Method and apparatus for recognizing previously unrecognized speech by requesting a predicted-category-related domain-dictionary-linking word

• US 5,797,116 A
• Filed: 01/21/1997
• Issued: 08/18/1998
• Est. Priority Date: 06/16/1993
• Status: Expired due to Term

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Yamada, Masayuki, Komori, Yasuhiro, Ohora, Yasunori
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): Smits, Talivaldis Ivars
• Application Number:
  US08/785,840
• Time in Patent Office: 574 Days
• Field of Search: 704/10, 704/244, 704/251
• US Class Current: 704/10
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A voice communication method includes the steps of inputting speech into an apparatus, recognizing the input speech using a first dictionary, predicting the category of an unrecognized word included in the input speech based on the recognition of the input speech in the recognition step, outputting a question to be asked to an operator requesting the operator to input a word which is included in the first dictionary and which can specify a second dictionary for recognizing the unrecognized word, based on the predicted category, and re-recognizing the unrecognized word with the second dictionary specified in response to the word inputted by the operator. The invention also relates to an apparatus performing these functions and to a computer program product instructing a computer to perform these functions.

Claims

1. A voice communication method comprising the steps of:

inputting speech into an apparatus;

recognizing the input speech using a first dictionary;

predicting the category of an unrecognized word included in the input speech based on the recognition of the input speech in said recognition step;

outputting a question to be asked to an operator requesting the operator to input a word which is included in the first dictionary and which can specify a second dictionary for recognizing the unrecognized word, based on the predicted category; and

re-recognizing the unrecognized word with the second dictionary specified in response to the word inputted by the operator.

2. A method according to claim 1, wherein said determining step determines the question in expectation that the operator will utter a word in the predicted vocabulary.

3. A method according to claim 2, further comprising the steps of:

inputting additional speech into the apparatus comprising the word uttered by the operator in reply to the output question;

recognizing the additional speech; and

recognizing the unrecognized speech portion using a vocabulary predicted in response to a recognition made in said step of recognizing the additional speech.

4. A method according to claim 2, further comprising the steps of:

inputting additional speech into the apparatus comprising the word uttered by the operator in reply to the output question;

recognizing the additional speech; and

recognizing the input speech using recognition results for the input speech and the additional speech.

5. A method according to claim 1, wherein said determining step determines the question in expectation that the operator will utter a word forming speech different from the unrecognized speech portion.

6. A method according to claim 1, wherein said predicting step predicts the vocabulary of the unrecognized speech portion from a sentence structure of a recognized speech portion.

7. A method according to claim 1, further comprising the steps of:

deriving a plurality of sentence structures which can be formed by the input speech from a recognition result for the input speech; and

selecting one of the plurality of derived sentence structures which is acoustically closest to the input speech.

8. A method according to claim 1, wherein said inputting step inputs the speech with a microphone.

9. A method according to claim 1, wherein said recognizing step recognizes the input speech using a memory storing a dictionary for speech recognition.

10. A method according to claim 1, further comprising the step of storing recognition results for the recognized speech portion and the unrecognized speech portion.

11. A speech communication apparatus comprising:

input means for inputting speech into said apparatus;

recognizing means for recognizing the speech input by said input means using a first dictionary;

predicting means for predicting the category of an unrecognized word included in the input speech based on the recognition of the input speech by said recognizing means;

outputting means for outputting a question to be asked of an operator requesting the operator to input a word which is included in the first dictionary and which can specify a second dictionary for recognizing the unrecognized word, based on the predicted category; and

re-recognizing means for re-recognizing the unrecognized word with the second dictionary specified in response to the word inputted by the operator.

12. An apparatus according to claim 11, wherein said determining means determines the question in expectation that the operator will utter a word in the predicted vocabulary.

13. An apparatus according to claim 12, wherein said input means inputs additional speech comprising the word uttered by the operator in reply to the output question, wherein said recognizing means recognizes the input additional speech, and wherein said recognizing means recognizes the unrecognized speech portion using a vocabulary predicted in response to a recognition of the additional speech.

14. An apparatus according to claim 12, wherein said input means inputs additional speech comprising the word uttered by the operator in reply to the output question, wherein said recognizing means recognizes the additional speech, and wherein said recognizing means recognizes the input speech using recognition results for the input speech and the additional speech.

15. An apparatus according to claim 11, wherein said determining means determines the question in expectation that the operator will utter a word forming a speech different from the unrecognized speech portion.

16. An apparatus according to claim 11, wherein said predicting means predicts the vocabulary of the unrecognized speech portion from a sentence structure of a recognized speech portion.

17. An apparatus according to claim 11, further comprising:

deriving means for deriving a plurality of sentence structures which can be formed by the input speech from a recognition result for the input speech; and

selecting means for selecting one of the plurality of derived sentence structures which is acoustically closest to the input speech.

18. An apparatus according to claim 11, wherein said input means comprises a microphone.

19. An apparatus according to claim 11, further comprising means for storing a dictionary for speech recognition.

20. An apparatus according to claim 11 further comprising means for storing recognition results for the recognized speech portion and the unrecognized speech portion.

21. A computer program product comprising a computer usable medium having computer readable program code means embodied in said medium for enabling speech communication by a computer, said computer program product comprising:

first computer readable program code means for causing said computer to input speech thereinto;

second computer readable program code means for causing said computer to recognize the input speech using a first dictionary;

third computer readable program code means for causing said computer to predict the category of an unrecognized word included in the input speech based on the recognition of the input speech caused by said second computer readable program code means;

fourth computer readable program code means for causing said computer to output a question to be asked of an operator requesting the operator to input a word which is included in the first dictionary and which can specify a second dictionary for recognizing the unrecognized word, based on the predicted category; and

fifth computer readable program code means for causing said computer to re-recognize the unrecognized word with the second dictionary specified in response to the word inputted by the operator.

22. A computer program product to claim 21, wherein said fourth computer readable program code means causes said computer to determine the question in expectation that the operator will utter the word.

23. A computer program product according to claim 21, wherein said first computer readable program code means causes said computer to input additional speech comprising the word uttered by the operator in reply to the output question, wherein said second computer readable program code means causes said computer to recognize the input additional speech, and wherein said second computer readable program code means causes said computer to recognize the unrecognized word using a vocabulary predicted in response to recognition of the additional speech.

24. A computer program product according to claim 21, wherein said first computer readable program code means causes said computer to input additional speech comprising the word uttered by the operator in reply to the output question, wherein said second computer readable program code means causes said computer to recognize the additional speech, and wherein said second computer readable program code means causes said computer to recognize the input speech using recognition results for the input speech and the additional speech.

25. A computer program product according to claim 21, wherein said fourth computer readable program code means causes said computer to determine the question in expectation that the operator will utter a word forming a speech different from the unrecognized word.

26. A computer program product according to claim 21, wherein said third computer readable program code means causes said computer to predict the unrecognized word from a sentence structure of the recognized input speech.

27. A computer program product according to claim 21, further comprising:

sixth computer readable program code means for causing said computer to derive a plurality of sentence structures which can be formed by the input speech from a recognition result for the input speech; and

seventh computer readable program code means for causing said computer to select one of the plurality of derived sentence structures which is acoustically closest to the input speech.

28. A computer program product according to claim 21, wherein said first computer readable program code means causes said computer to input speech using a microphone.

29. A computer program product according to claim 21, further comprising means for storing the first and second dictionaries.

30. A computer program product according to claim 21, further comprising sixth computer readable program code means for causing said computer to store recognition results for the recognized speech and the unrecognized word.