Method and apparatus for transmitting artificial intelligence-based data by using distributed network

1 . An operating method of a first agent, the operating method comprising:

transmitting a message including a key frame of an input video received from a first client corresponding to the first agent to a second agent; and

receiving, from the first client, key points extracted from the input video as features for every frame after the key frame and transmitting the received key points to the second agent, wherein

the first agent and the second agent communicate through a distributed network,

wherein the message includes:

a version field including information representing a version of a protocol for a distributed networking;

a type field including information representing an encoding type of the message;

a length field including length information of a header field;

the header field including header information for classifying the message and length information of an extension header field and a payload field;

the extension header field including information of a binary data contained in the payload field; and

the payload field including one of the key frame, the key points, or control data for a session for transmission of the input video,

wherein the input video is generated based on the key frame and a frame generated from the key points received through the second agent by a second client corresponding to the second agent.

2 . The operating method of claim 1 , wherein the key frame is a first frame when an angle and/or position of an object comprised in the input video changes.

3 . The operating method of claim 1 , wherein the key points comprise a coordinate value for at least one of an eye, nose, mouth, ear, or jawline of a face comprised in the input video.

4 . The operating method of claim 1 , further comprising opening a session for video transmission between the first client and a hybrid overlay management server (HOMS) configured to manage the first agent and the second agent, wherein

a distributed network between the first agent and the second agent is generated in response to the second agent joining the session.

5 . A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the operating method of claim 1 .

6 . A first agent comprising one or more processors, wherein the one or more processors are configured to:

transmit a message including a key frame of an input video received from a first client corresponding to the first agent to a second agent, and

receive, from the first client, key points extracted from the input video as features for every frame after the key frame and transmit the received key points to the second agent, wherein

the first agent and the second agent communicate through a distributed network,

wherein the message includes:

a version field including information representing a version of a protocol for a distributed networking;

a type field including information representing an encoding type of the message;

a length field including length information of a header field;

the header field including header information for classifying the message and length information of an extension header field and a payload field;

the extension header field including information of a binary data contained in the payload field; and

the payload field including one of the key frame, the key points, or control data for a session for transmission of the input video,

wherein the input video is generated based on the key frame and a frame generated from the key points received through the second agent by a second client corresponding to the second agent.

7 . The first agent of claim 6 , wherein the key frame is a first frame when an angle and/or position of an object comprised in the input video changes.

8 . The first agent of claim 6 , wherein the key points comprise a coordinate value for at least one of an eye, nose, mouth, ear, or jawline of a face comprised in the input video.

9 . The first agent of claim 6 , wherein the one or more processors are further configured to open a session for video transmission between the first client and a hybrid overlay management server (HOMS) configured to manage the first agent and the second agent, wherein

a distributed network between the first agent and the second agent is generated in response to the second agent joining the session.

10 . The operating method of claim 1 , wherein the payload field further includes a signature for a hash value of the key frame.