Method Of Voice Recognition And Electronic Apparatus

• US 20140163984A1
• Filed: 12/10/2013
• Published: 06/12/2014
• Est. Priority Date: 12/10/2012
• Status: Active Grant

• Current Assignee: Lenovo (Beijing) Limited (Lenovo Group Ltd.), Lenovo Company Limited (Lenovo Group Ltd.)
• Original Assignee: Lenovo (Beijing) Limited (Lenovo Group Ltd.), Lenovo Company Limited (Lenovo Group Ltd.)
• Inventors: Dai, Haisheng, Wang, Qianying, Wang, Hao
• Granted Patent: US 10,068,570 B2
• Time in Patent Office: Days
• US Class Current: 704/246
• CPC Class Codes: G10L 15/04 Segmentation; Word boundary detection
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 25/87 Detection of discrete points within a voice signal

Abstract

A method of voice recognition and an electronic apparatus are described with the method of voice recognition being applied in an electronic apparatus. The method includes taking i=1 and detecting corresponding i-th voice sub-information at a moment Ti when the electronic apparatus detects that a user starts to talk at a moment T0, wherein the i-th voice sub-information is corresponding voice information from the moment T0 to the moment Ti, the i-th voice sub-information is partial voice information of voice information with integral semantic corresponding to a moment Tj after the moment T0 to the moment Ti, and i is an integer greater than or equal to 1; and analyzing the i-th voice sub-information to obtain M results of analysis, M being an integer greater than or equal to 1.

Claims

1. A method of voice recognition applied in an electronic apparatus comprising:

taking i=1 and detecting corresponding i-th voice sub-information at a moment Ti when the electronic apparatus detects that a user starts to talk at a moment T0, wherein the i-th voice sub-information is corresponding voice information from the moment T0 to the moment Ti, the i-th voice sub-information is partial voice information of voice information with integral semantic corresponding to a moment Tj after the moment T0 to the moment Ti, and i is an integer greater than or equal to 1; and

analyzing the i-th voice sub-information to obtain M results of analysis, M being an integer greater than or equal to 1.

2. The method according to claim 1, wherein analyzing the i-th voice sub-information to obtain M results of analysis comprises analyzing the i-th voice sub-information to obtain M keyword text information, the M keyword text information being exactly the M results of analysis.

3. The method according to claim 1, wherein analyzing the i-th voice sub-information to obtain M results of analysis comprises:

analyzing the i-th voice sub-information to obtain N keyword text information, wherein N is an integer greater than or equal to 1; and

obtaining M results of matching whose match degree with any of the N keyword text information is larger than a first preset threshold based on the N keyword text information, the M results of matching being exactly the M results of analysis.

4. The method according to claim 1, wherein after analyzing the i-th voice sub-information to obtain M results of analysis, the method further comprises displaying M1 results of analysis in the M results of analysis on a display interface to receive a selection instruction of the user, M1 being an integer greater than or equal to 1.

5. The method according to claim 4, wherein after analyzing the i-th voice sub-information to obtain M results of analysis, the method further comprises:

deciding whether there is a first control instruction for stopping detecting whether the user is talking;

stopping detecting whether the user is talking when there is the first control instruction; and

taking i=i+1 and executing a step of detecting the corresponding i-th voice sub-information at a moment Ti when there is no first control instruction.

6. The method according to claim 5, wherein deciding whether there is a first control instruction for stopping detecting whether the user is talking further comprises deciding whether there is a selection operation of selecting a first result of analysis in the M1 results of analysis to obtain a decision result, wherein when the result of decision indicates that there is the selection operation, a first control instruction is generated.

7. The method according to claim 6, wherein the selection operation comprises an operation carried out through a first input mode, the first input mode being an input mode different from the voice input mode.

8. The method according to claim 4, wherein before displaying the M1 results of analysis in the M results of analysis on the display interface, the method further comprises:

scoring the M results of analysis according to the match degree to obtain M first point value;

or

scoring the M results of analysis according to the historical selection information to obtain M second point values;

or

scoring the M results of analysis according to the match degree and the historical selection information to obtain M third point values.

9. The method according to claim 8, wherein before displaying the M1 results of analysis in the M results of analysis on the display interface, the method further comprises:

acquiring M1 results of analysis having the highest point value among the first point values, the second point values or the third point values from the M results of analysis; and

displaying the M1 results of analysis on the display interface.

10. The method according to claim 1, wherein analyzing the i-th voice sub-information to obtain M results of analysis comprises carrying out voiceprint analysis to the i-th voice sub-information to obtain the M results of analysis.

11. An electronic apparatus comprising:

a detecting chip for taking i=1 and detecting corresponding i-th voice sub-information at a moment Ti when the electronic apparatus detects that a user starts to talk at a moment T0, wherein the i-th voice sub-information is corresponding voice information from the moment T0 to the moment Ti, the i-th voice sub-information is partial voice information of voice information with integral semantic corresponding to a moment Tj after the moment T0 to the moment Ti, and i is an integer greater than or equal to 1; and

an analyzing chip for analyzing the i-th voice sub-information to obtain M results of analysis, M being an integer greater than or equal to 1.

12. The electronic apparatus according to claim 11, wherein the analyzing chip comprises analyzing the i-th voice sub-information to obtain M keyword text information, the M keyword text information being exactly the M results of analysis.

13. The electronic apparatus according to claim 11, wherein the analyzing chip comprises:

an analyzing sub-chip for analyzing the i-th voice sub-information to obtain N keyword text information, N being an integer greater than or equal to 1; and

a matching sub-chip for obtaining M results of matching whose match degree with any of the N keyword text information is larger than a first preset threshold based on the N keyword text information, the M results of matching being exactly the M results of analysis.

14. The electronic apparatus according to claim 11, wherein the electronic apparatus further comprises a displaying chip for displaying the M1 results of analysis in the M results of analysis on a display interface to receive a selection instruction of the user after analyzing the i-th voice sub-information to obtain the M results of analysis, M1 being an integer greater than or equal to 1.

15. The electronic apparatus according to claim 14, wherein the electronic apparatus further comprises:

a deciding chip for deciding whether there is a first control instruction for stopping detecting whether the user is talking after analyzing the i-th voice sub-information to obtain the M results of analysis;

a stopping chip for stopping detecting whether the user is talking when there is the first control instruction; and

a returning chip for taking i=i+1 and executing a step of detecting the corresponding i-th voice sub-information at a moment Ti when there is no first control instruction.

16. The electronic apparatus according to claim 15, wherein the deciding chip comprises:

a deciding sub-chip for deciding whether there is a selection operation of selecting a first result of analysis in the M1 results of analysis to obtain a result of decision; and

a generating sub-chip for generating the first control instruction when the result of decision indicates that there is the selection operation.

17. The electronic apparatus according to claim 16, wherein the selection operation includes an operation carried out through a first input mode, the first input mode being an input mode different from the voice input mode.

18. The electronic apparatus according to claim 14, wherein the electronic apparatus further comprises:

a scoring chip for scoring the M results of analysis according to match degree to obtain M first point values before displaying the M1 results of analysis in the M results of analysis on the display interface;

or

scoring the M results of analysis according to historical selection information to obtain M second point values;

or

scoring the M results of analysis according to the match degree and historical selection information to obtain M third point values.

19. The electronic apparatus according to claim 18, wherein the display chip comprises:

an acquiring sub-chip for acquiring M1 results of analysis having the highest point value among the first point values, the second point values or the third point values from the M results of analysis; and

a displaying sub-chip for displaying the M1 results of analysis on the display interface.

20. The electronic apparatus according to claim 11, wherein the analyzing chip comprises carrying out voiceprint analysis to the i-th voice sub-information to obtain the M results of analysis.