SPEECH RECOGNITION SYSTEM AND METHOD FOR OPERATING A SPEECH RECOGNITION SYSTEM WITH A MOBILE UNIT AND AN EXTERNAL SERVER

• US 20160372115A1
• Filed: 06/16/2016
• Published: 12/22/2016
• Est. Priority Date: 06/17/2015
• Status: Active Grant

• Current Assignee: Volkswagen AG
• Original Assignee: Volkswagen AG
• Inventors: LDDEKE, Daniel, SPIKA, Marius, SCHNEIDER, Jens, SCHOLZ, Stephan, BERNER, Eva
• Granted Patent: US 10,170,121 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G10L 15/00 Speech recognition G10L17/00 takes precedence
• CPC Class Codes: G10L 15/065 Adaptation
• CPC Class Codes: G10L 15/18 using natural language modelling
• 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
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A voice recognition system having a mobile unit and an external server. The mobile unit includes a memory unit that stores voice model data having at least one expression set with expressions, a voice recognition unit, and a data interface that can set up a data-oriented connection to a data interface of the external server. The external server includes a database with event data having associated time data and expressions. The external server can, by comparing the time data associated with the event data with a current time, produce updated data that includes at least expressions associated with the event data of the database. The update data can be transmitted to the mobile unit. The update data can also be taken as a basis for adding the transmitted expressions to the expression set. Also disclosed is a method for operating a voice recognition system.

Claims

1. A voice recognition system having a mobile unit in communication with an external server,

wherein the mobile unit comprises:

a memory unit that stores voice model data that comprise at least one expression set with expressions;

a voice recognition unit configured to use the voice model data as a basis for producing a recognized text for captured voice input data; and

a data interface that at least intermittently sets up a data-oriented connection to a data interface of the external server,

wherein the external server comprises a database with event data having associated time data and expressions, the external server being configured to produce update for the voice model data by comparing the time data associated with the event data with a current time,

wherein the update data for the voice model data comprises at least expressions associated with the event data of the database,

wherein external server is further configured to transmit the update data from the data interface of the external server to the data interface of the mobile unit, and

wherein the transmitted expressions are added to the expression set based on the update data.

2. The voice recognition system of claim 1, wherein a subset of expressions of the expression set are erased based on the update data.

3. The voice recognition system of claim 1, wherein the transmitted expressions further comprise time data associated with the event data.

4. The voice recognition system of claim 1, wherein the mobile unit further comprises an update unit configured to produce status data about the voice model data and to transmit the status data to the external server, wherein the status data comprise information about at least one foregone change in the expression set by update data of the server and/or about a current volume of the expression set, and the update data is produced based on the status data.

5. The voice recognition system of claim 1, wherein the mobile unit further comprises a context capture unit configured to capture context data and transmit the context data to the external server,

wherein the update data are further produced based on the context data.

6. The voice recognition system of claim 5, wherein the context data is configured to be captured by further devices.

7. The voice recognition system of claim 1, wherein the event data of the database further have associated geographical position data, wherein the voice recognition system further includes a position finding unit configured to determine a current position of the mobile unit, wherein position finding unit is configured to transmit the current mobile unit'"'"'s position to the external server, and wherein the update data are further produced by the external server comparing the geographical position data associated with the event data with the mobile unit'"'"'s position.

8. The voice recognition system of claim 1, wherein the voice recognition unit is further configured to capture statistical data about the captured voice input data and transmit the statistical data to the external server, and wherein the updated data is produced by the external server based on the statistical data.

9. A method for operating a voice recognition system having a mobile unit and an external server, the method comprising:

storing voice model data that comprise at least one expression set with expressions in a memory unit of the mobile unit;

producing a recognized text for captured voice input data based on the voice model data; and

setting up a data-oriented connection to a data interface of the external server at least intermittently, wherein the external server comprises a database with event data that have associated time data and expressions;

producing, by the external server, update data for the voice model data by comparing the time data associated with the event data with a current time, wherein the update data for the voice model data comprises at least expressions associated with the event data of the database; and

transmitting, by the external server, the update data to the mobile unit; and

adding the transmitted expressions to the expression set based on the update data.

10. The method of claim 9, further comprises erasing a subset of expressions of the expression set based on the update data.

11. The method of claim 9, wherein the transmitted expressions further comprise time data associated with the event data.

12. The method of claim 9, wherein the mobile unit further comprises an update unit configured to produce status data about the voice model data and to transmit the status data to the external server, wherein the status data comprise information about at least one foregone change in the expression set by update data of the server and/or about a current volume of the expression set, and the update data is produced based on the status data.

13. The method of claim 9, wherein the mobile unit includes a context capture unit and the method further comprises capturing and transmitting context data, by the context capture unit, to the external server, and producing update data, by the external server, based on the context data.

14. The method of claim 13, wherein the context data is configured to be captured by further devices.

15. The method of claim 9, wherein the event data of the database further have associated geographical position data, wherein the voice recognition system further includes a position finding unit and the method further comprises the position finding unit determining and transmitting a current position of the mobile unit to the external server, and comparing the geographical position data associated with the event data with the mobile unit'"'"'s position by the external server to produce the update data.

16. The method of claim 9, further comprising, the voice recognition unit, capturing statistical data about the captured voice input data and transmitting the statistical data to the external server, and producing, by the external server, the updated data by the external server based on the statistical data.