Keyword model generation for detecting user-defined keyword
First Claim
1. A method, performed in an electronic device, for generating a keyword model associated with a user-defined keyword, the method comprising:
- receiving a speech input from a user, at a sensor of the electronic device, in response to a prompt;
extracting a speech segment from the speech input;
determining a sequence of subwords representative of the speech segment of the speech input;
obtaining at least one subword model based on the sequence of subwords, wherein the at least one subword model includes a plurality of subword units and model parameters associated with each of the plurality of subword units;
generating the keyword model associated with the user-defined keyword by;
modifying the plurality of subword units of the at least one subword model based on the sequence of subwords representative of the speech segment of the speech input; and
adjusting the model parameters of the at least one subword model based on acoustic features extracted from the speech segment of the speech input; and
providing the keyword model associated with the user-defined keyword to a voice activation unit.
1 Assignment
0 Petitions
Accused Products
Abstract
According to an aspect of the present disclosure, a method for generating a keyword model of a user-defined keyword in an electronic device is disclosed. The method includes receiving at least one input indicative of the user-defined keyword, determining a sequence of subwords from the at least one input, generating the keyword model associated with the user-defined keyword based on the sequence of subwords and a subword model of the subwords, wherein the subword model is configured to model a plurality of acoustic features of the subwords based on a speech database, and providing the keyword model associated with the user-defined keyword to a voice activation unit configured with a keyword model associated with a predetermined keyword.
-
Citations
30 Claims
-
1. A method, performed in an electronic device, for generating a keyword model associated with a user-defined keyword, the method comprising:
-
receiving a speech input from a user, at a sensor of the electronic device, in response to a prompt; extracting a speech segment from the speech input; determining a sequence of subwords representative of the speech segment of the speech input; obtaining at least one subword model based on the sequence of subwords, wherein the at least one subword model includes a plurality of subword units and model parameters associated with each of the plurality of subword units; generating the keyword model associated with the user-defined keyword by; modifying the plurality of subword units of the at least one subword model based on the sequence of subwords representative of the speech segment of the speech input; and adjusting the model parameters of the at least one subword model based on acoustic features extracted from the speech segment of the speech input; and providing the keyword model associated with the user-defined keyword to a voice activation unit. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. An electronic device for generating a keyword model associated with a user-defined keyword, comprising:
-
an input unit configured to receive a speech input from a user, at a sensor of the electronic device, in response to a prompt; a subword recognition unit configured to extract a speech segment from the speech input; and a user-defined keyword model generation unit configured to; determine a sequence of subwords representative of the speech segment of the speech input; obtain at least one subword model of a plurality of subword models based on the sequence of subwords, wherein the at least one subword model includes a plurality of subword units and model parameters associated with each of the plurality of subword units; generating the keyword model associated with the user-defined keyword by; modify the plurality of subword units of the at least one subword model based on the sequence of subwords representative of the speech segment of the speech input; and adjust the model parameters of the at least one subword model based on acoustic features extracted from the speech segment of the speech input; and provide the keyword model associated with the user-defined keyword to a voice activation unit. - View Dependent Claims (19, 20, 21, 22, 23, 24, 25, 26)
-
-
27. A non-transitory computer-readable storage medium storing instructions for generating a keyword model associated with a user-defined keyword in an electronic device, the instructions causing a processor to perform operations of:
-
receiving a speech input from a user, at a sensor of the electronic device, in response to a prompt; extracting a speech segment from the speech input; determining a sequence of subwords representative of the speech segment of the speech input; obtaining at least one subword model based on the sequence of subwords, wherein the at least one subword model includes a plurality of subword units and model parameters associated with each of the plurality of subword units; generating the keyword model associated with the user-defined keyword by; modifying the plurality of subword units of the at least one subword model based on the sequence of subwords representative of the speech segment of the speech input; and adjusting the model parameters of the at least one subword model based on acoustic features extracted from the speech segment of the speech input; and providing the keyword model associated with the user-defined keyword to a voice activation unit.
-
-
28. An electronic device for generating a keyword model associated with a user-defined keyword, comprising:
-
means for receiving a speech input from a user, at a sensor of the electronic device, in response to a prompt; means for extracting a speech segment from the speech input; means for determining a sequence of subwords representative of the speech segment of the speech input; means for obtaining at least one subword model of a plurality of subword models based on the sequence of subwords, wherein the at least one subword model includes a plurality of subword units and model parameters associated with each of the plurality of subword units; means for generating the keyword model associated with the user-defined keyword by; modifying the plurality of subword units of the at least one subword model based on the sequence of subwords representative of the speech segment of the speech input; and adjusting the model parameters of the at least one subword model based on acoustic features extracted from the speech segment of the speech input; and means for providing the keyword model associated with the user-defined keyword to a voice activation unit. - View Dependent Claims (29, 30)
-
Specification