Method and apparatus for controlling multiple speech engines in an in-vehicle speech recognition system

• US 6,230,138 B1
• Filed: 06/28/2000
• Issued: 05/08/2001
• Est. Priority Date: 06/28/2000
• Status: Expired due to Fees

• Current Assignee: Visteon Global Technologies Incorporated (Visteon Corporation)
• Original Assignee: Visteon Global Technologies Incorporated (Visteon Corporation)
• Inventors: Everhart, Charles Allen
• Primary Examiner(s): Smits, Talivaldis I.
• Assistant Examiner(s): Nolan, Daniel A
• Application Number:
  US09/605,253
• Time in Patent Office: 314 Days
• Field of Search: 704/500, 704/231, 704/272, 704/270, 704/271, 704/275
• US Class Current: 704/275
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 15/32 Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command
• CPC Class Codes: G10L 2015/228 of application context

Abstract

Disclosed herein is a method and apparatus for controlling a speech recognition system on board an automobile. The automobile has one or more voice activated accessories and a passenger cabin with a number of seating locations. The speech recognition system has a plurality of microphones and push-to-talk controls corresponding to the seating locations for inputting speech commands and location identifying signals, respectively. The speech recognition system also includes multiple speech engines recognizing speech commands for operating the voice activated accessories. A selector is coupled to the speech engines and push-to-talk controls for selecting the speech engine best suited for the current speaking location. A speech processor coupled to the speech engine selector is used to recognize the speech commands and transmit the commands to the voice activated accessory.

Claims

1. A speech recognition system in an automobile having at least one voice activated accessory and a passenger cabin having seats defining a plurality of speaking locations, comprising:

a plurality of push-to-talk controls mounted within said passenger cabin so as to correspond to said plurality of speaking locations, each of said plurality of push-to-talk controls supplying a location signal when operated;

a plurality of microphones mounted within said passenger cabin so as to correspond to said plurality of speaking locations, each of said plurality of microphones being capable of receiving a spoken utterance from a speaker at said plurality of speaking locations and providing a corresponding audio signal;

a plurality of speech engines, each having different recognition parameters, for recognizing said audio signal as a speech command for controlling said at least one voice activated accessory; and

a selector coupled to said plurality of push-to-talk controls, microphones and speech engines, said selector selecting one of said plurality of speech engines according to said location signal and providing said audio signal to said selected speech engine.

2. The system of claim 1, wherein said selector is coupled to said plurality of push-to-talk controls via a vehicle network and wherein said selector processes said location signal to determine said selected one of said plurality of speech engines.

3. The system of claim 1, wherein said selector has separate inputs for each of said plurality of push-to-talk controls and wherein said inputs are coupled to a particular one of said plurality of speech engines.

4. The system of claim 1, wherein said selector provides a selection signal to said selected speech engine corresponding to said location signal.

5. The system of claim 4, wherein at least one of said plurality of speech engines contains a plurality of grammar sets each containing different sets of voice commands.

6. The system of claim 5, wherein said selector selects an active grammar according to said location signal.

7. The system of claim 6, wherein said selection signal indicates said selected active grammar set.

8. The system of claim 1, wherein said speech recognition characteristics are set according to acoustic modeling parameters.

9. The system of claim 8, wherein each of said plurality of speech processors have different acoustic modeling parameters for processing audio signals corresponding to spoken utterances originating at speaking locations having at least one of different acoustic characteristics and microphone placement.

10. The system of claim 1, wherein said selector and said plurality of speech engines are included in a central control unit.

11. The system of claim 10, wherein said voice central control unit includes a memory for storing voice commands.

12. The system of claim 1, wherein said selected speech engine provides a speech command to a controller for said at least one voice activated accessory.

13. In an automobile having a passenger cabin with a plurality of speaking locations and a speech recognition system with multiple speech engines for controlling at least one voice activated accessory, a method for operating said at least one voice activated accessory via a spoken utterance issue by a speaker at one of said plurality of speaking locations, comprising the steps of:

receiving a location signal from one of a plurality of push-to-talk controls mounted within said passenger cabin and operated by said speaker at said one of said plurality of speaking locations;

selecting, according to said location signal, one of said multiple speech engines having different recognition parameters and having speech commands for operating said at least one voice activated accessory;

supplying said selected speech processor an audio signal corresponding to said spoken utterance received by one of a plurality of microphones mounted within said passenger cabin and corresponding to said plurality of speaking locations;

processing said spoken utterance to identify a corresponding speech command; and

outputting said corresponding speech command to a control for said at least one voice activated accessory.

14. The method of claim 13, further comprising the step of supplying a selection signal corresponding to said location signal to said selected speech engine.

15. The system of claim 14, wherein at least one of said multiple speech engines contains a plurality of grammar sets each containing different sets of voice commands.

16. The system of claim 15, further comprising the step of selecting an active grammar set according to said location signal.

17. The system of claim 16, wherein said selection signal indicates said selected active grammar set.

18. The method of claim 13, wherein said speech recognition characteristics are set according to acoustic modeling parameters.

19. The system of claim 18, wherein each of said multiple speech engines have different acoustic modeling parameters for processing audio signals corresponding to spoken utterances originating at speaking locations having at least one of different acoustic characteristics and microphone placement.