Methods and apparatus for interpreting received speech data using speech recognition

Patent metadata

Field	Value
Publication number	US-9418679-B2
Application number	US-201414457809-A
Country	US
Kind code	B2
Filing date	Aug 12, 2014
Priority date	Aug 12, 2014
Publication date	Aug 16, 2016
Grant date	Aug 16, 2016

Abstract

A method for processing a received set of speech data, wherein the received set of speech data comprises an utterance, is provided. The method executes a process to generate a plurality of confidence scores, wherein each of the plurality of confidence scores is associated with one of a plurality of candidate utterances; determines a plurality of difference values, each of the plurality of difference values comprising a difference between two of the plurality of confidence scores; and compares the plurality of difference values to determine at least one disparity.

First claim

What is claimed is:

1. A method for processing a received set of speech data, wherein the received set of speech data comprises an utterance,
the method comprising:
receiving a set of speech data;
executing a process to calculate a plurality of probabilities based on the received set of speech data, wherein each of the plurality of probabilities indicates a statistical likelihood that the set of speech data comprises one of a plurality of candidate utterances;
determining a plurality of difference values, each of the plurality of difference values comprising a difference between two of the plurality of probabilities; and comparing the plurality of difference values to determine at least one disparity;
wherein the determining step further comprises:
calculating a first difference between a first probability and a second probability;
calculating a second difference between a third probability and a fourth probability; and
comparing the first difference to the second difference to determine the at least one disparity;
wherein the plurality of probabilities comprises the first probability, the second probability, the third probability, and the fourth probability; and
wherein the plurality of difference values comprises the first difference and the second difference.

2. The method of claim 1 , further comprising:
when the disparity is greater than a disparity threshold,
identifying a correct candidate utterance associated with one of the plurality of probabilities, wherein the correct candidate utterance comprises one of the plurality of candidate utterances; and
completing a task associated with the correct candidate utterance.

3. The method of claim 1 , further comprising:
when the disparity is less than a disparity threshold, present a user with a request to repeat the utterance.

4. The method of claim 1 , further comprising:
when the disparity comprises a value within a range of predetermined values, present a user with at least one of the plurality of candidate utterances for verification.

5. The method of claim 1 , wherein each of the plurality of probabilities comprises a calculated probability that a particular string of phonemes corresponds to the received set of speech data.

6. The method of claim 1 , further comprising:
identifying one or more probabilities above a probability threshold; and determining the plurality of difference values using the identified one or more probabilities.

7. A system for processing a received set of speech data, wherein the received set of speech data comprises an utterance,
the system comprising: a user interface module, configured to receive the set of speech data;
a probability module, configured to calculate a plurality of probabilities based on the received set of speech data, each of the calculated plurality of probabilities indicating a statistical likelihood that the set of speech data comprises one of a plurality of candidate utterances; and
a data analysis module, configured to:
calculate a plurality of difference values, each of the plurality of difference values comprising a difference between two of the plurality of probabilities; and compare the plurality of difference values to determine at least one disparity;
wherein the data analysis module is further configured to:
calculate a first difference between a first probability and a second probability;
calculate a second difference between a third probability and a fourth probability; and
compare the first difference to the second difference to determine the at least one disparity:
wherein the plurality of probabilities comprises the first probability, the second probability, the third probability, and the fourth probability; and
wherein the plurality of difference values comprises the first difference and the second difference.

8. The system of claim 7 , wherein, when the disparity is greater than a disparity threshold, the data analysis module is further configured to:
identify a correct candidate utterance associated with one of the plurality of probabilities,
wherein the correct candidate utterance comprises one of the plurality of candidate utterances; and
complete a task associated with the correct candidate utterance.

9. The system of claim 7 , wherein, when the disparity is less than a disparity threshold, the user interface module is further configured to present a user with a request to repeat the utterance.

10. The system of claim 7 , wherein, when the at least one disparity comprises a value within a range of predetermined values, the user interface module is further configured to present a user with at least one of a plurality of candidate utterances for verification.

11. The system of claim 7 , wherein the data analysis module is further configured to: identify one or more probabilities above a probability threshold; and determine the plurality of difference values using the identified one or more probabilities.

12. A non-transitory, computer-readable medium containing instructions thereon, which, when executed by a processor, perform a method comprising:
receiving a set of speech data, wherein an utterance comprises the set of speech data;
calculating a plurality of probabilities,
wherein each of the plurality of probabilities indicates a statistical likelihood that the set of speech data comprises one of a plurality of candidate utterances;
determining a plurality of difference values, each of the plurality of difference values comprising a difference between two of the plurality of probabilities; and comparing the plurality of difference values to determine at least one disparity;
wherein the method further comprises:
calculating a first difference between a first probability and a second probability;
calculating a second difference between a third probability and a fourth probability; and
comparing the first difference to the second difference to determine the at least one disparity:
wherein the plurality of probabilities comprises the first probability, the second probability, the third probability, and the fourth probability; and
wherein the plurality of difference values comprises the first difference and the second difference.

13. The non-transitory, computer-readable medium of claim 12 , wherein the method further comprises:
when the disparity is greater than a disparity threshold,
identifying a correct candidate utterance associated with one of the plurality of probabilities, wherein the candidate utterance comprises one of the plurality of candidate utterances; and
completing a task associated with the correct candidate utterance.

14. The non-transitory, computer-readable medium of claim 12 , wherein the method further comprises:
when the disparity is less than a disparity threshold, presenting a user with a request to repeat the utterance.

15. The non-transitory, computer-readable medium of claim 12 , wherein the method further comprises:
when the disparity comprises a value within a range of predetermined values, presenting a user with at least one of the plurality of candidate utterances for verification.

16. The non-transitory, computer-readable medium of claim 12 , wherein the method further comprises:
identifying one or more probabilities above a probability threshold; and
determining the plurality of difference values using the identified one or more probabilities.

17. The non-transitory, computer-readable medium of claim 12 , wherein each of the plurality of probabilities comprises a calculated probability that a particular string of codewords corresponds to the received set

Assignees

• Honeywell Int Inc

Inventors

• Nelson Erik T

Classifications

• G10L15/08 Primary

  Speech classification or search · CPC title

• G10L15/02

  Feature extraction for speech recognition; Selection of recognition unit · CPC title

• G10L15/22

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title

• G10L25/51 Primary

  for comparison or discrimination · CPC title

• G10L15/142

  Hidden Markov Models [HMMs] · CPC title