Process and apparatus for real-time verbal input of a target address of a target address system

• US 6,230,132 B1
• Filed: 03/10/1998
• Issued: 05/08/2001
• Est. Priority Date: 03/10/1997
• Status: Expired due to Term

• Current Assignee: Harman Becker Automotive Systems GmbH (Samsung Electronics Co. Ltd.)
• Original Assignee: Daimler Chrysler Company LLC (Stellantis NV)
• Inventors: Class, Fritz, Kuhn, Thomas, Moeller, Carsten-Uwe, Reh, Frank, Nuessle, Gerhard
• Primary Examiner(s): Zele, Krista
• Assistant Examiner(s): Opsasnick, Michael N.
• Application Number:
  US09/038,147
• Time in Patent Office: 1,155 Days
• Field of Search: 704/270, 704/276, 701/210
• US Class Current: 704/270
• CPC Class Codes: G01C 21/3608 using speech input, e.g. using speech recognition
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

In a method for real time speech input of a destination address into a navigation system, the speech statements that are entered by a user are recognized by a speech recognition device and classified in accordance with their recognition probability. The speech statement with the greatest recognition probability is identified as the input speech statement, with at least one speech statement being an admissible speech command that activates the operating functions of the navigation system associated with this speech command. (All the admissible speech statements being stored in at least one database.) According to the invention, at least one operating function of the navigation system comprises an input dialogue. Following activation of that operating function, depending on the input dialogue, at least one lexicon is generated in real time from the admissible speech statements stored in at least one database, and the generated lexicon is loaded as vocabulary into the speech recognition device.

Claims

1. Method for speech input of a destination address into a navigation system in real time, in which the entered speech of a user is recognized by a speech recognition device, with at least one speech statement being an admissible speech statement that activates at least an associated operating function of the navigation system, all of the admissible speech statements being stored in at least one database, comprising:

activating at least one operating function of the navigation system and an input dialogue mode for communicating with the navigation system, by means of an admissible speech statement;

generating at least one lexicon in real time following activation of the at least one operating function, a word content of the at least one lexicon being selected to include only words which can be used to communicate via said activated input dialogue mode and said speech recognition device;

loading the at least one lexicon as a vocabulary into the speech recognition device; and

using words in said lexicon to input information into said navigation system via said speech recognition device.

2. Method according to claim 1 wherein:

the at least one lexicon is generated from admissible speech statements stored in at least one database in an off-line editing mode; and

the at least one lexicon generated in the off-line editing mode is loaded in real time as the vocabulary into the speech recognition device after activation of the at least one operating function of the navigation system depending on the at least one input dialogue.

3. Method according to claim 1 wherein:

at least one speech statement is at least one of a place name and a street name, with all admissible place names being stored in a destination file, and all admissible street names for at least one admissible place name being stored in a street list.

4. Method according to claim 1 wherein the speech recognition device comprises at least one speaker-independent speech recognition engine and at least one speaker-dependent additional speech recognition engine, whereby, based on an input dialogue, the speaker-independent speech recognition engine for recognizing place names, street names, or letters spoken one at a time or in groups or parts of words is used, and the speaker-dependent additional speech recognition engine is used to recognize at least one spoken keyword.

5. Method according to claim 4 wherein a particular destination address is assigned to the at least one keyword, with the at least one spoken keyword being stored in a personal address list, a name lexicon being generated from the personal address list and loaded into the speech recognition device.

6. Method according to claim 2 wherein a basic lexicon generated in the off-line editing mode contains a predetermined number of the largest places in a geographic area.

7. Method according to claim 6 wherein the basic lexicon is stored in an internal nonvolatile memory of the navigation system.

8. Method according to claim 1 wherein an environment lexicon generated in real time contains a predetermined number of locations in an area of current vehicle location, with the environment lexicon being updated at regular intervals.

9. Method according to claim 8 wherein the environment lexicon is stored in an internal, nonvolatile memory of the navigation system.

10. Method according to claim 3 wherein:

following activation of an input dialogue “

spell destination location,”

a partial-word lexicon for letter recognition is loaded into the speech recognition device;

the user then enters individual letters and/or letter groups as speech statements, which are compared in the speech recognition device with the partial-word lexicon, a hypothesis list with word hypotheses being formed from the recognized letters and/or letter groups;

a predetermined number of the word hypotheses are then compared with a destination file and a whole-word lexicon is generated from the result of the comparison, and is loaded into the speech recognition device for whole-word recognition; and

a stored acoustic value in the speech recognition device is then compared with the whole-word lexicon for whole-word recognition, with this acoustic value being generated from a speech statement spoken as a whole word prior to the loading of the partial-word lexicon.

11. Method according to claim 1 wherein:

following recognition of a “

coarse destination”

entered by means of an input dialogue “

enter coarse destination,”

the navigation system calculates in real time a preset number of locations in the area around the location “

coarse destination”

; and

from the preset number of locations a fine destination lexicon is generated and loaded into the speech recognition device.

12. Apparatus for speech input of destination information into a navigation system, said apparatus comprising:

a speech recognition device, for recognizing a spoken speech statement as an admissible speech command;

a dialogue and process control which activates an operating function of the navigation system that is associated with a particular speech command, in response to recognition by said speech recognition device of a speech statement spoken by a user, as the particular speech command; and

at least one database storing all admissible speech statements;

wherein

each operating function has at least one dialogue mode associated therewith; and

in response to recognition by said speech recognition device of an input dialogue mode associated with the activated operating function, the dialogue and process control can generate at least one lexicon in real time, a word content of said lexicon be selected from admissible speech statements stored in the at least one database, and being limited to words which can be to communicate via the recognized dialogue mode which lexicon can be loaded as vocabulary into the speech recognition device.

13. Device according to claim 12 wherein:

the dialogue and process control generates the at least one lexicon in an off-line editing mode; and

the at least one lexicon is stored in at least one database, and is loaded in real time as vocabulary into the speech recognition device.

14. Method for voice actuated entry of input information into a computer which is programmed to perform operating functions, comprising:

storing all admissible speech statements in a memory;

processing entered speech of a user by means of a speech recognition device which classifies said speech as at least one admissible speech statement according to its recognition probability;

identifying a speech statement with the greatest recognition probability as speech which was entered, at least one speech statement being an admissible speech statement for activating an associated operating function in said computer;

activating at least one operating function in said computer in response to identification of an input speech statement,

generating at least one lexicon in real time following activation of said at least one operating function in said computer, in response to an input speech statement identified from among said admissible speech statements, said lexicon comprising a subset of admissible speech statements which are selected from admissible speech statements stored in said memory, and which can be used to communicate information for said activated operating function;

entering said lexicon into said speech recognition device; and

using said lexicon as a vocabulary in said speech recognition device, for controlling implementation of said operating function.

15. A method of operating a vehicle navigation system having a speech recognition unit and a stored permissible vocabulary for communication via said speech recognition unit, said method comprising:

providing a plurality of dialogue modes for interactive voice communication of information between an operator and the vehicle navigation system via the speech recognition unit;

providing in each particular dialogue mode measures for establishing at least one associated lexicon comprising a limited subset of vocabulary words from said permissible vocabulary, said subset including only words which are usable for communicating information according to the particular dialogue mode;

selecting a dialogue mode in response to entry of a voice command into said vehicle navigation system via said speech recognition unit;

generating an associated lexicon for the selected dialogue mode;

loading said associated lexicon into said speech recognition system; and

entering information into said navigation system via said speech recognition system according to said lexicon and said selected dialogue mode.

16. The method according to claim 15 wherein, in an initialization step, a basic vocabulary and a vocabulary of speech recognition commands are first loaded into said speech recognition unit.

17. The method according to claim 16, wherein said basic vocabulary contains words descriptive of a set of largest cities in a geographic area of interest.