System and Program

Abstract

An object of the present invention is to provide a system and a program capable of reducing a communication load.SOLUTION: Each of the one or more second computers includes a second generation AI. The one or more first computers acquire generation information indicating a text to be input to the first generation AI and generation information indicating a text to be generated by the first generation AI, and transmit the generation information to the one or more second computers. The one or more second computers acquire the generation information and cause the second generator to generate the second image AI based on the generation information.SELECTED DRAWING: Figure 5

Description

Technical Field

[0001] The present invention relates to a system and a program.

Background Art

[0002] As an invention related to a conventional system, for example, a program described in Patent Document 1 is known. In this program, a server distributes the video being captured by a terminal to a plurality of participant terminals. Thereby, a plurality of participants can view the live video of the distributor.

[0003] By the way, in the field of the program described in Patent Document 1, there is a system in which a video of an avatar linked to the movement of a distributor is distributed. In this case, the video data of the avatar is transmitted from a server to a plurality of distributor terminals.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] In the above-mentioned system, there is a desire to reduce the communication load.

[0006] Therefore, an object of the present invention is to provide a system and a program capable of reducing the communication load.

Means for Solving the Problems

[0007] The first aspect is a system including one or more first computers and one or more second computers, each of the one or more second computers includes a second generation AI, The one or more first computers mentioned above are: The system acquires generation information, which is information indicating the text to be input to the first generation AI when generating the first image data, and / or generation information, which is information indicating the text generated by the first generation AI when generating the first image data. The generated information is transmitted to the one or more second computers. The aforementioned one or more second computers are The generated information is obtained, Based on the generated information, the second generation AI is made to generate the second image data. It is a system.

[0008] The second form is, The one or more first computers mentioned above are: To generate the first image data, a first prompt is input to the first generation AI, The first seed value generated by the first generation AI is obtained by inputting the first prompt. In the process of obtaining the aforementioned generation information, the second seed value related to the first seed value is obtained as the aforementioned generation information. In the process of transmitting the generated information, the second seed value is transmitted to the one or more second computers. The aforementioned one or more second computers are In the process of generating the second image data, the second generation AI is made to generate the second image data based on a third seed value associated with the second seed value. This is the system described in the first form.

[0009] The third form is, The second seed value in the second generation AI corresponds to the first seed value in the first generation AI. The third seed value is the same as the second seed value. This is the system described in the second form.

[0010] The fourth form is, The one or more first computers mentioned above are: It stores correspondence information indicating the correspondence between the seed value in the first generation AI and the seed value in the second generation AI. It specifies the second seed value corresponding to the first seed value based on the correspondence information. It is the system according to the third aspect.

[0011] The fifth aspect is There are the second generation AIs of multiple types of models. The correspondence information indicates the correspondence between the seed value in the first generation AI and the seed values in the second generation AIs of the multiple types of models. The one or more first computers acquire model information indicating the model of the second generation AI from the one or more second computers, and specify the second seed value corresponding to the first seed value generated by the first generation AI based on the model information and the correspondence information. It is the system according to the fourth aspect.

[0012] The sixth aspect is There are the second generation AIs of multiple versions. The correspondence information indicates the correspondence between the seed value in the first generation AI and the seed values in the second generation AIs of the multiple versions. The one or more first computers acquire version information indicating the version of the second generation AI from the one or more second computers, and specify the second seed value corresponding to the first seed value generated by the first generation AI based on the version information and the correspondence information. It is the system according to the fourth or fifth aspect.

[0013] The seventh aspect is The one or more first computers In the process of acquiring the generation information, it acquires the first prompt and the second seed value as the generation information. The above one or more second computers In the process of generating the second image data, based on the third seed value and the first prompt, the second generation AI is made to generate the second image data. The system according to any one of the second to sixth aspects.

[0014] The eighth aspect is The above one or more second computers Store the first prompt, In the process of generating the second image data, based on the third seed value and the first prompt, the second generation AI is made to generate the second image data. The system according to any one of the second to sixth aspects.

[0015] The ninth aspect is The above one or more first computers Obtain motion information indicating the motion of a person, Input the first prompt into the first generation AI, Obtain a first seed value generated by the first generation AI by inputting the first prompt, [[ID=3l]]In the process of obtaining the generation information, obtain a second seed value related to the first seed value as the generation information, In the process of transmitting the generation information, transmit the motion information and the second seed value to the above one or more second computers, The above one or more second computers In the process of generating the second image data, based on the third seed value and the motion information, make the second generation AI generate video data in which a plurality of second images indicated by the second image data are arranged in time series. The system according to any one of the second to eighth aspects.

[0016] The tenth aspect is The above one or more first computers include a third computer and a fourth computer. The aforementioned fourth computer is equipped with the aforementioned first generation AI, The aforementioned third computer, Generate video data showing the video of the aforementioned person, The motion information indicating the actions of the person included in the aforementioned person video data is acquired, The aforementioned operation information is transmitted to the fourth computer. The fourth computer is, In the process of acquiring the aforementioned operation information, the process of acquiring the aforementioned operation information is performed. This is the system described in Form 9.

[0017] The 11th form is, The fourth computer is, A second prompt is obtained to cause the second generation AI to generate video data showing a video of an object linked to the movements of the person, based on the movement information. In the process of transmitting the generated information, the second prompt is sent to the one or more second computers. The aforementioned one or more second computers are In the process of generating the second image data, the second generation AI is instructed to generate the video data based on the third seed value, the operation information, and the second prompt. This is the system described in the 10th form.

[0018] The 12th form is, The aforementioned one or more second computers further comprises a fifth computer and a sixth computer, The fifth computer is, The viewer action information generated in response to the viewer's actions is transmitted to the fourth computer. The fourth computer is, Obtain the aforementioned viewer Action information, A third prompt is obtained that causes a change in the object in accordance with the aforementioned viewer action information. In the process of transmitting the generated information, the third prompt is sent to the sixth computer. The sixth computer is, In the process of generating the second image data, based on the third prompt, the second generation AI is instructed to generate the video data in which changes occur in the object according to the viewer action information. This is the system described in the 11th form.

[0019] The 13th form is, A program to be executed on a second computer, The aforementioned second computer is equipped with a second generative AI, The aforementioned program, The system obtains generated information transmitted from the first computer, which is generated information that indicates the text to be input to the first generation AI when generating the first image data, and / or generated information that indicates the text generated by the first generation AI when generating the first image data. Based on the generated information, the second generation AI generates image data. Based on the aforementioned image data, the image is displayed. To cause the second computer to perform the operation, It is a program. [Effects of the Invention]

[0020] According to this disclosure, the communication load can be reduced. [Brief explanation of the drawing]

[0021] [Figure 1] Figure 1 is a block diagram of systems 1,1a to 1d. [Figure 2] Figure 2 is an explanatory diagram of the operation of System 1. [Figure 3] Figure 3 is an explanatory diagram of the operation of System 1. [Figure 4] Figure 4 is an explanatory diagram of the operation of System 1. [Figure 5] Figure 5 shows the image displayed on the viewer terminal 210-1. [Figure 6]Figure 6 is a block diagram of the broadcaster terminal 10. [Figure 7] Figure 7 is a block diagram of server 110. [Figure 8] Figure 8 is a block diagram of the viewer terminal 210-1. [Figure 9] Figure 9 is a flowchart showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1. [Figure 10] Figure 10 is a flowchart showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1. [Figure 11] Figure 11 is a seed value conversion table. [Figure 12] Figure 12 is a flowchart showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1. [Figure 13] Figure 13 is the seed value conversion table. [Figure 14] Figure 14 is a flowchart showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1. [Figure 15] Figure 15 is a flowchart showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1. [Figure 16] Figure 16 shows the image displayed on the viewer terminal 210-1. [Figure 17] Figure 17 is a flowchart showing the operation of the control unit 212 of viewer terminal 210-1, the control unit 112 of server 110, and the control unit 212 of viewer terminal 210-2. [Figure 18] Figure 18 is a block diagram of system 1e. Figure 19 is a flowchart showing the operation of the control unit 212 of the viewer terminal 210-1 and the control unit 112 of the server 110. [Figure 19]Figure 19 is a flowchart showing the operation of the control unit 212 of the viewer terminal 210-1 and the control unit 112 of the server 110. [Figure 20] Figure 20 is a table showing the relationship between the first to third seed values ​​in this disclosure and the first to third seed values ​​in systems 1,1a to 1b. [Modes for carrying out the invention]

[0022] (Embodiment) System 1 according to an embodiment of this disclosure will be described with reference to the drawings.

[0023] [System 1 Overview] First, I will explain the overall configuration of System 1 with reference to the diagram. Figure 1 is a block diagram of System 1, 1a to 1d.

[0024] System 1, shown in Figure 1, comprises a broadcaster terminal 10, a server 110, and viewer terminals 210-1 to 210-N. N is a natural number. The broadcaster terminal 10, the server 110, and the viewer terminals 210-1 to 210-N can communicate with each other via a communication network. The network can be the internet, an intranet, or the like.

[0025] The broadcaster terminal 10 is an information processing device used by the video broadcaster. The broadcaster terminal 10 is, for example, a smartphone, a tablet, a home game console, a portable game console, a personal computer, or a standalone virtual reality (VR) head-mounted display.

[0026] Each of the viewer terminals 210-1 to 210-N is an information processing device used by the first viewer or the Nth viewer. Viewer terminals 210-1 to 210-N are, for example, smartphones, tablet devices, home game consoles, portable game consoles, or personal computers.

[0027] Server 110 is an information processing device used by the operator of a video distribution service. Server 110 distributes video to viewer terminals 210-1 to 210-N. Server 110 is a computer.

[0028] [System 1 Operation Overview] Next, we will explain the operation overview of System 1. Figures 2 to 4 are explanatory diagrams of the operation of System 1. Figure 5 is an image displayed on the viewer terminal 210-1.

[0029] In System 1, the following two actions are performed: (1) Character determination (2) Distribution of video data

[0030] (1) Character determination The character determination process is the action performed by System 1 when a streamer decides which character to use when streaming a video. Specifically, Server 110 (the fourth computer) is equipped with a first generative AI. A generative AI refers to a large-scale artificial intelligence model used in the field of natural language processing (NLP). These models learn from large amounts of text data (web pages, books, articles, etc.) to understand patterns in human language and effectively perform natural language generation (NLG) tasks.

[0031] Generative AI is used in many NLP tasks, such as generating responses to specific questions, automatically generating text, summarizing text, translation, sentiment analysis, and image generation. It can also be used in a variety of applications, including education, entertainment, customer service, and product development. Generative AI encompasses the following models. In this specification, machine learning models, including large-scale language models that primarily output text information and images, are described as a type of generative AI. OpenAI GPT-4o, Sora Google Gemini Stable Diffusion Midjourney

[0032] As shown in Figure 2, the broadcaster inputs the first prompt P1 by operating the broadcaster terminal 10. The first prompt P1 describes the characteristics of the character to be used as the broadcaster's avatar. In this embodiment, the first prompt P1 is text information that reads, "Please create a human-shaped character with the face of a Shiba Inu." However, the first prompt P1 may be information that has been compressed, vectorized, or binaryized from the text information. The second prompt P2 and third prompt P3, described later, may also be text information, or information that has been compressed, vectorized, or binaryized from the text information, similar to the first prompt P1. In this specification, text information and information that has been compressed, vectorized, or binaryized from the text information are referred to as information that represents text. The broadcaster terminal 10 sends the first prompt P1 to the server 110.

[0033] Next, the server 110 inputs the first prompt P1 to the first generating AI. Accordingly, the server 110 generates the first seed value S1 when generating the first image data D1. Then, based on the first seed value S1 and the first prompt P1, the first generating AI generates the first image data D1 shown in Figure 2. The first image data D1 shows an image of a humanoid character with the face of a Shiba Inu.

[0034] The seed value is the initial value of the random number generator included in the first generation AI. If the seed value is different, even if the same prompt is input to the first generation AI, the first generation AI will generate image data of different characters. In other words, if the seed value is not changed, it is possible to keep the characters in the image data generated by the first generation AI unchanged. In this embodiment, the first generation AI generates "12345" as the first seed value S1. For the sake of explanation, the seed value is exemplified using only numbers, but the actual first seed value S1 and the second to fourth seed values ​​S2 to S4 described later may include letters and symbols. In this embodiment, the first seed value S1 is text information written as "12345". However, the first seed value S1 may be information that has been compressed, vectorized, or binaryized from text information. The second to fourth seed values ​​S2 to S4 described later may also be text information, or information that has been compressed, vectorized, or binaryized from text information, similar to the first seed value S1.

[0035] Next, the server 110 sends the first image data D1 and the first seed value S1 to the broadcaster terminal 10. The broadcaster terminal 10 displays the image of the character indicated by the first image data D1. This allows the broadcaster to confirm the image of the character. If the broadcaster likes the image of the character, they perform a character selection process by operating the broadcaster terminal 10. Once the character selection process is complete, the server 110 stores the first seed value S1 and the first prompt P1.

[0036] On the other hand, if the broadcaster does not like the character image, they have the broadcaster terminal 10 send the first prompt P1 shown in Figure 2 to the server 110 again. The server 110 generates the first image data D1 again based on a first seed value S1 and the first prompt P1 that is different from "12345". The server 110 then sends the first image data D1 and the first seed value S1 to the broadcaster terminal 10. In this way, the broadcaster can have the first generation AI repeat the generation of the first image data D1 until they obtain an image of the character they like.

[0037] (2) Distribution of video data The operation of distributing video data is an operation performed by System 1 when distributing a video of the streamer's character. Specifically, the streamer records a video of themselves by operating the streamer terminal 10. As a result, as shown in Figure 3, the streamer terminal 10 generates person video data D0, which represents the streamer's (person's) video.

[0038] Next, the broadcaster terminal 10 generates motion information A1 that shows the movements of the broadcaster (person) contained in the person video data D0. More specifically, as shown in Figure 3, the broadcaster terminal 10 extracts and analyzes the skeleton of the broadcaster contained in the person video data D0. Then, the broadcaster terminal 10 generates motion information A1 that shows the temporal changes in the coordinates of the broadcaster's skeleton. Such processing is realized by motion capture such as OpenPose. Then, the broadcaster terminal 10 transmits the motion information A1 to the server 110.

[0039] Server 110 stores the second prompt P2. The second prompt P2 is text information that reads, "Please move the character generated by the first prompt and the first seed value according to the operation information." Server 110 also stores the first prompt P1 and the first seed value S1 during the character setting operation. Then, as shown in Figure 4, Server 110 transmits the operation information A1, the first prompt P1, the second prompt P2, and the first seed value S1 to viewer terminals 210-1 to 210-N.

[0040] Each of the viewer terminals 210-1 to 210-N (one or more second computers) is equipped with a second generation AI. As shown in Figure 4, each of the viewer terminals 210-1 to 210-N causes the second generation AI to generate video data D00 based on operation information A1, first prompt P1, second prompt P2, and first seed value S1. As described above, the second prompt P2 states, "Please move the character generated by the first prompt and the first seed value according to the operation information." Therefore, in the video shown by video data D00, the character generated by the first prompt P1 and the first seed value S1 performs the operation indicated by operation information A1.

[0041] Incidentally, the characters contained in video data D00 are substantially identical or similar to the characters contained in first image data D1. However, the characters contained in video data D00 do not have to be exactly identical to the characters contained in first image data D1, or they may be similar. This is because even if the same prompt and the same seed value are input to the first generation AI and the second generation AI, the characters contained in the image data generated by the first generation AI and the characters contained in the image data generated by the second generation AI may not be an exact match.

[0042] Finally, each of the viewer terminals 210-1 to 210-N displays a video based on the video data D00, as shown in Figure 5. This allows each of the first to the Nth viewer to watch a video of the broadcaster's character using viewer terminals 210-1 to 210-N.

[0043] [Structure of the broadcaster's device 10] The structure of the broadcaster terminal 10 will be explained with reference to the diagram. Figure 6 is a block diagram of the broadcaster terminal 10.

[0044] As shown in Figure 6, the broadcaster terminal 10 includes a control unit 12, a storage unit 14, a network interface 16, a camera 17, a graphics processing unit 18, a display 20, an audio processing unit 22, a speaker 24, and an operation unit 26.

[0045] The memory unit 14 stores programs and data. The memory unit 14 is, for example, a combination of ROM (Read Only Memory), RAM (Random Access Memory), and storage (for example, flash memory or hard disk).

[0046] The program includes, for example, the following: • OS (Operating System) programs • Programs for applications that perform information processing (e.g., web browsers, or target applications described later)

[0047] The data includes, for example, the following: • Databases referenced in information processing • Data obtained by performing information processing (i.e., the results of performing information processing)

[0048] The control unit 12 implements the functions of the broadcaster terminal 10 by executing the program stored in the memory unit 14. The control unit 12 is, for example, at least one of the following: ·CPU(Central Processing Unit) ·GPU(Graphic Processing Unit) ·ASIC(Application Specific Integrated Circuit) ·FPGA(Field Programmable Gate Array)

[0049] The control unit 12 includes a data generation unit 30, an information acquisition unit 32, a communication control unit 34, and a display control unit 36 ​​as functional blocks.

[0050] The network interface 16 controls communication between the broadcaster terminal 10 and an external device. The external device is a server 110.

[0051] Camera 17 captures the surroundings of the broadcaster terminal 10 and generates image data. In this embodiment, camera 17 captures the broadcaster.

[0052] The graphics processing unit 18 displays an image on the display 20 based on the image data generated by the control unit 12. The display 20 is either a liquid crystal display or an organic EL (Electro-Luminescence) display.

[0053] The audio processing unit 22 causes the speaker 24 to output sound based on the audio data generated by the control unit 12.

[0054] The operation unit 26 generates an operation signal based on the operator's operation and outputs the operation signal to the control unit 12.

[0055] [Structure of Server 110] Next, the structure of server 110 will be explained with reference to the diagram. Figure 7 is a block diagram of server 110.

[0056] Server 110 can communicate with the broadcaster terminal 10 and the viewer terminals 210-1 to 210-N. As shown in Figure 7, server 110 includes a control unit 112, a storage unit 114, and a network interface 116.

[0057] The memory unit 114 stores programs and data. The memory unit 114 is, for example, a combination of ROM (Read Only Memory), RAM (Random Access Memory), and storage (for example, flash memory or hard disk). The program includes, for example, the following: • OS (Operating System) programs • Programs for applications that perform information processing (e.g., web browsers, or target applications described later)

[0058] The data includes, for example, the following: • Databases referenced in information processing • Data obtained by performing information processing (i.e., the results of performing information processing)

[0059] The control unit 112 implements the functions of the server 110 by executing the program stored in the storage unit 114. The control unit 112 is, for example, at least one of the following: ·CPU(Central Processing Unit) ·GPU(Graphic Processing Unit) ·ASIC(Application Specific Integrated Circuit) ·FPGA(Field Programmable Gate Array)

[0060] The control unit 112 includes, as functional blocks, an information acquisition unit 120, a communication control unit 122, a prompt input unit 124, a seed value acquisition unit 126, a seed value identification unit 128, and a prompt acquisition unit 130.

[0061] The network interface 116 controls communication between the server 110 and external devices. The external devices are the broadcaster terminal 10 and the viewer terminals 210-1 to 210-N.

[0062] [Structure of viewer terminals 210-1 to 210-N] Next, the structure of viewer terminals 210-1 to 210-N will be explained with reference to the diagrams. Figure 8 is a block diagram of viewer terminal 210-1.

[0063] As shown in Figure 8, the viewer terminal 210-1 includes a control unit 212, a storage unit 214, a network interface 216, a camera 217, a graphics processing unit 218, a display 220, an audio processing unit 222, a speaker 224, and an operation unit 226.

[0064] The memory unit 214 stores programs and data. The memory unit 214 is, for example, a combination of ROM (Read Only Memory), RAM (Random Access Memory), and storage (for example, flash memory or hard disk).

[0065] The program includes, for example, the following: • OS (Operating System) programs • Programs for applications that perform information processing (e.g., web browsers, or target applications described later)

[0066] The data includes, for example, the following: • Databases referenced in information processing • Data obtained by performing information processing (i.e., the results of performing information processing)

[0067] The control unit 212 implements the functions of the viewer terminal 210-1 by executing the program stored in the memory unit 214. The control unit 212 is, for example, at least one of the following: ·CPU(Central Processing Unit) ·GPU(Graphic Processing Unit) ·ASIC(Application Specific Integrated Circuit) ·FPGA(Field Programmable Gate Array)

[0068] The control unit 212 includes a data generation unit 230, an information acquisition unit 234, and a display control unit 236 as functional blocks.

[0069] The network interface 216 controls communication between the viewer terminal 210-1 and an external device. The external device is the server 110.

[0070] Camera 217 captures the surroundings of the viewer terminal 210-1 and generates image data.

[0071] The graphics processing unit 218 displays an image on the display 220 based on the image data generated by the control unit 212. The display 220 is either a liquid crystal display or an organic EL (Electro-Luminescence) display.

[0072] The audio processing unit 222 causes the speaker 224 to output sound based on the audio data generated by the control unit 212.

[0073] The operation unit 226 generates an operation signal based on the operation of the first viewer and outputs the operation signal to the control unit 212.

[0074] Note that the structure of viewer terminals 210-2 to 210-N is the same as that of viewer terminal 210-1, so we will omit the explanation.

[0075] [System 1 Operation] Next, the operation of System 1 will be explained with reference to the diagrams. Figures 9 and 10 are flowcharts showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1.

[0076] The control unit 12 of the broadcaster terminal 10 reads the programs stored in the storage unit 14, causing these programs to execute the operations described below in the broadcaster terminal 10. The programs then make the broadcaster terminal 10 function as a data generation unit 30, an information acquisition unit 32, a communication control unit 34, and a display control unit 36.

[0077] The control unit 112 of the server 110 reads the programs stored in the storage unit 114, causing these programs to perform the operations described below on the server 110. The programs then make the server 110 function as an information acquisition unit 120, a communication control unit 122, a prompt input unit 124, a seed value acquisition unit 126, a seed value identification unit 128, and a prompt acquisition unit 130.

[0078] The control unit 212 of the viewer terminal 210-1 reads the programs stored in the storage unit 214, causing these programs to execute the operations described below on the viewer terminal 210-1. The programs then make the viewer terminal 210-1 function as a data generation unit 230, an information acquisition unit 234, and a display control unit 236.

[0079] (1) Character determination First, the broadcaster inputs the first prompt P1 by operating the control unit 26 of the broadcaster terminal 10. As shown in Figure 2, the first prompt P1 is generation information that is input to the first generation AI when generating the first image data D1. The first prompt P1 is information that represents text. The first prompt P1 is text information that says, "Please create a human-shaped character with the face of a Shiba Inu." In this way, the first prompt P1 describes the characteristics of the character in the first image represented by the first image data D1. Accordingly, the control unit 12 of the broadcaster terminal 10 acquires the first prompt P1 (step S1).

[0080] Next, the control unit 12 of the broadcaster terminal 10 sends the first prompt P1 to the server 110 via the network interface 16 (step S2). In response, the network interface 116 of the server 110 receives the first prompt P1 and outputs the first prompt P1 to the control unit 112. As a result, the control unit 112 (information acquisition unit 120) of the server 110 (one or more first computers) acquires the first prompt P1 (generated information) (step S11).

[0081] Next, the control unit 112 (prompt input unit 124) of the server 110 (one or more first computers) inputs a first prompt P1 to the first generation AI in order to generate the first image data D1 (step S12). Accordingly, the control unit 112 (seed value acquisition unit 126) of the server 110 (one or more first computers) generates a first seed value S1, which is generation information generated by the first generation AI when generating the first image data D1 (step S13). As a result, the control unit 112 (seed value acquisition unit 126) of the server 110 (one or more first computers) acquires the first seed value S1 generated by the first generation AI in response to the input of the first prompt P1. The first seed value S1 is information that represents text. In this embodiment, the first seed value S1 is "12345".

[0082] Furthermore, the control unit 112 of server 110 (one or more first computers) causes the first generating AI to generate first image data D1 based on the first seed value S1 and the first prompt P1 (step S14). Specifically, the first generating AI analyzes the first prompt P1. In this process, the first generating AI uses natural language processing (NLP) techniques to grasp the content of the text. Next, the first generating AI maps the text prompt to a learned latent space. This latent space represents patterns and features learned from the training dataset. Next, the first generating AI generates first image data D1 from the latent space based on the first prompt P1 and the first seed value S1. In this process, the first generating AI utilizes the features and patterns it has learned to create the first image data D1 that is most suitable for the first prompt P1. The resulting first image data D1, as shown in Figure 2, represents an image of a humanoid character with the face of a Shiba Inu.

[0083] Next, the control unit 112 (communication control unit 122) of the server 110 transmits the first image data D1 and the first seed value S1 to the distributor terminal 10 via the network interface 116 (step S15). Accordingly, the network interface 16 of the distributor terminal 10 receives the first image data D1 and the first seed value S1 and outputs the first image data D1 and the first seed value S1 to the control unit 12. As a result, the control unit 12 of the distributor terminal 10 acquires the first image data D1 and the first seed value S1 (step S3).

[0084] Next, the control unit 12 (display control unit 36) of the broadcaster terminal 10 displays the character image (first image) indicated by the first image data D1 on the display 20 (step S4). Then, the control unit 12 of the broadcaster terminal 10 determines whether or not to select the first image data D1 (step S5). In step S5, the broadcaster inputs whether or not to select a character by operating the operation unit 26 of the broadcaster terminal 10. If the broadcaster inputs whether or not to select a character, the control unit 12 of the broadcaster terminal 10 determines that the first image data D1 is selected. In this case, the process proceeds to step S6. If the broadcaster does not input whether or not to select a character, the control unit 12 of the broadcaster terminal 10 determines that the first image data D1 is not selected. In this case, the process returns to step S1.

[0085] When determining the first image data D1, the control unit 12 of the distributor terminal 10 sends a determination notification to the server 110 via the network interface 16 indicating that the first image data D1 has been determined (step S6). Accordingly, the network interface 116 of the server 110 receives the determination notification and outputs it to the control unit 112. As a result, the control unit 112 of the server 110 acquires the determination notification (step S16).

[0086] The control unit 112 of the server 110 stores the first prompt P1 and the first seed value S1 obtained in steps S11 and S13 in the storage unit 114 (step S17). This completes the character determination operation.

[0087] (2) Distribution of video data First, the broadcaster initiates the process of starting video streaming by operating the control unit 26 of the broadcaster terminal 10. In response, the control unit 12 of the broadcaster terminal 10 sends a streaming start notification to the server 110 via the network interface 16, indicating that video streaming has begun (step S101). In response, the network interface 116 of the server 110 receives the streaming start notification and outputs it to the control unit 112. As a result, the control unit 112 of the server 110 acquires the streaming start notification (step S201).

[0088] Here, the storage unit 114 of the server 110 stores the first prompt P1, the second prompt P2, and the first seed value S1. The second prompt P2 is a prompt to cause the second generation AI to generate video data D00, which shows a video of a character (object) linked to the actions of the broadcaster (person), based on the action information A1. The second prompt P2 is information that represents text. In this embodiment, the second prompt P2 is text information that says, "Please move the character generated by the first prompt and the first seed value according to the action information." The control unit 112 (information acquisition unit 120, prompt acquisition unit 130) of the server 110 (one or more first computers and fourth computers) acquires the first prompt P1 (generation information), the second prompt P2, and the first seed value S1 (second seed value related to generation information and first seed value) by reading the first prompt P1, the second prompt P2, and the first seed value S1 from the storage unit 114 (step S202).

[0089] Next, the control unit 112 (communication control unit 122) of server 110 (one or more first computers) transmits the first prompt P1 (generated information), the second prompt P2, and the first seed value S1 (a second seed value related to the generated information and the first seed value) to the viewer terminal 210-1 (one or more second computers) via the network interface 116 (step S203). Accordingly, the network interface 216 of viewer terminal 210-1 receives the first prompt P1, the second prompt P2, and the first seed value S1, and outputs the first prompt P1, the second prompt P2, and the first seed value S1 to the control unit 212. As a result, the control unit 112 (information acquisition unit 234) of viewer terminal 210-1 (one or more second computers) acquires the first prompt P1 (generated information), the second prompt P2, and the first seed value S1 (a second seed value related to the generated information and the first seed value) (step S301).

[0090] Next, the broadcaster records their own video by operating the control unit 26 of the broadcaster terminal 10. Specifically, the camera 17 records the broadcaster. As a result, as shown in Figure 3, the control unit 12 (data generation unit 30) of the broadcaster terminal 10 (third computer) generates person video data D0 representing the broadcaster's video (step S102).

[0091] Next, the control unit 12 (information acquisition unit 32) of the broadcaster terminal 10 (third computer) generates (acquires) motion information A1 that shows the movements of the person included in the person video data D0 (step S103). More specifically, as shown in Figure 3, the control unit 12 of the broadcaster terminal 10 extracts and analyzes the skeleton of the broadcaster included in the person video data D0. Then, the control unit 12 of the broadcaster terminal 10 generates motion information A1 that shows the temporal changes in the coordinates of the broadcaster's skeleton. Then, the control unit 12 (communication control unit 34) of the broadcaster terminal 10 (third computer) transmits the motion information A1 to the server 110 (fourth computer) via the network interface 16 (step S104). Accordingly, the network interface 116 of the server 110 receives the motion information A1 and outputs the motion information A1 to the control unit 112. As a result, the control unit 112 (information acquisition unit 120) of the server 110 (fourth computer) acquires the motion information A1 (step S204).

[0092] Next, the control unit 112 (communication control unit 122) of server 110 (one or more first computers) transmits operation information A1 to viewer terminal 210-1 (one or more second computers) via network interface 116 (step S205). Accordingly, the network interface 216 of viewer terminal 210-1 receives operation information A1 and outputs operation information A1 to control unit 212. As a result, control unit 212 of viewer terminal 210-1 acquires operation information A1 (step S302).

[0093] Next, the control unit 212 (data generation unit 230) of the viewer terminal 210-1 (one or more second computers) causes the second generation AI to generate video data D00 (second image data) based on the operation information A1, the first prompt P1 (generation information), the second prompt P2, and the first seed value S1 (third seed value / generation information related to the second seed value), as shown in Figure 4 (step S303). Video data D00 is data in which multiple second images indicated by the second image data D2 are arranged in chronological order. The second image data D2 is data indicating the second image generated by the second generation AI based on the first prompt P1 and the first seed value S1. The second prompt P2 is text information that reads, "Please move the character generated by the first prompt and the first seed value according to the operation information." Therefore, in the video data D00, the character generated by the first prompt P1 and the first seed value S1 performs the action indicated by the action information A1.

[0094] Next, the control unit 212 (display control unit 236) of the viewer terminal 210-1 displays the video (image) on the display 220 based on the video data D00 (image data), as shown in Figure 5 (step S304). As a result, the first viewer can watch the video of the broadcaster's character using the viewer terminal 210-1.

[0095] Next, the control unit 12 of the broadcaster terminal 10 determines whether or not to terminate the video distribution (step S105). If the video distribution is to be terminated, the process proceeds to step S106. If the video distribution is not to be terminated, the process returns to step S102.

[0096] When video distribution is to be terminated, the control unit 12 of the distributor terminal 10 sends a distribution termination notification to the server 110 via the network interface 216 (step S106). Accordingly, the network interface 116 of the server 110 receives the distribution termination notification and outputs the distribution termination notification to the control unit 112.

[0097] The control unit 112 of the server 110 determines whether or not it has received a delivery completion notification (step S206). If the control unit 112 has received a delivery completion notification, the process proceeds to step S207. If the control unit 112 has not received a delivery completion notification, the process returns to step S204.

[0098] When the control unit 112 receives a delivery termination notification, the control unit 112 of the server 110 sends the delivery termination notification to the viewer terminal 210-1 via the network interface 116 (step S207). Accordingly, the network interface 216 of the viewer terminal 210-1 receives the delivery termination notification and outputs it to the control unit 212.

[0099] The control unit 212 of the viewer terminal 210-1 determines whether or not it has received a distribution end notification (step S305). If the control unit 212 has received a distribution end notification, this process ends. If the control unit 212 has not received a distribution end notification, this process returns to step S302.

[0100] Although the flowchart in Figure 10 only describes the operation of the control unit 212 of viewer terminal 210-1, the control units 212 of viewer terminals 210-2 to 210-N perform the same operation as the control unit 212 of viewer terminal 210-1.

[0101] [effect] According to System 1, the communication load can be reduced. More specifically, the control unit 112 of the server 110 obtains a first prompt P1 to be input to the first generation AI when generating the first image data D1 and / or a first seed value S1 generated by the first generation AI when generating the first image data D1. In this embodiment, the control unit 112 of the server 110 obtains a first prompt P1 to be input to the first generation AI when generating the first image data D1 and a first seed value S1 generated by the first generation AI when generating the first image data D1. Then, the control unit 112 of the server 110 transmits the first prompt P1 and / or the first seed value S1 to the viewer terminal 210-1. In this embodiment, the control unit 112 of the server 110 transmits the first prompt P1 and the first seed value S1 to the viewer terminal 210-1. Here, the first prompt P1 and the first seed value S1 are information representing text with a relatively small amount of data. Therefore, the first image data D1, which has a relatively large amount of data, does not need to be transmitted from the server 110 to the viewer terminal 210-1. As a result, the communication load can be reduced according to system 1.

[0102] In System 1, the first image data D1 does not need to be transmitted from Server 110 to Viewer Terminal 210-1. Therefore, Server 110 does not need to perform video data encoding, and Viewer Terminal 210-1 does not need to perform video data decoding. As a result, the processing load on Server 110 is reduced.

[0103] In System 1, the first prompt P1 and the first seed value S1 are sent from Server 110 to Viewer Terminal 210-1, enabling Viewer Terminal 210-1 to generate the first image data D1. Therefore, Viewer Terminal 210-1 does not need to download the character image data included in the first image data D1 from Server 110.

[0104] In System 1, the control unit 212 of the viewer terminal 210-1 can generate video data D00. More specifically, the control unit 112 of the server 110 acquires motion information A1 indicating a person's movements and transmits the motion information A1 to the viewer terminal 210-1 via the network interface 116. Based on the first seed value S1 and the motion information A1, the control unit 212 of the viewer terminal 210-1 causes the second generation AI to generate video data D00, in which multiple second images indicated by the second image data D2 are arranged in chronological order. Thus, in System 1, the control unit 212 of the viewer terminal 210-1 can generate video data D00 without the first image data D1 being transmitted from the server 110 to the viewer terminal 210-1.

[0105] In System 1, the communication load can be reduced for the following reasons. More specifically, the control unit 12 of the broadcaster terminal 10 generates person video data D0 that shows the broadcaster's video, and generates action information A1 that shows the broadcaster's actions included in the person video data D0. The control unit 12 of the broadcaster terminal 10 then transmits the action information A1 to the server 110 via the network interface 116. As a result, the control unit 112 of the server 110 acquires the action information A1. Therefore, the person video data D0, which has a relatively large amount of data, is not transmitted from the broadcaster terminal 10 to the server 110, and the control unit 112 of the server 110 can acquire the action information A1. As a result, the communication load can be reduced in System 1.

[0106] In System 1, the communication load can be reduced for the following reasons. More specifically, the control unit 112 of the server 110 transmits a second prompt P2 to the viewer terminal 210-1 via the network interface 116. The second prompt P2 is a prompt to cause the second generation AI to generate video data D00, which shows a video of a character linked to the broadcaster's actions, based on the action information A1. Then, the control unit 212 of the viewer terminal 210-1 causes the second generation AI to generate the video data D00 based on the first seed value S1, the action information A1, and the second prompt P2. In this way, because the second prompt P2, which has a relatively small amount of data, is transmitted from the server 110 to the viewer terminal 210-1, the control unit 212 of the viewer terminal 210-1 can generate the video data D00. Therefore, the communication load can be reduced in System 1.

[0107] (First variation) The following describes system 1a related to the first modified example. Figure 11 is a seed value conversion table.

[0108] System 1a differs from System 1 in that multiple types of second generation AI models exist for each of the multiple viewer terminals 210-1 to 210-N, and multiple versions of the second generation AI also exist. The generation AI model indicates the type of generation AI. Examples of generation AI models include GPT-4o, Sora, Gemini, Bard, Stable Diffusion, and midjourney. If the second generation AI model is different, the seed value of the second generation AI corresponding to the seed value generated by the first generation AI will be a different value.

[0109] The Generator AI version number is updated whenever the Generator AI model is updated. When the version of the second Generator AI changes, the seed value of the second Generator AI corresponding to the seed value generated by the first Generator AI changes.

[0110] Therefore, the storage unit 114 of the server 110 stores correspondence information that shows the correspondence between the seed value in the first generation AI and the seed value in the second generation AI. More specifically, the storage unit 114 of the server 110 stores correspondence information that shows the correspondence between the seed value in the first generation AI and the seed values ​​in the second generation AI of multiple types of models, and also shows the correspondence between the seed value in the first generation AI and the seed values ​​in the second generation AI of multiple versions.

[0111] The seed value conversion table shown in Figure 11 is an example of correspondence information. The second seed value S2 is a value related to the first seed value S1. Specifically, the second seed value S2 in the second generation AI corresponds to the first seed value S1 in the first generation AI. Therefore, the second image generated by the second generation AI based on the second seed value S2 and the first prompt P1 will be substantially identical or similar to the first image generated by the first generation AI based on the first seed value S1 and the first prompt P1. In the seed value conversion table shown in Figure 11, the second seed value S2 is described for each model of the second generation AI and for each version of the second generation AI.

[0112] Next, the operation of system 1a will be explained with reference to Figure 12. Figure 12 is a flowchart showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1. Note that the character determination operation in system 1a is the same as the character determination operation in system 1, so the explanation will be omitted.

[0113] The following describes the operation of video data distribution in System 1a. The operation of video data distribution in System 1a differs from the operation of video data distribution in System 1 in that steps S210-S212, S311, and S312 are added. Therefore, the explanation will focus on these differences.

[0114] When step S201 is completed, the control unit 112 of the server 110 sends a distribution start notification to the viewer terminal 210-1 via the network interface 116 (step S210). Accordingly, the network interface 216 of the viewer terminal 210-1 receives the distribution start notification and outputs the distribution start notification to the control unit 212. As a result, the control unit 212 of the viewer terminal 210-1 acquires the distribution start notification (step S311).

[0115] The memory unit 214 of the viewer terminal 210-1 stores model information M and version information V. Model information M is information indicating the model of the second-generation AI. Version information V is information indicating the version of the second-generation AI. The control unit 212 of the viewer terminal 210-1 reads the model information M and version information V and transmits the model information M and version information V to the server 110 via the network interface 216 (step S311). Accordingly, the network interface 116 of the server 110 receives the model information M and version information V and outputs the model information M and version information V to the control unit 112. As a result, the control unit 112 (information acquisition unit 120) of the server 110 (one or more first computers and fourth computers) acquires the model information M and version information V from the viewer terminal 210-1 (one or more second computers) (step S211).

[0116] Next, the control unit 112 (seed value identification unit 128) of the server 110 (one or more first and fourth computers) identifies the second seed value S2 corresponding to the first seed value S1 generated by the first generating AI in step S13, based on the model information M, version information V, and seed value conversion table (correspondence relationship information) (step S212). For example, if the model of the second generating AI is generating AI-X, the version of the second generating AI is ver.2, and the first seed value S1 is "12345", the control unit 112 of the server 110 identifies the second seed value S2 as "12945". After this, the process proceeds to step S202.

[0117] Note that the processing from step S202 onward in system 1a is the same as in system 1, except that the second seed value S2 is used instead of the first seed value S1, so the explanation is omitted.

[0118] System 1a produces the same effect as System 1.

[0119] Furthermore, in system 1a, even if the model or version of the second generation AI changes, the control unit 212 of the viewer terminal 210-1 can appropriately generate the video data D00.

[0120] (Second variation) The following describes system 1b related to the second modified example. Figure 13 is a seed value conversion table.

[0121] System 1b differs from System 1a in that the control unit 212 of the viewer terminal 210-1 identifies the third seed value S3 corresponding to the first seed value S1. Therefore, the storage unit 214 of the viewer terminal 210-1 stores the seed value conversion table shown in Figure 13. The third seed value S3 is a value related to the first seed value S1 (second seed value). Specifically, the third seed value S3 in the second generation AI corresponds to the first seed value S1 in the first generation AI. Therefore, the second image generated by the second generation AI based on the third seed value S3 and the first prompt P1 will be substantially identical or similar to the first image generated by the first generation AI based on the first seed value S1 and the first prompt P1. In the seed value conversion table shown in Figure 13, the third seed value S3 is described for each model of the second generation AI and for each version of the second generation AI.

[0122] Next, the operation of system 1b will be explained with reference to Figure 14. Figure 14 is a flowchart showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1. Note that the character determination operation in system 1b is the same as the character determination operation in system 1, so the explanation will be omitted.

[0123] The following describes the operation of video data distribution in System 1b. The operation of video data distribution in System 1b differs from the operation of video data distribution in System 1 in that step S321 is added. Therefore, the explanation will focus on this difference.

[0124] When step S301 is completed, the control unit 112 of the viewer terminal 210-1 identifies the third seed value S3 corresponding to the first seed value S1 (the second seed value associated with the first seed value) obtained in step S301, based on the model information M, version information V, and seed value conversion table (correspondence relationship information) (step S321). For example, if the model of the second generation AI is generation AI·X, the version of the second generation AI is ver.2, and the first seed value S1 is "12345", the control unit 212 of the viewer terminal 210-1 identifies the second seed value S2 as "12945". After this, the process proceeds to step S302.

[0125] Note that the processing from step S302 onward in system 1b is the same as in system 1, except that the third seed value S3 is used instead of the first seed value S1, so the explanation is omitted.

[0126] System 1b produces the same effect as System 1.

[0127] Furthermore, in system 1b, even if the model or version of the second generation AI changes, the control unit 112 of the viewer terminal 210-1 can appropriately generate the video data D00.

[0128] (Third variation) The following describes system 1c related to the third modified example.

[0129] System 1c differs from System 1 in that the memory unit 114 of the viewer terminal 210-1 (one or more second computers) stores the first prompt P1 and the second prompt P2. Therefore, the first prompt P1 and the second prompt P2 are not transmitted from the server 110 to the viewer terminal 210-1. Figure 15 is a flowchart showing the operation of the control unit 12 of the broadcaster terminal 10, the control unit 112 of the server 110, and the control unit 212 of the viewer terminal 210-1. Note that the character determination operation in System 1c is the same as the character determination operation in System 1, so the explanation is omitted.

[0130] The operation of video data distribution in System 1c is described below. The operation of video data distribution in System 1c differs from the operation of video data distribution in System 1 in that steps S202, S203, and S301 are replaced by steps S252, S253, and S351, and step S352 is added. Therefore, the explanation will focus on these differences.

[0131] When step S201 is completed, the control unit 112 of the server 110 acquires the first seed value S1 by reading it from the storage unit 114 (step S252). Then, the control unit 112 of the server 110 transmits the first seed value S1 to the viewer terminal 210-1 via the network interface 116 (step S253). Accordingly, the network interface 216 of the viewer terminal 210-1 receives the first seed value S1 and outputs the first seed value S1 to the control unit 212. As a result, the control unit 212 of the viewer terminal 210-1 acquires the first seed value S1 (step S351).

[0132] Next, the control unit 212 of the viewer terminal 210-1 obtains the first prompt P1 and the second prompt P2 by reading them from the storage unit 114 (step S352). After this, the process proceeds to step S302.

[0133] Note that the processing from step S302 onward in system 1c is the same as in system 1, so the explanation is omitted. System 1c has the same effect as system 1.

[0134] (Fourth variation) The following describes system 1d related to the fourth modified example. Figure 16 is an image displayed on the viewer terminal 210-1.

[0135] System 1d differs from System 1 in that each of the first to the Nth viewers can perform viewer actions. More specifically, the first to the Nth viewers can perform viewer actions while watching the video. Viewer actions are actions taken by the first to the Nth viewers on the video during its distribution. Specific examples of viewer actions are explained below with reference to Figure 16.

[0136] The image shown in Figure 16 includes a heart-shaped button, a gift-shaped button, and a text input box. The heart-shaped button is called the "Like" button. The gift-shaped button is called the "Gift" button. The text input box is called the "Input Box." Viewer actions include the action of the first to the Nth viewer pressing the "Like" button, the action of the first to the Nth viewer pressing the "Gift" button, and the action of the first to the Nth viewer sending a comment. When any of the first to the Nth viewer presses the "Like" button, a heart-shaped icon is displayed on viewer terminals 210-1 to 210-N, as shown in the image in Figure 16. When any of the first to the Nth viewer presses the "Gift" button, a "Gift" icon is displayed on viewer terminals 210-1 to 210-N, as shown in the image in Figure 16, and a gift is sent to the broadcaster by any of the first to the Nth viewer. The gift is a hat. As shown in Figure 16, the character image and the hat image are combined so that the character wears the hat. Furthermore, when either the first viewer or the Nth viewer enters a comment into the input box and presses the submit button, the comment is displayed on viewer terminals 210-1 to 210-N, as shown in the image in Figure 16.

[0137] The following describes the operation of the first viewer performing a viewer action in system 1c. The viewer action is the action of the first viewer pressing the gift button. This operation is performed in parallel with the operation of video data distribution. Figure 17 is a flowchart showing the operation of the control unit 212 of viewer terminal 210-1, the operation of the control unit 112 of server 110, and the operation of the control unit 212 of viewer terminal 210-2.

[0138] The first viewer performs a viewer action by operating the operation unit 26 of the viewer terminal 210-1. In this embodiment, the first viewer presses the gift button displayed on the display 20. As a result, the control unit 212 of the viewer terminal 210-1 acquires viewer action information AC (step S). The viewer action information AC indicates that the first viewer is giving a hat object to the broadcaster. The control unit 212 (communication control unit 122) of the viewer terminal 210-1 (fifth computer) then transmits the viewer action information AC, which is generated in response to the first viewer's operation, to the server 110 (fourth computer) via the network interface 216 (step S402). Accordingly, the network interface 116 of the server 110 receives the viewer action information AC and outputs the viewer action information AC to the control unit 112. As a result, the control unit 112 (information acquisition unit 120) of the server 110 (fourth computer) acquires the viewer action information AC (step S501).

[0139] Next, the control unit 112 (prompt acquisition unit 130) of server 110 (fourth computer) acquires a third prompt P3 that causes a change in the character (object) according to the viewer action information AC (step S502). More specifically, the viewer action information AC indicates that the first viewer will give the broadcaster a hat object. Therefore, the control unit 112 of server 110 generates a third prompt P3, which is text information that says, "Put a hat on the character's head." based on the viewer action information AC.

[0140] Next, the control unit 112 (communication control unit 122) of server 110 (fourth computer) transmits the third prompt P3 to the viewer terminal 210-2 (sixth computer) via the network interface 116 (step S503). In response, the network interface 216 of viewer terminal 210-2 receives the third prompt P3 and outputs the third prompt P3 to the control unit 212. As a result, the control unit 212 (information acquisition unit 234) of viewer terminal 210-2 (sixth computer) acquires the third prompt P3 (step S601).

[0141] Next, the control unit 212 (data generation unit 230) of the viewer terminal 210-2 (sixth computer) causes the second generation AI to generate video data D00, which represents a change in the character (object) corresponding to the viewer action information AC, based on the third prompt P3 (step S602). More specifically, in step S303, the control unit 212 of the viewer terminal 210-1 causes the second generation AI to generate video data D00 based on the operation information A1, the first prompt P1, the second prompt P2, and the first seed value S1, as shown in Figure 4. The control unit 212 of the viewer terminal 210-2 performs the processing of step S602 instead of step S303. Therefore, the control unit 212 of the viewer terminal 210-1 causes the second generation AI to generate video data D00 based on the operation information A1, the first prompt P1, the second prompt P2, the third prompt P3, and the first seed value S1, as shown in Figure 4. The second prompt P2 states, as mentioned above, "Please move the character generated by the first prompt and the first seed value according to the operation information." Furthermore, the third prompt P3 states, "Put a hat on the character's head." Therefore, in the video shown in video data D00, the character generated by the first prompt P1 and the first seed value S1 is wearing a hat and performing the operation indicated by operation information A1.

[0142] Next, the control unit 212 of the viewer terminal 210-2 displays the video (image) on the display 220 based on the video data D00 (image data), as shown in Figure 16 (step S603). As a result, the second viewer can watch the video of the character wearing a hat using the viewer terminal 210-1.

[0143] According to system 1d, server 110 generates a third prompt P3 based on viewer action information AC. Server 110 then sends the third prompt P3 to viewer terminal 210-2. This allows viewer terminal 210-2 to have the second generation AI generate video data D00 in which changes occur in the object according to the viewer action. Therefore, according to system 1d, the communication load can be reduced because the image data of the hat, which has a relatively large amount of data, is not transmitted.

[0144] (Fifth variation) The following describes system 1e according to the fifth modified example. Figure 18 is a block diagram of system 1e. Figure 19 is a flowchart showing the operation of the control unit 212 of the viewer terminal 210-1 and the control unit 112 of the server 110.

[0145] System 1d differs from System 1 in that it does not have a broadcaster terminal 10, as shown in Figure 18. That is, instead of live video from the broadcaster being streamed, pre-recorded video is streamed. The storage unit 114 of the server 110 stores the first prompt P1, the second prompt P2, the first seed value S1 (the second seed value related to the first seed value), and operation information A1.

[0146] The operation of video data distribution will be explained below with reference to Figure 19. The control unit 112 of the server 110 sends a distribution start notification to the viewer terminal 210-1 via the network interface 16 to indicate that video distribution has started (step S261). Accordingly, the network interface 216 of the viewer terminal 210-1 receives the distribution start notification and outputs the distribution start notification to the control unit 212. As a result, the control unit 212 of the viewer terminal 210-1 acquires the distribution start notification (step S361).

[0147] As described above, the storage unit 114 of the server 110 stores the first prompt P1, the second prompt P2, and the first seed value S1. The second prompt P2 is a prompt to cause the second generation AI to generate video data D00, which shows a video of an object linked to the actions of the broadcaster (person), based on the action information A1. The second prompt P2 is information that represents text. In this embodiment, the second prompt P2 is text information that says, "Please move the character generated by the first prompt and the first seed value according to the action of the action information." The control unit 112 (information acquisition unit 120, prompt acquisition unit 130) of the server 110 (one or more first computers and fourth computers) acquires the first prompt P1, the second prompt P2, and the first seed value S1 (the second seed value related to the first seed value) by reading the first prompt P1, the second prompt P2, and the first seed value S1 from the storage unit 114 (step S202).

[0148] Next, the control unit 112 (communication control unit 122) of server 110 (one or more first computers) transmits the first prompt P1 (generated information), the second prompt P2, and the first seed value S1 (the second seed value related to the first seed value) to the viewer terminal 210-1 via the network interface 116 (step S203). Accordingly, the network interface 216 of viewer terminal 210-1 receives the first prompt P1, the second prompt P2, and the first seed value S1, and outputs the first prompt P1, the second prompt P2, and the first seed value S1 to the control unit 212. As a result, the control unit 212 (information acquisition unit 234) of viewer terminal 210-1 acquires the first prompt P1 (generated information), the second prompt P2, and the first seed value S1 (the second seed value related to the first seed value) (step S301).

[0149] Next, the storage unit 114 of the server 110 stores the operation information A1 as described above. The control unit 112 of the server 110 acquires the operation information A1 by reading it from the storage unit 114 (step S204).

[0150] Next, the control unit 112 of the server 110 transmits operation information A1 to the viewer terminal 210-1 via the network interface 116 (step S205). Accordingly, the network interface 216 of the viewer terminal 210-1 receives the operation information A1 and outputs the operation information A1 to the control unit 212. As a result, the control unit 212 of the viewer terminal 210-1 acquires the operation information A1 (step S302). After this, the process proceeds to step S303.

[0151] Note that the processing from step S303 onwards in system 1e is the same as the processing from step S303 onwards in system 1, so the explanation will be omitted.

[0152] System 1e can achieve the same effect as System 1.

[0153] (supplement) The relationship between the first to third seed values ​​in this disclosure and the first to third seed values ​​in systems 1,1a to 1b will be explained below with reference to the drawings. Figure 20 is a table showing the relationship between the first to third seed values ​​in this disclosure and the first to third seed values ​​in systems 1,1a to 1b.

[0154] In this disclosure, one or more first computers obtain a first seed value and a second seed value associated with the first seed value. Then, one or more first computers transmit the second seed value to one or more second computers. Meanwhile, one or more second computers obtain the second seed value and generate second image data based on a third seed value associated with the second seed value.

[0155] Here, we will summarize and explain the relationship between the first seed value S1 to the third seed value S3 and the first to third seed value in each embodiment.

[0156] In System 1, Server 110 (one or more first computers) generates (acquires) a first seed value S1 and transmits it to Viewer Terminal 210-1 (one or more second computers). Viewer Terminal 210-1 (one or more second computers) receives (acquires) the first seed value S1 and generates second image data D2 based on the first seed value S1. Thus, in System 1, the second seed value is the same as the first seed value, and the third seed value is the same as the second seed value. This is because, in System 1, the model and version of the second generating AI are the same as the model and version of the first generating AI.

[0157] In system 1a, server 110 (one or more first computers) generates (acquires) a first seed value S1 and obtains a second seed value S2 associated with the first seed value S1. Then, server 110 (one or more first computers) transmits the second seed value S2 to viewer terminal 210-1 (one or more second computers). Viewer terminal 210-1 (one or more second computers) receives the second seed value S2 and generates second image data D2 based on the second seed value S2. Thus, in system 1a, the second seed value S2 in the second generation AI corresponds to the first seed value S1 in the first generation AI. The third seed value is the same as the second seed value. This is because, in system 1a, the model and version of the second generation AI are different from those of the first generation AI.

[0158] In system 1b, server 110 (one or more first computers) generates (acquires) a first seed value S1 and transmits the first seed value S1 to viewer terminal 210-1 (one or more second computers). Viewer terminal 210-1 (one or more second computers) receives the first seed value S1 and acquires a third seed value S3 associated with the first seed value S1. Viewer terminal 210-1 (one or more second computers) generates second image data D2 based on the third seed value S3. Thus, in system 1b, the third seed value S3 in the second generation AI corresponds to the first seed value S1 in the first generation AI. Furthermore, the second seed value is the same as the first seed value. This is because, in system 1, the model and version of the second generation AI are different from those of the first generation AI.

[0159] (Other embodiments) The various control means and processing procedures described in the above embodiments are examples and are not intended to limit the scope of the present invention, its applications, or its uses. The various control means and processing procedures can be modified as appropriate without altering the essence of the present invention.

[0160] Furthermore, the configurations of systems 1,1a to 1e may be combined in any way.

[0161] Note that the server 110 in systems 1,1a to 1e may include multiple computers.

[0162] In addition, in systems 1,1a to 1e, the generated information transmitted from server 110 to viewer terminal 210-1 does not have to be the first seed value S1 or the second seed value S2. The generated information can be any information generated by the first generation AI when the first image data D1 is generated, and it is sufficient if it contains information equivalent to the seed value.

[0163] In systems 1,1a to 1e, the first prompt P1 may include information corresponding to the first seed value S1 or the second seed value S2. In this case, the first seed value S1 or the second seed value S2 is not sent from server 110 to viewer terminal 210-1, and the first prompt P1 is sent from server 110 to viewer terminal 210-1.

[0164] In systems 1, 1a to 1e, an embodiment was described in which the server 110 is equipped with a first generation AI. However, if, for example, the broadcasting terminal 10 is equipped with a first generation AI, the functions of the present invention may be achieved using the first generation AI equipped in the broadcasting terminal 10 instead of the server 110. In this case, the server 110 may not be equipped with a first generation AI, and the broadcasting terminal 10 may be equipped with a first generation AI. Furthermore, a computer provided separately from systems 1, 1a to 1e may be equipped with a first generation AI.

[0165] The broadcaster terminal 10 sends the first prompt P1 to the server 110, and the server 110 sends the first prompt P1 to the viewer terminal 210-1. However, the broadcaster terminal 10 may send the source image data that will become the first image data D1 to the server 110 instead of the first prompt P1, and the server 110 may send the source image data that will become the first image data D1 along with the first seed value S1 to the viewer terminal 210-1 instead of the first prompt P1. Alternatively, the broadcaster terminal 10 may send the first prompt P1 and the source image data that will become the first image data D1 to the server 110, and the server 110 may send the first prompt P1 and the source image data that will become the first image data D1 to the viewer terminal 210-1. The viewer terminal 210-1 generates video data D00 based on the first prompt P1, the second prompt P2, and the source image data that will become the first image data D1.

[0166] The second prompt may also be a command statement instructing the second generation AI to generate video data D00 by continuously generating the second image data D2 a predetermined number of times or over a predetermined period of time.

[0167] Furthermore, the viewer terminal 210-1 may store the video data D00 generated by the second generation AI retrospectively to a predetermined time. This allows the viewer terminal 210-1 to rewind and play the video, or to fast-forward and play the video.

[0168] While systems 1,1a to 1e are applied to video streaming, they can also be applied to virtual spaces with multiple players, for example. In this case, there are multiple player terminals that serve as both the streamer terminal 10 and the viewer terminal 210-1.

[0169] When systems 1,1a to 1e are applied to the virtual space, the character will be displayed in three-dimensional space. In this case, the motion information A1 in the three-dimensional space includes not only the character's movement, but also the character's coordinates in the virtual space, the direction the character is facing in the virtual space, and the character's three-dimensional shape.

[0170] When systems 1,1a to 1e are applied to the virtual space, the first player terminal generates operation information A1 and sends operation information A1 to the server 110. The server 110 sends the first seed value S1, the first prompt P1, and operation information A1 to the second player terminal. This allows the second player terminal to generate and display the character of the first player terminal. The second player terminal also generates operation information A1 and sends it to the server 110. The server 110 sends the first seed value S1, the first prompt P1, and operation information A1 to the first player terminal. This allows the first player terminal to generate and display the character of the second player terminal.

[0171] In this specification, images include videos. That is, videos are generated by arranging multiple images in chronological order.

[0172] Furthermore, the characters included in the first image data D1 are not limited to animals. The characters may also be non-animal objects or anime characters.

[0173] The second generation AI may generate a second image data D2 instead of the video data D00. In this case, the server 110 distributes the image to viewer terminals 210-1 to 210-N.

[0174] In system 1c, the memory unit 114 of the viewer terminal 210-1 does not need to store the first prompt P1. In this case, the first prompt P1, along with the first seed value S1, is transmitted from the broadcaster terminal 10 to the server 110, and also from the server 110 to the viewer terminal 210-1.

[0175] The effects and advantages of this embodiment will also be achieved when these other embodiments are adopted. Furthermore, it is possible to combine this embodiment with other embodiments, and other embodiments with each other, as appropriate. [Explanation of Symbols]

[0176] 1,1a~1e: System 10: Streamer's device 12: Control Unit 14: Storage part 16: Network Interface 17: Camera 18: Graphics Processing Unit 20: Display 22: Audio Processing Unit 24: Speaker 26:Operation section 30: Data Generation Unit 32: Information acquisition department 34: Communication Control Unit 36: Display Control Unit 110: Server 112: Control Unit 114: Storage section 116: Network Interface 120: Information acquisition department 122: Communication Control Unit 124: Prompt input section 126: Seed value acquisition section 128: Seed value identification section 130: Prompt acquisition unit 210-1~210-N: Viewer terminals 212: Control Unit 214: Storage section 216: Network Interface 217: Camera 218: Graphics Processing Unit 220: Display 222: Audio Processing Unit 224: Speaker 226:Operation unit 230: Data Generation Unit 232: Communication Control Unit 234: Information acquisition department 236: Display Control Unit

Claims

1. A system comprising one or more first computers and one or more second computers, Each of the one or more second computers is equipped with a second generation AI, The one or more first computers mentioned above are: The system acquires generation information, which is information indicating the text to be input to the first generation AI when generating the first image data, and / or generation information, which is information indicating the text generated by the first generation AI when generating the first image data. The generated information is transmitted to the one or more second computers. The one or more second computers mentioned above are: The generated information is obtained, Based on the generated information, the second generation AI is made to generate the second image data. The first seed value is the initial value of the random number generator included in the first generation AI, The one or more first computers mentioned above are: To generate the first image data, a first prompt is input to the first generation AI, The first seed value generated by the first generation AI is obtained by inputting the first prompt. In the process of obtaining the generation information, the second seed value related to the first seed value is obtained as the generation information. In the process of transmitting the generated information, the second seed value is transmitted to the one or more second computers. The one or more second computers mentioned above are: In the process of generating the second image data, the second generation AI is made to generate the second image data based on a third seed value related to the second seed value. system.

2. The second seed value in the second generation AI corresponds to the first seed value in the first generation AI. The third seed value is the same as the second seed value. The system according to claim 1.

3. The one or more first computers mentioned above are: The system stores correspondence information showing the correspondence between the seed value in the first generation AI and the seed value in the second generation AI. The second seed value corresponding to the first seed value is identified based on the correspondence relationship information. The system according to claim 2.

4. There are multiple types of the aforementioned second generation AI models, The aforementioned correspondence information shows the correspondence between the seed value in the first generation AI and the seed value in the second generation AI of the multiple types of models. The one or more first computers mentioned above are: Obtain model information indicating the model of the second generated AI from the one or more second computers. Based on the model information and the correspondence information, the second seed value corresponding to the first seed value generated by the first generation AI is identified. The system according to claim 3.

5. There are multiple versions of the aforementioned second generation AI, The aforementioned correspondence information shows the correspondence between the seed value in the first generation AI and the seed value in multiple versions of the second generation AI. The one or more first computers mentioned above are: Version information indicating the version of the second generated AI is obtained from the one or more second computers. Based on the version information and the correspondence information, the second seed value corresponding to the first seed value generated by the first generation AI is identified. The system according to claim 3 or claim 4.

6. The one or more first computers mentioned above are: In the process of obtaining the generated information, the first prompt and the second seed value are obtained as the generated information. The one or more second computers mentioned above are: In the process of generating the second image data, the second generation AI is instructed to generate the second image data based on the third seed value and the first prompt. The system according to any one of claims 1 to 4.

7. The one or more second computers mentioned above are: The first prompt mentioned above is stored, In the process of generating the second image data, the second generation AI is instructed to generate the second image data based on the third seed value and the first prompt. The system according to any one of claims 1 to 4.

8. The one or more first computers mentioned above are: We obtain motion information that shows a person's movements, The first prompt is input to the first generating AI, The first seed value generated by the first generation AI is obtained by inputting the first prompt. In the process of obtaining the generation information, the second seed value related to the first seed value is obtained as the generation information. In the process of transmitting the generated information, the operation information and the second seed value are transmitted to the one or more second computers. The one or more second computers mentioned above are: In the process of generating the second image data, the second generation AI is instructed to generate video data in which a plurality of second images represented by the second image data are arranged in chronological order, based on the third seed value and the operation information. The system according to any one of claims 1 to 4.

9. The aforementioned one or more first computers include a third computer and a fourth computer, The fourth computer is equipped with the first generation AI, The aforementioned third computer, Generate video data showing the video of the aforementioned person, The motion information indicating the actions of the person included in the aforementioned person video data is acquired, The aforementioned operation information is transmitted to the fourth computer. The fourth computer is, In the process of acquiring the aforementioned operation information, the process of acquiring the aforementioned operation information is performed. The system according to claim 8.

10. The fourth computer is, A second prompt is obtained to cause the second generation AI to generate video data showing a video of an object linked to the movements of the person, based on the movement information. In the process of transmitting the generated information, the second prompt is sent to the one or more second computers. The one or more second computers mentioned above are: In the process of generating the second image data, the second generation AI is instructed to generate the video data based on the third seed value, the operation information, and the second prompt. The system according to claim 9.

11. The aforementioned one or more second computers further comprises a fifth computer and a sixth computer, The fifth computer is, The viewer action information generated in response to the viewer's actions is transmitted to the fourth computer. The fourth computer is, Obtain the aforementioned viewer Action information, A third prompt is obtained that causes a change in the object in accordance with the aforementioned viewer action information. In the process of transmitting the generated information, the third prompt is sent to the sixth computer. The sixth computer is, In the process of generating the second image data, based on the third prompt, the second generation AI is instructed to generate the video data in which changes occur in the object according to the viewer action information. The system according to claim 10.

12. A program to be executed on the second computer, The aforementioned second computer is equipped with a second generation AI, The aforementioned program, The system obtains generated information transmitted from the first computer, which is generated information indicating the text to be input to the first generation AI when generating the first image data, and / or generated information indicating the text generated by the first generation AI when generating the first image data. Based on the generated information, the second generation AI generates the second image data. Based on the aforementioned second image data, the second image is displayed. The operation is performed by the second computer. The first seed value is the initial value of the random number generator included in the first generation AI, The first computer is, To generate the first image data, a first prompt is input to the first generation AI, The first seed value generated by the first generation AI is obtained by inputting the first prompt. The second seed value related to the first seed value is obtained as the generation information, and the second seed value is transmitted to the second computer as the generation information. The aforementioned program, In the process of generating the second image data, the second generation AI is made to generate the second image data based on a third seed value related to the second seed value. The operation is to be performed by the second computer. program.

Images