Method, module, device and server for voice recognition

• US 20050102142A1
• Filed: 02/12/2002
• Published: 05/12/2005
• Est. Priority Date: 02/13/2001
• Status: Active Grant

• Current Assignee: InterDigital Patent Holdings, Inc. (InterDigital, Inc.)
• Original Assignee: Thomson Licensing (Vantiva SA)
• Inventors: Soufflet, Frederic, Tazine, Nour-Eddine
• Granted Patent: US 7,983,911 B2
• Time in Patent Office: Days
• US Class Current: 704/246
• CPC Class Codes: G10L 15/18 using natural language modelling
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications

Abstract

The invention concerns a voice recognition method implemented in at least a terminal, the voice recognition method using a language model, comprising the following steps: detecting at least a non-recognised expression in one of the terminals; recording in the terminal data representing the non-recognised expression; transmission by the terminal of the recorded data to a remote server; analysis, at the remote server, of said data and generation of data for correcting said language model taking into account at least part of the non-recognised expressions; and transmitting from the server to at least a terminal correction data, so as to enable subsequent recognition of at least some of the non-recognised expressions. The invention also concerns corresponding modules, devices and remote server.

Claims

1. Voice recognition process implemented in at least one terminal, the said voice recognition process using a language model, wherein it comprises the following steps:

detection of at least one unrecognized expression in one of the said terminals;

recording in the said terminal of data representative of the said unrecognized expression;

transmission by the said terminal of the said recorded data to a remote server, via a first transmission channel;

analysis, at the level of the said remote server, of the said data and generation of information for correcting the said language model taking account of at least one part of the said unrecognized expressions; and

transmission via a second transmission channel from the said server to at least one terminal of the said correcting information, so as to allow future recognition of at least certain of the said unrecognized expressions.

2. Process according to claim 1, wherein the said data representative of the said unrecognized expressions comprise a compressed voice recording representative of parameters descriptive of the acoustic signal.

3. Process according to claim 1, wherein during the said step of transmission by the said terminal, the latter furthermore transmits to the said server at least one of the items of information forming part of the group comprising:

information of context of use of the said voice recognition process when an expression has not been recognized; and

information relating to the speaker who has uttered an unrecognized expression.

4. Process according to claim 1, wherein it implements an encryption and/or a scrambling of the said recorded data and/or of the said correcting information.

5. Voice recognition module using a language model, wherein it comprises:

an analyser detecting unrecognized expressions;

a recorder of data representative of at least one unrecognized expression;

a transmitter transmitting the said recorded data to a remote server; and

a receiver of correcting information allowing the correcting of the said language model transmitted to the said module allowing future recognition of at least certain of the said unrecognized expressions by the said module, the correcting information having been transmitted by the said remote server after analysis at the level of the said remote server of the said data, and after generation of information for correcting the said language model taking account of at least one part of the unrecognized expressions.

6. Voice recognition device using a language model, wherein it comprises:

an analyser detecting unrecognized expressions;

a recorder of data representative of at least one unrecognized expression;

a transmitter transmitting the said recorded data to a remote server; and

a receiver of correcting information allowing the correcting of the said language model transmitted to the said device allowing future recognition of at least certain of the said unrecognized expressions by the said device, the correcting information having been transmitted by the said remote server after analysis at the level of the said remote server of the said data, and after generation of information for correcting the said language model taking account of at least one part of the unrecognized expressions.

7. Voice recognition server, the said recognition being implemented in a set of at least one remote terminal, using a language model, wherein it comprises the following means:

a receiver of data representative of at least one expression unrecognized by at least one terminal forming part of the said set and having detected the said unrecognized expression during a voice recognition operation; and

a sender sending to the said set of at least one remote terminal correcting information obtained on the basis of an analysis of the said data received at the level of the said server, the said correcting information allowing the correcting by each of the terminals of the said set, of the said language model allowing future recognition of at least one part of the unrecognized expressions.