AUTOMATIC NORMALIZATION OF SPOKEN SYLLABLE DURATION

• US 20110282650A1
• Filed: 05/17/2010
• Published: 11/17/2011
• Est. Priority Date: 05/17/2010
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Avaya Incorporated
• Inventors: Jennings, Terry, Michaelis, Paul Roller
• Granted Patent: US 8,401,856 B2
• Time in Patent Office: Days
• US Class Current: 704/8
• CPC Class Codes: G10L 15/005 Language recognition
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/1807 using prosody or stress
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: G10L 19/04 using predictive techniques
• CPC Class Codes: G10L 19/12 the excitation function being a code excitation, e.g. in code excited linear prediction [CELP] vocoders
• CPC Class Codes: G10L 21/04 Time compression or expansion
• CPC Class Codes: G10L 25/00 Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 muting semiconductor-based amplifiers when some special characteristics of a signal are sensed by a speech detector, e.g. sensing when no signal is present, H03G3/34

Abstract

A very common problem is when people speak a language other than the language which they are accustomed, syllables can be spoken for longer or shorter than the listener would regard as appropriate. An example of this can be observed when people who have a heavy Japanese accent speak English. Since Japanese words end with vowels, there is a tendency for native Japanese to add a vowel sound to the end of English words that should end with a consonant. Illustratively, native Japanese speakers often pronounce “orange” as “orenji.” An aspect provides an automatic speech-correcting process that would not necessarily need to know that fruit is being discussed; the system would only need to know that the speaker is accustomed to Japanese, that the listener is accustomed to English, that “orenji” is not a word in English, and that “orenji” is a typical Japanese mispronunciation of the English word “orange.”

Claims

1. A method to improve communications understandability comprising:

receiving speech;

identifying one or more distinct speech events in the received speech;

representing one or more of the one or more distinct speech events as an adjustable speech production parameter;

utilizing a knowledge base to determine an incorrect syllable duration caused by a mispronunciation; and

adjusting the mispronunciation to one or more of lengthen, shorten, emphasize or deemphasize a misspoken syllable.

2. The method of claim 1, further comprising detecting a language being spoken.

3. The method of claim 1, further comprising detecting a native language of a speaker.

4. The method of claim 1, further comprising utilizing pronunciation patterns and vocabularies of a language being spoken and a native language of a speaker to detect the incorrect syllable duration caused by the mispronunciation.

5. The method of claim 1, further comprising using modified speech product parameters to regenerate and present speech with a corrected syllabic timing to one or more listeners.

6. The method of claim 1, further comprising determining whether an utterance is a legitimate word.

7. The method of claim 1, further comprising determining if an utterance is a common mispronunciation.

8. The method of claim 1, further comprising providing feedback to a speaker.

9. One or more means for performing the functionality of claim 1.

10. A non-transitory computer readable information storage media having stored thereon instructions that, if executed by a processor, cause to be performed the method of claim 1.

11. A system that improves communications understandability comprising:

an analysis module that receives speech;

a distinct speech event recognition module cooperating with an encoding and compression module to identify one or more distinct speech events in the received speech and represent one or more of the one or more distinct speech events as an adjustable speech production parameter; and

a modification module that utilizes a knowledge base to determine an incorrect syllable duration caused by a mispronunciation and adjusts the mispronunciation to one or more of lengthen, shorten, emphasize or deemphasize a misspoken syllable.

12. The system of claim 11, further comprising a language detection module that detects a language being spoken.

13. The system of claim 11, further comprising a language detection module that detects a native language of a speaker.

14. The system of claim 11, wherein a repository including pronunciation patterns and vocabularies of a language being spoken and a native language of a speaker is used to detect the incorrect syllable duration caused by the mispronunciation.

15. The system of claim 11, wherein modified speech product parameters are used to regenerate and present speech with a corrected syllabic timing to one or more listeners.

16. The system of claim 11, further comprising a processor that determines whether an utterance is a legitimate word.

17. The system of claim 11, wherein the analysis module further determines if an utterance is a common mispronunciation.

18. The system of claim 11, further comprising a feedback module that provides feedback to a speaker.

19. The system of claim 11, wherein a participant can select via a feedback module one or more of a modified and unmodified stream to listen to.

20. The system of claim 11, wherein a further determination is made as to whether a modified word is inappropriate.