User adaptive speech recognition method and apparatus

• US 7,996,218 B2
• Filed: 02/16/2006
• Issued: 08/09/2011
• Est. Priority Date: 03/07/2005
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Kim, Jung-eun, Kim, Jeong-su
• Primary Examiner(s): Hudspeth; David R
• Assistant Examiner(s): Neway; Samuel G
• Application Number:
  US11/354,942
• Publication Number:
  US 20060200347A1
• Time in Patent Office: 2,000 Days
• Field of Search: 704231-257
• US Class Current: 704/236
• 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 2015/0631 Creating reference templates; Clustering

Abstract

A user adaptive speech recognition method and apparatus is disclosed that controls user confirmation of a recognition candidate using a new threshold value adapted to a user. The user adaptive speech recognition method includes calculating a confidence score of a recognition candidate according to the result of speech recognition, setting a new threshold value adapted to the user based on a result of user confirmation of the recognition candidate and the confidence score of the recognition candidate, and outputting a corresponding recognition candidate as a result of the speech recognition if the calculated confidence score is higher than the new threshold value. Thus, the need for user confirmation of the result of speech recognition is reduced and the probability of speech recognition success is increased.

Claims

1. A user adaptive speech recognition method comprising:

calculating confidence scores of respective recognition candidates according to results of speech recognition of respective user input speech;

setting a new threshold value adapted to one or more users based on results of respective user confirmations by the one or more users of the respective recognition candidates according to a previous threshold value, wherein the setting of the new threshold value includes selectively setting the new threshold value so as to have a value according to a select range, selected from plural ranges differently defining respective ranges, based upon a determination indicating that a first confidence score of a first recognition candidate, that should be rejected, is greater than a confidence score of a second recognition candidate, that should be accepted, or a confidence score of a third recognition candidate, that should be accepted, is less than a confidence score of a fourth recognition candidate, that should be rejected; and

outputting a corresponding recognition candidate as a result of the speech recognition when a calculated confidence score of the corresponding recognition candidate is determined to meet the new threshold value.

2. The method as claimed in claim 1, wherein the setting a new threshold value is performed when the number of speech inputs is greater than a threshold number of speech inputs.

3. The method as claimed in claim 1, wherein the new threshold value is set within the select range according to a predetermined threshold ratio, thereby controlling a permitted ratio of recognition candidates incorrectly accepted according to the new threshold value and recognition candidates incorrectly rejected according to the new threshold value.

4. The method as claimed in claim 1, wherein the differently defined ranges include at least a first range and a second range,

with the first range including values not greater than a lesser of a lowest confidence score of available recognition candidates incorrectly classified as being rejected and a lowest confidence score of recognition candidates correctly classified as being accepted and not less than a greater of a confidence score of available first recognition candidates incorrectly classified as being accepted and a greatest confidence score of recognition candidates correctly classified as being rejected, and

with the second range including values not greater than a greater of a greatest confidence score of the available recognition candidates incorrectly classified as being accepted and a confidence score of the recognition candidate correctly classified as being rejected, greater than a greatest confidence score of the available recognition candidates incorrectly classified as being rejected, and not less than a lesser of the lowest confidence score of the available recognition candidates incorrectly classified as being rejected and a confidence score of the recognition candidates correctly classified as being accepted, less than a lowest confidence score of the available recognition candidates incorrectly classified as being accepted.

5. A user adaptive speech recognition method comprising:

calculating a confidence score of a recognition candidate according to a result of speech recognition of speech input from a user;

setting a new threshold value adapted to the user based on a result of user confirmation of the recognition candidate and the confidence score of the recognition candidate; and

outputting a corresponding recognition candidate as the result of the speech recognition when the calculated confidence score is higher than the new threshold value, wherein the setting a new threshold value comprises;

classifying recognition candidates selected as correct answers by the user into a first group and classifying recognition candidates selected as incorrect answers by the user into a second group; and

calculating the new threshold value having a value in a range that is higher than the highest confidence score of the recognition candidates classified into the second group and that is lower than the lowest confidence score of the recognition candidates classified into the first group, when a first confidence score section being the confidence scores of the recognition candidates classified into the first group and a second confidence score section being the confidence scores of the recognition candidates classified into the second group do not overlap.

6. The method as claimed in claim 5, wherein the new threshold value is an average value of the lowest confidence score of the recognition candidates classified into the first group and the highest confidence score of the recognition candidates classified into the second group.

7. The method as claimed in claim 5, further comprising calculating the new threshold value so as to have a value in a range that is higher than the lowest confidence score of the recognition candidates classified into the first group and lower than the highest confidence score of the recognition candidates classified into the second group, when the first confidence score section and the second confidence score section overlap.

8. The method as claimed in claim 7, wherein the new threshold value is included in a specified confidence score range, and

wherein the confidence score range is set so that a ratio of a number of the recognition candidates having confidence scores below a lower threshold value of the confidence score range among the recognition candidates classified into the first group to a number of recognition candidates having the confidence scores above an upper threshold value of the confidence score range among the recognition candidates classified into the second group is the same as a threshold ratio.

9. The method as claimed in claim 8, wherein the new threshold value is an average value of the lowest confidence score of a recognition candidate that is higher than the upper threshold value of the confidence score range and the highest confidence score of a recognition candidate that is lower than the lower threshold value of the confidence score range.

10. A user adaptive speech recognition apparatus including at least one processing device, the apparatus comprising:

a confidence score calculation unit, using the at least one processing device, calculating confidence scores of respective recognition candidates according to results of speech recognition of respective user input speech;

a threshold value setup unit setting a new threshold value adapted to one or more users based on results of respective user confirmations by the one or more users of the respective recognition candidates according to a previous threshold value, wherein the setting of the new threshold value includes selectively setting the new threshold value so as to have a value according to a select range, selected from plural ranges differently defining respective ranges, based upon a determination indicating that a first confidence score of a first recognition candidate, that should be rejected, is greater than a confidence score of a second recognition candidate, that should be accepted, or a confidence score of a third recognition candidate, that should be accepted, is less than a confidence score of a fourth recognition candidate, that should be rejected; and

a control unit outputting a corresponding recognition candidate as a result of the speech recognition when a calculated confidence score of the corresponding recognition candidate is determined to meet the new threshold value.

11. The apparatus as claimed in claim 10, wherein the threshold value setup unit sets the new threshold value when the number of speech inputs is greater than a threshold number of speech inputs.

12. The apparatus as claimed in claim 10, wherein, wherein the new threshold value is set within the select range according to a predetermined threshold ratio, thereby controlling a permitted ratio of recognition candidates incorrectly accepted according to the new threshold value and recognition candidates incorrectly rejected according to the new threshold value.

13. A user adaptive speech recognition apparatus, including at least one processing device, the apparatus comprising:

a confidence score calculation unit, using the at least one processing device, calculating a confidence score of a recognition candidate according to the result of speech recognition of speech input by a user;

a threshold value setup unit setting a new threshold value adapted to the user based on a result of user confirmation of the recognition candidate and the confidence score of the recognition candidate; and

a control unit outputting a corresponding recognition candidate as the result of the speech recognition when the calculated confidence score is higher than the new threshold value, wherein the threshold value setup unit comprises;

a classification unit classifying recognition candidates selected as correct answers by the user into a first group and classifying recognition candidates selected as incorrect answers by the user into a second group; and

a threshold value calculation unit calculating the new threshold value having a value in a range that is higher than the highest confidence score of the recognition candidates classified into the second group and that is lower than the lowest confidence score of the recognition candidates classified into the first group, when a first confidence score section being the confidence scores of the recognition candidates classified into the first group and a second confidence score section being the confidence scores of the recognition candidates classified into the second group do not overlap.

14. The apparatus as claimed in claim 13, wherein the new threshold value is an average value of the lowest confidence score of the recognition candidates classified into the first group and the highest confidence score of the recognition candidates classified into the second group.

15. The apparatus as claimed in claim 13, wherein the threshold value calculation unit calculates the new threshold value so as to have a value in a range that is higher than the lowest confidence score of the recognition candidates classified into the first group and that is lower than the highest confidence score of the recognition candidates classified into the second group, when the first confidence score section and the second confidence score section overlap.

16. The apparatus as claimed in claim 15, wherein the new threshold value is included in a specified confidence score range, and

wherein the confidence score range is set so that a ratio of a number of the recognition candidates having confidence scores below a lower threshold value of the confidence score range among the recognition candidates classified into the first group to a number of recognition candidates having the confidence scores above an upper threshold value of a confidence score range among the recognition candidates classified into the second group is the same as a threshold ratio.

17. The apparatus as claimed in claim 16, wherein the new threshold value is an average value of the lowest confidence score of a recognition candidate that is higher than the upper threshold value of the confidence score range and the highest confidence score of a recognition candidate that is lower than the lower threshold value of the confidence score range.

18. A non-transitory computer-readable recording medium encoded with processing instructions for causing a processor to execute a user adaptive speech recognition method, the method comprising:

calculating confidence scores of respective recognition candidates according to results of speech recognition of respective user input speech;

setting a new threshold value adapted to one or more users based on results of respective user confirmations by the one or more users of the respective recognition candidates according to a previous threshold value, wherein the setting of the new threshold value includes selectively setting the new threshold value so as to have a value according to a select range, selected from plural ranges differently defining respective ranges, based upon a determination indicating that a first confidence score of a first recognition candidate, that should be rejected, is greater than a confidence score of a second recognition candidate, that should be accepted, or a confidence score of a third recognition candidate, that should be accepted, is less than a confidence score of a fourth recognition candidate, that should be rejected; and

outputting a corresponding recognition candidate as a result of the speech recognition when a calculated confidence score of the corresponding recognition candidate is determined to meet the new threshold value.

19. The medium as claimed in claim 18, wherein the new threshold value is set within the select range according to a predetermined threshold ratio, thereby controlling a permitted ratio of recognition candidates incorrectly accepted according to the new threshold value and recognition candidates incorrectly rejected according to the new threshold value.

20. A method of setting a new threshold value adapted to a user of a speech recognition apparatus, the method comprising:

determining whether a recognition candidate presented to the user is indicated by the user as correct;

classifying, using at least one processing device, the recognition candidate based on a result of the determining, the recognition candidate being classified in a first group when the recognition candidate is correct, the recognition candidate being classified in a second group when the recognition candidate is incorrect;

determining whether a range of probabilities of validity of the recognition candidates of the first group and a range of probabilities of validity of the recognition candidates of the second group overlap;

calculating the new threshold value in a range between the range of the probabilities of validity of the recognition candidates of the first group and the range of the probabilities of validity of the recognition candidates of the second group, when the range of the first group and the range of the second group do not overlap; and

calculating the new threshold value in a range where the range of the probabilities of validity of the recognition candidates of the first group and the range of the probabilities of validity of the recognition candidates of the second groups overlap, when the range of the first group and the range of the second group overlap.

21. A user adaptive speech recognition method comprising:

calculating confidence scores of respective recognition candidates according to results of speech recognition of respective user input speech;

setting a new threshold value adapted to one or more users based on results of respective user confirmations by the one or more users of the respective recognition candidates classified according to a previous threshold value, wherein the setting of the new threshold value includes selectively setting the new threshold value so as to have a value according to one of at least a first range and a second range, with the first range including values not greater than a lesser of a lowest confidence score of available recognition candidates incorrectly classified as being rejected and a lowest confidence score of recognition candidates correctly classified as being accepted and not less than a greater of a confidence score of available first recognition candidates incorrectly classified as being accepted and a greatest confidence score of recognition candidates correctly classified as being rejected, and with the second range including values not greater than a greater of a greatest confidence score of the available recognition candidates incorrectly classified as being accepted and a confidence score of the recognition candidate correctly classified as being rejected, greater than a greatest confidence score of the available recognition candidates incorrectly classified as being rejected, and not less than a lesser of the lowest confidence score of the available recognition candidates incorrectly classified as being rejected and a confidence score of the recognition candidates correctly classified as being accepted, less than a lowest confidence score of the available recognition candidates incorrectly classified as being accepted; and

outputting a corresponding recognition candidate as a result of speech recognition when a calculated confidence score of the corresponding recognition candidate is determined to meet the new threshold value.