Speech recognition system and speech file recording system

• US 7,979,278 B2
• Filed: 11/01/2002
• Issued: 07/12/2011
• Est. Priority Date: 02/28/2002
• Status: Expired due to Fees

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Matsuo, Naoshi
• Primary Examiner(s): Saint Cyr; Leonard
• Application Number:
  US10/285,482
• Publication Number:
  US 20030163308A1
• Time in Patent Office: 3,175 Days
• Field of Search: None
• US Class Current: 704/251
• CPC Class Codes: G10L 15/07 to the speaker

Abstract

A user term information extraction unit extracts term information of a user out of information that has been input by the user to an application for use other than speech recording beforehand, and a speech recognition dictionary management unit expands a vocabulary of a speech recognition dictionary according to the term information of the user. Next, the user inputs speech via a speech input unit, and a speech recognition unit executes speech recognition using the speech recognition dictionary. A representative term information selection unit extracts the term information of the user contained in the speech recognition result, and selects one or a plurality of pieces of representative term information from the term information of the user. A speech file recording unit records the speech data as a speech file, and renders a file name of the speech file according to the representative term information. Thus, a speech recognition and speech file recording system is provided that records input data as a speech file and automatically renders a file name thereto with which a content of the file can be grasped immediately.

Claims

1. A speech recognition system comprising:

a user term information extraction unit extracting automatically a user word that a user frequently uses or that is unique to the user from terms contained in data of an application for use other than speech recording when new data are stored upon use of the application;

a speech recognition list management unit having a first speech recognition list of recognition candidates, wherein said speech recognition list management unit adds a new recognition candidate to the first speech recognition list according to the user word;

a speech input unit, including at least one microphone, through which speech data of the user are input;

a speech recognition unit, comprising a server, executing speech recognition of the speech data using the first speech recognition list;

a speech file recording unit recording the input speech data as a speech file;

a speech file server storing the speech file on a network, the speech file recording unit transferring the speech file to the speech file server via the network so that the speech file is stored in the speech file server, the file name of the speech file and a network address of the speech file server being recorded in the speech file recording unit; and

a digest speech data generation unit that, prior to the transfer of the speech file to the speech file server by the speech file recording unit, extracts a part of the speech data recorded in the speech file and generates digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name of the speech file and the network address of the speech file server.

2. The speech recognition system according to claim 1, further comprising a speech recognition server on the network executing speech recognition using a second speech recognition list having a larger number of recognition candidates than that of the first speech recognition list,

wherein, when the speech recognition unit exhibits a lower recognition accuracy or reliability than a predetermined value, the speech recognition unit transmits the speech data to the speech recognition server via the network, and

the speech recognition server executes speech recognition of the speech data and transmits a result of the speech recognition to the speech input unit.

3. The speech recognition system according to claim 2, wherein the speech recognition list management unit adds the result of the speech recognition by the speech recognition server as a new recognition candidate to the first speech recognition list.

4. The speech recognition system according to claim 3, further comprising a selection unit with which the user selects whether or not an addition of a new recognition candidate to the first speech recognition list by the speech recognition list management unit is executed.

5. A speech file recording system comprising:

a user term information extraction unit extracting automatically a user word that a user frequently uses or that is unique to the user from terms contained in data of an application for use other than speech recording when new data are stored upon use of the application;

a speech recognition list management unit having a first speech recognition list of recognition candidates, wherein said speech recognition list management unit adds a new recognition candidate to the first speech recognition list according to the user word;

a speech input unit, including at least one microphone, through which speech data of the user are input;

a speech file recording unit recording the input speech data as a speech file;

a speech recognition unit, comprising a server, executing speech recognition of the speech data using the first speech recognition list;

a representative term information selection unit extracting a word used by the user contained in a result of the speech recognition, and selecting, as a representative word used by the user, one or a plurality of pieces of information out of the word used by the user;

a speech file server storing the speech file on a network, the speech file recording unit transferring the speech file to the speech file server via the network so that the speech file is stored in the speech file server, the file name of the speech file and a network address of the speech file server being recorded in the speech file recording unit; and

a digest speech data generation unit that, prior to the transfer of the speech file to the speech file server by the speech file recording unit, extracts a part of the speech data recorded in the speech file and generates digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name of the speech file and the network address of the speech file server, and

wherein the speech file recording unit renders a file name of the speech file according to the representative word used by the user.

6. The speech file recording system according to claim 5, further comprising a speech recognition server on the network executing speech recognition using a second speech recognition list having a larger number of recognition candidates than that of the first speech recognition list,

wherein, when the speech recognition unit exhibits a lower recognition accuracy or reliability than a predetermined value, the speech file recording unit transmits the speech file to the speech recognition server via the network,

the speech recognition server executes speech recognition of the speech data, and

the representative term information selection unit receives the result of the speech recognition from the speech recognition server via the network, extracts the word used by the user contained in the result of the speech recognition by the speech recognition server, and selects, as the representative word used by the user, one or a plurality of pieces of information out of the word used by the user.

7. The speech file recording system according to claim 6, further comprising an electronic mail generating unit for generating an electronic mail composed of a title and a text, assigning a content of a recognition result of the speech data of the speech file as the text of the electronic mail, assigning the word used by the user contained in the recognition result of the speech data of the speech file as the title of the electronic mail, and rendering a file name of the electronic mail according to the representative word used by the user.

8. The speech file recording system according to claim 7, further comprising an electronic mail generating unit for generating an electronic mail composed of a title and a text, and assigning the file name of the speech file and a network address of the speech file server as the text of the electronic mail, and the word used by the user contained in the recognition result of the speech data of the speech file as the title of the electronic mail.

9. The speech file recording system according to claim 5, further comprising an electronic mail generating unit generating an electronic mail composed of a title and a text, assigning a content of a recognition result of the speech data of the speech file as the text of the electronic mail, assigning the word used by the user contained in the recognition result of the speech data of the speech file as the title of the electronic mail, and rendering a file name of the electronic mail according to the representative word used by the user.

10. The speech file recording system according to claim 9, further comprising an electronic mail generating unit for generating an electronic mail composed of a title and a text, and assigning the file name of the speech file and a network address of the speech file server as the text of the electronic mail, and the word used by the user contained in the recognition result of the speech data of the speech file as the title of the electronic mail.

11. The speech recognition system according to claim 5, wherein the user term information extraction unit extracts the user word used by the user from information input by the user to the application for use other than speech recording.

12. The speech recognition system according to claim 11, wherein the application for use other than speech recording is a personal information manager application.

13. A speech file recording system comprising:

a user term information extraction unit for extracting term information of a user from information input by the user to a certain application;

a speech recognition dictionary management unit having a first speech recognition dictionary, wherein said speech recognition dictionary management unit adds vocabulary to the first speech recognition dictionary according to the term information of the user;

a speech input unit, including at least one microphone, through which speech data of the user are input;

a speech file recording unit recording the input speech data as a speech file;

a speech recognition unit, executing speech recognition of the speech data using the first speech recognition dictionary; and

a representative term information selection unit extracting term information of the user contained in a result of the speech recognition, and selecting, as a representative term information of the user, one or a plurality of pieces of information out of term information of the user,

wherein the speech file recording unit renders a file name of the speech file according to the representative word used by term information of the user;

a speech file server for storing the speech file on the network, wherein

the speech file recording unit transfers the speech file and a retrieval keyword to the speech file server via the network so that the speech file is stored in the speech file server, and

the file name and retrieval keyword of the speech file, and a network address of the speech file server are recorded in the speech file recording unit;

a digest speech data generation unit for, prior to the transfer of the speech file to the speech file server by the speech file recording unit, extracting a part of the speech data recorded in the speech file, and generating digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name and retrieval keyword of the speech file, and the network address of the speech file server.

14. A speech file recording system comprising:

a user term information extraction unit extracting a word used by a user from information input by the user to a certain application;

a speech recognition dictionary management unit having a first speech recognition dictionary, wherein said speech recognition dictionary management unit adds a word to be output as a recognition result to the first speech recognition dictionary according to the word used by the user;

a speech input unit, including at least one microphone, through which speech data of the user are input;

a speech file recording unit recording the input speech data as a speech file;

a speech recognition unit executing speech recognition of the speech data using the first speech recognition dictionary; and

a representative term information selection unit extracting a term information of the user contained in a result of the speech recognition, and selecting, as a representative term information of the user, one or a plurality of pieces of information out of the user,

wherein the speech file recording unit renders a file name of the speech file according to the representative term information of the user;

a speech file server for storing the speech file on the network, wherein

the speech file recording unit transfers the speech file and a retrieval keyword to the speech file server via the network so that the speech file is stored in the speech file server, and

the file name and retrieval keyword of the speech file, and a network address of the speech file server are recorded in the speech file recording unit;

a digest speech data generation unit for, prior to the transfer of the speech file to the speech file server by the speech file recording unit, extracting a part of the speech data recorded in the speech file, and generating digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name and retrieval keyword of the speech file, and the network address of the speech file server.

15. A speech recognition method comprising:

extracting automatically a user word that a user frequently uses or that is unique to the user from terms contained in data of an application for use other than speech recording when new data are stored upon use of the application;

holding a speech recognition list of recognition candidates, and adding a new recognition candidate to the speech recognition list, which is held for speech recognition, according to the extracted user word;

inputting speech data of the user;

executing speech recognition, using a server, of the speech data using the speech recognition list;

recording the input speech data as a speech file in a speech file recording unit;

storing the speech file by a speech file server on a network, the speech file recording unit transferring the speech file to the speech file server via the network so that the speech file is stored in the speech file server, the file name of the speech file and a network address of the speech file server being recorded in the speech file recording unit; and

extracting, prior to the transfer of the speech file to the speech file server by the speech file recording unit, a part of the speech data recorded in the speech file and generating digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name of the speech file and the network address of the speech file server.

16. The speech recognition method according to claim 15, wherein the user term information extraction unit extracts the user word used by the user from information input by the user to the application for use other than speech recording.

17. The speech recognition method according to claim 16, wherein the application for use other than speech recording is a personal information manager application.

18. A non-transitory computer-readable recording medium storing a speech recognizing operation program recognizing input speech data, the program causing a computer to execute:

extracting automatically a user word that a user frequently uses or that is unique to the user from terms contained in data of an application for use other than speech recording when new data are stored upon use of the application;

holding a speech recognition list of recognition candidates and adding a new recognition candidate to the speech recognition list according to the extracted user word;

inputting speech data of the user;

executing speech recognition of the speech data using the speech recognition list;

recording the input speech data as a speech file in a speech file recording unit;

storing the speech file by a speech file server on a network, the speech file recording unit transferring the speech file to the speech file server via the network so that the speech file is stored in the speech file server, the file name of the speech file and a network address of the speech file server being recorded in the speech file recording unit; and

extracting, prior to the transfer of the speech file to the speech file server by the speech file recording unit, a part of the speech data recorded in the speech file and generating digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name of the speech file and the network address of the speech file server.

19. The non-transitory computer-readable recording medium storing a speech recognizing operation program according to claim 18, further comprising the program codes for extracting the user word used by the user from information input by the user to the application for use other than speech recording.

20. The non-transitory computer-readable recording medium storing a speech recognizing operation program according to claim 19, wherein the application for use other than speech recording is a personal information manager application.

21. A speech file recording method comprising:

extracting automatically a user word that a user frequently uses or that is unique to the user from terms contained in data of an application for use other than speech recording when new data are stored upon use of the application;

holding a speech recognition list of recognition candidates, and adding a new recognition candidate to the speech recognition list according to the extracted user word;

inputting speech data of the user;

recording the input speech data as a speech file in speech file recording unit;

executing speech recognition, using a server, of the speech data using the speech recognition list; and

extracting a word used by the user contained in a result of the speech recognition, and selecting, as a representative word used by the user, one or a plurality of pieces of information out of the word used by the user;

storing the speech file by a speech file server on a network, the speech file recording unit transferring the speech file to the speech file server via the network so that the speech file is stored in the speech file server, the file name of the speech file and a network address of the speech file server being recorded in the speech file recording unit; and

extracting, prior to the transfer of the speech file to the speech file server by the speech file recording unit, a part of the speech data recorded in the speech file and generating digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name of the speech file and the network address of the speech file server, and

wherein a file name of the speech file is rendered according to the representative word used by the user.

22. The speech file recording method according to claim 21, wherein the extracted user word used by the user is carried out from information input by the user to the application for use other than speech recording.

23. The speech file recording method according to claim 22, wherein the application for use other than speech recording is a personal information manager application.

24. A non-transitory computer-readable recording medium storing a speech file recording program for recording input speech data as a speech file, the program causing a computer to execute:

extracting automatically a user word that a user frequently uses or that is unique to the user from terms contained in data of an application for use other than speech recording when new data are stored upon use of the application;

holding a speech recognition list of recognition candidates and adding a new recognition candidate to the speech recognition list according to the extracted user word;

inputting speech data of the user;

recording the input speech data as a speech file in speech file recording unit;

executing speech recognition of the speech data using the speech recognition list;

extracting a word used by the user contained in a result of the speech recognition, and selecting, as a representative word used by the user, one or a plurality of pieces of information out of the word used by the user; and

rendering a file name of the speech file according to the representative word used by the user;

storing the speech file by a speech file server on a network, the speech file recording unit transferring the speech file to the speech file server via the network so that the speech file is stored in the speech file server, the file name of the speech file and a network address of the speech file server being recorded in the speech file recording unit; and

extracting, prior to the transfer of the speech file to the speech file server by the speech file recording unit, a part of the speech data recorded in the speech file and generating digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name of the speech file and the network address of the speech file server, and.

25. The non-transitory computer-readable recording medium storing a speech file recording program for recording input speech data according to claim 24, further comprising the program codes for extracting the user word used by the user from information input by the user to the application for use other than speech recording.

26. The non-transitory computer-readable recording medium storing a speech file recording program for recording input speech data according to claim 25, wherein the application for use other than speech recording is a personal information manager application.

27. A speech recognition system comprising:

a user term information extraction unit for extracting automatiocally a user word that a user frequently uses or that is unique to the user from terms contained in data of an application for use other than speech recording when new data are stored upon use of the application;

a speech recognition list management unit having a first speech recognition list of recognition candidates, wherein said speech recognition list management unit adds a new recognition candidate to the first speech recognition list according to the user word;

a speech input unit, including at least one microphone, through which speech data of the user are input;

a speech recognition unit, comprising a server, for executing speech recognition of the speech data using the first speech recognition list, wherein the application for use other than speech recording is a personal information manager application;

a speech file recording unit recording the input speech data as a speech file;

a speech file server storing the speech file on a network, the speech file recording unit transferring the speech file to the speech file server via the network so that the speech file is stored in the speech file server, the file name of the speech file and a network address of the speech file server being recorded in the speech file recording unit; and

a digest speech data generation unit that, prior to the transfer of the speech file to the speech file server by the speech file recording unit, extracts a part of the speech data recorded in the speech file and generates digest speech data of the speech file,

wherein the speech file recording unit records the digest speech data in addition to the file name of the speech file and the network address of the speech file server.