JPH0267889A - Displayed person identifying communication conference system - Google Patents

Abstract

PURPOSE:To manage a conference efficient in the preparation of minutes, etc., by recognizing automatically the face picture of an attendee in a video conference on a display and superimposing and displaying automatically the name and various attributes of the attendee on a display screen. CONSTITUTION:The title system is equipped with a displayed person recognizing part 4 to identify a person among conference attendees displayed on a display 8 and a data base 5 to hold information about the conference attendees. The person displayed on the display 8 is recognized by the recognizing part 4, who he is among the conference attendees is automatically discriminated, the information from the data base 5 is suitably added and the information about the person displayed on the display 8 is presented to a receiver who watches the display. Thus, who the displayed person is among the conference attendees can be discriminated by the video conference system and the inconvenience in the proceeding of the conference and the preparation of the minutes can be eliminated.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field] The present invention relates to a display person identification teleconferencing system that is highly functional in a TV conference system.

[Prior Art] In conventional TV conference systems, it is not always possible to identify who among the conference participants is the person shown on the display, which may be inconvenient for progressing the conference or creating minutes.
When you participate in an executive meeting using a TV conference as a non-regular member, it is often difficult to tell who the executive is on the display! II + to the secretariat after the meeting to prepare the minutes! The disadvantage was that it caused inconveniences such as recognition.

[Object of the present invention]

In order to eliminate the above-mentioned drawbacks, the present invention recognizes the person displayed on the display using a facial image recognition device to automatically determine who among the participants is, and also uses information from the participant database as appropriate. In addition, information regarding the person displayed on the display is provided to the recipient viewing the display.

In addition, in order to improve recognition accuracy, 1
1! In addition to image recognition, it also uses voice recognition.

〔Example〕

FIG. 1 is a diagram showing a first embodiment of the present invention. In communication conferences, two-way communication or communication between multiple points is common, but only two-way communication is shown here for simplicity. Bidirectional or point-to-point communication can be achieved by combining these unidirectional systems.

In the figure, 1 is one of the participants in the communication conference, and the image of 0 displayer 1 is captured by camera 2 (hereinafter referred to as the display person), and the image of 0 displayer 1 is on the display side 7? The image is displayed on the receiving side display of 8 through the 111 part like a human image of 9. The signal from camera 2 is 3
The face image processing unit in 4 extracts the target face image and extracts its features, and the face image recognition unit in 4 compares it with the meeting participant database in 5 to identify the person displaying the image.The identification result is coded as a name code. etc., is sent to the display sound information providing section 6.

The display sound information providing unit 6 uses a name code etc. as a key,
The conference participant database 5 is searched for the name, affiliation, position, etc. corresponding to the name code, and the search results are sent to the display control section 7. The display control section 7 converts the information from the one-displayer information providing section 6 into information to be provided to the recipient. For example, if a name is to be superimposed on the reception screen, this process is performed.

As a result, the name ``Yamada'', such as the name 10, is superimposed and displayed on the receiving side display 8 together with the human image 9. The face image processing section 3 can be realized by a technique such as "Face Image Feature Point Extraction Method" described in Technical Research Report PRL80-8 of the Institute of Electronics and Communication Engineers.

FIG. 2 is a diagram showing a second embodiment of the present invention. In this embodiment, in addition to the configuration shown in FIG. 1, 11 microphones and 12 audio processing units are added. In order to improve the recognition rate of the display person, the voice of the display person is collected by the microphone 11, the voice processing unit 12 extracts the features, and the display person recognition unit 13 collects the face obtained by the image processing unit 3. The following process for identifying one display person using the 14 conference participant databases based on image characteristics and audio characteristics is the same as that shown in FIG. 1.

FIG. 3 is a diagram showing an example of displaying information regarding two displayers. In Fig. 3 (A), 20 is a display.

21 is an image (person image) of the display person, and 22 is the name of the superimposed display person. FIG. 3(B) shows another display example, where 23 is the display, 24.25 is the first . This is the second window. The first window 24 has an image of the person displaying it.

The second window 25 displays the name and affiliation of the person being displayed.
Information such as job title is displayed.

FIG. 4 is a diagram showing a third embodiment of the present invention. 1st
．． The recognition unit shown in the second embodiment can be installed on the transmitting terminal side, the receiving terminal side, two communication conference nodes, etc., and the installation location is not limited, as shown in FIG. 4. is an example of installing a recognition unit etc. in a communication conference node. 3
1, 32, 33 are video telephones or video conference terminals, 34
is a communication network such as a local area network or a public network, and 35 is a communication conference node. The communication network 34 may be a circuit switching network, a bucket switching network, or an asynchronous transfer network.

An example of a circuit switched network is shown. In other words, a plurality of subscriber lines corresponding to the maximum number of multipoint conferences are connected to one teleconference node 35. Terminals 31, 32, and 33 are connected to one teleconference node 35 via five communication networks 34, and a conference is being held between three points.

FIG. 5 is a diagram showing an example of the configuration of a teleconference node. 41
, 42, and 43 are receiving units connected to the terminals 31, 32, and 33, respectively, and 44 is a screen switching unit. Terminal 31,
The video information from 32 and 33 is switched manually or automatically at a switching unit 44 according to an instruction from a host terminal (for example, 31) or the volume of the voice of the speaker at each terminal. The switched information is sent to the transmitter 50 via the display controller 49. From the transmitter, further terminal 31.32゜3
The information is sent to the necessary terminals among the three through the communication network. The output video signal of the switching unit 44 is simultaneously subjected to sampling and feature extraction of a target face image in a face image processing unit 45.
By the facial image recognition section 6.

The 9 displayers are identified by checking against the 47 conference participant databases. The identification results are sent to 48 display sound information providing units using a name code or the like. The display sound information providing section 48 uses the name code as a key to search the conference participant data field 47 for the name, affiliation, position, etc. corresponding to the name code, and sends the search results to the display control section 49 . 1 in the display control section 49
The information from the displayer information providing unit 48 is converted into information to be provided to the recipient. For example, if a name is to be superimposed on the reception screen, this process is performed. The processing results are sent to the receiving terminal via the transmitter 50.

Face image processing section 45. The face image recognition unit 461 and the display person information providing unit 48 are restarted when the switching unit 44 switches the line or when there is a scene change on the screen of a specific line, and performs recognition processing of the face image of the new display person. and provide new information.

FIG. 6 shows another example of the configuration of one teleconference node.

61, 62, and 63 are reception units connected to the terminals 31, 32, and 33, respectively, and 69 is a display control unit that synthesizes screens from a maximum of four locations, for example. The synthesized screen is sent via a transmission unit 70. , and are transmitted to the necessary terminals among the terminals 3L, 32, and 33 via the communication network. 64 is a screen switching section.

The video information from the terminals 31, 32, and 33 is transferred to the switching unit 6.
4, the screens of the receiving units 61, 62, and 63 are sequentially inspected, and when there is a scene change or by remote control from a host terminal, etc., the following recognition processing unit 65 is activated. The R image processing unit 65 performs facial image sampling and feature extraction, and the facial image recognition unit 66 performs facial image sampling and feature extraction.

1 The display person is identified by comparing it with the conference participant database 67. The identification result is sent to the display sound information providing section 68 using the 2 name code etc. 0 The display person information providing section 68 uses the 9 name code etc. as a key. Then, the conference participant database 67 is searched for the name, affiliation, position, etc. corresponding to the name code, and the search results are sent to the display control section 69.
Converting information from the displayer information providing unit 68 into information to be provided to the recipient. For example, when superimposing a name on the reception screen, this process is performed. The processing results are sent to the receiving terminal via the transmitter 70.

FIG. 7 shows an example of the output of the display control unit 69, and 80 is a display on the receiving side, which shows an example of reception by the terminal 33, for example. Terminal 3
It is assumed that Mr. A, Mr. B, and Mr. C are facing their respective cameras as conference participants at 1, 32, and 33.

Mr. A is on the small screen 81.

Mr. B himself is displayed in 82, and Mr. C himself is displayed in 83, and their names are also superimposed and displayed in each small screen.

〔Effect of the invention〕

As described above, according to the present invention, by automatically recognizing the face images of participants on the display during a TV conference, the names and various attributes of the participants are automatically superimposed on the display screen, ! This method has the advantage of enabling efficient meeting management in the preparation of i-minutes and the like.

FIG. 3 is a diagram showing an example of displaying information regarding a display person.

FIG. 4 is a diagram showing a third embodiment of the present invention.

FIG. 5 is a diagram showing an example of the configuration of a communication conference node.

FIG. 6 shows a circuit 7 showing another example of the configuration of the communication conference node. FIG. 6 shows an example of the display of the display control section.

In the figure, 1 is the display person, 2 is the camera, 3 is the face image processing unit, 4
5 is a face image recognition unit, 5 is a conference participant database, 6 is a display person information acquisition unit, and 7 is a display control unit.

8 represents a display.

Patent applicant: Nippon Telegraph and Telephone Corporation

Claims (1)

[Claims] The TV conference system includes a display person recognition unit that identifies a person displayed on a display among conference participants by a face image recognition unit and/or a voice recognition unit, and a database that holds information regarding the conference participants, A display person identification teleconference system characterized in that the database is indexed based on the result identified by the display person recognition unit, and information regarding the displayed person is provided to the recipient.