Method for presenting a person

FIELD: computer network communication means.

SUBSTANCE: method includes conversion of speech to electric digital signal, transfer of said signal to sound-playing device, conversion of person face to electric digital signal, recognition of face, its characteristic areas and their movement parameters, transfer of information along communication channels to graphic information output device, control of shape changes and space direction of artificial three-dimensional object and its characteristic areas. Method additionally includes detecting errors in face recognition and accompanying parameters by detecting mismatches between configurations of face areas and characteristics of movement thereof for speaking person in electric digital signals, and correction of mistakes before visualization of artificial three-dimensional object by forming control commands on basis of previously recorded shape signs and orientation of three-dimensional object and its characteristic areas for speech characteristics.

EFFECT: higher reliability and precision.

3 cl, 1 dwg

 

The invention relates to the field of telecommunications via electronic means, in particular computer networks. More specifically the invention relates to a method of representation of the person on the device for displaying graphical information.

Known methods of representation of the person on the computer display when the telecommunications, computer networks via artificial three-dimensional objects, the so-called animated chat (see, for example, R.Lea, Y.Honda, K.Matsuda, and S.Matsuda. Community Place: Architecture and Performance, in Proceedings of the VRML'97 Symposium, ACM SIGGRAPH, 1997, p. 41-49).

Closest to the present invention is a method of representation of the person by means of an artificial three-dimensional object on the basis of video and audio information in telecommunication networks (see, for example, http://www.worldsaway.com)

This method is chosen as a prototype. Prototype method includes converting using the microphone audio, including human speech, into a digital electrical signal, the operation of transmitting the signal via the communication channels to the audio device, the conversion operation using a camera image of a scene including a person's face, in an electric digital signal, the operation of the recognition of the human face, its characteristic regions and the characteristics of their movement, the transfer operation of the above info is of rmacie over the communication channel, operation render artificial three-dimensional object on the graphical output device information, operations, change control forms, spatial orientation artificial three-dimensional object and its characteristic regions.

In the method-prototype digital signal corresponding to the human face, its characteristic regions and the characteristics of their traffic is routed directly to the visualization of artificial three-dimensional object and with its help you can manage changes in the shape and spatial orientation of an artificial three-dimensional object. Because it contains, in addition to reliable information, and error detection associated with the imperfection of the method and hardware recognition, prototype method has the following disadvantages:

- form distortion artificial three-dimensional object as a whole;

- distortion characteristic areas of artificial three-dimensional object;

the distortion of the movements, facial expressions and gestures artificial three-dimensional object.

The above-mentioned disadvantages lead to the fact that artificial three-dimensional object inadequately reflects the person and his behavior and, accordingly, the prototype method has poor accuracy and reliability of the representations in telecommunications.

The technical result of the proposed the procedure of representation of the person when telecommunications is a more accurate reflection of human emotional States in the process of telecommunications.

Another technical result of the proposed method is to increase its reliability by eliminating the loss of information associated with the imperfection of technical means and methods of detection based on video information.

These technical results achieved in the way of representation of the person by means of an artificial three-dimensional object on the graphical output device information on the basis of audio and video in telecommunications in computer networks, including: converts audio information containing human speech using a microphone into an electrical digital signal; an operation of transmitting the signal via the communication channels to the reproducing device; converts an image of a scene containing a human face, with the camera into an electrical digital signal; an operation detection in the scene mentioned in the human face, its characteristic regions and the characteristics of their movement; the transfer operation of the above-mentioned detected information communication channels to the output device graphics data; operation visualization artificial three-dimensional object on said graphical output device information; an operation control changes in the shape and spatial orientation of an artificial three-dimensional about the project and its characteristic regions, which further comprises: an operation detecting errors in recognizing a human face, its characteristic regions and the characteristics of their movement by identifying discrepancies between configurations characteristic regions of the human person and the characteristics of their movement speaking person, contained in the above-mentioned electrical digital signals; and an operation of correcting the above errors before rendering artificial three-dimensional object by generating a control command using a pre-recorded signs of shape and spatial orientation of an artificial three-dimensional object and its characteristic regions corresponding to the characteristics of speech.

Differences method according to the invention are also in the fact that the operation of detecting errors in the recognition of the human face, its characteristic regions and the characteristics of their movement is carried out by determining inconsistencies configurations characteristic of the face of the person and characteristics of their movement in human speech contained in the digital signals corresponding to the video and audio information, and an operation to repair the above-mentioned error detection before rendering the artificial three-dimensional object is produced by forming mentioned at least part of the control commands based on the Audi the information.

Differences of the second variant of the method according to the invention are also in the fact that the operation of detecting errors in the recognition of the human face, its characteristic regions and the characteristics of their movement is carried out by determining inconsistencies configurations characteristic of the face of the person and characteristics of their movement in human speech contained in the digital signals corresponding to the video information and behavior model of an artificial three-dimensional object, comprising a set of characteristic gestures and facial expressions, and surgery to repair the above-mentioned error detection before rendering the artificial three-dimensional object is produced by forming the above-mentioned control commands based on those behaviors.

The way of man through artificial three-dimensional object according to the invention illustrated by the drawing.

The way of man according to the present invention involves a preliminary operation 1 create a model of the behavior of an artificial three-dimensional object. Model behavior may be a configuration sets the characteristic of the face of the artificial three-dimensional object, for example, multiple configurations of the mouth associated with the different possible emotional States artificial three-dimensional object is. In the behavior model can be included and additional relative position of the different characteristic areas of the face artificial three-dimensional object, for example, eyes can always be located above the mouth, and eyebrows above the eyes. In the behavior model may also include the values of maximum permissible velocities and angles of rotation of the face of the artificial three-dimensional object in different directions. The above are not limited to, all possible behaviors. Behaviors can present various combinations of static and dynamic parameters. The model can be represented in the form of digital codes recorded in the storage device (step 2) and read (operation 3) from the storage device in the form of digital electrical signals.

The method includes the following operations:

activity 4 receiving in consecutive moments of time video images of a scene, comprising, at least one person, the first party's telecommunication;

activity 5 receiving audio information, including the speech of the first party's telecommunication;

activity 6 face detection is the first participant of telecommunications and the recognition of characteristic regions in each of the successive points in time;

activity 7 definition configurabledirectory areas of the human face from the audio information;

activity 8 the formation of the digital signal corresponding to the detected based on the video information of the human face and its characteristic areas;

activity 9 the formation of the digital signal corresponding to the detected based on the audio information of the human face and its characteristic areas;

activity 10 error detection face recognition of human rights and its specific areas (e.g., mouth, eyes etc) and characteristics of their movement (e.g., direction, speed, angle);

activity 11 fixes the above error.

activity 12 visualization through artificial three-dimensional object on the graphical output device information (for example, a computer display). According to the first variant of the proposed method operation 10 error detection recognition of the human face, its characteristic regions and the characteristics of their movement is carried out by determining inconsistencies configurations characteristic of the face of the person and characteristics of the movement of human speech is contained in the digital signals corresponding to video and audio, and operation 11 hotfix mentioned recognition errors before surgery visualization 12 artificial three-dimensional object is produced by forming mentioned at least part of the control commands on the basis of the audio information. For example, if during a certain period of time, the audio data contains characteristics of speech, and the video does not contain such (for example, the configuration of the mouth is not changed at this time), then generates a command to change the configuration of the mouth. If the audio contains signs of laughter, then generates a command to change the configuration of the mouth corresponding to smile. Using pre-recorded in a persistent storage device signs the form, the spatial orientation of the artificial three-dimensional object and its specific areas, such as mouth and eyes, the relevant characteristics of speech. Recorded in a persistent storage device information includes sets of mutual combinations of speech characteristics and geometric shapes, for example, if the audio contains signs of laughter, the video information should contain signs of a smile. If coming to interpret information such there is no match, then, following this rule, issue a display command of a smile when rendering a three-dimensional artificial object. If the audio contains signs of sorrow, and videos of such signs does not contain, then select the appropriate configuration of the mouth and other areas of the face. According to the second variant of the proposed method Opera is July 10 error detection is carried out by determining inconsistencies configurations characteristic of the face of the person and characteristics of the movement of human speech, contained in the digital signals corresponding to the video information and behavior model of an artificial three-dimensional object created at operation 1, the set of characteristic gestures and facial expressions. When this operation fixes the above error detection before rendering the artificial three-dimensional object is produced by forming the above-mentioned control commands based on said model behavior using pre-recorded forms, spatial orientation artificial three-dimensional object and its characteristic regions corresponding to the characteristics of speech. For example, if the movement of a human face on the scene identified from the video occurs at a rate that exceeds the allowable, that is, appropriate behaviors, then when rendering the speed set in accordance with a behavior model of an artificial three-dimensional object. In the same way correct and other recognition errors in the rotation angles of the faces, gestures and facial expressions.

The method according to the invention can be used, for example for exchanging information between at least two persons at a great distance from each other, and connected to any communication channels. As a communication channel can be used, for example, computer network, Internet. Participants telecommunication the AI should be equipped with technical means, including a video camera, a microphone, a computer with software to support telecommunications. Software in addition to the operating system should include, for example, a program that allows real-time exchange audio information and video information between at least two participants in a computer network.

The method according to the invention can be used, for example, in telecommunications as follows. The first participant telecommunications is in the field of view of the camera and lens in front of the microphone and makes a speech message, followed by movements. Camcorder in discrete successive moments of time creates a video image of a scene including a face of the first participant telecommunications. Simultaneously, the microphone generates an audio data including speech first participant telecommunications and produces a corresponding digital signal. On the technical means of the first party's telecommunication using software to perform a discovery operation on a complex background scene in the field of view of the camera faces the first participant telecommunications in each of the successive moments of time. The operation of the face detection is performed, for example, by the method described in Jean-Christophe Terrillon, Mahdad N. Shirazi, Mhamed Sadek, Hideo Fukamachi, Shigeru Akamatsu “Invariant Face Detection with Support Vector Machines”, (p.4210, International Conference on Pattern Recognition (international champion pain relief'00)-Volume 4, September 03-08, 2000, Barcelona, Spain). In the result of the operation of the face detection and tracking his movements and turns in each of the successive moments of time is formed by a digital signal carrying information about the position of the face on the scene and configurations characteristic of the face (mouth, eyes etc). Using hardware and software of the first party's telecommunication mentioned digital signals are transmitted via communication channels on the hardware of the second party telecommunications. Concurrently, the communication channels on the hardware of the second party telecommunications transmitted digital signal representing audio data, including voice messages, the first participant telecommunications. Further technical equipment and software of the second participant telecommunications is mentioned operations identify and correct recognition errors and visualization of the first participant telecommunications. The present example, however, does not exhaust all possible applications of the proposed method view through artificial three-dimensional object. The method can be widely used in various computer technologies.

1. Way to depict what Alenia man through artificial three-dimensional object on the graphical output device information on the basis of audio and video in telecommunications in computer networks includes the conversion of audio information containing human speech using a microphone into an electrical digital signal; an operation of transmitting the signal via the communication channels to the reproducing device; converts an image of a scene containing a human face, with the camera into an electrical digital signal; an operation detection in the scene mentioned in the human face, its characteristic regions and the characteristics of their movement; the transfer operation of the above-mentioned detected channel information due to the graphical output device information; an operation visualization artificial three-dimensional object on said graphical output device information; an operation control changes in the shape and spatial orientation of an artificial three-dimensional object and its characteristic regions, characterized in that it further comprises an operation of detecting errors in recognizing a human face, its characteristic regions and the characteristics of their movement by identifying discrepancies between configurations characteristic regions of the human person and the characteristics of their movement speaking person, contained in the above-mentioned electrical digital signals; an operation of correcting the above errors before rendering artif the th three-dimensional object by generating a control command using a pre-recorded signs of shape and spatial orientation of an artificial three-dimensional object and its characteristic regions, relevant characteristics of speech.

2. The way of man through artificial three-dimensional object according to claim 1, wherein the operation of detecting errors in the recognition of the human face, its characteristic regions and the characteristics of their movement is carried out by determining inconsistencies configurations characteristic of the face of the person and characteristics of their movement in human speech contained in the digital signals corresponding to the video and audio information, and an operation to repair the above-mentioned error detection before rendering the artificial three-dimensional object is produced by forming mentioned at least part of the control commands based on the audio information.

3. The way of man through artificial three-dimensional object according to claim 1, wherein the operation of detecting errors in the recognition of the human face, its characteristic regions and the characteristics of their movement is carried out by determining inconsistencies configurations characteristic of the face of the person and characteristics of their movement in human speech contained in the digital signals corresponding to the video information and behavior model of an artificial three-dimensional object, comprising a set of characteristic gestures and facial expressions, and surgery to repair the above-mentioned recognition errors n the ed visualization artificial three-dimensional object is produced by forming the above-mentioned control commands based on those behaviors.



 

Same patents:

The invention relates to computer animation images

The invention relates to the field of hairdressing services and can be used for hair care, during a haircut, shave, manicure and similar cosmetic treatment

The invention relates to sports training equipment and can be used in the organization of sporting events

The invention relates to the field of creation of the animation effects associated with the color image, in particular, three-dimensional, intensifying the emotional impact when accompanied by appropriate sound works, and can be used when conducting entertainment events

FIELD: computer network communication means.

SUBSTANCE: method includes conversion of speech to electric digital signal, transfer of said signal to sound-playing device, conversion of person face to electric digital signal, recognition of face, its characteristic areas and their movement parameters, transfer of information along communication channels to graphic information output device, control of shape changes and space direction of artificial three-dimensional object and its characteristic areas. Method additionally includes detecting errors in face recognition and accompanying parameters by detecting mismatches between configurations of face areas and characteristics of movement thereof for speaking person in electric digital signals, and correction of mistakes before visualization of artificial three-dimensional object by forming control commands on basis of previously recorded shape signs and orientation of three-dimensional object and its characteristic areas for speech characteristics.

EFFECT: higher reliability and precision.

3 cl, 1 dwg

FIELD: information technology.

SUBSTANCE: invention relates to animation systems, and particularly to communication protocol for supporting information and time synchronisation between several animation systems. Proposed is a communication protocol, which controls asynchronous data exchange between a high-level animation system and a low-level animation system. The high-level animation system has varying, average frame frequency and is optimised for interactivity. The low-level animation system has constant, high frame frequency and is optimised for high frequency of frame updating. The communication protocol contains messages which can be sent by the high-level animation system to the low-level animation system so as to describe animation and how animation should change for after given period of time. As a result, the low-level system can display information at high rate of updating, even if animation data are not being received from the high-level system for each frame.

EFFECT: more efficient exchange of required data in an animation system.

26 cl, 9 dwg

FIELD: information technologies.

SUBSTANCE: three-dimensional icon configuration file is read for current user; at least one set of three-dimensional elements specified in file of three-dimensional icon configuration is read; at least one set of three-dimensional elements is displayed in accordance with three-dimensional icon configuration file. Also client of immediate messages exchange, server and system of three-dimensional icon display are suggested. Due to mentioned technical features, three-dimensional icon is displayed into immediate messages exchange client, which may be used by user to demonstrate various personal images, freely selecting a set of three-dimensional elements.

EFFECT: reduced delay in communication network in process of messages exchange with three-dimensional icons in terminals of information networks with existing throughput capacity of networks.

25 cl, 20 dwg

Up!