Single input voice authentication
AssigneeCisco Technology, Inc.
InventorEric Yi-Hua Chen
Life of this patent
Estimating the term…
Abstract
A system and method for authenticating a user via a single voice audio input is disclosed. The method includes obtaining voice audio from a user, converting the voice audio to text, comparing at least a portion of the text to a database, determining whether a user profile exists in the database based on the comparing, the user profile including a voiceprint, in response to determining that the user profile exists in the database, analyzing the voice audio against the voiceprint of the user profile; and authenticating the user based on the voice audio substantially matching the voiceprint of the user profile.
Claims20
A method comprising:
- obtaining voice audio from a user;
- converting the voice audio to text;
- determining an attempt to access a resource based on the text;
- determining an identifier included in the text;
- comparing the identifier to a plurality of identifiers stored in a database in response to determining the identifier, wherein each identifier of the plurality of identifiers corresponds to a respective user profile in the database;
- retrieving a user profile based on the user profile corresponding to the identifier stored in the database, the user profile including a voiceprint;
- in response to retrieving the user profile, comparing the voice audio to the voiceprint of the user profile;
- authenticating the user based on the voice audio substantially matching the voiceprint of the user profile;
- and
- providing access to the resource in response to authenticating the user.
An apparatus comprising:
- a non-transitory computer readable storage medium;
- and
- a processor configured to:
- obtain voice audio from a user;
- convert the voice audio to text;
- determine a login attempt based on the text;
- determine an identifier included in the text;
- compare the identifier to a plurality of identifiers stored in a database in response to determining the identifier, wherein each identifier of the plurality of identifiers corresponds to a respective user profile in the database;
- retrieve a user profile based on the user profile corresponding to the identifier stored in the database, the user profile including a voiceprint;
- in response to retrieving the user profile, compare the voice audio to the voiceprint of the user profile;
- determine the voice audio substantially matches the voiceprint;
- and
- allow the login attempt based on the voice audio substantially matching the voiceprint of the user profile.
One or more non-transitory computer readable storage media encoded with instructions that, when executed by a processor, cause the processor to perform operations including:
- obtaining voice audio from a user;
- converting the voice audio to text;
- determining a request to access a resource based on the text;
- extracting an identifier included in the text;
- comparing the identifier to a plurality of identifiers stored in a database in response to extracting the identifier, wherein each identifier of the plurality of identifiers corresponds to a respective user profile in the database;
- retrieving a user profile based on the user profile corresponding to the identifier stored in the database, the user profile including a voiceprint;
- in response to retrieving the user profile, comparing the voice audio to the voiceprint of the user profile;
- and
- providing access to the resource based on the voice audio substantially matching the voiceprint of the user profile.
Dependencies are inferred from the claim text.