Back to results
US 2014/0004826 A1Published application
System and method of speaker recognition
AssigneeHoneywell International Inc.
InventorsKenneth L. Addy, Ronald K. Rothman
Life of this patent
Estimating the term…
Abstract
An authentication and authorization apparatus combines a unique identifier for a communications device with pre-stored voice recognition information. Incoming audio, associated with the unique identifier is processed using the pre-stored vice recognition information to authenticate the speaker. In response to successful authentication, a requested function or action embedded in the audio can be recognized and, if authorized, implemented.
Claims20
An apparatus comprising:
- circuits that sense selected identification information for a wireless communication device;
- circuits that select pre-stored voice related information associated with the identification information;
- circuits that carry out an authentication process of incoming audio using the selected information;
- and
- circuits that respond to the results of the authentication process.
2 more elements
A speech authentication and authorization combination comprising:
- circuitry to associate a unique identifier for a communications device with pre-stored voice recognition information for the authorized user of that device;
- circuitry to receive a stream of incoming audio from a user requesting that a function be implemented in a displaced system;
- circuitry to determine if at least part of the incoming audio corresponds to the pre-stored recognition information associated with the unique identifier;
- and
- circuitry to implement the requested function.
2 more elements
A method comprising:
- establishing a data base having identification indicia linked to voice recognition information for each member of a plurality of persons;
- receiving a stream of audio from a person;
- associating an identifier, connected to the person, with the stream of audio;
- retrieving voice recognition information from the data base, using the identifier,
- carrying out voice recognition processing between the retrieved voice recognition information, and the representation of the stream of audio;
- and
- responsive to the results of the processing, determining if the voice recognition information and the representation of the stream of audio correspond to the same person.
4 more elements
Dependencies are inferred from the claim text.