JP2007233075A - Minutes preparation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a minutes preparation device that makes it possible to easily recognize whose utterance each of utterance contents is. <P>SOLUTION: A speech recognition unit 4 recognizes a speech of a conferee based upon sound recording contents picked up by a microphone 3, and a CPU 2 prepares minutes data from the recognized speech. A speaker specifying unit 5 specifies speakers of respective utterance contents through voiceprint authentication. Minutes data wherein utterance contents of the prepared minutes data and the specified speakers are made to correspond to each other are displayed at a display unit 8. The utterance contents of the speakers are displayed in different colors by the speakers. Speaker information is added in front of the respective utterance contents. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a minutes creation apparatus that automatically creates minutes of a meeting using a voice recognition function.

  Record the voice of the speaker during the meeting, use the voice recognition function to recognize the content of the speech, convert the recognized content to text information, display it on a display device, or print it on paper Proposals for creating minutes have been proposed (Patent Documents 1, 2, etc.).

Patent Document 1 discloses a minutes creation device that outputs the contents of a statement by changing the size of characters and symbols according to the voice level of the speaker. Patent Document 2 discloses a conference support system that discriminates a speaker's emotion when speaking and generates a minutes that records the speaker's emotion in addition to the content of the statement.
JP 2002-222186 A JP 2005-277462 A

  In the conventional minutes creation apparatus including Patent Documents 1 and 2, the content of the remarks is recorded, but it is not specified who is the remark among the attendees for each remark content. There is a problem in that it is inconvenient for a person who has not attended to see who has spoken even when looking at the minutes.

  The present invention has been made in view of such circumstances, and provides a minutes creation device that allows a third party to recognize who made a statement by identifying a speaker and creating a minutes. For the purpose.

  Another object of the present invention is to provide a minutes creation device that can easily identify who is speaking by changing the display form according to the speaker.

  Still another object of the present invention is to provide a minutes creation device that can easily confirm important matters of a meeting by emphasizing the contents of statements including keywords or the contents of statements before and after that. It is in.

  The minutes creation device according to the present invention is a device for creating minutes of a meeting, a creation means for creating minutes data using speech recognition, a specifying means for specifying a speaker, and the creation means It is characterized by comprising means for associating and storing the message content in the minutes data and the speaker specified by the specifying means.

  In the minutes creation apparatus of the present invention, the minutes data created by using speech recognition from the recording result of the meeting and the speaker specified for each statement content are stored in association with each other. Therefore, in the minutes that have been created, the speakers are identified for the content of each statement, so not only participants in the conference but also third parties who did not attend the conference can see who is speaking. .

  The minutes creation apparatus according to the present invention is characterized by comprising output means for outputting the minutes data so that the contents of the statements are displayed in different forms depending on the speakers.

  In the minutes creation apparatus of the present invention, the content of a statement is displayed in different forms (for example, different colors, different character sizes, etc.) depending on the speaker. Therefore, it is easy to understand the content of a statement made by a specific speaker.

  The minutes generating device according to the present invention comprises means for storing speaker information relating to a speaker, and the output means is configured to output the minutes data with the speaker information added thereto. To do.

  In the minutes preparing apparatus of the present invention, the speaker data (name, title, nickname, photo, etc.) for identifying the speaker is added and the minutes data is output. Therefore, it is possible to easily recognize who is speaking.

  The minutes generating apparatus according to the present invention comprises means for registering attendees of a meeting based on the speaker specified by the specifying means.

  In the minutes preparing apparatus of the present invention, the attendees of the conference are registered based on the result of specifying the speaker. Therefore, it is possible to automatically register who attended the meeting.

  The minutes generating apparatus according to the present invention includes means for storing a predetermined keyword and detecting means for detecting the presence of the keyword in the content of a statement, and the output means includes a means for detecting the presence of the keyword. When a message is detected, the content of the utterance including the keyword and / or the content of the utterance before and after the content of the keyword is emphasized and the minutes data is output.

  In the minutes creation apparatus of the present invention, the content of a statement including a predetermined keyword (such as “decision items” and “conclusion”), the content of a statement that does not include a keyword, the content of a statement immediately before the keyword, and / or the keyword The content of the comment immediately after the keyword is emphasized and output. Therefore, it becomes easy to understand important matters of the conference.

  In the present invention, the minutes data is created by using voice recognition, the speaker of each statement content is specified, and the statement content and the specified speaker in the generated minutes data are stored in association with each other. In the minutes, the speakers are identified for the content of each statement, so that not only attendees of the conference but also third parties who did not attend the conference can know the speakers of each statement. Can do. Further, it is possible to easily search for the contents of remarks for each attendee.

  In the present invention, since the minutes data is output so that the content of the speech is displayed in a different form depending on the speaker, the content of the speech by a specific speaker can be easily identified.

  Further, in the present invention, since the speaker information relating to the speaker is added and the minutes data is output, it is possible to easily recognize who was speaking.

  In the present invention, since the attendees of the conference are registered based on the specified speaker, the attendees of the conference can be automatically registered.

  Further, in the present invention, since the content of the utterance including the keyword and / or the content of the utterance before and after the content of the utterance including the keyword is emphasized and the minutes data is output, the important matter in the meeting is simply displayed. Can be found.

  Hereinafter, the present invention will be described in detail with reference to the drawings showing embodiments thereof. In addition, this invention is not limited to the following embodiment.

  FIG. 1 is a diagram showing a configuration of an example of a minutes creation apparatus according to the present invention. 1 includes a CPU 2, a microphone 3, a voice recognition unit 4, a speaker identification unit 5, a voice print data storage unit 6, an input operation unit 7, a display unit 8, and printing. A unit 9, a ROM 10, and a RAM 11 are provided.

  The microphone 3 collects and records the sound at the conference. The voice recognition unit 4 recognizes the voice of the conference attendee based on the sound collection result by the microphone 3. The CPU 2 creates minutes data from the voice recognized by the voice recognition unit 4. On the other hand, the speaker specifying unit 5 compares and contrasts the voiceprint data of the voice recognized by the voice recognition unit 4 with the voiceprint data of the conference attendees stored in advance in the voiceprint data storage unit 6, and the contents of each statement Identify the speakers of The CPU 2 associates the content of the utterance in the created minutes data with the speaker identified by the speaker identification unit 5 and stores the associated minutes data in the RAM 11.

  The input operation unit 7 accepts user operation input, and accepts user edit processing for minutes such as deletion, correction, and addition of the contents of the created minutes. The display unit 8 is, for example, a liquid crystal display device, and displays minutes data in association with the content of the speech and the speaker. The printing unit 9 is, for example, an electrophotographic printer device, and prints minutes data on a sheet by associating the content of a statement with the speaker.

  The ROM 10 stores various software control programs necessary for the operation of the minutes creation apparatus 1. The RAM 11 stores temporary data generated when the software is executed. The RAM 11 stores the minutes data in which the content of the speech and the speaker are associated with each other as described above, speaker information (name, title, nickname, photo, etc.) for each speaker, and a predetermined keyword ( "Decision items", "conclusion", etc.)

  In the configuration as described above, the CPU 1 functions as a creation unit that creates minutes data using voice recognition. The CPU 1 also functions as a detection unit that detects the presence of a keyword in the content of a statement and a unit that registers attendees of the conference. The speaker specifying unit 5 functions as specifying means for specifying a speaker. The display unit 8 and the printing unit 9 function as output means for outputting the minutes data. The RAM 11 functions as means for storing the minutes data by associating the content of the comment and the speaker, a means for storing the speaker information, and a means for storing a predetermined keyword.

  Next, operation | movement of the minutes production apparatus 1 of this invention is demonstrated. Hereinafter, a case where the minutes data is displayed in color on the display unit 8 will be described as an example.

  Based on the recorded content collected by the microphone 3, the voice recognition unit 4 recognizes the voice of the attendee of the conference, and the minutes data is created from the recognized voice by the CPU 2. On the other hand, the recognized voice is output to the speaker specifying unit 5, and the speaker specifying unit 5 recognizes the voice print data of the recognized conference attendee and the conference attendee stored in the voice print data storage unit 6. By comparing and contrasting the voiceprint data, the speaker of each content is specified.

  The CPU 2 associates the content of the statement in the created minutes data with the identified speaker, and the associated minutes data is temporarily stored in the RAM 11. Then, the minutes data stored in the RAM 11 is displayed on the display unit 8. FIG. 2 is a diagram illustrating an example of the display of minutes data on the display unit 8.

  In FIG. 2, the minutes in which the contents of each statement are associated with the speaker are displayed. At this time, the content of the speech of the speaker is displayed in different colors for each speaker. For example, the speech content of the attendee A is displayed in red, the speech content of the attendee B is displayed in blue, and the speech content of the attendee C is displayed in green. Therefore, the content of a specific participant's remarks can be easily found.

  Further, a name (A, B, C) and a title (section manager), which are speaker information, are added in front of each comment content. Therefore, it is possible to recognize at a glance who is speaking. As the speaker information, in addition to the name and position, a nickname of the speaker, a photograph (image), or the like may be used.

  In the present invention, since the speaker can be specified, the attendee of the conference can be known from the specified information. Therefore, the names of the attendees who are found are registered in the minutes. Therefore, registration of attendees of the conference can be performed automatically.

  In addition, the content of a message including the keyword “decision item” set in advance is displayed underlined. Specifically, in the display example of FIG. 2, the term “determined matter” is included in the content of the remarks of the section manager Mr. C, and the content of the remarks is underlined. Therefore, since the items determined in the conference are highlighted and displayed, the important matters of the conference can be easily understood. Although it is emphasized by underlining it, it can be emphasized by displaying it with a larger character than other utterance contents, emphasizing with a special symbol, displaying with a special font, etc. good. In addition, since the content of the speech before and after the content of the speech including the keyword is considered important, the content of the speech before and after the content may be highlighted. In other words, the content of the message immediately before the keyword that does not include the keyword and / or the content of the message immediately after the keyword that does not include the keyword may be highlighted. Such keyword setting may be performed before the conference or may be performed after the conference ends.

  FIG. 3 is a diagram showing the configuration of another example of the minutes creation apparatus according to the present invention. In this example, the method for identifying a speaker is different from the above-described example (FIG. 1). The minutes creation device 21 includes a plurality of microphones 3a, 3b,. In FIG. 3, the same or similar parts as those in FIG.

  In this example, one microphone is installed in front of the meeting attendee. The voice recognition unit 4 recognizes the voice of the attendee based on the sound collection results of the plurality of microphones 3a, 3b, ..., 3n. The recognized voice for each microphone is output to the speaker specifying unit 5.

  The speaker specifying unit 5 specifies a speaker of each utterance content from the sound of each microphone. Comparing the input level and phase (time delay) of a certain voice (speech content) between microphones, selecting a microphone with a high input level and a small phase (time delay), and attendance corresponding to the selected microphone Identifies the person as the speaker.

  Other operations (correspondence between the contents of a statement and a speaker, display of associated minutes data, and the like) are the same as those in the above-described example, and a description thereof will be omitted.

  Although the case where the minutes data is displayed and output on the display unit 8 has been described, it is a matter of course that the same can be done when the minutes data is printed on paper by the printing unit 9 and output. Further, in the above-described embodiment, the content of speech according to each speaker is identified by color coding. However, as a different output mode according to the speaker, in addition to the color, the size of the character, the font, etc. Available. In addition, a function for changing the character form according to the sound level (voice volume) and a function for changing the output form of the content of the speech according to the determined emotion of the speaker are further added to the minutes creating apparatus 1. You may do it.

  By the way, a plurality of voices (speech) may be mixed. The correspondence in this case will be described. When a plurality of voices are input at the same time, if a plurality of utterance contents can be separated from the mixed voice and a speaker can be specified, a normal output as described above is added with a mark indicating that the utterances are simultaneously uttered. (Output the minutes data that correlates the content of the utterance and the speaker), and when the content of multiple utterances cannot be separated, a mark indicating that they are uttered at the same time is attached and “content unreadable” is output ( Display or print). Such a “content unreadable” part can be corrected via the input operation unit 7.

It is a figure which shows the structure of an example of the minutes production apparatus which concerns on this invention. It is a figure which shows an example of a display of the minutes data in a display part. It is a figure which shows the structure of the other example of the minutes production apparatus which concerns on this invention.

Explanation of symbols

1,21 Minutes creation device 2 CPU
3, 3 a, 3 b,..., 3 n Microphone 4 Voice recognition unit 5 Speaker identification unit 6 Voiceprint data storage unit 7 Input operation unit 8 Display unit 9 Printing unit 10 ROM
11 RAM

Claims (5)

  In an apparatus for creating a minutes of a meeting, a creation means for creating minutes data using speech recognition, a specifying means for specifying a speaker, a statement content in the minutes data created by the creation means, and the specifying means Means for associating and storing the speaker specified by the person.   2. The minutes creation apparatus according to claim 1, further comprising output means for outputting minutes data so that the contents of the comments are displayed in different forms depending on the speaker.   3. A minutes creation apparatus according to claim 2, further comprising means for storing speaker information relating to a speaker, wherein the output means outputs the minutes data by adding the speaker information. .   4. The minutes creation apparatus according to claim 1, further comprising means for registering a meeting attendee based on a speaker identified by the identification means. Means for storing a predetermined keyword, and detecting means for detecting the presence of the keyword in a statement content, and the output means includes the keyword when the detecting means detects the presence of the keyword. 4. The apparatus according to claim 2, wherein the content of the utterance and / or the content of the utterance before and after the utterance content including the keyword is subjected to emphasis processing and the minutes data is output. 5. Minutes creation device.

Images