Dynamic insertion of a speech recognition engine within a distributed speech recognition system

• US 7,133,829 B2
• Filed: 10/31/2001
• Issued: 11/07/2006
• Est. Priority Date: 10/31/2001
• Status: Expired due to Term

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Dictaphone Corporation (Microsoft Corporation)
• Inventors: Green, Channell, Hold, Martin, Kuhnen, Regina, MacGinite, Andrew, Cyr, James
• Primary Examiner(s): Hudspeth, David
• Assistant Examiner(s): ALBERTALLI, BRIAN LOUIS
• Application Number:
  US09/984,873
• Publication Number:
  US 20030083879A1
• Time in Patent Office: 1,833 Days
• Field of Search: 704/270.1, 704/270, 704/231, 704227-229, 709227-229, 709201-203, 713 1- 2, 719/330
• US Class Current: 704/270.1
• CPC Class Codes: G10L 15/285 Memory allocation or algorithm optimisation to reduce hardware requirements
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications

Abstract

A distributed speech recognition system provides for the convenient and expedited insertion of a new speech recognition engine. The system includes a central speech processor linked to at least one speech recognition engine. The at least one speech recognition engine includes means for addressing the central speech processor upon installation within the system. The means for addressing transmits to the central speech processor a message indicating its operating protocol and location to permit the central speech processor to identify and begin transmitting work thereto.

Claims

1. A distributed speech recognition system providing for the convenient and expedited insertion of a new speech recognition engine into said system without rebooting, comprising:

a central speech processor, said central speech processor comprising;

a first server component, said first server component configured to receive a call from a speech recognition engine wrapper, said call containing an address of said speech recognition engine wrapper;

a first client component, said first client component configured to transmit a call to said speech recognition engine wrapper, said call containing an address of said central speech processor; and

a first address memory, said first address memory configured to store said address of said speech recognition engine wrapper;

said speech recognition engine wrapper associated with said central speech processor and with at least one speech recognition engine, said speech recognition engine wrapper comprising;

a second client component, said second client component configured to transmit said call to said central speech processor;

a second server component, said second server component configured to receive said call from said central speech processor;

a second address memory, said second address memory configured to store said address of said central speech processor; and

at least one software application, said at least one software application corresponding with said at least one speech recognition engine;

said at least one speech recognition engine associated with said speech recognition engine wrapper, said at least one speech recognition engine containing at least one recognition grammar corresponding to a speech command;

said at least one software application of said speech recognition engine wrapper configured to communicate with said at least one speech recognition engine upon installation of said at least one speech recognition engine, said at least one software application further configured to alert said second client component of said installation;

said second client component configured to call said first server component, said second client component further configured to provide said first server component with said address of said speech recognition engine wrapper;

said first server component configured to receive said address of said speech recognition engine wrapper and store said address in said first address memory, said first server component further configured to alert said first client component of said stored address;

said first client component configured to call said second server component based on said address stored in said first address memory, said first client component further configured to provide said second server component with any changes with said address of said central speech processor;

said second server component configured to receive any changes with said address of said central speech processor and store said address in said second address memory, said second server component further configured to alert said second client component of said stored address;

said central speech processor configured to recognize said speech recognition engine based on said call by said second client component; and

said central speech processor further configured to transmit work to said speech recognition engine.

2. The distributed speech recognition system according to claim 1, wherein the at least one speech recognition engine is configured to communicate with the central speech processor using remote procedure call.

3. The distributed speech recognition system according to claim 1, wherein the speech recognition engine wrapper is configured to function as an interface between the central speech processor and the at least one speech recognition engine.

4. The distributed speech recognition system to claim 3, wherein the at least one speech recognition engine is configured to communicate communicates with the central speech processor using remote procedure call.

5. The distributed speech recognition system according to claim 1, wherein the at least one speech recognition engine is configured to communicate with the central speech processor via remote procedure call.

6. The distributed speech recognition system according to claim 1, wherein a plurality of speech recognition engines are associated with said speech recognition engine wrapper.

7. The distributed speech recognition system according to claim 1, wherein the speech recognition engine wrapper has a first interface for communication with the at least one speech recognition engine and a second interface for communication with the central speech processor.

8. A method for the convenient and expedited insertion of a new speech recognition engine within a distributed speech recognition system without rebooting, comprising the following steps:

providing a central speech processor, said central speech processor comprising;

a first server component, said first server component configured to receive a call from a speech recognition engine wrapper, said call containing an address of said speech recognition engine wrapper;

a first client component, said first client component configured to transmit a call to said speech recognition engine wrapper, said call containing an address of said central speech processor; and

a first address memory, said first address memory configured to store said address of said speech recognition engine wrapper;

associating said central speech processor with said speech recognition engine wrapper, said speech recognition engine wrapper comprising;

a second client component, said second client component configured to transmit said call to said central speech processor;

a second server component, said second server component configured to receive said call from said central speech processor;

a second address memory, said second address memory configured to store said address of said central speech processor; and

at least one software application, said at least one software application corresponding with at least one speech recognition engine;

associating said speech recognition engine wrapper with said at least one speech recognition engine, said at least one speech recognition engine containing at least one recognition grammar corresponding to a speech command;

said at least one software application of said speech recognition engine wrapper communicating with said at least one speech recognition engine upon installation of said at least one speech recognition engine, said at least one software application alerting said second client component of said installation;

said second client component calling said first server component, said second client component providing said first server component with said address of said speech recognition engine wrapper;

said first server component receiving said address of said speech recognition engine wrapper and storing said address in said first address memory, said first server component alerting said first client component of said stored address;

said first client component calling said second server component based on said address stored in said first address memory, said first client component providing said second server component with any changes with said address of said central speech processor;

said second server component receiving any changes with said address of said central speech processor and storing said address in said second address memory, said second server component alerting said second client component of said stored address;

said central speech processor recogizing said speech recognition engine based on said call by said second client component; and

said central speech processor transmitting work to said speech recognition engine.

9. The method according to claim 8, wherein the at least one speech recognition engine communicates with the central speech processor using remote procedure call.

10. The method according to claim 8, wherein the speech recognition engine wrapper functions as an interface between the central speech processor and the at least one speech recognition engine.

11. The method according to claim 10, wherein at least one speech recognition engine communicates with the central speech processor using remote procedure call.

12. The method according to claim 8, wherein the at least one speech recognition engine communicates with the central speech processor via remote procedure call.

13. The method according to claim 8, wherein the speech recognition engine wrapper has a first and a second interface such that the at least one speech recognition engine is configured to communicate with the speech recognition engine wrapper via the first interface, which is configured to communicate with the central speech processor via the second interface.

14. A distributed speech recognition system providing for the convenient and expedited insertion of a new speech recognition engine without rebooting, comprising:

a speech recognition engine wrapper having at least a first and a second interface, said speech recognition engine wrapper comprising;

a client component, said client component configured to transmit a call to a central speech processor, said call containing an address of said speech recognition engine wrapper;

a server component, said server component configured to receive a call from said central speech processor, said call containing an address of said central speech processor;

an address memory, said address memory configured to store said address of said central speech processor; and

at least one software application, said at least one software application corresponding with said at least one speech recognition engine;

said central speech processor linked to said first interface, said central speech processor comprising;

a server component, said server component configured to receive said call from said speech recognition engine wrapper;

a client component, said client component configured to transmit said call to said speech recognition engine wrapper; and

an address memory, said address memory configured to store said address of said speech recognition engine wrapper;

said at least one speech recognition engine linked to the second interface of the wrapper, said at least one speech recognition engine containing at least one recognition grammar corresponding to a speech command;

said at least one software application of said speech recognition engine wrapper configured to communicate with said at least one speech recognition engine upon installation of said at least one speech recognition engine, said at least one software application further configured to alert said client component of said speech recognition engine wrapper of said installation;

said client component of said speech recognition engine wrapper configured to call said server component of said central speech processor, said client component farther configured to provide said server component with said address of said speech recognition engine wrapper;

said server component of said central speech processor configured to receive said address of said speech recognition engine wrapper and store said address in said address memory of said central speech processor, said server component further configured to alert said client component of said central speech processor of said stored address;

said client component of said central speech processor configured to call said server component of said speech recognition engine wrapper based on said address stored in said address memory of said central speech processor, said client component further configured to provide said server component with any changes with said address of said central speech processor;

said server component of said speech recognition engine wrapper configured to receive any changes with said address of said central speech processor and store said address in said address memory of said speech recognition engine wrapper, said server component further configured to alert said client component of said speech recognition engine wrapper of said stored address;

said central speech processor configured to recognize said speech recognition engine based on said call by said client component of said speech recognition engine wrapper; and

said central speech processor further configured to transmit work to said speech recognition engine.

15. The distributed speech recognition system according to claim 14, wherein the at least one speech recognition engine is configured to communicate with the speech recognition engine wrapper using remote procedure call.

16. The distributed speech recognition system according to claim 14, wherein the speech recognition engine wrapper is configured to communicate with the central speech processor using remote procedure call.