Work Files Saved Searches
   My Account                                                  Search:   Quick/Number   Boolean   Advanced       Help   


 The Delphion Integrated View

  Buy Now:   Buy PDF- 14pp  PDF  |   File History  |   Other choices   
  Tools:  Citation Link  |  Add to Work File:    
  View:  Expand Details   |  INPADOC   |  Jump to: 
 
 Email this to a friend  Email this to a friend 
       
Title: US5233681: Context-dependent speech recognizer using estimated next word context
[ Derwent Title ]


Country: US United States of America

View Images High
Resolution

 Low
 Resolution

 
14 pages

 
Inventor: Bahl, Lalit R.; Amawalk, NY
De Souza, Peter V.; Mahopac Falls, NY
Gopalakrishnan, Ponani S.; Croton-on-Hudson, NY
Picheny, Michael A.; White Plains, NY

Assignee: International Business Machines Corporation, Armonk, NY
other patents from INTERNATIONAL BUSINESS MACHINES CORPORATION (280070) (approx. 44,393)
 News, Profiles, Stocks and More about this company

Published / Filed: 1993-08-03 / 1992-04-24

Application Number: US1992000874271

IPC Code: Advanced: G10L 15/10; G10L 15/18; G10L 15/28;
IPC-7: G10L 91/00;

ECLA Code: G10L15/19; S10L15/193;

U.S. Class: Current: 704/251; 704/252; 704/255; 704/E15.021;
Original: 395/002; 381/043; 381/041;

Field of Search: 381/041,43,51 395/002

Priority Number:
1992-04-24  US1992000874271

Abstract: A speech recognition apparatus and method estimates the next word context for each current candidate word in a speech hypothesis. An initial model of each speech hypothesis comprises a model of a partial hypothesis of zero or more words followed by a model of a candidate word. An initial hypothesis score for each speech hypothesis comprises an estimate of the closeness of a match between the initial model of the speech hypothesis and a sequence of coded representations of the utterance. The speech hypotheses having the best initial hypothesis scores form an initial subset. For each speech hypothesis in the initial subset, the word which is most likely to follow the speech hypothesis is estimated. A revised model of each speech hypothesis in the initial subset comprises a model of the partial hypothesis followed by a revised model of the candidate word. The revised candidate word model is dependent at least on the word which is estimated to be most likely to follow the speech hypothesis. A revised hypothesis score for each speech hypothesis in the initial subset comprises an estimate of the closeness of a match between the revised model of the speech hypothesis and the sequence of coded representations of the utterance. The speech hypotheses from the initial subset which have the best revised match scores are stored as a reduced subset. At least one word of one or more of the speech hypotheses in the reduced subset is output as a speech recognition result.

Attorney, Agent or Firm: Schechter, Marc D. ;

Primary / Asst. Examiners: Fleming, Michael R.; Hafie, Tariq R.

Maintenance Status: E2 Expired  Check current status

INPADOC Legal Status: Show legal status actions          Buy Now: Family Legal Status Report

Family: Show 7 known family members

First Claim:
Show all 31 claims
We claim:     1. A speech recognition apparatus comprising:
  • means for generating a set of two or more speech hypotheses, each speech hypothesis comprising a partial hypothesis of zero or more words followed by a candidate word selected from a vocabulary of candidate words;
  • means for storing a set of word models, each word model representing one or more possible coded representations of an utterance of a word;
  • means for generating an initial model of each speech hypothesis, each initial model comprising a model of the partial hypothesis followed by a model of the candidate word;
  • an acoustic processor for generating a sequence of coded representations of an utterance to be recognized;
  • means for generating an initial hypothesis score for each speech hypothesis, each initial hypothesis score comprising an estimate of the closeness of a match between the initial model of the speech hypothesis and the sequence of coded representations of the utterance;
  • means for storing an initial subset of one or more speech hypotheses, from the set of speech hypotheses, having the best initial hypothesis scores;
  • next context estimating means for estimating, for each speech hypothesis in the initial subset, a likely word, from the vocabulary of words, which is likely to follow the speech hypothesis;
  • means for generating a revised model of each speech hypothesis in the initial subset, each revised model comprising a model of the partial hypothesis followed by a revised model of the candidate word, the revised candidate word model being dependent at least on the word which is estimated to be likely to follow the speech hypothesis;
  • means for generating a revised hypothesis score for each speech hypothesis in the initial subset, each revised hypothesis score comprising an estimate of the closeness of a match between the revised model of the speech hypothesis and the sequence of coded representations of the utterance;
  • means for storing a reduced subset of one or more speech hypotheses, from the initial subset of speech hypotheses, having the best revised match scores; and
  • means for outputting at least one word of one or more of the speech hypotheses in the reduced subset.


Background / Summary: Show background / summary

Drawing Descriptions: Show drawing descriptions

Description: Show description

Forward References: Show 45 U.S. patent(s) that reference this one

       
U.S. References: Go to Result Set: All U.S. references   |  Forward references (45)   |   Backward references (11)   |   Citation Link

Buy
PDF
Patent  Pub.Date  Inventor Assignee   Title
Get PDF - 54pp US4748670  1988-05 Bahl et al.  International Business Machines Corporation Apparatus and method for determining a likely word sequence from labels generated by an acoustic processor
Get PDF - 12pp US4751737  1988-06 Gerson et al.  Motorola Inc. Template generation method in a speech recognition system
Get PDF - 35pp US4759068  1988-07 Rahl et al.  International Business Machines Corporation Constructing Markov models of words from multiple utterances
Get PDF - 33pp US4783804  1988-11 Juang et al.  American Telephone and Telegraph Company, AT&T Bell Laboratories Hidden Markov model speech recognition arrangement
Get PDF - 50pp US4977599  1990-12 Bohl et al.  International Business Machines Corporation Speech recognition employing a set of Markov models that includes Markov models representing transitions to and from silence
Get PDF - 45pp US4980918  1990-12 Bahl et al.  International Business Machines Corporation Speech recognition system with efficient storage and rapid assembly of phonological graphs
Get PDF - 31pp US5033087  1991-07 Bahl et al.  International Business Machines Corp. Method and apparatus for the automatic determination of phonological rules as for a continuous speech recognition system
Get PDF - 16pp US5054074  1991-10 Bukis  International Business Machines Corporation Optimized speech recognition system and method
Get PDF - 28pp US5072452  1991-12 Brown et al.  International Business Machines Corporation Automatic determination of labels and Markov word models in a speech recognition system
Get PDF - 16pp US5129001  1992-07 Buhl et al.  International Business Machines Corporation Method and apparatus for modeling words with multi-arc markov models
Get PDF - 24pp US5131043  1992-07 Fujii et al.  Matsushita Electric Industrial Co., Ltd. Method of and apparatus for speech recognition wherein decisions are made based on phonemes
       
Foreign References: None

Other References:
  • Bahl, L. R., et al. "Context Dependent Modeling of Phones in Continuous Speech Using Decision Trees", pp. 264-269.
  • Bahl, L. R., et al. "A Maximum Likelihood Approach to Continuous Speech Recognition", IEEE Transactions On Pattern Analysis and Machine Intelligence, vol. PAMI-5, No. 2 Mar. 1983, V33, N9, Feb. 1991, pp. 179-190. (12 pages) Cited by 42 patents
  • Bahl, L. R., "Speech Recognition Apparatus Having a Speech Coder Outputting Acoustic Prototype Ranks," U.S. Patent Application Ser. No. 781,440, filed Oct. 23, 1992.
  • Bahl, L. R. et al. "Apparatus and Method of Grouping Utterances of a Phoneme Into Context-Dependent Categories Based on Sound-Similarity for Automatic Speech Recognition," U.S. Pat. Application Ser. No. 468,546, filed Jan. 23, 1990.


  • Inquire Regarding Licensing

    Powered by Verity


    Plaques from Patent Awards      Gallery of Obscure PatentsNominate this for the Gallery...

    Thomson Reuters Copyright © 1997-2014 Thomson Reuters 
    Subscriptions  |  Web Seminars  |  Privacy  |  Terms & Conditions  |  Site Map  |  Contact Us  |  Help