Operating method for voice function and electronic device supporting the same

Patent metadata

Field	Value
Publication number	US-2016232893-A1
Application number	US-201615017957-A
Country	US
Kind code	A1
Filing date	Feb 8, 2016
Priority date	Feb 11, 2015
Publication date	Aug 11, 2016
Grant date	—

Abstract

An electronic device is provided. The electronic device includes a memory configured to store at least a portion of a plurality of pieces of speech information used for voice recognition, and a processor operatively connected to the memory, wherein the processor selects speaker speech information from at least a portion of the plurality of pieces of speech information based on mutual similarity, and generates voice recognition information to be registered as personalized voice information based on the speaker speech information.

First claim

What is claimed is:

1 . An electronic device comprising:
a memory configured to store a plurality of pieces of speech information used for voice recognition; and a processor including processing circuitry, the processor functionally connected to the memory, wherein the processor is configured to select speaker speech information from at least portions of the plurality of pieces of speech information based on mutual similarity, and to generate voice recognition information to be registered as personalized voice information based on the speaker speech information.

2 . The electronic device of claim 1 , wherein the processor is configured to output a message providing a notification that an operation of applying the voice recognition information to voice recognition is being performed.

3 . The electronic device of claim 1 , wherein the processor is configured to collect the pieces of speech information for at least one of a specified time or until a specified number of the pieces of speech information are collected.

4 . The electronic device of claim 1 , wherein the processor is configured to generate multi-condition training models to which at least one of at least a part of a noise or a specified sound effect is applied to the plurality of pieces of speech information, and to use the multi-condition training models to determine the voice recognition information to be registered as the personalized voice information.

5 . The electronic device of claim 1 , wherein the processor is configured to generate multi-condition training models to which at least one of a noise or a specified sound effect is applied to pieces of the speaker speech information, and to determine, based on the multi-condition training models, the voice recognition information to be registered as the personalized voice information.

6 . The electronic device of claim 1 , wherein the processor is configured to collect other speech information input by a specific speaker corresponding to the personalized voice information and to adapt a model of the personalized voice information.

7 . The electronic device of claim 6 , wherein the processor is configured to extract a phonemic sample corresponding to a registered phonemic model included in the personalized voice information from the speech information input from the specific speaker, and to adapt the registered phonemic model using the phonemic sample.

8 . The electronic device of claim 1 , wherein, when new speech information newly input is not a speech of a specific speaker corresponding to the personalized voice information, the processor is configured to output a message of unavailability of execution of a function requested by the new speech information, or to selectively execute the function based on a type of the function requested by the new speech information.

9 . The electronic device of claim 8 , wherein the processor is configured to not perform the function if the function is a specified secure function or to perform the function if the function is a non-secure function.

10 . The electronic device of claim 1 , wherein the processor is configured to output a setting screen for setting at least one function item to be executed based on a voice function in response to a speech information input from a speaker specified based on the personalized voice information.

11 . A voice function operating method comprising:
storing a plurality of pieces of speech information used for voice recognition; selecting speaker speech information from at least portions of the plurality of pieces of speech information based on mutual similarity; and generating voice recognition information to be registered as personalized voice information based on the speaker speech information selected.

12 . The method of claim 11 , further comprising at least one of:
collecting the speech information for a specified time; or collecting the speech information until a specified number of candidate data is collected.

13 . The method of claim 11 , further comprising outputting a message providing a notification that an operation of applying the voice recognition information to the voice recognition is being performed.

14 . The method of claim 11 , further comprising:
generating multi-condition training models to which at least one of at least a part of a noise or a specified sound effect is applied to the plurality of pieces of speech information; and applying the multi-condition training models to determine the voice recognition information to be registered as the personalized voice information.

15 . The method of claim 11 , wherein the generating comprises:
generating multi-condition training models to which at least a part of a noise or a specified sound effect is applied to pieces of the speaker speech information; and applying the multi-condition training models to determine the voice recognition information to be registered as the personalized voice information.

16 . The method of claim 11 , further comprising:
collecting other speech information input by a specific speaker corresponding to the personalized voice information; and adapting a model of the personalized voice information using the other speech information of the specific speaker.

17 . The method of claim 16 , wherein the adapting comprises extracting, from the speech information input from the specific speaker, a phonemic sample corresponding to a registered phonemic model included in the personalized voice information, and to use the phonemic sample in adapting the registered phonemic model.

18 . The method of claim 11 , further comprising:
outputting, if new speech information requesting a function is not a speech of the specific speaker corresponding to the personalized voice information, a message of unavailability of execution of the function based on the new speech information; and selectively executing the function based on a type of the function requested by the new speech information.

19 . The method of claim 18 , wherein the executing the function comprises:
execution of the function is not performed if the function is a specified secure function; and execution of the function is performed if the function is a non-secure function not specified.

20 . The method of claim 11 , further comprising at least one of:
outputting a setting screen for setting at least one function item to be executed based on a voice function in response to a speech information input by a speaker specified based on the personalized voice information; or outputting the generated voice recognition information.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Subhojit Chakladar

Classifications

• G10L15/07 Primary

  to the speaker · CPC title

• G10L2015/025

  Phonemes, fenemes or fenones being the recognition units · CPC title

• G10L17/02

  Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction · CPC title

• G10L15/063 Primary

  Training · CPC title

• G10L17/22

  Interactive procedures; Man-machine interfaces · CPC title