US5819221A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Speech recognition using clustered between word and/or phrase coarticulation
Assignee: Texas Instruments Inc
•
Filing: 1994-08-31
•
Granted: 1998-10-06
•
Expires: 2014-08-31
Abstract
Abstract Improved speech recognition is achieved according to the present invention by use of between word and/or between phrase coarticulation. The increase in the number of phonetic models required to model this additional vocabulary is reduced by clustering 19, 20 the inter-word/phrase models and grammar into only a few classes. By using one class for consonant inter-word context and two classes for vowel contexts, the accuracy for Japanese was almost as good as for unclustered models while the number of models was reduced more than half.
Inventors
- 👤 Kazuhiro Kondo
- 👤 Ikuo Kudo
- 👤 Yu-Hung Kao
- 👤 Barbara J. Wheatley
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-04 09:04
CPC Classifications
G; G10; G10L; G10L15/00; G10L15/08; G10L15/18; G10L15/183; G10L15/187; G10L15/06; G10L15/063; G10L2015/0631
Claims Analysis
Total Claims: 6
Claim 1
Independent Claim
141 words
A speech recognizer comprising: first storage for storing first phonetic models comprising models with within word phonetic word contexts clustered into a first set of given classes and for storing first speech recognition grammar with phonetic within word contexts clustered according to said first set of given classes; second storage for storing second phonetic models comprising models with phonetic between word or phrase contexts clustered into a second set of given classes being generic classes and being significantly fewer classes than said first set of given classes and for storing second speech recognition grammars with phonetic between word or phrase contexts clustered according to said second set of given classes; and means for comparing incoming speech to said first phonetic models and said second phonetic models and said first grammar and said second grammar to provide a best match output therefrom.
Claim 2
Dependent Claim
141 words
A speech recognizer comprising: first storage for storing first phonetic models comprising models with within word phonetic word contexts clustered into a first set of given classes and for storing first speech recognition grammar with phonetic within word contexts clustered according to said first set of given classes; second storage for storing second phonetic models comprising models with phonetic between word or phrase contexts clustered into a second set of given classes being generic classes and being significantly fewer classes than said first set of given classes and for storing second speech recognition grammars with phonetic between word or phrase contexts clustered according to said second set of given classes; and means for comparing incoming speech to said first phonetic models and said second phonetic models and said first grammar and said second grammar to provide a best match output therefrom.
Claim 3
Dependent Claim
45 words
The recognizer of claim 1 wherein one of said second set of given classes is for silence context; a second of said second set of said given classes is for consonants; and a third of said second set of said given classes is for vowels.
Claim 4
Dependent Claim
59 words
The recognizer of claim 2 wherein phonetic between word or phrase context dependent models to the right of a word/phrase includes a silence class, a vowel class, and a consonant class, and for phonetic context dependent models to the left of a word/phrase includes a silence class, a high vowel class, a mid-low vowel class, and a nasal class.
Claim 5
Independent Claim
137 words
A method for speech recognition comprising the steps of: providing models comprising within word context models; providing models comprising phonetic between word or phrase context models; clustering said phonetic between word or phrase context models according to linguistic knowledge classes of silence, consonants and vowels to form generic clustered models for silence, consonants and vowels; providing speech recognition grammar; phonetic between word or phrase contexts expanding of said speech recognition grammar; clustering said expanded phonetic between word or phrase contexts of said speech recognition application grammar according to generic classes of silence, consonants, and vowels to form grammars with clustered phonetic between word or phrase contexts; and comparing input speech to said models and said generic clustered models and said clustered phonetic between word or phrase contexts of said speech recognition grammar to identify a best match.
Claim 6
Independent Claim
137 words
A method for speech recognition comprising the steps of: providing models comprising within word context models; providing models comprising phonetic between word or phrase context models; clustering said phonetic between word or phrase context models according to linguistic knowledge classes of silence, consonants and vowels to form generic clustered models for silence, consonants and vowels; providing speech recognition grammar; phonetic between word or phrase contexts expanding of said speech recognition grammar; clustering said expanded phonetic between word or phrase contexts of said speech recognition application grammar according to generic classes of silence, consonants, and vowels to form grammars with clustered phonetic between word or phrase contexts; and comparing input speech to said models and said generic clustered models and said clustered phonetic between word or phrase contexts of said speech recognition grammar to identify a best match.
References Cited (Backward Citations)
US4908865A
Cited
US4852279A
Cited
US5438511A
Cited
US4984178A
Cited
US4977598A
Cited
US5418717A
Cited
US5388183A
Cited
US5511213A
Cited
US5384892A
Cited
US5390279A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2001-10-05 | 2002-04-06 | 2002-10-06 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2005-10-05 | 2006-04-06 | 2006-10-06 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2009-10-05 | 2010-04-06 | 2010-10-06 | $7700.00 | Window Closed / Lapsed |