system

Abstract

We provide the system. [Solution] A means for acquiring video data and detecting an object from said video data, A means of superimposing an indistinguishable alternative display onto the detected object, Means for monitoring the behavior of the object and detecting abnormal behavior, Means for generating a notification based on the detection of the abnormal behavior, Means for canceling the alternative display and reconstructing the original image as needed, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, and includes steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, 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

Summary of the Invention

Problems to be Solved by the Invention

[0004] In modern society, many video surveillance systems installed to ensure public safety have a serious problem of privacy infringement. While real-time crime prevention is required, appropriate personal information protection is indispensable. However, conventional security cameras often capture personal characteristics as they are and do not provide sufficient privacy protection for surveillance. Also, it is not easy to accurately detect abnormal behavior from video and respond appropriately. Due to these problems, the balance between crime prevention and personal privacy protection has emerged as a difficult issue.

Means for Solving the Problems

[0005] This invention provides a system that detects objects from video data and overlays an illegible substitute display onto the detected object to achieve a security function while ensuring individual privacy. Specifically, a video data acquisition means detects the object, and based on this detection, a substitute display means overlays an avatar or other substitute display onto the object. Furthermore, an abnormal behavior detection means monitors abnormal behavior by comparing it with predefined behavior patterns, and if such abnormal behavior is detected, a notification generation means notifies the administrator of the information. In the event of a specific situation or after appropriate legal procedures have been followed, the original video can be reconstructed by a substitute display removal means. In this way, an efficient security system is realized while respecting privacy.

[0006] "Video data" refers to digital data of visual information acquired by a camera or other recording device.

[0007] "Object" refers to people or objects detected within the video data.

[0008] A "difficult-to-identify alternative representation" is a graphical or digital representation that is superimposed on an object to prevent a third party from easily identifying that object.

[0009] "Abnormal behavior" refers to movements or actions that differ from predefined normal behavioral patterns and indicate a security risk.

[0010] A "notification" is a report message generated when abnormal behavior is detected, intended to inform the system administrator of that information.

[0011] "Removing substitute displays" is a process that removes substitute displays that are difficult to identify and restores the original video data. [Brief explanation of the drawing]

[0012] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine. [Modes for carrying out the invention]

[0013] Hereinafter, an example of an embodiment of the system relating to the technology of this disclosure will be described with reference to the attached drawings.

[0014] First, the terms used in the following description will be explained.

[0015] In the following embodiments, the labeled processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0016] In the following embodiments, the labeled RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0017] In the following embodiments, the labeled storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0018] In the following embodiments, the labeled communication I / F (Interface) is an interface including a communication processor and 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), or Bluetooth (registered trademark), and the like.

[0019] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0020] [First Embodiment]

[0021] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0022] As shown in Figure 1, the 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.

[0023] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0024] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0025] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and 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.

[0026] 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 perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0027] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0028] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0029] As shown in Figure 2, in the data processing device 12, a specific processing 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" related to the technology of this 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 according to the specific processing program 56 executed on the RAM 30.

[0030] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0031] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0032] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0033] This invention is implemented as an AI-powered security system to ensure public safety while protecting privacy. The core of this system lies in a technology that detects abnormal behavior through real-time video data analysis while making it difficult to identify individuals.

[0034] System implementation example:

[0035] The server acquires video data captured by security cameras in real time. The video data is immediately processed by an AI agent to detect people and objects present in the video. Predefined avatars or other graphical representations are overlaid on detected objects to make them difficult to identify. This process is performed in real time in response to the movement of the objects, thus providing continuous privacy protection.

[0036] Abnormal behavior is detected by an AI agent on the server, which compares the actual behavior with pre-trained normal behavior patterns. If abnormal behavior is detected through this comparison, the server immediately sends a notification to the administrator's terminal. The notification includes the location, time, and specific details of the abnormal behavior.

[0037] For example, suppose there is a camera system installed in a shopping mall. The server monitors the entrances and exits of each store in the mall and tracks people coming and going. If the system detects the same person repeatedly entering and exiting a particular area during a specific time period, this behavior is considered abnormal and is deemed unusual. An AI agent detects this anomaly and notifies the administrator of the anomaly while concealing the movements of the person in the relevant video footage with an avatar.

[0038] For users, i.e., visitors to the mall, the fact that these security cameras are operated without infringing on their privacy can be a source of reassurance. On the other hand, for administrators who operate the servers and terminals, they are a powerful tool for continuous and efficient security.

[0039] This system can be customized according to the usage environment, and camera placement, types of alternative displays, and criteria for abnormal behavior are set during system installation. This allows for flexible adaptation to facilities of various sizes and types.

[0040] The following describes the processing flow.

[0041] Step 1:

[0042] The server collects video data in real time from the installed security cameras. The video data is continuously divided into frames and sent as data for analysis by the AI ​​agent.

[0043] Step 2:

[0044] The AI ​​agent on the server uses an image recognition algorithm to detect objects in each received frame. These objects include people and specific objects.

[0045] Step 3:

[0046] The AI ​​agent automatically overlays a difficult-to-identify alternative representation onto the detected object. This alternative representation is created, for example, using an avatar or special filters.

[0047] Step 4:

[0048] The server records video data with alternative displays applied while maintaining real-time data for monitoring specific behaviors. At this stage, data is generated in a way that ensures privacy.

[0049] Step 5:

[0050] The AI ​​agent continuously monitors the movement of the target object and determines whether there is any abnormal behavior. Abnormal behavior is defined by comparing it to a pre-defined normal behavior pattern.

[0051] Step 6:

[0052] When abnormal behavior is detected, the server immediately sends a notification to the administrator's terminal. The notification includes the time, location, and nature of the abnormal behavior.

[0053] Step 7:

[0054] The device receives the notification and provides information for the administrator to review the details of the anomaly and take appropriate action if necessary.

[0055] Step 8:

[0056] Under certain circumstances or court orders, the server will deactivate the alternative display, allowing access to the original video data. The process of deactivating the alternative display is performed only under strict access control.

[0057] (Example 1)

[0058] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0059] The goal is to create a security system that ensures public safety while protecting privacy. Conventional video surveillance systems make it easy to identify individuals, leading to privacy concerns. Furthermore, detecting abnormal behavior from large amounts of video data relies on human intervention, limiting efficiency and accuracy.

[0060] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0061] In this invention, the server includes means for acquiring video information and detecting a target from the video information, means for overlaying and displaying an alternative display that is difficult to identify onto the detected target, and means for managing the behavior of the target and detecting abnormal behavior. This makes it possible to automatically detect abnormal behavior efficiently and accurately while protecting privacy.

[0062] "Video information" refers to video data acquired by cameras and other imaging devices.

[0063] "Target" refers to a specific object, such as a person or thing, detected within the video information.

[0064] "Alternative representation" refers to avatars or graphical display formats that are superimposed on a subject to make it difficult to identify the individual.

[0065] "Abnormal operation" refers to behavior that deviates from a pre-set normal operating pattern.

[0066] "Notification" refers to a message sent to the administrator to convey relevant information when abnormal operation is detected.

[0067] A "machine learning model" refers to the technology that forms the basis of algorithms and programs used in the analysis of video information.

[0068] "Identification information" refers to information that includes characteristics used to recognize a specific individual or object.

[0069] This invention is a security system designed to ensure public safety while protecting privacy. The system operates primarily from a server and uses video information acquired from security cameras to detect targets and monitor for abnormal activity in real time.

[0070] Server role:

[0071] The server first receives video information acquired from security cameras, and an AI agent immediately analyzes it. This analysis uses machine learning models such as TENSORFLOW® and PyTorch. The AI ​​agent detects objects such as people and objects from the video information and overlays difficult-to-identify alternative displays onto them using libraries such as OpenCV. These alternative displays are designed to conceal personal identification information and protect privacy.

[0072] Detection and notification of abnormal operation:

[0073] The server detects abnormal behavior by comparing pre-configured behavioral patterns with real-time actions. If abnormal behavior is detected, a notification is immediately sent to the administrator's terminal. The notification includes the location, time, and specific actions taken when the abnormality occurred.

[0074] Specific example:

[0075] For example, if security cameras are installed in a shopping mall, the server monitors each entrance and exit within the mall and tracks the movements of visitors. If an AI agent determines that the same individual is frequently entering and exiting the mall within a specific time period is abnormal, it notifies the administrator with video footage overlaid with alternative displays. Meanwhile, users can use the mall in a safe environment, feeling secure knowing that their privacy is protected by this system.

[0076] Example of a prompt:

[0077] "Design an AI program for a shopping mall's security camera system that detects behavioral patterns where the same person enters and exits the building multiple times within a specific period. This program should use avatars to make it difficult to identify individuals in the video footage, and should notify administrators when abnormal behavior is detected."

[0078] This system is designed to adapt to diverse environments, allowing for flexible configuration of camera placement, alternative display formats, and criteria for abnormal behavior, thereby achieving efficient security.

[0079] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0080] Step 1:

[0081] The server acquires video data from security cameras in real time. The input is the video data transmitted from the cameras, which is a raw video stream. The server captures this data and temporarily stores it to prepare for subsequent processing. This step also involves converting the video data into a format suitable for capture.

[0082] Step 2:

[0083] The server sends the acquired video data to the AI ​​agent and begins analysis. The input is the video data acquired in step 1. The AI ​​agent uses a machine learning model to detect objects in the video. Specifically, it uses image processing algorithms to identify people and objects frame by frame. This analysis outputs a list of specific objects present in the video.

[0084] Step 3:

[0085] The server applies a difficult-to-identify alternative display to the objects detected by the AI ​​agent. The input is a list of objects detected in step 2 and their location information. The server uses libraries such as OpenCV to overlay avatars or graphical displays on the objects. This output is a video stream with personal information protected. Specifically, for example, processing is performed such as overlaying an anime character mask on the faces of people.

[0086] Step 4:

[0087] The server monitors the target's actions within the video data and checks for abnormal behavior. The input consists of an anonymized video stream and the analysis data from step 2. The server compares this to a pre-configured normal behavior pattern and uses a machine learning algorithm to detect anomalies. If abnormal behavior is detected, detailed information about it is output.

[0088] Step 5:

[0089] If the server detects abnormal behavior, it sends a notification to the administrator's terminal. The input is the detailed information of the abnormal behavior output in step 4. Based on this information, the server generates a notification message and sends it to the terminal via email or a dedicated application. The output is a notification message for the administrator and includes the specific nature, location, and time of the abnormality.

[0090] Step 6:

[0091] The administrator's terminal receives the notification and displays it on the screen. The terminal receives the notification message from the server as input and displays an alert on the screen to notify the administrator of the occurrence of an anomaly. This output is a visual notification to the administrator, allowing them to quickly take necessary security measures.

[0092] (Application Example 1)

[0093] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0094] In modern urban environments, balancing enhanced security with privacy protection is a challenging task. While security camera systems improve public safety, they also carry the potential to infringe on individual privacy. To address this contradiction, there is a need to provide a secure method that makes it difficult to identify individuals on screen while rapidly detecting abnormal behavior and providing effective security notifications.

[0095] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0096] In this invention, the server includes means for acquiring video information and detecting an object from the video information, means for visualizing the detected object by superimposing an alternative visual representation that is difficult to identify, and means for monitoring the movement of the object and detecting abnormal behavior. This enables rapid detection of abnormal behavior to ensure public safety and real-time information provision while securely protecting users' personal information.

[0097] "Visual information" refers to visual data acquired from cameras and recording devices.

[0098] "Target" refers to a person or object detected within the video information.

[0099] "Difficult-to-identify alternative visual representations" refer to visual substitutes used to make it difficult to identify an object.

[0100] "Trends" refers to the state or changes in the behavior or movement of an object.

[0101] "Abnormal behavior" refers to actions that are detected as deviating from normal patterns.

[0102] "Communication" refers to notifications and messages generated based on the detection of abnormal operation.

[0103] A "personal device" refers to a portable electronic device owned by a user.

[0104] "Visualization" refers to the process of representing information visually to aid its understanding.

[0105] This invention is a security system that balances safety and privacy. This system consists of multiple components and is configured as follows:

[0106] The server first acquires video information from surveillance cameras and recording devices. This acquired video information is analyzed in real time using AI technology. AI video analysis models such as TensorFlow and PyTorch are used for this analysis. The AI ​​agent detects objects from the video information. Alternative visual representations are overlaid on these detected objects to make identification difficult. This process protects individual privacy.

[0107] The server continuously monitors the target's activity and detects abnormal behavior. To detect abnormal behavior, a comparison is made with a pre-configured normal operating pattern. This comparison involves data stream processing in the cloud, and as soon as an anomaly is detected, a communication is immediately sent to the user's personal device. Cloud services such as AWS® Lambda and Firebase Cloud Messaging may be used for this communication.

[0108] Users can receive real-time notifications through their personal devices and obtain information about safety in public places. For example, if unusual behavior is detected in a shopping mall, the user will receive a communication stating, "There is unusual activity in a specific area of ​​the mall." This allows the user to immediately understand the situation and take safe actions.

[0109] An example of a generative AI model or prompt related to this system is, "How can I develop an application that detects abnormal behavior in real time within a shopping mall and sends notifications to users?"

[0110] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0111] Step 1:

[0112] The server acquires video information from surveillance cameras and recording devices. This input data is a real-time video feed captured by the cameras. The server prepares this video information as a dataset for AI analysis and converts it to a format to be used in the next step.

[0113] Step 2:

[0114] The server detects objects from the acquired video information. At this stage, an AI video analysis model using TensorFlow or PyTorch is executed to detect objects (such as people or vehicles). The input is the prepared video information, and the output is the coordinate data and feature data of the detected objects. Alternative visual representations that are difficult to identify are superimposed on the detected objects.

[0115] Step 3:

[0116] The server monitors the target's movements and detects abnormal behavior. The input consists of the detected target's coordinate and feature data. Data processing is performed by comparing the data with normal behavior patterns. If an anomaly is detected, the result is passed to the next step. The output is flag information indicating whether abnormal behavior was detected.

[0117] Step 4:

[0118] If abnormal behavior is detected, the server uses a cloud service to send a communication to the user's personal device. The input is flag information about the abnormal behavior and the location of the affected device. The output is a notification sent through a platform such as Firebase Cloud Messaging. This notification will describe the area where the abnormality occurred and what kind of abnormality it is.

[0119] Step 5:

[0120] The user's device receives notifications from the server and displays the information visually. The device analyzes this input data and generates a screen that notifies the user of warnings and recommended actions. The output is detailed information about the abnormal behavior displayed on the device's screen.

[0121] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0122] This invention provides a system that achieves more advanced crime prevention and abnormal behavior detection by recognizing emotions while protecting user privacy. This system incorporates an emotion engine that recognizes user emotions in conjunction with analyzing video data from security cameras. The emotion engine uses image analysis technology to analyze the user's facial expressions and movements, and evaluates their emotional state in real time. This evaluation result is used to determine abnormal behavior and apply alternative displays.

[0123] System implementation example:

[0124] The server first collects video data from security cameras and uses an AI agent to detect people. A substitute image is then overlaid on the detected individuals to make identification difficult. This substitute image, for example, covers the person's appearance as an avatar, thus protecting their privacy.

[0125] Furthermore, the emotion engine analyzes the person's facial expressions and movements, recognizing emotional states such as joy, anger, sadness, and happiness in real time. These recognized emotions contribute to the identification of abnormal behavior. For example, even if the behavior is normal, if feelings of anxiety or fear are recognized, it is logged by the server as a type of abnormal behavior, and a notification is sent to the terminal.

[0126] As a concrete example, consider a security system within a train station. This system constantly monitors passenger movements and analyzes their behavior. If a passenger frequently exhibits unstable emotions according to the emotion engine, the server determines this behavior is abnormal by comparing it to pre-configured behavioral patterns and sends a detailed notification to the administrator's terminal. It is also possible to dynamically change alternative displays based on changes in emotion. For example, if fear persists, a stronger alternative display can be applied to conceal the fear.

[0127] By utilizing an emotion engine, this system goes beyond simple motion detection, enabling appropriate responses tailored to individual emotions and achieving a more accurate crime prevention system. This dramatically improves public safety while simultaneously ensuring thorough consideration of user privacy.

[0128] The following describes the processing flow.

[0129] Step 1:

[0130] The server acquires video data from security cameras in real time. The video data is divided into frames and sent to the AI ​​agent in a format that is easy to analyze.

[0131] Step 2:

[0132] The AI ​​agent on the server processes the received frames based on image recognition technology to detect people and objects in the video. In the process, it extracts their location information and movement data.

[0133] Step 3:

[0134] Each detected object is overlaid with a difficult-to-identify alternative representation. The server's alternative representation means cover the object with avatars or filters to protect individual privacy.

[0135] Step 4:

[0136] The server also plays a role in continuously monitoring the movement of the target object and detecting abnormal behavior. It collects patterns of movement, and if they deviate from pre-configured normal behavior patterns, they are marked as abnormal behavior.

[0137] Step 5:

[0138] The emotion engine analyzes the facial expressions and movements of people in the video and evaluates their emotional state. The server uses this information to further evaluate abnormal behavior in detail.

[0139] Step 6:

[0140] If behavior is deemed abnormal based on the emotional state, the server sends a notification to the administrator's terminal. The notification includes the type of abnormal behavior, the time it occurred, and emotional information.

[0141] Step 7:

[0142] The terminal processes notifications from the server and displays relevant video and emotion data to enable administrators to respond quickly. Based on this display, it helps determine whether or not to take action on-site.

[0143] Step 8:

[0144] If necessary, the server will activate the alternative display deactivation function, providing access to the original video under certain conditions. For example, in legal requests or emergencies, the video can be viewed with all privacy restrictions removed.

[0145] (Example 2)

[0146] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0147] Conventional security systems have difficulty detecting abnormal behavior without identifying individuals in the video, posing a challenge in accurately detecting anomalies while protecting individual privacy. Furthermore, abnormal behavior is judged based solely on actions and does not take emotional states into consideration, leading to the possibility of false positives.

[0148] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0149] In this invention, the server includes means for acquiring video data and detecting a target from that data, means for overlaying a substitute display that is difficult to identify onto the detected target, and means for analyzing the emotional state of the target and detecting abnormal behavior based on that emotional state. This makes it possible to accurately detect abnormal behavior that takes emotional state into account while protecting the privacy of individuals.

[0150] "Video data" refers to visual information represented in digital format, acquired by devices such as cameras for security and surveillance purposes.

[0151] A "subject" is something that exhibits a specific action or characteristic within the video data and is detected as a person or object.

[0152] "Alternative displays" are visual overlays used to protect the privacy of detected objects, and their role is to make identification difficult.

[0153] "Emotional state" refers to the internal psychological state inferred from the subject's facial expressions and actions, and includes sensory reactions such as joy, anger, sadness, and happiness.

[0154] "Abnormal behavior" refers to movements or states that deviate from normal behavioral patterns, indicating that the person being monitored may be experiencing some kind of danger or instability.

[0155] A "notification" is information sent to the relevant parties or systems upon detection of abnormal behavior, intended to prompt appropriate action.

[0156] This invention provides a system that utilizes security cameras installed in public places to detect abnormal behavior with high accuracy through video data. The system enables advanced surveillance using sentiment analysis while protecting user privacy.

[0157] The server collects video data from security cameras in real time. The collected video data is then used to detect people via an AI agent. In this case, commonly used image recognition technologies include TensorFlow and OpenCV.

[0158] Individuals who are detected will have alternative markers overlaid to make identification difficult. These alternative markers are provided in the form of mosaics or avatars. This ensures user privacy and provides a sense of security that they are being monitored.

[0159] Furthermore, the server uses an emotion engine to analyze a person's facial expressions and movements, and evaluates their emotional state in real time. Emotional states include specific expressions such as anger, joy, and fear.

[0160] Abnormal behavior is detected by comparing pre-configured behavioral patterns with analyzed emotional states. Even normal actions are considered abnormal if emotions such as anxiety or fear are recognized, and are logged on the server. If the state of the target is determined to be abnormal, a notification is sent to the terminal.

[0161] A concrete example of its use is in a security system within a train station. This system has the function of monitoring and constantly analyzing passengers' behavior and emotions. If a passenger frequently exhibits unstable emotions, the emotion engine will determine this to be an anomaly and provide a notification with details to the administrator's terminal.

[0162] Examples of prompts for the generative AI model include: "Please describe a system that identifies emotions from video data. How does this system detect abnormal behavior while protecting privacy?" and "Please give specific examples of the role and practical applications of emotion engines in security cameras."

[0163] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0164] Step 1:

[0165] The server collects video data from security cameras in real time. The input is the camera stream. This data is converted to a digital format and stored within the server. Specifically, the video is formatted for processing frame by frame, and pre-processing is performed to remove noise. The output is video data in an analyzable format.

[0166] Step 2:

[0167] The server uses an AI agent to detect people from video data. The input here is the processed video data from the previous step. A person detection model (e.g., HOG + SVM using OpenCV, or a deep learning model using TensorFlow) is applied to this data. As a result, the output is data showing the location and size of the detected people.

[0168] Step 3:

[0169] The server overlays a substitute image onto the detected person. The input is the person's location data obtained in step 2. Using image processing techniques, mosaics or avatars are overlaid on the person's face and body to make identification difficult. This process generates edited frames. The output is privacy-protected video data.

[0170] Step 4:

[0171] The server uses an emotion engine to analyze emotions from privacy-protected video data. The data from step 3 is used as input. The engine analyzes the subject's facial expressions and movements and applies image recognition and motion analysis algorithms to identify specific emotions (such as joy, anger, or anxiety). As a result, the output is obtained as the analyzed emotional state.

[0172] Step 5:

[0173] The server detects abnormal behavior based on the emotion analysis results. The input consists of emotional state and behavior pattern data obtained in step 4. The server compares the emotional state with a pre-configured normal behavior pattern to determine if there is an abnormality. The output is notification data if abnormal behavior is detected.

[0174] Step 6:

[0175] The server sends a notification to the terminal when an anomaly is detected. The input includes the notification data generated in step 5. From this data, a message is created that describes the specific abnormal behavior in detail. The output is a notification message displayed on the user's or administrator's terminal. This notification allows for prompt action to be taken as needed.

[0176] (Application Example 2)

[0177] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0178] In modern surveillance systems, balancing privacy protection and security is a crucial challenge. Conventional technologies only monitor the movement of objects to detect anomalies, making it difficult to detect abnormal behavior accompanied by unstable emotional states early. Furthermore, methods using alternative displays to protect personal information in video footage lack the flexibility to adapt to different situations. This has made it difficult to achieve both improved public safety and protection of individual privacy.

[0179] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0180] In this invention, the server includes means for acquiring video data and detecting an object from the video data, means for overlaying and displaying a difficult-to-identify alternative display on the detected object, and means for recognizing the emotional state of the object. This enables early detection of abnormal behavior based on emotional state and enhanced privacy protection according to the situation.

[0181] "Video data" refers to a collection of visual information acquired by cameras and other visual sensors.

[0182] "Object" refers to a person or object that exists within the video data and is detected based on specific conditions.

[0183] "Detection" refers to the process of identifying and distinguishing an object from video data.

[0184] "Alternative displays" are visual substitutes used to conceal personal information or privacy-related elements in the original video.

[0185] "Emotional state" refers to the psychological elements and characteristics recognized to express human emotions.

[0186] "Abnormal behavior" refers to actions that have characteristics different from the norm and require attention in monitoring systems.

[0187] A "notification" is a message generated by a system to inform about a specific event or state.

[0188] "Dynamic modification" means changing the display or behavior in response to changes in circumstances or parameters.

[0189] A "pre-configured behavioral pattern" is a pre-programmed behavioral profile used as a criterion for detecting abnormal behavior.

[0190] The system for implementing this invention is a security system that uses video data to detect objects and recognize their emotional states. The server acquires video data using cameras and other visual sensors and detects objects from it. Image processing libraries such as OpenCV are used for object detection.

[0191] The server uses an emotion engine powered by a generative AI model to analyze the emotional state of objects in real time from video data. This allows the emotion engine to recognize emotions such as joy, anger, sadness, and happiness, and detect abnormal behavior based on this data. This information is compared with pre-configured behavioral and emotional patterns, and if an anomaly is detected, a notification is sent to the user's device.

[0192] To protect user privacy, the server overlays a substitute display on detected objects. This substitute display is processed using classes such as PrivacyOverlay and is dynamically modified depending on the context. This dynamic modification allows for adjustment of its intensity based on the perceived emotional state.

[0193] As a concrete example, a home camera system may have a function that recognizes the emotions of family members and notifies them if they are experiencing persistent feelings of depression, suggesting that they may need support. In this case, a generative AI model is used to prompt the implementation of a specific emotion recognition algorithm.

[0194] An example of a prompt might be, "Implement an algorithm that recognizes an individual's emotions in real time while protecting their privacy when they are photographed." This prompt is used to provide generative AI models with implementation guidelines for emotion recognition and privacy protection features.

[0195] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0196] Step 1:

[0197] The server acquires video data in real time from cameras and vision sensors. This video data serves as input for processing. The input images are converted into a format that can be analyzed using OpenCV.

[0198] Step 2:

[0199] The server detects objects from the acquired video data. Here, image processing algorithms are used to identify people and objects in the video and output information about the detected objects. Contour recognition and face recognition technologies are used for image analysis.

[0200] Step 3:

[0201] The server overlays a substitute display onto the detected object. The input uses the object's location and shape data, and the PrivacyOverlay class is used to apply a substitute display that hides personal information. This process results in an output where a portion of the video is modified while protecting privacy.

[0202] Step 4:

[0203] The server recognizes the emotional state of the subject from the displayed alternative video. Using a generative AI model, it analyzes facial expressions and movements in the input video and outputs emotional data. A pre-trained emotion recognition algorithm is used to classify the emotional state.

[0204] Step 5:

[0205] The server detects abnormal behavior by comparing emotional states with previously recorded behavioral patterns. Emotional state data is used as input, and abnormal behavior is output only if it matches the conditions for being judged as abnormal. At this stage, pattern matching is performed for both emotions and behavior.

[0206] Step 6:

[0207] When abnormal behavior is detected, the server generates a notification and sends it to the designated device. The notification includes information such as the type and time of the abnormality and the emotional state, and a warning message is displayed to the user. This output is then fed into the alert system.

[0208] Step 7:

[0209] Furthermore, the alternative display is dynamically changed according to the detected emotional state. Based on the emotional data as input, the PrivacyOverlay class is executed again, and a new alternative display is applied to the video. As a result, a video that makes the user feel more secure is output.

[0210] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating 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.

[0211] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0212] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0213] [Second Embodiment]

[0214] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0215] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0216] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0217] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0218] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0219] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0220] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0221] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0222] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0223] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0224] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0225] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. 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".

[0226] This invention is implemented as an AI-powered security system to ensure public safety while protecting privacy. The core of this system lies in a technology that detects abnormal behavior through real-time video data analysis while making it difficult to identify individuals.

[0227] System implementation example:

[0228] The server acquires video data captured by security cameras in real time. The video data is immediately processed by an AI agent to detect people and objects present in the video. Predefined avatars or other graphical representations are overlaid on detected objects to make them difficult to identify. This process is performed in real time in response to the movement of the objects, thus providing continuous privacy protection.

[0229] Abnormal behavior is detected by an AI agent on the server, which compares the actual behavior with pre-trained normal behavior patterns. If abnormal behavior is detected through this comparison, the server immediately sends a notification to the administrator's terminal. The notification includes the location, time, and specific details of the abnormal behavior.

[0230] For example, suppose there is a camera system installed in a shopping mall. The server monitors the entrances and exits of each store in the mall and tracks people coming and going. If the system detects the same person repeatedly entering and exiting a particular area during a specific time period, this behavior is considered abnormal and is deemed unusual. An AI agent detects this anomaly and notifies the administrator of the anomaly while concealing the movements of the person in the relevant video footage with an avatar.

[0231] For users, i.e., visitors to the mall, the fact that these security cameras are operated without infringing on their privacy can be a source of reassurance. On the other hand, for administrators who operate the servers and terminals, they are a powerful tool for continuous and efficient security.

[0232] This system can be customized according to the usage environment, and camera placement, types of alternative displays, and criteria for abnormal behavior are set during system installation. This allows for flexible adaptation to facilities of various sizes and types.

[0233] The following describes the processing flow.

[0234] Step 1:

[0235] The server collects video data in real time from the installed security cameras. The video data is continuously divided into frames and sent as data for analysis by the AI ​​agent.

[0236] Step 2:

[0237] The AI ​​agent on the server uses an image recognition algorithm to detect objects in each received frame. These objects include people and specific objects.

[0238] Step 3:

[0239] The AI ​​agent automatically overlays a difficult-to-identify alternative representation onto the detected object. This alternative representation is created, for example, using an avatar or special filters.

[0240] Step 4:

[0241] The server records video data with alternative displays applied while maintaining real-time data for monitoring specific behaviors. At this stage, data is generated in a way that ensures privacy.

[0242] Step 5:

[0243] The AI ​​agent continuously monitors the movement of the target object and determines whether there is any abnormal behavior. Abnormal behavior is defined by comparing it to a pre-defined normal behavior pattern.

[0244] Step 6:

[0245] When abnormal behavior is detected, the server immediately sends a notification to the administrator's terminal. The notification includes the time, location, and nature of the abnormal behavior.

[0246] Step 7:

[0247] The device receives the notification and provides information for the administrator to review the details of the anomaly and take appropriate action if necessary.

[0248] Step 8:

[0249] Under certain circumstances or court orders, the server will deactivate the alternative display, allowing access to the original video data. The process of deactivating the alternative display is performed only under strict access control.

[0250] (Example 1)

[0251] Next, we will describe Example 1. 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."

[0252] The goal is to create a security system that ensures public safety while protecting privacy. Conventional video surveillance systems make it easy to identify individuals, leading to privacy concerns. Furthermore, detecting abnormal behavior from large amounts of video data relies on human intervention, limiting efficiency and accuracy.

[0253] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0254] In this invention, the server includes means for acquiring video information and detecting a target from the video information, means for overlaying and displaying an alternative display that is difficult to identify onto the detected target, and means for managing the behavior of the target and detecting abnormal behavior. This makes it possible to automatically detect abnormal behavior efficiently and accurately while protecting privacy.

[0255] "Video information" refers to video data acquired by cameras and other imaging devices.

[0256] "Target" refers to a specific object, such as a person or thing, detected within the video information.

[0257] "Alternative representation" refers to avatars or graphical display formats that are superimposed on a subject to make it difficult to identify the individual.

[0258] "Abnormal operation" refers to behavior that deviates from a pre-set normal operating pattern.

[0259] "Notification" refers to a message sent to the administrator to convey relevant information when abnormal operation is detected.

[0260] A "machine learning model" refers to the technology that forms the basis of algorithms and programs used in the analysis of video information.

[0261] "Identification information" refers to information that includes characteristics used to recognize a specific individual or object.

[0262] This invention is a security system designed to ensure public safety while protecting privacy. The system operates primarily from a server and uses video information acquired from security cameras to detect targets and monitor for abnormal activity in real time.

[0263] Server role:

[0264] The server first receives video information acquired from security cameras, and an AI agent immediately performs analysis. This analysis uses machine learning models such as TensorFlow and PyTorch. The AI ​​agent detects objects such as people and objects from the video information and overlays difficult-to-identify alternative displays onto them using libraries such as OpenCV. These alternative displays are designed to conceal personal identification information and protect privacy.

[0265] Detection and notification of abnormal operation:

[0266] The server detects abnormal behavior by comparing pre-configured behavioral patterns with real-time actions. If abnormal behavior is detected, a notification is immediately sent to the administrator's terminal. The notification includes the location, time, and specific actions taken when the abnormality occurred.

[0267] Specific example:

[0268] For example, if security cameras are installed in a shopping mall, the server monitors each entrance and exit within the mall and tracks the movements of visitors. If an AI agent determines that the same individual is frequently entering and exiting the mall within a specific time period is abnormal, it notifies the administrator with video footage overlaid with alternative displays. Meanwhile, users can use the mall in a safe environment, feeling secure knowing that their privacy is protected by this system.

[0269] Example of a prompt:

[0270] "Design an AI program for a shopping mall's security camera system that detects behavioral patterns where the same person enters and exits the building multiple times within a specific period. This program should use avatars to make it difficult to identify individuals in the video footage, and should notify administrators when abnormal behavior is detected."

[0271] This system is designed to adapt to diverse environments, allowing for flexible configuration of camera placement, alternative display formats, and criteria for abnormal behavior, thereby achieving efficient security.

[0272] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0273] Step 1:

[0274] The server acquires video data from security cameras in real time. The input is the video data transmitted from the cameras, which is a raw video stream. The server captures this data and temporarily stores it to prepare for subsequent processing. This step also involves converting the video data into a format suitable for capture.

[0275] Step 2:

[0276] The server sends the acquired video data to the AI ​​agent and begins analysis. The input is the video data acquired in step 1. The AI ​​agent uses a machine learning model to detect objects in the video. Specifically, it uses image processing algorithms to identify people and objects frame by frame. This analysis outputs a list of specific objects present in the video.

[0277] Step 3:

[0278] The server applies a difficult-to-identify alternative display to the objects detected by the AI ​​agent. The input is a list of objects detected in step 2 and their location information. The server uses libraries such as OpenCV to overlay avatars or graphical displays on the objects. This output is a video stream with personal information protected. Specifically, for example, processing is performed such as overlaying an anime character mask on the faces of people.

[0279] Step 4:

[0280] The server monitors the target's actions within the video data and checks for abnormal behavior. The input consists of an anonymized video stream and the analysis data from step 2. The server compares this to a pre-configured normal behavior pattern and uses a machine learning algorithm to detect anomalies. If abnormal behavior is detected, detailed information about it is output.

[0281] Step 5:

[0282] When an abnormal operation is detected in the server, a notification is sent to the administrator's terminal. The input is the detailed information of the abnormal operation output in step 4. Based on this information, the server generates a notification message and sends it to the terminal through email or a dedicated application. The output is a notification message for the administrator, including the specific content, location, and time of the abnormality.

[0283] Step 6:

[0284] The administrator's terminal receives the notification and displays it on the screen. The terminal receives the notification message from the server as input and notifies the occurrence of an abnormality by displaying an alert on the screen. This output is a visual notification for the administrator. The administrator can thus take necessary safety measures promptly.

[0285] (Application Example 1)

[0286] Next, Application Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0287] In a modern urban environment, it is a difficult task to balance security enhancement and privacy protection. While the security camera system improves public safety, it also poses a potential risk of personal privacy infringement. To address this contradiction, it is necessary to provide a safe method that can quickly detect abnormal behavior, issue effective security notifications, and make it difficult to identify individuals on the screen.

[0288] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0289] In this invention, the server includes means for acquiring video information and detecting an object from the video information, means for visualizing the detected object by superimposing an alternative visual representation that is difficult to identify, and means for monitoring the movement of the object and detecting abnormal behavior. This enables rapid detection of abnormal behavior to ensure public safety and real-time information provision while securely protecting users' personal information.

[0290] "Visual information" refers to visual data acquired from cameras and recording devices.

[0291] "Target" refers to a person or object detected within the video information.

[0292] "Difficult-to-identify alternative visual representations" refer to visual substitutes used to make it difficult to identify an object.

[0293] "Trends" refers to the state or changes in the behavior or movement of an object.

[0294] "Abnormal behavior" refers to actions that are detected as deviating from normal patterns.

[0295] "Communication" refers to notifications and messages generated based on the detection of abnormal operation.

[0296] A "personal device" refers to a portable electronic device owned by a user.

[0297] "Visualization" refers to the process of representing information visually to aid its understanding.

[0298] This invention is a security system that balances safety and privacy. This system consists of multiple components and is configured as follows:

[0299] The server first acquires video information from surveillance cameras and recording devices. This acquired video information is analyzed in real time using AI technology. AI video analysis models such as TensorFlow and PyTorch are used for this analysis. The AI ​​agent detects objects from the video information. Alternative visual representations are overlaid on these detected objects to make identification difficult. This process protects individual privacy.

[0300] The server continuously monitors the target's activity and detects abnormal behavior. To detect abnormal behavior, a comparison is made with a pre-configured normal operating pattern. This comparison involves data stream processing in the cloud, and as soon as an anomaly is detected, a communication is immediately sent to the user's personal device. Cloud services such as AWS Lambda and Firebase Cloud Messaging may be used for this communication.

[0301] Users can receive real-time notifications through their personal devices and obtain information about safety in public places. For example, if unusual behavior is detected in a shopping mall, the user will receive a communication stating, "There is unusual activity in a specific area of ​​the mall." This allows the user to immediately understand the situation and take safe actions.

[0302] An example of a generative AI model or prompt related to this system is, "How can I develop an application that detects abnormal behavior in real time within a shopping mall and sends notifications to users?"

[0303] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0304] Step 1:

[0305] The server acquires video information from surveillance cameras and recording devices. This input data is a real-time video feed captured by the cameras. The server prepares this video information as a dataset for AI analysis and converts it into a format to be used in the next step.

[0306] Step 2:

[0307] The server detects objects from the acquired video information. At this stage, an AI video analysis model using TensorFlow or PyTorch is executed to detect objects (such as people and vehicles). The input is the prepared video information, and the output is the coordinate data and feature data of the detected objects. Alternative visual representations that are difficult to identify are overlaid on the detected objects.

[0308] Step 3:

[0309] The server monitors the movement trends of the objects and detects abnormal operations. The input is the coordinate data and feature data of the detected objects. Data processing is performed by comparing with normal operation patterns. If an abnormality is determined, the result is passed to the next step. The output is flag information indicating whether an abnormal operation has been detected.

[0310] Step 4:

[0311] If an abnormal operation is detected, the server uses cloud services to send a communication to the user's personal terminal. The input is the flag information of the abnormal operation and the location information of the object. The output is a notification sent through a platform such as Firebase Cloud Messaging. This notification describes where and what abnormality has occurred.

[0312] Step 5:

[0313] The user's device receives notifications from the server and displays the information visually. The device analyzes this input data and generates a screen that notifies the user of warnings and recommended actions. The output is detailed information about the abnormal behavior displayed on the device's screen.

[0314] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0315] This invention provides a system that achieves more advanced crime prevention and abnormal behavior detection by recognizing emotions while protecting user privacy. This system incorporates an emotion engine that recognizes user emotions in conjunction with analyzing video data from security cameras. The emotion engine uses image analysis technology to analyze the user's facial expressions and movements, and evaluates their emotional state in real time. This evaluation result is used to determine abnormal behavior and apply alternative displays.

[0316] System implementation example:

[0317] The server first collects video data from security cameras and uses an AI agent to detect people. A substitute image is then overlaid on the detected individuals to make identification difficult. This substitute image, for example, covers the person's appearance as an avatar, thus protecting their privacy.

[0318] Furthermore, the emotion engine analyzes the person's facial expressions and movements, recognizing emotional states such as joy, anger, sadness, and happiness in real time. These recognized emotions contribute to the identification of abnormal behavior. For example, even if the behavior is normal, if feelings of anxiety or fear are recognized, it is logged by the server as a type of abnormal behavior, and a notification is sent to the terminal.

[0319] As a concrete example, consider a security system within a train station. This system constantly monitors passenger movements and analyzes their behavior. If a passenger frequently exhibits unstable emotions according to the emotion engine, the server determines this behavior is abnormal by comparing it to pre-configured behavioral patterns and sends a detailed notification to the administrator's terminal. It is also possible to dynamically change alternative displays based on changes in emotion. For example, if fear persists, a stronger alternative display can be applied to conceal the fear.

[0320] By utilizing an emotion engine, this system goes beyond simple motion detection, enabling appropriate responses tailored to individual emotions and achieving a more accurate crime prevention system. This dramatically improves public safety while simultaneously ensuring thorough consideration of user privacy.

[0321] The following describes the processing flow.

[0322] Step 1:

[0323] The server acquires video data from security cameras in real time. The video data is divided into frames and sent to the AI ​​agent in a format that is easy to analyze.

[0324] Step 2:

[0325] The AI ​​agent on the server processes the received frames based on image recognition technology to detect people and objects in the video. In the process, it extracts their location information and movement data.

[0326] Step 3:

[0327] Each detected object is overlaid with a difficult-to-identify alternative representation. The server's alternative representation means cover the object with avatars or filters to protect individual privacy.

[0328] Step 4:

[0329] The server also plays a role in continuously monitoring the movement of the target object and detecting abnormal behavior. It collects patterns of movement, and if they deviate from pre-configured normal behavior patterns, they are marked as abnormal behavior.

[0330] Step 5:

[0331] The emotion engine analyzes the facial expressions and movements of people in the video and evaluates their emotional state. The server uses this information to further evaluate abnormal behavior in detail.

[0332] Step 6:

[0333] If behavior is deemed abnormal based on the emotional state, the server sends a notification to the administrator's terminal. The notification includes the type of abnormal behavior, the time it occurred, and emotional information.

[0334] Step 7:

[0335] The terminal processes notifications from the server and displays relevant video and emotion data to enable administrators to respond quickly. Based on this display, it helps determine whether or not to take action on-site.

[0336] Step 8:

[0337] If necessary, the server will activate the alternative display deactivation function, providing access to the original video under certain conditions. For example, in legal requests or emergencies, the video can be viewed with all privacy restrictions removed.

[0338] (Example 2)

[0339] Next, we will describe Example 2. 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".

[0340] Conventional security systems have difficulty detecting abnormal behavior without identifying individuals in the video, posing a challenge in accurately detecting anomalies while protecting individual privacy. Furthermore, abnormal behavior is judged based solely on actions and does not take emotional states into consideration, leading to the possibility of false positives.

[0341] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0342] In this invention, the server includes means for acquiring video data and detecting a target from that data, means for overlaying a substitute display that is difficult to identify onto the detected target, and means for analyzing the emotional state of the target and detecting abnormal behavior based on that emotional state. This makes it possible to accurately detect abnormal behavior that takes emotional state into account while protecting the privacy of individuals.

[0343] "Video data" refers to visual information represented in digital format, acquired by devices such as cameras for security and surveillance purposes.

[0344] A "subject" is something that exhibits a specific action or characteristic within the video data and is detected as a person or object.

[0345] "Alternative displays" are visual overlays used to protect the privacy of detected objects, and their role is to make identification difficult.

[0346] "Emotional state" refers to the internal psychological state inferred from the subject's facial expressions and actions, and includes sensory reactions such as joy, anger, sadness, and happiness.

[0347] "Abnormal behavior" refers to movements or states that deviate from normal behavioral patterns, indicating that the person being monitored may be experiencing some kind of danger or instability.

[0348] A "notification" is information sent to the relevant parties or systems upon detection of abnormal behavior, intended to prompt appropriate action.

[0349] This invention provides a system that utilizes security cameras installed in public places to detect abnormal behavior with high accuracy through video data. The system enables advanced surveillance using sentiment analysis while protecting user privacy.

[0350] The server collects video data from security cameras in real time. The collected video data is then used to detect people via an AI agent. In this case, commonly used image recognition technologies include TensorFlow and OpenCV.

[0351] Individuals who are detected will have alternative markers overlaid to make identification difficult. These alternative markers are provided in the form of mosaics or avatars. This ensures user privacy and provides a sense of security that they are being monitored.

[0352] Furthermore, the server uses an emotion engine to analyze a person's facial expressions and movements, and evaluates their emotional state in real time. Emotional states include specific expressions such as anger, joy, and fear.

[0353] Abnormal behavior is detected by comparing pre-configured behavioral patterns with analyzed emotional states. Even normal actions are considered abnormal if emotions such as anxiety or fear are recognized, and are logged on the server. If the state of the target is determined to be abnormal, a notification is sent to the terminal.

[0354] A concrete example of its use is in a security system within a train station. This system has the function of monitoring and constantly analyzing passengers' behavior and emotions. If a passenger frequently exhibits unstable emotions, the emotion engine will determine this to be an anomaly and provide a notification with details to the administrator's terminal.

[0355] Examples of prompts for the generative AI model include: "Please describe a system that identifies emotions from video data. How does this system detect abnormal behavior while protecting privacy?" and "Please give specific examples of the role and practical applications of emotion engines in security cameras."

[0356] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0357] Step 1:

[0358] The server collects video data from security cameras in real time. The input is the camera stream. This data is converted to a digital format and stored within the server. Specifically, the video is formatted for processing frame by frame, and pre-processing is performed to remove noise. The output is video data in an analyzable format.

[0359] Step 2:

[0360] The server uses an AI agent to detect people from video data. The input here is the processed video data from the previous step. A person detection model (e.g., HOG + SVM using OpenCV, or a deep learning model using TensorFlow) is applied to this data. As a result, the output is data showing the location and size of the detected people.

[0361] Step 3:

[0362] The server overlays a substitute image onto the detected person. The input is the person's location data obtained in step 2. Using image processing techniques, mosaics or avatars are overlaid on the person's face and body to make identification difficult. This process generates edited frames. The output is privacy-protected video data.

[0363] Step 4:

[0364] The server uses an emotion engine to analyze emotions from privacy-protected video data. The data from step 3 is used as input. The engine analyzes the subject's facial expressions and movements and applies image recognition and motion analysis algorithms to identify specific emotions (such as joy, anger, or anxiety). As a result, the output is obtained as the analyzed emotional state.

[0365] Step 5:

[0366] The server detects abnormal behavior based on the emotion analysis results. The input consists of emotional state and behavior pattern data obtained in step 4. The server compares the emotional state with a pre-configured normal behavior pattern to determine if there is an abnormality. The output is notification data if abnormal behavior is detected.

[0367] Step 6:

[0368] The server sends a notification to the terminal when an anomaly is detected. The input includes the notification data generated in step 5. From this data, a message is created that describes the specific abnormal behavior in detail. The output is a notification message displayed on the user's or administrator's terminal. This notification allows for prompt action to be taken as needed.

[0369] (Application Example 2)

[0370] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0371] In modern surveillance systems, balancing privacy protection and security is a crucial challenge. Conventional technologies only monitor the movement of objects to detect anomalies, making it difficult to detect abnormal behavior accompanied by unstable emotional states early. Furthermore, methods using alternative displays to protect personal information in video footage lack the flexibility to adapt to different situations. This has made it difficult to achieve both improved public safety and protection of individual privacy.

[0372] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0373] In this invention, the server includes means for acquiring video data and detecting an object from the video data, means for overlaying and displaying a difficult-to-identify alternative display on the detected object, and means for recognizing the emotional state of the object. This enables early detection of abnormal behavior based on emotional state and enhanced privacy protection according to the situation.

[0374] "Video data" refers to a collection of visual information acquired by cameras and other visual sensors.

[0375] "Object" refers to a person or object that exists within the video data and is detected based on specific conditions.

[0376] "Detection" refers to the process of identifying and distinguishing an object from video data.

[0377] "Alternative displays" are visual substitutes used to conceal personal information or privacy-related elements in the original video.

[0378] "Emotional state" refers to the psychological elements and characteristics recognized to express human emotions.

[0379] "Abnormal behavior" refers to actions that have characteristics different from the norm and require attention in monitoring systems.

[0380] A "notification" is a message generated by a system to inform about a specific event or state.

[0381] "Dynamic modification" means changing the display or behavior in response to changes in circumstances or parameters.

[0382] A "pre-configured behavioral pattern" is a pre-programmed behavioral profile used as a criterion for detecting abnormal behavior.

[0383] The system for implementing this invention is a security system that uses video data to detect objects and recognize their emotional states. The server acquires video data using cameras and other visual sensors and detects objects from it. Image processing libraries such as OpenCV are used for object detection.

[0384] The server uses an emotion engine powered by a generative AI model to analyze the emotional state of objects in real time from video data. This allows the emotion engine to recognize emotions such as joy, anger, sadness, and happiness, and detect abnormal behavior based on this data. This information is compared with pre-configured behavioral and emotional patterns, and if an anomaly is detected, a notification is sent to the user's device.

[0385] To protect user privacy, the server overlays a substitute display on detected objects. This substitute display is processed using classes such as PrivacyOverlay and is dynamically modified depending on the context. This dynamic modification allows for adjustment of its intensity based on the perceived emotional state.

[0386] As a concrete example, a home camera system may have a function that recognizes the emotions of family members and notifies them if they are experiencing persistent feelings of depression, suggesting that they may need support. In this case, a generative AI model is used to prompt the implementation of a specific emotion recognition algorithm.

[0387] An example of a prompt might be, "Implement an algorithm that recognizes an individual's emotions in real time while protecting their privacy when they are photographed." This prompt is used to provide generative AI models with implementation guidelines for emotion recognition and privacy protection features.

[0388] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0389] Step 1:

[0390] The server acquires video data in real time from cameras and vision sensors. This video data serves as input for processing. The input images are converted into a format that can be analyzed using OpenCV.

[0391] Step 2:

[0392] The server detects objects from the acquired video data. Here, image processing algorithms are used to identify people and objects in the video and output information about the detected objects. Contour recognition and face recognition technologies are used for image analysis.

[0393] Step 3:

[0394] The server overlays a substitute display onto the detected object. The input uses the object's location and shape data, and the PrivacyOverlay class is used to apply a substitute display that hides personal information. This process results in an output where a portion of the video is modified while protecting privacy.

[0395] Step 4:

[0396] The server recognizes the emotional state of the subject from the displayed alternative video. Using a generative AI model, it analyzes facial expressions and movements in the input video and outputs emotional data. A pre-trained emotion recognition algorithm is used to classify the emotional state.

[0397] Step 5:

[0398] The server detects abnormal behavior by comparing emotional states with previously recorded behavioral patterns. Emotional state data is used as input, and abnormal behavior is output only if it matches the conditions for being judged as abnormal. At this stage, pattern matching is performed for both emotions and behavior.

[0399] Step 6:

[0400] When abnormal behavior is detected, the server generates a notification and sends it to the designated device. The notification includes information such as the type and time of the abnormality and the emotional state, and a warning message is displayed to the user. This output is then fed into the alert system.

[0401] Step 7:

[0402] Furthermore, the alternative display is dynamically changed according to the detected emotional state. Based on the emotional data as input, the PrivacyOverlay class is executed again, and a new alternative display is applied to the video. As a result, a video that makes the user feel more secure is output.

[0403] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0404] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0405] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0406] [Third Embodiment]

[0407] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0408] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0409] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0410] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0411] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0412] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0413] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0414] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0415] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0416] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0417] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0418] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0419] This invention is implemented as an AI-powered security system to ensure public safety while protecting privacy. The core of this system lies in a technology that detects abnormal behavior through real-time video data analysis while making it difficult to identify individuals.

[0420] System implementation example:

[0421] The server acquires video data captured by security cameras in real time. The video data is immediately processed by an AI agent to detect people and objects present in the video. Predefined avatars or other graphical representations are overlaid on detected objects to make them difficult to identify. This process is performed in real time in response to the movement of the objects, thus providing continuous privacy protection.

[0422] Abnormal behavior is detected by an AI agent on the server, which compares the actual behavior with pre-trained normal behavior patterns. If abnormal behavior is detected through this comparison, the server immediately sends a notification to the administrator's terminal. The notification includes the location, time, and specific details of the abnormal behavior.

[0423] For example, suppose there is a camera system installed in a shopping mall. The server monitors the entrances and exits of each store in the mall and tracks people coming and going. If the system detects the same person repeatedly entering and exiting a particular area during a specific time period, this behavior is considered abnormal and is deemed unusual. An AI agent detects this anomaly and notifies the administrator of the anomaly while concealing the movements of the person in the relevant video footage with an avatar.

[0424] For users, i.e., visitors to the mall, the fact that these security cameras are operated without infringing on their privacy can be a source of reassurance. On the other hand, for administrators who operate the servers and terminals, they are a powerful tool for continuous and efficient security.

[0425] This system can be customized according to the usage environment, and camera placement, types of alternative displays, and criteria for abnormal behavior are set during system installation. This allows for flexible adaptation to facilities of various sizes and types.

[0426] The following describes the processing flow.

[0427] Step 1:

[0428] The server collects video data in real time from the installed security cameras. The video data is continuously divided into frames and sent as data for analysis by the AI ​​agent.

[0429] Step 2:

[0430] The AI ​​agent on the server uses an image recognition algorithm to detect objects in each received frame. These objects include people and specific objects.

[0431] Step 3:

[0432] The AI ​​agent automatically overlays a difficult-to-identify alternative representation onto the detected object. This alternative representation is created, for example, using an avatar or special filters.

[0433] Step 4:

[0434] The server records video data with alternative displays applied while maintaining real-time data for monitoring specific behaviors. At this stage, data is generated in a way that ensures privacy.

[0435] Step 5:

[0436] The AI ​​agent continuously monitors the movement of the target object and determines whether there is any abnormal behavior. Abnormal behavior is defined by comparing it to a pre-defined normal behavior pattern.

[0437] Step 6:

[0438] When abnormal behavior is detected, the server immediately sends a notification to the administrator's terminal. The notification includes the time, location, and nature of the abnormal behavior.

[0439] Step 7:

[0440] The device receives the notification and provides information for the administrator to review the details of the anomaly and take appropriate action if necessary.

[0441] Step 8:

[0442] Under certain circumstances or court orders, the server will deactivate the alternative display, allowing access to the original video data. The process of deactivating the alternative display is performed only under strict access control.

[0443] (Example 1)

[0444] Next, we will describe Example 1. 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."

[0445] The goal is to create a security system that ensures public safety while protecting privacy. Conventional video surveillance systems make it easy to identify individuals, leading to privacy concerns. Furthermore, detecting abnormal behavior from large amounts of video data relies on human intervention, limiting efficiency and accuracy.

[0446] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0447] In this invention, the server includes means for acquiring video information and detecting a target from the video information, means for overlaying and displaying an alternative display that is difficult to identify onto the detected target, and means for managing the behavior of the target and detecting abnormal behavior. This makes it possible to automatically detect abnormal behavior efficiently and accurately while protecting privacy.

[0448] "Video information" refers to video data acquired by cameras and other imaging devices.

[0449] "Target" refers to a specific object, such as a person or thing, detected within the video information.

[0450] "Alternative representation" refers to avatars or graphical display formats that are superimposed on a subject to make it difficult to identify the individual.

[0451] "Abnormal operation" refers to behavior that deviates from a pre-set normal operating pattern.

[0452] "Notification" refers to a message sent to the administrator to convey relevant information when abnormal operation is detected.

[0453] A "machine learning model" refers to the technology that forms the basis of algorithms and programs used in the analysis of video information.

[0454] "Identification information" refers to information that includes characteristics used to recognize a specific individual or object.

[0455] This invention is a security system designed to ensure public safety while protecting privacy. The system operates primarily from a server and uses video information acquired from security cameras to detect targets and monitor for abnormal activity in real time.

[0456] Server role:

[0457] The server first receives video information acquired from security cameras, and an AI agent immediately performs analysis. This analysis uses machine learning models such as TensorFlow and PyTorch. The AI ​​agent detects objects such as people and objects from the video information and overlays difficult-to-identify alternative displays onto them using libraries such as OpenCV. These alternative displays are designed to conceal personal identification information and protect privacy.

[0458] Detection and notification of abnormal operation:

[0459] The server detects abnormal behavior by comparing pre-configured behavioral patterns with real-time actions. If abnormal behavior is detected, a notification is immediately sent to the administrator's terminal. The notification includes the location, time, and specific actions taken when the abnormality occurred.

[0460] Specific example:

[0461] For example, if security cameras are installed in a shopping mall, the server monitors each entrance and exit within the mall and tracks the movements of visitors. If an AI agent determines that the same individual is frequently entering and exiting the mall within a specific time period is abnormal, it notifies the administrator with video footage overlaid with alternative displays. Meanwhile, users can use the mall in a safe environment, feeling secure knowing that their privacy is protected by this system.

[0462] Example of a prompt:

[0463] "Design an AI program for a shopping mall's security camera system that detects behavioral patterns where the same person enters and exits the building multiple times within a specific period. This program should use avatars to make it difficult to identify individuals in the video footage, and should notify administrators when abnormal behavior is detected."

[0464] This system is designed to adapt to diverse environments, allowing for flexible configuration of camera placement, alternative display formats, and criteria for abnormal behavior, thereby achieving efficient security.

[0465] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0466] Step 1:

[0467] The server acquires video data from security cameras in real time. The input is the video data transmitted from the cameras, which is a raw video stream. The server captures this data and temporarily stores it to prepare for subsequent processing. This step also involves converting the video data into a format suitable for capture.

[0468] Step 2:

[0469] The server sends the acquired video data to the AI ​​agent and begins analysis. The input is the video data acquired in step 1. The AI ​​agent uses a machine learning model to detect objects in the video. Specifically, it uses image processing algorithms to identify people and objects frame by frame. This analysis outputs a list of specific objects present in the video.

[0470] Step 3:

[0471] The server applies a difficult-to-identify alternative display to the objects detected by the AI ​​agent. The input is a list of objects detected in step 2 and their location information. The server uses libraries such as OpenCV to overlay avatars or graphical displays on the objects. This output is a video stream with personal information protected. Specifically, for example, processing is performed such as overlaying an anime character mask on the faces of people.

[0472] Step 4:

[0473] The server monitors the target's actions within the video data and checks for abnormal behavior. The input consists of an anonymized video stream and the analysis data from step 2. The server compares this to a pre-configured normal behavior pattern and uses a machine learning algorithm to detect anomalies. If abnormal behavior is detected, detailed information about it is output.

[0474] Step 5:

[0475] If the server detects abnormal behavior, it sends a notification to the administrator's terminal. The input is the detailed information of the abnormal behavior output in step 4. Based on this information, the server generates a notification message and sends it to the terminal via email or a dedicated application. The output is a notification message for the administrator and includes the specific nature, location, and time of the abnormality.

[0476] Step 6:

[0477] The administrator's terminal receives the notification and displays it on the screen. The terminal receives the notification message from the server as input and displays an alert on the screen to notify the administrator of the occurrence of an anomaly. This output is a visual notification to the administrator, allowing them to quickly take necessary security measures.

[0478] (Application Example 1)

[0479] Next, we will explain Application Example 1. In the following explanation, 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."

[0480] In modern urban environments, balancing enhanced security with privacy protection is a challenging task. While security camera systems improve public safety, they also carry the potential to infringe on individual privacy. To address this contradiction, there is a need to provide a secure method that makes it difficult to identify individuals on screen while rapidly detecting abnormal behavior and providing effective security notifications.

[0481] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0482] In this invention, the server includes means for acquiring video information and detecting an object from the video information, means for visualizing the detected object by superimposing an alternative visual representation that is difficult to identify, and means for monitoring the movement of the object and detecting abnormal behavior. This enables rapid detection of abnormal behavior to ensure public safety and real-time information provision while securely protecting users' personal information.

[0483] "Visual information" refers to visual data acquired from cameras and recording devices.

[0484] "Target" refers to a person or object detected within the video information.

[0485] "Difficult-to-identify alternative visual representations" refer to visual substitutes used to make it difficult to identify an object.

[0486] "Trends" refers to the state or changes in the behavior or movement of an object.

[0487] "Abnormal behavior" refers to actions that are detected as deviating from normal patterns.

[0488] "Communication" refers to notifications and messages generated based on the detection of abnormal operation.

[0489] A "personal device" refers to a portable electronic device owned by a user.

[0490] "Visualization" refers to the process of representing information visually to aid its understanding.

[0491] This invention is a security system that balances safety and privacy. This system consists of multiple components and is configured as follows:

[0492] The server first acquires video information from surveillance cameras and recording devices. This acquired video information is analyzed in real time using AI technology. AI video analysis models such as TensorFlow and PyTorch are used for this analysis. The AI ​​agent detects objects from the video information. Alternative visual representations are overlaid on these detected objects to make identification difficult. This process protects individual privacy.

[0493] The server continuously monitors the target's activity and detects abnormal behavior. To detect abnormal behavior, a comparison is made with a pre-configured normal operating pattern. This comparison involves data stream processing in the cloud, and as soon as an anomaly is detected, a communication is immediately sent to the user's personal device. Cloud services such as AWS Lambda and Firebase Cloud Messaging may be used for this communication.

[0494] Users can receive real-time notifications through their personal devices and obtain information about safety in public places. For example, if unusual behavior is detected in a shopping mall, the user will receive a communication stating, "There is unusual activity in a specific area of ​​the mall." This allows the user to immediately understand the situation and take safe actions.

[0495] An example of a generative AI model or prompt related to this system is, "How can I develop an application that detects abnormal behavior in real time within a shopping mall and sends notifications to users?"

[0496] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0497] Step 1:

[0498] The server acquires video information from surveillance cameras and recording devices. This input data is a real-time video feed captured by the cameras. The server prepares this video information as a dataset for AI analysis and converts it to a format to be used in the next step.

[0499] Step 2:

[0500] The server detects objects from the acquired video information. At this stage, an AI video analysis model using TensorFlow or PyTorch is executed to detect objects (such as people or vehicles). The input is the prepared video information, and the output is the coordinate data and feature data of the detected objects. Alternative visual representations that are difficult to identify are superimposed on the detected objects.

[0501] Step 3:

[0502] The server monitors the target's movements and detects abnormal behavior. The input consists of the detected target's coordinate and feature data. Data processing is performed by comparing the data with normal behavior patterns. If an anomaly is detected, the result is passed to the next step. The output is flag information indicating whether abnormal behavior was detected.

[0503] Step 4:

[0504] If abnormal behavior is detected, the server uses a cloud service to send a communication to the user's personal device. The input is flag information about the abnormal behavior and the location of the affected device. The output is a notification sent through a platform such as Firebase Cloud Messaging. This notification will describe the area where the abnormality occurred and what kind of abnormality it is.

[0505] Step 5:

[0506] The user's device receives notifications from the server and displays the information visually. The device analyzes this input data and generates a screen that notifies the user of warnings and recommended actions. The output is detailed information about the abnormal behavior displayed on the device's screen.

[0507] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0508] This invention provides a system that achieves more advanced crime prevention and abnormal behavior detection by recognizing emotions while protecting user privacy. This system incorporates an emotion engine that recognizes user emotions in conjunction with analyzing video data from security cameras. The emotion engine uses image analysis technology to analyze the user's facial expressions and movements, and evaluates their emotional state in real time. This evaluation result is used to determine abnormal behavior and apply alternative displays.

[0509] System implementation example:

[0510] The server first collects video data from security cameras and uses an AI agent to detect people. A substitute image is then overlaid on the detected individuals to make identification difficult. This substitute image, for example, covers the person's appearance as an avatar, thus protecting their privacy.

[0511] Furthermore, the emotion engine analyzes the person's facial expressions and movements, recognizing emotional states such as joy, anger, sadness, and happiness in real time. These recognized emotions contribute to the identification of abnormal behavior. For example, even if the behavior is normal, if feelings of anxiety or fear are recognized, it is logged by the server as a type of abnormal behavior, and a notification is sent to the terminal.

[0512] As a concrete example, consider a security system within a train station. This system constantly monitors passenger movements and analyzes their behavior. If a passenger frequently exhibits unstable emotions according to the emotion engine, the server determines this behavior is abnormal by comparing it to pre-configured behavioral patterns and sends a detailed notification to the administrator's terminal. It is also possible to dynamically change alternative displays based on changes in emotion. For example, if fear persists, a stronger alternative display can be applied to conceal the fear.

[0513] By utilizing an emotion engine, this system goes beyond simple motion detection, enabling appropriate responses tailored to individual emotions and achieving a more accurate crime prevention system. This dramatically improves public safety while simultaneously ensuring thorough consideration of user privacy.

[0514] The following describes the processing flow.

[0515] Step 1:

[0516] The server acquires video data from security cameras in real time. The video data is divided into frames and sent to the AI ​​agent in a format that is easy to analyze.

[0517] Step 2:

[0518] The AI ​​agent on the server processes the received frames based on image recognition technology to detect people and objects in the video. In the process, it extracts their location information and movement data.

[0519] Step 3:

[0520] Each detected object is overlaid with a difficult-to-identify alternative representation. The server's alternative representation means cover the object with avatars or filters to protect individual privacy.

[0521] Step 4:

[0522] The server also plays a role in continuously monitoring the movement of the target object and detecting abnormal behavior. It collects patterns of movement, and if they deviate from pre-configured normal behavior patterns, they are marked as abnormal behavior.

[0523] Step 5:

[0524] The emotion engine analyzes the facial expressions and movements of people in the video and evaluates their emotional state. The server uses this information to further evaluate abnormal behavior in detail.

[0525] Step 6:

[0526] If behavior is deemed abnormal based on the emotional state, the server sends a notification to the administrator's terminal. The notification includes the type of abnormal behavior, the time it occurred, and emotional information.

[0527] Step 7:

[0528] The terminal processes notifications from the server and displays relevant video and emotion data to enable administrators to respond quickly. Based on this display, it helps determine whether or not to take action on-site.

[0529] Step 8:

[0530] If necessary, the server will activate the alternative display deactivation function, providing access to the original video under certain conditions. For example, in legal requests or emergencies, the video can be viewed with all privacy restrictions removed.

[0531] (Example 2)

[0532] Next, we will describe Example 2. 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."

[0533] Conventional security systems have difficulty detecting abnormal behavior without identifying individuals in the video, posing a challenge in accurately detecting anomalies while protecting individual privacy. Furthermore, abnormal behavior is judged based solely on actions and does not take emotional states into consideration, leading to the possibility of false positives.

[0534] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0535] In this invention, the server includes means for acquiring video data and detecting a target from that data, means for overlaying a substitute display that is difficult to identify onto the detected target, and means for analyzing the emotional state of the target and detecting abnormal behavior based on that emotional state. This makes it possible to accurately detect abnormal behavior that takes emotional state into account while protecting the privacy of individuals.

[0536] "Video data" refers to visual information represented in digital format, acquired by devices such as cameras for security and surveillance purposes.

[0537] A "subject" is something that exhibits a specific action or characteristic within the video data and is detected as a person or object.

[0538] "Alternative displays" are visual overlays used to protect the privacy of detected objects, and their role is to make identification difficult.

[0539] "Emotional state" refers to the internal psychological state inferred from the subject's facial expressions and actions, and includes sensory reactions such as joy, anger, sadness, and happiness.

[0540] "Abnormal behavior" refers to movements or states that deviate from normal behavioral patterns, indicating that the person being monitored may be experiencing some kind of danger or instability.

[0541] A "notification" is information sent to the relevant parties or systems upon detection of abnormal behavior, intended to prompt appropriate action.

[0542] This invention provides a system that utilizes security cameras installed in public places to detect abnormal behavior with high accuracy through video data. The system enables advanced surveillance using sentiment analysis while protecting user privacy.

[0543] The server collects video data from security cameras in real time. The collected video data is then used to detect people via an AI agent. In this case, commonly used image recognition technologies include TensorFlow and OpenCV.

[0544] Individuals who are detected will have alternative markers overlaid to make identification difficult. These alternative markers are provided in the form of mosaics or avatars. This ensures user privacy and provides a sense of security that they are being monitored.

[0545] Furthermore, the server uses an emotion engine to analyze a person's facial expressions and movements, and evaluates their emotional state in real time. Emotional states include specific expressions such as anger, joy, and fear.

[0546] Abnormal behavior is detected by comparing pre-configured behavioral patterns with analyzed emotional states. Even normal actions are considered abnormal if emotions such as anxiety or fear are recognized, and are logged on the server. If the state of the target is determined to be abnormal, a notification is sent to the terminal.

[0547] A concrete example of its use is in a security system within a train station. This system has the function of monitoring and constantly analyzing passengers' behavior and emotions. If a passenger frequently exhibits unstable emotions, the emotion engine will determine this to be an anomaly and provide a notification with details to the administrator's terminal.

[0548] Examples of prompts for the generative AI model include: "Please describe a system that identifies emotions from video data. How does this system detect abnormal behavior while protecting privacy?" and "Please give specific examples of the role and practical applications of emotion engines in security cameras."

[0549] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0550] Step 1:

[0551] The server collects video data from security cameras in real time. The input is the camera stream. This data is converted to a digital format and stored within the server. Specifically, the video is formatted for processing frame by frame, and pre-processing is performed to remove noise. The output is video data in an analyzable format.

[0552] Step 2:

[0553] The server uses an AI agent to detect people from video data. The input here is the processed video data from the previous step. A person detection model (e.g., HOG + SVM using OpenCV, or a deep learning model using TensorFlow) is applied to this data. As a result, the output is data showing the location and size of the detected people.

[0554] Step 3:

[0555] The server overlays a substitute image onto the detected person. The input is the person's location data obtained in step 2. Using image processing techniques, mosaics or avatars are overlaid on the person's face and body to make identification difficult. This process generates edited frames. The output is privacy-protected video data.

[0556] Step 4:

[0557] The server uses an emotion engine to analyze emotions from privacy-protected video data. The data from step 3 is used as input. The engine analyzes the subject's facial expressions and movements and applies image recognition and motion analysis algorithms to identify specific emotions (such as joy, anger, or anxiety). As a result, the output is obtained as the analyzed emotional state.

[0558] Step 5:

[0559] The server detects abnormal behavior based on the emotion analysis results. The input consists of emotional state and behavior pattern data obtained in step 4. The server compares the emotional state with a pre-configured normal behavior pattern to determine if there is an abnormality. The output is notification data if abnormal behavior is detected.

[0560] Step 6:

[0561] The server sends a notification to the terminal when an anomaly is detected. The input includes the notification data generated in step 5. From this data, a message is created that describes the specific abnormal behavior in detail. The output is a notification message displayed on the user's or administrator's terminal. This notification allows for prompt action to be taken as needed.

[0562] (Application Example 2)

[0563] Next, we will explain application example 2. In the following explanation, 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."

[0564] In modern surveillance systems, balancing privacy protection and security is a crucial challenge. Conventional technologies only monitor the movement of objects to detect anomalies, making it difficult to detect abnormal behavior accompanied by unstable emotional states early. Furthermore, methods using alternative displays to protect personal information in video footage lack the flexibility to adapt to different situations. This has made it difficult to achieve both improved public safety and protection of individual privacy.

[0565] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0566] In this invention, the server includes means for acquiring video data and detecting an object from the video data, means for overlaying and displaying a difficult-to-identify alternative display on the detected object, and means for recognizing the emotional state of the object. This enables early detection of abnormal behavior based on emotional state and enhanced privacy protection according to the situation.

[0567] "Video data" refers to a collection of visual information acquired by cameras and other visual sensors.

[0568] "Object" refers to a person or object that exists within the video data and is detected based on specific conditions.

[0569] "Detection" refers to the process of identifying and distinguishing an object from video data.

[0570] "Alternative displays" are visual substitutes used to conceal personal information or privacy-related elements in the original video.

[0571] "Emotional state" refers to the psychological elements and characteristics recognized to express human emotions.

[0572] "Abnormal behavior" refers to actions that have characteristics different from the norm and require attention in monitoring systems.

[0573] A "notification" is a message generated by a system to inform about a specific event or state.

[0574] "Dynamic modification" means changing the display or behavior in response to changes in circumstances or parameters.

[0575] A "pre-configured behavioral pattern" is a pre-programmed behavioral profile used as a criterion for detecting abnormal behavior.

[0576] The system for implementing this invention is a security system that uses video data to detect objects and recognize their emotional states. The server acquires video data using cameras and other visual sensors and detects objects from it. Image processing libraries such as OpenCV are used for object detection.

[0577] The server uses an emotion engine powered by a generative AI model to analyze the emotional state of objects in real time from video data. This allows the emotion engine to recognize emotions such as joy, anger, sadness, and happiness, and detect abnormal behavior based on this data. This information is compared with pre-configured behavioral and emotional patterns, and if an anomaly is detected, a notification is sent to the user's device.

[0578] To protect user privacy, the server overlays a substitute display on detected objects. This substitute display is processed using classes such as PrivacyOverlay and is dynamically modified depending on the context. This dynamic modification allows for adjustment of its intensity based on the perceived emotional state.

[0579] As a concrete example, a home camera system may have a function that recognizes the emotions of family members and notifies them if they are experiencing persistent feelings of depression, suggesting that they may need support. In this case, a generative AI model is used to prompt the implementation of a specific emotion recognition algorithm.

[0580] An example of a prompt might be, "Implement an algorithm that recognizes an individual's emotions in real time while protecting their privacy when they are photographed." This prompt is used to provide generative AI models with implementation guidelines for emotion recognition and privacy protection features.

[0581] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0582] Step 1:

[0583] The server acquires video data in real time from cameras and vision sensors. This video data serves as input for processing. The input images are converted into a format that can be analyzed using OpenCV.

[0584] Step 2:

[0585] The server detects objects from the acquired video data. Here, image processing algorithms are used to identify people and objects in the video and output information about the detected objects. Contour recognition and face recognition technologies are used for image analysis.

[0586] Step 3:

[0587] The server overlays a substitute display onto the detected object. The input uses the object's location and shape data, and the PrivacyOverlay class is used to apply a substitute display that hides personal information. This process results in an output where a portion of the video is modified while protecting privacy.

[0588] Step 4:

[0589] The server recognizes the emotional state of the subject from the displayed alternative video. Using a generative AI model, it analyzes facial expressions and movements in the input video and outputs emotional data. A pre-trained emotion recognition algorithm is used to classify the emotional state.

[0590] Step 5:

[0591] The server detects abnormal behavior by comparing emotional states with previously recorded behavioral patterns. Emotional state data is used as input, and abnormal behavior is output only if it matches the conditions for being judged as abnormal. At this stage, pattern matching is performed for both emotions and behavior.

[0592] Step 6:

[0593] When abnormal behavior is detected, the server generates a notification and sends it to the designated device. The notification includes information such as the type and time of the abnormality and the emotional state, and a warning message is displayed to the user. This output is then fed into the alert system.

[0594] Step 7:

[0595] Furthermore, the alternative display is dynamically changed according to the detected emotional state. Based on the emotional data as input, the PrivacyOverlay class is executed again, and a new alternative display is applied to the video. As a result, a video that makes the user feel more secure is output.

[0596] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0597] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0598] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0599] [Fourth Embodiment]

[0600] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0601] As shown in Figure 7, the 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.

[0602] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0603] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0604] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0605] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0606] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0607] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive 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 robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0608] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0609] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0610] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0611] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0612] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0613] This invention is implemented as an AI-powered security system to ensure public safety while protecting privacy. The core of this system lies in a technology that detects abnormal behavior through real-time video data analysis while making it difficult to identify individuals.

[0614] System implementation example:

[0615] The server acquires video data captured by security cameras in real time. The video data is immediately processed by an AI agent to detect people and objects present in the video. Predefined avatars or other graphical representations are overlaid on detected objects to make them difficult to identify. This process is performed in real time in response to the movement of the objects, thus providing continuous privacy protection.

[0616] Abnormal behavior is detected by an AI agent on the server, which compares the actual behavior with pre-trained normal behavior patterns. If abnormal behavior is detected through this comparison, the server immediately sends a notification to the administrator's terminal. The notification includes the location, time, and specific details of the abnormal behavior.

[0617] For example, suppose there is a camera system installed in a shopping mall. The server monitors the entrances and exits of each store in the mall and tracks people coming and going. If the system detects the same person repeatedly entering and exiting a particular area during a specific time period, this behavior is considered abnormal and is deemed unusual. An AI agent detects this anomaly and notifies the administrator of the anomaly while concealing the movements of the person in the relevant video footage with an avatar.

[0618] For users, i.e., visitors to the mall, the fact that these security cameras are operated without infringing on their privacy can be a source of reassurance. On the other hand, for administrators who operate the servers and terminals, they are a powerful tool for continuous and efficient security.

[0619] This system can be customized according to the usage environment, and camera placement, types of alternative displays, and criteria for abnormal behavior are set during system installation. This allows for flexible adaptation to facilities of various sizes and types.

[0620] The following describes the processing flow.

[0621] Step 1:

[0622] The server collects video data in real time from the installed security cameras. The video data is continuously divided into frames and sent as data for analysis by the AI ​​agent.

[0623] Step 2:

[0624] The AI ​​agent on the server uses an image recognition algorithm to detect objects in each received frame. These objects include people and specific objects.

[0625] Step 3:

[0626] The AI ​​agent automatically overlays a difficult-to-identify alternative representation onto the detected object. This alternative representation is created, for example, using an avatar or special filters.

[0627] Step 4:

[0628] The server records video data with alternative displays applied while maintaining real-time data for monitoring specific behaviors. At this stage, data is generated in a way that ensures privacy.

[0629] Step 5:

[0630] The AI ​​agent continuously monitors the movement of the target object and determines whether there is any abnormal behavior. Abnormal behavior is defined by comparing it to a pre-defined normal behavior pattern.

[0631] Step 6:

[0632] When abnormal behavior is detected, the server immediately sends a notification to the administrator's terminal. The notification includes the time, location, and nature of the abnormal behavior.

[0633] Step 7:

[0634] The device receives the notification and provides information for the administrator to review the details of the anomaly and take appropriate action if necessary.

[0635] Step 8:

[0636] Under certain circumstances or court orders, the server will deactivate the alternative display, allowing access to the original video data. The process of deactivating the alternative display is performed only under strict access control.

[0637] (Example 1)

[0638] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0639] The goal is to create a security system that ensures public safety while protecting privacy. Conventional video surveillance systems make it easy to identify individuals, leading to privacy concerns. Furthermore, detecting abnormal behavior from large amounts of video data relies on human intervention, limiting efficiency and accuracy.

[0640] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0641] In this invention, the server includes means for acquiring video information and detecting a target from the video information, means for overlaying and displaying an alternative display that is difficult to identify onto the detected target, and means for managing the behavior of the target and detecting abnormal behavior. This makes it possible to automatically detect abnormal behavior efficiently and accurately while protecting privacy.

[0642] "Video information" refers to video data acquired by cameras and other imaging devices.

[0643] "Target" refers to a specific object, such as a person or thing, detected within the video information.

[0644] "Alternative representation" refers to avatars or graphical display formats that are superimposed on a subject to make it difficult to identify the individual.

[0645] "Abnormal operation" refers to behavior that deviates from a pre-set normal operating pattern.

[0646] "Notification" refers to a message sent to the administrator to convey relevant information when abnormal operation is detected.

[0647] A "machine learning model" refers to the technology that forms the basis of algorithms and programs used in the analysis of video information.

[0648] "Identification information" refers to information that includes characteristics used to recognize a specific individual or object.

[0649] This invention is a security system designed to ensure public safety while protecting privacy. The system operates primarily from a server and uses video information acquired from security cameras to detect targets and monitor for abnormal activity in real time.

[0650] Server role:

[0651] The server first receives video information acquired from security cameras, and an AI agent immediately performs analysis. This analysis uses machine learning models such as TensorFlow and PyTorch. The AI ​​agent detects objects such as people and objects from the video information and overlays difficult-to-identify alternative displays onto them using libraries such as OpenCV. These alternative displays are designed to conceal personal identification information and protect privacy.

[0652] Detection and notification of abnormal operation:

[0653] The server detects abnormal behavior by comparing pre-configured behavioral patterns with real-time actions. If abnormal behavior is detected, a notification is immediately sent to the administrator's terminal. The notification includes the location, time, and specific actions taken when the abnormality occurred.

[0654] Specific example:

[0655] For example, if security cameras are installed in a shopping mall, the server monitors each entrance and exit within the mall and tracks the movements of visitors. If an AI agent determines that the same individual is frequently entering and exiting the mall within a specific time period is abnormal, it notifies the administrator with video footage overlaid with alternative displays. Meanwhile, users can use the mall in a safe environment, feeling secure knowing that their privacy is protected by this system.

[0656] Example of a prompt:

[0657] "Design an AI program for a shopping mall's security camera system that detects behavioral patterns where the same person enters and exits the building multiple times within a specific period. This program should use avatars to make it difficult to identify individuals in the video footage, and should notify administrators when abnormal behavior is detected."

[0658] This system is designed to adapt to diverse environments, allowing for flexible configuration of camera placement, alternative display formats, and criteria for abnormal behavior, thereby achieving efficient security.

[0659] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0660] Step 1:

[0661] The server acquires video data from security cameras in real time. The input is the video data transmitted from the cameras, which is a raw video stream. The server captures this data and temporarily stores it to prepare for subsequent processing. This step also involves converting the video data into a format suitable for capture.

[0662] Step 2:

[0663] The server sends the acquired video data to the AI ​​agent and begins analysis. The input is the video data acquired in step 1. The AI ​​agent uses a machine learning model to detect objects in the video. Specifically, it uses image processing algorithms to identify people and objects frame by frame. This analysis outputs a list of specific objects present in the video.

[0664] Step 3:

[0665] The server applies a difficult-to-identify alternative display to the objects detected by the AI ​​agent. The input is a list of objects detected in step 2 and their location information. The server uses libraries such as OpenCV to overlay avatars or graphical displays on the objects. This output is a video stream with personal information protected. Specifically, for example, processing is performed such as overlaying an anime character mask on the faces of people.

[0666] Step 4:

[0667] The server monitors the target's actions within the video data and checks for abnormal behavior. The input consists of an anonymized video stream and the analysis data from step 2. The server compares this to a pre-configured normal behavior pattern and uses a machine learning algorithm to detect anomalies. If abnormal behavior is detected, detailed information about it is output.

[0668] Step 5:

[0669] If the server detects abnormal behavior, it sends a notification to the administrator's terminal. The input is the detailed information of the abnormal behavior output in step 4. Based on this information, the server generates a notification message and sends it to the terminal via email or a dedicated application. The output is a notification message for the administrator and includes the specific nature, location, and time of the abnormality.

[0670] Step 6:

[0671] The administrator's terminal receives the notification and displays it on the screen. The terminal receives the notification message from the server as input and displays an alert on the screen to notify the administrator of the occurrence of an anomaly. This output is a visual notification to the administrator, allowing them to quickly take necessary security measures.

[0672] (Application Example 1)

[0673] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0674] In modern urban environments, balancing enhanced security with privacy protection is a challenging task. While security camera systems improve public safety, they also carry the potential to infringe on individual privacy. To address this contradiction, there is a need to provide a secure method that makes it difficult to identify individuals on screen while rapidly detecting abnormal behavior and providing effective security notifications.

[0675] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0676] In this invention, the server includes means for acquiring video information and detecting an object from the video information, means for visualizing the detected object by superimposing an alternative visual representation that is difficult to identify, and means for monitoring the movement of the object and detecting abnormal behavior. This enables rapid detection of abnormal behavior to ensure public safety and real-time information provision while securely protecting users' personal information.

[0677] "Visual information" refers to visual data acquired from cameras and recording devices.

[0678] "Target" refers to a person or object detected within the video information.

[0679] "Difficult-to-identify alternative visual representations" refer to visual substitutes used to make it difficult to identify an object.

[0680] "Trends" refers to the state or changes in the behavior or movement of an object.

[0681] "Abnormal behavior" refers to actions that are detected as deviating from normal patterns.

[0682] "Communication" refers to notifications and messages generated based on the detection of abnormal operation.

[0683] A "personal device" refers to a portable electronic device owned by a user.

[0684] "Visualization" refers to the process of representing information visually to aid its understanding.

[0685] This invention is a security system that balances safety and privacy. This system consists of multiple components and is configured as follows:

[0686] The server first acquires video information from surveillance cameras and recording devices. This acquired video information is analyzed in real time using AI technology. AI video analysis models such as TensorFlow and PyTorch are used for this analysis. The AI ​​agent detects objects from the video information. Alternative visual representations are overlaid on these detected objects to make identification difficult. This process protects individual privacy.

[0687] The server continuously monitors the target's activity and detects abnormal behavior. To detect abnormal behavior, a comparison is made with a pre-configured normal operating pattern. This comparison involves data stream processing in the cloud, and as soon as an anomaly is detected, a communication is immediately sent to the user's personal device. Cloud services such as AWS Lambda and Firebase Cloud Messaging may be used for this communication.

[0688] Users can receive real-time notifications through their personal devices and obtain information about safety in public places. For example, if unusual behavior is detected in a shopping mall, the user will receive a communication stating, "There is unusual activity in a specific area of ​​the mall." This allows the user to immediately understand the situation and take safe actions.

[0689] An example of a generative AI model or prompt related to this system is, "How can I develop an application that detects abnormal behavior in real time within a shopping mall and sends notifications to users?"

[0690] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0691] Step 1:

[0692] The server acquires video information from surveillance cameras and recording devices. This input data is a real-time video feed captured by the cameras. The server prepares this video information as a dataset for AI analysis and converts it to a format to be used in the next step.

[0693] Step 2:

[0694] The server detects objects from the acquired video information. At this stage, an AI video analysis model using TensorFlow or PyTorch is executed to detect objects (such as people or vehicles). The input is the prepared video information, and the output is the coordinate data and feature data of the detected objects. Alternative visual representations that are difficult to identify are superimposed on the detected objects.

[0695] Step 3:

[0696] The server monitors the target's movements and detects abnormal behavior. The input consists of the detected target's coordinate and feature data. Data processing is performed by comparing the data with normal behavior patterns. If an anomaly is detected, the result is passed to the next step. The output is flag information indicating whether abnormal behavior was detected.

[0697] Step 4:

[0698] If abnormal behavior is detected, the server uses a cloud service to send a communication to the user's personal device. The input is flag information about the abnormal behavior and the location of the affected device. The output is a notification sent through a platform such as Firebase Cloud Messaging. This notification will describe the area where the abnormality occurred and what kind of abnormality it is.

[0699] Step 5:

[0700] The user's device receives notifications from the server and displays the information visually. The device analyzes this input data and generates a screen that notifies the user of warnings and recommended actions. The output is detailed information about the abnormal behavior displayed on the device's screen.

[0701] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0702] This invention provides a system that achieves more advanced crime prevention and abnormal behavior detection by recognizing emotions while protecting user privacy. This system incorporates an emotion engine that recognizes user emotions in conjunction with analyzing video data from security cameras. The emotion engine uses image analysis technology to analyze the user's facial expressions and movements, and evaluates their emotional state in real time. This evaluation result is used to determine abnormal behavior and apply alternative displays.

[0703] System implementation example:

[0704] The server first collects video data from security cameras and uses an AI agent to detect people. A substitute image is then overlaid on the detected individuals to make identification difficult. This substitute image, for example, covers the person's appearance as an avatar, thus protecting their privacy.

[0705] Furthermore, the emotion engine analyzes the person's facial expressions and movements, recognizing emotional states such as joy, anger, sadness, and happiness in real time. These recognized emotions contribute to the identification of abnormal behavior. For example, even if the behavior is normal, if feelings of anxiety or fear are recognized, it is logged by the server as a type of abnormal behavior, and a notification is sent to the terminal.

[0706] As a concrete example, consider a security system within a train station. This system constantly monitors passenger movements and analyzes their behavior. If a passenger frequently exhibits unstable emotions according to the emotion engine, the server determines this behavior is abnormal by comparing it to pre-configured behavioral patterns and sends a detailed notification to the administrator's terminal. It is also possible to dynamically change alternative displays based on changes in emotion. For example, if fear persists, a stronger alternative display can be applied to conceal the fear.

[0707] By utilizing an emotion engine, this system goes beyond simple motion detection, enabling appropriate responses tailored to individual emotions and achieving a more accurate crime prevention system. This dramatically improves public safety while simultaneously ensuring thorough consideration of user privacy.

[0708] The following describes the processing flow.

[0709] Step 1:

[0710] The server acquires video data from security cameras in real time. The video data is divided into frames and sent to the AI ​​agent in a format that is easy to analyze.

[0711] Step 2:

[0712] The AI ​​agent on the server processes the received frames based on image recognition technology to detect people and objects in the video. In the process, it extracts their location information and movement data.

[0713] Step 3:

[0714] Each detected object is overlaid with a difficult-to-identify alternative representation. The server's alternative representation means cover the object with avatars or filters to protect individual privacy.

[0715] Step 4:

[0716] The server also plays a role in continuously monitoring the movement of the target object and detecting abnormal behavior. It collects patterns of movement, and if they deviate from pre-configured normal behavior patterns, they are marked as abnormal behavior.

[0717] Step 5:

[0718] The emotion engine analyzes the facial expressions and movements of people in the video and evaluates their emotional state. The server uses this information to further evaluate abnormal behavior in detail.

[0719] Step 6:

[0720] If behavior is deemed abnormal based on the emotional state, the server sends a notification to the administrator's terminal. The notification includes the type of abnormal behavior, the time it occurred, and emotional information.

[0721] Step 7:

[0722] The terminal processes notifications from the server and displays relevant video and emotion data to enable administrators to respond quickly. Based on this display, it helps determine whether or not to take action on-site.

[0723] Step 8:

[0724] If necessary, the server will activate the alternative display deactivation function, providing access to the original video under certain conditions. For example, in legal requests or emergencies, the video can be viewed with all privacy restrictions removed.

[0725] (Example 2)

[0726] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0727] Conventional security systems have difficulty detecting abnormal behavior without identifying individuals in the video, posing a challenge in accurately detecting anomalies while protecting individual privacy. Furthermore, abnormal behavior is judged based solely on actions and does not take emotional states into consideration, leading to the possibility of false positives.

[0728] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0729] In this invention, the server includes means for acquiring video data and detecting a target from that data, means for overlaying a substitute display that is difficult to identify onto the detected target, and means for analyzing the emotional state of the target and detecting abnormal behavior based on that emotional state. This makes it possible to accurately detect abnormal behavior that takes emotional state into account while protecting the privacy of individuals.

[0730] "Video data" refers to visual information represented in digital format, acquired by devices such as cameras for security and surveillance purposes.

[0731] A "subject" is something that exhibits a specific action or characteristic within the video data and is detected as a person or object.

[0732] "Alternative displays" are visual overlays used to protect the privacy of detected objects, and their role is to make identification difficult.

[0733] "Emotional state" refers to the internal psychological state inferred from the subject's facial expressions and actions, and includes sensory reactions such as joy, anger, sadness, and happiness.

[0734] "Abnormal behavior" refers to movements or states that deviate from normal behavioral patterns, indicating that the person being monitored may be experiencing some kind of danger or instability.

[0735] A "notification" is information sent to the relevant parties or systems upon detection of abnormal behavior, intended to prompt appropriate action.

[0736] This invention provides a system that utilizes security cameras installed in public places to detect abnormal behavior with high accuracy through video data. The system enables advanced surveillance using sentiment analysis while protecting user privacy.

[0737] The server collects video data from security cameras in real time. The collected video data is then used to detect people via an AI agent. In this case, commonly used image recognition technologies include TensorFlow and OpenCV.

[0738] Individuals who are detected will have alternative markers overlaid to make identification difficult. These alternative markers are provided in the form of mosaics or avatars. This ensures user privacy and provides a sense of security that they are being monitored.

[0739] Furthermore, the server uses an emotion engine to analyze a person's facial expressions and movements, and evaluates their emotional state in real time. Emotional states include specific expressions such as anger, joy, and fear.

[0740] Abnormal behavior is detected by comparing pre-configured behavioral patterns with analyzed emotional states. Even normal actions are considered abnormal if emotions such as anxiety or fear are recognized, and are logged on the server. If the state of the target is determined to be abnormal, a notification is sent to the terminal.

[0741] A concrete example of its use is in a security system within a train station. This system has the function of monitoring and constantly analyzing passengers' behavior and emotions. If a passenger frequently exhibits unstable emotions, the emotion engine will determine this to be an anomaly and provide a notification with details to the administrator's terminal.

[0742] Examples of prompts for the generative AI model include: "Please describe a system that identifies emotions from video data. How does this system detect abnormal behavior while protecting privacy?" and "Please give specific examples of the role and practical applications of emotion engines in security cameras."

[0743] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0744] Step 1:

[0745] The server collects video data from security cameras in real time. The input is the camera stream. This data is converted to a digital format and stored within the server. Specifically, the video is formatted for processing frame by frame, and pre-processing is performed to remove noise. The output is video data in an analyzable format.

[0746] Step 2:

[0747] The server uses an AI agent to detect people from video data. The input here is the processed video data from the previous step. A person detection model (e.g., HOG + SVM using OpenCV, or a deep learning model using TensorFlow) is applied to this data. As a result, the output is data showing the location and size of the detected people.

[0748] Step 3:

[0749] The server overlays a substitute image onto the detected person. The input is the person's location data obtained in step 2. Using image processing techniques, mosaics or avatars are overlaid on the person's face and body to make identification difficult. This process generates edited frames. The output is privacy-protected video data.

[0750] Step 4:

[0751] The server uses an emotion engine to analyze emotions from privacy-protected video data. The data from step 3 is used as input. The engine analyzes the subject's facial expressions and movements and applies image recognition and motion analysis algorithms to identify specific emotions (such as joy, anger, or anxiety). As a result, the output is obtained as the analyzed emotional state.

[0752] Step 5:

[0753] The server detects abnormal behavior based on the emotion analysis results. The input consists of emotional state and behavior pattern data obtained in step 4. The server compares the emotional state with a pre-configured normal behavior pattern to determine if there is an abnormality. The output is notification data if abnormal behavior is detected.

[0754] Step 6:

[0755] The server sends a notification to the terminal when an anomaly is detected. The input includes the notification data generated in step 5. From this data, a message is created that describes the specific abnormal behavior in detail. The output is a notification message displayed on the user's or administrator's terminal. This notification allows for prompt action to be taken as needed.

[0756] (Application Example 2)

[0757] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0758] In modern surveillance systems, balancing privacy protection and security is a crucial challenge. Conventional technologies only monitor the movement of objects to detect anomalies, making it difficult to detect abnormal behavior accompanied by unstable emotional states early. Furthermore, methods using alternative displays to protect personal information in video footage lack the flexibility to adapt to different situations. This has made it difficult to achieve both improved public safety and protection of individual privacy.

[0759] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0760] In this invention, the server includes means for acquiring video data and detecting an object from the video data, means for overlaying and displaying a difficult-to-identify alternative display on the detected object, and means for recognizing the emotional state of the object. This enables early detection of abnormal behavior based on emotional state and enhanced privacy protection according to the situation.

[0761] "Video data" refers to a collection of visual information acquired by cameras and other visual sensors.

[0762] "Object" refers to a person or object that exists within the video data and is detected based on specific conditions.

[0763] "Detection" refers to the process of identifying and distinguishing an object from video data.

[0764] "Alternative displays" are visual substitutes used to conceal personal information or privacy-related elements in the original video.

[0765] "Emotional state" refers to the psychological elements and characteristics recognized to express human emotions.

[0766] "Abnormal behavior" refers to actions that have characteristics different from the norm and require attention in monitoring systems.

[0767] A "notification" is a message generated by a system to inform about a specific event or state.

[0768] "Dynamic modification" means changing the display or behavior in response to changes in circumstances or parameters.

[0769] A "pre-configured behavioral pattern" is a pre-programmed behavioral profile used as a criterion for detecting abnormal behavior.

[0770] The system for implementing this invention is a security system that uses video data to detect objects and recognize their emotional states. The server acquires video data using cameras and other visual sensors and detects objects from it. Image processing libraries such as OpenCV are used for object detection.

[0771] The server uses an emotion engine powered by a generative AI model to analyze the emotional state of objects in real time from video data. This allows the emotion engine to recognize emotions such as joy, anger, sadness, and happiness, and detect abnormal behavior based on this data. This information is compared with pre-configured behavioral and emotional patterns, and if an anomaly is detected, a notification is sent to the user's device.

[0772] To protect user privacy, the server overlays a substitute display on detected objects. This substitute display is processed using classes such as PrivacyOverlay and is dynamically modified depending on the context. This dynamic modification allows for adjustment of its intensity based on the perceived emotional state.

[0773] As a concrete example, a home camera system may have a function that recognizes the emotions of family members and notifies them if they are experiencing persistent feelings of depression, suggesting that they may need support. In this case, a generative AI model is used to prompt the implementation of a specific emotion recognition algorithm.

[0774] An example of a prompt might be, "Implement an algorithm that recognizes an individual's emotions in real time while protecting their privacy when they are photographed." This prompt is used to provide generative AI models with implementation guidelines for emotion recognition and privacy protection features.

[0775] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0776] Step 1:

[0777] The server acquires video data in real time from cameras and vision sensors. This video data serves as input for processing. The input images are converted into a format that can be analyzed using OpenCV.

[0778] Step 2:

[0779] The server detects objects from the acquired video data. Here, image processing algorithms are used to identify people and objects in the video and output information about the detected objects. Contour recognition and face recognition technologies are used for image analysis.

[0780] Step 3:

[0781] The server overlays a substitute display onto the detected object. The input uses the object's location and shape data, and the PrivacyOverlay class is used to apply a substitute display that hides personal information. This process results in an output where a portion of the video is modified while protecting privacy.

[0782] Step 4:

[0783] The server recognizes the emotional state of the subject from the displayed alternative video. Using a generative AI model, it analyzes facial expressions and movements in the input video and outputs emotional data. A pre-trained emotion recognition algorithm is used to classify the emotional state.

[0784] Step 5:

[0785] The server detects abnormal behavior by comparing emotional states with previously recorded behavioral patterns. Emotional state data is used as input, and abnormal behavior is output only if it matches the conditions for being judged as abnormal. At this stage, pattern matching is performed for both emotions and behavior.

[0786] Step 6:

[0787] When abnormal behavior is detected, the server generates a notification and sends it to the designated device. The notification includes information such as the type and time of the abnormality and the emotional state, and a warning message is displayed to the user. This output is then fed into the alert system.

[0788] Step 7:

[0789] Furthermore, the alternative display is dynamically changed according to the detected emotional state. Based on the emotional data as input, the PrivacyOverlay class is executed again, and a new alternative display is applied to the video. As a result, a video that makes the user feel more secure is output.

[0790] 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 controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0791] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0792] 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 this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0793] Furthermore, the emotion identification model 59, acting 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 a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0794] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0795] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0796] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0797] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0798] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is 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 the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0799] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0800] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0801] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0802] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0803] 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.

[0804] Furthermore, it is not necessary to store the entirety 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 the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0805] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0806] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of 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). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0807] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0808] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0809] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0810] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0811] The following is further disclosed regarding the embodiments described above.

[0812] (Claim 1)

[0813] A means for acquiring video data and detecting an object from said video data,

[0814] A means of superimposing an indistinguishable alternative display onto the detected object,

[0815] Means for monitoring the behavior of the object and detecting abnormal behavior,

[0816] Means for generating a notification based on the detection of the abnormal behavior,

[0817] Means for canceling the alternative display and reconstructing the original image as needed,

[0818] A system that includes this.

[0819] (Claim 2)

[0820] The system according to claim 1, wherein the alternative representation is selected from a predefined set.

[0821] (Claim 3)

[0822] The system according to claim 1, wherein the detection of the abnormal behavior is compared with a pre-set behavior pattern.

[0823] "Example 1"

[0824] (Claim 1)

[0825] A means for acquiring video information and detecting an object from said video information,

[0826] A means of superimposing an indistinguishable alternative display onto the detected object,

[0827] Means for managing the operation of the target and detecting abnormal operation,

[0828] Means for generating a notification based on the detection of the abnormal operation,

[0829] Means for canceling the alternative display and reconstructing the original image as needed,

[0830] A means of using a machine learning model in the analysis of the video information,

[0831] The alternative representation is a means used to conceal a person's identifying information,

[0832] A system that includes this.

[0833] (Claim 2)

[0834] The system according to claim 1, wherein the alternative display is selected from a predefined display format.

[0835] (Claim 3)

[0836] The system according to claim 1, wherein the detection of the abnormal operation is compared with a pre-set operation pattern.

[0837] "Application Example 1"

[0838] (Claim 1)

[0839] A means for acquiring video information and detecting an object from said video information,

[0840] A means for visualizing the detected object by superimposing an alternative visual representation that is difficult to identify,

[0841] A means for monitoring the behavior of the target and detecting abnormal operation,

[0842] Means for generating communication based on the detection of the abnormal operation,

[0843] Means for reconstructing the original image by deactivating the alternative visual representation as needed,

[0844] A means of providing information to a personal device when abnormal operation is detected,

[0845] A system that includes this.

[0846] (Claim 2)

[0847] The system according to claim 1, wherein the alternative visual representation is selected from a predefined set.

[0848] (Claim 3)

[0849] The system according to claim 1, wherein the detection of the abnormal operation is compared with a pre-set operation pattern.

[0850] "Example 2 of combining an emotion engine"

[0851] (Claim 1)

[0852] A means for acquiring video data and detecting an object from said video data,

[0853] A means for superimposing an alternative display that is difficult to identify onto the detected object,

[0854] A means for analyzing the emotional state of the subject and detecting abnormal behavior based on the analyzed emotional state,

[0855] Means for generating a notification based on the detection of the abnormal behavior,

[0856] Means for dynamically changing the alternative display in response to the change in the emotional state,

[0857] ...

[0858] A system that includes this.

[0859] (Claim 2)

[0860] The system according to claim 1, wherein the alternative display is selected from a predefined set and dynamically changed according to the emotional state.

[0861] (Claim 3)

[0862] The system according to claim 1, which, in detecting the abnormal behavior, compares a pre-set behavioral pattern with a change in emotional state.

[0863] "Application example 2 when combining with an emotional engine"

[0864] (Claim 1)

[0865] A means for acquiring video data and detecting an object from said video data,

[0866] A means of superimposing an indistinguishable alternative display onto the detected object,

[0867] Means for recognizing the emotional state of the object,

[0868] A means for detecting abnormal behavior based on the emotional state and for generating a notification based on the detection of abnormal behavior,

[0869] Means for dynamically changing the alternative display according to the emotional state as needed, and for reconstructing the original video,

[0870] A system that includes this.

[0871] (Claim 2)

[0872] The system according to claim 1, wherein the alternative display uses a dynamically changeable display method and is selected from a predefined set.

[0873] (Claim 3)

[0874] The system according to claim 1, which detects abnormal behavior based on the emotional state by comparing it with a pre-set behavior pattern and emotional pattern. [Explanation of symbols]

[0875] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for acquiring video data and detecting an object from said video data, A means of superimposing an indistinguishable alternative display onto the detected object, Means for monitoring the behavior of the object and detecting abnormal behavior, Means for generating a notification based on the detection of the abnormal behavior, Means for canceling the alternative display and reconstructing the original image as needed, A system that includes this.

2. The system according to claim 1, wherein the alternative representation is selected from a predefined set.

3. The system according to claim 1, wherein the detection of the abnormal behavior is compared with a pre-set behavior pattern.

Images