System

Abstract

A system is provided.SOLUTION: A system comprising: means for uploading an image from a terminal of a user; means for analyzing the uploaded image to identify an object or motion in the image; means for generating a live commentary associated with a designated event based on an analysis result; means for transmitting the generated live commentary to the terminal of the user; and means for displaying the transmitted live commentary on the terminal of the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] The goal is to provide an automated system that can appropriately assess the image situation in real time and provide users with easy-to-understand, high-quality live broadcasts. Specifically, it is necessary to realize fast and accurate live broadcasts of sports games and everyday events without using professional announcers. To achieve this, a system that combines image analysis technology and generative AI is required. [Means for solving the problem]

[0005] To solve the above problem, we provide a system that includes the following elements: means for uploading images from a user's device, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating live commentary related to a specified event based on the analysis results, means for transmitting the generated live commentary to the user's device, and means for displaying the transmitted live commentary on the user's device. Furthermore, the system uses natural language generation technology to generate live commentary based on the analysis results, thereby achieving high-quality broadcasting, and performs highly accurate image analysis using an object detection model.

[0006] "Image uploading means" is a function including communication and an interface for sending image files from a user's terminal to the system.

[0007] "Image analysis means" means the use of image recognition algorithms and object detection models to identify objects and movements within uploaded images.

[0008] The "means for generating live commentary" is a function that uses natural language generation technology to generate live commentary in appropriate natural language based on the analysis results.

[0009] The "live comment sending means" is a function for sending the generated live comment to the user's terminal via communication.

[0010] The "live comment display means" is a function including an interface for displaying the live comment sent on the user's terminal in an easy-to-read format.

[0011] An "object detection model" is a machine learning model trained to detect specific objects (e.g., people, balls, etc.) in images.

[0012] "Natural language generation technology" refers to machine learning algorithms and rule-based systems that generate appropriate and easy-to-understand sentences based on the results of image analysis. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0014] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0015] First, the terms used in the following description will be explained.

[0016] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0017] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0018] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0019] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0020] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0021] [First embodiment]

[0022] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0023] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0024] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0025] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0026] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0027] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0028] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0029] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0030] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0031] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0032] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0033] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0034] This invention relates to an automated system that analyzes the status of an image in real time and generates and provides appropriate live commentary. The system mainly comprises a user terminal, a server, and a means for communicating between them. The details of the embodiment are described below.

[0035] 1. Image input from the user

[0036] A user uploads an image to the system using a device. The user selects an image file using a specific web interface or application and taps the upload button. This action causes the device to send the image file to the server as an HTTP request.

[0037] 2. Image analysis on the server

[0038] After receiving the uploaded image file, the server performs image analysis. Here, an object detection model is used to identify the main objects in the image (e.g., athletes, balls, etc.). The object detection model uses a pre-trained machine learning model that can recognize the type and location of objects with high accuracy. This allows detailed information about each element in the image to be obtained.

[0039] 3. Live comment generation on the server

[0040] The server generates commentary based on the analysis results using natural language generation technology. For example, when an image of a soccer game is input, analysis can yield information such as "Player A is dribbling the ball." Based on this information, the generation AI generates commentary such as "Player A is dribbling the ball."

[0041] 4. Output and display of live commentary

[0042] The generated commentary is sent from the server to the user's device. The server returns the commentary as an HTTP response. The device uses a dedicated UI component (e.g., text view) to display the received commentary to the user. This allows the user to experience high-quality live broadcasting in real time.

[0043] Specific examples

[0044] Example: Soccer image

[0045] 1. A user uploads images of a soccer match from their device to the system.

[0046] The user uses the image selection interface to select an image and then taps the upload button.

[0047] 2. The server receives the uploaded images and performs image analysis, using object detection models to identify players, the ball, goals, etc.

[0048] 3. Based on the analysis results, the server obtains the information that "Player A has the ball." Based on this information, the generation AI generates a commentary such as "Player A is dribbling the ball."

[0049] 4. The server sends the generated commentary to the user's device, which then displays it to the user, allowing the user to understand the situation of the game in real time.

[0050] This system allows users to receive accurate live coverage in real time, eliminating the need for specialized announcers and significantly reducing costs and effort. At the same time, it is expected to improve the user experience.

[0051] The processing flow will be explained below.

[0052] Step 1:

[0053] The user selects an image from the device and clicks the upload button. This operation causes the user device to attach the selected image file to an HTTP request and send it to the server. Specifically, when an image file is selected using the image selection interface and the send button is pressed, the device generates and sends an HTTP POST request including the image data to the server.

[0054] Step 2:

[0055] The server saves the received image file. The server analyzes the HTTP request, extracts the file data, and saves the image file in the specified directory. This makes the image to be analyzed accessible on the server.

[0056] Step 3:

[0057] The server analyzes the saved images. The server loads an object detection model and analyzes the saved images to identify key objects and movements within the images. Specifically, the image is input into the object detection model to detect the position and type of players, balls, etc. This provides detailed information about the scene within the image.

[0058] Step 4:

[0059] The server generates commentary based on the analysis results. The server uses natural language generation technology to generate appropriate commentary based on the analyzed objects and their movements. For example, based on the analysis result that "Player A has the ball," the server generates the commentary "Player A is dribbling the ball."

[0060] Step 5:

[0061] The server sends the generated commentary to the user's terminal. The server returns the generated commentary to the user's terminal as an HTTP response. This response includes the generated commentary in text format.

[0062] Step 6:

[0063] The device displays the commentary received to the user. The device receives the HTTP response and displays the commentary in a UI component on the screen. Specifically, it uses a user interface such as a text view to display the commentary in a format that is easy for the user to read.

[0064] This allows users to experience high-quality live broadcasts based on real-time images.

[0065] Example 1

[0066] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0067] In recent years, there has been a demand for real-time information provision for sports and event broadcasts, but deploying professional announcers and facilitators is costly and difficult. This invention aims to solve this problem by providing high-quality commentary in real time through an automated system. In particular, by combining image analysis and natural language generation technology, it aims to provide users with an experience similar to that of professional commentary.

[0068] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0069] In this invention, the server includes means for uploading images from a user's terminal, means for receiving and saving the uploaded images, means for analyzing the saved images using an object detection model, means for generating live commentary comments using natural language generation technology based on the analysis results, means for transmitting the generated live commentary comments from the server to the user's terminal, and means for displaying the transmitted live commentary comments on the user's terminal, thereby enabling automatic provision of high-quality live commentary comments in real time.

[0070] A "user's terminal" is a device used by a user to upload images, and includes, for example, a smartphone, a personal computer, a tablet terminal, etc.

[0071] "Server" refers to a central processing unit that receives and stores images sent from user terminals, and further analyzes and generates live comments.

[0072] "Upload" refers to the process of sending image data from a user's device to a server.

[0073] "Image analysis" refers to the process by which the server uses object detection models to identify objects and movements in uploaded images.

[0074] An "object detection model" is a machine learning model used to identify dominant objects in an image, and can use techniques such as YOLO or SSD.

[0075] "Natural language generation technology" refers to technology for generating sentences in a form that can be understood by humans based on information obtained from analysis, and examples include generative AI models (e.g., GPT-3 (registered trademark)).

[0076] "Live commentary" refers to explanatory text about movements and events within an image that is generated based on the analysis results.

[0077] A "generative AI model" is an artificial intelligence model used to generate natural language text from a given prompt and analysis results.

[0078] A "prompt sentence" refers to a sentence that is input into a generative AI model to instruct it on what kind of text to generate.

[0079] An "HTTP request" refers to a form of communication protocol that requests data transmission from a user terminal to a server.

[0080] "HTTP response" refers to a form of communication protocol that returns data from a server to a user's terminal.

[0081] MODE FOR CARRYING OUT THE INVENTION

[0082] The present invention relates to a system in which a user uploads an image from a terminal, a server analyzes the image, and generates live commentary in real time. The present invention is comprised of the following steps as an embodiment.

[0083] Required Hardware and Software

[0084] User devices: smartphones, computers, tablets, etc.

[0085] Server: Central Processing Unit

[0086] Web interface or application: Software for uploading images

[0087] Object detection models: YOLO, SSD, etc.

[0088] Generative AI models: GPT-3, BERT, etc.

[0089] Communication method: HTTP protocol

[0090] Specific processing and data calculations

[0091] 1. Image input from the user

[0092] Users use devices such as smartphones or PCs to select and upload images through a web interface or application. When they tap the upload button, the device sends the image data to the server as an HTTP POST request.

[0093] 2. Receiving images on the server

[0094] The server receives the HTTP POST request sent from the user's device and saves the image file. This data is stored in the server's temporary directory.

[0095] 3. Image analysis on the server

[0096] The server analyzes the stored image files, using a pre-trained object detection model (e.g., YOLO or SSD) to identify the dominant objects in the image. The analysis results in the location and type of object in the image. This information is stored in an internal data structure.

[0097] 4. Live comment generation on the server

[0098] The server generates commentary based on the results of image analysis. Specifically, it extracts key information from the analysis (e.g., "Player A has the ball") and inputs the prompt and analysis results into a generative AI model (e.g., GPT-3). The generated commentary is then translated into natural language.

[0099] Example prompt sentence:

[0100] "Image analysis results of a soccer game: Player A has the ball. Please generate commentary based on this information."

[0101] 5. Sending live commentary from the server to the device

[0102] The server sends the generated live commentary to the user's terminal. The server creates a live commentary as an HTTP response and sends it to the user's terminal.

[0103] 6. Displaying live comments on user devices

[0104] The user's device displays the received live commentary. It receives the HTTP response, extracts the live commentary from the response body, and displays it to the user in real time using a dedicated UI component (e.g., text view).

[0105] In this way, the user can experience high quality real-time live broadcasting. The above is a specific embodiment for carrying out the present invention.

[0106] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0107] Step 1:

[0108] A user uploads an image using a device such as a smartphone or PC. Specifically, they first open a web interface or application, click or tap the image selection button to select an image file from the device's storage, and then click or tap the upload button. This causes the device to send the selected image file to the server as an HTTP POST request.

[0109] Input: An image file selected by the user

[0110] Output: Image data is sent to the server as an HTTP POST request

[0111] Step 2:

[0112] The server receives the HTTP POST request sent from the user's device and saves the image file. Specifically, it extracts the image data from the body of the received request and saves it in a temporary directory on the server.

[0113] Input: Image data sent as an HTTP POST request

[0114] Output: Image file saved in a temporary directory on the server

[0115] Step 3:

[0116] The server analyzes the stored image files and identifies the dominant object in the image using a pre-trained object detection model (e.g., YOLO or SSD). Specifically, the image data is input to the object detection model, and the location and type of the dominant object are obtained as output. The analysis results are stored in an internal data structure.

[0117] Input: Image file saved in the temporary directory on the server

[0118] Output: Analysis results on the location and type of objects in the image

[0119] Step 4:

[0120] The server generates commentary based on the results of image analysis. Specifically, it extracts key information from the analysis results (e.g., "Player A has the ball") and inputs the prompt and analysis results into a generative AI model (e.g., GPT-3). The generative AI model generates commentary in natural language based on the input.

[0121] Input: Analysis results obtained from the object detection model

[0122] Output: Generated natural language commentary

[0123] Example prompt sentence:

[0124] "Image analysis results of a soccer game: Player A has the ball. Please generate commentary based on this information."

[0125] Step 5:

[0126] The server sends the generated commentary to the user's terminal. Specifically, the server creates the commentary as an HTTP response and sends it to the user's terminal.

[0127] Input: Generated commentary

[0128] Output: Commentary sent to the user's device as an HTTP response

[0129] Step 6:

[0130] The user's device displays the received live commentary. Specifically, it receives the HTTP response, extracts the live commentary from the response body, and displays it to the user using a dedicated UI component (e.g., text view).

[0131] Input: Comments sent as an HTTP response

[0132] Output: Commentary displayed on the user's device

[0133] The above is the detailed flow of the processing of this program.

[0134] (Application example 1)

[0135] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0136] Conventional factory monitoring systems have made it difficult to quickly grasp the operating status and abnormalities of equipment, resulting in delayed response when an abnormality occurs. Furthermore, constant monitoring by an operator with specialized knowledge is required, which increases costs. The present invention aims to solve these problems by providing a system that can monitor the status of equipment in a factory in real time and quickly take appropriate action.

[0137] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0138] In this invention, the server includes means for uploading images from a user's device, means for analyzing the uploaded images and identifying objects and actions in the images, means for generating commentary related to a specified situation based on the analysis results, means for transmitting the generated commentary to the user's device, means for displaying the transmitted commentary on the user's device, and means for analyzing camera footage within the factory in real time and automatically generating and notifying the status and abnormalities of equipment as commentary, thereby enabling factory supervisors and operators to grasp the status of equipment in real time and take appropriate action promptly.

[0139] An "image" is digital data containing visual information that is uploaded from a user's device.

[0140] "User's Device" is the terminal equipment used to upload images to the system.

[0141] "Uploading" is the act of sending an image from a user's device to a server.

[0142] "Analysis" is the process of processing uploaded images to identify objects and actions within them.

[0143] "Target" refers to an item to be detected, such as an object, person, or device present in an image.

[0144] "Action" refers to a specific action or change of state performed by an object in an image.

[0145] The "specified situation" refers to the state of an object or action that the system determines based on the analysis results.

[0146] "Live commentary" is a natural language description of an object or action that is generated based on the analysis results.

[0147] "Generation" is the process of creating live commentary based on the analysis results.

[0148] "Transmitting" refers to the act of transferring the generated live commentary to the user's device using a communication means.

[0149] "Display" refers to the act of visually showing the live commentary sent on the user's device.

[0150] "Camera footage" refers to real-time video data captured by cameras installed within the factory.

[0151] "Equipment status" refers to the operating status and operational status of various equipment within the factory.

[0152] "Abnormality" refers to a state in which the equipment is not operating normally or an error has occurred.

[0153] "Notification" refers to the act of notifying the user of the generated live commentary in real time on the user's device.

[0154] This invention realizes a system that analyzes camera images in a factory in real time and automatically generates and notifies the status and abnormalities of equipment as live comments. A specific embodiment of this system is described below.

[0155] Image input from the user

[0156] Users transmit real-time images from cameras installed in the factory to the server, which provides high-resolution images that are then received in real time by the user's device (such as a tablet or PC).

[0157] Image analysis on the server

[0158] The server uses an object detection model to analyze the video received from the user's device in real time. This object detection model applies a pre-trained model using a deep learning framework such as TENSORFLOW®. The server can analyze the operating status of the equipment in the image (e.g., operating normally, stopped, abnormality, etc.) with high accuracy.

[0159] Live commentary generation on the server

[0160] Based on the analysis results, the server generates a commentary in natural language using a generative AI model such as GPT-3. For example, if the server obtains the analysis result "The conveyor belt has stopped," the generative AI model will generate a specific commentary such as "The conveyor belt has stopped. Please check error code 105."

[0161] Output and display of live commentary

[0162] The server sends the generated commentary to the user's device. The commentary is returned as an HTTP response and displayed in real time on the user's device. A dedicated UI component (e.g., a text view) is used to convey visual information to the user.

[0163] Hardware and software used

[0164] Hardware: Camera (high-resolution industrial camera), server (with high-performance CPU and GPU), user device (tablet, PC, etc.)

[0165] Software: Python, TensorFlow, Transformers library, OpenCV, HTTP communication library

[0166] Specific examples

[0167] As a specific example, camera footage from inside a factory is analyzed, and the server analyzes footage of a stopped conveyor belt. In this case, the server uses an object detection model to detect the conveyor belt stopping, and then uses a generative AI model to generate a live commentary such as "The conveyor belt has stopped. Please check error code 105." The generated live commentary is sent to the user's device and displayed in real time.

[0168] Prompt Sentence Examples

[0169] An example of a prompt sentence to input to the generative AI model is as follows:

[0170] Generate a comment for the following machine state: stopped.

[0171] An example of the commentary that is output is as follows:

[0172] The machine has stopped. Please check Error Code 105.

[0173] This invention enables factory supervisors and operators to grasp the status of equipment in real time and take appropriate action quickly.

[0174] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0175] Step 1:

[0176] Users send real-time video to a server via cameras installed in their factories. Specifically, the video captured by the camera is received by the user's device (tablet, PC, etc.) and sent to the server as an HTTP request. The input is the real-time camera video, and the output is the video data sent to the server.

[0177] Step 2:

[0178] The server analyzes video data received from the user's device in real time. This analysis uses an object detection model using a deep learning framework such as TensorFlow. The input is camera video, which is analyzed to detect and identify the operating status of equipment (e.g., conveyor belts). The output is the identified operating status information.

[0179] Step 3:

[0180] Based on the analysis results, the server uses a generative AI model such as GPT-3 to generate commentary in natural language. The input is the analysis result (e.g., conveyor belt stopped), and a prompt sentence is generated based on this information. The generative AI model responds to the prompt sentence and generates an appropriate commentary. The output is the generated commentary.

[0181] Step 4:

[0182] The server transmits the generated commentary to the user's device. Specifically, the server returns the generated commentary to the user's device as an HTTP response. The input is the generated commentary, and the output is the commentary to be transmitted to the user's device.

[0183] Step 5:

[0184] The user's device receives the commentary and displays it in real time. Specifically, the commentary received as an HTTP response is displayed using a dedicated UI component such as a text view. The input is the commentary sent from the server, and the output is the visually displayed commentary.

[0185] The specific actions taken at each step enable a series of processes to work together seamlessly, enabling real-time monitoring of the status and abnormalities of equipment within the factory and appropriate notification.

[0186] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0187] This invention relates to an image analysis and commentary generation system that combines an emotion engine that recognizes user emotions. The main components of the system are a user terminal, a server, an emotion engine, and a means for communication between them. The following describes the details of the embodiment.

[0188] 1. Image input from the user

[0189] A user uploads an image to the system using a device. The user selects an image file using a specific web interface or application and taps the upload button. This action causes the device to send the image file to the server as an HTTP request.

[0190] 2. Image analysis on the server

[0191] The server stores the received image files and performs image analysis, using an object detection model to identify key objects in the image (e.g., athletes, balls, etc.). The object detection model uses a pre-trained machine learning model and can recognize the type and location of objects with high accuracy.

[0192] 3. Live comment generation on the server

[0193] The server generates commentary based on the analysis results using natural language generation technology. For example, when an image of a soccer game is input, analysis can yield information such as "Player A is dribbling the ball." Based on this information, the generation AI generates commentary such as "Player A is dribbling the ball."

[0194] 4. Emotion Recognition on the Server

[0195] The server uses an emotion engine to recognize the user's emotional state. The emotion engine can identify emotions by analyzing the user's facial expressions, voice, or biometric information. This information is obtained through a camera, microphone, or biometric sensor.

[0196] 5. Moderating commentary based on emotions

[0197] The server automatically adjusts the content and expression of the commentary based on the user's emotional state obtained from the emotion engine. For example, if the user is excited, the server can make the commentary more energetic.

[0198] 6. Output and display of live commentary

[0199] The generated commentary is sent from the server to the user's device. The server returns the commentary as an HTTP response. The device uses a dedicated UI component (e.g., text view) to display the received commentary to the user. This allows the user to experience high-quality live broadcasting in real time.

[0200] Specific examples

[0201] Example: Soccer image

[0202] 1. A user uploads soccer game images from their device to the system. The user uses the image selection interface to select an image and then taps the upload button.

[0203] 2. The server receives the uploaded images and performs image analysis, using object detection models to identify players, the ball, goals, etc.

[0204] 3. Based on the analysis results, the server obtains the information that "Player A has the ball." Based on this information, the generation AI generates a commentary such as "Player A is dribbling the ball."

[0205] 4. The server uses the camera images and audio data acquired from the user's device to analyze the user's emotions using an emotion engine. For example, if the user is smiling, it will determine that the user is excited.

[0206] 5. The server adjusts the commentary based on the user's emotional state to "Player A is moving forward with a great dribble!"

[0207] 6. The server sends the generated commentary to the user's device, which then displays the commentary to the user, allowing the user to grasp the game situation in real time and enjoy a commentary experience that suits their emotions.

[0208] This system allows users to not only receive accurate commentary in real time, but also experience commentary that is customized according to their emotional state, providing a more personalized and high-quality user experience.

[0209] The processing flow will be explained below.

[0210] Step 1:

[0211] The user selects an image from the device and clicks the upload button. This operation causes the user device to attach the selected image file to an HTTP request and send it to the server. Specifically, the user uses the image selection interface to select an image file, and then presses the send button, causing the device to send the image data to the server via an HTTP POST request.

[0212] Step 2:

[0213] The server saves the received image file. The server analyzes the HTTP request, extracts the sent file data, and saves it as an image file in the specified directory. This operation makes the image to be analyzed accessible on the server.

[0214] Step 3:

[0215] The server analyzes the saved images. The server loads an object detection model and analyzes the saved images to identify key objects and movements within the images. Specifically, the image is input into the object detection model to detect the location and type of objects, such as players and the ball.

[0216] Step 4:

[0217] The server generates commentary based on the analysis results. The server generates commentary using natural language generation technology based on the analyzed objects and their movements. For example, if the analysis result is "Player A is dribbling the ball," the server generates commentary such as "Player A is dribbling the ball."

[0218] Step 5:

[0219] The user transmits emotion data from the device. The user uses the device's camera and microphone to transmit their facial expressions and voice to the system. The device then transmits this data to the server in real time.

[0220] Step 6:

[0221] The server analyzes the user's emotions using an emotion engine. The server inputs the received user's facial expressions and voice into the emotion engine to determine the user's current emotional state. For example, if the server receives image data of the user smiling, it determines that the user is excited.

[0222] Step 7:

[0223] The server adjusts the commentary based on the user's emotional state. The tone and content of the commentary is changed based on the emotional information obtained from the emotion engine. For example, if the user is excited, the commentary will be adjusted from "Player A is dribbling the ball" to an energetic expression such as "Player A is moving forward with a great dribble!"

[0224] Step 8:

[0225] The server sends the adjusted live commentary to the user's terminal. The server sends the adjusted live commentary to the user's terminal as an HTTP response.

[0226] Step 9:

[0227] The device displays the commentary received to the user. The device receives the HTTP response and displays the commentary in a UI component on the screen. Specifically, the device uses an interface such as a text view to display the commentary in a format that is easy for the user to see.

[0228] This process flow allows users to not only receive accurate live coverage in real time, but also enjoy a live experience that is customized according to their emotional state.

[0229] Example 2

[0230] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0231] Conventional image analysis systems were able to generate live commentary related to a specific event based on image information provided by the user, but were unable to provide personalized live commentary that took the user's emotional state into account. This resulted in a uniform user experience, making it difficult to provide high-quality live coverage that was tailored to each individual user's emotional state. The present invention aims to solve this problem by generating live commentary customized according to the user's emotional state, thereby providing a personalized, high-quality live experience.

[0232] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0233] In this invention, the server includes means for uploading images from a user's terminal, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating live commentary related to a specified event based on the analysis results, means for transmitting the generated live commentary to the user's terminal, means for displaying the transmitted live commentary on the user's terminal, means for acquiring emotional data for recognizing the user's emotional state, and means for adjusting the content and expression of the live commentary based on the acquired emotional data. This makes it possible to provide customized live commentary that takes the user's emotional state into consideration in real time, thereby providing a more personalized and high-quality live commentary experience.

[0234] "Means for uploading images from the user's terminal" is a function that allows the user to send image files from their own terminal to the server.

[0235] "Means of analyzing uploaded images and identifying objects and movements within the images" refers to a function that uses image analysis technology on received images to recognize and identify objects and their movements within the images.

[0236] The "means for generating live commentary related to a specified event based on the analysis results" is a function that automatically generates sentences related to an event based on data obtained from image analysis.

[0237] The "means for transmitting the generated live commentary to the user's terminal" is a function for transmitting the automatically generated live commentary from the server to the user's terminal.

[0238] The "means for displaying the transmitted live commentary on the user's terminal" is a function for displaying the live commentary received on the user's terminal.

[0239] The "means for acquiring emotional data to recognize the user's emotional state" is a function for collecting emotional data such as the user's facial expressions and voice through sensors, cameras, etc.

[0240] The "means for adjusting the content and expression of the commentary commentary based on the acquired emotional data" is a function for analyzing the collected emotional data and changing the tone and content of the commentary commentary based on the results.

[0241] The present invention relates to an image analysis and commentary generation system that combines an emotion engine that recognizes user emotions. The following describes in detail the embodiments of the present invention.

[0242] The system of the present invention is mainly composed of a user terminal, a server, an emotion engine, and means for communicating among them.

[0243] Hardware and software used

[0244] The user's device may be a smartphone, tablet, or PC, and may be equipped with a camera, microphone, and, if necessary, biometric sensors. The server may be a cloud environment or a dedicated computer system, and includes the following main software components:

[0245] 1. Image analysis module: This module implements an object detection model to identify objects and movements in images, using computer vision libraries (e.g., OpenCV, TensorFlow).

[0246] 2. Natural Language Generation Module: Uses a generative AI model (e.g., GPT-3) to generate commentary based on the analysis results.

[0247] 3. Emotion Engine: Machine learning models are implemented to recognize the user's emotional state, using facial recognition and voice analysis models.

[0248] 4. Communication module: Processes HTTP requests and responses to send and receive data between the user's device and the server.

[0249] Example of a system

[0250] Specific example using soccer images

[0251] 1. A user uses a device to upload an image of a soccer match to the system. The user uses the image selection interface to select an image and then taps the upload button. This action sends the image file to the server as an HTTP request.

[0252] 2. The server stores the received image files and uses an image analysis module to identify objects in the images (e.g., players, ball, goal). The object detection model utilizes computer vision libraries to perform highly accurate analysis.

[0253] 3. Based on the results of the image analysis, the server uses a natural language generation module to generate a commentary such as "Player A is dribbling the ball" based on the information that "Player A has the ball." The following prompt is input to the generation AI model:

[0254] "Enter a football image."

[0255] "Analyze your emotional state and energetically adjust the tone of the play-by-play commentary you generate."

[0256] 4. The user's device uses a camera, microphone, and biometric sensors to acquire emotional data such as the user's facial expressions, voice, and heart rate, and sends this data to the server as an HTTP request.

[0257] 5. The server uses the emotion engine to analyze the received emotion data and identify the user's emotional state. For example, if the user is smiling, it determines that the user is excited.

[0258] 6. Based on the results of the emotion engine, the server adjusts the generated commentary to "Player A is moving forward with a great dribble!" This adjusted commentary is sent to the user's device as an HTTP response.

[0259] 7. The device displays the received live commentary using a dedicated UI component (e.g., text view), allowing the user to experience high-quality live broadcasts in real time.

[0260] Through the above-described method, users can obtain commentary that is customized according to their emotional state, and can enjoy a more personalized and high-quality commentary experience.

[0261] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0262] Step 1:

[0263] A user uses a device to open a specific web interface or application. They select an image file and tap the upload button to send the image to the system. The input is an image file stored on the device, and the output is that the image file is sent to the server as an HTTP request. The device generates header information for the HTTP request and attaches the image file.

[0264] Step 2:

[0265] The server receives the HTTP request. It saves the received image file and generates a path for the save location. Here, it checks the image format and size, and converts the file format or adjusts the size as necessary. The input is the image file sent from the terminal, and the output is the image file path saved in a specific directory on the server. The server extracts the image file data from the HTTP request and saves it in the specified directory.

[0266] Step 3:

[0267] The server inputs the saved image file into an object detection model and performs analysis. It uses an image analysis module to identify objects and movements and outputs the analysis results in JSON format. The input is the saved image file path, and the output is the JSON data of the analysis results. The server uses an object detection model (e.g., OpenCV, TensorFlow) to process and recognize objects and movements in the image.

[0268] Step 4:

[0269] Based on the analysis results, the server inputs the results as prompts into a generative AI model (e.g., GPT-3) to generate play-by-play comments. For example, based on the result "Player A has the ball," the server generates the comment "Player A is dribbling the ball." The input is the JSON data of the analysis results, and the output is play-by-play comments in natural language. The server creates appropriate prompts for the generative AI model and inputs them into the model.

[0270] Step 5:

[0271] The user's device uses a camera, microphone, and biometric sensors to collect the user's facial expressions, voice, heart rate, and other emotional data. The input is the user's various sensor data, and the output is emotional data sent to the server via an HTTP request. The device formats the collected data and prepares it to be sent to the server.

[0272] Step 6:

[0273] The server inputs the received emotional data into the emotion engine for analysis. The emotion engine determines whether the user is happy, excited, or sad. The input is the emotional data sent by the user, and the output is the analysis result (in JSON format) about the user's emotional state. The server uses the emotion engine to analyze the data using facial expression recognition models and voice analysis models.

[0274] Step 7:

[0275] The server readjusts the commentary based on the analysis results of the emotion engine. For example, if the user is excited, the tone of the commentary can be adjusted to be more energetic, such as "Player A is moving forward with a great dribble!" The input is the analysis results of the emotion engine and the initial commentary, and the output is the adjusted commentary. The server again inputs the prompt into the generative AI model to generate the adjusted commentary.

[0276] Step 8:

[0277] The server sends the final commentary to the user's device as an HTTP response. The input is the adjusted commentary, and the output is the HTTP response it sends to the device. The server sets the appropriate response headers and sends the commentary.

[0278] Step 9:

[0279] The device displays the received live commentary to the user. Specifically, it displays the live commentary in real time using a dedicated UI component (e.g., text view). The input is the live commentary obtained from the HTTP response, and the output is the live commentary displayed on the screen. The device performs UI update processing to provide the user with a real-time live commentary experience.

[0280] (Application example 2)

[0281] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0282] In autonomous vehicles, it is difficult to grasp the driver's emotional state in real time and provide appropriate information and alerts while driving. In addition, there is a lack of technology to personalize road information and driving advice according to the driver's emotions.

[0283] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for uploading images from a user's terminal, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating a commentary related to a specified event based on the analysis results, means for recognizing the user's emotional state, means for adjusting the content and tone of the commentary based on the recognized emotional state, means for transmitting the generated commentary to the user's terminal, and means for displaying the transmitted commentary on the user's terminal. This makes it possible to analyze the driver's emotions in real time and provide commentary and situation reports accordingly.

[0284] "Means for uploading images from the user's terminal" refers to an interface or function that allows the user to select image data from their own terminal and transfer it to the system.

[0285] "Means for analyzing uploaded images and identifying objects and movements within the images" refers to technology for analyzing received image data and extracting information about objects present within the images and their movements.

[0286] "Means for generating live commentary related to a specified event based on the analysis results" refers to a function that automatically generates commentary and narration related to a specific event based on the results of image analysis.

[0287] "Means for recognizing the user's emotional state" refers to technology that uses sensors such as cameras and microphones to identify emotions from the user's facial expressions and voice.

[0288] "Means for adjusting the content and tone of live commentary based on the recognized emotional state" refers to a technology that utilizes the user's emotional information to change the wording and expression of the live commentary that is generated.

[0289] The "means for transmitting the generated live commentary to the user's terminal" is a function for transferring the generated live commentary to the user's terminal via a communication means such as the Internet.

[0290] "Means for displaying the transmitted live commentary on the user's terminal" refers to a function for displaying and playing the received live commentary on the user's terminal by screen or audio.

[0291] This invention is a system that recognizes a user's emotions and generates and provides live commentary in real time based on those emotions. A specific example for realizing this invention is shown below.

[0292] 1. Upload an image

[0293] Users upload images to the system using their own devices. To upload images, a dedicated application or a web interface is used. The user selects an image file and taps the upload button, which sends the image data from the device to the server.

[0294] 2. Image Analysis

[0295] The server receives and analyzes images uploaded by users. This analysis uses an object detection model, a machine learning model. This model is pre-trained and can detect objects and their movements in images with high accuracy. For example, in an image of a soccer game, it can recognize players, the ball, goals, etc.

[0296] 3. Generating live commentary

[0297] Based on the analysis results, the server generates commentary using natural language generation technology. Examples of generative AI models used include GPT-3. The following prompt sentence is input to this model to generate commentary:

[0298] Road conditions: Congested. Driver's emotional state: Tired. Commentary:

[0299] 4. Emotion recognition

[0300] The server uses an emotion recognition engine to recognize the user's emotions in real time. This engine analyzes the user's facial expressions and voice data to identify their emotional state. Cameras and microphones installed in the vehicle are used to acquire the data.

[0301] 5. Adjustments to commentary

[0302] Based on the emotional information obtained by the emotion recognition engine, the server adjusts the content and tone of the live commentary. For example, if the user is tired, it uses expressions that encourage relaxation.

[0303] 6. Output and display of live commentary

[0304] The generated commentary is sent from the server to the user's device, which displays the commentary on the screen or plays it back as audio. This allows the user to receive customized commentary in real time, tailored to their emotional state.

[0305] Specific examples

[0306] soccer match

[0307] A user uploads an image of a soccer game, and the analysis results provide information such as "Player A is dribbling the ball." Furthermore, if the user is recognized as being tired, a commentary using the generative AI model is generated, such as "Player A is moving forward with a relaxed dribble!", and displayed on the user's device.

[0308] Autonomous vehicles

[0309] If the driver is tired, the road conditions are analyzed as "congested." The following prompt sentence is input into the generative AI model to generate a running commentary.

[0310] Road conditions: Congested. Driver's emotional state: Tired. Commentary:

[0311] The generated comment will be "The roads are currently congested. Please relax and continue driving," and will be personalized according to the driver's emotions.

[0312] In this way, the system combines image analysis and emotion recognition to provide users with more personalized, high-quality information.

[0313] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0314] Step 1:

[0315] Users upload images to the system through their own devices. When a user selects an image file and taps the upload button, the image data is sent from the device to the server as an HTTP request. The input is the image data from the user device, and the output is the transfer of the image data to the server.

[0316] Step 2:

[0317] The server analyzes the image data received from the user. For image analysis, it uses an object detection model, which is a machine learning model. Specifically, it analyzes the information of each pixel in the image and identifies the main objects and their movements. The input is the image data uploaded by the user, and the output is the detection results of the objects and movements.

[0318] Step 3:

[0319] The server generates commentary based on the analysis results using the generative AI model GPT-3. During the generation process, contextual information from the analysis results is input as a prompt, and the model generates commentary in natural language format. The input is the image analysis results and the prompt, and the output is the generated commentary.

[0320] Step 4:

[0321] The server recognizes the user's emotional state in real time. It uses an emotion recognition engine to analyze the user's facial and voice data from a camera and microphone to identify the user's emotional state. The input is the facial and voice data obtained from the user, and the output is emotional state information from the emotion recognition engine.

[0322] Step 5:

[0323] The server adjusts the content and tone of the commentary based on the user's emotional state. It also changes the wording and tone of the commentary based on the emotional data obtained from the emotion recognition engine. The input is the emotional state information and the commentary generated in the initial stage, and the output is the adjusted commentary.

[0324] Step 6:

[0325] The server sends the adjusted live commentary to the user's device. The live commentary is delivered to the user's device using an HTTP response as the communication method. The input is the adjusted live commentary, and the output is the comment sent to the user's device.

[0326] Step 7:

[0327] The terminal displays the received live commentary. The live commentary is provided to the user in real time in the form of a screen display or audio playback. The input is the live commentary received from the server, and the output is information provided to the user.

[0328] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0329] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0330] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0331] [Second embodiment]

[0332] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0333] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0334] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0335] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0336] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0337] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0338] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0339] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0340] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0341] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0342] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0343] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0344] This invention relates to an automated system that analyzes the status of an image in real time and generates and provides appropriate live commentary. The system mainly comprises a user terminal, a server, and a means for communicating between them. The details of the embodiment are described below.

[0345] 1. Image input from the user

[0346] A user uploads an image to the system using a device. The user selects an image file using a specific web interface or application and taps the upload button. This action causes the device to send the image file to the server as an HTTP request.

[0347] 2. Image analysis on the server

[0348] After receiving the uploaded image file, the server performs image analysis. Here, an object detection model is used to identify the main objects in the image (e.g., athletes, balls, etc.). The object detection model uses a pre-trained machine learning model that can recognize the type and location of objects with high accuracy. This allows detailed information about each element in the image to be obtained.

[0349] 3. Live comment generation on the server

[0350] The server generates commentary based on the analysis results using natural language generation technology. For example, when an image of a soccer game is input, analysis can yield information such as "Player A is dribbling the ball." Based on this information, the generation AI generates commentary such as "Player A is dribbling the ball."

[0351] 4. Output and display of live commentary

[0352] The generated commentary is sent from the server to the user's device. The server returns the commentary as an HTTP response. The device uses a dedicated UI component (e.g., text view) to display the received commentary to the user. This allows the user to experience high-quality live broadcasting in real time.

[0353] Specific examples

[0354] Example: Soccer image

[0355] 1. A user uploads images of a soccer match from their device to the system.

[0356] The user uses the image selection interface to select an image and then taps the upload button.

[0357] 2. The server receives the uploaded images and performs image analysis, using object detection models to identify players, the ball, goals, etc.

[0358] 3. Based on the analysis results, the server obtains the information that "Player A has the ball." Based on this information, the generation AI generates a commentary such as "Player A is dribbling the ball."

[0359] 4. The server sends the generated commentary to the user's device, which then displays it to the user, allowing the user to understand the situation of the game in real time.

[0360] This system allows users to receive accurate live coverage in real time, eliminating the need for specialized announcers and significantly reducing costs and effort. At the same time, it is expected to improve the user experience.

[0361] The processing flow will be explained below.

[0362] Step 1:

[0363] The user selects an image from the device and clicks the upload button. This operation causes the user device to attach the selected image file to an HTTP request and send it to the server. Specifically, when an image file is selected using the image selection interface and the send button is pressed, the device generates and sends an HTTP POST request including the image data to the server.

[0364] Step 2:

[0365] The server saves the received image file. The server analyzes the HTTP request, extracts the file data, and saves the image file in the specified directory. This makes the image to be analyzed accessible on the server.

[0366] Step 3:

[0367] The server analyzes the saved images. The server loads an object detection model and analyzes the saved images to identify key objects and movements within the images. Specifically, the image is input into the object detection model to detect the position and type of players, balls, etc. This provides detailed information about the scene within the image.

[0368] Step 4:

[0369] The server generates commentary based on the analysis results. The server uses natural language generation technology to generate appropriate commentary based on the analyzed objects and their movements. For example, based on the analysis result that "Player A has the ball," the server generates the commentary "Player A is dribbling the ball."

[0370] Step 5:

[0371] The server sends the generated commentary to the user's terminal. The server returns the generated commentary to the user's terminal as an HTTP response. This response includes the generated commentary in text format.

[0372] Step 6:

[0373] The device displays the commentary received to the user. The device receives the HTTP response and displays the commentary in a UI component on the screen. Specifically, it uses a user interface such as a text view to display the commentary in a format that is easy for the user to read.

[0374] This allows users to experience high-quality live broadcasts based on real-time images.

[0375] Example 1

[0376] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0377] In recent years, there has been a demand for real-time information provision for sports and event broadcasts, but deploying professional announcers and facilitators is costly and difficult. This invention aims to solve this problem by providing high-quality commentary in real time through an automated system. In particular, by combining image analysis and natural language generation technology, it aims to provide users with an experience similar to that of professional commentary.

[0378] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0379] In this invention, the server includes means for uploading images from a user's terminal, means for receiving and saving the uploaded images, means for analyzing the saved images using an object detection model, means for generating live commentary comments using natural language generation technology based on the analysis results, means for transmitting the generated live commentary comments from the server to the user's terminal, and means for displaying the transmitted live commentary comments on the user's terminal, thereby enabling automatic provision of high-quality live commentary comments in real time.

[0380] A "user's terminal" is a device used by a user to upload images, and includes, for example, a smartphone, a personal computer, a tablet terminal, etc.

[0381] "Server" refers to a central processing unit that receives and stores images sent from user terminals, and further analyzes and generates live comments.

[0382] "Upload" refers to the process of sending image data from a user's device to a server.

[0383] "Image analysis" refers to the process by which the server uses object detection models to identify objects and movements in uploaded images.

[0384] An "object detection model" is a machine learning model used to identify dominant objects in an image, and can use techniques such as YOLO or SSD.

[0385] "Natural language generation technology" refers to technology for generating sentences in a form that can be understood by humans based on information obtained from analysis, and examples include generative AI models (such as GPT-3).

[0386] "Live commentary" refers to explanatory text about movements and events within an image that is generated based on the analysis results.

[0387] A "generative AI model" is an artificial intelligence model used to generate natural language text from a given prompt and analysis results.

[0388] A "prompt sentence" refers to a sentence that is input into a generative AI model to instruct it on what kind of text to generate.

[0389] An "HTTP request" refers to a form of communication protocol that requests data transmission from a user terminal to a server.

[0390] "HTTP response" refers to a form of communication protocol that returns data from a server to a user's terminal.

[0391] MODE FOR CARRYING OUT THE INVENTION

[0392] The present invention relates to a system in which a user uploads an image from a terminal, a server analyzes the image, and generates live commentary in real time. The present invention is comprised of the following steps as an embodiment.

[0393] Required Hardware and Software

[0394] User devices: smartphones, computers, tablets, etc.

[0395] Server: Central Processing Unit

[0396] Web interface or application: Software for uploading images

[0397] Object detection models: YOLO, SSD, etc.

[0398] Generative AI models: GPT-3, BERT, etc.

[0399] Communication method: HTTP protocol

[0400] Specific processing and data calculations

[0401] 1. Image input from the user

[0402] Users use devices such as smartphones or PCs to select and upload images through a web interface or application. When they tap the upload button, the device sends the image data to the server as an HTTP POST request.

[0403] 2. Receiving images on the server

[0404] The server receives the HTTP POST request sent from the user's device and saves the image file. This data is stored in the server's temporary directory.

[0405] 3. Image analysis on the server

[0406] The server analyzes the stored image files, using a pre-trained object detection model (e.g., YOLO or SSD) to identify the dominant objects in the image. The analysis results in the location and type of object in the image. This information is stored in an internal data structure.

[0407] 4. Live comment generation on the server

[0408] The server generates commentary based on the results of image analysis. Specifically, it extracts key information from the analysis (e.g., "Player A has the ball") and inputs the prompt and analysis results into a generative AI model (e.g., GPT-3). The generated commentary is then translated into natural language.

[0409] Example prompt sentence:

[0410] "Image analysis results of a soccer game: Player A has the ball. Please generate commentary based on this information."

[0411] 5. Sending live commentary from the server to the device

[0412] The server sends the generated live commentary to the user's terminal. The server creates a live commentary as an HTTP response and sends it to the user's terminal.

[0413] 6. Displaying live comments on user devices

[0414] The user's device displays the received live commentary. It receives the HTTP response, extracts the live commentary from the response body, and displays it to the user in real time using a dedicated UI component (e.g., text view).

[0415] In this way, the user can experience high quality real-time live broadcasting. The above is a specific embodiment for carrying out the present invention.

[0416] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0417] Step 1:

[0418] A user uploads an image using a device such as a smartphone or PC. Specifically, they first open a web interface or application, click or tap the image selection button to select an image file from the device's storage, and then click or tap the upload button. This causes the device to send the selected image file to the server as an HTTP POST request.

[0419] Input: An image file selected by the user

[0420] Output: Image data is sent to the server as an HTTP POST request

[0421] Step 2:

[0422] The server receives the HTTP POST request sent from the user's device and saves the image file. Specifically, it extracts the image data from the body of the received request and saves it in a temporary directory on the server.

[0423] Input: Image data sent as an HTTP POST request

[0424] Output: Image file saved in a temporary directory on the server

[0425] Step 3:

[0426] The server analyzes the stored image files and identifies the dominant object in the image using a pre-trained object detection model (e.g., YOLO or SSD). Specifically, the image data is input to the object detection model, and the location and type of the dominant object are obtained as output. The analysis results are stored in an internal data structure.

[0427] Input: Image file saved in the temporary directory on the server

[0428] Output: Analysis results on the location and type of objects in the image

[0429] Step 4:

[0430] The server generates commentary based on the results of image analysis. Specifically, it extracts key information from the analysis results (e.g., "Player A has the ball") and inputs the prompt and analysis results into a generative AI model (e.g., GPT-3). The generative AI model generates commentary in natural language based on the input.

[0431] Input: Analysis results obtained from the object detection model

[0432] Output: Generated natural language commentary

[0433] Example prompt sentence:

[0434] "Image analysis results of a soccer game: Player A has the ball. Please generate commentary based on this information."

[0435] Step 5:

[0436] The server sends the generated commentary to the user's terminal. Specifically, the server creates the commentary as an HTTP response and sends it to the user's terminal.

[0437] Input: Generated commentary

[0438] Output: Commentary sent to the user's device as an HTTP response

[0439] Step 6:

[0440] The user's device displays the received live commentary. Specifically, it receives the HTTP response, extracts the live commentary from the response body, and displays it to the user using a dedicated UI component (e.g., text view).

[0441] Input: Comments sent as an HTTP response

[0442] Output: Commentary displayed on the user's device

[0443] The above is the detailed flow of the processing of this program.

[0444] (Application example 1)

[0445] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0446] Conventional factory monitoring systems have made it difficult to quickly grasp the operating status and abnormalities of equipment, resulting in delayed response when an abnormality occurs. Furthermore, constant monitoring by an operator with specialized knowledge is required, which increases costs. The present invention aims to solve these problems by providing a system that can monitor the status of equipment in a factory in real time and quickly take appropriate action.

[0447] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0448] In this invention, the server includes means for uploading images from a user's device, means for analyzing the uploaded images and identifying objects and actions in the images, means for generating commentary related to a specified situation based on the analysis results, means for transmitting the generated commentary to the user's device, means for displaying the transmitted commentary on the user's device, and means for analyzing camera footage within the factory in real time and automatically generating and notifying the status and abnormalities of equipment as commentary, thereby enabling factory supervisors and operators to grasp the status of equipment in real time and take appropriate action promptly.

[0449] An "image" is digital data containing visual information that is uploaded from a user's device.

[0450] "User's Device" is the terminal equipment used to upload images to the system.

[0451] "Uploading" is the act of sending an image from a user's device to a server.

[0452] "Analysis" is the process of processing uploaded images to identify objects and actions within them.

[0453] "Target" refers to an item to be detected, such as an object, person, or device present in an image.

[0454] "Action" refers to a specific action or change of state performed by an object in an image.

[0455] The "specified situation" refers to the state of an object or action that the system determines based on the analysis results.

[0456] "Live commentary" is a natural language description of an object or action that is generated based on the analysis results.

[0457] "Generation" is the process of creating live commentary based on the analysis results.

[0458] "Transmitting" refers to the act of transferring the generated live commentary to the user's device using a communication means.

[0459] "Display" refers to the act of visually showing the live commentary sent on the user's device.

[0460] "Camera footage" refers to real-time video data captured by cameras installed within the factory.

[0461] "Equipment status" refers to the operating status and operational status of various equipment within the factory.

[0462] "Abnormality" refers to a state in which the equipment is not operating normally or an error has occurred.

[0463] "Notification" refers to the act of notifying the user of the generated live commentary in real time on the user's device.

[0464] This invention realizes a system that analyzes camera images in a factory in real time and automatically generates and notifies the status and abnormalities of equipment as live comments. A specific embodiment of this system is described below.

[0465] Image input from the user

[0466] Users transmit real-time images from cameras installed in the factory to the server, which provides high-resolution images that are then received in real time by the user's device (such as a tablet or PC).

[0467] Image analysis on the server

[0468] The server uses an object detection model to analyze the video received from the user's device in real time. This object detection model applies a pre-trained model using a deep learning framework such as TensorFlow. The server can accurately analyze the operating status of the equipment in the image (e.g., operating normally, stopped, abnormality, etc.).

[0469] Live commentary generation on the server

[0470] Based on the analysis results, the server generates a commentary in natural language using a generative AI model such as GPT-3. For example, if the server obtains the analysis result "The conveyor belt has stopped," the generative AI model will generate a specific commentary such as "The conveyor belt has stopped. Please check error code 105."

[0471] Output and display of live commentary

[0472] The server sends the generated commentary to the user's device. The commentary is returned as an HTTP response and displayed in real time on the user's device. A dedicated UI component (e.g., a text view) is used to convey visual information to the user.

[0473] Hardware and software used

[0474] Hardware: Camera (high-resolution industrial camera), server (with high-performance CPU and GPU), user device (tablet, PC, etc.)

[0475] Software: Python, TensorFlow, Transformers library, OpenCV, HTTP communication library

[0476] Specific examples

[0477] As a specific example, camera footage from inside a factory is analyzed, and the server analyzes footage of a stopped conveyor belt. In this case, the server uses an object detection model to detect the conveyor belt stopping, and then uses a generative AI model to generate a live commentary such as "The conveyor belt has stopped. Please check error code 105." The generated live commentary is sent to the user's device and displayed in real time.

[0478] Prompt Sentence Examples

[0479] An example of a prompt sentence to input to the generative AI model is as follows:

[0480] Generate a comment for the following machine state: stopped.

[0481] An example of the commentary that is output is as follows:

[0482] The machine has stopped. Please check Error Code 105.

[0483] This invention enables factory supervisors and operators to grasp the status of equipment in real time and take appropriate action quickly.

[0484] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0485] Step 1:

[0486] Users send real-time video to a server via cameras installed in their factories. Specifically, the video captured by the camera is received by the user's device (tablet, PC, etc.) and sent to the server as an HTTP request. The input is the real-time camera video, and the output is the video data sent to the server.

[0487] Step 2:

[0488] The server analyzes video data received from the user's device in real time. This analysis uses an object detection model using a deep learning framework such as TensorFlow. The input is camera video, which is analyzed to detect and identify the operating status of equipment (e.g., conveyor belts). The output is the identified operating status information.

[0489] Step 3:

[0490] Based on the analysis results, the server uses a generative AI model such as GPT-3 to generate commentary in natural language. The input is the analysis result (e.g., conveyor belt stopped), and a prompt sentence is generated based on this information. The generative AI model responds to the prompt sentence and generates an appropriate commentary. The output is the generated commentary.

[0491] Step 4:

[0492] The server transmits the generated commentary to the user's device. Specifically, the server returns the generated commentary to the user's device as an HTTP response. The input is the generated commentary, and the output is the commentary to be transmitted to the user's device.

[0493] Step 5:

[0494] The user's device receives the commentary and displays it in real time. Specifically, the commentary received as an HTTP response is displayed using a dedicated UI component such as a text view. The input is the commentary sent from the server, and the output is the visually displayed commentary.

[0495] The specific actions taken at each step enable a series of processes to work together seamlessly, enabling real-time monitoring of the status and abnormalities of equipment within the factory and appropriate notification.

[0496] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0497] This invention relates to an image analysis and commentary generation system that combines an emotion engine that recognizes user emotions. The main components of the system are a user terminal, a server, an emotion engine, and a means for communication between them. The following describes the details of the embodiment.

[0498] 1. Image input from the user

[0499] A user uploads an image to the system using a device. The user selects an image file using a specific web interface or application and taps the upload button. This action causes the device to send the image file to the server as an HTTP request.

[0500] 2. Image analysis on the server

[0501] The server stores the received image files and performs image analysis, using an object detection model to identify key objects in the image (e.g., athletes, balls, etc.). The object detection model uses a pre-trained machine learning model and can recognize the type and location of objects with high accuracy.

[0502] 3. Live comment generation on the server

[0503] The server generates commentary based on the analysis results using natural language generation technology. For example, when an image of a soccer game is input, analysis can yield information such as "Player A is dribbling the ball." Based on this information, the generation AI generates commentary such as "Player A is dribbling the ball."

[0504] 4. Emotion Recognition on the Server

[0505] The server uses an emotion engine to recognize the user's emotional state. The emotion engine can identify emotions by analyzing the user's facial expressions, voice, or biometric information. This information is obtained through a camera, microphone, or biometric sensor.

[0506] 5. Moderating commentary based on emotions

[0507] The server automatically adjusts the content and expression of the commentary based on the user's emotional state obtained from the emotion engine. For example, if the user is excited, the server can make the commentary more energetic.

[0508] 6. Output and display of live commentary

[0509] The generated commentary is sent from the server to the user's device. The server returns the commentary as an HTTP response. The device uses a dedicated UI component (e.g., text view) to display the received commentary to the user. This allows the user to experience high-quality live broadcasting in real time.

[0510] Specific examples

[0511] Example: Soccer image

[0512] 1. A user uploads soccer game images from their device to the system. The user uses the image selection interface to select an image and then taps the upload button.

[0513] 2. The server receives the uploaded images and performs image analysis, using object detection models to identify players, the ball, goals, etc.

[0514] 3. Based on the analysis results, the server obtains the information that "Player A has the ball." Based on this information, the generation AI generates a commentary such as "Player A is dribbling the ball."

[0515] 4. The server uses the camera images and audio data acquired from the user's device to analyze the user's emotions using an emotion engine. For example, if the user is smiling, it will determine that the user is excited.

[0516] 5. The server adjusts the commentary based on the user's emotional state to "Player A is moving forward with a great dribble!"

[0517] 6. The server sends the generated commentary to the user's device, which then displays the commentary to the user, allowing the user to grasp the game situation in real time and enjoy a commentary experience that suits their emotions.

[0518] This system allows users to not only receive accurate commentary in real time, but also experience commentary that is customized according to their emotional state, providing a more personalized and high-quality user experience.

[0519] The processing flow will be explained below.

[0520] Step 1:

[0521] The user selects an image from the device and clicks the upload button. This operation causes the user device to attach the selected image file to an HTTP request and send it to the server. Specifically, the user uses the image selection interface to select an image file, and then presses the send button, causing the device to send the image data to the server via an HTTP POST request.

[0522] Step 2:

[0523] The server saves the received image file. The server analyzes the HTTP request, extracts the sent file data, and saves it as an image file in the specified directory. This operation makes the image to be analyzed accessible on the server.

[0524] Step 3:

[0525] The server analyzes the saved images. The server loads an object detection model and analyzes the saved images to identify key objects and movements within the images. Specifically, the image is input into the object detection model to detect the location and type of objects, such as players and the ball.

[0526] Step 4:

[0527] The server generates commentary based on the analysis results. The server generates commentary using natural language generation technology based on the analyzed objects and their movements. For example, if the analysis result is "Player A is dribbling the ball," the server generates commentary such as "Player A is dribbling the ball."

[0528] Step 5:

[0529] The user transmits emotion data from the device. The user uses the device's camera and microphone to transmit their facial expressions and voice to the system. The device then transmits this data to the server in real time.

[0530] Step 6:

[0531] The server analyzes the user's emotions using an emotion engine. The server inputs the received user's facial expressions and voice into the emotion engine to determine the user's current emotional state. For example, if the server receives image data of the user smiling, it determines that the user is excited.

[0532] Step 7:

[0533] The server adjusts the commentary based on the user's emotional state. The tone and content of the commentary is changed based on the emotional information obtained from the emotion engine. For example, if the user is excited, the commentary will be adjusted from "Player A is dribbling the ball" to an energetic expression such as "Player A is moving forward with a great dribble!"

[0534] Step 8:

[0535] The server sends the adjusted live commentary to the user's terminal. The server sends the adjusted live commentary to the user's terminal as an HTTP response.

[0536] Step 9:

[0537] The device displays the commentary received to the user. The device receives the HTTP response and displays the commentary in a UI component on the screen. Specifically, the device uses an interface such as a text view to display the commentary in a format that is easy for the user to see.

[0538] This process flow allows users to not only receive accurate live coverage in real time, but also enjoy a live experience that is customized according to their emotional state.

[0539] Example 2

[0540] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0541] Conventional image analysis systems were able to generate live commentary related to a specific event based on image information provided by the user, but were unable to provide personalized live commentary that took the user's emotional state into account. This resulted in a uniform user experience, making it difficult to provide high-quality live coverage that was tailored to each individual user's emotional state. The present invention aims to solve this problem by generating live commentary customized according to the user's emotional state, thereby providing a personalized, high-quality live experience.

[0542] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0543] In this invention, the server includes means for uploading images from a user's terminal, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating live commentary related to a specified event based on the analysis results, means for transmitting the generated live commentary to the user's terminal, means for displaying the transmitted live commentary on the user's terminal, means for acquiring emotional data for recognizing the user's emotional state, and means for adjusting the content and expression of the live commentary based on the acquired emotional data. This makes it possible to provide customized live commentary that takes the user's emotional state into consideration in real time, thereby providing a more personalized and high-quality live commentary experience.

[0544] "Means for uploading images from the user's terminal" is a function that allows the user to send image files from their own terminal to the server.

[0545] "Means of analyzing uploaded images and identifying objects and movements within the images" refers to a function that uses image analysis technology on received images to recognize and identify objects and their movements within the images.

[0546] The "means for generating live commentary related to a specified event based on the analysis results" is a function that automatically generates sentences related to an event based on data obtained from image analysis.

[0547] The "means for transmitting the generated live commentary to the user's terminal" is a function for transmitting the automatically generated live commentary from the server to the user's terminal.

[0548] The "means for displaying the transmitted live commentary on the user's terminal" is a function for displaying the live commentary received on the user's terminal.

[0549] The "means for acquiring emotional data to recognize the user's emotional state" is a function for collecting emotional data such as the user's facial expressions and voice through sensors, cameras, etc.

[0550] The "means for adjusting the content and expression of the commentary commentary based on the acquired emotional data" is a function for analyzing the collected emotional data and changing the tone and content of the commentary commentary based on the results.

[0551] The present invention relates to an image analysis and commentary generation system that combines an emotion engine that recognizes user emotions. The following describes in detail the embodiments of the present invention.

[0552] The system of the present invention is mainly composed of a user terminal, a server, an emotion engine, and means for communicating among them.

[0553] Hardware and software used

[0554] The user's device may be a smartphone, tablet, or PC, and may be equipped with a camera, microphone, and, if necessary, biometric sensors. The server may be a cloud environment or a dedicated computer system, and includes the following main software components:

[0555] 1. Image analysis module: This module implements an object detection model to identify objects and movements in images, using computer vision libraries (e.g., OpenCV, TensorFlow).

[0556] 2. Natural Language Generation Module: Uses a generative AI model (e.g., GPT-3) to generate commentary based on the analysis results.

[0557] 3. Emotion Engine: Machine learning models are implemented to recognize the user's emotional state, using facial recognition and voice analysis models.

[0558] 4. Communication module: Processes HTTP requests and responses to send and receive data between the user's device and the server.

[0559] Example of a system

[0560] Specific example using soccer images

[0561] 1. A user uses a device to upload an image of a soccer match to the system. The user uses the image selection interface to select an image and then taps the upload button. This action sends the image file to the server as an HTTP request.

[0562] 2. The server stores the received image files and uses an image analysis module to identify objects in the images (e.g., players, ball, goal). The object detection model utilizes computer vision libraries to perform highly accurate analysis.

[0563] 3. Based on the results of the image analysis, the server uses a natural language generation module to generate a commentary such as "Player A is dribbling the ball" based on the information that "Player A has the ball." The following prompt is input to the generation AI model:

[0564] "Enter a football image."

[0565] "Analyze your emotional state and energetically adjust the tone of the play-by-play commentary you generate."

[0566] 4. The user's device uses a camera, microphone, and biometric sensors to acquire emotional data such as the user's facial expressions, voice, and heart rate, and sends this data to the server as an HTTP request.

[0567] 5. The server uses the emotion engine to analyze the received emotion data and identify the user's emotional state. For example, if the user is smiling, it determines that the user is excited.

[0568] 6. Based on the results of the emotion engine, the server adjusts the generated commentary to "Player A is moving forward with a great dribble!" This adjusted commentary is sent to the user's device as an HTTP response.

[0569] 7. The device displays the received live commentary using a dedicated UI component (e.g., text view), allowing the user to experience high-quality live broadcasts in real time.

[0570] Through the above-described method, users can obtain commentary that is customized according to their emotional state, and can enjoy a more personalized and high-quality commentary experience.

[0571] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0572] Step 1:

[0573] A user uses a device to open a specific web interface or application. They select an image file and tap the upload button to send the image to the system. The input is an image file stored on the device, and the output is that the image file is sent to the server as an HTTP request. The device generates header information for the HTTP request and attaches the image file.

[0574] Step 2:

[0575] The server receives the HTTP request. It saves the received image file and generates a path for the save location. Here, it checks the image format and size, and converts the file format or adjusts the size as necessary. The input is the image file sent from the terminal, and the output is the image file path saved in a specific directory on the server. The server extracts the image file data from the HTTP request and saves it in the specified directory.

[0576] Step 3:

[0577] The server inputs the saved image file into an object detection model and performs analysis. It uses an image analysis module to identify objects and movements and outputs the analysis results in JSON format. The input is the saved image file path, and the output is the JSON data of the analysis results. The server uses an object detection model (e.g., OpenCV, TensorFlow) to process and recognize objects and movements in the image.

[0578] Step 4:

[0579] Based on the analysis results, the server inputs the results as prompts into a generative AI model (e.g., GPT-3) to generate play-by-play comments. For example, based on the result "Player A has the ball," the server generates the comment "Player A is dribbling the ball." The input is the JSON data of the analysis results, and the output is play-by-play comments in natural language. The server creates appropriate prompts for the generative AI model and inputs them into the model.

[0580] Step 5:

[0581] The user's device uses a camera, microphone, and biometric sensors to collect the user's facial expressions, voice, heart rate, and other emotional data. The input is the user's various sensor data, and the output is emotional data sent to the server via an HTTP request. The device formats the collected data and prepares it to be sent to the server.

[0582] Step 6:

[0583] The server inputs the received emotional data into the emotion engine for analysis. The emotion engine determines whether the user is happy, excited, or sad. The input is the emotional data sent by the user, and the output is the analysis result (in JSON format) about the user's emotional state. The server uses the emotion engine to analyze the data using facial expression recognition models and voice analysis models.

[0584] Step 7:

[0585] The server readjusts the commentary based on the analysis results of the emotion engine. For example, if the user is excited, the tone of the commentary can be adjusted to be more energetic, such as "Player A is moving forward with a great dribble!" The input is the analysis results of the emotion engine and the initial commentary, and the output is the adjusted commentary. The server again inputs the prompt into the generative AI model to generate the adjusted commentary.

[0586] Step 8:

[0587] The server sends the final commentary to the user's device as an HTTP response. The input is the adjusted commentary, and the output is the HTTP response it sends to the device. The server sets the appropriate response headers and sends the commentary.

[0588] Step 9:

[0589] The device displays the received live commentary to the user. Specifically, it displays the live commentary in real time using a dedicated UI component (e.g., text view). The input is the live commentary obtained from the HTTP response, and the output is the live commentary displayed on the screen. The device performs UI update processing to provide the user with a real-time live commentary experience.

[0590] (Application example 2)

[0591] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0592] In autonomous vehicles, it is difficult to grasp the driver's emotional state in real time and provide appropriate information and alerts while driving. In addition, there is a lack of technology to personalize road information and driving advice according to the driver's emotions.

[0593] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for uploading images from a user's terminal, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating a commentary related to a specified event based on the analysis results, means for recognizing the user's emotional state, means for adjusting the content and tone of the commentary based on the recognized emotional state, means for transmitting the generated commentary to the user's terminal, and means for displaying the transmitted commentary on the user's terminal. This makes it possible to analyze the driver's emotions in real time and provide commentary and situation reports accordingly.

[0594] "Means for uploading images from the user's terminal" refers to an interface or function that allows the user to select image data from their own terminal and transfer it to the system.

[0595] "Means for analyzing uploaded images and identifying objects and movements within the images" refers to technology for analyzing received image data and extracting information about objects present within the images and their movements.

[0596] "Means for generating live commentary related to a specified event based on the analysis results" refers to a function that automatically generates commentary and narration related to a specific event based on the results of image analysis.

[0597] "Means for recognizing the user's emotional state" refers to technology that uses sensors such as cameras and microphones to identify emotions from the user's facial expressions and voice.

[0598] "Means for adjusting the content and tone of live commentary based on the recognized emotional state" refers to a technology that utilizes the user's emotional information to change the wording and expression of the live commentary that is generated.

[0599] The "means for transmitting the generated live commentary to the user's terminal" is a function for transferring the generated live commentary to the user's terminal via a communication means such as the Internet.

[0600] "Means for displaying the transmitted live commentary on the user's terminal" refers to a function for displaying and playing the received live commentary on the user's terminal by screen or audio.

[0601] This invention is a system that recognizes a user's emotions and generates and provides live commentary in real time based on those emotions. A specific example for realizing this invention is shown below.

[0602] 1. Upload an image

[0603] Users upload images to the system using their own devices. To upload images, a dedicated application or a web interface is used. The user selects an image file and taps the upload button, which sends the image data from the device to the server.

[0604] 2. Image Analysis

[0605] The server receives and analyzes images uploaded by users. This analysis uses an object detection model, a machine learning model. This model is pre-trained and can detect objects and their movements in images with high accuracy. For example, in an image of a soccer game, it can recognize players, the ball, goals, etc.

[0606] 3. Generating live commentary

[0607] Based on the analysis results, the server generates commentary using natural language generation technology. Examples of generative AI models used include GPT-3. The following prompt sentence is input to this model to generate commentary:

[0608] Road conditions: Congested. Driver's emotional state: Tired. Commentary:

[0609] 4. Emotion recognition

[0610] The server uses an emotion recognition engine to recognize the user's emotions in real time. This engine analyzes the user's facial expressions and voice data to identify their emotional state. Cameras and microphones installed in the vehicle are used to acquire the data.

[0611] 5. Adjustments to commentary

[0612] Based on the emotional information obtained by the emotion recognition engine, the server adjusts the content and tone of the live commentary. For example, if the user is tired, it uses expressions that encourage relaxation.

[0613] 6. Output and display of live commentary

[0614] The generated commentary is sent from the server to the user's device, which displays the commentary on the screen or plays it back as audio. This allows the user to receive customized commentary in real time, tailored to their emotional state.

[0615] Specific examples

[0616] soccer match

[0617] A user uploads an image of a soccer game, and the analysis results provide information such as "Player A is dribbling the ball." Furthermore, if the user is recognized as being tired, a commentary using the generative AI model is generated, such as "Player A is moving forward with a relaxed dribble!", and displayed on the user's device.

[0618] Autonomous vehicles

[0619] If the driver is tired, the road conditions are analyzed as "congested." The following prompt sentence is input into the generative AI model to generate a running commentary.

[0620] Road conditions: Congested. Driver's emotional state: Tired. Commentary:

[0621] The generated comment will be "The roads are currently congested. Please relax and continue driving," and will be personalized according to the driver's emotions.

[0622] In this way, the system combines image analysis and emotion recognition to provide users with more personalized, high-quality information.

[0623] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0624] Step 1:

[0625] Users upload images to the system through their own devices. When a user selects an image file and taps the upload button, the image data is sent from the device to the server as an HTTP request. The input is the image data from the user device, and the output is the transfer of the image data to the server.

[0626] Step 2:

[0627] The server analyzes the image data received from the user. For image analysis, it uses an object detection model, which is a machine learning model. Specifically, it analyzes the information of each pixel in the image and identifies the main objects and their movements. The input is the image data uploaded by the user, and the output is the detection results of the objects and movements.

[0628] Step 3:

[0629] The server generates commentary based on the analysis results using the generative AI model GPT-3. During the generation process, contextual information from the analysis results is input as a prompt, and the model generates commentary in natural language format. The input is the image analysis results and the prompt, and the output is the generated commentary.

[0630] Step 4:

[0631] The server recognizes the user's emotional state in real time. It uses an emotion recognition engine to analyze the user's facial and voice data from a camera and microphone to identify the user's emotional state. The input is the facial and voice data obtained from the user, and the output is emotional state information from the emotion recognition engine.

[0632] Step 5:

[0633] The server adjusts the content and tone of the commentary based on the user's emotional state. It also changes the wording and tone of the commentary based on the emotional data obtained from the emotion recognition engine. The input is the emotional state information and the commentary generated in the initial stage, and the output is the adjusted commentary.

[0634] Step 6:

[0635] The server sends the adjusted live commentary to the user's device. The live commentary is delivered to the user's device using an HTTP response as the communication method. The input is the adjusted live commentary, and the output is the comment sent to the user's device.

[0636] Step 7:

[0637] The terminal displays the received live commentary. The live commentary is provided to the user in real time in the form of a screen display or audio playback. The input is the live commentary received from the server, and the output is information provided to the user.

[0638] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0639] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0640] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0641] [Third embodiment]

[0642] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0643] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0644] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0645] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0646] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0647] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0648] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0649] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0650] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0651] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0652] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0653] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0654] This invention relates to an automated system that analyzes the status of an image in real time and generates and provides appropriate live commentary. The system mainly comprises a user terminal, a server, and a means for communicating between them. The details of the embodiment are described below.

[0655] 1. Image input from the user

[0656] A user uploads an image to the system using a device. The user selects an image file using a specific web interface or application and taps the upload button. This action causes the device to send the image file to the server as an HTTP request.

[0657] 2. Image analysis on the server

[0658] After receiving the uploaded image file, the server performs image analysis. Here, an object detection model is used to identify the main objects in the image (e.g., athletes, balls, etc.). The object detection model uses a pre-trained machine learning model that can recognize the type and location of objects with high accuracy. This allows detailed information about each element in the image to be obtained.

[0659] 3. Live comment generation on the server

[0660] The server generates commentary based on the analysis results using natural language generation technology. For example, when an image of a soccer game is input, analysis can yield information such as "Player A is dribbling the ball." Based on this information, the generation AI generates commentary such as "Player A is dribbling the ball."

[0661] 4. Output and display of live commentary

[0662] The generated commentary is sent from the server to the user's device. The server returns the commentary as an HTTP response. The device uses a dedicated UI component (e.g., text view) to display the received commentary to the user. This allows the user to experience high-quality live broadcasting in real time.

[0663] Specific examples

[0664] Example: Soccer image

[0665] 1. A user uploads images of a soccer match from their device to the system.

[0666] The user uses the image selection interface to select an image and then taps the upload button.

[0667] 2. The server receives the uploaded images and performs image analysis, using object detection models to identify players, the ball, goals, etc.

[0668] 3. Based on the analysis results, the server obtains the information that "Player A has the ball." Based on this information, the generation AI generates a commentary such as "Player A is dribbling the ball."

[0669] 4. The server sends the generated commentary to the user's device, which then displays it to the user, allowing the user to understand the situation of the game in real time.

[0670] This system allows users to receive accurate live coverage in real time, eliminating the need for specialized announcers and significantly reducing costs and effort. At the same time, it is expected to improve the user experience.

[0671] The processing flow will be explained below.

[0672] Step 1:

[0673] The user selects an image from the device and clicks the upload button. This operation causes the user device to attach the selected image file to an HTTP request and send it to the server. Specifically, when an image file is selected using the image selection interface and the send button is pressed, the device generates and sends an HTTP POST request including the image data to the server.

[0674] Step 2:

[0675] The server saves the received image file. The server analyzes the HTTP request, extracts the file data, and saves the image file in the specified directory. This makes the image to be analyzed accessible on the server.

[0676] Step 3:

[0677] The server analyzes the saved images. The server loads an object detection model and analyzes the saved images to identify key objects and movements within the images. Specifically, the image is input into the object detection model to detect the position and type of players, balls, etc. This provides detailed information about the scene within the image.

[0678] Step 4:

[0679] The server generates commentary based on the analysis results. The server uses natural language generation technology to generate appropriate commentary based on the analyzed objects and their movements. For example, based on the analysis result that "Player A has the ball," the server generates the commentary "Player A is dribbling the ball."

[0680] Step 5:

[0681] The server sends the generated commentary to the user's terminal. The server returns the generated commentary to the user's terminal as an HTTP response. This response includes the generated commentary in text format.

[0682] Step 6:

[0683] The device displays the commentary received to the user. The device receives the HTTP response and displays the commentary in a UI component on the screen. Specifically, it uses a user interface such as a text view to display the commentary in a format that is easy for the user to read.

[0684] This allows users to experience high-quality live broadcasts based on real-time images.

[0685] Example 1

[0686] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0687] In recent years, there has been a demand for real-time information provision for sports and event broadcasts, but deploying professional announcers and facilitators is costly and difficult. This invention aims to solve this problem by providing high-quality commentary in real time through an automated system. In particular, by combining image analysis and natural language generation technology, it aims to provide users with an experience similar to that of professional commentary.

[0688] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0689] In this invention, the server includes means for uploading images from a user's terminal, means for receiving and saving the uploaded images, means for analyzing the saved images using an object detection model, means for generating live commentary comments using natural language generation technology based on the analysis results, means for transmitting the generated live commentary comments from the server to the user's terminal, and means for displaying the transmitted live commentary comments on the user's terminal, thereby enabling automatic provision of high-quality live commentary comments in real time.

[0690] A "user's terminal" is a device used by a user to upload images, and includes, for example, a smartphone, a personal computer, a tablet terminal, etc.

[0691] "Server" refers to a central processing unit that receives and stores images sent from user terminals, and further analyzes and generates live comments.

[0692] "Upload" refers to the process of sending image data from a user's device to a server.

[0693] "Image analysis" refers to the process by which the server uses object detection models to identify objects and movements in uploaded images.

[0694] An "object detection model" is a machine learning model used to identify dominant objects in an image, and can use techniques such as YOLO or SSD.

[0695] "Natural language generation technology" refers to technology for generating sentences in a form that can be understood by humans based on information obtained from analysis, and examples include generative AI models (such as GPT-3).

[0696] "Live commentary" refers to explanatory text about movements and events within an image that is generated based on the analysis results.

[0697] A "generative AI model" is an artificial intelligence model used to generate natural language text from a given prompt and analysis results.

[0698] A "prompt sentence" refers to a sentence that is input into a generative AI model to instruct it on what kind of text to generate.

[0699] An "HTTP request" refers to a form of communication protocol that requests data transmission from a user terminal to a server.

[0700] "HTTP response" refers to a form of communication protocol that returns data from a server to a user's terminal.

[0701] MODE FOR CARRYING OUT THE INVENTION

[0702] The present invention relates to a system in which a user uploads an image from a terminal, a server analyzes the image, and generates live commentary in real time. The present invention is comprised of the following steps as an embodiment.

[0703] Required Hardware and Software

[0704] User devices: smartphones, computers, tablets, etc.

[0705] Server: Central Processing Unit

[0706] Web interface or application: Software for uploading images

[0707] Object detection models: YOLO, SSD, etc.

[0708] Generative AI models: GPT-3, BERT, etc.

[0709] Communication method: HTTP protocol

[0710] Specific processing and data calculations

[0711] 1. Image input from the user

[0712] Users use devices such as smartphones or PCs to select and upload images through a web interface or application. When they tap the upload button, the device sends the image data to the server as an HTTP POST request.

[0713] 2. Receiving images on the server

[0714] The server receives the HTTP POST request sent from the user's device and saves the image file. This data is stored in the server's temporary directory.

[0715] 3. Image analysis on the server

[0716] The server analyzes the stored image files, using a pre-trained object detection model (e.g., YOLO or SSD) to identify the dominant objects in the image. The analysis results in the location and type of object in the image. This information is stored in an internal data structure.

[0717] 4. Live comment generation on the server

[0718] The server generates commentary based on the results of image analysis. Specifically, it extracts key information from the analysis (e.g., "Player A has the ball") and inputs the prompt and analysis results into a generative AI model (e.g., GPT-3). The generated commentary is then translated into natural language.

[0719] Example prompt sentence:

[0720] "Image analysis results of a soccer game: Player A has the ball. Please generate commentary based on this information."

[0721] 5. Sending live commentary from the server to the device

[0722] The server sends the generated live commentary to the user's terminal. The server creates a live commentary as an HTTP response and sends it to the user's terminal.

[0723] 6. Displaying live comments on user devices

[0724] The user's device displays the received live commentary. It receives the HTTP response, extracts the live commentary from the response body, and displays it to the user in real time using a dedicated UI component (e.g., text view).

[0725] In this way, the user can experience high quality real-time live broadcasting. The above is a specific embodiment for carrying out the present invention.

[0726] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0727] Step 1:

[0728] A user uploads an image using a device such as a smartphone or PC. Specifically, they first open a web interface or application, click or tap the image selection button to select an image file from the device's storage, and then click or tap the upload button. This causes the device to send the selected image file to the server as an HTTP POST request.

[0729] Input: An image file selected by the user

[0730] Output: Image data is sent to the server as an HTTP POST request

[0731] Step 2:

[0732] The server receives the HTTP POST request sent from the user's device and saves the image file. Specifically, it extracts the image data from the body of the received request and saves it in a temporary directory on the server.

[0733] Input: Image data sent as an HTTP POST request

[0734] Output: Image file saved in a temporary directory on the server

[0735] Step 3:

[0736] The server analyzes the stored image files and identifies the dominant object in the image using a pre-trained object detection model (e.g., YOLO or SSD). Specifically, the image data is input to the object detection model, and the location and type of the dominant object are obtained as output. The analysis results are stored in an internal data structure.

[0737] Input: Image file saved in the temporary directory on the server

[0738] Output: Analysis results on the location and type of objects in the image

[0739] Step 4:

[0740] The server generates commentary based on the results of image analysis. Specifically, it extracts key information from the analysis results (e.g., "Player A has the ball") and inputs the prompt and analysis results into a generative AI model (e.g., GPT-3). The generative AI model generates commentary in natural language based on the input.

[0741] Input: Analysis results obtained from the object detection model

[0742] Output: Generated natural language commentary

[0743] Example prompt sentence:

[0744] "Image analysis results of a soccer game: Player A has the ball. Please generate commentary based on this information."

[0745] Step 5:

[0746] The server sends the generated commentary to the user's terminal. Specifically, the server creates the commentary as an HTTP response and sends it to the user's terminal.

[0747] Input: Generated commentary

[0748] Output: Commentary sent to the user's device as an HTTP response

[0749] Step 6:

[0750] The user's device displays the received live commentary. Specifically, it receives the HTTP response, extracts the live commentary from the response body, and displays it to the user using a dedicated UI component (e.g., text view).

[0751] Input: Comments sent as an HTTP response

[0752] Output: Commentary displayed on the user's device

[0753] The above is the detailed flow of the processing of this program.

[0754] (Application example 1)

[0755] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0756] Conventional factory monitoring systems have made it difficult to quickly grasp the operating status and abnormalities of equipment, resulting in delayed response when an abnormality occurs. Furthermore, constant monitoring by an operator with specialized knowledge is required, which increases costs. The present invention aims to solve these problems by providing a system that can monitor the status of equipment in a factory in real time and quickly take appropriate action.

[0757] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0758] In this invention, the server includes means for uploading images from a user's device, means for analyzing the uploaded images and identifying objects and actions in the images, means for generating commentary related to a specified situation based on the analysis results, means for transmitting the generated commentary to the user's device, means for displaying the transmitted commentary on the user's device, and means for analyzing camera footage within the factory in real time and automatically generating and notifying the status and abnormalities of equipment as commentary, thereby enabling factory supervisors and operators to grasp the status of equipment in real time and take appropriate action promptly.

[0759] An "image" is digital data containing visual information that is uploaded from a user's device.

[0760] "User's Device" is the terminal equipment used to upload images to the system.

[0761] "Uploading" is the act of sending an image from a user's device to a server.

[0762] "Analysis" is the process of processing uploaded images to identify objects and actions within them.

[0763] "Target" refers to an item to be detected, such as an object, person, or device present in an image.

[0764] "Action" refers to a specific action or change of state performed by an object in an image.

[0765] The "specified situation" refers to the state of an object or action that the system determines based on the analysis results.

[0766] "Live commentary" is a natural language description of an object or action that is generated based on the analysis results.

[0767] "Generation" is the process of creating live commentary based on the analysis results.

[0768] "Transmitting" refers to the act of transferring the generated live commentary to the user's device using a communication means.

[0769] "Display" refers to the act of visually showing the live commentary sent on the user's device.

[0770] "Camera footage" refers to real-time video data captured by cameras installed within the factory.

[0771] "Equipment status" refers to the operating status and operational status of various equipment within the factory.

[0772] "Abnormality" refers to a state in which the equipment is not operating normally or an error has occurred.

[0773] "Notification" refers to the act of notifying the user of the generated live commentary in real time on the user's device.

[0774] This invention realizes a system that analyzes camera images in a factory in real time and automatically generates and notifies the status and abnormalities of equipment as live comments. A specific embodiment of this system is described below.

[0775] Image input from the user

[0776] Users transmit real-time images from cameras installed in the factory to the server, which provides high-resolution images that are then received in real time by the user's device (such as a tablet or PC).

[0777] Image analysis on the server

[0778] The server uses an object detection model to analyze the video received from the user's device in real time. This object detection model applies a pre-trained model using a deep learning framework such as TensorFlow. The server can accurately analyze the operating status of the equipment in the image (e.g., operating normally, stopped, abnormality, etc.).

[0779] Live commentary generation on the server

[0780] Based on the analysis results, the server generates a commentary in natural language using a generative AI model such as GPT-3. For example, if the server obtains the analysis result "The conveyor belt has stopped," the generative AI model will generate a specific commentary such as "The conveyor belt has stopped. Please check error code 105."

[0781] Output and display of live commentary

[0782] The server sends the generated commentary to the user's device. The commentary is returned as an HTTP response and displayed in real time on the user's device. A dedicated UI component (e.g., a text view) is used to convey visual information to the user.

[0783] Hardware and software used

[0784] Hardware: Camera (high-resolution industrial camera), server (with high-performance CPU and GPU), user device (tablet, PC, etc.)

[0785] Software: Python, TensorFlow, Transformers library, OpenCV, HTTP communication library

[0786] Specific examples

[0787] As a specific example, camera footage from inside a factory is analyzed, and the server analyzes footage of a stopped conveyor belt. In this case, the server uses an object detection model to detect the conveyor belt stopping, and then uses a generative AI model to generate a live commentary such as "The conveyor belt has stopped. Please check error code 105." The generated live commentary is sent to the user's device and displayed in real time.

[0788] Prompt Sentence Examples

[0789] An example of a prompt sentence to input to the generative AI model is as follows:

[0790] Generate a comment for the following machine state: stopped.

[0791] An example of the commentary that is output is as follows:

[0792] The machine has stopped. Please check Error Code 105.

[0793] This invention enables factory supervisors and operators to grasp the status of equipment in real time and take appropriate action quickly.

[0794] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0795] Step 1:

[0796] Users send real-time video to a server via cameras installed in their factories. Specifically, the video captured by the camera is received by the user's device (tablet, PC, etc.) and sent to the server as an HTTP request. The input is the real-time camera video, and the output is the video data sent to the server.

[0797] Step 2:

[0798] The server analyzes video data received from the user's device in real time. This analysis uses an object detection model using a deep learning framework such as TensorFlow. The input is camera video, which is analyzed to detect and identify the operating status of equipment (e.g., conveyor belts). The output is the identified operating status information.

[0799] Step 3:

[0800] Based on the analysis results, the server uses a generative AI model such as GPT-3 to generate commentary in natural language. The input is the analysis result (e.g., conveyor belt stopped), and a prompt sentence is generated based on this information. The generative AI model responds to the prompt sentence and generates an appropriate commentary. The output is the generated commentary.

[0801] Step 4:

[0802] The server transmits the generated commentary to the user's device. Specifically, the server returns the generated commentary to the user's device as an HTTP response. The input is the generated commentary, and the output is the commentary to be transmitted to the user's device.

[0803] Step 5:

[0804] The user's device receives the commentary and displays it in real time. Specifically, the commentary received as an HTTP response is displayed using a dedicated UI component such as a text view. The input is the commentary sent from the server, and the output is the visually displayed commentary.

[0805] The specific actions taken at each step enable a series of processes to work together seamlessly, enabling real-time monitoring of the status and abnormalities of equipment within the factory and appropriate notification.

[0806] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0807] This invention relates to an image analysis and commentary generation system that combines an emotion engine that recognizes user emotions. The main components of the system are a user terminal, a server, an emotion engine, and a means for communication between them. The following describes the details of the embodiment.

[0808] 1. Image input from the user

[0809] A user uploads an image to the system using a device. The user selects an image file using a specific web interface or application and taps the upload button. This action causes the device to send the image file to the server as an HTTP request.

[0810] 2. Image analysis on the server

[0811] The server stores the received image files and performs image analysis, using an object detection model to identify key objects in the image (e.g., athletes, balls, etc.). The object detection model uses a pre-trained machine learning model and can recognize the type and location of objects with high accuracy.

[0812] 3. Live comment generation on the server

[0813] The server generates commentary based on the analysis results using natural language generation technology. For example, when an image of a soccer game is input, analysis can yield information such as "Player A is dribbling the ball." Based on this information, the generation AI generates commentary such as "Player A is dribbling the ball."

[0814] 4. Emotion Recognition on the Server

[0815] The server uses an emotion engine to recognize the user's emotional state. The emotion engine can identify emotions by analyzing the user's facial expressions, voice, or biometric information. This information is obtained through a camera, microphone, or biometric sensor.

[0816] 5. Moderating commentary based on emotions

[0817] The server automatically adjusts the content and expression of the commentary based on the user's emotional state obtained from the emotion engine. For example, if the user is excited, the server can make the commentary more energetic.

[0818] 6. Output and display of live commentary

[0819] The generated commentary is sent from the server to the user's device. The server returns the commentary as an HTTP response. The device uses a dedicated UI component (e.g., text view) to display the received commentary to the user. This allows the user to experience high-quality live broadcasting in real time.

[0820] Specific examples

[0821] Example: Soccer image

[0822] 1. A user uploads soccer game images from their device to the system. The user uses the image selection interface to select an image and then taps the upload button.

[0823] 2. The server receives the uploaded images and performs image analysis, using object detection models to identify players, the ball, goals, etc.

[0824] 3. Based on the analysis results, the server obtains the information that "Player A has the ball." Based on this information, the generation AI generates a commentary such as "Player A is dribbling the ball."

[0825] 4. The server uses the camera images and audio data acquired from the user's device to analyze the user's emotions using an emotion engine. For example, if the user is smiling, it will determine that the user is excited.

[0826] 5. The server adjusts the commentary based on the user's emotional state to "Player A is moving forward with a great dribble!"

[0827] 6. The server sends the generated commentary to the user's device, which then displays the commentary to the user, allowing the user to grasp the game situation in real time and enjoy a commentary experience that suits their emotions.

[0828] This system allows users to not only receive accurate commentary in real time, but also experience commentary that is customized according to their emotional state, providing a more personalized and high-quality user experience.

[0829] The processing flow will be explained below.

[0830] Step 1:

[0831] The user selects an image from the device and clicks the upload button. This operation causes the user device to attach the selected image file to an HTTP request and send it to the server. Specifically, the user uses the image selection interface to select an image file, and then presses the send button, causing the device to send the image data to the server via an HTTP POST request.

[0832] Step 2:

[0833] The server saves the received image file. The server analyzes the HTTP request, extracts the sent file data, and saves it as an image file in the specified directory. This operation makes the image to be analyzed accessible on the server.

[0834] Step 3:

[0835] The server analyzes the saved images. The server loads an object detection model and analyzes the saved images to identify key objects and movements within the images. Specifically, the image is input into the object detection model to detect the location and type of objects, such as players and the ball.

[0836] Step 4:

[0837] The server generates commentary based on the analysis results. The server generates commentary using natural language generation technology based on the analyzed objects and their movements. For example, if the analysis result is "Player A is dribbling the ball," the server generates commentary such as "Player A is dribbling the ball."

[0838] Step 5:

[0839] The user transmits emotion data from the device. The user uses the device's camera and microphone to transmit their facial expressions and voice to the system. The device then transmits this data to the server in real time.

[0840] Step 6:

[0841] The server analyzes the user's emotions using an emotion engine. The server inputs the received user's facial expressions and voice into the emotion engine to determine the user's current emotional state. For example, if the server receives image data of the user smiling, it determines that the user is excited.

[0842] Step 7:

[0843] The server adjusts the commentary based on the user's emotional state. The tone and content of the commentary is changed based on the emotional information obtained from the emotion engine. For example, if the user is excited, the commentary will be adjusted from "Player A is dribbling the ball" to an energetic expression such as "Player A is moving forward with a great dribble!"

[0844] Step 8:

[0845] The server sends the adjusted live commentary to the user's terminal. The server sends the adjusted live commentary to the user's terminal as an HTTP response.

[0846] Step 9:

[0847] The device displays the commentary received to the user. The device receives the HTTP response and displays the commentary in a UI component on the screen. Specifically, the device uses an interface such as a text view to display the commentary in a format that is easy for the user to see.

[0848] This process flow allows users to not only receive accurate live coverage in real time, but also enjoy a live experience that is customized according to their emotional state.

[0849] Example 2

[0850] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0851] Conventional image analysis systems were able to generate live commentary related to a specific event based on image information provided by the user, but were unable to provide personalized live commentary that took the user's emotional state into account. This resulted in a uniform user experience, making it difficult to provide high-quality live coverage that was tailored to each individual user's emotional state. The present invention aims to solve this problem by generating live commentary customized according to the user's emotional state, thereby providing a personalized, high-quality live experience.

[0852] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0853] In this invention, the server includes means for uploading images from a user's terminal, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating live commentary related to a specified event based on the analysis results, means for transmitting the generated live commentary to the user's terminal, means for displaying the transmitted live commentary on the user's terminal, means for acquiring emotional data for recognizing the user's emotional state, and means for adjusting the content and expression of the live commentary based on the acquired emotional data. This makes it possible to provide customized live commentary that takes the user's emotional state into consideration in real time, thereby providing a more personalized and high-quality live commentary experience.

[0854] "Means for uploading images from the user's terminal" is a function that allows the user to send image files from their own terminal to the server.

[0855] "Means of analyzing uploaded images and identifying objects and movements within the images" refers to a function that uses image analysis technology on received images to recognize and identify objects and their movements within the images.

[0856] The "means for generating live commentary related to a specified event based on the analysis results" is a function that automatically generates sentences related to an event based on data obtained from image analysis.

[0857] The "means for transmitting the generated live commentary to the user's terminal" is a function for transmitting the automatically generated live commentary from the server to the user's terminal.

[0858] The "means for displaying the transmitted live commentary on the user's terminal" is a function for displaying the live commentary received on the user's terminal.

[0859] The "means for acquiring emotional data to recognize the user's emotional state" is a function for collecting emotional data such as the user's facial expressions and voice through sensors, cameras, etc.

[0860] The "means for adjusting the content and expression of the commentary commentary based on the acquired emotional data" is a function for analyzing the collected emotional data and changing the tone and content of the commentary commentary based on the results.

[0861] The present invention relates to an image analysis and commentary generation system that combines an emotion engine that recognizes user emotions. The following describes in detail the embodiments of the present invention.

[0862] The system of the present invention is mainly composed of a user terminal, a server, an emotion engine, and means for communicating among them.

[0863] Hardware and software used

[0864] The user's device may be a smartphone, tablet, or PC, and may be equipped with a camera, microphone, and, if necessary, biometric sensors. The server may be a cloud environment or a dedicated computer system, and includes the following main software components:

[0865] 1. Image analysis module: This module implements an object detection model to identify objects and movements in images, using computer vision libraries (e.g., OpenCV, TensorFlow).

[0866] 2. Natural Language Generation Module: Uses a generative AI model (e.g., GPT-3) to generate commentary based on the analysis results.

[0867] 3. Emotion Engine: Machine learning models are implemented to recognize the user's emotional state, using facial recognition and voice analysis models.

[0868] 4. Communication module: Processes HTTP requests and responses to send and receive data between the user's device and the server.

[0869] Example of a system

[0870] Specific example using soccer images

[0871] 1. A user uses a device to upload an image of a soccer match to the system. The user uses the image selection interface to select an image and then taps the upload button. This action sends the image file to the server as an HTTP request.

[0872] 2. The server stores the received image files and uses an image analysis module to identify objects in the images (e.g., players, ball, goal). The object detection model utilizes computer vision libraries to perform highly accurate analysis.

[0873] 3. Based on the results of the image analysis, the server uses a natural language generation module to generate a commentary such as "Player A is dribbling the ball" based on the information that "Player A has the ball." The following prompt is input to the generation AI model:

[0874] "Enter a football image."

[0875] "Analyze your emotional state and energetically adjust the tone of the play-by-play commentary you generate."

[0876] 4. The user's device uses a camera, microphone, and biometric sensors to acquire emotional data such as the user's facial expressions, voice, and heart rate, and sends this data to the server as an HTTP request.

[0877] 5. The server uses the emotion engine to analyze the received emotion data and identify the user's emotional state. For example, if the user is smiling, it determines that the user is excited.

[0878] 6. Based on the results of the emotion engine, the server adjusts the generated commentary to "Player A is moving forward with a great dribble!" This adjusted commentary is sent to the user's device as an HTTP response.

[0879] 7. The device displays the received live commentary using a dedicated UI component (e.g., text view), allowing the user to experience high-quality live broadcasts in real time.

[0880] Through the above-described method, users can obtain commentary that is customized according to their emotional state, and can enjoy a more personalized and high-quality commentary experience.

[0881] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0882] Step 1:

[0883] A user uses a device to open a specific web interface or application. They select an image file and tap the upload button to send the image to the system. The input is an image file stored on the device, and the output is that the image file is sent to the server as an HTTP request. The device generates header information for the HTTP request and attaches the image file.

[0884] Step 2:

[0885] The server receives the HTTP request. It saves the received image file and generates a path for the save location. Here, it checks the image format and size, and converts the file format or adjusts the size as necessary. The input is the image file sent from the terminal, and the output is the image file path saved in a specific directory on the server. The server extracts the image file data from the HTTP request and saves it in the specified directory.

[0886] Step 3:

[0887] The server inputs the saved image file into an object detection model and performs analysis. It uses an image analysis module to identify objects and movements and outputs the analysis results in JSON format. The input is the saved image file path, and the output is the JSON data of the analysis results. The server uses an object detection model (e.g., OpenCV, TensorFlow) to process and recognize objects and movements in the image.

[0888] Step 4:

[0889] Based on the analysis results, the server inputs the results as prompts into a generative AI model (e.g., GPT-3) to generate play-by-play comments. For example, based on the result "Player A has the ball," the server generates the comment "Player A is dribbling the ball." The input is the JSON data of the analysis results, and the output is play-by-play comments in natural language. The server creates appropriate prompts for the generative AI model and inputs them into the model.

[0890] Step 5:

[0891] The user's device uses a camera, microphone, and biometric sensors to collect the user's facial expressions, voice, heart rate, and other emotional data. The input is the user's various sensor data, and the output is emotional data sent to the server via an HTTP request. The device formats the collected data and prepares it to be sent to the server.

[0892] Step 6:

[0893] The server inputs the received emotional data into the emotion engine for analysis. The emotion engine determines whether the user is happy, excited, or sad. The input is the emotional data sent by the user, and the output is the analysis result (in JSON format) about the user's emotional state. The server uses the emotion engine to analyze the data using facial expression recognition models and voice analysis models.

[0894] Step 7:

[0895] The server readjusts the commentary based on the analysis results of the emotion engine. For example, if the user is excited, the tone of the commentary can be adjusted to be more energetic, such as "Player A is moving forward with a great dribble!" The input is the analysis results of the emotion engine and the initial commentary, and the output is the adjusted commentary. The server again inputs the prompt into the generative AI model to generate the adjusted commentary.

[0896] Step 8:

[0897] The server sends the final commentary to the user's device as an HTTP response. The input is the adjusted commentary, and the output is the HTTP response it sends to the device. The server sets the appropriate response headers and sends the commentary.

[0898] Step 9:

[0899] The device displays the received live commentary to the user. Specifically, it displays the live commentary in real time using a dedicated UI component (e.g., text view). The input is the live commentary obtained from the HTTP response, and the output is the live commentary displayed on the screen. The device performs UI update processing to provide the user with a real-time live commentary experience.

[0900] (Application example 2)

[0901] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0902] In autonomous vehicles, it is difficult to grasp the driver's emotional state in real time and provide appropriate information and alerts while driving. In addition, there is a lack of technology to personalize road information and driving advice according to the driver's emotions.

[0903] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for uploading images from a user's terminal, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating a commentary related to a specified event based on the analysis results, means for recognizing the user's emotional state, means for adjusting the content and tone of the commentary based on the recognized emotional state, means for transmitting the generated commentary to the user's terminal, and means for displaying the transmitted commentary on the user's terminal. This makes it possible to analyze the driver's emotions in real time and provide commentary and situation reports accordingly.

[0904] "Means for uploading images from the user's terminal" refers to an interface or function that allows the user to select image data from their own terminal and transfer it to the system.

[0905] "Means for analyzing uploaded images and identifying objects and movements within the images" refers to technology for analyzing received image data and extracting information about objects present within the images and their movements.

[0906] "Means for generating live commentary related to a specified event based on the analysis results" refers to a function that automatically generates commentary and narration related to a specific event based on the results of image analysis.

[0907] "Means for recognizing the user's emotional state" refers to technology that uses sensors such as cameras and microphones to identify emotions from the user's facial expressions and voice.

[0908] "Means for adjusting the content and tone of live commentary based on the recognized emotional state" refers to a technology that utilizes the user's emotional information to change the wording and expression of the live commentary that is generated.

[0909] The "means for transmitting the generated live commentary to the user's terminal" is a function for transferring the generated live commentary to the user's terminal via a communication means such as the Internet.

[0910] "Means for displaying the transmitted live commentary on the user's terminal" refers to a function for displaying and playing the received live commentary on the user's terminal by screen or audio.

[0911] This invention is a system that recognizes a user's emotions and generates and provides live commentary in real time based on those emotions. A specific example for realizing this invention is shown below.

[0912] 1. Upload an image

[0913] Users upload images to the system using their own devices. To upload images, a dedicated application or a web interface is used. The user selects an image file and taps the upload button, which sends the image data from the device to the server.

[0914] 2. Image Analysis

[0915] The server receives and analyzes images uploaded by users. This analysis uses an object detection model, a machine learning model. This model is pre-trained and can detect objects and their movements in images with high accuracy. For example, in an image of a soccer game, it can recognize players, the ball, goals, etc.

[0916] 3. Generating live commentary

[0917] Based on the analysis results, the server generates commentary using natural language generation technology. Examples of generative AI models used include GPT-3. The following prompt sentence is input to this model to generate commentary:

[0918] Road conditions: Congested. Driver's emotional state: Tired. Commentary:

[0919] 4. Emotion recognition

[0920] The server uses an emotion recognition engine to recognize the user's emotions in real time. This engine analyzes the user's facial expressions and voice data to identify their emotional state. Cameras and microphones installed in the vehicle are used to acquire the data.

[0921] 5. Adjustments to commentary

[0922] Based on the emotional information obtained by the emotion recognition engine, the server adjusts the content and tone of the live commentary. For example, if the user is tired, it uses expressions that encourage relaxation.

[0923] 6. Output and display of live commentary

[0924] The generated commentary is sent from the server to the user's device, which displays the commentary on the screen or plays it back as audio. This allows the user to receive customized commentary in real time, tailored to their emotional state.

[0925] Specific examples

[0926] soccer match

[0927] A user uploads an image of a soccer game, and the analysis results provide information such as "Player A is dribbling the ball." Furthermore, if the user is recognized as being tired, a commentary using the generative AI model is generated, such as "Player A is moving forward with a relaxed dribble!", and displayed on the user's device.

[0928] Autonomous vehicles

[0929] If the driver is tired, the road conditions are analyzed as "congested." The following prompt sentence is input into the generative AI model to generate a running commentary.

[0930] Road conditions: Congested. Driver's emotional state: Tired. Commentary:

[0931] The generated comment will be "The roads are currently congested. Please relax and continue driving," and will be personalized according to the driver's emotions.

[0932] In this way, the system combines image analysis and emotion recognition to provide users with more personalized, high-quality information.

[0933] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0934] Step 1:

[0935] Users upload images to the system through their own devices. When a user selects an image file and taps the upload button, the image data is sent from the device to the server as an HTTP request. The input is the image data from the user device, and the output is the transfer of the image data to the server.

[0936] Step 2:

[0937] The server analyzes the image data received from the user. For image analysis, it uses an object detection model, which is a machine learning model. Specifically, it analyzes the information of each pixel in the image and identifies the main objects and their movements. The input is the image data uploaded by the user, and the output is the detection results of the objects and movements.

[0938] Step 3:

[0939] The server generates commentary based on the analysis results using the generative AI model GPT-3. During the generation process, contextual information from the analysis results is input as a prompt, and the model generates commentary in natural language format. The input is the image analysis results and the prompt, and the output is the generated commentary.

[0940] Step 4:

[0941] The server recognizes the user's emotional state in real time. It uses an emotion recognition engine to analyze the user's facial and voice data from a camera and microphone to identify the user's emotional state. The input is the facial and voice data obtained from the user, and the output is emotional state information from the emotion recognition engine.

[0942] Step 5:

[0943] The server adjusts the content and tone of the commentary based on the user's emotional state. It also changes the wording and tone of the commentary based on the emotional data obtained from the emotion recognition engine. The input is the emotional state information and the commentary generated in the initial stage, and the output is the adjusted commentary.

[0944] Step 6:

[0945] The server sends the adjusted live commentary to the user's device. The live commentary is delivered to the user's device using an HTTP response as the communication method. The input is the adjusted live commentary, and the output is the comment sent to the user's device.

[0946] Step 7:

[0947] The terminal displays the received live commentary. The live commentary is provided to the user in real time in the form of a screen display or audio playback. The input is the live commentary received from the server, and the output is information provided to the user.

[0948] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0949] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0950] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[0951] [Fourth embodiment]

[0952] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0953] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0954] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0955] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[0956] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0957] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0958] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0959] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[0960] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0961] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0962] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0963] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0964] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0965] This invention relates to an automated system that analyzes the status of an image in real time and generates and provides appropriate live commentary. The system mainly comprises a user terminal, a server, and a means for communicating between them. The details of the embodiment are described below.

[0966] 1. Image input from the user

[0967] A user uploads an image to the system using a device. The user selects an image file using a specific web interface or application and taps the upload button. This action causes the device to send the image file to the server as an HTTP request.

[0968] 2. Image analysis on the server

[0969] After receiving the uploaded image file, the server performs image analysis. Here, an object detection model is used to identify the main objects in the image (e.g., athletes, balls, etc.). The object detection model uses a pre-trained machine learning model that can recognize the type and location of objects with high accuracy. This allows detailed information about each element in the image to be obtained.

[0970] 3. Live comment generation on the server

[0971] The server generates commentary based on the analysis results using natural language generation technology. For example, when an image of a soccer game is input, analysis can yield information such as "Player A is dribbling the ball." Based on this information, the generation AI generates commentary such as "Player A is dribbling the ball."

[0972] 4. Output and display of live commentary

[0973] The generated commentary is sent from the server to the user's device. The server returns the commentary as an HTTP response. The device uses a dedicated UI component (e.g., text view) to display the received commentary to the user. This allows the user to experience high-quality live broadcasting in real time.

[0974] Specific examples

[0975] Example: Soccer image

[0976] 1. A user uploads images of a soccer match from their device to the system.

[0977] The user uses the image selection interface to select an image and then taps the upload button.

[0978] 2. The server receives the uploaded images and performs image analysis, using object detection models to identify players, the ball, goals, etc.

[0979] 3. Based on the analysis results, the server obtains the information that "Player A has the ball." Based on this information, the generation AI generates a commentary such as "Player A is dribbling the ball."

[0980] 4. The server sends the generated commentary to the user's device, which then displays it to the user, allowing the user to understand the situation of the game in real time.

[0981] This system allows users to receive accurate live coverage in real time, eliminating the need for specialized announcers and significantly reducing costs and effort. At the same time, it is expected to improve the user experience.

[0982] The processing flow will be explained below.

[0983] Step 1:

[0984] The user selects an image from the device and clicks the upload button. This operation causes the user device to attach the selected image file to an HTTP request and send it to the server. Specifically, when an image file is selected using the image selection interface and the send button is pressed, the device generates and sends an HTTP POST request including the image data to the server.

[0985] Step 2:

[0986] The server saves the received image file. The server analyzes the HTTP request, extracts the file data, and saves the image file in the specified directory. This makes the image to be analyzed accessible on the server.

[0987] Step 3:

[0988] The server analyzes the saved images. The server loads an object detection model and analyzes the saved images to identify key objects and movements within the images. Specifically, the image is input into the object detection model to detect the position and type of players, balls, etc. This provides detailed information about the scene within the image.

[0989] Step 4:

[0990] The server generates commentary based on the analysis results. The server uses natural language generation technology to generate appropriate commentary based on the analyzed objects and their movements. For example, based on the analysis result that "Player A has the ball," the server generates the commentary "Player A is dribbling the ball."

[0991] Step 5:

[0992] The server sends the generated commentary to the user's terminal. The server returns the generated commentary to the user's terminal as an HTTP response. This response includes the generated commentary in text format.

[0993] Step 6:

[0994] The device displays the commentary received to the user. The device receives the HTTP response and displays the commentary in a UI component on the screen. Specifically, it uses a user interface such as a text view to display the commentary in a format that is easy for the user to read.

[0995] This allows users to experience high-quality live broadcasts based on real-time images.

[0996] Example 1

[0997] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0998] In recent years, there has been a demand for real-time information provision for sports and event broadcasts, but deploying professional announcers and facilitators is costly and difficult. This invention aims to solve this problem by providing high-quality commentary in real time through an automated system. In particular, by combining image analysis and natural language generation technology, it aims to provide users with an experience similar to that of professional commentary.

[0999] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1000] In this invention, the server includes means for uploading images from a user's terminal, means for receiving and saving the uploaded images, means for analyzing the saved images using an object detection model, means for generating live commentary comments using natural language generation technology based on the analysis results, means for transmitting the generated live commentary comments from the server to the user's terminal, and means for displaying the transmitted live commentary comments on the user's terminal, thereby enabling automatic provision of high-quality live commentary comments in real time.

[1001] A "user's terminal" is a device used by a user to upload images, and includes, for example, a smartphone, a personal computer, a tablet terminal, etc.

[1002] "Server" refers to a central processing unit that receives and stores images sent from user terminals, and further analyzes and generates live comments.

[1003] "Upload" refers to the process of sending image data from a user's device to a server.

[1004] "Image analysis" refers to the process by which the server uses object detection models to identify objects and movements in uploaded images.

[1005] An "object detection model" is a machine learning model used to identify dominant objects in an image, and can use techniques such as YOLO or SSD.

[1006] "Natural language generation technology" refers to technology for generating sentences in a form that can be understood by humans based on information obtained from analysis, and examples include generative AI models (such as GPT-3).

[1007] "Live commentary" refers to explanatory text about movements and events within an image that is generated based on the analysis results.

[1008] A "generative AI model" is an artificial intelligence model used to generate natural language text from a given prompt and analysis results.

[1009] A "prompt sentence" refers to a sentence that is input into a generative AI model to instruct it on what kind of text to generate.

[1010] An "HTTP request" refers to a form of communication protocol that requests data transmission from a user terminal to a server.

[1011] "HTTP response" refers to a form of communication protocol that returns data from a server to a user's terminal.

[1012] MODE FOR CARRYING OUT THE INVENTION

[1013] The present invention relates to a system in which a user uploads an image from a terminal, a server analyzes the image, and generates live commentary in real time. The present invention is comprised of the following steps as an embodiment.

[1014] Required Hardware and Software

[1015] User devices: smartphones, computers, tablets, etc.

[1016] Server: Central Processing Unit

[1017] Web interface or application: Software for uploading images

[1018] Object detection models: YOLO, SSD, etc.

[1019] Generative AI models: GPT-3, BERT, etc.

[1020] Communication method: HTTP protocol

[1021] Specific processing and data calculations

[1022] 1. Image input from the user

[1023] Users use devices such as smartphones or PCs to select and upload images through a web interface or application. When they tap the upload button, the device sends the image data to the server as an HTTP POST request.

[1024] 2. Receiving images on the server

[1025] The server receives the HTTP POST request sent from the user's device and saves the image file. This data is stored in the server's temporary directory.

[1026] 3. Image analysis on the server

[1027] The server analyzes the stored image files, using a pre-trained object detection model (e.g., YOLO or SSD) to identify the dominant objects in the image. The analysis results in the location and type of object in the image. This information is stored in an internal data structure.

[1028] 4. Live comment generation on the server

[1029] The server generates commentary based on the results of image analysis. Specifically, it extracts key information from the analysis (e.g., "Player A has the ball") and inputs the prompt and analysis results into a generative AI model (e.g., GPT-3). The generated commentary is then translated into natural language.

[1030] Example prompt sentence:

[1031] "Image analysis results of a soccer game: Player A has the ball. Please generate commentary based on this information."

[1032] 5. Sending live commentary from the server to the device

[1033] The server sends the generated live commentary to the user's terminal. The server creates a live commentary as an HTTP response and sends it to the user's terminal.

[1034] 6. Displaying live comments on user devices

[1035] The user's device displays the received live commentary. It receives the HTTP response, extracts the live commentary from the response body, and displays it to the user in real time using a dedicated UI component (e.g., text view).

[1036] In this way, the user can experience high quality real-time live broadcasting. The above is a specific embodiment for carrying out the present invention.

[1037] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1038] Step 1:

[1039] A user uploads an image using a device such as a smartphone or PC. Specifically, they first open a web interface or application, click or tap the image selection button to select an image file from the device's storage, and then click or tap the upload button. This causes the device to send the selected image file to the server as an HTTP POST request.

[1040] Input: An image file selected by the user

[1041] Output: Image data is sent to the server as an HTTP POST request

[1042] Step 2:

[1043] The server receives the HTTP POST request sent from the user's device and saves the image file. Specifically, it extracts the image data from the body of the received request and saves it in a temporary directory on the server.

[1044] Input: Image data sent as an HTTP POST request

[1045] Output: Image file saved in a temporary directory on the server

[1046] Step 3:

[1047] The server analyzes the stored image files and identifies the dominant object in the image using a pre-trained object detection model (e.g., YOLO or SSD). Specifically, the image data is input to the object detection model, and the location and type of the dominant object are obtained as output. The analysis results are stored in an internal data structure.

[1048] Input: Image file saved in the temporary directory on the server

[1049] Output: Analysis results on the location and type of objects in the image

[1050] Step 4:

[1051] The server generates commentary based on the results of image analysis. Specifically, it extracts key information from the analysis results (e.g., "Player A has the ball") and inputs the prompt and analysis results into a generative AI model (e.g., GPT-3). The generative AI model generates commentary in natural language based on the input.

[1052] Input: Analysis results obtained from the object detection model

[1053] Output: Generated natural language commentary

[1054] Example prompt sentence:

[1055] "Image analysis results of a soccer game: Player A has the ball. Please generate commentary based on this information."

[1056] Step 5:

[1057] The server sends the generated commentary to the user's terminal. Specifically, the server creates the commentary as an HTTP response and sends it to the user's terminal.

[1058] Input: Generated commentary

[1059] Output: Commentary sent to the user's device as an HTTP response

[1060] Step 6:

[1061] The user's device displays the received live commentary. Specifically, it receives the HTTP response, extracts the live commentary from the response body, and displays it to the user using a dedicated UI component (e.g., text view).

[1062] Input: Comments sent as an HTTP response

[1063] Output: Commentary displayed on the user's device

[1064] The above is the detailed flow of the processing of this program.

[1065] (Application example 1)

[1066] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1067] Conventional factory monitoring systems have made it difficult to quickly grasp the operating status and abnormalities of equipment, resulting in delayed response when an abnormality occurs. Furthermore, constant monitoring by an operator with specialized knowledge is required, which increases costs. The present invention aims to solve these problems by providing a system that can monitor the status of equipment in a factory in real time and quickly take appropriate action.

[1068] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1069] In this invention, the server includes means for uploading images from a user's device, means for analyzing the uploaded images and identifying objects and actions in the images, means for generating commentary related to a specified situation based on the analysis results, means for transmitting the generated commentary to the user's device, means for displaying the transmitted commentary on the user's device, and means for analyzing camera footage within the factory in real time and automatically generating and notifying the status and abnormalities of equipment as commentary, thereby enabling factory supervisors and operators to grasp the status of equipment in real time and take appropriate action promptly.

[1070] An "image" is digital data containing visual information that is uploaded from a user's device.

[1071] "User's Device" is the terminal equipment used to upload images to the system.

[1072] "Uploading" is the act of sending an image from a user's device to a server.

[1073] "Analysis" is the process of processing uploaded images to identify objects and actions within them.

[1074] "Target" refers to an item to be detected, such as an object, person, or device present in an image.

[1075] "Action" refers to a specific action or change of state performed by an object in an image.

[1076] The "specified situation" refers to the state of an object or action that the system determines based on the analysis results.

[1077] "Live commentary" is a natural language description of an object or action that is generated based on the analysis results.

[1078] "Generation" is the process of creating live commentary based on the analysis results.

[1079] "Transmitting" refers to the act of transferring the generated live commentary to the user's device using a communication means.

[1080] "Display" refers to the act of visually showing the live commentary sent on the user's device.

[1081] "Camera footage" refers to real-time video data captured by cameras installed within the factory.

[1082] "Equipment status" refers to the operating status and operational status of various equipment within the factory.

[1083] "Abnormality" refers to a state in which the equipment is not operating normally or an error has occurred.

[1084] "Notification" refers to the act of notifying the user of the generated live commentary in real time on the user's device.

[1085] This invention realizes a system that analyzes camera images in a factory in real time and automatically generates and notifies the status and abnormalities of equipment as live comments. A specific embodiment of this system is described below.

[1086] Image input from the user

[1087] Users transmit real-time images from cameras installed in the factory to the server, which provides high-resolution images that are then received in real time by the user's device (such as a tablet or PC).

[1088] Image analysis on the server

[1089] The server uses an object detection model to analyze the video received from the user's device in real time. This object detection model applies a pre-trained model using a deep learning framework such as TensorFlow. The server can accurately analyze the operating status of the equipment in the image (e.g., operating normally, stopped, abnormality, etc.).

[1090] Live commentary generation on the server

[1091] Based on the analysis results, the server generates a commentary in natural language using a generative AI model such as GPT-3. For example, if the server obtains the analysis result "The conveyor belt has stopped," the generative AI model will generate a specific commentary such as "The conveyor belt has stopped. Please check error code 105."

[1092] Output and display of live commentary

[1093] The server sends the generated commentary to the user's device. The commentary is returned as an HTTP response and displayed in real time on the user's device. A dedicated UI component (e.g., a text view) is used to convey visual information to the user.

[1094] Hardware and software used

[1095] Hardware: Camera (high-resolution industrial camera), server (with high-performance CPU and GPU), user device (tablet, PC, etc.)

[1096] Software: Python, TensorFlow, Transformers library, OpenCV, HTTP communication library

[1097] Specific examples

[1098] As a specific example, camera footage from inside a factory is analyzed, and the server analyzes footage of a stopped conveyor belt. In this case, the server uses an object detection model to detect the conveyor belt stopping, and then uses a generative AI model to generate a live commentary such as "The conveyor belt has stopped. Please check error code 105." The generated live commentary is sent to the user's device and displayed in real time.

[1099] Prompt Sentence Examples

[1100] An example of a prompt sentence to input to the generative AI model is as follows:

[1101] Generate a comment for the following machine state: stopped.

[1102] An example of the commentary that is output is as follows:

[1103] The machine has stopped. Please check Error Code 105.

[1104] This invention enables factory supervisors and operators to grasp the status of equipment in real time and take appropriate action quickly.

[1105] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1106] Step 1:

[1107] Users send real-time video to a server via cameras installed in their factories. Specifically, the video captured by the camera is received by the user's device (tablet, PC, etc.) and sent to the server as an HTTP request. The input is the real-time camera video, and the output is the video data sent to the server.

[1108] Step 2:

[1109] The server analyzes video data received from the user's device in real time. This analysis uses an object detection model using a deep learning framework such as TensorFlow. The input is camera video, which is analyzed to detect and identify the operating status of equipment (e.g., conveyor belts). The output is the identified operating status information.

[1110] Step 3:

[1111] Based on the analysis results, the server uses a generative AI model such as GPT-3 to generate commentary in natural language. The input is the analysis result (e.g., conveyor belt stopped), and a prompt sentence is generated based on this information. The generative AI model responds to the prompt sentence and generates an appropriate commentary. The output is the generated commentary.

[1112] Step 4:

[1113] The server transmits the generated commentary to the user's device. Specifically, the server returns the generated commentary to the user's device as an HTTP response. The input is the generated commentary, and the output is the commentary to be transmitted to the user's device.

[1114] Step 5:

[1115] The user's device receives the commentary and displays it in real time. Specifically, the commentary received as an HTTP response is displayed using a dedicated UI component such as a text view. The input is the commentary sent from the server, and the output is the visually displayed commentary.

[1116] The specific actions taken at each step enable a series of processes to work together seamlessly, enabling real-time monitoring of the status and abnormalities of equipment within the factory and appropriate notification.

[1117] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1118] This invention relates to an image analysis and commentary generation system that combines an emotion engine that recognizes user emotions. The main components of the system are a user terminal, a server, an emotion engine, and a means for communication between them. The following describes the details of the embodiment.

[1119] 1. Image input from the user

[1120] A user uploads an image to the system using a device. The user selects an image file using a specific web interface or application and taps the upload button. This action causes the device to send the image file to the server as an HTTP request.

[1121] 2. Image analysis on the server

[1122] The server stores the received image files and performs image analysis, using an object detection model to identify key objects in the image (e.g., athletes, balls, etc.). The object detection model uses a pre-trained machine learning model and can recognize the type and location of objects with high accuracy.

[1123] 3. Live comment generation on the server

[1124] The server generates commentary based on the analysis results using natural language generation technology. For example, when an image of a soccer game is input, analysis can yield information such as "Player A is dribbling the ball." Based on this information, the generation AI generates commentary such as "Player A is dribbling the ball."

[1125] 4. Emotion Recognition on the Server

[1126] The server uses an emotion engine to recognize the user's emotional state. The emotion engine can identify emotions by analyzing the user's facial expressions, voice, or biometric information. This information is obtained through a camera, microphone, or biometric sensor.

[1127] 5. Moderating commentary based on emotions

[1128] The server automatically adjusts the content and expression of the commentary based on the user's emotional state obtained from the emotion engine. For example, if the user is excited, the server can make the commentary more energetic.

[1129] 6. Output and display of live commentary

[1130] The generated commentary is sent from the server to the user's device. The server returns the commentary as an HTTP response. The device uses a dedicated UI component (e.g., text view) to display the received commentary to the user. This allows the user to experience high-quality live broadcasting in real time.

[1131] Specific examples

[1132] Example: Soccer image

[1133] 1. A user uploads soccer game images from their device to the system. The user uses the image selection interface to select an image and then taps the upload button.

[1134] 2. The server receives the uploaded images and performs image analysis, using object detection models to identify players, the ball, goals, etc.

[1135] 3. Based on the analysis results, the server obtains the information that "Player A has the ball." Based on this information, the generation AI generates a commentary such as "Player A is dribbling the ball."

[1136] 4. The server uses the camera images and audio data acquired from the user's device to analyze the user's emotions using an emotion engine. For example, if the user is smiling, it will determine that the user is excited.

[1137] 5. The server adjusts the commentary based on the user's emotional state to "Player A is moving forward with a great dribble!"

[1138] 6. The server sends the generated commentary to the user's device, which then displays the commentary to the user, allowing the user to grasp the game situation in real time and enjoy a commentary experience that suits their emotions.

[1139] This system allows users to not only receive accurate commentary in real time, but also experience commentary that is customized according to their emotional state, providing a more personalized and high-quality user experience.

[1140] The processing flow will be explained below.

[1141] Step 1:

[1142] The user selects an image from the device and clicks the upload button. This operation causes the user device to attach the selected image file to an HTTP request and send it to the server. Specifically, the user uses the image selection interface to select an image file, and then presses the send button, causing the device to send the image data to the server via an HTTP POST request.

[1143] Step 2:

[1144] The server saves the received image file. The server analyzes the HTTP request, extracts the sent file data, and saves it as an image file in the specified directory. This operation makes the image to be analyzed accessible on the server.

[1145] Step 3:

[1146] The server analyzes the saved images. The server loads an object detection model and analyzes the saved images to identify key objects and movements within the images. Specifically, the image is input into the object detection model to detect the location and type of objects, such as players and the ball.

[1147] Step 4:

[1148] The server generates commentary based on the analysis results. The server generates commentary using natural language generation technology based on the analyzed objects and their movements. For example, if the analysis result is "Player A is dribbling the ball," the server generates commentary such as "Player A is dribbling the ball."

[1149] Step 5:

[1150] The user transmits emotion data from the device. The user uses the device's camera and microphone to transmit their facial expressions and voice to the system. The device then transmits this data to the server in real time.

[1151] Step 6:

[1152] The server analyzes the user's emotions using an emotion engine. The server inputs the received user's facial expressions and voice into the emotion engine to determine the user's current emotional state. For example, if the server receives image data of the user smiling, it determines that the user is excited.

[1153] Step 7:

[1154] The server adjusts the commentary based on the user's emotional state. The tone and content of the commentary is changed based on the emotional information obtained from the emotion engine. For example, if the user is excited, the commentary will be adjusted from "Player A is dribbling the ball" to an energetic expression such as "Player A is moving forward with a great dribble!"

[1155] Step 8:

[1156] The server sends the adjusted live commentary to the user's terminal. The server sends the adjusted live commentary to the user's terminal as an HTTP response.

[1157] Step 9:

[1158] The device displays the commentary received to the user. The device receives the HTTP response and displays the commentary in a UI component on the screen. Specifically, the device uses an interface such as a text view to display the commentary in a format that is easy for the user to see.

[1159] This process flow allows users to not only receive accurate live coverage in real time, but also enjoy a live experience that is customized according to their emotional state.

[1160] Example 2

[1161] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1162] Conventional image analysis systems were able to generate live commentary related to a specific event based on image information provided by the user, but were unable to provide personalized live commentary that took the user's emotional state into account. This resulted in a uniform user experience, making it difficult to provide high-quality live coverage that was tailored to each individual user's emotional state. The present invention aims to solve this problem by generating live commentary customized according to the user's emotional state, thereby providing a personalized, high-quality live experience.

[1163] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1164] In this invention, the server includes means for uploading images from a user's terminal, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating live commentary related to a specified event based on the analysis results, means for transmitting the generated live commentary to the user's terminal, means for displaying the transmitted live commentary on the user's terminal, means for acquiring emotional data for recognizing the user's emotional state, and means for adjusting the content and expression of the live commentary based on the acquired emotional data. This makes it possible to provide customized live commentary that takes the user's emotional state into consideration in real time, thereby providing a more personalized and high-quality live commentary experience.

[1165] "Means for uploading images from the user's terminal" is a function that allows the user to send image files from their own terminal to the server.

[1166] "Means of analyzing uploaded images and identifying objects and movements within the images" refers to a function that uses image analysis technology on received images to recognize and identify objects and their movements within the images.

[1167] The "means for generating live commentary related to a specified event based on the analysis results" is a function that automatically generates sentences related to an event based on data obtained from image analysis.

[1168] The "means for transmitting the generated live commentary to the user's terminal" is a function for transmitting the automatically generated live commentary from the server to the user's terminal.

[1169] The "means for displaying the transmitted live commentary on the user's terminal" is a function for displaying the live commentary received on the user's terminal.

[1170] The "means for acquiring emotional data to recognize the user's emotional state" is a function for collecting emotional data such as the user's facial expressions and voice through sensors, cameras, etc.

[1171] The "means for adjusting the content and expression of the commentary commentary based on the acquired emotional data" is a function for analyzing the collected emotional data and changing the tone and content of the commentary commentary based on the results.

[1172] The present invention relates to an image analysis and commentary generation system that combines an emotion engine that recognizes user emotions. The following describes in detail the embodiments of the present invention.

[1173] The system of the present invention is mainly composed of a user terminal, a server, an emotion engine, and means for communicating among them.

[1174] Hardware and software used

[1175] The user's device may be a smartphone, tablet, or PC, and may be equipped with a camera, microphone, and, if necessary, biometric sensors. The server may be a cloud environment or a dedicated computer system, and includes the following main software components:

[1176] 1. Image analysis module: This module implements an object detection model to identify objects and movements in images, using computer vision libraries (e.g., OpenCV, TensorFlow).

[1177] 2. Natural Language Generation Module: Uses a generative AI model (e.g., GPT-3) to generate commentary based on the analysis results.

[1178] 3. Emotion Engine: Machine learning models are implemented to recognize the user's emotional state, using facial recognition and voice analysis models.

[1179] 4. Communication module: Processes HTTP requests and responses to send and receive data between the user's device and the server.

[1180] Example of a system

[1181] Specific example using soccer images

[1182] 1. A user uses a device to upload an image of a soccer match to the system. The user uses the image selection interface to select an image and then taps the upload button. This action sends the image file to the server as an HTTP request.

[1183] 2. The server stores the received image files and uses an image analysis module to identify objects in the images (e.g., players, ball, goal). The object detection model utilizes computer vision libraries to perform highly accurate analysis.

[1184] 3. Based on the results of the image analysis, the server uses a natural language generation module to generate a commentary such as "Player A is dribbling the ball" based on the information that "Player A has the ball." The following prompt is input to the generation AI model:

[1185] "Enter a football image."

[1186] "Analyze your emotional state and energetically adjust the tone of the play-by-play commentary you generate."

[1187] 4. The user's device uses a camera, microphone, and biometric sensors to acquire emotional data such as the user's facial expressions, voice, and heart rate, and sends this data to the server as an HTTP request.

[1188] 5. The server uses the emotion engine to analyze the received emotion data and identify the user's emotional state. For example, if the user is smiling, it determines that the user is excited.

[1189] 6. Based on the results of the emotion engine, the server adjusts the generated commentary to "Player A is moving forward with a great dribble!" This adjusted commentary is sent to the user's device as an HTTP response.

[1190] 7. The device displays the received live commentary using a dedicated UI component (e.g., text view), allowing the user to experience high-quality live broadcasts in real time.

[1191] Through the above-described method, users can obtain commentary that is customized according to their emotional state, and can enjoy a more personalized and high-quality commentary experience.

[1192] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1193] Step 1:

[1194] A user uses a device to open a specific web interface or application. They select an image file and tap the upload button to send the image to the system. The input is an image file stored on the device, and the output is that the image file is sent to the server as an HTTP request. The device generates header information for the HTTP request and attaches the image file.

[1195] Step 2:

[1196] The server receives the HTTP request. It saves the received image file and generates a path for the save location. Here, it checks the image format and size, and converts the file format or adjusts the size as necessary. The input is the image file sent from the terminal, and the output is the image file path saved in a specific directory on the server. The server extracts the image file data from the HTTP request and saves it in the specified directory.

[1197] Step 3:

[1198] The server inputs the saved image file into an object detection model and performs analysis. It uses an image analysis module to identify objects and movements and outputs the analysis results in JSON format. The input is the saved image file path, and the output is the JSON data of the analysis results. The server uses an object detection model (e.g., OpenCV, TensorFlow) to process and recognize objects and movements in the image.

[1199] Step 4:

[1200] Based on the analysis results, the server inputs the results as prompts into a generative AI model (e.g., GPT-3) to generate play-by-play comments. For example, based on the result "Player A has the ball," the server generates the comment "Player A is dribbling the ball." The input is the JSON data of the analysis results, and the output is play-by-play comments in natural language. The server creates appropriate prompts for the generative AI model and inputs them into the model.

[1201] Step 5:

[1202] The user's device uses a camera, microphone, and biometric sensors to collect the user's facial expressions, voice, heart rate, and other emotional data. The input is the user's various sensor data, and the output is emotional data sent to the server via an HTTP request. The device formats the collected data and prepares it to be sent to the server.

[1203] Step 6:

[1204] The server inputs the received emotional data into the emotion engine for analysis. The emotion engine determines whether the user is happy, excited, or sad. The input is the emotional data sent by the user, and the output is the analysis result (in JSON format) about the user's emotional state. The server uses the emotion engine to analyze the data using facial expression recognition models and voice analysis models.

[1205] Step 7:

[1206] The server readjusts the commentary based on the analysis results of the emotion engine. For example, if the user is excited, the tone of the commentary can be adjusted to be more energetic, such as "Player A is moving forward with a great dribble!" The input is the analysis results of the emotion engine and the initial commentary, and the output is the adjusted commentary. The server again inputs the prompt into the generative AI model to generate the adjusted commentary.

[1207] Step 8:

[1208] The server sends the final commentary to the user's device as an HTTP response. The input is the adjusted commentary, and the output is the HTTP response it sends to the device. The server sets the appropriate response headers and sends the commentary.

[1209] Step 9:

[1210] The device displays the received live commentary to the user. Specifically, it displays the live commentary in real time using a dedicated UI component (e.g., text view). The input is the live commentary obtained from the HTTP response, and the output is the live commentary displayed on the screen. The device performs UI update processing to provide the user with a real-time live commentary experience.

[1211] (Application example 2)

[1212] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1213] In autonomous vehicles, it is difficult to grasp the driver's emotional state in real time and provide appropriate information and alerts while driving. In addition, there is a lack of technology to personalize road information and driving advice according to the driver's emotions.

[1214] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for uploading images from a user's terminal, means for analyzing the uploaded images and identifying objects and movements in the images, means for generating a commentary related to a specified event based on the analysis results, means for recognizing the user's emotional state, means for adjusting the content and tone of the commentary based on the recognized emotional state, means for transmitting the generated commentary to the user's terminal, and means for displaying the transmitted commentary on the user's terminal. This makes it possible to analyze the driver's emotions in real time and provide commentary and situation reports accordingly.

[1215] "Means for uploading images from the user's terminal" refers to an interface or function that allows the user to select image data from their own terminal and transfer it to the system.

[1216] "Means for analyzing uploaded images and identifying objects and movements within the images" refers to technology for analyzing received image data and extracting information about objects present within the images and their movements.

[1217] "Means for generating live commentary related to a specified event based on the analysis results" refers to a function that automatically generates commentary and narration related to a specific event based on the results of image analysis.

[1218] "Means for recognizing the user's emotional state" refers to technology that uses sensors such as cameras and microphones to identify emotions from the user's facial expressions and voice.

[1219] "Means for adjusting the content and tone of live commentary based on the recognized emotional state" refers to a technology that utilizes the user's emotional information to change the wording and expression of the live commentary that is generated.

[1220] The "means for transmitting the generated live commentary to the user's terminal" is a function for transferring the generated live commentary to the user's terminal via a communication means such as the Internet.

[1221] "Means for displaying the transmitted live commentary on the user's terminal" refers to a function for displaying and playing the received live commentary on the user's terminal by screen or audio.

[1222] This invention is a system that recognizes a user's emotions and generates and provides live commentary in real time based on those emotions. A specific example for realizing this invention is shown below.

[1223] 1. Upload an image

[1224] Users upload images to the system using their own devices. To upload images, a dedicated application or a web interface is used. The user selects an image file and taps the upload button, which sends the image data from the device to the server.

[1225] 2. Image Analysis

[1226] The server receives and analyzes images uploaded by users. This analysis uses an object detection model, a machine learning model. This model is pre-trained and can detect objects and their movements in images with high accuracy. For example, in an image of a soccer game, it can recognize players, the ball, goals, etc.

[1227] 3. Generating live commentary

[1228] Based on the analysis results, the server generates commentary using natural language generation technology. Examples of generative AI models used include GPT-3. The following prompt sentence is input to this model to generate commentary:

[1229] Road conditions: Congested. Driver's emotional state: Tired. Commentary:

[1230] 4. Emotion recognition

[1231] The server uses an emotion recognition engine to recognize the user's emotions in real time. This engine analyzes the user's facial expressions and voice data to identify their emotional state. Cameras and microphones installed in the vehicle are used to acquire the data.

[1232] 5. Adjustments to commentary

[1233] Based on the emotional information obtained by the emotion recognition engine, the server adjusts the content and tone of the live commentary. For example, if the user is tired, it uses expressions that encourage relaxation.

[1234] 6. Output and display of live commentary

[1235] The generated commentary is sent from the server to the user's device, which displays the commentary on the screen or plays it back as audio. This allows the user to receive customized commentary in real time, tailored to their emotional state.

[1236] Specific examples

[1237] soccer match

[1238] A user uploads an image of a soccer game, and the analysis results provide information such as "Player A is dribbling the ball." Furthermore, if the user is recognized as being tired, a commentary using the generative AI model is generated, such as "Player A is moving forward with a relaxed dribble!", and displayed on the user's device.

[1239] Autonomous vehicles

[1240] If the driver is tired, the road conditions are analyzed as "congested." The following prompt sentence is input into the generative AI model to generate a running commentary.

[1241] Road conditions: Congested. Driver's emotional state: Tired. Commentary:

[1242] The generated comment will be "The roads are currently congested. Please relax and continue driving," and will be personalized according to the driver's emotions.

[1243] In this way, the system combines image analysis and emotion recognition to provide users with more personalized, high-quality information.

[1244] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1245] Step 1:

[1246] Users upload images to the system through their own devices. When a user selects an image file and taps the upload button, the image data is sent from the device to the server as an HTTP request. The input is the image data from the user device, and the output is the transfer of the image data to the server.

[1247] Step 2:

[1248] The server analyzes the image data received from the user. For image analysis, it uses an object detection model, which is a machine learning model. Specifically, it analyzes the information of each pixel in the image and identifies the main objects and their movements. The input is the image data uploaded by the user, and the output is the detection results of the objects and movements.

[1249] Step 3:

[1250] The server generates commentary based on the analysis results using the generative AI model GPT-3. During the generation process, contextual information from the analysis results is input as a prompt, and the model generates commentary in natural language format. The input is the image analysis results and the prompt, and the output is the generated commentary.

[1251] Step 4:

[1252] The server recognizes the user's emotional state in real time. It uses an emotion recognition engine to analyze the user's facial and voice data from a camera and microphone to identify the user's emotional state. The input is the facial and voice data obtained from the user, and the output is emotional state information from the emotion recognition engine.

[1253] Step 5:

[1254] The server adjusts the content and tone of the commentary based on the user's emotional state. It also changes the wording and tone of the commentary based on the emotional data obtained from the emotion recognition engine. The input is the emotional state information and the commentary generated in the initial stage, and the output is the adjusted commentary.

[1255] Step 6:

[1256] The server sends the adjusted live commentary to the user's device. The live commentary is delivered to the user's device using an HTTP response as the communication method. The input is the adjusted live commentary, and the output is the comment sent to the user's device.

[1257] Step 7:

[1258] The terminal displays the received live commentary. The live commentary is provided to the user in real time in the form of a screen display or audio playback. The input is the live commentary received from the server, and the output is information provided to the user.

[1259] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1260] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1261] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1262] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1263] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1264] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1265] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1266] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1267] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1268] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1269] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1270] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1271] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1272] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1273] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1274] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1275] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1276] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1277] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1278] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1279] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1280] The following is further disclosed regarding the above embodiment.

[1281] (Claim 1)

[1282] a means for uploading images from a user's device;

[1283] A means for analyzing uploaded images to identify objects and movements within the images; and

[1284] means for generating commentary relating to a specified event based on the analysis results;

[1285] means for transmitting the generated commentary comments to a user's terminal;

[1286] The system includes a means for displaying the transmitted live commentary on the user's terminal.

[1287] (Claim 2)

[1288] The system according to claim 1, wherein a live commentary is generated based on the analysis results using natural language generation technology.

[1289] (Claim 3)

[1290] The system of claim 1, wherein the image analysis is performed using an object detection model.

[1291] "Example 1"

[1292] (Claim 1)

[1293] a means for uploading images from a user's device;

[1294] A means for the server to receive and store the uploaded images;

[1295] A means for the server to analyze the stored image using an object detection model;

[1296] A means for the server to generate live commentary using natural language generation technology based on the analysis results;

[1297] means for transmitting the generated live commentary from the server to the user's terminal;

[1298] The system includes a means for displaying the transmitted live commentary on the user's terminal.

[1299] (Claim 2)

[1300] The system of claim 1 extracts key information from the analysis results, inputs the prompt sentence and the analysis results into a generative AI model, and generates live comments.

[1301] (Claim 3)

[1302] 10. The system of claim 1, wherein the image analysis is performed using a pre-trained object detection model.

[1303] "Application Example 1"

[1304] (Claim 1)

[1305] means for uploading images from a user's device;

[1306] A means for analyzing uploaded images to identify objects and actions within the images; and

[1307] means for generating commentary relating to a specified situation based on the analysis results;

[1308] means for transmitting the generated commentary to a user device;

[1309] means for displaying the transmitted live commentary on a user's device;

[1310] This system analyzes camera footage from within the factory in real time and automatically generates and notifies users of the equipment's status and any abnormalities as live commentary.

[1311] (Claim 2)

[1312] The system according to claim 1, wherein a live commentary is generated based on the analysis results using natural language generation technology.

[1313] (Claim 3)

[1314] The system of claim 1, wherein the image analysis is performed using an object detection model.

[1315] "Example 2: Combining Emotion Engines"

[1316] (Claim 1)

[1317] a means for uploading images from a user's device;

[1318] A means for analyzing uploaded images to identify objects and movements within the images; and

[1319] means for generating commentary relating to a specified event based on the analysis results;

[1320] means for transmitting the generated commentary comments to a user's terminal;

[1321] A means for displaying the transmitted live commentary on the user's terminal;

[1322] means for acquiring emotional data for recognizing an emotional state of a user;

[1323] The system includes a means for adjusting the content and expression of live commentary based on acquired emotional data.

[1324] (Claim 2)

[1325] The system according to claim 1, wherein a live commentary is generated based on the analysis results using natural language generation technology.

[1326] (Claim 3)

[1327] The system of claim 1, wherein the image analysis is performed using an object detection model.

[1328] "Application example 2 when combining emotion engines"

[1329] (Claim 1)

[1330] a means for uploading images from a user's device;

[1331] A means for analyzing uploaded images to identify objects and movements within the images; and

[1332] means for generating commentary relating to a specified event based on the analysis results;

[1333] means for recognizing the emotional state of a user;

[1334] a means for adjusting the content and tone of commentary based on the perceived emotional state;

[1335] means for transmitting the generated commentary comments to a user's terminal;

[1336] The system includes a means for displaying the transmitted live commentary on the user's terminal.

[1337] (Claim 2)

[1338] The system according to claim 1, wherein a live commentary is generated based on the analysis results using natural language generation technology.

[1339] (Claim 3)

[1340] The system of claim 1, wherein the image analysis is performed using an object detection model. [Explanation of symbols]

[1341] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for uploading images from a user's device; A means for analyzing uploaded images to identify objects and movements within the images; and means for generating commentary relating to a specified event based on the analysis results; means for transmitting the generated commentary comments to a user's terminal; The system includes a means for displaying the transmitted live commentary on the user's terminal.

2. The system according to claim 1, wherein a live commentary is generated based on the analysis result using natural language generation technology.

3. The system of claim 1 , wherein the image analysis is performed using an object detection model.

Images