METHOD AND DEVICE FOR THE NATURAL-LANGUAGE RECOGNITION OF A VOCAL EXPRESSION

• US 20100114577A1
• Filed: 06/14/2007
• Published: 05/06/2010
• Est. Priority Date: 06/27/2006
• Status: Active Grant

• Current Assignee: Deutsche Telekom AG
• Original Assignee: Deutsche Telekom AG
• Inventors: Markefka, Guntbert, Hayn, Ekkehard, Liedtke, Klaus-Dieter
• Granted Patent: US 9,208,787 B2
• Time in Patent Office: Days
• US Class Current: 704/257
• CPC Class Codes: G10L 15/19 Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules
• CPC Class Codes: G10L 15/32 Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems

Abstract

The invention relates to a method and a device for the natural-language recognition of a vocal expression. A vocal expression of a person is detected and converted into a voice signal to be processed by a voice recognition device. Afterwards, the voice signal is analyzed at the same time or sequentially in a plurality of voice recognition branches of the voice recognition device using a plurality of grammars, wherein the recognition process is successfully completed if the analysis of the voice signal in at least one voice recognition branch supplies a positive recognition result.

Claims

1. A method for the natural language recognition of a vocal expression, with the steps:

detection of the vocal expression and conversion into a voice signal to be processed by a voice recognition device,

sequential analysis of the voice signal in a plurality of voice recognition branches of the voice recognition device using a plurality of grammars and

successful completion of the recognition process of the vocal expression, in case the analysis of the voice signal supplies a positive recognition result in at least one voice recognition branch, characterized by the steps;

a) feeding of the voice signal to a first voice recognition branch comprising a first grammar for analysis of the voice signal,

b) analysis of the voice signal by the first grammar, wherein in the case of a recognition of the vocal expression a positive first recognition is generated and the recognition process is concluded and in the case of a non-recognition of the vocal expression a negative first recognition result is generated,

c) wherein in the case of negative recognition result the voice signal is fed to a further voice recognition branch comprising a further grammar.

d) analysis of the voice signal by the further grammar, wherein in the case of a recognition of the vocal expression a positive recognition is generated and the recognition process is concluded and in the case of a non-recognition of the vocal expression a negative first recognition result is generated,

e) wherein in the case of a negative recognition result the method continues with step (c) until the grammars of all existing voice recognition branches have been run.

2. The method according to claim 1, characterized in that the sets of rules of the grammars do not exhibit a common intersection.

3. The method according to claim 1, characterized in that a first grammar analyzes frequently occurring vocal expressions, a second grammar analyzes less frequently occurring vocal expressions and any further grammar analyzes even less frequently occurring vocal expressions.

4. The method according to claim 1, characterized in that when both the first and the second recognition result are positive, the recognition result supplied by the first grammar is used.

5. The method according to claim 1, characterized in that when both the first and the second recognition result are positive, the recognition result whose recognition reliability is the greatest is used.

7. A computer program with a program code which executes on a computer a method according to claim 1.

8. A computer program product which comprises program code executable on a computer for the carrying out of the method according to claim 1.

9. The method according to claim 2, characterized in that a first grammar analyzes frequently occurring vocal expressions, a second grammar analyzes less frequently occurring vocal expressions and any further grammar analyzes even less frequently occurring vocal expressions.

10. The method according to claim 2, characterized in that when both the first and the second recognition result are positive, the recognition result supplied by the first grammar is used.

11. The method according to claim 3, characterized in that when both the first and the second recognition result are positive, the recognition result supplied by the first grammar is used.

12. The method according to claim 2, characterized in that when both the first and the second recognition result are positive, the recognition result whose recognition reliability is the greatest is used.

13. The method according to claim 3, characterized in that when both the first and the second recognition result are positive, the recognition result whose recognition reliability is the greatest is used.

14. A computer program with a program code which executes on a computer a method according to claim 2.

15. A computer program with a program code which executes on a computer a method according to claim 3.

16. A computer program with a program code which executes on a computer a method according to claim 4.

17. A computer program with a program code which executes on a computer a method according to claim 5.

18. A computer program product which comprises program code executable on a computer for the carrying out of the method according to claim 2.

19. A computer program product which comprises program code executable on a computer for the carrying out of the method according to claim 3.

20. A computer program product which comprises program code executable on a computer for the carrying out of the method according to claim 4.

6. A device for the natural language recognition of a vocal expression which comprises:

means for detection of the vocal expression and for conversion into a voice signal to be processed by a voice recognition device,

a voice recognition device with a plurality of voice recognition branches, wherein each voice recognition branch exhibits a grammar for the analysis of the voice signal, wherein the voice signal is fed to the voice recognition branches sequentially, and

means for the control and evaluation of the recognition process in dependency on the recognition result of at least one voice recognition branch.