains transitions between acoustic elements phonemically representing the one or more syllables according to one or more phonotactic rules of an acoustic speech model; and
generating, on the electronic device, an interpretation of the utterance that includes the one or more syllables in the syllable series mapped to the recognized phoneme stream.
5. A system for interpreting natural language utterances using out-of-vocabulary and noise toleration capabilities, comprising:
an input device configured to receive an utterance; and
a speech interpretation engine configured to:
recognize a phoneme stream contained in the received utterance;
map the recognized phoneme stream to a syllable series that includes one or more syllables that an acoustic grammar phonemically represents in accordance with an acoustic speech model; and
generate an interpretation of the utterance, wherein the generated interpretation includes the one or more syllables in the syllable series mapped to the recognized phoneme stream.
6. The system of claim 5 , wherein the acoustic speech model phonemically represents the one or more syllables with acoustic elements for an onset, a nucleus, and a coda.
7. The system of claim 5 , wherein the acoustic speech model includes an unstressed central vowel that links sequential phonemic elements in the acoustic speech model.
8. A system for interpreting natural language utterances using out-of-vocabulary and noise toleration capabilities, comprising:
an input device configured to receive an utterance; and
a speech interpretation engine configured to:
recognize a phoneme stream contained in the received utterance;
map the recognized phoneme stream to a syllable series that includes one or more syllables using an acoustic grammar that constrains transitions between acoustic elements phonemically representing the one or more according to one or more phonotactic rules of an acoustic speech model; and
generate an interpretation of the utterance, wherein the generated interpretation includes the one or more syllables in the syllable series mapped to the recognized phoneme stream.