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


 The Delphion Integrated View

  Buy Now:   Buy PDF- 11pp  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: US6421641: Methods and apparatus for fast adaptation of a band-quantized speech decoding system
[ Derwent Title ]


Country: US United States of America

View Images High
Resolution

 Low
 Resolution

 
11 pages

 
Inventor: Huang, Jing; Ossining, NY
Padmanabhan, Mukund; 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: 2002-07-16 / 1999-11-12

Application Number: US1999000438932

IPC Code: Advanced: G10L 15/06;
IPC-7: G10L 17/00;

ECLA Code: G10L15/07;

U.S. Class: 704/250; 704/230; 704/249;

Field of Search: 714/246,255,256,257,240,250,230,249,200

Priority Number:
1999-11-12  US1999000438932

Abstract:     A method of performing speaker adaptation of acoustic models in a band-quantized speech recognition system, wherein the system including one or more acoustic models represented by a feature space of multi-dimensional gaussians, whose dimensions are partitioned into bands, and the gaussian means and covariances within each band are quantized into atoms, comprises the following steps. A decoded segment of a speech signal associated with a particular speaker is obtained. Then, at least one adaptation mapping based on the decoded segment is computed. Lastly, the at least one adaptation mapping is applied to the atoms of the acoustic models to generate one or more acoustic models adapted to the particular speaker. Accordingly, a fast speaker adaptation methodology is provided for use in real-time applications.

Attorney, Agent or Firm: Otterstedt, Paul J.Ryan, Mason & Lewis, LLP ;

Primary / Asst. Examiners: Dorvil, Richemond;

INPADOC Legal Status: Show legal status actions

Family: None

First Claim:
Show all 24 claims
What is claimed is:     1. A method of performing speaker adaptation of acoustic models in a band-quantized speech recognition system, the system including one or more acoustic models represented by a feature space of multi-dimensional gaussians, whose dimensions are partitioned into bands, and gaussian means and covariances within each band are quantized into atoms, the method comprising of the steps of:
  • obtaining a decoded segment of a speech signal associated with a particular speaker;
  • computing at least one adaptation mapping based on the decoded segment; and
  • applying the at least one adaptation mapping to the atoms of the acoustic models to generate one or more acoustic models adapted to the particular speaker.


Background / Summary: Show background / summary

Drawing Descriptions: Show drawing descriptions

Description: Show description

Forward References: Show 7 U.S. patent(s) that reference this one

       
U.S. References: Go to Result Set: All U.S. references   |  Forward references (7)   |   Backward references (5)   |   Citation Link

Buy
PDF
Patent  Pub.Date  Inventor Assignee   Title
Get PDF - 16pp US5199077  1993-03 Wilcox et al.  Xerox Corporation Wordspotting for voice editing and indexing
Get PDF - 17pp US5522011  1996-05 Epstein et al.  International Business Machines Corporation Speech coding apparatus and method using classification rules
Get PDF - 26pp US5793891  1998-08 Takahashi et al.  Nippon Telegraph and Telephone Corporation Adaptive training method for pattern recognition
Get PDF - 25pp US5839105  1998-11 Ostendorf et al.  ATR Interpreting Telecommunications Research Laboratories Speaker-independent model generation apparatus and speech recognition apparatus each equipped with means for splitting state having maximum increase in likelihood
Get PDF - 10pp US6023673  2000-02 Bakis et al.  International Business Machines Corporation Hierarchical labeler in a speech recognition system
       
Foreign References: None

Other References:
  • C.J. Leggetter et al., "Flexible Speaker Adaptation Using Maximum Likelihood Linear Regression," Proceedings of ARPA Spoken Language Technology Workshop, Barton Creek, pp. 1-6, 1995.
  • J-L. Gauvain et al., "Maximum a Posteriori Estimation For Multivate Gaussian Mixture Observations of Markov Chains," IEEE Transactions of Speech and Audio Processing, vol. 2, No. 2, pp. 291-298, Apr. 1994.


  • 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