Pronunciation discovery for spoken words

• US 20050143970A1
• Filed: 09/13/2004
• Published: 06/30/2005
• Est. Priority Date: 09/11/2003
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Voice Signal Technologies Incorporated (Microsoft Corporation)
• Inventors: Shire, Michael, Roth, Daniel L., Gillick, Laurence S.
• Granted Patent: US 8,577,681 B2
• Time in Patent Office: Days
• US Class Current: 704/4
• CPC Class Codes: G10L 15/06 Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice G10L15/14 takes precedence
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams

Abstract

A method of generating an alternative pronunciation for a word or phrase, given an initial pronunciation and a spoken example of the word or phrase, includes providing the initial pronunciation of the word or phrase, and generating the alternative pronunciation by searching a neighborhood of pronunciations about the initial pronunciation via a constrained hypothesis, wherein the neighborhood includes pronunciations that differ from the initial pronunciation by at most one phoneme. The method further includes selecting a highest scoring pronunciation within the neighborhood of pronunciations.

Claims

1. A method of generating an alternative pronunciation for a word or phrase given an initial pronunciation and a spoken example of the word or phrase, comprising:

providing the initial pronunciation of the word or phrase; and

,

generating the alternative pronunciation by searching a neighborhood of pronunciations about the initial pronunciation and selecting a highest scoring pronunciation within the neighborhood of pronunciations.

2. The method of claim 1, further including searching the neighborhood of pronunciations that differ from the initial pronunciation in a limited number of phonemes.

3. The method of claim 1, further including searching the neighborhood of pronunciations that differ from the initial pronunciation by at most one phoneme.

4. The method of claim 1, further including providing the initial pronunciation of the word or phrase by accepting a recognized transcript from a speech recognizer.

5. The method of claim 1, further including searching the neighborhood by using a speech recognition system to perform phoneme recognition with a constraint.

6. The method of claim 1, further including generating the initial pronunciation by accepting a user selection of a recognized transcript.

7. The method of claim 1, further including generating the initial pronunciations by accepting a user selection from a list of alternatives to a recognized transcript.

8. The method of claim 1, further including using a phonetic recognizer to associate a score with each of the alternative pronunciation and/or the initial pronunciation, and using one or both of these scores to decide whether to add the alternative pronunciation to a lexicon.

9. The method of claim 1, further including updating a lexicon by replacing the initial pronunciation with the highest-scoring alternative pronunciation.

10. The method of claim 9, further including providing a user with an option of allowing or disallowing the update of the lexicon

11. A method of generating an alternative pronunciation of a initial pronunciation, comprising:

generating an initial pronunciation corresponding to a spoken utterance;

generating one or more potential alternative pronunciations by changing the initial pronunciation by one phoneme; and

,

selecting a highest scoring potential alternative pronunciation with respect to the spoken utterance as the alternative pronunciation of the initial pronunciation.

12. The method of claim 11, further including updating a lexicon by adding a phonetic representation of the highest-scoring potential alternative pronunciations to the lexicon.

13. The method of claim 12, further including providing a user with an option of allowing or disallowing the update of the lexicon.

14. A computer readable medium including stored instructions adapted for generating an alternative pronunciation of a initial pronunciation, comprising:

instructions for generating an initial pronunciation corresponding to a spoken utterance;

instructions for generating one or more potential alternative pronunciations by changing the initial pronunciation by one phoneme; and

,

instructions for selecting a highest scoring potential alternative pronunciation with respect to the spoken utterance as the alternative pronunciation of the initial pronunciation.

15. The computer readable medium of claim 14, wherein the one or more alternative pronunciations are generated by constructing one or more constrained hypothesis.

16. The computer readable medium of claim 15, wherein the one or more hypothesis are constrained to match the initial hypothesis except for one state of each hypothesis.

17. The computer readable medium of claim 14, wherein the initial pronunciation is generated by accepting a user selection of a recognized transcript.

18. The computer readable medium of claim 14, wherein the initial pronunciations are generated by accepting a user selection from a list of alternatives to a recognized transcript.

19. The computer readable medium of claim 14, wherein a phonetic recognizer is used to associate a score with each of potential alternative pronunciations

20. The computer readable medium of claim 14, further including updating a lexicon by replacing a phonetic representation of the initial pronunciation with the highest-scoring potential alternative pronunciation.

21. A method of updating a lexicon used by a speech recognizer, comprising;

selecting a phonetic representation of a spoken utterance;

generating a set of alternate phonetic representations by changing one or more phonemes in the phonetic representation;

scoring the set of alternate phonetic representations as to how well each one matches the spoken utterance, so as to produce a highest-scoring phonetic representation; and

,

updating the lexicon with the highest scoring phonetic representation.

22. A method of generating an alternative pronunciation for a word or phrase given an initial pronunciation and a spoken example of the word or phrase, comprising:

providing the initial pronunciation of the word or phrase;

generating the alternative pronunciation by searching a neighborhood of pronunciations about the initial pronunciation via a constrained search, wherein the neighborhood includes pronunciations that differ from the initial pronunciation by at most one phoneme; and

,

selecting a highest scoring pronunciation within the neighborhood of pronunciations.

23. A method of generating an alternative pronunciation of an initial pronunciation, comprising:

generating an initial pronunciation corresponding to a spoken utterance;

generating one or more potential alternative pronunciations by constructing one or more hypotheses constrained so as to match the initial pronunciation except for phoneme; and

,

selecting a highest scoring potential alternative pronunciation with respect to the spoken utterance as the alternative pronunciation of the initial pronunciation.