System

Abstract

A system is provided.SOLUTION: A system comprising: means for obtaining video data from an on-board camera; means for analyzing the obtained video data on a frame-by-frame basis to recognize a traffic condition; means for automatically generating a narrative based on the analyzed traffic condition; and means for providing the generated narrative to a user device.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] Currently, there is a lack of generative AI that can accurately describe traffic conditions, which means that creating training data for AI development requires a great deal of effort and time. Furthermore, efficient data generation technology is needed to smoothly advance the development of autonomous driving and advanced driver assistance systems (ADAS), which require accurate understanding of traffic conditions. [Means for solving the problem]

[0005] The present invention provides a system that includes a means for acquiring video data from an in-vehicle camera, a means for analyzing the acquired video data frame by frame to recognize traffic conditions, a means for automatically generating explanatory text based on the analyzed traffic conditions, and a means for providing the generated explanatory text to a user device. This system enables accurate and efficient generation of explanatory text about traffic conditions, thereby supporting the development of autonomous driving and ADAS. Specifically, the system acquires in-vehicle camera video, analyzes it to detect the status of vehicles, pedestrians, and traffic lights, and generates explanatory text from the analysis results using a generative AI model. This allows developers to efficiently create training data.

[0006] An "in-vehicle camera" is a camera device installed in a vehicle, and is a device for capturing images of external road conditions and obstacles.

[0007] "Video data" refers to real-time video information acquired by an in-vehicle camera, and is a collection of frames to be analyzed.

[0008] A "frame" is an individual still image that makes up video data, and is the unit of analysis.

[0009] "Traffic conditions" refers to the general road conditions, including the status and movement of vehicles, pedestrians, and traffic lights on the road.

[0010] "Recognition" is the process of identifying and locating objects and situations in video data.

[0011] The "description" is a natural language text generated based on the video data, which describes the traffic situation in detail.

[0012] A "user device" is an electronic device, such as a smartphone or PC, that a user uses to receive and display information.

[0013] A "generative AI model" is an artificial intelligence model that automatically generates natural language descriptions from the results of traffic situation analysis. [Brief explanation of the drawings]

[0014] [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

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] The present invention is a system that automatically converts in-vehicle camera images into explanatory text about traffic conditions, and aims to streamline the development of autonomous driving and advanced driver assistance systems (ADAS). This system includes the following main means:

[0036] Acquiring video data

[0037] The user starts the in-vehicle camera and starts capturing video data. The in-vehicle camera is installed in front of the vehicle and records road conditions in real time.

[0038] Receiving and analyzing video data

[0039] The server receives video data sent from the in-vehicle camera. The received video data is divided into frames and pre-processed to analyze the traffic situation within each frame. This pre-processing includes noise removal, color correction, and resolution adjustment.

[0040] Traffic situation recognition

[0041] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. For example, it recognizes whether a vehicle is about to turn left, whether a pedestrian is crossing the street, or whether a traffic light has changed from red to green.

[0042] Generating Descriptions

[0043] The server generates a natural language description based on the analysis results. Using a generative AI model, detailed text about traffic conditions is automatically created, allowing users to understand real-time traffic conditions without having to check the video.

[0044] Providing a description

[0045] The server then provides the generated description to the user device, which is an electronic device such as a smartphone or PC that receives and displays the description, allowing the user to check detailed information about traffic conditions in real time.

[0046] Specific examples

[0047] For example, in a scenario describing traffic conditions on a given day, the following sequence of actions occurs:

[0048] 1. The user turns on the dashcam and begins capturing video.

[0049] 2. The server receives the video data and analyzes it frame by frame.

[0050] 3. The server recognizes that there are four vehicles ahead, the traffic light is red, and detects that a pedestrian is beginning to cross the street.

[0051] 4. The server passes the analysis results to the generation AI, which automatically generates a description such as "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0052] 5. The server generates a description and sends it to the user's device, where the user can view this information in real time.

[0053] This allows users to understand traffic conditions in detail through text without having to look at the video directly. This system enables the efficient generation of learning data in the development of autonomous driving and ADAS, greatly streamlining the development process.

[0054] The above is a detailed description of the embodiment of the present invention.

[0055] The processing flow will be explained below.

[0056] Step 1:

[0057] The user activates the dashcam, which then starts capturing video data. Specifically, the user turns on the power of the camera device in the vehicle, and a real-time stream of video data begins.

[0058] Step 2:

[0059] The server receives video data sent from the dashcam in real time. The server establishes a connection based on the specified camera ID and continuously acquires data.

[0060] Step 3:

[0061] The server divides the received video data into frames, converting the video data into frames for each small time unit, and these frames are then used for subsequent analysis processing.

[0062] Step 4:

[0063] The server performs pre-processing on each frame, including noise reduction, color correction, and resolution adjustment, to improve the quality of the frame data.

[0064] Step 5:

[0065] The server performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This is done using object detection and motion analysis algorithms.

[0066] Step 6:

[0067] The server recognizes traffic conditions based on the results of object detection and motion analysis, including vehicle position, traffic light status, and pedestrian movement.

[0068] Step 7:

[0069] The server passes the analysis results to a generative AI model, which then generates a natural language explanation. The generative AI model converts the analyzed traffic situation data into text and automatically creates a detailed explanation.

[0070] Step 8:

[0071] The server then sends the generated explanatory text to the user device, which may be a smartphone or PC, and receives the explanatory text from the server and displays it to the user.

[0072] Step 9:

[0073] Users can view the description on their device and get a detailed understanding of the real-time traffic situation. The textual information allows users to intuitively understand the situation ahead of the vehicle and other important traffic information.

[0074] Example 1

[0075] 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."

[0076] In the development of autonomous driving and advanced driver assistance systems (ADAS), it is important to accurately and efficiently grasp traffic conditions. However, conventional methods require manual analysis of vast amounts of video data, which is time-consuming and costly. Furthermore, it is difficult to grasp traffic conditions in real time, which reduces development efficiency. This invention aims to solve these problems and significantly streamline the development process for autonomous driving and ADAS.

[0077] 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.

[0078] In this invention, the server includes means for acquiring video data from an on-board camera, means for preprocessing the acquired video data frame by frame, means for recognizing traffic conditions for each preprocessed frame, means for automatically generating explanatory text based on the recognized traffic conditions, and means for providing the generated explanatory text to a user device. This enables real-time analysis of video data and efficient development of autonomous driving and ADAS.

[0079] An "in-vehicle camera" is a camera installed in a vehicle to record road and traffic conditions.

[0080] "Video data" refers to video captured by a camera and stored in digital format.

[0081] A "frame" refers to each of the consecutive still images that make up video data.

[0082] "Preprocessing" refers to initial processing such as noise removal, color correction, and resolution adjustment that is performed to make video data easier to analyze.

[0083] "Traffic conditions" refers to information that indicates the overall state and movement of vehicles, pedestrians, traffic lights, etc. on the road.

[0084] "Recognition" is the process of detecting and understanding specific objects and their movements from video data.

[0085] A "description" is a text that describes the recognized traffic situation in natural language.

[0086] A "generative AI model" is an artificial intelligence model that automatically generates text based on input data.

[0087] A "user device" is an electronic device for displaying the generated description, and includes a smartphone, a PC, etc.

[0088] The system according to the present invention automatically converts images from an in-vehicle camera into explanatory text about traffic conditions, and aims to improve the efficiency of the development of autonomous driving and advanced driver assistance systems (ADAS). Specific embodiments of this system are described below.

[0089] First, the user activates the dashcam to begin capturing video data. The dashcam is installed in the front of the car and records road conditions in real time while the car is moving. This camera uses a standard high-resolution camera and captures video data at 30 frames per second (FPS).

[0090] The captured video data is then sent to a server in real time. The server divides the video data into frames and performs preprocessing. This preprocessing includes processes such as noise reduction using the OpenCV library, color correction, and resolution adjustment. For example, noise reduction involves applying a filter to remove unwanted background signals. Color correction involves adjusting the color balance of each frame to improve visibility.

[0091] After preprocessing, the server recognizes traffic conditions for each frame. During this process, it runs an object detection algorithm to detect objects such as vehicles, pedestrians, and traffic lights. For example, it can use the YOLO (You Only Look Once) model to detect objects quickly and accurately. It analyzes the position and movement of detected objects in each frame to understand traffic flow. Examples of identified objects include vehicles approaching an intersection, pedestrians crossing a crosswalk, and traffic lights changing from red to green.

[0092] Based on the analysis results, the server uses a generative AI model to generate a description in natural language. One example of a generative AI model that can be used is GPT-3. The server passes the analysis results to the generative AI as a prompt, which generates a description like the following: "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk." The generated description is formatted in a way that is easy for the user to understand.

[0093] Finally, the server sends the generated description in real time to the user device. The user device is an electronic device such as a smartphone or PC, which displays the description on its screen. The user can check the traffic situation in text form through the application. The display on the user device shows, "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0094] As a concrete example, here is a scenario describing traffic conditions on a given day:

[0095] 1. The user activates the dashcam using a smartphone app and begins capturing footage.

[0096] 2. The server receives the video data in real time and performs noise reduction, color correction, and resolution adjustment on each frame.

[0097] 3. The server uses the YOLO model to detect vehicles, pedestrians, and traffic lights ahead and analyze traffic conditions.

[0098] 4. The server uses GPT-3 to generate an explanation based on the analysis results, generating the sentence, "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0099] 5. The server generates a description and sends it to the smartphone, where the user can view the information in real time through the app.

[0100] This allows users to understand traffic conditions in detail through text without having to look at the video directly. This system enables efficient generation of learning data for the development of autonomous driving and ADAS, greatly streamlining the development process.

[0101] Example prompt sentence:

[0102] "Analyze the traffic situation ahead from the specified video data and generate a description including the status of vehicles, pedestrians, and traffic lights."

[0103] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0104] Step 1:

[0105] The user activates the dashcam and begins capturing video data. Specifically, the user presses the camera's activation button using a smartphone app, and video recording begins from that point. The camera is installed at the front of the car and records video at 30 frames per second (FPS). The input is the user's camera activation operation and the real-time video scene. The output is high-resolution video data obtained from the camera.

[0106] Step 2:

[0107] The server receives video data sent from the in-vehicle camera in real time. The received video data is 30 frames per second (FPS) and is received and stored on the server. The input is the video data from the in-vehicle camera. The output is the raw video data stored on the server.

[0108] Step 3:

[0109] The server pre-processes the received video data for each frame. Specifically, it uses the OpenCV library to perform noise reduction, color correction, and resolution adjustment. For example, noise reduction applies a Gaussian filter, color correction adjusts the RGB balance, and resolution adjustment resizes the data to a consistent size. The input is the raw video data (in frame format) stored on the server. The output is the pre-processed frame data.

[0110] Step 4:

[0111] The server recognizes traffic conditions for each preprocessed frame. Specifically, it uses the YOLO (You Only Look Once) model to detect objects such as vehicles, pedestrians, and traffic lights. It tracks the location and movement of each detected object to understand traffic flow and conditions. For example, it detects changes in the color of traffic lights and the direction of pedestrian movement. The input is the preprocessed frame data. The output is the analysis result of the traffic conditions for each frame.

[0112] Step 5:

[0113] The server uses a generative AI model based on the analysis results to generate a natural language description. Specifically, it utilizes a large-scale language model such as GPT-3 and provides the traffic situation analysis results obtained in the previous step to the model as a prompt. For example, it generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk." The input is the traffic situation analysis results and the prompt. The output is the generated natural language description.

[0114] Step 6:

[0115] The server sends the generated description to the user device in real time. The user device is an electronic device such as a smartphone or PC, which displays the received description on its screen. The input is the generated natural language description. The output is the description displayed on the user device. For example, the app's display might say, "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0116] (Application example 1)

[0117] 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."

[0118] Conventional autonomous vehicles and advanced driver assistance systems (ADAS) require drivers and developers to directly check video data to understand traffic conditions, which places a significant burden on the driver and developers. Additionally, it is difficult to grasp traffic conditions in real time, resulting in delayed reactions in situations where instant decisions are required. Furthermore, relying solely on vision means that information acquisition is incomplete, potentially increasing the risk of accidents and traffic violations.

[0119] 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.

[0120] In this invention, the server includes means for acquiring video data from an on-board camera, means for analyzing the acquired video data frame by frame and recognizing traffic conditions, means for automatically generating explanatory text based on the analyzed traffic conditions, means for providing the generated explanatory text to a user device, and means for displaying the generated explanatory text in real time on smart glasses, which are a type of user device. This enables drivers and developers of autonomous driving systems to grasp detailed and accurate traffic conditions in real time without directly viewing the video.

[0121] An "on-board camera" is a camera device mounted on a vehicle that captures road and surrounding conditions as video data in real time.

[0122] "Video data" refers to continuous image information captured by an on-board camera, and visually records traffic conditions and the surrounding environment.

[0123] A "frame" refers to a series of individual images of video data that are treated as a single still image.

[0124] "Traffic conditions" refers to the state of various elements on the road, such as vehicles, pedestrians, and traffic lights, and their interrelationships.

[0125] A "user device" is an electronic device that can receive explanatory text sent from a server and display it to a user.

[0126] "Analysis" is the process of processing video data to extract and understand specific information.

[0127] "Explanation" refers to text information written in natural language based on traffic situation information obtained through analysis.

[0128] "Real time" refers to processing and display occurring immediately without delay.

[0129] "Smart glasses" are glasses-type electronic devices that have the function of displaying images and text information, allowing the wearer to view information through the device.

[0130] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate text from given data.

[0131] This invention is a system that automatically converts in-vehicle camera footage into traffic situation descriptions and provides them to user devices (especially smart glasses) in real time. The purpose of this system is to streamline the development of autonomous vehicles and advanced driver assistance systems (ADAS) and to provide drivers with instant, detailed traffic information.

[0132] The server has a means to acquire video data from dashcams, which are attached to vehicles and record road conditions in real time. This video data is processed frame by frame and pre-processed to recognize traffic conditions. Pre-processing includes noise reduction, color correction, resolution adjustment, etc.

[0133] Next, the server uses a method to analyze traffic conditions for each frame of pre-processed video data. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. For example, it recognizes whether a vehicle is about to turn left, whether a pedestrian is crossing the street, or whether a traffic light has changed from red to green.

[0134] Based on the results of traffic situation analysis, the server has a means to generate a natural language description using a generative AI model. The generated description is detailed and specific, allowing users to understand the traffic situation in real time without having to check the video directly.

[0135] The generated explanatory text is sent from the server to the user device. A particularly distinctive feature of this invention is that it provides a means for displaying the explanatory text in real time on smart glasses. The smart glasses are a device that allows drivers to instantly check traffic conditions while ensuring safety while driving.

[0136] For example, GPS information can be added to provide information about the actual location. In this case, an example of a prompt sentence would be "Currently, there are four vehicles ahead at the specified location, the traffic light is red, and a pedestrian is starting to cross the street."

[0137] This system provides drivers and autonomous driving developers with detailed and accurate real-time traffic conditions, supporting driving decisions and streamlining development.

[0138] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0139] Step 1:

[0140] The onboard camera is activated to begin capturing video data. The onboard camera is installed in front of the vehicle and records road conditions. The input is visual information on the road, and the output is video data.

[0141] Step 2:

[0142] The server receives video data sent from the in-vehicle camera in real time. The received video data is divided into frames. The input is the video data from the in-vehicle camera, and the output is the divided frame data.

[0143] Step 3:

[0144] The server pre-processes the video data, specifically noise removal, color correction, resolution adjustment, etc. The input is the divided frame data, and the output is the pre-processed frame data.

[0145] Step 4:

[0146] The server analyzes the traffic situation for each pre-processed frame. It detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. The input is the pre-processed frame data, and the output is the analysis results (traffic situation data).

[0147] Step 5:

[0148] The server generates a natural language explanation using a generative AI model based on the analysis results. The input is traffic situation data, and the output is the generated explanation.

[0149] Step 6:

[0150] The server provides the generated explanatory text to the user device (smart glasses). The input is the generated explanatory text, and the output is the text information displayed on the smart glasses.

[0151] Step 7:

[0152] Users can check traffic conditions in real time through smart glasses, allowing them to understand detailed traffic conditions without having to view the video directly. The input is the explanatory text displayed on the smart glasses, and the output is the user's perception of the traffic conditions.

[0153] 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.

[0154] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[0155] Video data acquisition and processing

[0156] The user turns on the dashcam and starts capturing video data. The server receives the video data sent from the dashcam in real time and divides it into frames. Then, it performs pre-processing on each frame, which includes noise reduction, color correction, and resolution adjustment.

[0157] Traffic situation analysis

[0158] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This allows for a detailed understanding of the traffic situation.

[0159] Generating Descriptions

[0160] The server uses a generative AI model to generate natural language descriptions based on the analysis results, which can provide detailed descriptions of traffic conditions, such as vehicle movements on the road, traffic light status, and pedestrian locations.

[0161] Providing a description

[0162] The server provides the generated explanatory text to the user's device (e.g., a smartphone or PC). The user device receives the explanatory text sent from the server and displays it to the user.

[0163] Adding an Emotion Engine

[0164] The system further includes an emotion engine that recognizes the user's emotion by analyzing the user's voice data and facial expression data.

[0165] Adjusting descriptions based on emotion

[0166] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server can increase the user's sense of security by providing more detailed explanations or changing them to more positive expressions.

[0167] Specific examples

[0168] For example, consider the following scenario:

[0169] 1. The user turns on the dashcam and begins capturing video data.

[0170] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0171] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0172] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0173] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[0174] 6. The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[0175] 7. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[0176] In this way, the system of the present invention can improve the user experience by providing explanations that take the user's emotions into consideration.

[0177] The processing flow will be explained below.

[0178] Step 1:

[0179] The user activates the dashcam, which then starts capturing video data in real time.

[0180] Step 2:

[0181] The server receives video data transmitted from the dashcam in real time, establishes a connection based on the specified camera ID, and starts receiving the data stream.

[0182] Step 3:

[0183] The server continuously divides the received video data into frames, so that the video data is treated as individual frames in small, easily processable time units.

[0184] Step 4:

[0185] The server performs pre-processing on each frame, including noise reduction, color correction, and resolution adjustment, to improve the quality of the frame data and increase the accuracy of subsequent analysis.

[0186] Step 5:

[0187] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their positions and movements. Object detection and motion analysis algorithms are used.

[0188] Step 6:

[0189] The server recognizes traffic conditions based on the results of object detection and motion analysis, and the information it obtains makes it possible to grasp detailed information such as the relative positions of vehicles, the status of traffic lights, and the movements of pedestrians.

[0190] Step 7:

[0191] The server passes the analysis results to a generative AI model, which then generates a natural language explanation. The generative AI model then automatically creates an appropriate explanation based on the traffic condition data.

[0192] Step 8:

[0193] The server transmits the generated explanatory text to the user device, which is an electronic device such as a smartphone or PC, and receives and displays the explanatory text to the user.

[0194] Step 9:

[0195] The emotion engine analyzes the user's voice data and facial expression data, identifies emotions from the user's voice and facial expressions, and recognizes the user's emotional state.

[0196] Step 10:

[0197] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server will make the explanation more detailed or change it to more positive language to increase the user's sense of security.

[0198] Specific examples

[0199] Example 1: Normal traffic conditions

[0200] 1. The user turns on the dashcam and begins capturing video data.

[0201] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0202] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0203] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0204] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[0205] Example 2: When the user feels anxious

[0206] 1. The user turns on the dashcam and begins capturing video data.

[0207] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0208] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0209] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0210] 5. The emotion engine analyzes the user's voice and facial expressions and recognizes that they are feeling anxious.

[0211] 6. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[0212] 7. The server sends the adjusted explanation to the user's device, and the user checks the explanation and feels reassured.

[0213] Through these steps, users can understand traffic conditions in real time and receive information tailored to their emotions.

[0214] Example 2

[0215] 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."

[0216] Current in-vehicle camera systems have difficulty recognizing traffic conditions and automatically generating explanatory text. Furthermore, they are unable to provide personalized information that takes into account the user's emotions, which means that safety and the user experience are not sufficiently improved. Therefore, there is a need for the development of a system that can analyze traffic conditions in real time and generate and provide explanatory text that reflects the user's emotional state.

[0217] 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.

[0218] In this invention, the server includes: means for acquiring video data from an in-vehicle camera; preprocessing means for dividing the acquired video data into frames and performing noise removal, color correction, and resolution adjustment; means for analyzing traffic conditions for each preprocessed frame, detecting objects such as vehicles, pedestrians, and traffic lights, and analyzing their positions and movements; means for generating a natural language description using a generative AI model based on the analyzed traffic conditions; means for providing the generated description to a user device; an emotion engine for analyzing a user's voice data and facial expression data to recognize the user's emotional state; and means for adjusting the content of the generated description based on the recognized user's emotional state. This makes it possible to analyze traffic conditions in detail in real time and generate and provide appropriate description that takes the user's emotional state into consideration.

[0219] An "on-board camera" is a device that is mounted on a vehicle and captures images of the road and surrounding area.

[0220] "Video data" refers to continuous image information acquired by an on-board camera.

[0221] A "frame" is an individual still image that makes up video data.

[0222] "Noise reduction" is a process that removes unnecessary signals and disturbances from video data.

[0223] "Color correction" is a process that adjusts the color balance of video data to bring it closer to the original color tone.

[0224] "Resolution adjustment" is a process of changing the number of pixels in the video data to maintain an appropriate resolution.

[0225] "Preprocessing" refers to a series of initial data processing steps, including noise removal, color correction, and resolution adjustment.

[0226] "Traffic conditions" refers to information about the status and movement of vehicles, pedestrians, and traffic lights on the road.

[0227] A "generative AI model" is an artificial intelligence model that generates explanatory text in natural language from analysis results.

[0228] A "description" is a natural language text generated based on the analyzed traffic situation.

[0229] A "user device" is a computing device used by a user, such as a smartphone or PC.

[0230] "Voice data" refers to digital audio information that records the user's speech or voice.

[0231] "Facial expression data" is digital image information that records the facial expressions of a user.

[0232] The "emotion engine" is a system that analyzes voice data and facial expression data to recognize the user's emotional state.

[0233] An "emotional state" is the psychological state that the user is currently feeling (for example, anxiety or relief).

[0234] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[0235] Video data acquisition and preprocessing

[0236] The user activates the in-vehicle camera to capture video data of road conditions. The in-vehicle camera is attached to the vehicle and captures video in real time. The video data is then sent to the server.

[0237] The server receives video data sent from the vehicle-mounted camera in real time. The received video data is divided into frames and then pre-processed. This pre-processing includes noise reduction, color correction, and resolution adjustment. This improves the quality of the video and enables accurate analysis.

[0238] Traffic situation analysis

[0239] The server analyzes the traffic situation for each pre-processed frame. Specifically, it uses deep learning models to detect objects such as vehicles, pedestrians, and traffic lights in the frame and analyzes their positions and movements, providing a detailed understanding of the current traffic situation.

[0240] Generating Descriptions

[0241] The server uses a generative AI model based on the analysis results to convert the traffic situation into a natural language description. For example, if the analysis result is "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk," the server inputs this analysis result as a prompt into the generative AI model to generate a description. A specific description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk" is output.

[0242] Providing a description

[0243] The generated description is sent from the server to the user's device (e.g., a smartphone or PC). The user's device receives the description and displays it on the screen to inform the user of the current traffic situation.

[0244] Emotion engine integration

[0245] Furthermore, the system includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes the user's voice data and facial expression data to recognize the user's emotional state. This allows the system to respond in accordance with the user's emotions.

[0246] Adjusted description

[0247] The server adjusts the generated explanation based on the user's emotional state as recognized by the emotion engine. For example, if the user is feeling anxious, the explanation can be made more detailed to increase the sense of security. For example, it can be adjusted to say, "There are four vehicles ahead, the traffic light is red, and there are pedestrians at the crosswalk, so please proceed safely and with caution."

[0248] Specific examples

[0249] The user activates the in-vehicle camera and starts capturing video data.

[0250] The server receives the video data, performs preprocessing, and then analyzes the traffic situation.

[0251] The analysis results are input into a generative AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0252] The server sends the explanation to the user's device, which the user confirms.

[0253] The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[0254] The server adjusts the description to read, "Proceed safely. The light is red, please stop."

[0255] This allows users to accurately grasp the current traffic situation and receive detailed information tailored to their emotions, significantly improving user safety and experience.

[0256] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0257] Step 1:

[0258] The user activates the in-vehicle camera, which captures road conditions in real time and acquires video data.

[0259] Input: Real-time footage from a dashcam.

[0260] Specific operation: The user turns on the dashcam and launches the video capture application.

[0261] Step 2:

[0262] The server receives video data sent from the vehicle-mounted camera in real time and divides the received video data into frames.

[0263] Input: Real-time video data from a dashcam.

[0264] Output: Split frame data.

[0265] Specific operation: The server executes a script that divides the video data into frames and prepares each frame for processing.

[0266] Step 3:

[0267] The server performs pre-processing on the divided frames, which includes noise reduction, color correction, and resolution adjustment.

[0268] Input: Split frame data.

[0269] Output: Preprocessed frame data.

[0270] What happens: The server uses image processing algorithms to remove noise, perform color correction, and adjust the resolution of each frame.

[0271] Step 4:

[0272] The server analyzes the preprocessed frames and identifies traffic situations. Specifically, it uses deep learning models to detect objects such as vehicles, pedestrians, and traffic lights. It analyzes the position and movement of each object.

[0273] Input: Preprocessed frame data.

[0274] Output: Traffic condition analysis result data.

[0275] How it works: The server applies deep learning models to identify the location and movement of vehicles, pedestrians, and traffic lights in each frame.

[0276] Step 5:

[0277] The server uses a generative AI model based on the analysis results to generate a natural language description.

[0278] Input: Traffic analysis result data.

[0279] Output: The generated natural language description.

[0280] Specific operation: The server inputs the analysis results as a prompt sentence to the generative AI model and obtains the generated explanation sentence.

[0281] Example prompt: "There are four vehicles ahead, the light is red, and a pedestrian is crossing the street."

[0282] Step 6:

[0283] The server sends the generated explanation to the user's device, which receives the explanation and displays it to the user.

[0284] Input: The generated natural language description.

[0285] Output: Description that will be displayed on the user's device.

[0286] Specific operation: The server sends the explanatory text to the user's smartphone or PC, and the user device receives it and displays it on the screen.

[0287] Step 7:

[0288] The server uses an emotion engine to analyze the user's voice data and facial expression data to recognize emotions.

[0289] Input: User's voice and facial expression data.

[0290] Output: Perceived emotional state.

[0291] Specific operation: The user provides voice data and facial expression data using a microphone or camera, and the server analyzes it using an emotion engine.

[0292] Step 8:

[0293] The server adjusts the content of the generated narrative based on the recognized emotional state of the user.

[0294] Input: A recognized emotional state and a generated natural language description.

[0295] Output: A narrative tailored based on sentiment.

[0296] Specific operation: Based on the results of emotion analysis, the server adjusts the output of the generative AI model and changes the description to take the user's emotions into consideration.

[0297] This allows the user to accurately grasp the current traffic situation and receive detailed information tailored to their emotions.

[0298] (Application example 2)

[0299] 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."

[0300] In conventional autonomous vehicle systems, there are methods for analyzing in-vehicle camera footage to inform users of traffic conditions, but the feedback provided does not take into account the user's emotional state, which can lead to anxiety or a lack of understanding of the situation. Furthermore, the explanations of traffic conditions are uniform, making it difficult to provide information tailored to the needs of individual users.

[0301] The identification processing 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 acquiring video data from an in-vehicle camera, means for analyzing the acquired video data frame by frame and recognizing traffic conditions, means for automatically generating an explanatory text based on the analyzed traffic conditions, means for providing the generated explanatory text to a user device, means for recognizing the user's emotional state, and means for adjusting the content of the generated explanatory text based on the recognized emotional state. This allows the user to use an autonomous vehicle with peace of mind by providing detailed information about traffic conditions and feedback adjusted according to the user's emotional state.

[0302] An "on-board camera" is a camera device that is mounted on a vehicle and captures video data of the surrounding area in real time.

[0303] "Video data" refers to continuous image information acquired by an on-board camera, and is basic data for traffic situation analysis.

[0304] A "frame" is a unit of an individual still image that makes up video data.

[0305] "Traffic conditions" refers to objects on the road, such as vehicles, pedestrians, and traffic lights, as well as their condition and position.

[0306] The "explanation" is a sentence generated to convey the analyzed traffic situation to the user in natural language.

[0307] "User device" refers to a device owned by a user that can display and operate information, such as a smartphone or tablet.

[0308] "Emotional state" indicates the user's mental and sensory response, and is primarily recognized from voice and facial expression data.

[0309] A "server" is a computer system that processes and analyzes various types of data and is managed centrally.

[0310] A "generative AI model" is an artificial intelligence model that uses machine learning technology to automatically generate sentences corresponding to specific tasks.

[0311] "Adjusting" means changing the content or format of the product based on the perceived information, for example, modifying the generated description based on the user's emotional state.

[0312] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[0313] Video data acquisition and processing

[0314] The user starts the dashcam and starts capturing video data. The server receives the video data sent from the dashcam in real time and divides it into frames. Next, preprocessing is performed on each frame. This includes noise reduction, color correction, and resolution adjustment. Noise reduction, color correction, and resolution adjustment are achieved using libraries such as OpenCV.

[0315] Traffic situation analysis

[0316] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it uses a deep learning model (such as YOLO) to detect objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This allows for a detailed understanding of the traffic situation.

[0317] Generating Descriptions

[0318] The server uses a generative AI model (e.g., GPT-4) based on the analysis results to generate natural language descriptions, which can provide detailed descriptions of traffic situations, such as vehicle movements on the road, traffic light status, and pedestrian positions. Specific examples of prompts include the following:

[0319] Input: Number of vehicles: 4, Traffic light status: red, Pedestrian position: on crosswalk

[0320] Prompt: Generate a description based on four vehicles ahead, a red light, and pedestrians crossing the street.

[0321] Providing a description

[0322] The server provides the generated explanatory text to a user device such as a user's smartphone, tablet, etc. The user device receives the explanatory text sent from the server and displays it to the user.

[0323] Adding an Emotion Engine

[0324] This system also includes an emotion engine that recognizes the user's emotions. The emotion engine has the function of recognizing the user's emotional state by analyzing the user's voice data and facial expression data. Azure Face API is used to analyze facial expressions, and Amazon Polly is used to analyze voice data. As a result, the user's emotional state is recognized as "anxiety."

[0325] Adjusting descriptions based on emotion

[0326] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server can increase the user's sense of security by making the explanation more detailed or changing it to more positive language. Specific examples of prompts include the following:

[0327] Input: User emotion: Anxiety

[0328] Prompt: The user is feeling anxious. Please regenerate the description with more detail and positive wording.

[0329] Specific examples

[0330] For example, consider the following scenario:

[0331] 1. The user turns on the dashcam and begins capturing video data.

[0332] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0333] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0334] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0335] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[0336] 6. The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[0337] 7. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[0338] In this way, the system of the present invention can improve the user experience by providing explanations that take the user's emotions into consideration.

[0339] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0340] Step 1:

[0341] The user activates the in-vehicle camera and starts capturing video data.

[0342] Input: Start instruction for in-car camera

[0343] Output: Start of real-time video data acquisition

[0344] Specific operation: The user activates the dashcam from the smartphone app.

[0345] Step 2:

[0346] The server receives video data transmitted from the vehicle-mounted camera in real time and divides it into frames.

[0347] Input: Real-time video data

[0348] Output: List of frame data

[0349] Specific operation: Video data is divided into frames and a timestamp is stored for each frame.

[0350] Step 3:

[0351] The server performs pre-processing on each frame, which includes noise reduction, color correction, and resolution adjustment.

[0352] Input: List of frame data

[0353] Output: Preprocessed frame data

[0354] What it does: It uses the OpenCV library to perform noise reduction, color correction, and resolution adjustment on each frame.

[0355] Step 4:

[0356] The server performs traffic situation analysis for each frame after preprocessing. Specifically, it uses a deep learning model (e.g., YOLO) to detect objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object.

[0357] Input: Preprocessed frame data

[0358] Output: Information about detected objects (type, location, movement, etc.)

[0359] Specific operation: The YOLO model is applied to frame data, and the identification of objects indicating traffic conditions and their location information are output.

[0360] Step 5:

[0361] The server generates a natural language explanation based on the analysis results using a generative AI model (e.g., GPT-4).

[0362] Input: Object information (type, position, movement, etc.)

[0363] Output: Generated natural language description

[0364] Specific operation: Based on the analysis results, a prompt sentence is input into the AI ​​model to generate an explanatory sentence. Specific prompt sentence examples:

[0365] Input: Number of vehicles: 4, Traffic light status: red, Pedestrian position: on crosswalk

[0366] Prompt: Generate a description based on four vehicles ahead, a red light, and pedestrians crossing the street.

[0367] Step 6:

[0368] The server provides the generated explanation to the user device.

[0369] Input: Generated description

[0370] Output: Send to user device

[0371] Specific operation: The generated explanation is sent to the user's smartphone or tablet and displayed.

[0372] Step 7:

[0373] The server acquires the user's voice data and facial expression data and recognizes their emotional state using an emotion engine.

[0374] Input: Voice data, facial expression data

[0375] Output: Emotional state recognition result

[0376] Specific operation: The voice data and facial expression data collected by the user device are sent to the server and analyzed using an emotion engine (e.g., Azure Face API or Amazon Polly).

[0377] Step 8:

[0378] The content of the generated description is adjusted based on the emotional state recognized by the server.

[0379] Input: Emotional state recognition results, generated description

[0380] Output: Adjusted description

[0381] Specific behavior: Depending on the recognized emotional state, the generative AI model is re-prompted to generate an adjusted explanation. Specific prompt examples:

[0382] Input: User emotion: Anxiety

[0383] Prompt: The user is feeling anxious. Please regenerate the description with more detail and positive wording.

[0384] Step 9:

[0385] The server provides the adjusted explanation back to the user device.

[0386] Input: Adjusted description

[0387] Output: Retransmission to user device

[0388] Specific operation: The adjusted description is resent to the user's smartphone or tablet and displayed.

[0389] This provides detailed information about traffic conditions and tailored feedback based on the user's emotional state, allowing users to use self-driving vehicles with peace of mind.

[0390] 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.

[0391] 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.

[0392] 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.

[0393] [Second embodiment]

[0394] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0395] 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.

[0396] 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).

[0397] 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.

[0398] 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.

[0399] 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).

[0400] 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.

[0401] 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.

[0402] 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.

[0403] 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.

[0404] 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.

[0405] 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."

[0406] The present invention is a system that automatically converts in-vehicle camera images into explanatory text about traffic conditions, and aims to streamline the development of autonomous driving and advanced driver assistance systems (ADAS). This system includes the following main means:

[0407] Acquiring video data

[0408] The user starts the in-vehicle camera and starts capturing video data. The in-vehicle camera is installed in front of the vehicle and records road conditions in real time.

[0409] Receiving and analyzing video data

[0410] The server receives video data sent from the in-vehicle camera. The received video data is divided into frames and pre-processed to analyze the traffic situation within each frame. This pre-processing includes noise removal, color correction, and resolution adjustment.

[0411] Traffic situation recognition

[0412] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. For example, it recognizes whether a vehicle is about to turn left, whether a pedestrian is crossing the street, or whether a traffic light has changed from red to green.

[0413] Generating Descriptions

[0414] The server generates a natural language description based on the analysis results. Using a generative AI model, detailed text about traffic conditions is automatically created, allowing users to understand real-time traffic conditions without having to check the video.

[0415] Providing a description

[0416] The server then provides the generated description to the user device, which is an electronic device such as a smartphone or PC that receives and displays the description, allowing the user to check detailed information about traffic conditions in real time.

[0417] Specific examples

[0418] For example, in a scenario describing traffic conditions on a given day, the following sequence of actions occurs:

[0419] 1. The user turns on the dashcam and begins capturing video.

[0420] 2. The server receives the video data and analyzes it frame by frame.

[0421] 3. The server recognizes that there are four vehicles ahead, the traffic light is red, and detects that a pedestrian is beginning to cross the street.

[0422] 4. The server passes the analysis results to the generation AI, which automatically generates a description such as "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0423] 5. The server generates a description and sends it to the user's device, where the user can view this information in real time.

[0424] This allows users to understand traffic conditions in detail through text without having to look at the video directly. This system enables the efficient generation of learning data in the development of autonomous driving and ADAS, greatly streamlining the development process.

[0425] The above is a detailed description of the embodiment of the present invention.

[0426] The processing flow will be explained below.

[0427] Step 1:

[0428] The user activates the dashcam, which then starts capturing video data. Specifically, the user turns on the power of the camera device in the vehicle, and a real-time stream of video data begins.

[0429] Step 2:

[0430] The server receives video data sent from the dashcam in real time. The server establishes a connection based on the specified camera ID and continuously acquires data.

[0431] Step 3:

[0432] The server divides the received video data into frames, converting the video data into frames for each small time unit, and these frames are then used for subsequent analysis processing.

[0433] Step 4:

[0434] The server performs pre-processing on each frame, including noise reduction, color correction, and resolution adjustment, to improve the quality of the frame data.

[0435] Step 5:

[0436] The server performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This is done using object detection and motion analysis algorithms.

[0437] Step 6:

[0438] The server recognizes traffic conditions based on the results of object detection and motion analysis, including vehicle position, traffic light status, and pedestrian movement.

[0439] Step 7:

[0440] The server passes the analysis results to a generative AI model, which then generates a natural language explanation. The generative AI model converts the analyzed traffic situation data into text and automatically creates a detailed explanation.

[0441] Step 8:

[0442] The server then sends the generated explanatory text to the user device, which may be a smartphone or PC, and receives the explanatory text from the server and displays it to the user.

[0443] Step 9:

[0444] Users can view the description on their device and get a detailed understanding of the real-time traffic situation. The textual information allows users to intuitively understand the situation ahead of the vehicle and other important traffic information.

[0445] Example 1

[0446] 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."

[0447] In the development of autonomous driving and advanced driver assistance systems (ADAS), it is important to accurately and efficiently grasp traffic conditions. However, conventional methods require manual analysis of vast amounts of video data, which is time-consuming and costly. Furthermore, it is difficult to grasp traffic conditions in real time, which reduces development efficiency. This invention aims to solve these problems and significantly streamline the development process for autonomous driving and ADAS.

[0448] 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.

[0449] In this invention, the server includes means for acquiring video data from an on-board camera, means for preprocessing the acquired video data frame by frame, means for recognizing traffic conditions for each preprocessed frame, means for automatically generating explanatory text based on the recognized traffic conditions, and means for providing the generated explanatory text to a user device. This enables real-time analysis of video data and efficient development of autonomous driving and ADAS.

[0450] An "in-vehicle camera" is a camera installed in a vehicle to record road and traffic conditions.

[0451] "Video data" refers to video captured by a camera and stored in digital format.

[0452] A "frame" refers to each of the consecutive still images that make up video data.

[0453] "Preprocessing" refers to initial processing such as noise removal, color correction, and resolution adjustment that is performed to make video data easier to analyze.

[0454] "Traffic conditions" refers to information that indicates the overall state and movement of vehicles, pedestrians, traffic lights, etc. on the road.

[0455] "Recognition" is the process of detecting and understanding specific objects and their movements from video data.

[0456] A "description" is a text that describes the recognized traffic situation in natural language.

[0457] A "generative AI model" is an artificial intelligence model that automatically generates text based on input data.

[0458] A "user device" is an electronic device for displaying the generated description, and includes a smartphone, a PC, etc.

[0459] The system according to the present invention automatically converts images from an in-vehicle camera into explanatory text about traffic conditions, and aims to improve the efficiency of the development of autonomous driving and advanced driver assistance systems (ADAS). Specific embodiments of this system are described below.

[0460] First, the user activates the dashcam to begin capturing video data. The dashcam is installed in the front of the car and records road conditions in real time while the car is moving. This camera uses a standard high-resolution camera and captures video data at 30 frames per second (FPS).

[0461] The captured video data is then sent to a server in real time. The server divides the video data into frames and performs preprocessing. This preprocessing includes processes such as noise reduction using the OpenCV library, color correction, and resolution adjustment. For example, noise reduction involves applying a filter to remove unwanted background signals. Color correction involves adjusting the color balance of each frame to improve visibility.

[0462] After preprocessing, the server recognizes traffic conditions for each frame. During this process, it runs an object detection algorithm to detect objects such as vehicles, pedestrians, and traffic lights. For example, it can use the YOLO (You Only Look Once) model to detect objects quickly and accurately. It analyzes the position and movement of detected objects in each frame to understand traffic flow. Examples of identified objects include vehicles approaching an intersection, pedestrians crossing a crosswalk, and traffic lights changing from red to green.

[0463] Based on the analysis results, the server uses a generative AI model to generate a description in natural language. One example of a generative AI model that can be used is GPT-3. The server passes the analysis results to the generative AI as a prompt, which generates a description like the following: "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk." The generated description is formatted in a way that is easy for the user to understand.

[0464] Finally, the server sends the generated description in real time to the user device. The user device is an electronic device such as a smartphone or PC, which displays the description on its screen. The user can check the traffic situation in text form through the application. The display on the user device shows, "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0465] As a concrete example, here is a scenario describing traffic conditions on a given day:

[0466] 1. The user activates the dashcam using a smartphone app and begins capturing footage.

[0467] 2. The server receives the video data in real time and performs noise reduction, color correction, and resolution adjustment on each frame.

[0468] 3. The server uses the YOLO model to detect vehicles, pedestrians, and traffic lights ahead and analyze traffic conditions.

[0469] 4. The server uses GPT-3 to generate an explanation based on the analysis results, generating the sentence, "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0470] 5. The server generates a description and sends it to the smartphone, where the user can view the information in real time through the app.

[0471] This allows users to understand traffic conditions in detail through text without having to look at the video directly. This system enables efficient generation of learning data for the development of autonomous driving and ADAS, greatly streamlining the development process.

[0472] Example prompt sentence:

[0473] "Analyze the traffic situation ahead from the specified video data and generate a description including the status of vehicles, pedestrians, and traffic lights."

[0474] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0475] Step 1:

[0476] The user activates the dashcam and begins capturing video data. Specifically, the user presses the camera's activation button using a smartphone app, and video recording begins from that point. The camera is installed at the front of the car and records video at 30 frames per second (FPS). The input is the user's camera activation operation and the real-time video scene. The output is high-resolution video data obtained from the camera.

[0477] Step 2:

[0478] The server receives video data sent from the in-vehicle camera in real time. The received video data is 30 frames per second (FPS) and is received and stored on the server. The input is the video data from the in-vehicle camera. The output is the raw video data stored on the server.

[0479] Step 3:

[0480] The server pre-processes the received video data for each frame. Specifically, it uses the OpenCV library to perform noise reduction, color correction, and resolution adjustment. For example, noise reduction applies a Gaussian filter, color correction adjusts the RGB balance, and resolution adjustment resizes the data to a consistent size. The input is the raw video data (in frame format) stored on the server. The output is the pre-processed frame data.

[0481] Step 4:

[0482] The server recognizes traffic conditions for each preprocessed frame. Specifically, it uses the YOLO (You Only Look Once) model to detect objects such as vehicles, pedestrians, and traffic lights. It tracks the location and movement of each detected object to understand traffic flow and conditions. For example, it detects changes in the color of traffic lights and the direction of pedestrian movement. The input is the preprocessed frame data. The output is the analysis result of the traffic conditions for each frame.

[0483] Step 5:

[0484] The server uses a generative AI model based on the analysis results to generate a natural language description. Specifically, it utilizes a large-scale language model such as GPT-3 and provides the traffic situation analysis results obtained in the previous step to the model as a prompt. For example, it generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk." The input is the traffic situation analysis results and the prompt. The output is the generated natural language description.

[0485] Step 6:

[0486] The server sends the generated description to the user device in real time. The user device is an electronic device such as a smartphone or PC, which displays the received description on its screen. The input is the generated natural language description. The output is the description displayed on the user device. For example, the app's display might say, "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0487] (Application example 1)

[0488] 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."

[0489] Conventional autonomous vehicles and advanced driver assistance systems (ADAS) require drivers and developers to directly check video data to understand traffic conditions, which places a significant burden on the driver and developers. Additionally, it is difficult to grasp traffic conditions in real time, resulting in delayed reactions in situations where instant decisions are required. Furthermore, relying solely on vision means that information acquisition is incomplete, potentially increasing the risk of accidents and traffic violations.

[0490] 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.

[0491] In this invention, the server includes means for acquiring video data from an on-board camera, means for analyzing the acquired video data frame by frame and recognizing traffic conditions, means for automatically generating explanatory text based on the analyzed traffic conditions, means for providing the generated explanatory text to a user device, and means for displaying the generated explanatory text in real time on smart glasses, which are a type of user device. This enables drivers and developers of autonomous driving systems to grasp detailed and accurate traffic conditions in real time without directly viewing the video.

[0492] An "on-board camera" is a camera device mounted on a vehicle that captures road and surrounding conditions as video data in real time.

[0493] "Video data" refers to continuous image information captured by an on-board camera, and visually records traffic conditions and the surrounding environment.

[0494] A "frame" refers to a series of individual images of video data that are treated as a single still image.

[0495] "Traffic conditions" refers to the state of various elements on the road, such as vehicles, pedestrians, and traffic lights, and their interrelationships.

[0496] A "user device" is an electronic device that can receive explanatory text sent from a server and display it to a user.

[0497] "Analysis" is the process of processing video data to extract and understand specific information.

[0498] "Explanation" refers to text information written in natural language based on traffic situation information obtained through analysis.

[0499] "Real time" refers to processing and display occurring immediately without delay.

[0500] "Smart glasses" are glasses-type electronic devices that have the function of displaying images and text information, allowing the wearer to view information through the device.

[0501] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate text from given data.

[0502] This invention is a system that automatically converts in-vehicle camera footage into traffic situation descriptions and provides them to user devices (especially smart glasses) in real time. The purpose of this system is to streamline the development of autonomous vehicles and advanced driver assistance systems (ADAS) and to provide drivers with instant, detailed traffic situation information.

[0503] The server has a means to acquire video data from dashcams. The dashcams are attached to vehicles and record road conditions in real time. This video data is processed frame by frame and pre-processed to recognize traffic conditions. Pre-processing includes noise reduction, color correction, and resolution adjustment.

[0504] Next, the server uses a method to analyze traffic conditions for each frame of pre-processed video data. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. For example, it recognizes whether a vehicle is about to turn left, whether a pedestrian is crossing the street, or whether a traffic light has changed from red to green.

[0505] Based on the results of traffic situation analysis, the server has a means to generate a natural language description using a generative AI model. The generated description is detailed and specific, allowing users to understand the traffic situation in real time without having to check the video directly.

[0506] The generated explanatory text is sent from the server to the user device. A particularly distinctive feature of this invention is that it provides a means for displaying the explanatory text in real time on smart glasses. The smart glasses are a device that allows drivers to instantly check traffic conditions while ensuring safety while driving.

[0507] For example, GPS information can be added to provide information about the actual location. In this case, an example of a prompt sentence would be "Currently, there are four vehicles ahead at the specified location, the traffic light is red, and a pedestrian is starting to cross the street."

[0508] This system provides drivers and autonomous driving developers with detailed and accurate real-time traffic conditions, supporting driving decisions and streamlining development.

[0509] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0510] Step 1:

[0511] The onboard camera is activated to begin capturing video data. The onboard camera is installed in front of the vehicle and records road conditions. The input is visual information on the road, and the output is video data.

[0512] Step 2:

[0513] The server receives video data sent from the in-vehicle camera in real time. The received video data is divided into frames. The input is the video data from the in-vehicle camera, and the output is the divided frame data.

[0514] Step 3:

[0515] The server pre-processes the video data, specifically noise removal, color correction, resolution adjustment, etc. The input is the divided frame data, and the output is the pre-processed frame data.

[0516] Step 4:

[0517] The server analyzes the traffic situation for each pre-processed frame. It detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. The input is the pre-processed frame data, and the output is the analysis results (traffic situation data).

[0518] Step 5:

[0519] The server generates a natural language explanation using a generative AI model based on the analysis results. The input is traffic situation data, and the output is the generated explanation.

[0520] Step 6:

[0521] The server provides the generated explanatory text to the user device (smart glasses). The input is the generated explanatory text, and the output is the text information displayed on the smart glasses.

[0522] Step 7:

[0523] Users can check traffic conditions in real time through smart glasses, allowing them to understand detailed traffic conditions without having to view the video directly. The input is the explanatory text displayed on the smart glasses, and the output is the user's perception of the traffic conditions.

[0524] 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.

[0525] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[0526] Video data acquisition and processing

[0527] The user turns on the dashcam and starts capturing video data. The server receives the video data sent from the dashcam in real time and divides it into frames. Then, it performs pre-processing on each frame, which includes noise reduction, color correction, and resolution adjustment.

[0528] Traffic situation analysis

[0529] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This allows for a detailed understanding of the traffic situation.

[0530] Generating Descriptions

[0531] The server uses a generative AI model to generate natural language descriptions based on the analysis results, which can provide detailed descriptions of traffic conditions, such as vehicle movements on the road, traffic light status, and pedestrian locations.

[0532] Providing a description

[0533] The server provides the generated explanatory text to the user's device (e.g., a smartphone or PC). The user device receives the explanatory text sent from the server and displays it to the user.

[0534] Adding an Emotion Engine

[0535] The system further includes an emotion engine that recognizes the user's emotion by analyzing the user's voice data and facial expression data.

[0536] Adjusting descriptions based on emotion

[0537] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server can increase the user's sense of security by providing more detailed explanations or changing them to more positive expressions.

[0538] Specific examples

[0539] For example, consider the following scenario:

[0540] 1. The user turns on the dashcam and begins capturing video data.

[0541] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0542] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0543] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0544] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[0545] 6. The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[0546] 7. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[0547] In this way, the system of the present invention can improve the user experience by providing explanations that take the user's emotions into consideration.

[0548] The processing flow will be explained below.

[0549] Step 1:

[0550] The user activates the dashcam, which then starts capturing video data in real time.

[0551] Step 2:

[0552] The server receives video data transmitted from the dashcam in real time, establishes a connection based on the specified camera ID, and starts receiving the data stream.

[0553] Step 3:

[0554] The server continuously divides the received video data into frames, so that the video data is treated as individual frames in small, easily processable time units.

[0555] Step 4:

[0556] The server performs pre-processing on each frame, including noise reduction, color correction, and resolution adjustment, to improve the quality of the frame data and increase the accuracy of subsequent analysis.

[0557] Step 5:

[0558] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their positions and movements. Object detection and motion analysis algorithms are used.

[0559] Step 6:

[0560] The server recognizes traffic conditions based on the results of object detection and motion analysis, and the information it obtains makes it possible to grasp detailed information such as the relative positions of vehicles, the status of traffic lights, and the movements of pedestrians.

[0561] Step 7:

[0562] The server passes the analysis results to a generative AI model, which then generates a natural language explanation. The generative AI model then automatically creates an appropriate explanation based on the traffic condition data.

[0563] Step 8:

[0564] The server transmits the generated explanatory text to the user device, which is an electronic device such as a smartphone or PC, and receives and displays the explanatory text to the user.

[0565] Step 9:

[0566] The emotion engine analyzes the user's voice data and facial expression data, identifies emotions from the user's voice and facial expressions, and recognizes the user's emotional state.

[0567] Step 10:

[0568] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server will make the explanation more detailed or change it to more positive language to increase the user's sense of security.

[0569] Specific examples

[0570] Example 1: Normal traffic conditions

[0571] 1. The user turns on the dashcam and begins capturing video data.

[0572] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0573] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0574] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0575] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[0576] Example 2: When the user feels anxious

[0577] 1. The user turns on the dashcam and begins capturing video data.

[0578] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0579] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0580] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0581] 5. The emotion engine analyzes the user's voice and facial expressions and recognizes that they are feeling anxious.

[0582] 6. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[0583] 7. The server sends the adjusted explanation to the user's device, and the user checks the explanation and feels reassured.

[0584] Through these steps, users can understand traffic conditions in real time and receive information tailored to their emotions.

[0585] Example 2

[0586] 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."

[0587] Current in-vehicle camera systems have difficulty recognizing traffic conditions and automatically generating explanatory text. Furthermore, they are unable to provide personalized information that takes into account the user's emotions, which means that safety and the user experience are not sufficiently improved. Therefore, there is a need for the development of a system that can analyze traffic conditions in real time and generate and provide explanatory text that reflects the user's emotional state.

[0588] 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.

[0589] In this invention, the server includes: means for acquiring video data from an in-vehicle camera; preprocessing means for dividing the acquired video data into frames and performing noise removal, color correction, and resolution adjustment; means for analyzing traffic conditions for each preprocessed frame, detecting objects such as vehicles, pedestrians, and traffic lights, and analyzing their positions and movements; means for generating a natural language description using a generative AI model based on the analyzed traffic conditions; means for providing the generated description to a user device; an emotion engine for analyzing a user's voice data and facial expression data to recognize the user's emotional state; and means for adjusting the content of the generated description based on the recognized user's emotional state. This makes it possible to analyze traffic conditions in detail in real time and generate and provide appropriate description that takes the user's emotional state into consideration.

[0590] An "on-board camera" is a device that is mounted on a vehicle and captures images of the road and surrounding area.

[0591] "Video data" refers to continuous image information acquired by an on-board camera.

[0592] A "frame" is an individual still image that makes up video data.

[0593] "Noise reduction" is a process that removes unnecessary signals and disturbances from video data.

[0594] "Color correction" is a process that adjusts the color balance of video data to bring it closer to the original color tone.

[0595] "Resolution adjustment" is a process of changing the number of pixels in the video data to maintain an appropriate resolution.

[0596] "Preprocessing" refers to a series of initial data processing steps, including noise removal, color correction, and resolution adjustment.

[0597] "Traffic conditions" refers to information about the status and movement of vehicles, pedestrians, and traffic lights on the road.

[0598] A "generative AI model" is an artificial intelligence model that generates explanatory text in natural language from analysis results.

[0599] A "description" is a natural language text generated based on the analyzed traffic situation.

[0600] A "user device" is a computing device used by a user, such as a smartphone or PC.

[0601] "Voice data" refers to digital audio information that records the user's speech or voice.

[0602] "Facial expression data" is digital image information that records the facial expressions of a user.

[0603] The "emotion engine" is a system that analyzes voice data and facial expression data to recognize the user's emotional state.

[0604] An "emotional state" is the psychological state that the user is currently feeling (for example, anxiety or relief).

[0605] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[0606] Video data acquisition and preprocessing

[0607] The user activates the in-vehicle camera to capture video data of road conditions. The in-vehicle camera is attached to the vehicle and captures video in real time. The video data is then sent to the server.

[0608] The server receives video data sent from the vehicle-mounted camera in real time. The received video data is divided into frames and then pre-processed. This pre-processing includes noise reduction, color correction, and resolution adjustment. This improves the quality of the video and enables accurate analysis.

[0609] Traffic situation analysis

[0610] The server analyzes the traffic situation for each pre-processed frame. Specifically, it uses deep learning models to detect objects such as vehicles, pedestrians, and traffic lights in the frame and analyzes their positions and movements, providing a detailed understanding of the current traffic situation.

[0611] Generating Descriptions

[0612] The server uses a generative AI model based on the analysis results to convert the traffic situation into a natural language description. For example, if the analysis result is "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk," the server inputs this analysis result as a prompt into the generative AI model to generate a description. A specific description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk" is output.

[0613] Providing a description

[0614] The generated description is sent from the server to the user's device (e.g., a smartphone or PC). The user's device receives the description and displays it on the screen to inform the user of the current traffic situation.

[0615] Emotion engine integration

[0616] Furthermore, the system includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes the user's voice data and facial expression data to recognize the user's emotional state. This allows the system to respond in accordance with the user's emotions.

[0617] Adjusted description

[0618] The server adjusts the generated explanation based on the user's emotional state as recognized by the emotion engine. For example, if the user is feeling anxious, the explanation can be made more detailed to increase the sense of security. For example, it can be adjusted to say, "There are four vehicles ahead, the traffic light is red, and there are pedestrians at the crosswalk, so please proceed safely and with caution."

[0619] Specific examples

[0620] The user activates the in-vehicle camera and starts capturing video data.

[0621] The server receives the video data, performs preprocessing, and then analyzes the traffic situation.

[0622] The analysis results are input into a generative AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0623] The server sends the explanation to the user's device, which the user confirms.

[0624] The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[0625] The server adjusts the description to read, "Proceed safely. The light is red, please stop."

[0626] This allows users to accurately grasp the current traffic situation and receive detailed information tailored to their emotions, significantly improving user safety and experience.

[0627] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0628] Step 1:

[0629] The user activates the in-vehicle camera, which captures road conditions in real time and acquires video data.

[0630] Input: Real-time footage from a dashcam.

[0631] Specific operation: The user turns on the dashcam and launches the video capture application.

[0632] Step 2:

[0633] The server receives video data sent from the vehicle-mounted camera in real time and divides the received video data into frames.

[0634] Input: Real-time video data from a dashcam.

[0635] Output: Split frame data.

[0636] Specific operation: The server executes a script that divides the video data into frames and prepares each frame for processing.

[0637] Step 3:

[0638] The server performs pre-processing on the divided frames, which includes noise reduction, color correction, and resolution adjustment.

[0639] Input: Split frame data.

[0640] Output: Preprocessed frame data.

[0641] What happens: The server uses image processing algorithms to remove noise, perform color correction, and adjust the resolution of each frame.

[0642] Step 4:

[0643] The server analyzes the preprocessed frames and identifies traffic situations. Specifically, it uses deep learning models to detect objects such as vehicles, pedestrians, and traffic lights. It analyzes the position and movement of each object.

[0644] Input: Preprocessed frame data.

[0645] Output: Traffic condition analysis result data.

[0646] How it works: The server applies deep learning models to identify the location and movement of vehicles, pedestrians, and traffic lights in each frame.

[0647] Step 5:

[0648] The server uses a generative AI model based on the analysis results to generate a natural language description.

[0649] Input: Traffic analysis result data.

[0650] Output: The generated natural language description.

[0651] Specific operation: The server inputs the analysis results as a prompt sentence to the generative AI model and obtains the generated explanation sentence.

[0652] Example prompt: "There are four vehicles ahead, the light is red, and a pedestrian is crossing the street."

[0653] Step 6:

[0654] The server sends the generated explanation to the user's device, which receives the explanation and displays it to the user.

[0655] Input: The generated natural language description.

[0656] Output: Description that will be displayed on the user's device.

[0657] Specific operation: The server sends the explanatory text to the user's smartphone or PC, and the user device receives it and displays it on the screen.

[0658] Step 7:

[0659] The server uses an emotion engine to analyze the user's voice data and facial expression data to recognize emotions.

[0660] Input: User's voice and facial expression data.

[0661] Output: Perceived emotional state.

[0662] Specific operation: The user provides voice data and facial expression data using a microphone or camera, and the server analyzes it using an emotion engine.

[0663] Step 8:

[0664] The server adjusts the content of the generated narrative based on the recognized emotional state of the user.

[0665] Input: A recognized emotional state and a generated natural language description.

[0666] Output: A narrative tailored based on sentiment.

[0667] Specific operation: Based on the results of emotion analysis, the server adjusts the output of the generative AI model and changes the description to take the user's emotions into consideration.

[0668] This allows the user to accurately grasp the current traffic situation and receive detailed information tailored to their emotions.

[0669] (Application example 2)

[0670] 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."

[0671] In conventional autonomous vehicle systems, there are methods for analyzing in-vehicle camera footage to inform users of traffic conditions, but the feedback provided does not take into account the user's emotional state, which can lead to anxiety or a lack of understanding of the situation. Furthermore, the explanations of traffic conditions are uniform, making it difficult to provide information tailored to the needs of individual users.

[0672] The identification processing 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 acquiring video data from an in-vehicle camera, means for analyzing the acquired video data frame by frame and recognizing traffic conditions, means for automatically generating an explanatory text based on the analyzed traffic conditions, means for providing the generated explanatory text to a user device, means for recognizing the user's emotional state, and means for adjusting the content of the generated explanatory text based on the recognized emotional state. This allows the user to use an autonomous vehicle with peace of mind by providing detailed information about traffic conditions and feedback adjusted according to the user's emotional state.

[0673] An "on-board camera" is a camera device that is mounted on a vehicle and captures video data of the surrounding area in real time.

[0674] "Video data" refers to continuous image information acquired by an on-board camera, and is basic data for traffic situation analysis.

[0675] A "frame" is a unit of an individual still image that makes up video data.

[0676] "Traffic conditions" refers to objects on the road, such as vehicles, pedestrians, and traffic lights, as well as their condition and position.

[0677] The "explanation" is a sentence generated to convey the analyzed traffic situation to the user in natural language.

[0678] "User device" refers to a device owned by a user that can display and operate information, such as a smartphone or tablet.

[0679] "Emotional state" indicates the user's mental and sensory response, and is primarily recognized from voice and facial expression data.

[0680] A "server" is a computer system that processes and analyzes various types of data and is managed centrally.

[0681] A "generative AI model" is an artificial intelligence model that uses machine learning technology to automatically generate sentences corresponding to specific tasks.

[0682] "Adjusting" means changing the content or format of the product based on the perceived information, for example, modifying the generated description based on the user's emotional state.

[0683] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[0684] Video data acquisition and processing

[0685] The user starts the dashcam and starts capturing video data. The server receives the video data sent from the dashcam in real time and divides it into frames. Next, preprocessing is performed on each frame. This includes noise reduction, color correction, and resolution adjustment. Noise reduction, color correction, and resolution adjustment are achieved using libraries such as OpenCV.

[0686] Traffic situation analysis

[0687] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it uses a deep learning model (such as YOLO) to detect objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This allows for a detailed understanding of the traffic situation.

[0688] Generating Descriptions

[0689] The server uses a generative AI model (e.g., GPT-4) based on the analysis results to generate natural language descriptions, which can provide detailed descriptions of traffic situations, such as vehicle movements on the road, traffic light status, and pedestrian positions. Specific examples of prompts include the following:

[0690] Input: Number of vehicles: 4, Traffic light status: red, Pedestrian position: on crosswalk

[0691] Prompt: Generate a description based on four vehicles ahead, a red light, and pedestrians crossing the street.

[0692] Providing a description

[0693] The server provides the generated explanatory text to a user device such as a user's smartphone, tablet, etc. The user device receives the explanatory text sent from the server and displays it to the user.

[0694] Adding an Emotion Engine

[0695] This system also includes an emotion engine that recognizes the user's emotions. The emotion engine has the function of recognizing the user's emotional state by analyzing the user's voice data and facial expression data. Azure Face API is used to analyze facial expressions, and Amazon Polly is used to analyze voice data. As a result, the user's emotional state is recognized as "anxiety."

[0696] Adjusting descriptions based on emotion

[0697] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server can increase the user's sense of security by making the explanation more detailed or changing it to more positive language. Specific examples of prompts include the following:

[0698] Input: User emotion: Anxiety

[0699] Prompt: The user is feeling anxious. Please regenerate the description with more detail and positive wording.

[0700] Specific examples

[0701] For example, consider the following scenario:

[0702] 1. The user turns on the dashcam and begins capturing video data.

[0703] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0704] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0705] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0706] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[0707] 6. The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[0708] 7. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[0709] In this way, the system of the present invention can improve the user experience by providing explanations that take the user's emotions into consideration.

[0710] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0711] Step 1:

[0712] The user activates the in-vehicle camera and starts capturing video data.

[0713] Input: Start instruction for in-car camera

[0714] Output: Start of real-time video data acquisition

[0715] Specific operation: The user activates the dashcam from the smartphone app.

[0716] Step 2:

[0717] The server receives video data transmitted from the vehicle-mounted camera in real time and divides it into frames.

[0718] Input: Real-time video data

[0719] Output: List of frame data

[0720] Specific operation: Video data is divided into frames and a timestamp is stored for each frame.

[0721] Step 3:

[0722] The server performs pre-processing on each frame, which includes noise reduction, color correction, and resolution adjustment.

[0723] Input: List of frame data

[0724] Output: Preprocessed frame data

[0725] What it does: It uses the OpenCV library to perform noise reduction, color correction, and resolution adjustment on each frame.

[0726] Step 4:

[0727] The server performs traffic situation analysis for each frame after preprocessing. Specifically, it uses a deep learning model (e.g., YOLO) to detect objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object.

[0728] Input: Preprocessed frame data

[0729] Output: Information about detected objects (type, location, movement, etc.)

[0730] Specific operation: The YOLO model is applied to frame data, and the identification of objects indicating traffic conditions and their location information are output.

[0731] Step 5:

[0732] The server generates a natural language explanation based on the analysis results using a generative AI model (e.g., GPT-4).

[0733] Input: Object information (type, position, movement, etc.)

[0734] Output: Generated natural language description

[0735] Specific operation: Based on the analysis results, a prompt sentence is input into the AI ​​model to generate an explanatory sentence. Specific prompt sentence examples:

[0736] Input: Number of vehicles: 4, Traffic light status: red, Pedestrian position: on crosswalk

[0737] Prompt: Generate a description based on four vehicles ahead, a red light, and pedestrians crossing the street.

[0738] Step 6:

[0739] The server provides the generated explanation to the user device.

[0740] Input: Generated description

[0741] Output: Send to user device

[0742] Specific operation: The generated explanation is sent to the user's smartphone or tablet and displayed.

[0743] Step 7:

[0744] The server acquires the user's voice data and facial expression data and recognizes their emotional state using an emotion engine.

[0745] Input: Voice data, facial expression data

[0746] Output: Emotional state recognition result

[0747] Specific operation: The voice data and facial expression data collected by the user device are sent to the server and analyzed using an emotion engine (e.g., Azure Face API or Amazon Polly).

[0748] Step 8:

[0749] The content of the generated description is adjusted based on the emotional state recognized by the server.

[0750] Input: Emotional state recognition results, generated description

[0751] Output: Adjusted description

[0752] Specific behavior: Depending on the recognized emotional state, the generative AI model is re-prompted to generate an adjusted explanation. Specific prompt examples:

[0753] Input: User emotion: Anxiety

[0754] Prompt: The user is feeling anxious. Please regenerate the description with more detail and positive wording.

[0755] Step 9:

[0756] The server provides the adjusted explanation back to the user device.

[0757] Input: Adjusted description

[0758] Output: Retransmission to user device

[0759] Specific operation: The adjusted description is resent to the user's smartphone or tablet and displayed.

[0760] This provides detailed information about traffic conditions and tailored feedback based on the user's emotional state, allowing users to use self-driving vehicles with peace of mind.

[0761] 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.

[0762] 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.

[0763] 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.

[0764] [Third embodiment]

[0765] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0766] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0767] 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).

[0768] 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.

[0769] 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.

[0770] 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).

[0771] 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.

[0772] 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.

[0773] 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.

[0774] 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.

[0775] 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.

[0776] 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."

[0777] The present invention is a system that automatically converts in-vehicle camera images into explanatory text about traffic conditions, and aims to streamline the development of autonomous driving and advanced driver assistance systems (ADAS). This system includes the following main means:

[0778] Acquiring video data

[0779] The user starts the in-vehicle camera and starts capturing video data. The in-vehicle camera is installed in front of the vehicle and records road conditions in real time.

[0780] Receiving and analyzing video data

[0781] The server receives video data sent from the in-vehicle camera. The received video data is divided into frames and pre-processed to analyze the traffic situation within each frame. This pre-processing includes noise removal, color correction, and resolution adjustment.

[0782] Traffic situation recognition

[0783] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. For example, it recognizes whether a vehicle is about to turn left, whether a pedestrian is crossing the street, or whether a traffic light has changed from red to green.

[0784] Generating Descriptions

[0785] The server generates a natural language description based on the analysis results. Using a generative AI model, detailed text about traffic conditions is automatically created, allowing users to understand real-time traffic conditions without having to check the video.

[0786] Providing a description

[0787] The server then provides the generated description to the user device, which is an electronic device such as a smartphone or PC that receives and displays the description, allowing the user to check detailed information about traffic conditions in real time.

[0788] Specific examples

[0789] For example, in a scenario describing traffic conditions on a given day, the following sequence of actions occurs:

[0790] 1. The user turns on the dashcam and begins capturing video.

[0791] 2. The server receives the video data and analyzes it frame by frame.

[0792] 3. The server recognizes that there are four vehicles ahead, the traffic light is red, and detects that a pedestrian is beginning to cross the street.

[0793] 4. The server passes the analysis results to the generation AI, which automatically generates a description such as "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0794] 5. The server generates a description and sends it to the user's device, where the user can view this information in real time.

[0795] This allows users to understand traffic conditions in detail through text without having to look at the video directly. This system enables the efficient generation of learning data in the development of autonomous driving and ADAS, greatly streamlining the development process.

[0796] The above is a detailed description of the embodiment of the present invention.

[0797] The processing flow will be explained below.

[0798] Step 1:

[0799] The user activates the dashcam, which then starts capturing video data. Specifically, the user turns on the power of the camera device in the vehicle, and a real-time stream of video data begins.

[0800] Step 2:

[0801] The server receives video data sent from the dashcam in real time. The server establishes a connection based on the specified camera ID and continuously acquires data.

[0802] Step 3:

[0803] The server divides the received video data into frames, converting the video data into frames for each small time unit, and these frames are then used for subsequent analysis processing.

[0804] Step 4:

[0805] The server performs pre-processing on each frame, including noise reduction, color correction, and resolution adjustment, to improve the quality of the frame data.

[0806] Step 5:

[0807] The server performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This is done using object detection and motion analysis algorithms.

[0808] Step 6:

[0809] The server recognizes traffic conditions based on the results of object detection and motion analysis, including vehicle position, traffic light status, and pedestrian movement.

[0810] Step 7:

[0811] The server passes the analysis results to a generative AI model, which then generates a natural language explanation. The generative AI model converts the analyzed traffic situation data into text and automatically creates a detailed explanation.

[0812] Step 8:

[0813] The server then sends the generated explanatory text to the user device, which may be a smartphone or PC, and receives the explanatory text from the server and displays it to the user.

[0814] Step 9:

[0815] Users can view the description on their device and get a detailed understanding of the real-time traffic situation. The textual information allows users to intuitively understand the situation ahead of the vehicle and other important traffic information.

[0816] Example 1

[0817] 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."

[0818] In the development of autonomous driving and advanced driver assistance systems (ADAS), it is important to accurately and efficiently grasp traffic conditions. However, conventional methods require manual analysis of vast amounts of video data, which is time-consuming and costly. Furthermore, it is difficult to grasp traffic conditions in real time, which reduces development efficiency. This invention aims to solve these problems and significantly streamline the development process for autonomous driving and ADAS.

[0819] 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.

[0820] In this invention, the server includes means for acquiring video data from an on-board camera, means for preprocessing the acquired video data frame by frame, means for recognizing traffic conditions for each preprocessed frame, means for automatically generating explanatory text based on the recognized traffic conditions, and means for providing the generated explanatory text to a user device. This enables real-time analysis of video data and efficient development of autonomous driving and ADAS.

[0821] An "in-vehicle camera" is a camera installed in a vehicle to record road and traffic conditions.

[0822] "Video data" refers to video captured by a camera and stored in digital format.

[0823] A "frame" refers to each of the consecutive still images that make up video data.

[0824] "Preprocessing" refers to initial processing such as noise removal, color correction, and resolution adjustment that is performed to make video data easier to analyze.

[0825] "Traffic conditions" refers to information that indicates the overall state and movement of vehicles, pedestrians, traffic lights, etc. on the road.

[0826] "Recognition" is the process of detecting and understanding specific objects and their movements from video data.

[0827] A "description" is a text that describes the recognized traffic situation in natural language.

[0828] A "generative AI model" is an artificial intelligence model that automatically generates text based on input data.

[0829] A "user device" is an electronic device for displaying the generated description, and includes a smartphone, a PC, etc.

[0830] The system according to the present invention automatically converts images from an in-vehicle camera into explanatory text about traffic conditions, and aims to improve the efficiency of the development of autonomous driving and advanced driver assistance systems (ADAS). Specific embodiments of this system are described below.

[0831] First, the user activates the dashcam to begin capturing video data. The dashcam is installed in the front of the car and records road conditions in real time while the car is moving. This camera uses a standard high-resolution camera and captures video data at 30 frames per second (FPS).

[0832] The captured video data is then sent to a server in real time. The server divides the video data into frames and performs preprocessing. This preprocessing includes processes such as noise reduction using the OpenCV library, color correction, and resolution adjustment. For example, noise reduction involves applying a filter to remove unwanted background signals. Color correction involves adjusting the color balance of each frame to improve visibility.

[0833] After preprocessing, the server recognizes traffic conditions for each frame. During this process, it runs an object detection algorithm to detect objects such as vehicles, pedestrians, and traffic lights. For example, it can use the YOLO (You Only Look Once) model to detect objects quickly and accurately. It analyzes the position and movement of detected objects in each frame to understand traffic flow. Examples of identified objects include vehicles approaching an intersection, pedestrians crossing a crosswalk, and traffic lights changing from red to green.

[0834] Based on the analysis results, the server uses a generative AI model to generate a description in natural language. One example of a generative AI model that can be used is GPT-3. The server passes the analysis results to the generative AI as a prompt, which generates a description like the following: "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk." The generated description is formatted in a way that is easy for the user to understand.

[0835] Finally, the server sends the generated description in real time to the user device. The user device is an electronic device such as a smartphone or PC, which displays the description on its screen. The user can check the traffic situation in text form through the application. The display on the user device shows, "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0836] As a concrete example, here is a scenario describing traffic conditions on a given day:

[0837] 1. The user activates the dashcam using a smartphone app and begins capturing footage.

[0838] 2. The server receives the video data in real time and performs noise reduction, color correction, and resolution adjustment on each frame.

[0839] 3. The server uses the YOLO model to detect vehicles, pedestrians, and traffic lights ahead and analyze traffic conditions.

[0840] 4. The server uses GPT-3 to generate an explanation based on the analysis results, generating the sentence, "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0841] 5. The server generates a description and sends it to the smartphone, where the user can view the information in real time through the app.

[0842] This allows users to understand traffic conditions in detail through text without having to look at the video directly. This system enables efficient generation of learning data for the development of autonomous driving and ADAS, greatly streamlining the development process.

[0843] Example prompt sentence:

[0844] "Analyze the traffic situation ahead from the specified video data and generate a description including the status of vehicles, pedestrians, and traffic lights."

[0845] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0846] Step 1:

[0847] The user activates the dashcam and begins capturing video data. Specifically, the user presses the camera's activation button using a smartphone app, and video recording begins from that point. The camera is installed at the front of the car and records video at 30 frames per second (FPS). The input is the user's camera activation operation and the real-time video scene. The output is high-resolution video data obtained from the camera.

[0848] Step 2:

[0849] The server receives video data sent from the in-vehicle camera in real time. The received video data is 30 frames per second (FPS) and is received and stored on the server. The input is the video data from the in-vehicle camera. The output is the raw video data stored on the server.

[0850] Step 3:

[0851] The server pre-processes the received video data for each frame. Specifically, it uses the OpenCV library to perform noise reduction, color correction, and resolution adjustment. For example, noise reduction applies a Gaussian filter, color correction adjusts the RGB balance, and resolution adjustment resizes the data to a consistent size. The input is the raw video data (in frame format) stored on the server. The output is the pre-processed frame data.

[0852] Step 4:

[0853] The server recognizes traffic conditions for each preprocessed frame. Specifically, it uses the YOLO (You Only Look Once) model to detect objects such as vehicles, pedestrians, and traffic lights. It tracks the location and movement of each detected object to understand traffic flow and conditions. For example, it detects changes in the color of traffic lights and the direction of pedestrian movement. The input is the preprocessed frame data. The output is the analysis result of the traffic conditions for each frame.

[0854] Step 5:

[0855] The server uses a generative AI model based on the analysis results to generate a natural language description. Specifically, it utilizes a large-scale language model such as GPT-3 and provides the traffic situation analysis results obtained in the previous step to the model as a prompt. For example, it generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk." The input is the traffic situation analysis results and the prompt. The output is the generated natural language description.

[0856] Step 6:

[0857] The server sends the generated description to the user device in real time. The user device is an electronic device such as a smartphone or PC, which displays the received description on its screen. The input is the generated natural language description. The output is the description displayed on the user device. For example, the app's display might say, "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0858] (Application example 1)

[0859] 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."

[0860] Conventional autonomous vehicles and advanced driver assistance systems (ADAS) require drivers and developers to directly check video data to understand traffic conditions, which places a significant burden on the driver and developers. Additionally, it is difficult to grasp traffic conditions in real time, resulting in delayed reactions in situations where instant decisions are required. Furthermore, relying solely on vision means that information acquisition is incomplete, potentially increasing the risk of accidents and traffic violations.

[0861] 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.

[0862] In this invention, the server includes means for acquiring video data from an on-board camera, means for analyzing the acquired video data frame by frame and recognizing traffic conditions, means for automatically generating explanatory text based on the analyzed traffic conditions, means for providing the generated explanatory text to a user device, and means for displaying the generated explanatory text in real time on smart glasses, which are a type of user device. This enables drivers and developers of autonomous driving systems to grasp detailed and accurate traffic conditions in real time without directly viewing the video.

[0863] An "on-board camera" is a camera device mounted on a vehicle that captures road and surrounding conditions as video data in real time.

[0864] "Video data" refers to continuous image information captured by an on-board camera, and visually records traffic conditions and the surrounding environment.

[0865] A "frame" refers to a series of individual images of video data that are treated as a single still image.

[0866] "Traffic conditions" refers to the state of various elements on the road, such as vehicles, pedestrians, and traffic lights, and their interrelationships.

[0867] A "user device" is an electronic device that can receive explanatory text sent from a server and display it to a user.

[0868] "Analysis" is the process of processing video data to extract and understand specific information.

[0869] "Explanation" refers to text information written in natural language based on traffic situation information obtained through analysis.

[0870] "Real time" refers to processing and display occurring immediately without delay.

[0871] "Smart glasses" are glasses-type electronic devices that have the function of displaying images and text information, allowing the wearer to view information through the device.

[0872] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate text from given data.

[0873] This invention is a system that automatically converts in-vehicle camera footage into traffic situation descriptions and provides them to user devices (especially smart glasses) in real time. The purpose of this system is to streamline the development of autonomous vehicles and advanced driver assistance systems (ADAS) and to provide drivers with instant, detailed traffic information.

[0874] The server has a means to acquire video data from dashcams, which are attached to vehicles and record road conditions in real time. This video data is processed frame by frame and pre-processed to recognize traffic conditions. Pre-processing includes noise reduction, color correction, resolution adjustment, etc.

[0875] Next, the server uses a method to analyze traffic conditions for each frame of pre-processed video data. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. For example, it recognizes whether a vehicle is about to turn left, whether a pedestrian is crossing the street, or whether a traffic light has changed from red to green.

[0876] Based on the results of traffic situation analysis, the server has a means to generate a natural language description using a generative AI model. The generated description is detailed and specific, allowing users to understand the traffic situation in real time without having to check the video directly.

[0877] The generated explanatory text is sent from the server to the user device. A particularly distinctive feature of this invention is that it provides a means for displaying the explanatory text in real time on smart glasses. The smart glasses are a device that allows drivers to instantly check traffic conditions while ensuring safety while driving.

[0878] For example, GPS information can be added to provide information about the actual location. In this case, an example of a prompt sentence would be "Currently, there are four vehicles ahead at the specified location, the traffic light is red, and a pedestrian is starting to cross the street."

[0879] This system provides drivers and autonomous driving developers with detailed and accurate real-time traffic conditions, supporting driving decisions and streamlining development.

[0880] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0881] Step 1:

[0882] The onboard camera is activated to begin capturing video data. The onboard camera is installed in front of the vehicle and records road conditions. The input is visual information on the road, and the output is video data.

[0883] Step 2:

[0884] The server receives video data sent from the in-vehicle camera in real time. The received video data is divided into frames. The input is the video data from the in-vehicle camera, and the output is the divided frame data.

[0885] Step 3:

[0886] The server pre-processes the video data, specifically noise removal, color correction, resolution adjustment, etc. The input is the divided frame data, and the output is the pre-processed frame data.

[0887] Step 4:

[0888] The server analyzes the traffic situation for each pre-processed frame. It detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. The input is the pre-processed frame data, and the output is the analysis results (traffic situation data).

[0889] Step 5:

[0890] The server generates a natural language explanation using a generative AI model based on the analysis results. The input is traffic situation data, and the output is the generated explanation.

[0891] Step 6:

[0892] The server provides the generated explanatory text to the user device (smart glasses). The input is the generated explanatory text, and the output is the text information displayed on the smart glasses.

[0893] Step 7:

[0894] Users can check traffic conditions in real time through smart glasses, allowing them to understand detailed traffic conditions without having to view the video directly. The input is the explanatory text displayed on the smart glasses, and the output is the user's perception of the traffic conditions.

[0895] 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.

[0896] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[0897] Video data acquisition and processing

[0898] The user turns on the dashcam and starts capturing video data. The server receives the video data sent from the dashcam in real time and divides it into frames. Then, it performs pre-processing on each frame, which includes noise reduction, color correction, and resolution adjustment.

[0899] Traffic situation analysis

[0900] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This allows for a detailed understanding of the traffic situation.

[0901] Generating Descriptions

[0902] The server uses a generative AI model to generate natural language descriptions based on the analysis results, which can provide detailed descriptions of traffic conditions, such as vehicle movements on the road, traffic light status, and pedestrian locations.

[0903] Providing a description

[0904] The server provides the generated explanatory text to the user's device (e.g., a smartphone or PC). The user device receives the explanatory text sent from the server and displays it to the user.

[0905] Adding an Emotion Engine

[0906] The system further includes an emotion engine that recognizes the user's emotion by analyzing the user's voice data and facial expression data.

[0907] Adjusting descriptions based on emotion

[0908] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server can increase the user's sense of security by providing more detailed explanations or changing them to more positive expressions.

[0909] Specific examples

[0910] For example, consider the following scenario:

[0911] 1. The user turns on the dashcam and begins capturing video data.

[0912] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0913] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0914] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0915] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[0916] 6. The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[0917] 7. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[0918] In this way, the system of the present invention can improve the user experience by providing explanations that take the user's emotions into consideration.

[0919] The processing flow will be explained below.

[0920] Step 1:

[0921] The user activates the dashcam, which then starts capturing video data in real time.

[0922] Step 2:

[0923] The server receives video data transmitted from the dashcam in real time, establishes a connection based on the specified camera ID, and starts receiving the data stream.

[0924] Step 3:

[0925] The server continuously divides the received video data into frames, so that the video data is treated as individual frames in small, easily processable time units.

[0926] Step 4:

[0927] The server performs pre-processing on each frame, including noise reduction, color correction, and resolution adjustment, to improve the quality of the frame data and increase the accuracy of subsequent analysis.

[0928] Step 5:

[0929] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their positions and movements. Object detection and motion analysis algorithms are used.

[0930] Step 6:

[0931] The server recognizes traffic conditions based on the results of object detection and motion analysis, and the information it obtains makes it possible to grasp detailed information such as the relative positions of vehicles, the status of traffic lights, and the movements of pedestrians.

[0932] Step 7:

[0933] The server passes the analysis results to a generative AI model, which then generates a natural language explanation. The generative AI model then automatically creates an appropriate explanation based on the traffic condition data.

[0934] Step 8:

[0935] The server transmits the generated explanatory text to the user device, which is an electronic device such as a smartphone or PC, and receives and displays the explanatory text to the user.

[0936] Step 9:

[0937] The emotion engine analyzes the user's voice data and facial expression data, identifies emotions from the user's voice and facial expressions, and recognizes the user's emotional state.

[0938] Step 10:

[0939] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server will make the explanation more detailed or change it to more positive language to increase the user's sense of security.

[0940] Specific examples

[0941] Example 1: Normal traffic conditions

[0942] 1. The user turns on the dashcam and begins capturing video data.

[0943] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0944] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0945] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0946] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[0947] Example 2: When the user feels anxious

[0948] 1. The user turns on the dashcam and begins capturing video data.

[0949] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[0950] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[0951] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[0952] 5. The emotion engine analyzes the user's voice and facial expressions and recognizes that they are feeling anxious.

[0953] 6. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[0954] 7. The server sends the adjusted explanation to the user's device, and the user checks the explanation and feels reassured.

[0955] Through these steps, users can understand traffic conditions in real time and receive information tailored to their emotions.

[0956] Example 2

[0957] 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."

[0958] Current in-vehicle camera systems have difficulty recognizing traffic conditions and automatically generating explanatory text. Furthermore, they are unable to provide personalized information that takes into account the user's emotions, which means that safety and the user experience are not sufficiently improved. Therefore, there is a need for the development of a system that can analyze traffic conditions in real time and generate and provide explanatory text that reflects the user's emotional state.

[0959] 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.

[0960] In this invention, the server includes: means for acquiring video data from an in-vehicle camera; preprocessing means for dividing the acquired video data into frames and performing noise removal, color correction, and resolution adjustment; means for analyzing traffic conditions for each preprocessed frame, detecting objects such as vehicles, pedestrians, and traffic lights, and analyzing their positions and movements; means for generating a natural language description using a generative AI model based on the analyzed traffic conditions; means for providing the generated description to a user device; an emotion engine for analyzing a user's voice data and facial expression data to recognize the user's emotional state; and means for adjusting the content of the generated description based on the recognized user's emotional state. This makes it possible to analyze traffic conditions in detail in real time and generate and provide appropriate description that takes the user's emotional state into consideration.

[0961] An "on-board camera" is a device that is mounted on a vehicle and captures images of the road and surrounding area.

[0962] "Video data" refers to continuous image information acquired by an on-board camera.

[0963] A "frame" is an individual still image that makes up video data.

[0964] "Noise reduction" is a process that removes unnecessary signals and disturbances from video data.

[0965] "Color correction" is a process that adjusts the color balance of video data to bring it closer to the original color tone.

[0966] "Resolution adjustment" is a process of changing the number of pixels in the video data to maintain an appropriate resolution.

[0967] "Preprocessing" refers to a series of initial data processing steps, including noise removal, color correction, and resolution adjustment.

[0968] "Traffic conditions" refers to information about the status and movement of vehicles, pedestrians, and traffic lights on the road.

[0969] A "generative AI model" is an artificial intelligence model that generates explanatory text in natural language from analysis results.

[0970] A "description" is a natural language text generated based on the analyzed traffic situation.

[0971] A "user device" is a computing device used by a user, such as a smartphone or PC.

[0972] "Voice data" refers to digital audio information that records the user's speech or voice.

[0973] "Facial expression data" is digital image information that records the facial expressions of a user.

[0974] The "emotion engine" is a system that analyzes voice data and facial expression data to recognize the user's emotional state.

[0975] An "emotional state" is the psychological state that the user is currently feeling (for example, anxiety or relief).

[0976] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[0977] Video data acquisition and preprocessing

[0978] The user activates the in-vehicle camera to capture video data of road conditions. The in-vehicle camera is attached to the vehicle and captures video in real time. The video data is then sent to the server.

[0979] The server receives video data sent from the vehicle-mounted camera in real time. The received video data is divided into frames and then pre-processed. This pre-processing includes noise reduction, color correction, and resolution adjustment. This improves the quality of the video and enables accurate analysis.

[0980] Traffic situation analysis

[0981] The server analyzes the traffic situation for each pre-processed frame. Specifically, it uses deep learning models to detect objects such as vehicles, pedestrians, and traffic lights in the frame and analyzes their positions and movements, providing a detailed understanding of the current traffic situation.

[0982] Generating Descriptions

[0983] The server uses a generative AI model based on the analysis results to convert the traffic situation into a natural language description. For example, if the analysis result is "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk," the server inputs this analysis result as a prompt into the generative AI model to generate a description. A specific description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk" is output.

[0984] Providing a description

[0985] The generated description is sent from the server to the user's device (e.g., a smartphone or PC). The user's device receives the description and displays it on the screen to inform the user of the current traffic situation.

[0986] Emotion engine integration

[0987] Furthermore, the system includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes the user's voice data and facial expression data to recognize the user's emotional state. This allows the system to respond in accordance with the user's emotions.

[0988] Adjusted description

[0989] The server adjusts the generated explanation based on the user's emotional state as recognized by the emotion engine. For example, if the user is feeling anxious, the explanation can be made more detailed to increase the sense of security. For example, it can be adjusted to say, "There are four vehicles ahead, the traffic light is red, and there are pedestrians at the crosswalk, so please proceed safely and with caution."

[0990] Specific examples

[0991] The user activates the in-vehicle camera and starts capturing video data.

[0992] The server receives the video data, performs preprocessing, and then analyzes the traffic situation.

[0993] The analysis results are input into a generative AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[0994] The server sends the explanation to the user's device, which the user confirms.

[0995] The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[0996] The server adjusts the description to read, "Proceed safely. The light is red, please stop."

[0997] This allows users to accurately grasp the current traffic situation and receive detailed information tailored to their emotions, significantly improving user safety and experience.

[0998] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0999] Step 1:

[1000] The user activates the in-vehicle camera, which captures road conditions in real time and acquires video data.

[1001] Input: Real-time footage from a dashcam.

[1002] Specific operation: The user turns on the dashcam and launches the video capture application.

[1003] Step 2:

[1004] The server receives video data sent from the vehicle-mounted camera in real time and divides the received video data into frames.

[1005] Input: Real-time video data from a dashcam.

[1006] Output: Split frame data.

[1007] Specific operation: The server executes a script that divides the video data into frames and prepares each frame for processing.

[1008] Step 3:

[1009] The server performs pre-processing on the divided frames, which includes noise reduction, color correction, and resolution adjustment.

[1010] Input: Split frame data.

[1011] Output: Preprocessed frame data.

[1012] What happens: The server uses image processing algorithms to remove noise, perform color correction, and adjust the resolution of each frame.

[1013] Step 4:

[1014] The server analyzes the preprocessed frames and identifies traffic situations. Specifically, it uses deep learning models to detect objects such as vehicles, pedestrians, and traffic lights. It analyzes the position and movement of each object.

[1015] Input: Preprocessed frame data.

[1016] Output: Traffic condition analysis result data.

[1017] How it works: The server applies deep learning models to identify the location and movement of vehicles, pedestrians, and traffic lights in each frame.

[1018] Step 5:

[1019] The server uses a generative AI model based on the analysis results to generate a natural language description.

[1020] Input: Traffic analysis result data.

[1021] Output: The generated natural language description.

[1022] Specific operation: The server inputs the analysis results as a prompt sentence to the generative AI model and obtains the generated explanation sentence.

[1023] Example prompt: "There are four vehicles ahead, the light is red, and a pedestrian is crossing the street."

[1024] Step 6:

[1025] The server sends the generated explanation to the user's device, which receives the explanation and displays it to the user.

[1026] Input: The generated natural language description.

[1027] Output: Description that will be displayed on the user's device.

[1028] Specific operation: The server sends the explanatory text to the user's smartphone or PC, and the user device receives it and displays it on the screen.

[1029] Step 7:

[1030] The server uses an emotion engine to analyze the user's voice data and facial expression data to recognize emotions.

[1031] Input: User's voice and facial expression data.

[1032] Output: Perceived emotional state.

[1033] Specific operation: The user provides voice data and facial expression data using a microphone or camera, and the server analyzes it using an emotion engine.

[1034] Step 8:

[1035] The server adjusts the content of the generated narrative based on the recognized emotional state of the user.

[1036] Input: A recognized emotional state and a generated natural language description.

[1037] Output: A narrative tailored based on sentiment.

[1038] Specific operation: Based on the results of emotion analysis, the server adjusts the output of the generative AI model and changes the description to take the user's emotions into consideration.

[1039] This allows the user to accurately grasp the current traffic situation and receive detailed information tailored to their emotions.

[1040] (Application example 2)

[1041] 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."

[1042] In conventional autonomous vehicle systems, there are methods for analyzing in-vehicle camera footage to inform users of traffic conditions, but the feedback provided does not take into account the user's emotional state, which can lead to anxiety or a lack of understanding of the situation. Furthermore, the explanations of traffic conditions are uniform, making it difficult to provide information tailored to the needs of individual users.

[1043] The identification processing 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 acquiring video data from an in-vehicle camera, means for analyzing the acquired video data frame by frame and recognizing traffic conditions, means for automatically generating an explanatory text based on the analyzed traffic conditions, means for providing the generated explanatory text to a user device, means for recognizing the user's emotional state, and means for adjusting the content of the generated explanatory text based on the recognized emotional state. This allows the user to use an autonomous vehicle with peace of mind by providing detailed information about traffic conditions and feedback adjusted according to the user's emotional state.

[1044] An "on-board camera" is a camera device that is mounted on a vehicle and captures video data of the surrounding area in real time.

[1045] "Video data" refers to continuous image information acquired by an on-board camera, and is basic data for traffic situation analysis.

[1046] A "frame" is a unit of an individual still image that makes up video data.

[1047] "Traffic conditions" refers to objects on the road, such as vehicles, pedestrians, and traffic lights, as well as their condition and position.

[1048] The "explanation" is a sentence generated to convey the analyzed traffic situation to the user in natural language.

[1049] "User device" refers to a device owned by a user that can display and operate information, such as a smartphone or tablet.

[1050] "Emotional state" indicates the user's mental and sensory response, and is primarily recognized from voice and facial expression data.

[1051] A "server" is a computer system that processes and analyzes various types of data and is managed centrally.

[1052] A "generative AI model" is an artificial intelligence model that uses machine learning technology to automatically generate sentences corresponding to specific tasks.

[1053] "Adjusting" means changing the content or format of the product based on the perceived information, for example, modifying the generated description based on the user's emotional state.

[1054] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[1055] Video data acquisition and processing

[1056] The user starts the dashcam and starts capturing video data. The server receives the video data sent from the dashcam in real time and divides it into frames. Next, preprocessing is performed on each frame. This includes noise reduction, color correction, and resolution adjustment. Noise reduction, color correction, and resolution adjustment are achieved using libraries such as OpenCV.

[1057] Traffic situation analysis

[1058] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it uses a deep learning model (such as YOLO) to detect objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This allows for a detailed understanding of the traffic situation.

[1059] Generating Descriptions

[1060] The server uses a generative AI model (e.g., GPT-4) based on the analysis results to generate natural language descriptions, which can provide detailed descriptions of traffic situations, such as vehicle movements on the road, traffic light status, and pedestrian positions. Specific examples of prompts include the following:

[1061] Input: Number of vehicles: 4, Traffic light status: red, Pedestrian position: on crosswalk

[1062] Prompt: Generate a description based on four vehicles ahead, a red light, and pedestrians crossing the street.

[1063] Providing a description

[1064] The server provides the generated explanatory text to a user device such as a user's smartphone, tablet, etc. The user device receives the explanatory text sent from the server and displays it to the user.

[1065] Adding an Emotion Engine

[1066] This system also includes an emotion engine that recognizes the user's emotions. The emotion engine has the function of recognizing the user's emotional state by analyzing the user's voice data and facial expression data. Azure Face API is used to analyze facial expressions, and Amazon Polly is used to analyze voice data. As a result, the user's emotional state is recognized as "anxiety."

[1067] Adjusting descriptions based on emotion

[1068] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server can increase the user's sense of security by making the explanation more detailed or changing it to more positive language. Specific examples of prompts include the following:

[1069] Input: User emotion: Anxiety

[1070] Prompt: The user is feeling anxious. Please regenerate the description with more detail and positive wording.

[1071] Specific examples

[1072] For example, consider the following scenario:

[1073] 1. The user turns on the dashcam and begins capturing video data.

[1074] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[1075] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[1076] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[1077] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[1078] 6. The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[1079] 7. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[1080] In this way, the system of the present invention can improve the user experience by providing explanations that take the user's emotions into consideration.

[1081] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1082] Step 1:

[1083] The user activates the in-vehicle camera and starts capturing video data.

[1084] Input: Start instruction for in-car camera

[1085] Output: Start of real-time video data acquisition

[1086] Specific operation: The user activates the dashcam from the smartphone app.

[1087] Step 2:

[1088] The server receives video data transmitted from the vehicle-mounted camera in real time and divides it into frames.

[1089] Input: Real-time video data

[1090] Output: List of frame data

[1091] Specific operation: Video data is divided into frames and a timestamp is stored for each frame.

[1092] Step 3:

[1093] The server performs pre-processing on each frame, which includes noise reduction, color correction, and resolution adjustment.

[1094] Input: List of frame data

[1095] Output: Preprocessed frame data

[1096] What it does: It uses the OpenCV library to perform noise reduction, color correction, and resolution adjustment on each frame.

[1097] Step 4:

[1098] The server performs traffic situation analysis for each frame after preprocessing. Specifically, it uses a deep learning model (e.g., YOLO) to detect objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object.

[1099] Input: Preprocessed frame data

[1100] Output: Information about detected objects (type, location, movement, etc.)

[1101] Specific operation: The YOLO model is applied to frame data, and the identification of objects indicating traffic conditions and their location information are output.

[1102] Step 5:

[1103] The server generates a natural language explanation based on the analysis results using a generative AI model (e.g., GPT-4).

[1104] Input: Object information (type, position, movement, etc.)

[1105] Output: Generated natural language description

[1106] Specific operation: Based on the analysis results, a prompt sentence is input into the AI ​​model to generate an explanatory sentence. Specific prompt sentence examples:

[1107] Input: Number of vehicles: 4, Traffic light status: red, Pedestrian position: on crosswalk

[1108] Prompt: Generate a description based on four vehicles ahead, a red light, and pedestrians crossing the street.

[1109] Step 6:

[1110] The server provides the generated explanation to the user device.

[1111] Input: Generated description

[1112] Output: Send to user device

[1113] Specific operation: The generated explanation is sent to the user's smartphone or tablet and displayed.

[1114] Step 7:

[1115] The server acquires the user's voice data and facial expression data and recognizes their emotional state using an emotion engine.

[1116] Input: Voice data, facial expression data

[1117] Output: Emotional state recognition result

[1118] Specific operation: The voice data and facial expression data collected by the user device are sent to the server and analyzed using an emotion engine (e.g., Azure Face API or Amazon Polly).

[1119] Step 8:

[1120] The content of the generated description is adjusted based on the emotional state recognized by the server.

[1121] Input: Emotional state recognition results, generated description

[1122] Output: Adjusted description

[1123] Specific behavior: Depending on the recognized emotional state, the generative AI model is re-prompted to generate an adjusted explanation. Specific prompt examples:

[1124] Input: User emotion: Anxiety

[1125] Prompt: The user is feeling anxious. Please regenerate the description with more detail and positive wording.

[1126] Step 9:

[1127] The server provides the adjusted explanation back to the user device.

[1128] Input: Adjusted description

[1129] Output: Retransmission to user device

[1130] Specific operation: The adjusted description is resent to the user's smartphone or tablet and displayed.

[1131] This provides detailed information about traffic conditions and tailored feedback based on the user's emotional state, allowing users to use self-driving vehicles with peace of mind.

[1132] 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.

[1133] 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.

[1134] 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.

[1135] [Fourth embodiment]

[1136] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1137] 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.

[1138] 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).

[1139] 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.

[1140] 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.

[1141] 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).

[1142] 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.

[1143] 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.

[1144] 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.

[1145] 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.

[1146] 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.

[1147] 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.

[1148] 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."

[1149] The present invention is a system that automatically converts in-vehicle camera images into explanatory text about traffic conditions, and aims to streamline the development of autonomous driving and advanced driver assistance systems (ADAS). This system includes the following main means:

[1150] Acquiring video data

[1151] The user starts the in-vehicle camera and starts capturing video data. The in-vehicle camera is installed in front of the vehicle and records road conditions in real time.

[1152] Receiving and analyzing video data

[1153] The server receives video data sent from the in-vehicle camera. The received video data is divided into frames and pre-processed to analyze the traffic situation within each frame. This pre-processing includes noise removal, color correction, and resolution adjustment.

[1154] Traffic situation recognition

[1155] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. For example, it recognizes whether a vehicle is about to turn left, whether a pedestrian is crossing the street, or whether a traffic light has changed from red to green.

[1156] Generating Descriptions

[1157] The server generates a natural language description based on the analysis results. Using a generative AI model, detailed text about traffic conditions is automatically created, allowing users to understand real-time traffic conditions without having to check the video.

[1158] Providing a description

[1159] The server then provides the generated description to the user device, which is an electronic device such as a smartphone or PC that receives and displays the description, allowing the user to check detailed information about traffic conditions in real time.

[1160] Specific examples

[1161] For example, in a scenario describing traffic conditions on a given day, the following sequence of actions occurs:

[1162] 1. The user turns on the dashcam and begins capturing video.

[1163] 2. The server receives the video data and analyzes it frame by frame.

[1164] 3. The server recognizes that there are four vehicles ahead, the traffic light is red, and detects that a pedestrian is beginning to cross the street.

[1165] 4. The server passes the analysis results to the generation AI, which automatically generates a description such as "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[1166] 5. The server generates a description and sends it to the user's device, where the user can view this information in real time.

[1167] This allows users to understand traffic conditions in detail through text without having to look at the video directly. This system enables the efficient generation of learning data in the development of autonomous driving and ADAS, greatly streamlining the development process.

[1168] The above is a detailed description of the embodiment of the present invention.

[1169] The processing flow will be explained below.

[1170] Step 1:

[1171] The user activates the dashcam, which then starts capturing video data. Specifically, the user turns on the power of the camera device in the vehicle, and a real-time stream of video data begins.

[1172] Step 2:

[1173] The server receives video data sent from the dashcam in real time. The server establishes a connection based on the specified camera ID and continuously acquires data.

[1174] Step 3:

[1175] The server divides the received video data into frames, converting the video data into frames for each small time unit, and these frames are then used for subsequent analysis processing.

[1176] Step 4:

[1177] The server performs pre-processing on each frame, including noise reduction, color correction, and resolution adjustment, to improve the quality of the frame data.

[1178] Step 5:

[1179] The server performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This is done using object detection and motion analysis algorithms.

[1180] Step 6:

[1181] The server recognizes traffic conditions based on the results of object detection and motion analysis, including vehicle position, traffic light status, and pedestrian movement.

[1182] Step 7:

[1183] The server passes the analysis results to a generative AI model, which then generates a natural language explanation. The generative AI model converts the analyzed traffic situation data into text and automatically creates a detailed explanation.

[1184] Step 8:

[1185] The server then sends the generated explanatory text to the user device, which may be a smartphone or PC, and receives the explanatory text from the server and displays it to the user.

[1186] Step 9:

[1187] Users can view the description on their device and get a detailed understanding of the real-time traffic situation. The textual information allows users to intuitively understand the situation ahead of the vehicle and other important traffic information.

[1188] Example 1

[1189] 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."

[1190] In the development of autonomous driving and advanced driver assistance systems (ADAS), it is important to accurately and efficiently grasp traffic conditions. However, conventional methods require manual analysis of vast amounts of video data, which is time-consuming and costly. Furthermore, it is difficult to grasp traffic conditions in real time, which reduces development efficiency. This invention aims to solve these problems and significantly streamline the development process for autonomous driving and ADAS.

[1191] 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.

[1192] In this invention, the server includes means for acquiring video data from an on-board camera, means for preprocessing the acquired video data frame by frame, means for recognizing traffic conditions for each preprocessed frame, means for automatically generating explanatory text based on the recognized traffic conditions, and means for providing the generated explanatory text to a user device. This enables real-time analysis of video data and efficient development of autonomous driving and ADAS.

[1193] An "in-vehicle camera" is a camera installed in a vehicle to record road and traffic conditions.

[1194] "Video data" refers to video captured by a camera and stored in digital format.

[1195] A "frame" refers to each of the consecutive still images that make up video data.

[1196] "Preprocessing" refers to initial processing such as noise removal, color correction, and resolution adjustment that is performed to make video data easier to analyze.

[1197] "Traffic conditions" refers to information that indicates the overall state and movement of vehicles, pedestrians, traffic lights, etc. on the road.

[1198] "Recognition" is the process of detecting and understanding specific objects and their movements from video data.

[1199] A "description" is a text that describes the recognized traffic situation in natural language.

[1200] A "generative AI model" is an artificial intelligence model that automatically generates text based on input data.

[1201] A "user device" is an electronic device for displaying the generated description, and includes a smartphone, a PC, etc.

[1202] The system according to the present invention automatically converts images from an in-vehicle camera into explanatory text about traffic conditions, and aims to improve the efficiency of the development of autonomous driving and advanced driver assistance systems (ADAS). Specific embodiments of this system are described below.

[1203] First, the user activates the dashcam to begin capturing video data. The dashcam is installed in the front of the car and records road conditions in real time while the car is moving. This camera uses a standard high-resolution camera and captures video data at 30 frames per second (FPS).

[1204] The captured video data is then sent to a server in real time. The server divides the video data into frames and performs preprocessing. This preprocessing includes processes such as noise reduction using the OpenCV library, color correction, and resolution adjustment. For example, noise reduction involves applying a filter to remove unwanted background signals. Color correction involves adjusting the color balance of each frame to improve visibility.

[1205] After preprocessing, the server recognizes traffic conditions for each frame. During this process, it runs an object detection algorithm to detect objects such as vehicles, pedestrians, and traffic lights. For example, it can use the YOLO (You Only Look Once) model to detect objects quickly and accurately. It analyzes the position and movement of detected objects in each frame to understand traffic flow. Examples of identified objects include vehicles approaching an intersection, pedestrians crossing a crosswalk, and traffic lights changing from red to green.

[1206] Based on the analysis results, the server uses a generative AI model to generate a description in natural language. One example of a generative AI model that can be used is GPT-3. The server passes the analysis results to the generative AI as a prompt, which generates a description like the following: "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk." The generated description is formatted in a way that is easy for the user to understand.

[1207] Finally, the server sends the generated description in real time to the user device. The user device is an electronic device such as a smartphone or PC, which displays the description on its screen. The user can check the traffic situation in text form through the application. The display on the user device shows, "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[1208] As a concrete example, here is a scenario describing traffic conditions on a given day:

[1209] 1. The user activates the dashcam using a smartphone app and begins capturing footage.

[1210] 2. The server receives the video data in real time and performs noise reduction, color correction, and resolution adjustment on each frame.

[1211] 3. The server uses the YOLO model to detect vehicles, pedestrians, and traffic lights ahead and analyze traffic conditions.

[1212] 4. The server uses GPT-3 to generate an explanation based on the analysis results, generating the sentence, "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[1213] 5. The server generates a description and sends it to the smartphone, where the user can view the information in real time through the app.

[1214] This allows users to understand traffic conditions in detail through text without having to look at the video directly. This system enables efficient generation of learning data for the development of autonomous driving and ADAS, greatly streamlining the development process.

[1215] Example prompt sentence:

[1216] "Analyze the traffic situation ahead from the specified video data and generate a description including the status of vehicles, pedestrians, and traffic lights."

[1217] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1218] Step 1:

[1219] The user activates the dashcam and begins capturing video data. Specifically, the user presses the camera's activation button using a smartphone app, and video recording begins from that point. The camera is installed at the front of the car and records video at 30 frames per second (FPS). The input is the user's camera activation operation and the real-time video scene. The output is high-resolution video data obtained from the camera.

[1220] Step 2:

[1221] The server receives video data sent from the in-vehicle camera in real time. The received video data is 30 frames per second (FPS) and is received and stored on the server. The input is the video data from the in-vehicle camera. The output is the raw video data stored on the server.

[1222] Step 3:

[1223] The server pre-processes the received video data for each frame. Specifically, it uses the OpenCV library to perform noise reduction, color correction, and resolution adjustment. For example, noise reduction applies a Gaussian filter, color correction adjusts the RGB balance, and resolution adjustment resizes the data to a consistent size. The input is the raw video data (in frame format) stored on the server. The output is the pre-processed frame data.

[1224] Step 4:

[1225] The server recognizes traffic conditions for each preprocessed frame. Specifically, it uses the YOLO (You Only Look Once) model to detect objects such as vehicles, pedestrians, and traffic lights. It tracks the location and movement of each detected object to understand traffic flow and conditions. For example, it detects changes in the color of traffic lights and the direction of pedestrian movement. The input is the preprocessed frame data. The output is the analysis result of the traffic conditions for each frame.

[1226] Step 5:

[1227] The server uses a generative AI model based on the analysis results to generate a natural language description. Specifically, it utilizes a large-scale language model such as GPT-3 and provides the traffic situation analysis results obtained in the previous step to the model as a prompt. For example, it generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk." The input is the traffic situation analysis results and the prompt. The output is the generated natural language description.

[1228] Step 6:

[1229] The server sends the generated description to the user device in real time. The user device is an electronic device such as a smartphone or PC, which displays the received description on its screen. The input is the generated natural language description. The output is the description displayed on the user device. For example, the app's display might say, "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[1230] (Application example 1)

[1231] 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."

[1232] Conventional autonomous vehicles and advanced driver assistance systems (ADAS) require drivers and developers to directly check video data to understand traffic conditions, which places a significant burden on the driver and developers. Additionally, it is difficult to grasp traffic conditions in real time, resulting in delayed reactions in situations where instant decisions are required. Furthermore, relying solely on vision means that information acquisition is incomplete, potentially increasing the risk of accidents and traffic violations.

[1233] 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.

[1234] In this invention, the server includes means for acquiring video data from an on-board camera, means for analyzing the acquired video data frame by frame and recognizing traffic conditions, means for automatically generating explanatory text based on the analyzed traffic conditions, means for providing the generated explanatory text to a user device, and means for displaying the generated explanatory text in real time on smart glasses, which are a type of user device. This enables drivers and developers of autonomous driving systems to grasp detailed and accurate traffic conditions in real time without directly viewing the video.

[1235] An "on-board camera" is a camera device mounted on a vehicle that captures road and surrounding conditions as video data in real time.

[1236] "Video data" refers to continuous image information captured by an on-board camera, and visually records traffic conditions and the surrounding environment.

[1237] A "frame" refers to a series of individual images of video data that are treated as a single still image.

[1238] "Traffic conditions" refers to the state of various elements on the road, such as vehicles, pedestrians, and traffic lights, and their interrelationships.

[1239] A "user device" is an electronic device that can receive explanatory text sent from a server and display it to a user.

[1240] "Analysis" is the process of processing video data to extract and understand specific information.

[1241] "Explanation" refers to text information written in natural language based on traffic situation information obtained through analysis.

[1242] "Real time" refers to processing and display occurring immediately without delay.

[1243] "Smart glasses" are glasses-type electronic devices that have the function of displaying images and text information, allowing the wearer to view information through the device.

[1244] A "generative AI model" is an algorithm or system that uses artificial intelligence to generate text from given data.

[1245] This invention is a system that automatically converts in-vehicle camera footage into traffic situation descriptions and provides them to user devices (especially smart glasses) in real time. The purpose of this system is to streamline the development of autonomous vehicles and advanced driver assistance systems (ADAS) and to provide drivers with instant, detailed traffic information.

[1246] The server has a means to acquire video data from dashcams, which are attached to vehicles and record road conditions in real time. This video data is processed frame by frame and pre-processed to recognize traffic conditions. Pre-processing includes noise reduction, color correction, resolution adjustment, etc.

[1247] Next, the server uses a method to analyze traffic conditions for each frame of pre-processed video data. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. For example, it recognizes whether a vehicle is about to turn left, whether a pedestrian is crossing the street, or whether a traffic light has changed from red to green.

[1248] Based on the results of traffic situation analysis, the server has a means to generate a natural language description using a generative AI model. The generated description is detailed and specific, allowing users to understand the traffic situation in real time without having to check the video directly.

[1249] The generated explanatory text is sent from the server to the user device. A particularly distinctive feature of this invention is that it provides a means for displaying the explanatory text in real time on smart glasses. The smart glasses are a device that allows drivers to instantly check traffic conditions while ensuring safety while driving.

[1250] For example, GPS information can be added to provide information about the actual location. In this case, an example of a prompt sentence would be "Currently, there are four vehicles ahead at the specified location, the traffic light is red, and a pedestrian is starting to cross the street."

[1251] This system provides drivers and autonomous driving developers with detailed and accurate real-time traffic conditions, supporting driving decisions and streamlining development.

[1252] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1253] Step 1:

[1254] The onboard camera is activated to begin capturing video data. The onboard camera is installed in front of the vehicle and records road conditions. The input is visual information on the road, and the output is video data.

[1255] Step 2:

[1256] The server receives video data sent from the in-vehicle camera in real time. The received video data is divided into frames. The input is the video data from the in-vehicle camera, and the output is the divided frame data.

[1257] Step 3:

[1258] The server pre-processes the video data, specifically noise removal, color correction, resolution adjustment, etc. The input is the divided frame data, and the output is the pre-processed frame data.

[1259] Step 4:

[1260] The server analyzes the traffic situation for each pre-processed frame. It detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their relative positions and movements. The input is the pre-processed frame data, and the output is the analysis results (traffic situation data).

[1261] Step 5:

[1262] The server generates a natural language explanation using a generative AI model based on the analysis results. The input is traffic situation data, and the output is the generated explanation.

[1263] Step 6:

[1264] The server provides the generated explanatory text to the user device (smart glasses). The input is the generated explanatory text, and the output is the text information displayed on the smart glasses.

[1265] Step 7:

[1266] Users can check traffic conditions in real time through smart glasses, allowing them to understand detailed traffic conditions without having to view the video directly. The input is the explanatory text displayed on the smart glasses, and the output is the user's perception of the traffic conditions.

[1267] 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.

[1268] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[1269] Video data acquisition and processing

[1270] The user turns on the dashcam and starts capturing video data. The server receives the video data sent from the dashcam in real time and divides it into frames. Then, it performs pre-processing on each frame, which includes noise reduction, color correction, and resolution adjustment.

[1271] Traffic situation analysis

[1272] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This allows for a detailed understanding of the traffic situation.

[1273] Generating Descriptions

[1274] The server uses a generative AI model to generate natural language descriptions based on the analysis results, which can provide detailed descriptions of traffic conditions, such as vehicle movements on the road, traffic light status, and pedestrian locations.

[1275] Providing a description

[1276] The server provides the generated explanatory text to the user's device (e.g., a smartphone or PC). The user device receives the explanatory text sent from the server and displays it to the user.

[1277] Adding an Emotion Engine

[1278] The system further includes an emotion engine that recognizes the user's emotion by analyzing the user's voice data and facial expression data.

[1279] Adjusting descriptions based on emotion

[1280] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server can increase the user's sense of security by providing more detailed explanations or changing them to more positive expressions.

[1281] Specific examples

[1282] For example, consider the following scenario:

[1283] 1. The user turns on the dashcam and begins capturing video data.

[1284] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[1285] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[1286] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[1287] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[1288] 6. The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[1289] 7. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[1290] In this way, the system of the present invention can improve the user experience by providing explanations that take the user's emotions into consideration.

[1291] The processing flow will be explained below.

[1292] Step 1:

[1293] The user activates the dashcam, which then starts capturing video data in real time.

[1294] Step 2:

[1295] The server receives video data transmitted from the dashcam in real time, establishes a connection based on the specified camera ID, and starts receiving the data stream.

[1296] Step 3:

[1297] The server continuously divides the received video data into frames, so that the video data is treated as individual frames in small, easily processable time units.

[1298] Step 4:

[1299] The server performs pre-processing on each frame, including noise reduction, color correction, and resolution adjustment, to improve the quality of the frame data and increase the accuracy of subsequent analysis.

[1300] Step 5:

[1301] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it detects objects such as vehicles, pedestrians, and traffic lights, and analyzes their positions and movements. Object detection and motion analysis algorithms are used.

[1302] Step 6:

[1303] The server recognizes traffic conditions based on the results of object detection and motion analysis, and the information it obtains makes it possible to grasp detailed information such as the relative positions of vehicles, the status of traffic lights, and the movements of pedestrians.

[1304] Step 7:

[1305] The server passes the analysis results to a generative AI model, which then generates a natural language explanation. The generative AI model then automatically creates an appropriate explanation based on the traffic condition data.

[1306] Step 8:

[1307] The server transmits the generated explanatory text to the user device, which is an electronic device such as a smartphone or PC, and receives and displays the explanatory text to the user.

[1308] Step 9:

[1309] The emotion engine analyzes the user's voice data and facial expression data, identifies emotions from the user's voice and facial expressions, and recognizes the user's emotional state.

[1310] Step 10:

[1311] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server will make the explanation more detailed or change it to more positive language to increase the user's sense of security.

[1312] Specific examples

[1313] Example 1: Normal traffic conditions

[1314] 1. The user turns on the dashcam and begins capturing video data.

[1315] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[1316] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[1317] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[1318] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[1319] Example 2: When the user feels anxious

[1320] 1. The user turns on the dashcam and begins capturing video data.

[1321] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[1322] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[1323] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[1324] 5. The emotion engine analyzes the user's voice and facial expressions and recognizes that they are feeling anxious.

[1325] 6. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[1326] 7. The server sends the adjusted explanation to the user's device, and the user checks the explanation and feels reassured.

[1327] Through these steps, users can understand traffic conditions in real time and receive information tailored to their emotions.

[1328] Example 2

[1329] 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."

[1330] Current in-vehicle camera systems have difficulty recognizing traffic conditions and automatically generating explanatory text. Furthermore, they are unable to provide personalized information that takes into account the user's emotions, which means that safety and the user experience are not sufficiently improved. Therefore, there is a need for the development of a system that can analyze traffic conditions in real time and generate and provide explanatory text that reflects the user's emotional state.

[1331] 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.

[1332] In this invention, the server includes: means for acquiring video data from an in-vehicle camera; preprocessing means for dividing the acquired video data into frames and performing noise removal, color correction, and resolution adjustment; means for analyzing traffic conditions for each preprocessed frame, detecting objects such as vehicles, pedestrians, and traffic lights, and analyzing their positions and movements; means for generating a natural language description using a generative AI model based on the analyzed traffic conditions; means for providing the generated description to a user device; an emotion engine for analyzing a user's voice data and facial expression data to recognize the user's emotional state; and means for adjusting the content of the generated description based on the recognized user's emotional state. This makes it possible to analyze traffic conditions in detail in real time and generate and provide appropriate description that takes the user's emotional state into consideration.

[1333] An "on-board camera" is a device that is mounted on a vehicle and captures images of the road and surrounding area.

[1334] "Video data" refers to continuous image information acquired by an on-board camera.

[1335] A "frame" is an individual still image that makes up video data.

[1336] "Noise reduction" is a process that removes unnecessary signals and disturbances from video data.

[1337] "Color correction" is a process that adjusts the color balance of video data to bring it closer to the original color tone.

[1338] "Resolution adjustment" is a process of changing the number of pixels in the video data to maintain an appropriate resolution.

[1339] "Preprocessing" refers to a series of initial data processing steps, including noise removal, color correction, and resolution adjustment.

[1340] "Traffic conditions" refers to information about the status and movement of vehicles, pedestrians, and traffic lights on the road.

[1341] A "generative AI model" is an artificial intelligence model that generates explanatory text in natural language from analysis results.

[1342] A "description" is a natural language text generated based on the analyzed traffic situation.

[1343] A "user device" is a computing device used by a user, such as a smartphone or PC.

[1344] "Voice data" refers to digital audio information that records the user's speech or voice.

[1345] "Facial expression data" is digital image information that records the facial expressions of a user.

[1346] The "emotion engine" is a system that analyzes voice data and facial expression data to recognize the user's emotional state.

[1347] An "emotional state" is the psychological state that the user is currently feeling (for example, anxiety or relief).

[1348] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[1349] Video data acquisition and preprocessing

[1350] The user activates the in-vehicle camera to capture video data of road conditions. The in-vehicle camera is attached to the vehicle and captures video in real time. The video data is then sent to the server.

[1351] The server receives video data sent from the vehicle-mounted camera in real time. The received video data is divided into frames and then pre-processed. This pre-processing includes noise reduction, color correction, and resolution adjustment. This improves the quality of the video and enables accurate analysis.

[1352] Traffic situation analysis

[1353] The server analyzes the traffic situation for each pre-processed frame. Specifically, it uses deep learning models to detect objects such as vehicles, pedestrians, and traffic lights in the frame and analyzes their positions and movements, providing a detailed understanding of the current traffic situation.

[1354] Generating Descriptions

[1355] The server uses a generative AI model based on the analysis results to convert the traffic situation into a natural language description. For example, if the analysis result is "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk," the server inputs this analysis result as a prompt into the generative AI model to generate a description. A specific description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk" is output.

[1356] Providing a description

[1357] The generated description is sent from the server to the user's device (e.g., a smartphone or PC). The user's device receives the description and displays it on the screen to inform the user of the current traffic situation.

[1358] Emotion engine integration

[1359] Furthermore, the system includes an emotion engine that recognizes the user's emotions. The emotion engine analyzes the user's voice data and facial expression data to recognize the user's emotional state. This allows the system to respond in accordance with the user's emotions.

[1360] Adjusted description

[1361] The server adjusts the generated explanation based on the user's emotional state as recognized by the emotion engine. For example, if the user is feeling anxious, the explanation can be made more detailed to increase the sense of security. For example, it can be adjusted to say, "There are four vehicles ahead, the traffic light is red, and there are pedestrians at the crosswalk, so please proceed safely and with caution."

[1362] Specific examples

[1363] The user activates the in-vehicle camera and starts capturing video data.

[1364] The server receives the video data, performs preprocessing, and then analyzes the traffic situation.

[1365] The analysis results are input into a generative AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and a pedestrian is crossing the crosswalk."

[1366] The server sends the explanation to the user's device, which the user confirms.

[1367] The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[1368] The server adjusts the description to read, "Proceed safely. The light is red, please stop."

[1369] This allows users to accurately grasp the current traffic situation and receive detailed information tailored to their emotions, significantly improving user safety and experience.

[1370] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1371] Step 1:

[1372] The user activates the in-vehicle camera, which captures road conditions in real time and acquires video data.

[1373] Input: Real-time footage from a dashcam.

[1374] Specific operation: The user turns on the dashcam and launches the video capture application.

[1375] Step 2:

[1376] The server receives video data sent from the vehicle-mounted camera in real time and divides the received video data into frames.

[1377] Input: Real-time video data from a dashcam.

[1378] Output: Split frame data.

[1379] Specific operation: The server executes a script that divides the video data into frames and prepares each frame for processing.

[1380] Step 3:

[1381] The server performs pre-processing on the divided frames, which includes noise reduction, color correction, and resolution adjustment.

[1382] Input: Split frame data.

[1383] Output: Preprocessed frame data.

[1384] What happens: The server uses image processing algorithms to remove noise, perform color correction, and adjust the resolution of each frame.

[1385] Step 4:

[1386] The server analyzes the preprocessed frames and identifies traffic situations. Specifically, it uses deep learning models to detect objects such as vehicles, pedestrians, and traffic lights. It analyzes the position and movement of each object.

[1387] Input: Preprocessed frame data.

[1388] Output: Traffic condition analysis result data.

[1389] How it works: The server applies deep learning models to identify the location and movement of vehicles, pedestrians, and traffic lights in each frame.

[1390] Step 5:

[1391] The server uses a generative AI model based on the analysis results to generate a natural language description.

[1392] Input: Traffic analysis result data.

[1393] Output: The generated natural language description.

[1394] Specific operation: The server inputs the analysis results as a prompt sentence to the generative AI model and obtains the generated explanation sentence.

[1395] Example prompt: "There are four vehicles ahead, the light is red, and a pedestrian is crossing the street."

[1396] Step 6:

[1397] The server sends the generated explanation to the user's device, which receives the explanation and displays it to the user.

[1398] Input: The generated natural language description.

[1399] Output: Description that will be displayed on the user's device.

[1400] Specific operation: The server sends the explanatory text to the user's smartphone or PC, and the user device receives it and displays it on the screen.

[1401] Step 7:

[1402] The server uses an emotion engine to analyze the user's voice data and facial expression data to recognize emotions.

[1403] Input: User's voice and facial expression data.

[1404] Output: Perceived emotional state.

[1405] Specific operation: The user provides voice data and facial expression data using a microphone or camera, and the server analyzes it using an emotion engine.

[1406] Step 8:

[1407] The server adjusts the content of the generated narrative based on the recognized emotional state of the user.

[1408] Input: A recognized emotional state and a generated natural language description.

[1409] Output: A narrative tailored based on sentiment.

[1410] Specific operation: Based on the results of emotion analysis, the server adjusts the output of the generative AI model and changes the description to take the user's emotions into consideration.

[1411] This allows the user to accurately grasp the current traffic situation and receive detailed information tailored to their emotions.

[1412] (Application example 2)

[1413] 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."

[1414] In conventional autonomous vehicle systems, there are methods for analyzing in-vehicle camera footage to inform users of traffic conditions, but the feedback provided does not take into account the user's emotional state, which can lead to anxiety or a lack of understanding of the situation. Furthermore, the explanations of traffic conditions are uniform, making it difficult to provide information tailored to the needs of individual users.

[1415] The identification processing 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 acquiring video data from an on-board camera, means for analyzing the acquired video data frame by frame and recognizing traffic conditions, means for automatically generating an explanatory text based on the analyzed traffic conditions, means for providing the generated explanatory text to a user device, means for recognizing the user's emotional state, and means for adjusting the content of the generated explanatory text based on the recognized emotional state. This allows the user to use an autonomous vehicle with peace of mind by providing detailed information about traffic conditions and feedback adjusted according to the user's emotional state.

[1416] An "on-board camera" is a camera device that is mounted on a vehicle and captures video data of the surrounding area in real time.

[1417] "Video data" refers to continuous image information acquired by an on-board camera, and is basic data for traffic situation analysis.

[1418] A "frame" is a unit of an individual still image that makes up video data.

[1419] "Traffic conditions" refers to objects on the road, such as vehicles, pedestrians, and traffic lights, as well as their condition and position.

[1420] The "explanation" is a sentence generated to convey the analyzed traffic situation to the user in natural language.

[1421] "User device" refers to a device owned by a user that can display and operate information, such as a smartphone or tablet.

[1422] "Emotional state" indicates the user's mental and sensory response, and is primarily recognized from voice and facial expression data.

[1423] A "server" is a computer system that processes and analyzes various types of data and is managed centrally.

[1424] A "generative AI model" is an artificial intelligence model that uses machine learning technology to automatically generate sentences corresponding to specific tasks.

[1425] "Adjusting" means changing the content or format of the product based on the perceived information, for example, modifying the generated description based on the user's emotional state.

[1426] The present invention is a system that automatically analyzes traffic conditions from in-vehicle camera images, generates explanatory text in natural language, and provides it to users, and further combines it with an emotion engine that recognizes the user's emotions. This system includes the following main means.

[1427] Video data acquisition and processing

[1428] The user starts the dashcam and starts capturing video data. The server receives the video data sent from the dashcam in real time and divides it into frames. Next, preprocessing is performed on each frame. This includes noise reduction, color correction, and resolution adjustment. Noise reduction, color correction, and resolution adjustment are achieved using libraries such as OpenCV.

[1429] Traffic situation analysis

[1430] The server then performs traffic situation analysis for each pre-processed frame. Specifically, it uses a deep learning model (such as YOLO) to detect objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object. This allows for a detailed understanding of the traffic situation.

[1431] Generating Descriptions

[1432] The server uses a generative AI model (e.g., GPT-4) based on the analysis results to generate natural language descriptions, which can provide detailed descriptions of traffic situations, such as vehicle movements on the road, traffic light status, and pedestrian positions. Specific examples of prompts include the following:

[1433] Input: Number of vehicles: 4, Traffic light status: red, Pedestrian position: on crosswalk

[1434] Prompt: Generate a description based on four vehicles ahead, a red light, and pedestrians crossing the street.

[1435] Providing a description

[1436] The server provides the generated explanatory text to a user device such as a user's smartphone, tablet, etc. The user device receives the explanatory text sent from the server and displays it to the user.

[1437] Adding an Emotion Engine

[1438] This system also includes an emotion engine that recognizes the user's emotions. The emotion engine has the function of recognizing the user's emotional state by analyzing the user's voice data and facial expression data. Azure Face API is used to analyze facial expressions, and Amazon Polly is used to analyze voice data. As a result, the user's emotional state is recognized as "anxiety."

[1439] Adjusting descriptions based on emotion

[1440] The server adjusts the content of the generated explanation based on the user's emotions recognized by the emotion engine. For example, if the user is feeling anxious, the server can increase the user's sense of security by making the explanation more detailed or changing it to more positive language. Specific examples of prompts include the following:

[1441] Input: User emotion: Anxiety

[1442] Prompt: The user is feeling anxious. Please regenerate the description with more detail and positive wording.

[1443] Specific examples

[1444] For example, consider the following scenario:

[1445] 1. The user turns on the dashcam and begins capturing video data.

[1446] 2. The server receives the video data, divides it into frames, and performs preprocessing.

[1447] 3. The server analyzes the traffic conditions in each frame and detects the status of vehicles, pedestrians, traffic lights, etc.

[1448] 4. The server passes the analysis results to the generation AI model, which generates a description such as "There are four vehicles ahead, the traffic light is red, and pedestrians are crossing the crosswalk."

[1449] 5. The server sends the generated explanation to the user device, and the user confirms the explanation.

[1450] 6. The emotion engine analyzes the user's voice data and facial expression data and recognizes that the user is feeling anxious.

[1451] 7. The server adjusts the explanation based on the user's emotions, changing it to something like "Proceed safely. The light is red, so please stop."

[1452] In this way, the system of the present invention can improve the user experience by providing explanations that take the user's emotions into consideration.

[1453] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1454] Step 1:

[1455] The user activates the in-vehicle camera and starts capturing video data.

[1456] Input: Start instruction for in-car camera

[1457] Output: Start of real-time video data acquisition

[1458] Specific operation: The user activates the dashcam from the smartphone app.

[1459] Step 2:

[1460] The server receives video data transmitted from the vehicle-mounted camera in real time and divides it into frames.

[1461] Input: Real-time video data

[1462] Output: List of frame data

[1463] Specific operation: Video data is divided into frames and a timestamp is stored for each frame.

[1464] Step 3:

[1465] The server performs pre-processing on each frame, which includes noise reduction, color correction, and resolution adjustment.

[1466] Input: List of frame data

[1467] Output: Preprocessed frame data

[1468] What it does: It uses the OpenCV library to perform noise reduction, color correction, and resolution adjustment on each frame.

[1469] Step 4:

[1470] The server performs traffic situation analysis for each frame after preprocessing. Specifically, it uses a deep learning model (e.g., YOLO) to detect objects such as vehicles, pedestrians, and traffic lights, and analyzes the position and movement of each object.

[1471] Input: Preprocessed frame data

[1472] Output: Information about detected objects (type, location, movement, etc.)

[1473] Specific operation: The YOLO model is applied to frame data, and the identification of objects indicating traffic conditions and their location information are output.

[1474] Step 5:

[1475] The server generates a natural language explanation based on the analysis results using a generative AI model (e.g., GPT-4).

[1476] Input: Object information (type, position, movement, etc.)

[1477] Output: Generated natural language description

[1478] Specific operation: Based on the analysis results, a prompt sentence is input into the AI ​​model to generate an explanatory sentence. Specific prompt sentence examples:

[1479] Input: Number of vehicles: 4, Traffic light status: red, Pedestrian position: on crosswalk

[1480] Prompt: Generate a description based on four vehicles ahead, a red light, and pedestrians crossing the street.

[1481] Step 6:

[1482] The server provides the generated explanation to the user device.

[1483] Input: Generated description

[1484] Output: Send to user device

[1485] Specific operation: The generated explanation is sent to the user's smartphone or tablet and displayed.

[1486] Step 7:

[1487] The server acquires the user's voice data and facial expression data and recognizes their emotional state using an emotion engine.

[1488] Input: Voice data, facial expression data

[1489] Output: Emotional state recognition result

[1490] Specific operation: The voice data and facial expression data collected by the user device are sent to the server and analyzed using an emotion engine (e.g., Azure Face API or Amazon Polly).

[1491] Step 8:

[1492] The content of the generated description is adjusted based on the emotional state recognized by the server.

[1493] Input: Emotional state recognition results, generated description

[1494] Output: Adjusted description

[1495] Specific behavior: Depending on the recognized emotional state, the generative AI model is re-prompted to generate an adjusted explanation. Specific prompt examples:

[1496] Input: User emotion: Anxiety

[1497] Prompt: The user is feeling anxious. Please regenerate the description with more detail and positive wording.

[1498] Step 9:

[1499] The server provides the adjusted explanation back to the user device.

[1500] Input: Adjusted description

[1501] Output: Retransmission to user device

[1502] Specific operation: The adjusted description is resent to the user's smartphone or tablet and displayed.

[1503] This provides detailed information about traffic conditions and tailored feedback based on the user's emotional state, allowing users to use self-driving vehicles with peace of mind.

[1504] 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.

[1505] 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.

[1506] 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.

[1507] 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.

[1508] 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.

[1509] 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.

[1510] 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).

[1511] 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.

[1512] 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."

[1513] 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.

[1514] 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).

[1515] 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.

[1516] 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.

[1517] 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.

[1518] 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.

[1519] 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.

[1520] 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.

[1521] 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.

[1522] 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.

[1523] 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.

[1524] 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.

[1525] The following is further disclosed regarding the above embodiment.

[1526] (Claim 1)

[1527] A means for acquiring video data from an in-vehicle camera;

[1528] A means for analyzing the acquired video data frame by frame and recognizing traffic conditions;

[1529] A means for automatically generating a description based on the analyzed traffic conditions;

[1530] means for providing the generated narrative to a user device;

[1531] A system including:

[1532] (Claim 2)

[1533] 2. The system of claim 1, wherein the means for recognizing traffic conditions detects the status of vehicles, pedestrians, and traffic lights.

[1534] (Claim 3)

[1535] 2. The system of claim 1, wherein the means for generating the explanatory text generates the explanatory text using a generative AI model.

[1536] "Example 1"

[1537] (Claim 1)

[1538] A means for acquiring video data from an in-vehicle camera;

[1539] means for pre-processing the acquired video data frame by frame;

[1540] means for recognizing a traffic situation for each preprocessed frame;

[1541] means for automatically generating a description based on the recognized traffic situation;

[1542] means for providing the generated narrative to a user device;

[1543] A system including:

[1544] (Claim 2)

[1545] 2. The system according to claim 1, wherein the means for recognizing traffic conditions includes means for detecting the status of vehicles, pedestrians, and traffic lights, and analyzing their respective positional relationships and movements.

[1546] (Claim 3)

[1547] The system of claim 1, wherein the means for generating the explanatory text includes means for generating the explanatory text using a generative AI model and creating detailed text about traffic conditions based on the analysis results.

[1548] "Application Example 1"

[1549] (Claim 1)

[1550] A means for acquiring video data from an in-vehicle camera;

[1551] A means for analyzing the acquired video data frame by frame and recognizing traffic conditions;

[1552] A means for automatically generating a description based on the analyzed traffic conditions;

[1553] means for providing the generated narrative to a user device;

[1554] A means for displaying the generated explanatory text in real time on smart glasses, which are a type of user device;

[1555] A system including:

[1556] (Claim 2)

[1557] 2. The system of claim 1, wherein the means for recognizing traffic conditions detects the status of vehicles, pedestrians, and traffic lights.

[1558] (Claim 3)

[1559] 2. The system of claim 1, wherein the means for generating the explanatory text generates the explanatory text using a generative AI model.

[1560] "Example 2: Combining Emotion Engines"

[1561] (Claim 1)

[1562] A means for acquiring video data from an in-vehicle camera;

[1563] A pre-processing means for dividing the acquired video data into frames and performing noise removal, color correction, and resolution adjustment;

[1564] A means for analyzing traffic conditions for each preprocessed frame, detecting objects such as vehicles, pedestrians, and traffic lights, and analyzing their positions and movements;

[1565] a means for generating a natural language description using a generative AI model based on the analyzed traffic conditions;

[1566] means for providing the generated narrative to a user device;

[1567] An emotion engine that analyzes the user's voice data and facial expression data to recognize their emotional state;

[1568] means for adjusting the content of the generated narrative based on the recognized emotional state of the user;

[1569] A system including:

[1570] (Claim 2)

[1571] 2. The system according to claim 1, wherein the means for analyzing traffic conditions detects the status of vehicles, pedestrians, and traffic lights.

[1572] (Claim 3)

[1573] 2. The system of claim 1, wherein the means for generating the explanatory text generates the explanatory text using a generative AI model.

[1574] "Application example 2 when combining emotion engines"

[1575] (Claim 1)

[1576] A means for acquiring video data from an in-vehicle camera;

[1577] A means for analyzing the acquired video data frame by frame and recognizing traffic conditions;

[1578] A means for automatically generating a description based on the analyzed traffic conditions;

[1579] means for providing the generated narrative to a user device;

[1580] means for recognizing the emotional state of a user;

[1581] means for adjusting the content of the generated narrative based on the recognized emotional state;

[1582] A system including:

[1583] (Claim 2)

[1584] 10. The system of claim 1, wherein the means for recognizing traffic conditions detects the status of vehicles, pedestrians, and traffic lights.

[1585] (Claim 3)

[1586] 2. The system of claim 1, wherein the means for generating the explanatory text generates the explanatory text using a generative AI model. [Explanation of symbols]

[1587] 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 acquiring video data from an in-vehicle camera; A means for analyzing the acquired video data frame by frame and recognizing traffic conditions; A means for automatically generating a description based on the analyzed traffic conditions; means for providing the generated narrative to a user device; A system including:

2. 2. The system of claim 1, wherein the means for recognizing traffic conditions detects the status of vehicles, pedestrians, and traffic lights.

3. 2. The system of claim 1, wherein the means for generating the exposition generates the exposition using a generative AI model.

Images