JP2000333150A - Video conference system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate generation of a minute after a conference by allowing the system to take a special attention to major talkers and interrupt talkers in the case that many attendees simultaneously make a speech. SOLUTION: The video conference system is provided with a major talker decision means 2 that discriminates revision of major talkers on the basis of a sate a silencing state of other major talker during utterance of the major talker when the other major talker makes an utterance while the major talker displayed on a base image 51, an interrupt talker decision means 3 that decides an interrupt talker displayed on a small image 52 on the basis of it as to whether or not other utterance comes from the major talker displayed on the base menu 51, an image synthesis display means 6 that displays a decided major talker onto the base image 51 and displays the interrupt talker onto the small image 52 on the base image, and a minute capturing means 7 that stores information of the major talkers and the interrupt talkers together with time information or the like.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video conference system capable of easily collecting video and audio minutes.

[0002]

2. Description of the Related Art In a general video conference, a change in an image image is scarce and a sense of reality is given. Although there is a function to automatically zoom up the speaker, it cannot cope with a case where many people speak at the same time. In addition, as in the case of the recording, the transcription of the minutes by listening to the conversation from the recorded contents of each speaker after the meeting is very difficult for humans to concentrate on the speaker normally, so the speaker is very specific. It is difficult to do.

[0003]

Since the conventional video conference system is configured as described above, it is not possible to pay attention to the speaker as in a normal conference, and it is difficult to specify the speaker. There was a problem that it was not easy to create minutes by hearing after the fact.

[0004] Therefore, an object of the present invention is to focus on a main speaker who makes a main utterance and an interrupt speaker who makes a question or the like when a large number of people utter at the same time, as in a normal conference. The purpose of the present invention is to provide a video conference system that facilitates the creation of minutes.

[0005]

According to the video conference system of the present invention, when a main speaker displayed on the base screen among the video conference participants is speaking, there is another speech. Based on the state of silence in the speech of the main speaker, a determination is made as to the change of the main speaker, and whether or not the other speech is due to the main speaker displayed on the base screen Speaker determining means for determining an interrupting speaker to be displayed on a small screen based on the result, and displaying the main speaker on the base screen based on the determination result of the speaker determining means. Is displayed on the small screen on the base screen as a composite image, and the composite output means for outputting voices about utterances made by the main speaker and the interrupting speaker; and Key speaker and former Characterized by comprising the proceedings collecting means for storing the information of the interrupt speaker along with the time information, and the like.

According to the video conference system of the present invention, when a main speaker displayed on the base screen among the video conference participants is making a speech and there is another speech, the main conference speaker is speaking. Judge the change of the main speaker based on the state of silence, and further change to a small screen based on whether or not the other utterance is due to the main speaker displayed on the base screen. The interrupting speaker to be displayed is determined, and based on the determination results, the main speaker is displayed on the base screen, and the interrupting speaker is output to a small screen on the base screen as a synthesized image together with voice. Then, the information of the determined main speaker and the interrupted speaker is stored together with time information and the like, and based on this information, the main speaker and the interrupted talk when a large number of people simultaneously speak as in a normal meeting. The minutes of the meeting after the meeting.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below. A video conference system according to an embodiment of the present invention automatically detects a main speaker in a video conference and displays a screen of an interrupting speaker such as a questioner on a screen by a cut-in technique. In addition, by taking the record, the video and audio minutes with speaker annotations can be automatically collected.

FIG. 1 is a block diagram showing a configuration of a video conference system according to the present embodiment. A camera / microphone 1, a main speaker determination unit (speaker determination unit) 2, and an interrupt speaker determination unit (speaker) Person determination means) 3, base screen means 4, cut-in means 5, image synthesis display means (synthesis output means) 6, minutes collection means 7, and the like.

A camera / microphone 1 captures a video conference participant in a video conference at each location where the video conference participant is present and outputs it as, for example, a still image signal. Is output as an electric audio signal. The main speaker determination means 2 determines a person who has been talking for a long time as the main speaker using the algorithm shown in FIG. 2 based on the input to the microphone of the camera microphone 1. .

The interrupting speaker judging means 3 judges a person who speaks in a relatively short time, such as a questioner, as an interrupting speaker by the algorithm shown in FIG.

The base screen means 4 is for displaying the main speaker determined by the main speaker determination means 2 among the video conference participants photographed by the camera of the camera / microphone 1 as a base screen of an image of the video conference. Things.

The cut-in means 5 is for displaying, as a small screen, the interrupt speaker determined by the interrupt speaker determining means 3 among the video conference participants photographed by the camera of the camera / microphone 1.

The image synthesizing and displaying means 6 is for displaying the base screen and the small screen on the video conference screen as a synthetic image having a form as shown in FIG. In FIG. 4, reference numeral 51 denotes the base screen, and reference numeral 52 denotes a small screen.

The minutes collecting means 7 is a main speaker determining means 2
The information of the main speaker determined in step (1) and the information of the interrupting speaker determined by the interrupting speaker determination means 3 are stored together with time information and the like. This makes it possible to create minutes information with speaker annotations.

Next, the operation will be described. In a video conference, the main speaker tends to talk continuously for a long time, and questions and jiji are often asked by other participants as occasion demands. Utilizing this characteristic, a person who has been talking for a long time is determined as a main speaker by an algorithm shown in FIG. 4 is used as the base screen of the video conference image.

FIG. 2 is a flowchart showing the main speaker determination operation by the main speaker determination means 2. According to this flowchart, first, it is determined whether or not there is a person who is talking in a video conference (step S1).
As a result, if no one is speaking, the base screen 51 of the current video conference is maintained as it is (step S
5). On the other hand, if there is a speaker, it is determined whether the speaker is the speaker currently displayed on the base screen 51 (step S2). As a result of this determination, if the speaker is the speaker displayed on the current base screen 51, the base screen 51 of the current video conference is maintained as it is (step S5).

In step S2, if the speaker is not the speaker currently displayed on the base screen 51, the person displayed on the current base screen 51 is in a state of silence, for example, silence. It is determined whether the time has exceeded the defined value (step S3). In this case, if the silence time does not exceed the defined value and the speaker displayed on the base screen 51 starts speaking, the speaker displayed on the base screen 51 without being changed remains the main speaker. The speaker is displayed on the base screen 51 (step S5).

In step S3, when the silence time of the person currently displayed on the base screen 51 exceeds the defined value, the base screen 51 is changed and the main speaker displayed on the base screen 51 is changed. The person who is currently silent on the base screen is changed to the person who is currently talking (step S4).

Also, a person who speaks for a relatively short time, such as a questioner, is determined as an interrupting speaker by the interrupter determining means 3 by the algorithm shown in FIG. 3 and displayed as a small screen 52 by the cut-in means 5. .

FIG. 3 is a flowchart showing the operation of the interrupting speaker judging means 3 for judging the interrupting speaker. According to this flowchart, first, it is determined whether there is a person who is talking in the video conference (step S11). As a result, if there is no person talking, the small screen 52 of the current video conference is left as it is. Maintain (step S15). On the other hand, if there is a speaker, it is determined whether the speaker is the speaker currently displayed on the base screen 51 (step S12). As a result of the determination, if the speaker is the speaker displayed on the current base screen 51, the base screen 51 of the current video conference is maintained as it is (step S15).

In step S12, if the speaker is not the one currently displayed on the base screen 51, it is determined whether the speaker currently speaking is displayed on the small screen 52 (step S13). ). As a result, if the speaker is displayed on the small screen 52, the process proceeds to step S15, and the speaker displayed on the small screen 52 without being changed is displayed on the small screen 52 as an interrupting speaker (step S15). S15).

If the current speaker is not displayed on the small screen 52 in step S13, the current speaker is displayed on the small screen 52 (step S14).

As described above, a person who speaks for a relatively short time, such as a questioner, is determined as an interrupting speaker by the interrupting speaker determining means 3 and is displayed as a small image 52 by the cut-in means 5.

The images of the base screen 51 and the small screen 52 are displayed on the video conference screen as shown in FIG. Also, the speeches of the main speaker and the interrupting speaker are output as voice.

The minutes collecting means 7 is a main speaker judging means 2
The information on the main speaker determined in step (1) and the interrupt speaker determined by the interrupting speaker determination means 3 is stored together with time information and the like.

By comparing this with information such as images and sounds output from the image synthesizing and displaying means 6, minutes information with speaker annotations is created.

Therefore, according to the present embodiment, the main speaker and the interrupting speaker can be clearly distinguished and grasped by the base screen 51 and the small screen 52. Since the main speaker on the screen 51 and the person who talks for a short time such as a question can be identified and noticed as the interrupting speaker on the small screen 52, when a large number of people speak at the same time, the main speech is made as in a normal meeting. It is possible to provide a video conference system in which a speaker and an interrupting speaker who makes a question can be noticed and a minutes can be easily prepared by listening based on the recorded contents after the meeting.

In the embodiment described above, the information of the main speaker determined by the main speaker determining means 2 and the information of the interrupting speaker determined by the interrupting speaker determining means 3 are stored together with time information and the like. Then, the apparatus is provided with the minutes collecting means 7 which enables the generation of the minutes information with the speaker annotation by comparing with the synthesized image and the voice generated by the image synthesizing and displaying means 6. By adding a voice recognition function to the collection function 7, the construction of minutes may be semi-automated.

In the case of a large number of people, the speaker name may be superimposed on the base screen display means 4 and the cut-in means 5 so as to clarify the speaker.

[0030]

As described above, as described above, in the same way as the attention of the speaker in the ordinary meeting, when a large number of people speak at the same time, the speaker who makes a main speech and the interrupt speaker who makes a question etc. Attention can be paid, and the person who speaks in the video conference displays the image on either the base screen or the small screen, so that the image of the speaker is reliably displayed and the situation where the speaker can not be identified can be prevented. In addition, since the information about the speaker is automatically collected, it is possible to reduce the labor of preparing the minutes by hearing after the meeting based on the recorded contents after the meeting.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a video conference system according to an embodiment of the present invention.

FIG. 2 is a flowchart showing a main speaker determination operation by a main speaker determination unit in the video conference system according to one embodiment of the present invention;

FIG. 3 is a flowchart illustrating an interrupting speaker determination operation performed by an interrupting speaker determination unit in the video conference system according to the embodiment of the present invention;

FIG. 4 is an explanatory diagram showing display contents of a video conference screen in the video conference system according to one embodiment of the present invention;

[Explanation of symbols]

1 camera microphone 2 main speaker determination means (speaker determination means) 3 interrupt speaker determination means (speaker determination means) 4 base screen means 5 cut-in means,
6 ... image synthesis display means (synthesis output means) 7 ... minutes collection means

Claims (5)

[Claims] 1. A video conference system for a plurality of video conference participants to conduct a conference from each location via a video conference participant's image and voice, wherein the video conference participant is displayed on a base screen among the video conference participants. If there is another utterance while the main speaker is speaking, while making a decision on the change of the main speaker based on the state of silence during the speech of the main speaker, Speaker determination means for determining an interrupting speaker to be displayed on a small screen based on whether or not another utterance is from the main speaker displayed on the base screen; and a determination result of the speaker determination means Based on the above, while the main speaker is displayed on the base screen, the interrupting speaker is displayed as a composite image on a small screen on the base screen, and a statement made by the main speaker and the interrupting speaker is performed. of Synthesizing output means for outputting a voice; and minutes collecting means for storing information of the main speaker and the interrupting speaker determined by the speaker determining means together with time information and the like. Video conference system. 2. The combined minutes output from the combined output means, wherein the minutes collecting means outputs the information stored about the main speaker and the interrupt speaker determined by the speaker determining means,
2. The video conference system according to claim 1, wherein the minutes information with the speaker annotation can be created by comparing with the voice. 3. The speaker judging means, when a main speaker among the video conference participants displayed on the base screen is making a speech, and there is another speech, the speech of the main speaker is made. A determination is made as to the change of the main speaker based on the time during which the main speaker is considered to be silent, and whether or not the other utterance is due to the main speaker being displayed on the base screen is determined. 3. The video conference system according to claim 1, wherein the interrupting speaker to be displayed on the small screen is determined. 4. The speaker judging means, when a main speaker displayed on the base screen among the video conference participants is making a speech, and there is another speech, the speech of the main speaker is made. Main speaker determination means for determining the change of the main speaker based on the time considered to be in silence,
Interrupt speaker determining means for determining an interrupting speaker to be displayed on a small screen based on whether or not the other utterance is due to the main speaker displayed on the base screen. The video conference system according to claim 3, wherein: 5. A base screen means for displaying, as a base screen of a video conference image, a main speaker determined by the main speaker determination means among the video conference participants, and 5. The video conference system according to claim 4, further comprising cut-in means for displaying, as a small screen, the interrupting speaker determined by said interrupting speaker determining means.