System

Abstract

A system is provided.SOLUTION: The system includes a means for receiving video data from a security camera, a processing means including multi-modal artificial intelligence for analyzing the received video data in real time and detecting an abnormal behavior or a criminal act, and a notification means for notifying when the abnormal behavior or the criminal act is detected.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] The purpose of this invention is to improve crime prevention and security in places with high traffic, such as commercial facilities, office buildings, public facilities, schools, and hospitals. Conventional security cameras are effective in deterring crime and collecting evidence after a crime has occurred, but 24-hour real-time monitoring requires a large number of personnel and is expensive. Furthermore, it is difficult to detect and report abnormal behavior or criminal activity in real time, and an effective means to solve this problem is needed. Furthermore, there is an issue that they lack the ability to predict future crises using past data, resulting in insufficient advance countermeasures. [Means for solving the problem]

[0005] The present invention provides a system that includes a means for receiving video data from security cameras, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, and a notification means for issuing a report when abnormal behavior or criminal activity is detected. The system also includes a means for learning from past data from the received video data and predicting future crises and crimes, and a means for connecting to the security cameras and performing system maintenance and subscription management. The system also includes a means for predicting dangers at specific times and locations based on the learned data and notifying the results, and a means for temporarily processing security camera data on an edge device or local server and transmitting it to a central server. This not only enables efficient monitoring of large volumes of video data and immediate response, but also enables future crisis prediction, contributing to the creation of a safer society.

[0006] A "security camera" is a surveillance camera device that uses captured images for the purposes of monitoring and recording.

[0007] "Video data" refers to digital images or video captured by security cameras.

[0008] "Real-time" refers to data processing and analysis occurring simultaneously with actual events.

[0009] "Analysis" is the act of interpreting video data and detecting specific patterns or abnormal behavior.

[0010] "Abnormal behavior" refers to behavior that is unusual or dangerous.

[0011] "Criminal activity" means an act that violates the law.

[0012] "Multimodal AI" refers to AI that can simultaneously analyze and understand multiple data formats, such as text, video, and audio.

[0013] "Processing means" refers to a combination of hardware and software for receiving and analyzing video data.

[0014] "Notification means" refers to a system or device that provides notification when abnormal or criminal activity is detected.

[0015] "Learning" refers to the process of analyzing past data and using the results to make future predictions, etc.

[0016] "Prediction" refers to determining the likelihood of future events based on past data.

[0017] "Maintenance" refers to the act of regularly inspecting and repairing systems and equipment to keep them operating normally.

[0018] "Subscription management" refers to the process of managing contract information when providing fixed-term contract services, and performing operations including renewal and cancellation.

[0019] "Edge device" refers to a computer or piece of equipment that collects and processes data on the actual device.

[0020] A "local server" refers to a server device that processes and stores data within a specific facility.

[0021] "Central server" refers to a main server device that centrally manages and processes data received from multiple facilities and devices.

[0022] "Notification system" refers to a system for sending alerts to relevant parties when abnormal or criminal behavior is detected. [Brief explanation of the drawings]

[0023] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2]1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0024] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0025] First, the terms used in the following description will be explained.

[0026] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0027] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0028] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0029] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0030] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0031] [First embodiment]

[0032] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0033] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0034] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0035] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0036] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0037] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0038] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0039] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0040] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0041] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0042] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0043] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0044] This invention is a system that receives video data from security cameras, analyzes the data in real time, and detects and reports abnormal behavior and criminal activity. This system uses multimodal artificial intelligence (AI) to efficiently monitor large amounts of video data. It also has the ability to learn from past data and predict potential future crises. This improves safety while reducing costs and labor.

[0045] System configuration

[0046] Terminals: Security cameras and edge devices

[0047] Server: A processing server that analyzes and learns from video data, and a notification server that notifies administrators and users of the results.

[0048] Users: Administrators and security personnel who use the system.

[0049] What the program does

[0050] Receiving data from security cameras

[0051] The device receives real-time video data captured by the security camera. The edge device processes the data and sends it to a local server, which then compresses and encrypts the data and securely transfers it to a central server.

[0052] Real-time analysis of video data

[0053] The server decompresses the video data received on the central server and inputs it into the multimodal AI, which analyzes the video data in real time to detect abnormal behavior or criminal activity, such as detecting a person lying down or multiple people entering the same store at the same time.

[0054] Reporting when abnormal behavior is detected

[0055] When the server detects abnormal behavior or criminal activity, it immediately generates an alert and sends it to security personnel or administrators via the notification system. The alert is sent to each device, allowing administrators and security personnel to take the necessary action promptly.

[0056] Learning from past data and predicting future crises

[0057] The server periodically collects and saves past video data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal activity based on past behavioral patterns. For example, it can predict dangers at specific times or locations and notify administrators of the results.

[0058] Subscriptions and Maintenance

[0059] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. Furthermore, as part of system maintenance management, it performs regular system checks and self-diagnosis, and promptly notifies the maintenance team if any abnormalities are discovered.

[0060] Specific examples

[0061] For commercial facilities

[0062] The manager of a commercial facility installs the "system of the present invention." Security cameras capture video data within the facility in real time and send the data to a central server via terminals. Multimodal AI on the central server analyzes the video data and detects abnormal behavior (for example, multiple people entering the same store at the same time). Upon detection, the notification system sends an alert to the manager of the commercial facility, and a security guard is immediately dispatched, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week, and preventive measures can be taken by notifying the manager.

[0063] As a result, the present invention significantly improves safety in commercial facilities, reduces costs and labor, and realizes efficient crime prevention measures.

[0064] The processing flow will be explained below.

[0065] Processing steps for receiving data from security cameras and analyzing it in real time

[0066] Step 1:

[0067] The device captures video data captured by security cameras in real time, and metadata such as a timestamp and camera ID is attached to the video data.

[0068] Step 2:

[0069] The device sends the captured video data to an edge device or local server, which performs simple data processing to extract and compress the important parts.

[0070] Step 3:

[0071] The device encrypts the processed video data on a local server and securely transfers it to a central server.

[0072] Step 4:

[0073] The server decompresses the video data received by the central server and sends it to the AI ​​analysis module.

[0074] Step 5:

[0075] The server then begins real-time analysis of the received video data using multimodal AI, which analyzes people's movements and behaviors to detect patterns of abnormal or criminal behavior.

[0076] Step 6:

[0077] If the AI ​​analysis detects abnormal behavior or criminal activity, the server immediately generates an alert, which includes information such as the type of abnormality, the time of detection, and the camera ID.

[0078] Step 7:

[0079] The server sends alerts to administrators and users through a notification system, which can be sent via email, SMS, or a dedicated application.

[0080] Step 8:

[0081] Users receive alerts and can respond quickly to any detected abnormal or criminal activity (e.g., dispatch security personnel to the scene or notify the police).

[0082] Processing steps for learning from past data and predicting future crises

[0083] Step 1:

[0084] The server periodically collects video data from the past few years and stores it in a database.

[0085] Step 2:

[0086] The server preprocesses each video data, removes noise, and converts it into a format suitable for analysis.

[0087] Step 3:

[0088] The server inputs the preprocessed data into the multimodal AI and trains it to learn patterns of crime and abnormal behavior.

[0089] Step 4:

[0090] The server periodically provides new data to the AI ​​and updates the learning model.

[0091] Step 5:

[0092] The server uses the trained model to predict abnormal behavior and criminal acts that are likely to occur in the future.

[0093] Step 6:

[0094] The server notifies the administrator and user of the prediction results (e.g., recommended times and locations for increased security).

[0095] Subscription and Maintenance Process Steps

[0096] Step 1:

[0097] Users can sign up for a subscription through a web portal or a dedicated application.

[0098] Step 2:

[0099] The server receives new contract, renewal and cancellation information and stores it in a database.

[0100] Step 3:

[0101] The server connects and provides security cameras and services based on the contract (e.g., adding new cameras and updating existing cameras).

[0102] Step 4:

[0103] The server periodically schedules system self-diagnosis to check the operational status of security cameras and edge devices.

[0104] Step 5:

[0105] The terminal performs periodic self-diagnosis and notifies the server if an abnormality is detected.

[0106] Step 6:

[0107] If an abnormality is detected, the server will promptly notify the maintenance team and arrange for repair or replacement work.

[0108] The above is a specific operation of each processing step of the present invention.

[0109] Example 1

[0110] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0111] In recent years, the rise in crime and abnormal behavior has led to a growing need to strengthen security in commercial facilities and public spaces. Conventional security systems have difficulty detecting abnormal behavior in real time, and they rarely use past data to predict future crises. As a result, despite the significant cost and effort involved, effective security measures have yet to be implemented.

[0112] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0113] In this invention, the server includes: means for receiving video data from security cameras; means including an edge device that temporarily processes, compresses, and encrypts the received video data; means for transferring the compressed and encrypted data to a central server; means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity; means including a notification system that issues a report when abnormal behavior or criminal activity is detected; means for saving past video data and using it to predict future crises; and means for managing subscription contract information and performing system maintenance and management. This makes it possible to quickly detect abnormal behavior or criminal activity in real time and take effective reports and preventive measures.

[0114] A "security camera" is a photographic device for monitoring a specific area and capturing video data.

[0115] An "edge device" is a computing device that temporarily processes received data and performs complex calculations, data compression, and encryption locally.

[0116] A "local server" is an intermediate server that processes data received from edge devices and securely forwards it to the central server.

[0117] A "central server" is a high-performance computing environment for managing, analyzing, and storing large amounts of data.

[0118] "Multimodal artificial intelligence" is an artificial intelligence system that integrates and analyzes a variety of data sources (such as video, audio, and text) to perform highly accurate detection and prediction.

[0119] A "notification system" is a communication system that quickly notifies administrators and security personnel when abnormal behavior or criminal activity is detected.

[0120] "Historical video data" refers to video recordings previously collected and stored by security cameras.

[0121] "Crisis forecasting" is the process of predicting potential future dangers and risks based on past data and current conditions.

[0122] A "subscription contract" is a contract for using a specific service or function for a certain period of time, and manages the user's registration status and contract details.

[0123] "System maintenance management" is the process of regularly checking and maintaining a system to ensure that it always operates properly.

[0124] "Real-time analytics" is the process of analyzing data as it is generated and transmitted.

[0125] This invention is a system that receives video data from security cameras, analyzes the data in real time to detect abnormal behavior or criminal activity, and issues a report if necessary. This system uses multimodal artificial intelligence (AI) to efficiently monitor large amounts of video data. It also has the ability to learn from past data and predict potential future crises. This improves safety while reducing costs and labor.

[0126] System configuration

[0127] Terminals: Security cameras and edge devices

[0128] Server: A processing server that analyzes and learns from video data, and a notification server that notifies administrators and users of the results.

[0129] Users: Administrators and security personnel who use the system

[0130] Program processing

[0131] Receiving video data from security cameras

[0132] The terminal receives video data captured by security cameras in real time. The edge device temporarily processes this data and forwards it to a local server, which then compresses and encrypts the data and securely transmits it to a central server.

[0133] Video data analysis and abnormal behavior detection

[0134] The server decompresses the video data received on the central server and inputs it into a multimodal AI, which analyzes the video data in real time to detect abnormal behavior or criminal activity.

[0135] For example, AI can detect abnormal behavior such as:

[0136] A scene where someone has fallen

[0137] A situation where multiple people enter the same store at the same time

[0138] Reporting Procedures

[0139] If the server detects any abnormal or criminal activity, it immediately generates an alert and sends it to security personnel or administrators through the notification system, allowing them to respond quickly.

[0140] Learning from past data and predicting future crises

[0141] The server periodically collects and saves past video data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal activity based on past behavioral patterns.

[0142] For example, the AI ​​predicts:

[0143] Risks are higher at certain times of the day (e.g., 10 PM to 2 AM)

[0144] Behavioral patterns in specific locations

[0145] These predictions are communicated to administrators to help them take preventative measures.

[0146] Subscriptions and Maintenance

[0147] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. It also performs periodic system checks and self-diagnosis as part of system maintenance, and notifies the maintenance team if any abnormalities are discovered.

[0148] Specific examples

[0149] Shopping mall surveillance

[0150] The manager of a commercial facility installs the system of this invention. Security cameras capture video data within the facility in real time and send the data to a central server via terminals. Multimodal AI on the central server analyzes the video data and detects abnormal behavior (for example, multiple people entering the same store at the same time). When abnormal behavior is detected, the notification system sends an alert to the manager, allowing security guards to respond promptly. Furthermore, the system can predict dangers on specific days of the week and at specific times of the day based on past data and notify the manager so that preventative measures can be taken.

[0151] As a result, the present invention significantly improves safety in commercial facilities, reduces costs and labor, and realizes efficient crime prevention measures.

[0152] Example prompts to input to the generative AI model

[0153] I would like to develop a system that monitors video data from security cameras in real time and detects abnormal behavior. In particular, I need a function that can detect behavior such as a person falling or multiple people entering at the same time and notify an administrator with an alert. I would like to know examples of multimodal AI models suitable for this purpose and the latest technologies.

[0154] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0155] Step 1: Receiving video data from security cameras

[0156] The terminal receives video data captured by security cameras in real time. The input is raw video frames. This video data is sent to the edge device and temporarily stored in a buffer. The edge device then quickly processes the data and prepares it for the next step.

[0157] Step 2: Compress and encrypt the data

[0158] The local server compresses the video data received from the edge device. The input is raw video data. A compression algorithm (e.g., H.264 or HEVC) is used to reduce the data size for efficient transmission. The data is then encrypted using an encryption algorithm such as AES-256. The output is compressed and encrypted data. This process reduces the risk of unauthorized access to the data during transmission.

[0159] Step 3: Send data to a central server

[0160] The local server sends compressed and encrypted data to the central server. The input is the compressed and encrypted video data. The output is secure data sent to the central server. This procedure uses a secure protocol (e.g., HTTPS or TLS) to send the data.

[0161] Step 4: Unpack and analyze the data

[0162] The central server decompresses the received compressed and encrypted data and restores it to the original video data. The input is the compressed and encrypted data. After decompression, this data is fed into the multimodal AI, which analyzes the video data in real time to detect abnormal behavior and criminal activity. The output is the analyzed results. This analysis step uses deep learning and machine learning algorithms to identify abnormal patterns in the video.

[0163] Step 5: Detecting Anomalous Behavior

[0164] Multimodal AI on a central server analyzes video data to detect abnormal behavior or criminal activity. The input is decompressed video data. The AI ​​detects, for example, a person lying on the ground or multiple people entering the same store at the same time. The output is information that abnormal behavior or criminal activity has been detected.

[0165] Step 6: Alerting and Notification

[0166] The central server immediately generates an alert if abnormal behavior or criminal activity is detected. The input is the abnormal behavior detection information from the multimodal AI. It then sends a notification to security personnel or administrators via the notification system. The output is an alert sent to each terminal or notification device, allowing administrators and security personnel to respond quickly.

[0167] Step 7: Collect and store historical data

[0168] The central server periodically collects past video data and stores it in a database. The input is the video data received and analyzed in real time, along with the analysis results. The stored data is used for subsequent learning and analysis. The output is the past video data stored in the database.

[0169] Step 8: Anticipate future crises

[0170] The central server predicts future crises and criminal activity based on stored historical video data. The input is the learning results of past data. Multimodal AI analyzes past behavioral patterns and predicts potential dangers that may occur at specific times and locations. The output is predicted crisis information. This information is notified to administrators, and appropriate preventive measures are taken.

[0171] Step 9: Subscriptions and Maintenance

[0172] The central server manages subscription contract information and processes new contracts, renewals, and cancellations for users. The input is the user's subscription information. In addition, as part of system maintenance, it performs periodic system checks and self-diagnosis, and if an abnormality is discovered, it promptly notifies the maintenance team. The output is the current subscription status and the normal operating status of the system.

[0173] keyword:

[0174] Generative AI model, prompt sentence

[0175] (Application example 1)

[0176] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0177] Commercial facilities and public spaces require a means to quickly and efficiently detect abnormal behavior and criminal activity and report it to the relevant parties. Conventional systems often do not analyze surveillance camera footage in real time, making it difficult to detect abnormal situations early. Furthermore, they lack the ability to use past data to predict future crises, making it difficult to implement preventative safety measures.

[0178] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0179] In this invention, the server includes a means for receiving video data from security cameras, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, a notification means for reporting when abnormal behavior or criminal activity is detected, a means for learning from past data to predict future crises and notifying of risks at specific times or locations, and a means for sending alerts to smartphones. This not only enables the rapid detection and reporting of abnormal behavior and criminal activity, but also enables the prediction of future crises, enabling more comprehensive security measures.

[0180] A "security camera" is a photographic device installed for crime prevention purposes to photograph a specific area.

[0181] The "means for receiving video data" refers to a device or system that receives video data transmitted from a security camera.

[0182] "Means for analyzing in real time" refers to a processing device or system that can analyze received video data immediately without delay.

[0183] "Abnormal or criminal behavior" refers to behavior that deviates from normal patterns of behavior and may be related to a crime.

[0184] "Multimodal AI" is an AI technology that can simultaneously analyze different types of data (e.g., video, audio, text).

[0185] "Means for reporting" refers to a means for quickly conveying information about detected abnormal behavior or criminal activity to relevant parties.

[0186] "Means for learning from past data to predict future crises" refers to a processing device or system for analyzing and predicting future risks using data recorded in the past.

[0187] "Means for notifying risks at specific times and locations" refers to means for communicating information about times and locations where predicted dangers are likely to occur to relevant parties.

[0188] "Means for sending alerts to smartphones" refers to means for notifying smartphones of detected abnormalities or predicted dangers.

[0189] System configuration

[0190] This invention is a system that receives video data from security cameras, analyzes the data in real time, and detects and reports abnormal behavior and criminal activity. The main components of the system are terminals, security cameras, edge devices, local servers, central servers, multimodal AI, notification servers, and smartphones.

[0191] Hardware used

[0192] Security cameras: High-resolution cameras (e.g., Sony IMX series)

[0193] Edge device: Nvidia Jetson Nano

[0194] Cloud server: AWS (registered trademark) EC2 instance

[0195] Database: PostgreSQL

[0196] Software used

[0197] Multimodal AI model: Custom model using TENSORFLOW®

[0198] Encryption: AES encryption

[0199] API: REST API

[0200] Data processing flow

[0201] 1. The device receives video data captured by the security camera. The edge device temporarily processes this data, adjusting the frame rate and reducing noise.

[0202] 2. The edge device sends the processed data to a local server, where it is AES encrypted to ensure security.

[0203] 3. The local server transfers the encrypted data to the central server.

[0204] 4. The central server decompresses the received video data and inputs it into a multimodal artificial intelligence model using TensorFlow, which analyzes and detects abnormal or criminal behavior in real time.

[0205] 5. The notification server generates alerts immediately when abnormal or criminal behavior is detected and sends notifications to the smartphones of administrators and security personnel.

[0206] 6. The central server periodically collects and stores past video data, which is then used by the multimodal AI model to learn and predict future crises and criminal activity.

[0207] 7. The entire system manages subscription contract information, processes new contracts, renewals, and cancellations, and also periodically performs system checks and self-diagnosis, promptly notifying the maintenance team of any abnormalities.

[0208] Specific examples

[0209] For example, let's take a specific example of how this system might be implemented in a commercial facility. The facility manager installs security cameras in several locations. These cameras capture video data within the facility in real time and send the data via terminals to a central server. Multimodal AI on the central server analyzes the video data and detects abnormal behavior, such as multiple people entering a specific store at the same time. If detected, the notification system immediately sends an alert to the facility manager, enabling security guards to be dispatched immediately. Additionally, past data can be used to predict dangers on specific days and times of the day, and preventive measures can be taken by notifying the manager.

[0210] Prompt Sentence Examples

[0211] Analyze video data in real time to detect patterns of multiple people entering a specific area at once. When this happens, send an alert to the security guard's smartphone to notify them of any suspicious activity detected, and use past data to predict the time of day when the next suspicious activity is likely to occur.

[0212] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0213] Step 1:

[0214] The terminal receives video data captured by security cameras in real time. The edge device temporarily processes this data, adjusting the frame rate and reducing noise. Specifically, the video data is input and a noise reduction filter is applied to generate high-quality image data. The output is processed, high-quality video data.

[0215] Step 2:

[0216] The edge device sends the processed data to a local server. At this time, the data is AES encrypted. Specifically, the frame-rate-adjusted video data is input, and the AES encryption algorithm is applied to generate encrypted data. The output is the encrypted video data.

[0217] Step 3:

[0218] The local server transfers the encrypted data to the central server. This transfer uses a secure protocol (e.g. HTTPS). The input is the encrypted video data, which is sent to the central server using a data transfer protocol. The output is the encrypted data that arrives at the central server.

[0219] Step 4:

[0220] The central server decompresses the received video data and inputs it into a multimodal artificial intelligence model using TensorFlow. Specifically, encrypted data is input and the original video data is extracted by applying the AES decryption algorithm. The extracted video data is then input into the multimodal AI model. The output is the analysis result after the AI ​​model has performed its analysis.

[0221] Step 5:

[0222] The notification server immediately generates an alert if any abnormal or criminal behavior is detected and sends a notification to the smartphone of the administrator or security officer. Specifically, the analysis results are input and it determines whether or not any abnormal behavior has been detected. If an abnormality is detected, an alert message is generated and sent to the smartphone using a messaging protocol (e.g., FCM - Firebase Cloud Messaging). The output is an alert notification that is displayed on the smartphone of the administrator or security officer.

[0223] Step 6:

[0224] The central server periodically collects and stores past video data, and the multimodal AI model uses this data for learning. Specifically, video data is input at regular intervals and stored in a database. The AI ​​model is trained based on this data. The output is an updated AI model.

[0225] Step 7:

[0226] The entire system manages subscription contract information and processes new contracts, renewals, and cancellations. It also periodically performs system checks and self-diagnosis, and promptly notifies the maintenance team of any abnormalities. Specifically, user contract information is entered and saved / updated in the management database. If an abnormality is detected during regular self-diagnosis, the maintenance team is notified via a ticket system (e.g., JIRA). The output is the contract information update status and notification information for the maintenance team.

[0227] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0228] This invention combines a system that receives video data from security cameras, analyzes the data in real time to detect and report abnormal behavior and criminal activity, with an emotion engine that recognizes user emotions. This system utilizes multimodal artificial intelligence (AI) to efficiently monitor large volumes of video data. It also has the ability to learn from past data and predict potential future crises, as well as monitor the user's emotional state, enabling more accurate detection of abnormal behavior and immediate response.

[0229] System configuration

[0230] Terminals: Security cameras and edge devices

[0231] Server: A processing server that analyzes and learns video data and emotion data, and a notification server that reports the results to administrators and users.

[0232] Users: Administrators and security personnel who use the system.

[0233] What the program does

[0234] Receiving data from security cameras and analyzing it in real time

[0235] The device captures video data captured by security cameras in real time. The edge device processes the data and sends it to a local server, which then compresses and encrypts the data and securely transfers it to a central server.

[0236] The server decompresses the video data received on the central server and sends it to the AI ​​analysis module. Multimodal AI analyzes the video data in real time to detect abnormal behavior and criminal activity, such as detecting a person lying down or multiple people entering at the same time.

[0237] If the server detects abnormal behavior or criminal activity as a result of AI analysis, it immediately generates an alert and sends it to security personnel or administrators via the notification system. The alert is sent to each device, allowing administrators and security personnel to take the necessary action promptly.

[0238] Recognizing user emotions with an emotion engine

[0239] The server uses an emotion engine to analyze the user's video data and recognize their emotional state. The emotion engine analyzes the user's emotions based on facial expressions, voice, body movements, etc., and detects emotional states such as stress, anxiety, and anger.

[0240] The server complements the detection results of abnormal behavior and criminal activity based on the detected emotional state. For example, if a user shows strong anxiety or tension, it will focus on monitoring the surrounding environment. By integrating the analysis of emotional state and abnormal behavior patterns, it becomes possible to detect abnormal behavior with greater accuracy.

[0241] Learning from past data and predicting future crises

[0242] The server periodically collects past video and emotional data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal acts based on past behavioral patterns and emotional states. For example, it can predict dangers at specific times or locations and notify administrators of the results, allowing them to take preventative measures.

[0243] Subscriptions and Maintenance

[0244] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. Furthermore, as part of system maintenance management, it performs regular system checks and self-diagnosis, and promptly notifies the maintenance team if any abnormalities are discovered.

[0245] Specific examples

[0246] For commercial facilities

[0247] The manager of a commercial facility installs the "system of the present invention." Security cameras capture video data within the facility in real time and send the data to a central server via terminals. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering at the same time). Upon detection, an emotion engine analyzes the emotional state of surrounding users and detects the occurrence of abnormal behavior and associated emotional states (e.g., anxiety or fear). Based on this, a notification system sends an alert to the manager of the commercial facility, immediately dispatching security guards, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week, and preventive measures can be taken by notifying the manager.

[0248] As a result, the present invention significantly improves safety in commercial facilities and realizes comprehensive crime prevention measures that also take into account the emotional state of users.

[0249] The processing flow will be explained below.

[0250] Processing steps of a system that combines emotion engines

[0251] Processing steps for receiving data from security cameras and analyzing it in real time

[0252] Step 1:

[0253] The device captures video data captured by security cameras in real time, and metadata such as a timestamp and camera ID is attached to the video data.

[0254] Step 2:

[0255] The device sends the captured video data to an edge device or local server, which performs simple data processing to extract and compress the important parts.

[0256] Step 3:

[0257] The device encrypts the processed video data on a local server and securely transfers it to a central server.

[0258] Step 4:

[0259] The server decompresses the video data received by the central server and sends it to the AI ​​analysis module.

[0260] Step 5:

[0261] The server then begins real-time analysis of the received video data using multimodal AI, which analyzes people's movements and behaviors to detect patterns of abnormal or criminal behavior.

[0262] Step 6:

[0263] If the AI ​​analysis detects abnormal behavior or criminal activity, the server immediately generates an alert, which includes information such as the type of abnormality, the time of detection, and the camera ID.

[0264] Step 7:

[0265] The server sends alerts to administrators and users through a notification system, which can be sent via email, SMS, or a dedicated application.

[0266] Step 8:

[0267] Users receive alerts and can respond quickly to any detected abnormal or criminal activity (e.g., dispatch security personnel to the scene or notify the police).

[0268] Processing steps for recognizing user emotions by the emotion engine

[0269] Step 1:

[0270] The device uses security cameras or dedicated sensors to capture video data of the user, including the user's facial expressions and movements.

[0271] Step 2:

[0272] The device sends the captured video data to the edge device for initial processing, which then detects the user's face and extracts features for emotion analysis.

[0273] Step 3:

[0274] The terminal encrypts the processed data and sends it to the central server.

[0275] Step 4:

[0276] The server analyzes the received data using an emotion engine, which performs facial expression recognition, voice analysis, and body movement analysis to determine the user's emotional state.

[0277] Step 5:

[0278] When abnormal behavior or abnormal emotion is detected, the server integrates the abnormal behavior detection results with the emotion analysis results.

[0279] Step 6:

[0280] The server generates an alert including the results of the emotion analysis and notifies the administrator. Information about the emotional state (e.g., anxiety, stress) is added to the alert.

[0281] Processing steps for learning from past data and predicting future crises

[0282] Step 1:

[0283] The server periodically collects video data and emotion data from the past few years and stores them in a database.

[0284] Step 2:

[0285] The server preprocesses each video and emotion data, removes noise, and converts it into a format suitable for analysis.

[0286] Step 3:

[0287] The server inputs the preprocessed data into a multimodal AI and trains it to learn patterns of criminal and abnormal behavior and emotional states.

[0288] Step 4:

[0289] The server periodically provides new data to the AI ​​and updates the learning model.

[0290] Step 5:

[0291] The server uses the trained model to predict abnormal behavior and criminal acts that are likely to occur in the future.

[0292] Step 6:

[0293] The server notifies the administrator and user of the prediction results (e.g., recommended times and locations for increased security).

[0294] Subscription and Maintenance Process Steps

[0295] Step 1:

[0296] Users can sign up for a subscription through a web portal or a dedicated application.

[0297] Step 2:

[0298] The server receives new contract, renewal and cancellation information and stores it in a database.

[0299] Step 3:

[0300] The server connects and provides security cameras and services based on the contract (e.g., adding new cameras and updating existing cameras).

[0301] Step 4:

[0302] The server periodically schedules system self-diagnosis to check the operational status of security cameras and edge devices.

[0303] Step 5:

[0304] The terminal performs periodic self-diagnosis and notifies the server if an abnormality is detected.

[0305] Step 6:

[0306] If an abnormality is detected, the server will promptly notify the maintenance team and arrange for repair or replacement work.

[0307] As a result, the present invention significantly improves safety in commercial facilities and public places, and by taking into account the user's emotional state, it enables more precise and prompt responses.

[0308] Example 2

[0309] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0310] Conventional security systems have limitations in the monitoring and abnormal behavior detection capabilities of security cameras, making it difficult to detect abnormal behavior or criminal activity early and respond immediately. Furthermore, because they do not take the user's emotional state into account, they may miss potential dangers or stressful situations. Furthermore, systems that utilize past data to predict future risks are also inadequate, making it difficult to take effective preventative measures.

[0311] The identification processing by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for receiving video data from a security camera, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, a means for analyzing the user's emotional state using an emotion engine and complementing the abnormal behavior detection result, a notification means for issuing a report when abnormal behavior or criminal activity is detected, a means for collecting and learning past video data and emotion data, and a means for predicting future crises and criminal activity. This enables the security system to detect abnormal behavior or criminal activity with higher accuracy and respond immediately based on the analysis of emotion data. Furthermore, it becomes possible to predict future risks and take preventive measures by learning from past data.

[0312] A "security camera" is a video capture device installed to monitor fraudulent or abnormal behavior.

[0313] "Video data" refers to digital data that includes information about images and videos captured by security cameras.

[0314] "Reception" is the process of acquiring video data transmitted from a security camera.

[0315] "Real-time analysis" means processing and analyzing the received video data immediately.

[0316] "Abnormal behavior" refers to movements that deviate from normal patterns of behavior, and includes, for example, a person falling down, suddenly running, or roughly handling objects.

[0317] "Criminal activity" refers to conduct that violates the law and includes, for example, acts such as theft, assault, and trespass.

[0318] "Multimodal AI" is an AI technology that integrates and analyzes multiple data modalities (e.g., video, audio, text, etc.).

[0319] "Processing means" is a general term for devices and software for performing specific processes or functions.

[0320] An "emotion engine" is an algorithm or system for analyzing a user's emotional state, recognizing emotions based on facial expressions, voice, body movements, etc.

[0321] "Notification means" refers to a system or device that notifies security personnel or administrators of abnormal behavior or criminal activity when such behavior or criminal activity is detected.

[0322] "Data collection" is the process of systematically gathering information about visual and emotional states.

[0323] "Learning" is the process by which artificial intelligence acquires new knowledge based on past data and improves its performance.

[0324] "Predicting future crises and criminal acts" is the process of analyzing past data and trends to estimate possible future risks and criminal acts in advance.

[0325] This invention combines a system that receives video data from security cameras, analyzes the data in real time to detect and report abnormal behavior and criminal activity, with an emotion engine that recognizes user emotions. This system utilizes multimodal artificial intelligence (AI) to efficiently monitor large volumes of video data. It also has the ability to learn from past data and predict potential future crises, as well as monitor the user's emotional state, enabling more accurate detection of abnormal behavior and immediate response.

[0326] The system configuration is as follows:

[0327] Terminal

[0328] The devices include security cameras and edge devices. The security cameras capture video from designated surveillance areas 24 hours a day. The edge devices temporarily store the captured video data, then compress and encrypt it before sending it to a local server. The local server decompresses the data and securely transfers it to a central server.

[0329] server

[0330] The server has the following modules:

[0331] 1. The central server is a platform for real-time analysis of received video data. Multimodal AI runs on this server, analyzes the video data, and detects abnormal behavior and criminal activity.

[0332] 2. The emotion engine module analyzes the user's video data and recognizes their emotional state (e.g., stress, anxiety, anger), which complements and improves the accuracy of abnormal behavior detection.

[0333] 3. The notification system generates alerts immediately when abnormal or criminal behavior is detected and sends notifications to administrators and security personnel via various media such as email and SMS.

[0334] 4. The data collection and learning module periodically collects historical video and emotion data and stores it in a database. This data is then used by the multimodal AI to learn and predict future crises and criminal activity.

[0335] User

[0336] Users are administrators and security personnel who monitor and manage the system. They receive alerts from the notification system and can take immediate action. Administrators can also receive forecast information based on past data and take measures in advance.

[0337] Specific examples

[0338] For commercial facilities

[0339] The manager of a commercial facility installs the system of the present invention. Security cameras capture video data within the facility in real time and transmit the data to a central server via terminals. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering at the same time). Upon detection, an emotion engine analyzes the emotional states of surrounding users and detects emotional states (e.g., anxiety or fear) associated with the occurrence of abnormal behavior. Based on this, a notification system sends an alert to the manager of the commercial facility, and a security guard is immediately dispatched, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week and notify the manager, allowing preventive measures to be taken in advance.

[0340] Prompt Sentence Examples

[0341] "Please explain a system that receives video data from security cameras, analyzes it in real time to detect abnormal behavior or criminal activity, and also analyzes the user's emotional state and reports it."

[0342] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0343] Step 1:

[0344] The terminal captures video data in real time using security cameras. The real-time video information is input, and the video data sent to the edge device is generated as output. The security cameras operate 24 hours a day, capturing the movements of people and the placement of objects.

[0345] Step 2:

[0346] The edge device of the terminal temporarily stores the captured video data. The input is video data sent from the security camera, and the edge device compresses and encrypts the data and sends it to the local server. The compressed and encrypted data is then transferred to the local server as the output.

[0347] Step 3:

[0348] The server receives video data from the local server. The input is compressed and encrypted data, which the server decompresses and decrypts. The decompressed data is then passed to the AI ​​analysis module. The output is the decompressed and decrypted video data.

[0349] Step 4:

[0350] The server uses an AI analysis module to analyze the decompressed video data in real time. The input is the decompressed video data, and the multimodal AI analyzes people's movements and behavior patterns. For example, it analyzes people's dwell time and movements in a specific area. The output generates detection results for abnormal behavior and criminal activity.

[0351] Step 5:

[0352] When the server detects abnormal behavior or criminal activity, it immediately generates an alert. The input is the detection results from the AI ​​analysis module, and the generated alert is sent as output to the terminals of administrators and security personnel via the notification system. The alert is sent by email or SMS to prompt a response.

[0353] Step 6:

[0354] The server uses an emotion engine to analyze the user's emotional state. The input is video data, and the emotion engine recognizes emotions based on facial expressions, voice, body movements, etc. For example, data showing the user's anxiety or tension is analyzed. The output is the analyzed emotional state.

[0355] Step 7:

[0356] The server performs an integrated analysis of the emotion data obtained by the emotion engine and the abnormal behavior data from the AI ​​analysis module. The input is emotion data and abnormal behavior data, which complements the analysis results and improves accuracy. For example, if there is abnormal behavior in an area where anxious emotions are detected, further detailed monitoring will be performed. The output is the integrated analysis results.

[0357] Step 8:

[0358] The server periodically collects past video data and emotion data and stores them in a database. The input is the periodically collected video data and emotion data, which the multimodal AI uses to learn. The output is a trained AI model.

[0359] Step 9:

[0360] The server uses past data to predict future crises and criminal activity. The input is a trained AI model and past data, and it predicts risks, for example, during specific times of the day or on specific days of the week. The output is the predicted information, which is notified to an administrator so that countermeasures can be taken.

[0361] Step 10:

[0362] The server manages subscription contract information and performs system maintenance. The input is contract information, and it processes new contracts, renewals, and cancellations. It also performs system checks and self-diagnosis, and notifies the maintenance team if an abnormality is discovered. The output is the latest contract information and maintenance status.

[0363] (Application example 2)

[0364] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0365] Conventional security systems were able to detect abnormal behavior and criminal activity by analyzing video data, but because they did not take into account the emotional state of the user, there were limitations to the accuracy of predictions and the speed of response.Furthermore, there was an issue that simply detecting abnormal behavior made it difficult to respond flexibly according to people's emotions and situations.

[0366] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes processing means including multimodal artificial intelligence that analyzes received video data in real time and detects abnormal behavior or criminal activity, means including an emotion engine that analyzes the user's video data and recognizes their emotional state, and means for complementing the detection results of abnormal behavior or criminal activity based on the detected emotional state. This enables more accurate detection of abnormal behavior that takes the user's emotional state into consideration and immediate response.

[0367] A "security camera" is a device installed for security purposes that captures video data in real time.

[0368] "Video data" is data that includes visual information captured by a security camera.

[0369] "Real-time analysis" refers to the process of analyzing video data immediately after it is received.

[0370] "Abnormal behavior" refers to behavior that deviates from normal patterns of behavior and may be potentially dangerous or criminal.

[0371] "Criminal activity" refers to any activity that violates the law and threatens security or public safety.

[0372] "Multimodal AI" is AI that integrates multiple data modalities (e.g., video, audio, text) to analyze and make decisions.

[0373] "Processing means" refers to a combination of software and hardware for analyzing received video data and detecting abnormal behavior or criminal activity.

[0374] An "emotion engine" is a system for analyzing a user's emotional state, and is a technology that analyzes based on facial expressions, voice, body movements, etc.

[0375] A "reporting means" is a system that has the function of sending notifications to administrators or security personnel when abnormal behavior or criminal activity is detected.

[0376] The "alert level" is an indicator of the degree of alertness that is set based on detected abnormal behavior and emotional state.

[0377] "Past data" refers to a collection of video data and emotion data that has been previously collected and analyzed.

[0378] "Predicting future crises and crimes" is the process of predicting potential future dangers and criminal acts based on past data.

[0379] The system of the present invention receives video data from security cameras, analyzes the data in real time to detect abnormal behavior or criminal activity, and recognizes the user's emotions and adjusts the alert level, thereby achieving rapid and highly accurate security measures.

[0380] System configuration

[0381] The system consists of the following main elements:

[0382] 1. Device:

[0383] Security camera: A device that captures video data in real time.

[0384] Edge device: Equipment that performs initial data processing.

[0385] 2. Server:

[0386] Processing server: Includes multimodal artificial intelligence (AI) that analyzes received video data and detects abnormal behavior or criminal activity.

[0387] Emotion engine: An engine that analyzes the user's emotional state.

[0388] Notification Server: A server that sends notifications to security personnel and administrators based on detection results.

[0389] 3. User:

[0390] Administrators and security personnel who use the system.

[0391] What the program does

[0392] The server analyzes video data received from security cameras and edge devices to detect abnormal behavior and criminal activity. Video data is captured and analyzed in real time using software such as OpenCV. Multimodal artificial intelligence (e.g., YourAnomalyDetectionLibrary) identifies patterns of abnormal behavior and detects anomalous behavior.

[0393] At the same time, an emotion engine (such as YourEmotionRecognitionLibrary) is used to analyze the user's emotional state. Emotional states are recognized from data such as facial expressions, voice, and body movements, and indicators such as stress, anxiety, and anger are derived. As a result, the detection results of abnormal behavior and criminal acts are supplemented, making it possible to set more accurate alert levels.

[0394] The notification server generates alerts based on detected abnormal behaviors and emotional states and sends real-time notifications to security personnel and administrators, ensuring a prompt response.

[0395] Specific examples

[0396] When the system of the present invention is implemented in a commercial facility, security cameras capture video data within the facility in real time, which is then transmitted to a central server via an edge device. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering the facility at the same time). At the same time, an emotion engine analyzes the emotional states of surrounding users (e.g., anxiety or fear) and detects the emotional states associated with the occurrence of abnormal behavior. Based on the results, a notification system sends an alert to the facility manager, who can immediately dispatch security guards.

[0397] Prompt Sentence Examples

[0398] Design a system that analyzes video data in real time to detect abnormal and criminal behavior, and also recognizes user emotions. Include the ability to generate and send instant notifications to security personnel based on the detection results. Demonstrate specific application examples in various locations (e.g., commercial facilities, public transportation, etc.).

[0399] As a result, the present invention realizes security measures that take into account the emotional state of the user, enabling highly accurate detection of abnormal behavior and rapid response.

[0400] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0401] Step 1:

[0402] Security camera video data capture

[0403] The device captures video data in real time using a security camera. The input is a continuous video stream from the security camera, and the output is the captured video data. This video data is sent to the edge device for initial processing.

[0404] Step 2:

[0405] Pre-processing and transfer of video data

[0406] The edge device of the terminal temporarily processes the captured video data and sends it to a local server. The input is the captured video data, and the output is the processed video data. The video data is compressed and encrypted before being transferred to the central server.

[0407] Step 3:

[0408] Video data decompression and analysis

[0409] The server decompresses the video data received on the central server and sends it to a multimodal artificial intelligence (AI) analysis module. The input is compressed and encrypted video data, and the output is video data ready for analysis. The AI ​​analysis module analyzes the video data in real time to detect abnormal behavior and criminal activity.

[0410] Step 4:

[0411] Detecting abnormal and criminal behavior

[0412] The server's AI analysis module analyzes the received video data and detects abnormal behavior such as a person falling or multiple people entering at the same time, as well as criminal activity. The input is the decompressed video data, and the output is the detection results.

[0413] Step 5:

[0414] Emotional Data Analysis

[0415] The server's emotion engine analyzes the user's facial expressions, voice, and body movements based on the video data to recognize the user's emotional state (e.g., anxiety, fear). The input is video data, and the output is emotional state data.

[0416] Step 6:

[0417] Integrated analysis of abnormal behavior and emotional states

[0418] The server integrates the abnormal behavior detection results from the AI ​​analysis module and the emotional state data from the emotion engine, and sets an alert level based on the occurrence of abnormal behavior and the associated emotional state. The input is the abnormal behavior detection results and emotional state data, and the output is the integrated analysis results.

[0419] Step 7:

[0420] Alert level notification

[0421] The server's notification system generates alerts based on the integrated analysis results and sends them to security personnel and administrators in real time. The input is the integrated analysis results and the output is an alert notification, enabling users to respond quickly.

[0422] Specifically, in the case of a commercial facility, for example, video data captured in step 1 is processed in step 2 to detect abnormal behavior or emotional states during specific times or locations, and an alert is sent to the administrator in step 7. Upon receiving this alert, the facility's security staff will quickly head to the scene.

[0423] The following prompts are available for the generative AI model:

[0424] "Design a system that analyzes video data in real time to detect abnormal and criminal behavior, and also recognizes user emotions. Include a function to instantly generate and send notifications to security personnel based on the detection results. Please also provide specific application examples in various locations (e.g., commercial facilities, public transportation, etc.)."

[0425] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0426] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0427] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0428] [Second embodiment]

[0429] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0430] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0431] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0432] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0433] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0434] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0435] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0436] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0437] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0438] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0439] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0440] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0441] This invention is a system that receives video data from security cameras, analyzes the data in real time, and detects and reports abnormal behavior and criminal activity. This system uses multimodal artificial intelligence (AI) to efficiently monitor large amounts of video data. It also has the ability to learn from past data and predict potential future crises. This improves safety while reducing costs and labor.

[0442] System configuration

[0443] Terminals: Security cameras and edge devices

[0444] Server: A processing server that analyzes and learns from video data, and a notification server that notifies administrators and users of the results.

[0445] Users: Administrators and security personnel who use the system.

[0446] What the program does

[0447] Receiving data from security cameras

[0448] The device receives real-time video data captured by the security camera. The edge device processes the data and sends it to a local server, which then compresses and encrypts the data and securely transfers it to a central server.

[0449] Real-time analysis of video data

[0450] The server decompresses the video data received on the central server and inputs it into the multimodal AI, which analyzes the video data in real time to detect abnormal behavior or criminal activity, such as detecting a person lying down or multiple people entering the same store at the same time.

[0451] Reporting when abnormal behavior is detected

[0452] When the server detects abnormal behavior or criminal activity, it immediately generates an alert and sends it to security personnel or administrators via the notification system. The alert is sent to each device, allowing administrators and security personnel to take the necessary action promptly.

[0453] Learning from past data and predicting future crises

[0454] The server periodically collects and saves past video data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal activity based on past behavioral patterns. For example, it can predict dangers at specific times or locations and notify administrators of the results.

[0455] Subscriptions and Maintenance

[0456] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. Furthermore, as part of system maintenance management, it performs regular system checks and self-diagnosis, and promptly notifies the maintenance team if any abnormalities are discovered.

[0457] Specific examples

[0458] For commercial facilities

[0459] The manager of a commercial facility installs the "system of the present invention." Security cameras capture video data within the facility in real time and send the data to a central server via terminals. Multimodal AI on the central server analyzes the video data and detects abnormal behavior (for example, multiple people entering the same store at the same time). Upon detection, the notification system sends an alert to the manager of the commercial facility, and a security guard is immediately dispatched, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week, and preventive measures can be taken by notifying the manager.

[0460] As a result, the present invention significantly improves safety in commercial facilities, reduces costs and labor, and realizes efficient crime prevention measures.

[0461] The processing flow will be explained below.

[0462] Processing steps for receiving data from security cameras and analyzing it in real time

[0463] Step 1:

[0464] The device captures video data captured by security cameras in real time, and metadata such as a timestamp and camera ID is attached to the video data.

[0465] Step 2:

[0466] The device sends the captured video data to an edge device or local server, which performs simple data processing to extract and compress the important parts.

[0467] Step 3:

[0468] The device encrypts the processed video data on a local server and securely transfers it to a central server.

[0469] Step 4:

[0470] The server decompresses the video data received by the central server and sends it to the AI ​​analysis module.

[0471] Step 5:

[0472] The server then begins real-time analysis of the received video data using multimodal AI, which analyzes people's movements and behaviors to detect patterns of abnormal or criminal behavior.

[0473] Step 6:

[0474] If the AI ​​analysis detects abnormal behavior or criminal activity, the server immediately generates an alert, which includes information such as the type of abnormality, the time of detection, and the camera ID.

[0475] Step 7:

[0476] The server sends alerts to administrators and users through a notification system, which can be sent via email, SMS, or a dedicated application.

[0477] Step 8:

[0478] Users receive alerts and can respond quickly to any detected abnormal or criminal activity (e.g., dispatch security personnel to the scene or notify the police).

[0479] Processing steps for learning from past data and predicting future crises

[0480] Step 1:

[0481] The server periodically collects video data from the past few years and stores it in a database.

[0482] Step 2:

[0483] The server preprocesses each video data, removes noise, and converts it into a format suitable for analysis.

[0484] Step 3:

[0485] The server inputs the preprocessed data into the multimodal AI and trains it to learn patterns of crime and abnormal behavior.

[0486] Step 4:

[0487] The server periodically provides new data to the AI ​​and updates the learning model.

[0488] Step 5:

[0489] The server uses the trained model to predict abnormal behavior and criminal acts that are likely to occur in the future.

[0490] Step 6:

[0491] The server notifies the administrator and user of the prediction results (e.g., recommended times and locations for increased security).

[0492] Subscription and Maintenance Process Steps

[0493] Step 1:

[0494] Users can sign up for a subscription through a web portal or a dedicated application.

[0495] Step 2:

[0496] The server receives new contract, renewal and cancellation information and stores it in a database.

[0497] Step 3:

[0498] The server connects and provides security cameras and services based on the contract (e.g., adding new cameras and updating existing cameras).

[0499] Step 4:

[0500] The server periodically schedules system self-diagnosis to check the operational status of security cameras and edge devices.

[0501] Step 5:

[0502] The terminal performs periodic self-diagnosis and notifies the server if an abnormality is detected.

[0503] Step 6:

[0504] If an abnormality is detected, the server will promptly notify the maintenance team and arrange for repair or replacement work.

[0505] The above is a specific operation of each processing step of the present invention.

[0506] Example 1

[0507] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0508] In recent years, the rise in crime and abnormal behavior has led to a growing need to strengthen security in commercial facilities and public spaces. Conventional security systems have difficulty detecting abnormal behavior in real time, and they rarely use past data to predict future crises. As a result, despite the significant cost and effort involved, effective security measures have yet to be implemented.

[0509] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0510] In this invention, the server includes: means for receiving video data from security cameras; means including an edge device that temporarily processes, compresses, and encrypts the received video data; means for transferring the compressed and encrypted data to a central server; means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity; means including a notification system that issues a report when abnormal behavior or criminal activity is detected; means for saving past video data and using it to predict future crises; and means for managing subscription contract information and performing system maintenance and management. This makes it possible to quickly detect abnormal behavior or criminal activity in real time and take effective reports and preventive measures.

[0511] A "security camera" is a photographic device for monitoring a specific area and capturing video data.

[0512] An "edge device" is a computing device that temporarily processes received data and performs complex calculations, data compression, and encryption locally.

[0513] A "local server" is an intermediate server that processes data received from edge devices and securely forwards it to the central server.

[0514] A "central server" is a high-performance computing environment for managing, analyzing, and storing large amounts of data.

[0515] "Multimodal artificial intelligence" is an artificial intelligence system that integrates and analyzes a variety of data sources (such as video, audio, and text) to perform highly accurate detection and prediction.

[0516] A "notification system" is a communication system that quickly notifies administrators and security personnel when abnormal behavior or criminal activity is detected.

[0517] "Historical video data" refers to video recordings previously collected and stored by security cameras.

[0518] "Crisis forecasting" is the process of predicting potential future dangers and risks based on past data and current conditions.

[0519] A "subscription contract" is a contract for using a specific service or function for a certain period of time, and manages the user's registration status and contract details.

[0520] "System maintenance management" is the process of regularly checking and maintaining a system to ensure that it always operates properly.

[0521] "Real-time analytics" is the process of analyzing data as it is generated and transmitted.

[0522] This invention is a system that receives video data from security cameras, analyzes the data in real time to detect abnormal behavior or criminal activity, and issues a report if necessary. This system uses multimodal artificial intelligence (AI) to efficiently monitor large amounts of video data. It also has the ability to learn from past data and predict potential future crises. This improves safety while reducing costs and labor.

[0523] System configuration

[0524] Terminals: Security cameras and edge devices

[0525] Server: A processing server that analyzes and learns from video data, and a notification server that notifies administrators and users of the results.

[0526] Users: Administrators and security personnel who use the system

[0527] Program processing

[0528] Receiving video data from security cameras

[0529] The terminal receives video data captured by security cameras in real time. The edge device temporarily processes this data and forwards it to a local server, which then compresses and encrypts the data and securely transmits it to a central server.

[0530] Video data analysis and abnormal behavior detection

[0531] The server decompresses the video data received on the central server and inputs it into a multimodal AI, which analyzes the video data in real time to detect abnormal behavior or criminal activity.

[0532] For example, AI can detect abnormal behavior such as:

[0533] A scene where someone has fallen

[0534] A situation where multiple people enter the same store at the same time

[0535] Reporting Procedures

[0536] If the server detects any abnormal or criminal activity, it immediately generates an alert and sends it to security personnel or administrators through the notification system, allowing them to respond quickly.

[0537] Learning from past data and predicting future crises

[0538] The server periodically collects and saves past video data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal activity based on past behavioral patterns.

[0539] For example, the AI ​​predicts:

[0540] Risks are higher at certain times of the day (e.g., 10 PM to 2 AM)

[0541] Behavioral patterns in specific locations

[0542] These predictions are communicated to administrators to help them take preventative measures.

[0543] Subscriptions and Maintenance

[0544] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. It also performs periodic system checks and self-diagnosis as part of system maintenance, and notifies the maintenance team if any abnormalities are discovered.

[0545] Specific examples

[0546] Shopping mall surveillance

[0547] The manager of a commercial facility installs the system of this invention. Security cameras capture video data within the facility in real time and send the data to a central server via terminals. Multimodal AI on the central server analyzes the video data and detects abnormal behavior (for example, multiple people entering the same store at the same time). When abnormal behavior is detected, the notification system sends an alert to the manager, allowing security guards to respond promptly. Furthermore, the system can predict dangers on specific days of the week and at specific times of the day based on past data and notify the manager so that preventative measures can be taken.

[0548] As a result, the present invention significantly improves safety in commercial facilities, reduces costs and labor, and realizes efficient crime prevention measures.

[0549] Example prompts to input to the generative AI model

[0550] I would like to develop a system that monitors video data from security cameras in real time and detects abnormal behavior. In particular, I need a function that can detect behavior such as a person falling or multiple people entering at the same time and notify an administrator with an alert. I would like to know examples of multimodal AI models suitable for this purpose and the latest technologies.

[0551] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0552] Step 1: Receiving video data from security cameras

[0553] The terminal receives video data captured by security cameras in real time. The input is raw video frames. This video data is sent to the edge device and temporarily stored in a buffer. The edge device then quickly processes the data and prepares it for the next step.

[0554] Step 2: Compress and encrypt the data

[0555] The local server compresses the video data received from the edge device. The input is raw video data. A compression algorithm (e.g., H.264 or HEVC) is used to reduce the data size for efficient transmission. The data is then encrypted using an encryption algorithm such as AES-256. The output is compressed and encrypted data. This process reduces the risk of unauthorized access to the data during transmission.

[0556] Step 3: Send data to a central server

[0557] The local server sends compressed and encrypted data to the central server. The input is the compressed and encrypted video data. The output is secure data sent to the central server. This procedure uses a secure protocol (e.g., HTTPS or TLS) to send the data.

[0558] Step 4: Unpack and analyze the data

[0559] The central server decompresses the received compressed and encrypted data and restores it to the original video data. The input is the compressed and encrypted data. After decompression, this data is fed into the multimodal AI, which analyzes the video data in real time to detect abnormal behavior and criminal activity. The output is the analyzed results. This analysis step uses deep learning and machine learning algorithms to identify abnormal patterns in the video.

[0560] Step 5: Detecting Anomalous Behavior

[0561] Multimodal AI on a central server analyzes video data to detect abnormal behavior or criminal activity. The input is decompressed video data. The AI ​​detects, for example, a person lying on the ground or multiple people entering the same store at the same time. The output is information that abnormal behavior or criminal activity has been detected.

[0562] Step 6: Alerting and Notification

[0563] The central server immediately generates an alert if abnormal behavior or criminal activity is detected. The input is the abnormal behavior detection information from the multimodal AI. It then sends a notification to security personnel or administrators via the notification system. The output is an alert sent to each terminal or notification device, allowing administrators and security personnel to respond quickly.

[0564] Step 7: Collect and store historical data

[0565] The central server periodically collects past video data and stores it in a database. The input is the video data received and analyzed in real time, along with the analysis results. The stored data is used for subsequent learning and analysis. The output is the past video data stored in the database.

[0566] Step 8: Anticipate future crises

[0567] The central server predicts future crises and criminal activity based on stored historical video data. The input is the learning results of past data. Multimodal AI analyzes past behavioral patterns and predicts potential dangers that may occur at specific times and locations. The output is predicted crisis information. This information is notified to administrators, and appropriate preventive measures are taken.

[0568] Step 9: Subscriptions and Maintenance

[0569] The central server manages subscription contract information and processes new contracts, renewals, and cancellations for users. The input is the user's subscription information. In addition, as part of system maintenance, it performs periodic system checks and self-diagnosis, and if an abnormality is discovered, it promptly notifies the maintenance team. The output is the current subscription status and the normal operating status of the system.

[0570] keyword:

[0571] Generative AI model, prompt sentence

[0572] (Application example 1)

[0573] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0574] Commercial facilities and public spaces require a means to quickly and efficiently detect abnormal behavior and criminal activity and report it to the relevant parties. Conventional systems often do not analyze surveillance camera footage in real time, making it difficult to detect abnormal situations early. Furthermore, they lack the ability to use past data to predict future crises, making it difficult to implement preventative safety measures.

[0575] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0576] In this invention, the server includes a means for receiving video data from security cameras, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, a notification means for reporting when abnormal behavior or criminal activity is detected, a means for learning from past data to predict future crises and notifying of risks at specific times or locations, and a means for sending alerts to smartphones. This not only enables the rapid detection and reporting of abnormal behavior and criminal activity, but also enables the prediction of future crises, enabling more comprehensive security measures.

[0577] A "security camera" is a photographic device installed for crime prevention purposes to photograph a specific area.

[0578] The "means for receiving video data" refers to a device or system that receives video data transmitted from a security camera.

[0579] "Means for analyzing in real time" refers to a processing device or system that can analyze received video data immediately without delay.

[0580] "Abnormal or criminal behavior" refers to behavior that deviates from normal patterns of behavior and may be related to a crime.

[0581] "Multimodal AI" is an AI technology that can simultaneously analyze different types of data (e.g., video, audio, text).

[0582] "Means for reporting" refers to a means for quickly conveying information about detected abnormal behavior or criminal activity to relevant parties.

[0583] "Means for learning from past data to predict future crises" refers to a processing device or system for analyzing and predicting future risks using data recorded in the past.

[0584] "Means for notifying risks at specific times and locations" refers to means for communicating information about times and locations where predicted dangers are likely to occur to relevant parties.

[0585] "Means for sending alerts to smartphones" refers to means for notifying smartphones of detected abnormalities or predicted dangers.

[0586] System configuration

[0587] This invention is a system that receives video data from security cameras, analyzes the data in real time, and detects and reports abnormal behavior and criminal activity. The main components of the system are terminals, security cameras, edge devices, local servers, central servers, multimodal AI, notification servers, and smartphones.

[0588] Hardware used

[0589] Security cameras: High-resolution cameras (e.g., Sony IMX series)

[0590] Edge device: Nvidia Jetson Nano

[0591] Cloud Server: AWS EC2 instance

[0592] Database: PostgreSQL

[0593] Software used

[0594] Multimodal AI Models: Custom Models with TensorFlow

[0595] Encryption: AES encryption

[0596] API: REST API

[0597] Data processing flow

[0598] 1. The device receives video data captured by the security camera. The edge device temporarily processes this data, adjusting the frame rate and reducing noise.

[0599] 2. The edge device sends the processed data to a local server, where it is AES encrypted to ensure security.

[0600] 3. The local server transfers the encrypted data to the central server.

[0601] 4. The central server decompresses the received video data and inputs it into a multimodal artificial intelligence model using TensorFlow, which analyzes and detects abnormal or criminal behavior in real time.

[0602] 5. The notification server generates alerts immediately when abnormal or criminal behavior is detected and sends notifications to the smartphones of administrators and security personnel.

[0603] 6. The central server periodically collects and stores past video data, which is then used by the multimodal AI model to learn and predict future crises and criminal activity.

[0604] 7. The entire system manages subscription contract information, processes new contracts, renewals, and cancellations, and also periodically performs system checks and self-diagnosis, promptly notifying the maintenance team of any abnormalities.

[0605] Specific examples

[0606] For example, let's take a specific example of how this system might be implemented in a commercial facility. The facility manager installs security cameras in several locations. These cameras capture video data within the facility in real time and send the data via terminals to a central server. Multimodal AI on the central server analyzes the video data and detects abnormal behavior, such as multiple people entering a specific store at the same time. If detected, the notification system immediately sends an alert to the facility manager, enabling security guards to be dispatched immediately. Additionally, past data can be used to predict dangers on specific days and times of the day, and preventive measures can be taken by notifying the manager.

[0607] Prompt Sentence Examples

[0608] Analyze video data in real time to detect patterns of multiple people entering a specific area at once. When this happens, send an alert to the security guard's smartphone to notify them of any suspicious activity detected, and use past data to predict the time of day when the next suspicious activity is likely to occur.

[0609] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0610] Step 1:

[0611] The terminal receives video data captured by security cameras in real time. The edge device temporarily processes this data, adjusting the frame rate and reducing noise. Specifically, the video data is input and a noise reduction filter is applied to generate high-quality image data. The output is processed, high-quality video data.

[0612] Step 2:

[0613] The edge device sends the processed data to a local server. At this time, the data is AES encrypted. Specifically, the frame-rate-adjusted video data is input, and the AES encryption algorithm is applied to generate encrypted data. The output is the encrypted video data.

[0614] Step 3:

[0615] The local server transfers the encrypted data to the central server. This transfer uses a secure protocol (e.g. HTTPS). The input is the encrypted video data, which is sent to the central server using a data transfer protocol. The output is the encrypted data that arrives at the central server.

[0616] Step 4:

[0617] The central server decompresses the received video data and inputs it into a multimodal artificial intelligence model using TensorFlow. Specifically, encrypted data is input and the original video data is extracted by applying the AES decryption algorithm. The extracted video data is then input into the multimodal AI model. The output is the analysis result after the AI ​​model has performed its analysis.

[0618] Step 5:

[0619] The notification server immediately generates an alert if any abnormal or criminal behavior is detected and sends a notification to the smartphone of the administrator or security officer. Specifically, the analysis results are input and it determines whether or not any abnormal behavior has been detected. If an abnormality is detected, an alert message is generated and sent to the smartphone using a messaging protocol (e.g., FCM - Firebase Cloud Messaging). The output is an alert notification that is displayed on the smartphone of the administrator or security officer.

[0620] Step 6:

[0621] The central server periodically collects and stores past video data, and the multimodal AI model uses this data for learning. Specifically, video data is input at regular intervals and stored in a database. The AI ​​model is trained based on this data. The output is an updated AI model.

[0622] Step 7:

[0623] The entire system manages subscription contract information and processes new contracts, renewals, and cancellations. It also periodically performs system checks and self-diagnosis, and promptly notifies the maintenance team of any abnormalities. Specifically, user contract information is entered and saved / updated in the management database. If an abnormality is detected during regular self-diagnosis, the maintenance team is notified via a ticket system (e.g., JIRA). The output is the contract information update status and notification information for the maintenance team.

[0624] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0625] This invention combines a system that receives video data from security cameras, analyzes the data in real time to detect and report abnormal behavior and criminal activity, with an emotion engine that recognizes user emotions. This system utilizes multimodal artificial intelligence (AI) to efficiently monitor large volumes of video data. It also has the ability to learn from past data and predict potential future crises, as well as monitor the user's emotional state, enabling more accurate detection of abnormal behavior and immediate response.

[0626] System configuration

[0627] Terminals: Security cameras and edge devices

[0628] Server: A processing server that analyzes and learns video data and emotion data, and a notification server that reports the results to administrators and users.

[0629] Users: Administrators and security personnel who use the system.

[0630] What the program does

[0631] Receiving data from security cameras and analyzing it in real time

[0632] The device captures video data captured by security cameras in real time. The edge device processes the data and sends it to a local server, which then compresses and encrypts the data and securely transfers it to a central server.

[0633] The server decompresses the video data received on the central server and sends it to the AI ​​analysis module. Multimodal AI analyzes the video data in real time to detect abnormal behavior and criminal activity, such as detecting a person lying down or multiple people entering at the same time.

[0634] If the server detects abnormal behavior or criminal activity as a result of AI analysis, it immediately generates an alert and sends it to security personnel or administrators via the notification system. The alert is sent to each device, allowing administrators and security personnel to take the necessary action promptly.

[0635] Recognizing user emotions with an emotion engine

[0636] The server uses an emotion engine to analyze the user's video data and recognize their emotional state. The emotion engine analyzes the user's emotions based on facial expressions, voice, body movements, etc., and detects emotional states such as stress, anxiety, and anger.

[0637] The server complements the detection results of abnormal behavior and criminal activity based on the detected emotional state. For example, if a user shows strong anxiety or tension, it will focus on monitoring the surrounding environment. By integrating the analysis of emotional state and abnormal behavior patterns, it becomes possible to detect abnormal behavior with greater accuracy.

[0638] Learning from past data and predicting future crises

[0639] The server periodically collects past video and emotional data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal acts based on past behavioral patterns and emotional states. For example, it can predict dangers at specific times or locations and notify administrators of the results, allowing them to take preventative measures.

[0640] Subscriptions and Maintenance

[0641] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. Furthermore, as part of system maintenance management, it performs regular system checks and self-diagnosis, and promptly notifies the maintenance team if any abnormalities are discovered.

[0642] Specific examples

[0643] For commercial facilities

[0644] The manager of a commercial facility installs the "system of the present invention." Security cameras capture video data within the facility in real time and send the data to a central server via terminals. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering at the same time). Upon detection, an emotion engine analyzes the emotional state of surrounding users and detects the occurrence of abnormal behavior and associated emotional states (e.g., anxiety or fear). Based on this, a notification system sends an alert to the manager of the commercial facility, immediately dispatching security guards, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week, and preventive measures can be taken by notifying the manager.

[0645] As a result, the present invention significantly improves safety in commercial facilities and realizes comprehensive crime prevention measures that also take into account the emotional state of users.

[0646] The processing flow will be explained below.

[0647] Processing steps of a system that combines emotion engines

[0648] Processing steps for receiving data from security cameras and analyzing it in real time

[0649] Step 1:

[0650] The device captures video data captured by security cameras in real time, and metadata such as a timestamp and camera ID is attached to the video data.

[0651] Step 2:

[0652] The device sends the captured video data to an edge device or local server, which performs simple data processing to extract and compress the important parts.

[0653] Step 3:

[0654] The device encrypts the processed video data on a local server and securely transfers it to a central server.

[0655] Step 4:

[0656] The server decompresses the video data received by the central server and sends it to the AI ​​analysis module.

[0657] Step 5:

[0658] The server then begins real-time analysis of the received video data using multimodal AI, which analyzes people's movements and behaviors to detect patterns of abnormal or criminal behavior.

[0659] Step 6:

[0660] If the AI ​​analysis detects abnormal behavior or criminal activity, the server immediately generates an alert, which includes information such as the type of abnormality, the time of detection, and the camera ID.

[0661] Step 7:

[0662] The server sends alerts to administrators and users through a notification system, which can be sent via email, SMS, or a dedicated application.

[0663] Step 8:

[0664] Users receive alerts and can respond quickly to any detected abnormal or criminal activity (e.g., dispatch security personnel to the scene or notify the police).

[0665] Processing steps for recognizing user emotions by the emotion engine

[0666] Step 1:

[0667] The device uses security cameras or dedicated sensors to capture video data of the user, including the user's facial expressions and movements.

[0668] Step 2:

[0669] The device sends the captured video data to the edge device for initial processing, which then detects the user's face and extracts features for emotion analysis.

[0670] Step 3:

[0671] The terminal encrypts the processed data and sends it to the central server.

[0672] Step 4:

[0673] The server analyzes the received data using an emotion engine, which performs facial expression recognition, voice analysis, and body movement analysis to determine the user's emotional state.

[0674] Step 5:

[0675] When abnormal behavior or abnormal emotion is detected, the server integrates the abnormal behavior detection results with the emotion analysis results.

[0676] Step 6:

[0677] The server generates an alert including the results of the emotion analysis and notifies the administrator. Information about the emotional state (e.g., anxiety, stress) is added to the alert.

[0678] Processing steps for learning from past data and predicting future crises

[0679] Step 1:

[0680] The server periodically collects video data and emotion data from the past few years and stores them in a database.

[0681] Step 2:

[0682] The server preprocesses each video and emotion data, removes noise, and converts it into a format suitable for analysis.

[0683] Step 3:

[0684] The server inputs the preprocessed data into a multimodal AI and trains it to learn patterns of criminal and abnormal behavior and emotional states.

[0685] Step 4:

[0686] The server periodically provides new data to the AI ​​and updates the learning model.

[0687] Step 5:

[0688] The server uses the trained model to predict abnormal behavior and criminal acts that are likely to occur in the future.

[0689] Step 6:

[0690] The server notifies the administrator and user of the prediction results (e.g., recommended times and locations for increased security).

[0691] Subscription and Maintenance Process Steps

[0692] Step 1:

[0693] Users can sign up for a subscription through a web portal or a dedicated application.

[0694] Step 2:

[0695] The server receives new contract, renewal and cancellation information and stores it in a database.

[0696] Step 3:

[0697] The server connects and provides security cameras and services based on the contract (e.g., adding new cameras and updating existing cameras).

[0698] Step 4:

[0699] The server periodically schedules system self-diagnosis to check the operational status of security cameras and edge devices.

[0700] Step 5:

[0701] The terminal performs periodic self-diagnosis and notifies the server if an abnormality is detected.

[0702] Step 6:

[0703] If an abnormality is detected, the server will promptly notify the maintenance team and arrange for repair or replacement work.

[0704] As a result, the present invention significantly improves safety in commercial facilities and public places, and by taking into account the user's emotional state, it enables more precise and prompt responses.

[0705] Example 2

[0706] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0707] Conventional security systems have limitations in the monitoring and abnormal behavior detection capabilities of security cameras, making it difficult to detect abnormal behavior or criminal activity early and respond immediately. Furthermore, because they do not take the user's emotional state into account, they may miss potential dangers or stressful situations. Furthermore, systems that utilize past data to predict future risks are also inadequate, making it difficult to take effective preventative measures.

[0708] The identification processing by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for receiving video data from a security camera, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, a means for analyzing the user's emotional state using an emotion engine and complementing the abnormal behavior detection result, a notification means for issuing a report when abnormal behavior or criminal activity is detected, a means for collecting and learning past video data and emotion data, and a means for predicting future crises and criminal activity. This enables the security system to detect abnormal behavior or criminal activity with higher accuracy and respond immediately based on the analysis of emotion data. Furthermore, it becomes possible to predict future risks and take preventive measures by learning from past data.

[0709] A "security camera" is a video capture device installed to monitor fraudulent or abnormal behavior.

[0710] "Video data" refers to digital data that includes information about images and videos captured by security cameras.

[0711] "Reception" is the process of acquiring video data transmitted from a security camera.

[0712] "Real-time analysis" means processing and analyzing the received video data immediately.

[0713] "Abnormal behavior" refers to movements that deviate from normal patterns of behavior, and includes, for example, a person falling down, suddenly running, or roughly handling objects.

[0714] "Criminal activity" refers to conduct that violates the law and includes, for example, acts such as theft, assault, and trespass.

[0715] "Multimodal AI" is an AI technology that integrates and analyzes multiple data modalities (e.g., video, audio, text, etc.).

[0716] "Processing means" is a general term for devices and software for performing specific processes or functions.

[0717] An "emotion engine" is an algorithm or system for analyzing a user's emotional state, recognizing emotions based on facial expressions, voice, body movements, etc.

[0718] "Notification means" refers to a system or device that notifies security personnel or administrators of abnormal behavior or criminal activity when such behavior or criminal activity is detected.

[0719] "Data collection" is the process of systematically gathering information about visual and emotional states.

[0720] "Learning" is the process by which artificial intelligence acquires new knowledge based on past data and improves its performance.

[0721] "Predicting future crises and criminal acts" is the process of analyzing past data and trends to estimate possible future risks and criminal acts in advance.

[0722] This invention combines a system that receives video data from security cameras, analyzes the data in real time to detect and report abnormal behavior and criminal activity, with an emotion engine that recognizes user emotions. This system utilizes multimodal artificial intelligence (AI) to efficiently monitor large volumes of video data. It also has the ability to learn from past data and predict potential future crises, as well as monitor the user's emotional state, enabling more accurate detection of abnormal behavior and immediate response.

[0723] The system configuration is as follows:

[0724] Terminal

[0725] The devices include security cameras and edge devices. The security cameras capture video from designated surveillance areas 24 hours a day. The edge devices temporarily store the captured video data, then compress and encrypt it before sending it to a local server. The local server decompresses the data and securely transfers it to a central server.

[0726] server

[0727] The server has the following modules:

[0728] 1. The central server is a platform for real-time analysis of received video data. Multimodal AI runs on this server, analyzes the video data, and detects abnormal behavior and criminal activity.

[0729] 2. The emotion engine module analyzes the user's video data and recognizes their emotional state (e.g., stress, anxiety, anger), which complements and improves the accuracy of abnormal behavior detection.

[0730] 3. The notification system generates alerts immediately when abnormal or criminal behavior is detected and sends notifications to administrators and security personnel via various media such as email and SMS.

[0731] 4. The data collection and learning module periodically collects historical video and emotion data and stores it in a database. This data is then used by the multimodal AI to learn and predict future crises and criminal activity.

[0732] User

[0733] Users are administrators and security personnel who monitor and manage the system. They receive alerts from the notification system and can take immediate action. Administrators can also receive forecast information based on past data and take measures in advance.

[0734] Specific examples

[0735] For commercial facilities

[0736] The manager of a commercial facility installs the system of the present invention. Security cameras capture video data within the facility in real time and transmit the data to a central server via terminals. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering at the same time). Upon detection, an emotion engine analyzes the emotional states of surrounding users and detects emotional states (e.g., anxiety or fear) associated with the occurrence of abnormal behavior. Based on this, a notification system sends an alert to the manager of the commercial facility, and a security guard is immediately dispatched, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week and notify the manager, allowing preventive measures to be taken in advance.

[0737] Prompt Sentence Examples

[0738] "Please explain a system that receives video data from security cameras, analyzes it in real time to detect abnormal behavior or criminal activity, and also analyzes the user's emotional state and reports it."

[0739] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0740] Step 1:

[0741] The terminal captures video data in real time using security cameras. The real-time video information is input, and the video data sent to the edge device is generated as output. The security cameras operate 24 hours a day, capturing the movements of people and the placement of objects.

[0742] Step 2:

[0743] The edge device of the terminal temporarily stores the captured video data. The input is video data sent from the security camera, and the edge device compresses and encrypts the data and sends it to the local server. The compressed and encrypted data is then transferred to the local server as the output.

[0744] Step 3:

[0745] The server receives video data from the local server. The input is compressed and encrypted data, which the server decompresses and decrypts. The decompressed data is then passed to the AI ​​analysis module. The output is the decompressed and decrypted video data.

[0746] Step 4:

[0747] The server uses an AI analysis module to analyze the decompressed video data in real time. The input is the decompressed video data, and the multimodal AI analyzes people's movements and behavior patterns. For example, it analyzes people's dwell time and movements in a specific area. The output generates detection results for abnormal behavior and criminal activity.

[0748] Step 5:

[0749] When the server detects abnormal behavior or criminal activity, it immediately generates an alert. The input is the detection results from the AI ​​analysis module, and the generated alert is sent as output to the terminals of administrators and security personnel via the notification system. The alert is sent by email or SMS to prompt a response.

[0750] Step 6:

[0751] The server uses an emotion engine to analyze the user's emotional state. The input is video data, and the emotion engine recognizes emotions based on facial expressions, voice, body movements, etc. For example, data showing the user's anxiety or tension is analyzed. The output is the analyzed emotional state.

[0752] Step 7:

[0753] The server performs an integrated analysis of the emotion data obtained by the emotion engine and the abnormal behavior data from the AI ​​analysis module. The input is emotion data and abnormal behavior data, which complements the analysis results and improves accuracy. For example, if there is abnormal behavior in an area where anxious emotions are detected, further detailed monitoring will be performed. The output is the integrated analysis results.

[0754] Step 8:

[0755] The server periodically collects past video data and emotion data and stores them in a database. The input is the periodically collected video data and emotion data, which the multimodal AI uses to learn. The output is a trained AI model.

[0756] Step 9:

[0757] The server uses past data to predict future crises and criminal activity. The input is a trained AI model and past data, and it predicts risks, for example, during specific times of the day or on specific days of the week. The output is the predicted information, which is notified to an administrator so that countermeasures can be taken.

[0758] Step 10:

[0759] The server manages subscription contract information and performs system maintenance. The input is contract information, and it processes new contracts, renewals, and cancellations. It also performs system checks and self-diagnosis, and notifies the maintenance team if an abnormality is discovered. The output is the latest contract information and maintenance status.

[0760] (Application example 2)

[0761] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0762] Conventional security systems were able to detect abnormal behavior and criminal activity by analyzing video data, but because they did not take into account the emotional state of the user, there were limitations to the accuracy of predictions and the speed of response.Furthermore, there was an issue that simply detecting abnormal behavior made it difficult to respond flexibly according to people's emotions and situations.

[0763] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes processing means including multimodal artificial intelligence that analyzes received video data in real time and detects abnormal behavior or criminal activity, means including an emotion engine that analyzes the user's video data and recognizes their emotional state, and means for complementing the detection results of abnormal behavior or criminal activity based on the detected emotional state. This enables more accurate detection of abnormal behavior that takes the user's emotional state into consideration and immediate response.

[0764] A "security camera" is a device installed for security purposes that captures video data in real time.

[0765] "Video data" is data that includes visual information captured by a security camera.

[0766] "Real-time analysis" refers to the process of analyzing video data immediately after it is received.

[0767] "Abnormal behavior" refers to behavior that deviates from normal patterns of behavior and may be potentially dangerous or criminal.

[0768] "Criminal activity" refers to any activity that violates the law and threatens security or public safety.

[0769] "Multimodal AI" is AI that integrates multiple data modalities (e.g., video, audio, text) to analyze and make decisions.

[0770] "Processing means" refers to a combination of software and hardware for analyzing received video data and detecting abnormal behavior or criminal activity.

[0771] An "emotion engine" is a system for analyzing a user's emotional state, and is a technology that analyzes based on facial expressions, voice, body movements, etc.

[0772] A "reporting means" is a system that has the function of sending notifications to administrators or security personnel when abnormal behavior or criminal activity is detected.

[0773] The "alert level" is an indicator of the degree of alertness that is set based on detected abnormal behavior and emotional state.

[0774] "Past data" refers to a collection of video data and emotion data that has been previously collected and analyzed.

[0775] "Predicting future crises and crimes" is the process of predicting potential future dangers and criminal acts based on past data.

[0776] The system of the present invention receives video data from security cameras, analyzes the data in real time to detect abnormal behavior or criminal activity, and recognizes the user's emotions and adjusts the alert level, thereby achieving rapid and highly accurate security measures.

[0777] System configuration

[0778] The system consists of the following main elements:

[0779] 1. Device:

[0780] Security camera: A device that captures video data in real time.

[0781] Edge device: Equipment that performs initial data processing.

[0782] 2. Server:

[0783] Processing server: Includes multimodal artificial intelligence (AI) that analyzes received video data and detects abnormal behavior or criminal activity.

[0784] Emotion engine: An engine that analyzes the user's emotional state.

[0785] Notification Server: A server that sends notifications to security personnel and administrators based on detection results.

[0786] 3. User:

[0787] Administrators and security personnel who use the system.

[0788] What the program does

[0789] The server analyzes video data received from security cameras and edge devices to detect abnormal behavior and criminal activity. Video data is captured and analyzed in real time using software such as OpenCV. Multimodal artificial intelligence (e.g., YourAnomalyDetectionLibrary) identifies patterns of abnormal behavior and detects anomalous behavior.

[0790] At the same time, an emotion engine (such as YourEmotionRecognitionLibrary) is used to analyze the user's emotional state. Emotional states are recognized from data such as facial expressions, voice, and body movements, and indicators such as stress, anxiety, and anger are derived. As a result, the detection results of abnormal behavior and criminal acts are supplemented, making it possible to set more accurate alert levels.

[0791] The notification server generates alerts based on detected abnormal behaviors and emotional states and sends real-time notifications to security personnel and administrators, ensuring a prompt response.

[0792] Specific examples

[0793] When the system of the present invention is implemented in a commercial facility, security cameras capture video data within the facility in real time, which is then transmitted to a central server via an edge device. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering the facility at the same time). At the same time, an emotion engine analyzes the emotional states of surrounding users (e.g., anxiety or fear) and detects the emotional states associated with the occurrence of abnormal behavior. Based on the results, a notification system sends an alert to the facility manager, who can immediately dispatch security guards.

[0794] Prompt Sentence Examples

[0795] Design a system that analyzes video data in real time to detect abnormal and criminal behavior, and also recognizes user emotions. Include the ability to generate and send instant notifications to security personnel based on the detection results. Demonstrate specific application examples in various locations (e.g., commercial facilities, public transportation, etc.).

[0796] As a result, the present invention realizes security measures that take into account the emotional state of the user, enabling highly accurate detection of abnormal behavior and rapid response.

[0797] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0798] Step 1:

[0799] Security camera video data capture

[0800] The device captures video data in real time using a security camera. The input is a continuous video stream from the security camera, and the output is the captured video data. This video data is sent to the edge device for initial processing.

[0801] Step 2:

[0802] Pre-processing and transfer of video data

[0803] The edge device of the terminal temporarily processes the captured video data and sends it to a local server. The input is the captured video data, and the output is the processed video data. The video data is compressed and encrypted before being transferred to the central server.

[0804] Step 3:

[0805] Video data decompression and analysis

[0806] The server decompresses the video data received on the central server and sends it to a multimodal artificial intelligence (AI) analysis module. The input is compressed and encrypted video data, and the output is video data ready for analysis. The AI ​​analysis module analyzes the video data in real time to detect abnormal behavior and criminal activity.

[0807] Step 4:

[0808] Detecting abnormal and criminal behavior

[0809] The server's AI analysis module analyzes the received video data and detects abnormal behavior such as a person falling or multiple people entering at the same time, as well as criminal activity. The input is the decompressed video data, and the output is the detection results.

[0810] Step 5:

[0811] Emotional Data Analysis

[0812] The server's emotion engine analyzes the user's facial expressions, voice, and body movements based on the video data to recognize the user's emotional state (e.g., anxiety, fear). The input is video data, and the output is emotional state data.

[0813] Step 6:

[0814] Integrated analysis of abnormal behavior and emotional states

[0815] The server integrates the abnormal behavior detection results from the AI ​​analysis module and the emotional state data from the emotion engine, and sets an alert level based on the occurrence of abnormal behavior and the associated emotional state. The input is the abnormal behavior detection results and emotional state data, and the output is the integrated analysis results.

[0816] Step 7:

[0817] Alert level notification

[0818] The server's notification system generates alerts based on the integrated analysis results and sends them to security personnel and administrators in real time. The input is the integrated analysis results and the output is an alert notification, enabling users to respond quickly.

[0819] Specifically, in the case of a commercial facility, for example, video data captured in step 1 is processed in step 2 to detect abnormal behavior or emotional states during specific times or locations, and an alert is sent to the administrator in step 7. Upon receiving this alert, the facility's security staff will quickly head to the scene.

[0820] The following prompts are available for the generative AI model:

[0821] "Design a system that analyzes video data in real time to detect abnormal and criminal behavior, and also recognizes user emotions. Include a function to instantly generate and send notifications to security personnel based on the detection results. Please also provide specific application examples in various locations (e.g., commercial facilities, public transportation, etc.)."

[0822] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0823] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0824] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0825] [Third embodiment]

[0826] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0827] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0828] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0829] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0830] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0831] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0832] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0833] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0834] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0835] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0836] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0837] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0838] This invention is a system that receives video data from security cameras, analyzes the data in real time, and detects and reports abnormal behavior and criminal activity. This system uses multimodal artificial intelligence (AI) to efficiently monitor large amounts of video data. It also has the ability to learn from past data and predict potential future crises. This improves safety while reducing costs and labor.

[0839] System configuration

[0840] Terminals: Security cameras and edge devices

[0841] Server: A processing server that analyzes and learns from video data, and a notification server that notifies administrators and users of the results.

[0842] Users: Administrators and security personnel who use the system.

[0843] What the program does

[0844] Receiving data from security cameras

[0845] The device receives real-time video data captured by the security camera. The edge device processes the data and sends it to a local server, which then compresses and encrypts the data and securely transfers it to a central server.

[0846] Real-time analysis of video data

[0847] The server decompresses the video data received on the central server and inputs it into the multimodal AI, which analyzes the video data in real time to detect abnormal behavior or criminal activity, such as detecting a person lying down or multiple people entering the same store at the same time.

[0848] Reporting when abnormal behavior is detected

[0849] When the server detects abnormal behavior or criminal activity, it immediately generates an alert and sends it to security personnel or administrators via the notification system. The alert is sent to each device, allowing administrators and security personnel to take the necessary action promptly.

[0850] Learning from past data and predicting future crises

[0851] The server periodically collects and saves past video data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal activity based on past behavioral patterns. For example, it can predict dangers at specific times or locations and notify administrators of the results.

[0852] Subscriptions and Maintenance

[0853] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. Furthermore, as part of system maintenance management, it performs regular system checks and self-diagnosis, and promptly notifies the maintenance team if any abnormalities are discovered.

[0854] Specific examples

[0855] For commercial facilities

[0856] The manager of a commercial facility installs the "system of the present invention." Security cameras capture video data within the facility in real time and send the data to a central server via terminals. Multimodal AI on the central server analyzes the video data and detects abnormal behavior (for example, multiple people entering the same store at the same time). Upon detection, the notification system sends an alert to the manager of the commercial facility, and a security guard is immediately dispatched, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week, and preventive measures can be taken by notifying the manager.

[0857] As a result, the present invention significantly improves safety in commercial facilities, reduces costs and labor, and realizes efficient crime prevention measures.

[0858] The processing flow will be explained below.

[0859] Processing steps for receiving data from security cameras and analyzing it in real time

[0860] Step 1:

[0861] The device captures video data captured by security cameras in real time, and metadata such as a timestamp and camera ID is attached to the video data.

[0862] Step 2:

[0863] The device sends the captured video data to an edge device or local server, which performs simple data processing to extract and compress the important parts.

[0864] Step 3:

[0865] The device encrypts the processed video data on a local server and securely transfers it to a central server.

[0866] Step 4:

[0867] The server decompresses the video data received by the central server and sends it to the AI ​​analysis module.

[0868] Step 5:

[0869] The server then begins real-time analysis of the received video data using multimodal AI, which analyzes people's movements and behaviors to detect patterns of abnormal or criminal behavior.

[0870] Step 6:

[0871] If the AI ​​analysis detects abnormal behavior or criminal activity, the server immediately generates an alert, which includes information such as the type of abnormality, the time of detection, and the camera ID.

[0872] Step 7:

[0873] The server sends alerts to administrators and users through a notification system, which can be sent via email, SMS, or a dedicated application.

[0874] Step 8:

[0875] Users receive alerts and can respond quickly to any detected abnormal or criminal activity (e.g., dispatch security personnel to the scene or notify the police).

[0876] Processing steps for learning from past data and predicting future crises

[0877] Step 1:

[0878] The server periodically collects video data from the past few years and stores it in a database.

[0879] Step 2:

[0880] The server preprocesses each video data, removes noise, and converts it into a format suitable for analysis.

[0881] Step 3:

[0882] The server inputs the preprocessed data into the multimodal AI and trains it to learn patterns of crime and abnormal behavior.

[0883] Step 4:

[0884] The server periodically provides new data to the AI ​​and updates the learning model.

[0885] Step 5:

[0886] The server uses the trained model to predict abnormal behavior and criminal acts that are likely to occur in the future.

[0887] Step 6:

[0888] The server notifies the administrator and user of the prediction results (e.g., recommended times and locations for increased security).

[0889] Subscription and Maintenance Process Steps

[0890] Step 1:

[0891] Users can sign up for a subscription through a web portal or a dedicated application.

[0892] Step 2:

[0893] The server receives new contract, renewal and cancellation information and stores it in a database.

[0894] Step 3:

[0895] The server connects and provides security cameras and services based on the contract (e.g., adding new cameras and updating existing cameras).

[0896] Step 4:

[0897] The server periodically schedules system self-diagnosis to check the operational status of security cameras and edge devices.

[0898] Step 5:

[0899] The terminal performs periodic self-diagnosis and notifies the server if an abnormality is detected.

[0900] Step 6:

[0901] If an abnormality is detected, the server will promptly notify the maintenance team and arrange for repair or replacement work.

[0902] The above is a specific operation of each processing step of the present invention.

[0903] Example 1

[0904] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0905] In recent years, the rise in crime and abnormal behavior has led to a growing need to strengthen security in commercial facilities and public spaces. Conventional security systems have difficulty detecting abnormal behavior in real time, and they rarely use past data to predict future crises. As a result, despite the significant cost and effort involved, effective security measures have yet to be implemented.

[0906] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0907] In this invention, the server includes: means for receiving video data from security cameras; means including an edge device that temporarily processes, compresses, and encrypts the received video data; means for transferring the compressed and encrypted data to a central server; means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity; means including a notification system that issues a report when abnormal behavior or criminal activity is detected; means for saving past video data and using it to predict future crises; and means for managing subscription contract information and performing system maintenance and management. This makes it possible to quickly detect abnormal behavior or criminal activity in real time and take effective reports and preventive measures.

[0908] A "security camera" is a photographic device for monitoring a specific area and capturing video data.

[0909] An "edge device" is a computing device that temporarily processes received data and performs complex calculations, data compression, and encryption locally.

[0910] A "local server" is an intermediate server that processes data received from edge devices and securely forwards it to the central server.

[0911] A "central server" is a high-performance computing environment for managing, analyzing, and storing large amounts of data.

[0912] "Multimodal artificial intelligence" is an artificial intelligence system that integrates and analyzes a variety of data sources (such as video, audio, and text) to perform highly accurate detection and prediction.

[0913] A "notification system" is a communication system that quickly notifies administrators and security personnel when abnormal behavior or criminal activity is detected.

[0914] "Historical video data" refers to video recordings previously collected and stored by security cameras.

[0915] "Crisis forecasting" is the process of predicting potential future dangers and risks based on past data and current conditions.

[0916] A "subscription contract" is a contract for using a specific service or function for a certain period of time, and manages the user's registration status and contract details.

[0917] "System maintenance management" is the process of regularly checking and maintaining a system to ensure that it always operates properly.

[0918] "Real-time analytics" is the process of analyzing data as it is generated and transmitted.

[0919] This invention is a system that receives video data from security cameras, analyzes the data in real time to detect abnormal behavior or criminal activity, and issues a report if necessary. This system uses multimodal artificial intelligence (AI) to efficiently monitor large amounts of video data. It also has the ability to learn from past data and predict potential future crises. This improves safety while reducing costs and labor.

[0920] System configuration

[0921] Terminals: Security cameras and edge devices

[0922] Server: A processing server that analyzes and learns from video data, and a notification server that notifies administrators and users of the results.

[0923] Users: Administrators and security personnel who use the system

[0924] Program processing

[0925] Receiving video data from security cameras

[0926] The terminal receives video data captured by security cameras in real time. The edge device temporarily processes this data and forwards it to a local server, which then compresses and encrypts the data and securely transmits it to a central server.

[0927] Video data analysis and abnormal behavior detection

[0928] The server decompresses the video data received on the central server and inputs it into a multimodal AI, which analyzes the video data in real time to detect abnormal behavior or criminal activity.

[0929] For example, AI can detect abnormal behavior such as:

[0930] A scene where someone has fallen

[0931] A situation where multiple people enter the same store at the same time

[0932] Reporting Procedures

[0933] If the server detects any abnormal or criminal activity, it immediately generates an alert and sends it to security personnel or administrators through the notification system, allowing them to respond quickly.

[0934] Learning from past data and predicting future crises

[0935] The server periodically collects and saves past video data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal activity based on past behavioral patterns.

[0936] For example, the AI ​​predicts:

[0937] Risks are higher at certain times of the day (e.g., 10 PM to 2 AM)

[0938] Behavioral patterns in specific locations

[0939] These predictions are communicated to administrators to help them take preventative measures.

[0940] Subscriptions and Maintenance

[0941] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. It also performs periodic system checks and self-diagnosis as part of system maintenance, and notifies the maintenance team if any abnormalities are discovered.

[0942] Specific examples

[0943] Shopping mall surveillance

[0944] The manager of a commercial facility installs the system of this invention. Security cameras capture video data within the facility in real time and send the data to a central server via terminals. Multimodal AI on the central server analyzes the video data and detects abnormal behavior (for example, multiple people entering the same store at the same time). When abnormal behavior is detected, the notification system sends an alert to the manager, allowing security guards to respond promptly. Furthermore, the system can predict dangers on specific days of the week and at specific times of the day based on past data and notify the manager so that preventative measures can be taken.

[0945] As a result, the present invention significantly improves safety in commercial facilities, reduces costs and labor, and realizes efficient crime prevention measures.

[0946] Example prompts to input to the generative AI model

[0947] I would like to develop a system that monitors video data from security cameras in real time and detects abnormal behavior. In particular, I need a function that can detect behavior such as a person falling or multiple people entering at the same time and notify an administrator with an alert. I would like to know examples of multimodal AI models suitable for this purpose and the latest technologies.

[0948] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0949] Step 1: Receiving video data from security cameras

[0950] The terminal receives video data captured by security cameras in real time. The input is raw video frames. This video data is sent to the edge device and temporarily stored in a buffer. The edge device then quickly processes the data and prepares it for the next step.

[0951] Step 2: Compress and encrypt the data

[0952] The local server compresses the video data received from the edge device. The input is raw video data. A compression algorithm (e.g., H.264 or HEVC) is used to reduce the data size for efficient transmission. The data is then encrypted using an encryption algorithm such as AES-256. The output is compressed and encrypted data. This process reduces the risk of unauthorized access to the data during transmission.

[0953] Step 3: Send data to a central server

[0954] The local server sends compressed and encrypted data to the central server. The input is the compressed and encrypted video data. The output is secure data sent to the central server. This procedure uses a secure protocol (e.g., HTTPS or TLS) to send the data.

[0955] Step 4: Unpack and analyze the data

[0956] The central server decompresses the received compressed and encrypted data and restores it to the original video data. The input is the compressed and encrypted data. After decompression, this data is fed into the multimodal AI, which analyzes the video data in real time to detect abnormal behavior and criminal activity. The output is the analyzed results. This analysis step uses deep learning and machine learning algorithms to identify abnormal patterns in the video.

[0957] Step 5: Detecting Anomalous Behavior

[0958] Multimodal AI on a central server analyzes video data to detect abnormal behavior or criminal activity. The input is decompressed video data. The AI ​​detects, for example, a person lying on the ground or multiple people entering the same store at the same time. The output is information that abnormal behavior or criminal activity has been detected.

[0959] Step 6: Alerting and Notification

[0960] The central server immediately generates an alert if abnormal behavior or criminal activity is detected. The input is the abnormal behavior detection information from the multimodal AI. It then sends a notification to security personnel or administrators via the notification system. The output is an alert sent to each terminal or notification device, allowing administrators and security personnel to respond quickly.

[0961] Step 7: Collect and store historical data

[0962] The central server periodically collects past video data and stores it in a database. The input is the video data received and analyzed in real time, along with the analysis results. The stored data is used for subsequent learning and analysis. The output is the past video data stored in the database.

[0963] Step 8: Anticipate future crises

[0964] The central server predicts future crises and criminal activity based on stored historical video data. The input is the learning results of past data. Multimodal AI analyzes past behavioral patterns and predicts potential dangers that may occur at specific times and locations. The output is predicted crisis information. This information is notified to administrators, and appropriate preventive measures are taken.

[0965] Step 9: Subscriptions and Maintenance

[0966] The central server manages subscription contract information and processes new contracts, renewals, and cancellations for users. The input is the user's subscription information. In addition, as part of system maintenance, it performs periodic system checks and self-diagnosis, and if an abnormality is discovered, it promptly notifies the maintenance team. The output is the current subscription status and the normal operating status of the system.

[0967] keyword:

[0968] Generative AI model, prompt sentence

[0969] (Application example 1)

[0970] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0971] Commercial facilities and public spaces require a means to quickly and efficiently detect abnormal behavior and criminal activity and report it to the relevant parties. Conventional systems often do not analyze surveillance camera footage in real time, making it difficult to detect abnormal situations early. Furthermore, they lack the ability to use past data to predict future crises, making it difficult to implement preventative safety measures.

[0972] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0973] In this invention, the server includes a means for receiving video data from security cameras, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, a notification means for reporting when abnormal behavior or criminal activity is detected, a means for learning from past data to predict future crises and notifying of risks at specific times or locations, and a means for sending alerts to smartphones. This not only enables the rapid detection and reporting of abnormal behavior and criminal activity, but also enables the prediction of future crises, enabling more comprehensive security measures.

[0974] A "security camera" is a photographic device installed for crime prevention purposes to photograph a specific area.

[0975] The "means for receiving video data" refers to a device or system that receives video data transmitted from a security camera.

[0976] "Means for analyzing in real time" refers to a processing device or system that can analyze received video data immediately without delay.

[0977] "Abnormal or criminal behavior" refers to behavior that deviates from normal patterns of behavior and may be related to a crime.

[0978] "Multimodal AI" is an AI technology that can simultaneously analyze different types of data (e.g., video, audio, text).

[0979] "Means for reporting" refers to a means for quickly conveying information about detected abnormal behavior or criminal activity to relevant parties.

[0980] "Means for learning from past data to predict future crises" refers to a processing device or system for analyzing and predicting future risks using data recorded in the past.

[0981] "Means for notifying risks at specific times and locations" refers to means for communicating information about times and locations where predicted dangers are likely to occur to relevant parties.

[0982] "Means for sending alerts to smartphones" refers to means for notifying smartphones of detected abnormalities or predicted dangers.

[0983] System configuration

[0984] This invention is a system that receives video data from security cameras, analyzes the data in real time, and detects and reports abnormal behavior and criminal activity. The main components of the system are terminals, security cameras, edge devices, local servers, central servers, multimodal AI, notification servers, and smartphones.

[0985] Hardware used

[0986] Security cameras: High-resolution cameras (e.g., Sony IMX series)

[0987] Edge device: Nvidia Jetson Nano

[0988] Cloud Server: AWS EC2 instance

[0989] Database: PostgreSQL

[0990] Software used

[0991] Multimodal AI Models: Custom Models with TensorFlow

[0992] Encryption: AES encryption

[0993] API: REST API

[0994] Data processing flow

[0995] 1. The device receives video data captured by the security camera. The edge device temporarily processes this data, adjusting the frame rate and reducing noise.

[0996] 2. The edge device sends the processed data to a local server, where it is AES encrypted to ensure security.

[0997] 3. The local server transfers the encrypted data to the central server.

[0998] 4. The central server decompresses the received video data and inputs it into a multimodal artificial intelligence model using TensorFlow, which analyzes and detects abnormal or criminal behavior in real time.

[0999] 5. The notification server generates alerts immediately when abnormal or criminal behavior is detected and sends notifications to the smartphones of administrators and security personnel.

[1000] 6. The central server periodically collects and stores past video data, which is then used by the multimodal AI model to learn and predict future crises and criminal activity.

[1001] 7. The entire system manages subscription contract information, processes new contracts, renewals, and cancellations, and also periodically performs system checks and self-diagnosis, promptly notifying the maintenance team of any abnormalities.

[1002] Specific examples

[1003] For example, let's take a specific example of how this system might be implemented in a commercial facility. The facility manager installs security cameras in several locations. These cameras capture video data within the facility in real time and send the data via terminals to a central server. Multimodal AI on the central server analyzes the video data and detects abnormal behavior, such as multiple people entering a specific store at the same time. If detected, the notification system immediately sends an alert to the facility manager, enabling security guards to be dispatched immediately. Additionally, past data can be used to predict dangers on specific days and times of the day, and preventive measures can be taken by notifying the manager.

[1004] Prompt Sentence Examples

[1005] Analyze video data in real time to detect patterns of multiple people entering a specific area at once. When this happens, send an alert to the security guard's smartphone to notify them of any suspicious activity detected, and use past data to predict the time of day when the next suspicious activity is likely to occur.

[1006] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1007] Step 1:

[1008] The terminal receives video data captured by security cameras in real time. The edge device temporarily processes this data, adjusting the frame rate and reducing noise. Specifically, the video data is input and a noise reduction filter is applied to generate high-quality image data. The output is processed, high-quality video data.

[1009] Step 2:

[1010] The edge device sends the processed data to a local server. At this time, the data is AES encrypted. Specifically, the frame-rate-adjusted video data is input, and the AES encryption algorithm is applied to generate encrypted data. The output is the encrypted video data.

[1011] Step 3:

[1012] The local server transfers the encrypted data to the central server. This transfer uses a secure protocol (e.g. HTTPS). The input is the encrypted video data, which is sent to the central server using a data transfer protocol. The output is the encrypted data that arrives at the central server.

[1013] Step 4:

[1014] The central server decompresses the received video data and inputs it into a multimodal artificial intelligence model using TensorFlow. Specifically, encrypted data is input and the original video data is extracted by applying the AES decryption algorithm. The extracted video data is then input into the multimodal AI model. The output is the analysis result after the AI ​​model has performed its analysis.

[1015] Step 5:

[1016] The notification server immediately generates an alert if any abnormal or criminal behavior is detected and sends a notification to the smartphone of the administrator or security officer. Specifically, the analysis results are input and it determines whether or not any abnormal behavior has been detected. If an abnormality is detected, an alert message is generated and sent to the smartphone using a messaging protocol (e.g., FCM - Firebase Cloud Messaging). The output is an alert notification that is displayed on the smartphone of the administrator or security officer.

[1017] Step 6:

[1018] The central server periodically collects and stores past video data, and the multimodal AI model uses this data for learning. Specifically, video data is input at regular intervals and stored in a database. The AI ​​model is trained based on this data. The output is an updated AI model.

[1019] Step 7:

[1020] The entire system manages subscription contract information and processes new contracts, renewals, and cancellations. It also periodically performs system checks and self-diagnosis, and promptly notifies the maintenance team of any abnormalities. Specifically, user contract information is entered and saved / updated in the management database. If an abnormality is detected during regular self-diagnosis, the maintenance team is notified via a ticket system (e.g., JIRA). The output is the contract information update status and notification information for the maintenance team.

[1021] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1022] This invention combines a system that receives video data from security cameras, analyzes the data in real time to detect and report abnormal behavior and criminal activity, with an emotion engine that recognizes user emotions. This system utilizes multimodal artificial intelligence (AI) to efficiently monitor large volumes of video data. It also has the ability to learn from past data and predict potential future crises, as well as monitor the user's emotional state, enabling more accurate detection of abnormal behavior and immediate response.

[1023] System configuration

[1024] Terminals: Security cameras and edge devices

[1025] Server: A processing server that analyzes and learns video data and emotion data, and a notification server that reports the results to administrators and users.

[1026] Users: Administrators and security personnel who use the system.

[1027] What the program does

[1028] Receiving data from security cameras and analyzing it in real time

[1029] The device captures video data captured by security cameras in real time. The edge device processes the data and sends it to a local server, which then compresses and encrypts the data and securely transfers it to a central server.

[1030] The server decompresses the video data received on the central server and sends it to the AI ​​analysis module. Multimodal AI analyzes the video data in real time to detect abnormal behavior and criminal activity, such as detecting a person lying down or multiple people entering at the same time.

[1031] If the server detects abnormal behavior or criminal activity as a result of AI analysis, it immediately generates an alert and sends it to security personnel or administrators via the notification system. The alert is sent to each device, allowing administrators and security personnel to take the necessary action promptly.

[1032] Recognizing user emotions with an emotion engine

[1033] The server uses an emotion engine to analyze the user's video data and recognize their emotional state. The emotion engine analyzes the user's emotions based on facial expressions, voice, body movements, etc., and detects emotional states such as stress, anxiety, and anger.

[1034] The server complements the detection results of abnormal behavior and criminal activity based on the detected emotional state. For example, if a user shows strong anxiety or tension, it will focus on monitoring the surrounding environment. By integrating the analysis of emotional state and abnormal behavior patterns, it becomes possible to detect abnormal behavior with greater accuracy.

[1035] Learning from past data and predicting future crises

[1036] The server periodically collects past video and emotional data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal acts based on past behavioral patterns and emotional states. For example, it can predict dangers at specific times or locations and notify administrators of the results, allowing them to take preventative measures.

[1037] Subscriptions and Maintenance

[1038] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. Furthermore, as part of system maintenance management, it performs regular system checks and self-diagnosis, and promptly notifies the maintenance team if any abnormalities are discovered.

[1039] Specific examples

[1040] For commercial facilities

[1041] The manager of a commercial facility installs the "system of the present invention." Security cameras capture video data within the facility in real time and send the data to a central server via terminals. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering at the same time). Upon detection, an emotion engine analyzes the emotional state of surrounding users and detects the occurrence of abnormal behavior and associated emotional states (e.g., anxiety or fear). Based on this, a notification system sends an alert to the manager of the commercial facility, immediately dispatching security guards, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week, and preventive measures can be taken by notifying the manager.

[1042] As a result, the present invention significantly improves safety in commercial facilities and realizes comprehensive crime prevention measures that also take into account the emotional state of users.

[1043] The processing flow will be explained below.

[1044] Processing steps of a system that combines emotion engines

[1045] Processing steps for receiving data from security cameras and analyzing it in real time

[1046] Step 1:

[1047] The device captures video data captured by security cameras in real time, and metadata such as a timestamp and camera ID is attached to the video data.

[1048] Step 2:

[1049] The device sends the captured video data to an edge device or local server, which performs simple data processing to extract and compress the important parts.

[1050] Step 3:

[1051] The device encrypts the processed video data on a local server and securely transfers it to a central server.

[1052] Step 4:

[1053] The server decompresses the video data received by the central server and sends it to the AI ​​analysis module.

[1054] Step 5:

[1055] The server then begins real-time analysis of the received video data using multimodal AI, which analyzes people's movements and behaviors to detect patterns of abnormal or criminal behavior.

[1056] Step 6:

[1057] If the AI ​​analysis detects abnormal behavior or criminal activity, the server immediately generates an alert, which includes information such as the type of abnormality, the time of detection, and the camera ID.

[1058] Step 7:

[1059] The server sends alerts to administrators and users through a notification system, which can be sent via email, SMS, or a dedicated application.

[1060] Step 8:

[1061] Users receive alerts and can respond quickly to any detected abnormal or criminal activity (e.g., dispatch security personnel to the scene or notify the police).

[1062] Processing steps for recognizing user emotions by the emotion engine

[1063] Step 1:

[1064] The device uses security cameras or dedicated sensors to capture video data of the user, including the user's facial expressions and movements.

[1065] Step 2:

[1066] The device sends the captured video data to the edge device for initial processing, which then detects the user's face and extracts features for emotion analysis.

[1067] Step 3:

[1068] The terminal encrypts the processed data and sends it to the central server.

[1069] Step 4:

[1070] The server analyzes the received data using an emotion engine, which performs facial expression recognition, voice analysis, and body movement analysis to determine the user's emotional state.

[1071] Step 5:

[1072] When abnormal behavior or abnormal emotion is detected, the server integrates the abnormal behavior detection results with the emotion analysis results.

[1073] Step 6:

[1074] The server generates an alert including the results of the emotion analysis and notifies the administrator. Information about the emotional state (e.g., anxiety, stress) is added to the alert.

[1075] Processing steps for learning from past data and predicting future crises

[1076] Step 1:

[1077] The server periodically collects video data and emotion data from the past few years and stores them in a database.

[1078] Step 2:

[1079] The server preprocesses each video and emotion data, removes noise, and converts it into a format suitable for analysis.

[1080] Step 3:

[1081] The server inputs the preprocessed data into a multimodal AI and trains it to learn patterns of criminal and abnormal behavior and emotional states.

[1082] Step 4:

[1083] The server periodically provides new data to the AI ​​and updates the learning model.

[1084] Step 5:

[1085] The server uses the trained model to predict abnormal behavior and criminal acts that are likely to occur in the future.

[1086] Step 6:

[1087] The server notifies the administrator and user of the prediction results (e.g., recommended times and locations for increased security).

[1088] Subscription and Maintenance Process Steps

[1089] Step 1:

[1090] Users can sign up for a subscription through a web portal or a dedicated application.

[1091] Step 2:

[1092] The server receives new contract, renewal and cancellation information and stores it in a database.

[1093] Step 3:

[1094] The server connects and provides security cameras and services based on the contract (e.g., adding new cameras and updating existing cameras).

[1095] Step 4:

[1096] The server periodically schedules system self-diagnosis to check the operational status of security cameras and edge devices.

[1097] Step 5:

[1098] The terminal performs periodic self-diagnosis and notifies the server if an abnormality is detected.

[1099] Step 6:

[1100] If an abnormality is detected, the server will promptly notify the maintenance team and arrange for repair or replacement work.

[1101] As a result, the present invention significantly improves safety in commercial facilities and public places, and by taking into account the user's emotional state, it enables more precise and prompt responses.

[1102] Example 2

[1103] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1104] Conventional security systems have limitations in the monitoring and abnormal behavior detection capabilities of security cameras, making it difficult to detect abnormal behavior or criminal activity early and respond immediately. Furthermore, because they do not take the user's emotional state into account, they may miss potential dangers or stressful situations. Furthermore, systems that utilize past data to predict future risks are also inadequate, making it difficult to take effective preventative measures.

[1105] The identification processing by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for receiving video data from a security camera, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, a means for analyzing the user's emotional state using an emotion engine and complementing the abnormal behavior detection result, a notification means for issuing a report when abnormal behavior or criminal activity is detected, a means for collecting and learning past video data and emotion data, and a means for predicting future crises and criminal activity. This enables the security system to detect abnormal behavior or criminal activity with higher accuracy and respond immediately based on the analysis of emotion data. Furthermore, it becomes possible to predict future risks and take preventive measures by learning from past data.

[1106] A "security camera" is a video capture device installed to monitor fraudulent or abnormal behavior.

[1107] "Video data" refers to digital data that includes information about images and videos captured by security cameras.

[1108] "Reception" is the process of acquiring video data transmitted from a security camera.

[1109] "Real-time analysis" means processing and analyzing the received video data immediately.

[1110] "Abnormal behavior" refers to movements that deviate from normal patterns of behavior, and includes, for example, a person falling down, suddenly running, or roughly handling objects.

[1111] "Criminal activity" refers to conduct that violates the law and includes, for example, acts such as theft, assault, and trespass.

[1112] "Multimodal AI" is an AI technology that integrates and analyzes multiple data modalities (e.g., video, audio, text, etc.).

[1113] "Processing means" is a general term for devices and software for performing specific processes or functions.

[1114] An "emotion engine" is an algorithm or system for analyzing a user's emotional state, recognizing emotions based on facial expressions, voice, body movements, etc.

[1115] "Notification means" refers to a system or device that notifies security personnel or administrators of abnormal behavior or criminal activity when such behavior or criminal activity is detected.

[1116] "Data collection" is the process of systematically gathering information about visual and emotional states.

[1117] "Learning" is the process by which artificial intelligence acquires new knowledge based on past data and improves its performance.

[1118] "Predicting future crises and criminal acts" is the process of analyzing past data and trends to estimate possible future risks and criminal acts in advance.

[1119] This invention combines a system that receives video data from security cameras, analyzes the data in real time to detect and report abnormal behavior and criminal activity, with an emotion engine that recognizes user emotions. This system utilizes multimodal artificial intelligence (AI) to efficiently monitor large volumes of video data. It also has the ability to learn from past data and predict potential future crises, as well as monitor the user's emotional state, enabling more accurate detection of abnormal behavior and immediate response.

[1120] The system configuration is as follows:

[1121] Terminal

[1122] The devices include security cameras and edge devices. The security cameras capture video from designated surveillance areas 24 hours a day. The edge devices temporarily store the captured video data, then compress and encrypt it before sending it to a local server. The local server decompresses the data and securely transfers it to a central server.

[1123] server

[1124] The server has the following modules:

[1125] 1. The central server is a platform for real-time analysis of received video data. Multimodal AI runs on this server, analyzes the video data, and detects abnormal behavior and criminal activity.

[1126] 2. The emotion engine module analyzes the user's video data and recognizes their emotional state (e.g., stress, anxiety, anger), which complements and improves the accuracy of abnormal behavior detection.

[1127] 3. The notification system generates alerts immediately when abnormal or criminal behavior is detected and sends notifications to administrators and security personnel via various media such as email and SMS.

[1128] 4. The data collection and learning module periodically collects historical video and emotion data and stores it in a database. This data is then used by the multimodal AI to learn and predict future crises and criminal activity.

[1129] User

[1130] Users are administrators and security personnel who monitor and manage the system. They receive alerts from the notification system and can take immediate action. Administrators can also receive forecast information based on past data and take measures in advance.

[1131] Specific examples

[1132] For commercial facilities

[1133] The manager of a commercial facility installs the system of the present invention. Security cameras capture video data within the facility in real time and transmit the data to a central server via terminals. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering at the same time). Upon detection, an emotion engine analyzes the emotional states of surrounding users and detects emotional states (e.g., anxiety or fear) associated with the occurrence of abnormal behavior. Based on this, a notification system sends an alert to the manager of the commercial facility, and a security guard is immediately dispatched, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week and notify the manager, allowing preventive measures to be taken in advance.

[1134] Prompt Sentence Examples

[1135] "Please explain a system that receives video data from security cameras, analyzes it in real time to detect abnormal behavior or criminal activity, and also analyzes the user's emotional state and reports it."

[1136] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1137] Step 1:

[1138] The terminal captures video data in real time using security cameras. The real-time video information is input, and the video data sent to the edge device is generated as output. The security cameras operate 24 hours a day, capturing the movements of people and the placement of objects.

[1139] Step 2:

[1140] The edge device of the terminal temporarily stores the captured video data. The input is video data sent from the security camera, and the edge device compresses and encrypts the data and sends it to the local server. The compressed and encrypted data is then transferred to the local server as the output.

[1141] Step 3:

[1142] The server receives video data from the local server. The input is compressed and encrypted data, which the server decompresses and decrypts. The decompressed data is then passed to the AI ​​analysis module. The output is the decompressed and decrypted video data.

[1143] Step 4:

[1144] The server uses an AI analysis module to analyze the decompressed video data in real time. The input is the decompressed video data, and the multimodal AI analyzes people's movements and behavior patterns. For example, it analyzes people's dwell time and movements in a specific area. The output generates detection results for abnormal behavior and criminal activity.

[1145] Step 5:

[1146] When the server detects abnormal behavior or criminal activity, it immediately generates an alert. The input is the detection results from the AI ​​analysis module, and the generated alert is sent as output to the terminals of administrators and security personnel via the notification system. The alert is sent by email or SMS to prompt a response.

[1147] Step 6:

[1148] The server uses an emotion engine to analyze the user's emotional state. The input is video data, and the emotion engine recognizes emotions based on facial expressions, voice, body movements, etc. For example, data showing the user's anxiety or tension is analyzed. The output is the analyzed emotional state.

[1149] Step 7:

[1150] The server performs an integrated analysis of the emotion data obtained by the emotion engine and the abnormal behavior data from the AI ​​analysis module. The input is emotion data and abnormal behavior data, which complements the analysis results and improves accuracy. For example, if there is abnormal behavior in an area where anxious emotions are detected, further detailed monitoring will be performed. The output is the integrated analysis results.

[1151] Step 8:

[1152] The server periodically collects past video data and emotion data and stores them in a database. The input is the periodically collected video data and emotion data, which the multimodal AI uses to learn. The output is a trained AI model.

[1153] Step 9:

[1154] The server uses past data to predict future crises and criminal activity. The input is a trained AI model and past data, and it predicts risks, for example, during specific times of the day or on specific days of the week. The output is the predicted information, which is notified to an administrator so that countermeasures can be taken.

[1155] Step 10:

[1156] The server manages subscription contract information and performs system maintenance. The input is contract information, and it processes new contracts, renewals, and cancellations. It also performs system checks and self-diagnosis, and notifies the maintenance team if an abnormality is discovered. The output is the latest contract information and maintenance status.

[1157] (Application example 2)

[1158] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1159] Conventional security systems were able to detect abnormal behavior and criminal activity by analyzing video data, but because they did not take into account the emotional state of the user, there were limitations to the accuracy of predictions and the speed of response.Furthermore, there was an issue that simply detecting abnormal behavior made it difficult to respond flexibly according to people's emotions and situations.

[1160] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes processing means including multimodal artificial intelligence that analyzes received video data in real time and detects abnormal behavior or criminal activity, means including an emotion engine that analyzes the user's video data and recognizes their emotional state, and means for complementing the detection results of abnormal behavior or criminal activity based on the detected emotional state. This enables more accurate detection of abnormal behavior that takes the user's emotional state into consideration and immediate response.

[1161] A "security camera" is a device installed for security purposes that captures video data in real time.

[1162] "Video data" is data that includes visual information captured by a security camera.

[1163] "Real-time analysis" refers to the process of analyzing video data immediately after it is received.

[1164] "Abnormal behavior" refers to behavior that deviates from normal patterns of behavior and may be potentially dangerous or criminal.

[1165] "Criminal activity" refers to any activity that violates the law and threatens security or public safety.

[1166] "Multimodal AI" is AI that integrates multiple data modalities (e.g., video, audio, text) to analyze and make decisions.

[1167] "Processing means" refers to a combination of software and hardware for analyzing received video data and detecting abnormal behavior or criminal activity.

[1168] An "emotion engine" is a system for analyzing a user's emotional state, and is a technology that analyzes based on facial expressions, voice, body movements, etc.

[1169] A "reporting means" is a system that has the function of sending notifications to administrators or security personnel when abnormal behavior or criminal activity is detected.

[1170] The "alert level" is an indicator of the degree of alertness that is set based on detected abnormal behavior and emotional state.

[1171] "Past data" refers to a collection of video data and emotion data that has been previously collected and analyzed.

[1172] "Predicting future crises and crimes" is the process of predicting potential future dangers and criminal acts based on past data.

[1173] The system of the present invention receives video data from security cameras, analyzes the data in real time to detect abnormal behavior or criminal activity, and recognizes the user's emotions and adjusts the alert level, thereby achieving rapid and highly accurate security measures.

[1174] System configuration

[1175] The system consists of the following main elements:

[1176] 1. Device:

[1177] Security camera: A device that captures video data in real time.

[1178] Edge device: Equipment that performs initial data processing.

[1179] 2. Server:

[1180] Processing server: Includes multimodal artificial intelligence (AI) that analyzes received video data and detects abnormal behavior or criminal activity.

[1181] Emotion engine: An engine that analyzes the user's emotional state.

[1182] Notification Server: A server that sends notifications to security personnel and administrators based on detection results.

[1183] 3. User:

[1184] Administrators and security personnel who use the system.

[1185] What the program does

[1186] The server analyzes video data received from security cameras and edge devices to detect abnormal behavior and criminal activity. Video data is captured and analyzed in real time using software such as OpenCV. Multimodal artificial intelligence (e.g., YourAnomalyDetectionLibrary) identifies patterns of abnormal behavior and detects anomalous behavior.

[1187] At the same time, an emotion engine (such as YourEmotionRecognitionLibrary) is used to analyze the user's emotional state. Emotional states are recognized from data such as facial expressions, voice, and body movements, and indicators such as stress, anxiety, and anger are derived. As a result, the detection results of abnormal behavior and criminal acts are supplemented, making it possible to set more accurate alert levels.

[1188] The notification server generates alerts based on detected abnormal behaviors and emotional states and sends real-time notifications to security personnel and administrators, ensuring a prompt response.

[1189] Specific examples

[1190] When the system of the present invention is implemented in a commercial facility, security cameras capture video data within the facility in real time, which is then transmitted to a central server via an edge device. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering the facility at the same time). At the same time, an emotion engine analyzes the emotional states of surrounding users (e.g., anxiety or fear) and detects the emotional states associated with the occurrence of abnormal behavior. Based on the results, a notification system sends an alert to the facility manager, who can immediately dispatch security guards.

[1191] Prompt Sentence Examples

[1192] Design a system that analyzes video data in real time to detect abnormal and criminal behavior, and also recognizes user emotions. Include the ability to generate and send instant notifications to security personnel based on the detection results. Demonstrate specific application examples in various locations (e.g., commercial facilities, public transportation, etc.).

[1193] As a result, the present invention realizes security measures that take into account the emotional state of the user, enabling highly accurate detection of abnormal behavior and rapid response.

[1194] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1195] Step 1:

[1196] Security camera video data capture

[1197] The device captures video data in real time using a security camera. The input is a continuous video stream from the security camera, and the output is the captured video data. This video data is sent to the edge device for initial processing.

[1198] Step 2:

[1199] Pre-processing and transfer of video data

[1200] The edge device of the terminal temporarily processes the captured video data and sends it to a local server. The input is the captured video data, and the output is the processed video data. The video data is compressed and encrypted before being transferred to the central server.

[1201] Step 3:

[1202] Video data decompression and analysis

[1203] The server decompresses the video data received on the central server and sends it to a multimodal artificial intelligence (AI) analysis module. The input is compressed and encrypted video data, and the output is video data ready for analysis. The AI ​​analysis module analyzes the video data in real time to detect abnormal behavior and criminal activity.

[1204] Step 4:

[1205] Detecting abnormal and criminal behavior

[1206] The server's AI analysis module analyzes the received video data and detects abnormal behavior such as a person falling or multiple people entering at the same time, as well as criminal activity. The input is the decompressed video data, and the output is the detection results.

[1207] Step 5:

[1208] Emotional Data Analysis

[1209] The server's emotion engine analyzes the user's facial expressions, voice, and body movements based on the video data to recognize the user's emotional state (e.g., anxiety, fear). The input is video data, and the output is emotional state data.

[1210] Step 6:

[1211] Integrated analysis of abnormal behavior and emotional states

[1212] The server integrates the abnormal behavior detection results from the AI ​​analysis module and the emotional state data from the emotion engine, and sets an alert level based on the occurrence of abnormal behavior and the associated emotional state. The input is the abnormal behavior detection results and emotional state data, and the output is the integrated analysis results.

[1213] Step 7:

[1214] Alert level notification

[1215] The server's notification system generates alerts based on the integrated analysis results and sends them to security personnel and administrators in real time. The input is the integrated analysis results and the output is an alert notification, enabling users to respond quickly.

[1216] Specifically, in the case of a commercial facility, for example, video data captured in step 1 is processed in step 2 to detect abnormal behavior or emotional states during specific times or locations, and an alert is sent to the administrator in step 7. Upon receiving this alert, the facility's security staff will quickly head to the scene.

[1217] The following prompts are available for the generative AI model:

[1218] "Design a system that analyzes video data in real time to detect abnormal and criminal behavior, and also recognizes user emotions. Include a function to instantly generate and send notifications to security personnel based on the detection results. Please also provide specific application examples in various locations (e.g., commercial facilities, public transportation, etc.)."

[1219] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1220] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1221] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1222] [Fourth embodiment]

[1223] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1224] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1225] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1226] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1227] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1228] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1229] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1230] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1231] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1232] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1233] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1234] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1235] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1236] This invention is a system that receives video data from security cameras, analyzes the data in real time, and detects and reports abnormal behavior and criminal activity. This system uses multimodal artificial intelligence (AI) to efficiently monitor large amounts of video data. It also has the ability to learn from past data and predict potential future crises. This improves safety while reducing costs and labor.

[1237] System configuration

[1238] Terminals: Security cameras and edge devices

[1239] Server: A processing server that analyzes and learns from video data, and a notification server that notifies administrators and users of the results.

[1240] Users: Administrators and security personnel who use the system.

[1241] What the program does

[1242] Receiving data from security cameras

[1243] The device receives real-time video data captured by the security camera. The edge device processes the data and sends it to a local server, which then compresses and encrypts the data and securely transfers it to a central server.

[1244] Real-time analysis of video data

[1245] The server decompresses the video data received on the central server and inputs it into the multimodal AI, which analyzes the video data in real time to detect abnormal behavior or criminal activity, such as detecting a person lying down or multiple people entering the same store at the same time.

[1246] Reporting when abnormal behavior is detected

[1247] When the server detects abnormal behavior or criminal activity, it immediately generates an alert and sends it to security personnel or administrators via the notification system. The alert is sent to each device, allowing administrators and security personnel to take the necessary action promptly.

[1248] Learning from past data and predicting future crises

[1249] The server periodically collects and saves past video data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal activity based on past behavioral patterns. For example, it can predict dangers at specific times or locations and notify administrators of the results.

[1250] Subscriptions and Maintenance

[1251] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. Furthermore, as part of system maintenance management, it performs regular system checks and self-diagnosis, and promptly notifies the maintenance team if any abnormalities are discovered.

[1252] Specific examples

[1253] For commercial facilities

[1254] The manager of a commercial facility installs the "system of the present invention." Security cameras capture video data within the facility in real time and send the data to a central server via terminals. Multimodal AI on the central server analyzes the video data and detects abnormal behavior (for example, multiple people entering the same store at the same time). Upon detection, the notification system sends an alert to the manager of the commercial facility, and a security guard is immediately dispatched, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week, and preventive measures can be taken by notifying the manager.

[1255] As a result, the present invention significantly improves safety in commercial facilities, reduces costs and labor, and realizes efficient crime prevention measures.

[1256] The processing flow will be explained below.

[1257] Processing steps for receiving data from security cameras and analyzing it in real time

[1258] Step 1:

[1259] The device captures video data captured by security cameras in real time, and metadata such as a timestamp and camera ID is attached to the video data.

[1260] Step 2:

[1261] The device sends the captured video data to an edge device or local server, which performs simple data processing to extract and compress the important parts.

[1262] Step 3:

[1263] The device encrypts the processed video data on a local server and securely transfers it to a central server.

[1264] Step 4:

[1265] The server decompresses the video data received by the central server and sends it to the AI ​​analysis module.

[1266] Step 5:

[1267] The server then begins real-time analysis of the received video data using multimodal AI, which analyzes people's movements and behaviors to detect patterns of abnormal or criminal behavior.

[1268] Step 6:

[1269] If the AI ​​analysis detects abnormal behavior or criminal activity, the server immediately generates an alert, which includes information such as the type of abnormality, the time of detection, and the camera ID.

[1270] Step 7:

[1271] The server sends alerts to administrators and users through a notification system, which can be sent via email, SMS, or a dedicated application.

[1272] Step 8:

[1273] Users receive alerts and can respond quickly to any detected abnormal or criminal activity (e.g., dispatch security personnel to the scene or notify the police).

[1274] Processing steps for learning from past data and predicting future crises

[1275] Step 1:

[1276] The server periodically collects video data from the past few years and stores it in a database.

[1277] Step 2:

[1278] The server preprocesses each video data, removes noise, and converts it into a format suitable for analysis.

[1279] Step 3:

[1280] The server inputs the preprocessed data into the multimodal AI and trains it to learn patterns of crime and abnormal behavior.

[1281] Step 4:

[1282] The server periodically provides new data to the AI ​​and updates the learning model.

[1283] Step 5:

[1284] The server uses the trained model to predict abnormal behavior and criminal acts that are likely to occur in the future.

[1285] Step 6:

[1286] The server notifies the administrator and user of the prediction results (e.g., recommended times and locations for increased security).

[1287] Subscription and Maintenance Process Steps

[1288] Step 1:

[1289] Users can sign up for a subscription through a web portal or a dedicated application.

[1290] Step 2:

[1291] The server receives new contract, renewal and cancellation information and stores it in a database.

[1292] Step 3:

[1293] The server connects and provides security cameras and services based on the contract (e.g., adding new cameras and updating existing cameras).

[1294] Step 4:

[1295] The server periodically schedules system self-diagnosis to check the operational status of security cameras and edge devices.

[1296] Step 5:

[1297] The terminal performs periodic self-diagnosis and notifies the server if an abnormality is detected.

[1298] Step 6:

[1299] If an abnormality is detected, the server will promptly notify the maintenance team and arrange for repair or replacement work.

[1300] The above is a specific operation of each processing step of the present invention.

[1301] Example 1

[1302] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1303] In recent years, the rise in crime and abnormal behavior has led to a growing need to strengthen security in commercial facilities and public spaces. Conventional security systems have difficulty detecting abnormal behavior in real time, and they rarely use past data to predict future crises. As a result, despite the significant cost and effort involved, effective security measures have yet to be implemented.

[1304] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1305] In this invention, the server includes: means for receiving video data from security cameras; means including an edge device that temporarily processes, compresses, and encrypts the received video data; means for transferring the compressed and encrypted data to a central server; means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity; means including a notification system that issues a report when abnormal behavior or criminal activity is detected; means for saving past video data and using it to predict future crises; and means for managing subscription contract information and performing system maintenance and management. This makes it possible to quickly detect abnormal behavior or criminal activity in real time and take effective reports and preventive measures.

[1306] A "security camera" is a photographic device for monitoring a specific area and capturing video data.

[1307] An "edge device" is a computing device that temporarily processes received data and performs complex calculations, data compression, and encryption locally.

[1308] A "local server" is an intermediate server that processes data received from edge devices and securely forwards it to the central server.

[1309] A "central server" is a high-performance computing environment for managing, analyzing, and storing large amounts of data.

[1310] "Multimodal artificial intelligence" is an artificial intelligence system that integrates and analyzes a variety of data sources (such as video, audio, and text) to perform highly accurate detection and prediction.

[1311] A "notification system" is a communication system that quickly notifies administrators and security personnel when abnormal behavior or criminal activity is detected.

[1312] "Historical video data" refers to video recordings previously collected and stored by security cameras.

[1313] "Crisis forecasting" is the process of predicting potential future dangers and risks based on past data and current conditions.

[1314] A "subscription contract" is a contract for using a specific service or function for a certain period of time, and manages the user's registration status and contract details.

[1315] "System maintenance management" is the process of regularly checking and maintaining a system to ensure that it always operates properly.

[1316] "Real-time analytics" is the process of analyzing data as it is generated and transmitted.

[1317] This invention is a system that receives video data from security cameras, analyzes the data in real time to detect abnormal behavior or criminal activity, and issues a report if necessary. This system uses multimodal artificial intelligence (AI) to efficiently monitor large amounts of video data. It also has the ability to learn from past data and predict potential future crises. This improves safety while reducing costs and labor.

[1318] System configuration

[1319] Terminals: Security cameras and edge devices

[1320] Server: A processing server that analyzes and learns from video data, and a notification server that notifies administrators and users of the results.

[1321] Users: Administrators and security personnel who use the system

[1322] Program processing

[1323] Receiving video data from security cameras

[1324] The terminal receives video data captured by security cameras in real time. The edge device temporarily processes this data and forwards it to a local server, which then compresses and encrypts the data and securely transmits it to a central server.

[1325] Video data analysis and abnormal behavior detection

[1326] The server decompresses the video data received on the central server and inputs it into a multimodal AI, which analyzes the video data in real time to detect abnormal behavior or criminal activity.

[1327] For example, AI can detect abnormal behavior such as:

[1328] A scene where someone has fallen

[1329] A situation where multiple people enter the same store at the same time

[1330] Reporting Procedures

[1331] If the server detects any abnormal or criminal activity, it immediately generates an alert and sends it to security personnel or administrators through the notification system, allowing them to respond quickly.

[1332] Learning from past data and predicting future crises

[1333] The server periodically collects and saves past video data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal activity based on past behavioral patterns.

[1334] For example, the AI ​​predicts:

[1335] Risks are higher at certain times of the day (e.g., 10 PM to 2 AM)

[1336] Behavioral patterns in specific locations

[1337] These predictions are communicated to administrators to help them take preventative measures.

[1338] Subscriptions and Maintenance

[1339] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. It also performs periodic system checks and self-diagnosis as part of system maintenance, and notifies the maintenance team if any abnormalities are discovered.

[1340] Specific examples

[1341] Shopping mall surveillance

[1342] The manager of a commercial facility installs the system of this invention. Security cameras capture video data within the facility in real time and send the data to a central server via terminals. Multimodal AI on the central server analyzes the video data and detects abnormal behavior (for example, multiple people entering the same store at the same time). When abnormal behavior is detected, the notification system sends an alert to the manager, allowing security guards to respond promptly. Furthermore, the system can predict dangers on specific days of the week and at specific times of the day based on past data and notify the manager so that preventative measures can be taken.

[1343] As a result, the present invention significantly improves safety in commercial facilities, reduces costs and labor, and realizes efficient crime prevention measures.

[1344] Example prompts to input to the generative AI model

[1345] I would like to develop a system that monitors video data from security cameras in real time and detects abnormal behavior. In particular, I need a function that can detect behavior such as a person falling or multiple people entering at the same time and notify an administrator with an alert. I would like to know examples of multimodal AI models suitable for this purpose and the latest technologies.

[1346] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1347] Step 1: Receiving video data from security cameras

[1348] The terminal receives video data captured by security cameras in real time. The input is raw video frames. This video data is sent to the edge device and temporarily stored in a buffer. The edge device then quickly processes the data and prepares it for the next step.

[1349] Step 2: Compress and encrypt the data

[1350] The local server compresses the video data received from the edge device. The input is raw video data. A compression algorithm (e.g., H.264 or HEVC) is used to reduce the data size for efficient transmission. The data is then encrypted using an encryption algorithm such as AES-256. The output is compressed and encrypted data. This process reduces the risk of unauthorized access to the data during transmission.

[1351] Step 3: Send data to a central server

[1352] The local server sends compressed and encrypted data to the central server. The input is the compressed and encrypted video data. The output is secure data sent to the central server. This procedure uses a secure protocol (e.g., HTTPS or TLS) to send the data.

[1353] Step 4: Unpack and analyze the data

[1354] The central server decompresses the received compressed and encrypted data and restores it to the original video data. The input is the compressed and encrypted data. After decompression, this data is fed into the multimodal AI, which analyzes the video data in real time to detect abnormal behavior and criminal activity. The output is the analyzed results. This analysis step uses deep learning and machine learning algorithms to identify abnormal patterns in the video.

[1355] Step 5: Detecting Anomalous Behavior

[1356] Multimodal AI on a central server analyzes video data to detect abnormal behavior or criminal activity. The input is decompressed video data. The AI ​​detects, for example, a person lying on the ground or multiple people entering the same store at the same time. The output is information that abnormal behavior or criminal activity has been detected.

[1357] Step 6: Alerting and Notification

[1358] The central server immediately generates an alert if abnormal behavior or criminal activity is detected. The input is the abnormal behavior detection information from the multimodal AI. It then sends a notification to security personnel or administrators via the notification system. The output is an alert sent to each terminal or notification device, allowing administrators and security personnel to respond quickly.

[1359] Step 7: Collect and store historical data

[1360] The central server periodically collects past video data and stores it in a database. The input is the video data received and analyzed in real time, along with the analysis results. The stored data is used for subsequent learning and analysis. The output is the past video data stored in the database.

[1361] Step 8: Anticipate future crises

[1362] The central server predicts future crises and criminal activity based on stored historical video data. The input is the learning results of past data. Multimodal AI analyzes past behavioral patterns and predicts potential dangers that may occur at specific times and locations. The output is predicted crisis information. This information is notified to administrators, and appropriate preventive measures are taken.

[1363] Step 9: Subscriptions and Maintenance

[1364] The central server manages subscription contract information and processes new contracts, renewals, and cancellations for users. The input is the user's subscription information. In addition, as part of system maintenance, it performs periodic system checks and self-diagnosis, and if an abnormality is discovered, it promptly notifies the maintenance team. The output is the current subscription status and the normal operating status of the system.

[1365] keyword:

[1366] Generative AI model, prompt sentence

[1367] (Application example 1)

[1368] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1369] Commercial facilities and public spaces require a means to quickly and efficiently detect abnormal behavior and criminal activity and report it to the relevant parties. Conventional systems often do not analyze surveillance camera footage in real time, making it difficult to detect abnormal situations early. Furthermore, they lack the ability to use past data to predict future crises, making it difficult to implement preventative safety measures.

[1370] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1371] In this invention, the server includes a means for receiving video data from security cameras, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, a notification means for reporting when abnormal behavior or criminal activity is detected, a means for learning from past data to predict future crises and notifying of risks at specific times or locations, and a means for sending alerts to smartphones. This not only enables the rapid detection and reporting of abnormal behavior and criminal activity, but also enables the prediction of future crises, enabling more comprehensive security measures.

[1372] A "security camera" is a photographic device installed for crime prevention purposes to photograph a specific area.

[1373] The "means for receiving video data" refers to a device or system that receives video data transmitted from a security camera.

[1374] "Means for analyzing in real time" refers to a processing device or system that can analyze received video data immediately without delay.

[1375] "Abnormal or criminal behavior" refers to behavior that deviates from normal patterns of behavior and may be related to a crime.

[1376] "Multimodal AI" is an AI technology that can simultaneously analyze different types of data (e.g., video, audio, text).

[1377] "Means for reporting" refers to a means for quickly conveying information about detected abnormal behavior or criminal activity to relevant parties.

[1378] "Means for learning from past data to predict future crises" refers to a processing device or system for analyzing and predicting future risks using data recorded in the past.

[1379] "Means for notifying risks at specific times and locations" refers to means for communicating information about times and locations where predicted dangers are likely to occur to relevant parties.

[1380] "Means for sending alerts to smartphones" refers to means for notifying smartphones of detected abnormalities or predicted dangers.

[1381] System configuration

[1382] This invention is a system that receives video data from security cameras, analyzes the data in real time, and detects and reports abnormal behavior and criminal activity. The main components of the system are terminals, security cameras, edge devices, local servers, central servers, multimodal AI, notification servers, and smartphones.

[1383] Hardware used

[1384] Security cameras: High-resolution cameras (e.g., Sony IMX series)

[1385] Edge device: Nvidia Jetson Nano

[1386] Cloud Server: AWS EC2 instance

[1387] Database: PostgreSQL

[1388] Software used

[1389] Multimodal AI Models: Custom Models with TensorFlow

[1390] Encryption: AES encryption

[1391] API: REST API

[1392] Data processing flow

[1393] 1. The device receives video data captured by the security camera. The edge device temporarily processes this data, adjusting the frame rate and reducing noise.

[1394] 2. The edge device sends the processed data to a local server, where it is AES encrypted to ensure security.

[1395] 3. The local server transfers the encrypted data to the central server.

[1396] 4. The central server decompresses the received video data and inputs it into a multimodal artificial intelligence model using TensorFlow, which analyzes and detects abnormal or criminal behavior in real time.

[1397] 5. The notification server generates alerts immediately when abnormal or criminal behavior is detected and sends notifications to the smartphones of administrators and security personnel.

[1398] 6. The central server periodically collects and stores past video data, which is then used by the multimodal AI model to learn and predict future crises and criminal activity.

[1399] 7. The entire system manages subscription contract information, processes new contracts, renewals, and cancellations, and also periodically performs system checks and self-diagnosis, promptly notifying the maintenance team of any abnormalities.

[1400] Specific examples

[1401] For example, let's take a specific example of how this system might be implemented in a commercial facility. The facility manager installs security cameras in several locations. These cameras capture video data within the facility in real time and send the data via terminals to a central server. Multimodal AI on the central server analyzes the video data and detects abnormal behavior, such as multiple people entering a specific store at the same time. If detected, the notification system immediately sends an alert to the facility manager, enabling security guards to be dispatched immediately. Additionally, past data can be used to predict dangers on specific days and times of the day, and preventive measures can be taken by notifying the manager.

[1402] Prompt Sentence Examples

[1403] Analyze video data in real time to detect patterns of multiple people entering a specific area at once. When this happens, send an alert to the security guard's smartphone to notify them of any suspicious activity detected, and use past data to predict the time of day when the next suspicious activity is likely to occur.

[1404] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1405] Step 1:

[1406] The terminal receives video data captured by security cameras in real time. The edge device temporarily processes this data, adjusting the frame rate and reducing noise. Specifically, the video data is input and a noise reduction filter is applied to generate high-quality image data. The output is processed, high-quality video data.

[1407] Step 2:

[1408] The edge device sends the processed data to a local server. At this time, the data is AES encrypted. Specifically, the frame-rate-adjusted video data is input, and the AES encryption algorithm is applied to generate encrypted data. The output is the encrypted video data.

[1409] Step 3:

[1410] The local server transfers the encrypted data to the central server. This transfer uses a secure protocol (e.g. HTTPS). The input is the encrypted video data, which is sent to the central server using a data transfer protocol. The output is the encrypted data that arrives at the central server.

[1411] Step 4:

[1412] The central server decompresses the received video data and inputs it into a multimodal artificial intelligence model using TensorFlow. Specifically, encrypted data is input and the original video data is extracted by applying the AES decryption algorithm. The extracted video data is then input into the multimodal AI model. The output is the analysis result after the AI ​​model has performed its analysis.

[1413] Step 5:

[1414] The notification server immediately generates an alert if any abnormal or criminal behavior is detected and sends a notification to the smartphone of the administrator or security officer. Specifically, the analysis results are input and it determines whether or not any abnormal behavior has been detected. If an abnormality is detected, an alert message is generated and sent to the smartphone using a messaging protocol (e.g., FCM - Firebase Cloud Messaging). The output is an alert notification that is displayed on the smartphone of the administrator or security officer.

[1415] Step 6:

[1416] The central server periodically collects and stores past video data, and the multimodal AI model uses this data for learning. Specifically, video data is input at regular intervals and stored in a database. The AI ​​model is trained based on this data. The output is an updated AI model.

[1417] Step 7:

[1418] The entire system manages subscription contract information and processes new contracts, renewals, and cancellations. It also periodically performs system checks and self-diagnosis, and promptly notifies the maintenance team of any abnormalities. Specifically, user contract information is entered and saved / updated in the management database. If an abnormality is detected during regular self-diagnosis, the maintenance team is notified via a ticket system (e.g., JIRA). The output is the contract information update status and notification information for the maintenance team.

[1419] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1420] This invention combines a system that receives video data from security cameras, analyzes the data in real time to detect and report abnormal behavior and criminal activity, with an emotion engine that recognizes user emotions. This system utilizes multimodal artificial intelligence (AI) to efficiently monitor large volumes of video data. It also has the ability to learn from past data and predict potential future crises, as well as monitor the user's emotional state, enabling more accurate detection of abnormal behavior and immediate response.

[1421] System configuration

[1422] Terminals: Security cameras and edge devices

[1423] Server: A processing server that analyzes and learns video data and emotion data, and a notification server that reports the results to administrators and users.

[1424] Users: Administrators and security personnel who use the system.

[1425] What the program does

[1426] Receiving data from security cameras and analyzing it in real time

[1427] The device captures video data captured by security cameras in real time. The edge device processes the data and sends it to a local server, which then compresses and encrypts the data and securely transfers it to a central server.

[1428] The server decompresses the video data received on the central server and sends it to the AI ​​analysis module. Multimodal AI analyzes the video data in real time to detect abnormal behavior and criminal activity, such as detecting a person lying down or multiple people entering at the same time.

[1429] If the server detects abnormal behavior or criminal activity as a result of AI analysis, it immediately generates an alert and sends it to security personnel or administrators via the notification system. The alert is sent to each device, allowing administrators and security personnel to take the necessary action promptly.

[1430] Recognizing user emotions with an emotion engine

[1431] The server uses an emotion engine to analyze the user's video data and recognize their emotional state. The emotion engine analyzes the user's emotions based on facial expressions, voice, body movements, etc., and detects emotional states such as stress, anxiety, and anger.

[1432] The server complements the detection results of abnormal behavior and criminal activity based on the detected emotional state. For example, if a user shows strong anxiety or tension, it will focus on monitoring the surrounding environment. By integrating the analysis of emotional state and abnormal behavior patterns, it becomes possible to detect abnormal behavior with greater accuracy.

[1433] Learning from past data and predicting future crises

[1434] The server periodically collects past video and emotional data and stores it in a database. This data is used by the multimodal AI to learn and predict future crises and criminal acts based on past behavioral patterns and emotional states. For example, it can predict dangers at specific times or locations and notify administrators of the results, allowing them to take preventative measures.

[1435] Subscriptions and Maintenance

[1436] The server manages subscription contract information and processes new contracts, renewals, and cancellations. It connects and provides security cameras and services based on the contract details. Furthermore, as part of system maintenance management, it performs regular system checks and self-diagnosis, and promptly notifies the maintenance team if any abnormalities are discovered.

[1437] Specific examples

[1438] For commercial facilities

[1439] The manager of a commercial facility installs the "system of the present invention." Security cameras capture video data within the facility in real time and send the data to a central server via terminals. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering at the same time). Upon detection, an emotion engine analyzes the emotional state of surrounding users and detects the occurrence of abnormal behavior and associated emotional states (e.g., anxiety or fear). Based on this, a notification system sends an alert to the manager of the commercial facility, immediately dispatching security guards, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week, and preventive measures can be taken by notifying the manager.

[1440] As a result, the present invention significantly improves safety in commercial facilities and realizes comprehensive crime prevention measures that also take into account the emotional state of users.

[1441] The processing flow will be explained below.

[1442] Processing steps of a system that combines emotion engines

[1443] Processing steps for receiving data from security cameras and analyzing it in real time

[1444] Step 1:

[1445] The device captures video data captured by security cameras in real time, and metadata such as a timestamp and camera ID is attached to the video data.

[1446] Step 2:

[1447] The device sends the captured video data to an edge device or local server, which performs simple data processing to extract and compress the important parts.

[1448] Step 3:

[1449] The device encrypts the processed video data on a local server and securely transfers it to a central server.

[1450] Step 4:

[1451] The server decompresses the video data received by the central server and sends it to the AI ​​analysis module.

[1452] Step 5:

[1453] The server then begins real-time analysis of the received video data using multimodal AI, which analyzes people's movements and behaviors to detect patterns of abnormal or criminal behavior.

[1454] Step 6:

[1455] If the AI ​​analysis detects abnormal behavior or criminal activity, the server immediately generates an alert, which includes information such as the type of abnormality, the time of detection, and the camera ID.

[1456] Step 7:

[1457] The server sends alerts to administrators and users through a notification system, which can be sent via email, SMS, or a dedicated application.

[1458] Step 8:

[1459] Users receive alerts and can respond quickly to any detected abnormal or criminal activity (e.g., dispatch security personnel to the scene or notify the police).

[1460] Processing steps for recognizing user emotions by the emotion engine

[1461] Step 1:

[1462] The device uses security cameras or dedicated sensors to capture video data of the user, including the user's facial expressions and movements.

[1463] Step 2:

[1464] The device sends the captured video data to the edge device for initial processing, which then detects the user's face and extracts features for emotion analysis.

[1465] Step 3:

[1466] The terminal encrypts the processed data and sends it to the central server.

[1467] Step 4:

[1468] The server analyzes the received data using an emotion engine, which performs facial expression recognition, voice analysis, and body movement analysis to determine the user's emotional state.

[1469] Step 5:

[1470] When abnormal behavior or abnormal emotion is detected, the server integrates the abnormal behavior detection results with the emotion analysis results.

[1471] Step 6:

[1472] The server generates an alert including the results of the emotion analysis and notifies the administrator. Information about the emotional state (e.g., anxiety, stress) is added to the alert.

[1473] Processing steps for learning from past data and predicting future crises

[1474] Step 1:

[1475] The server periodically collects video data and emotion data from the past few years and stores them in a database.

[1476] Step 2:

[1477] The server preprocesses each video and emotion data, removes noise, and converts it into a format suitable for analysis.

[1478] Step 3:

[1479] The server inputs the preprocessed data into a multimodal AI and trains it to learn patterns of criminal and abnormal behavior and emotional states.

[1480] Step 4:

[1481] The server periodically provides new data to the AI ​​and updates the learning model.

[1482] Step 5:

[1483] The server uses the trained model to predict abnormal behavior and criminal acts that are likely to occur in the future.

[1484] Step 6:

[1485] The server notifies the administrator and user of the prediction results (e.g., recommended times and locations for increased security).

[1486] Subscription and Maintenance Process Steps

[1487] Step 1:

[1488] Users can sign up for a subscription through a web portal or a dedicated application.

[1489] Step 2:

[1490] The server receives new contract, renewal and cancellation information and stores it in a database.

[1491] Step 3:

[1492] The server connects and provides security cameras and services based on the contract (e.g., adding new cameras and updating existing cameras).

[1493] Step 4:

[1494] The server periodically schedules system self-diagnosis to check the operational status of security cameras and edge devices.

[1495] Step 5:

[1496] The terminal performs periodic self-diagnosis and notifies the server if an abnormality is detected.

[1497] Step 6:

[1498] If an abnormality is detected, the server will promptly notify the maintenance team and arrange for repair or replacement work.

[1499] As a result, the present invention significantly improves safety in commercial facilities and public places, and by taking into account the user's emotional state, it enables more precise and prompt responses.

[1500] Example 2

[1501] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1502] Conventional security systems have limitations in the monitoring and abnormal behavior detection capabilities of security cameras, making it difficult to detect abnormal behavior or criminal activity early and respond immediately. Furthermore, because they do not take the user's emotional state into account, they may miss potential dangers or stressful situations. Furthermore, systems that utilize past data to predict future risks are also inadequate, making it difficult to take effective preventative measures.

[1503] The identification processing by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for receiving video data from a security camera, a processing means including multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity, a means for analyzing the user's emotional state using an emotion engine and complementing the abnormal behavior detection result, a notification means for issuing a report when abnormal behavior or criminal activity is detected, a means for collecting and learning past video data and emotion data, and a means for predicting future crises and criminal activity. This enables the security system to detect abnormal behavior or criminal activity with higher accuracy and respond immediately based on the analysis of emotion data. Furthermore, it becomes possible to predict future risks and take preventive measures by learning from past data.

[1504] A "security camera" is a video capture device installed to monitor fraudulent or abnormal behavior.

[1505] "Video data" refers to digital data that includes information about images and videos captured by security cameras.

[1506] "Reception" is the process of acquiring video data transmitted from a security camera.

[1507] "Real-time analysis" means processing and analyzing the received video data immediately.

[1508] "Abnormal behavior" refers to movements that deviate from normal patterns of behavior, and includes, for example, a person falling down, suddenly running, or roughly handling objects.

[1509] "Criminal activity" refers to conduct that violates the law and includes, for example, acts such as theft, assault, and trespass.

[1510] "Multimodal AI" is an AI technology that integrates and analyzes multiple data modalities (e.g., video, audio, text, etc.).

[1511] "Processing means" is a general term for devices and software for performing specific processes or functions.

[1512] An "emotion engine" is an algorithm or system for analyzing a user's emotional state, recognizing emotions based on facial expressions, voice, body movements, etc.

[1513] "Notification means" refers to a system or device that notifies security personnel or administrators of abnormal behavior or criminal activity when such behavior or criminal activity is detected.

[1514] "Data collection" is the process of systematically gathering information about visual and emotional states.

[1515] "Learning" is the process by which artificial intelligence acquires new knowledge based on past data and improves its performance.

[1516] "Predicting future crises and criminal acts" is the process of analyzing past data and trends to estimate possible future risks and criminal acts in advance.

[1517] This invention combines a system that receives video data from security cameras, analyzes the data in real time to detect and report abnormal behavior and criminal activity, with an emotion engine that recognizes user emotions. This system utilizes multimodal artificial intelligence (AI) to efficiently monitor large volumes of video data. It also has the ability to learn from past data and predict potential future crises, as well as monitor the user's emotional state, enabling more accurate detection of abnormal behavior and immediate response.

[1518] The system configuration is as follows:

[1519] Terminal

[1520] The devices include security cameras and edge devices. The security cameras capture video from designated surveillance areas 24 hours a day. The edge devices temporarily store the captured video data, then compress and encrypt it before sending it to a local server. The local server decompresses the data and securely transfers it to a central server.

[1521] server

[1522] The server has the following modules:

[1523] 1. The central server is a platform for real-time analysis of received video data. Multimodal AI runs on this server, analyzes the video data, and detects abnormal behavior and criminal activity.

[1524] 2. The emotion engine module analyzes the user's video data and recognizes their emotional state (e.g., stress, anxiety, anger), which complements and improves the accuracy of abnormal behavior detection.

[1525] 3. The notification system generates alerts immediately when abnormal or criminal behavior is detected and sends notifications to administrators and security personnel via various media such as email and SMS.

[1526] 4. The data collection and learning module periodically collects historical video and emotion data and stores it in a database. This data is then used by the multimodal AI to learn and predict future crises and criminal activity.

[1527] User

[1528] Users are administrators and security personnel who monitor and manage the system. They receive alerts from the notification system and can take immediate action. Administrators can also receive forecast information based on past data and take measures in advance.

[1529] Specific examples

[1530] For commercial facilities

[1531] The manager of a commercial facility installs the system of the present invention. Security cameras capture video data within the facility in real time and transmit the data to a central server via terminals. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering at the same time). Upon detection, an emotion engine analyzes the emotional states of surrounding users and detects emotional states (e.g., anxiety or fear) associated with the occurrence of abnormal behavior. Based on this, a notification system sends an alert to the manager of the commercial facility, and a security guard is immediately dispatched, enabling a rapid response. Furthermore, past data can be used to predict dangers on specific days and times of the week and notify the manager, allowing preventive measures to be taken in advance.

[1532] Prompt Sentence Examples

[1533] "Please explain a system that receives video data from security cameras, analyzes it in real time to detect abnormal behavior or criminal activity, and also analyzes the user's emotional state and reports it."

[1534] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1535] Step 1:

[1536] The terminal captures video data in real time using security cameras. The real-time video information is input, and the video data sent to the edge device is generated as output. The security cameras operate 24 hours a day, capturing the movements of people and the placement of objects.

[1537] Step 2:

[1538] The edge device of the terminal temporarily stores the captured video data. The input is video data sent from the security camera, and the edge device compresses and encrypts the data and sends it to the local server. The compressed and encrypted data is then transferred to the local server as the output.

[1539] Step 3:

[1540] The server receives video data from the local server. The input is compressed and encrypted data, which the server decompresses and decrypts. The decompressed data is then passed to the AI ​​analysis module. The output is the decompressed and decrypted video data.

[1541] Step 4:

[1542] The server uses an AI analysis module to analyze the decompressed video data in real time. The input is the decompressed video data, and the multimodal AI analyzes people's movements and behavior patterns. For example, it analyzes people's dwell time and movements in a specific area. The output generates detection results for abnormal behavior and criminal activity.

[1543] Step 5:

[1544] When the server detects abnormal behavior or criminal activity, it immediately generates an alert. The input is the detection results from the AI ​​analysis module, and the generated alert is sent as output to the terminals of administrators and security personnel via the notification system. The alert is sent by email or SMS to prompt a response.

[1545] Step 6:

[1546] The server uses an emotion engine to analyze the user's emotional state. The input is video data, and the emotion engine recognizes emotions based on facial expressions, voice, body movements, etc. For example, data showing the user's anxiety or tension is analyzed. The output is the analyzed emotional state.

[1547] Step 7:

[1548] The server performs an integrated analysis of the emotion data obtained by the emotion engine and the abnormal behavior data from the AI ​​analysis module. The input is emotion data and abnormal behavior data, which complements the analysis results and improves accuracy. For example, if there is abnormal behavior in an area where anxious emotions are detected, further detailed monitoring will be performed. The output is the integrated analysis results.

[1549] Step 8:

[1550] The server periodically collects past video data and emotion data and stores them in a database. The input is the periodically collected video data and emotion data, which the multimodal AI uses to learn. The output is a trained AI model.

[1551] Step 9:

[1552] The server uses past data to predict future crises and criminal activity. The input is a trained AI model and past data, and it predicts risks, for example, during specific times of the day or on specific days of the week. The output is the predicted information, which is notified to an administrator so that countermeasures can be taken.

[1553] Step 10:

[1554] The server manages subscription contract information and performs system maintenance. The input is contract information, and it processes new contracts, renewals, and cancellations. It also performs system checks and self-diagnosis, and notifies the maintenance team if an abnormality is discovered. The output is the latest contract information and maintenance status.

[1555] (Application example 2)

[1556] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1557] Conventional security systems were able to detect abnormal behavior and criminal activity by analyzing video data, but because they did not take into account the emotional state of the user, there were limitations to the accuracy of predictions and the speed of response.Furthermore, there was an issue that simply detecting abnormal behavior made it difficult to respond flexibly according to people's emotions and situations.

[1558] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes processing means including multimodal artificial intelligence that analyzes received video data in real time and detects abnormal behavior or criminal activity, means including an emotion engine that analyzes the user's video data and recognizes their emotional state, and means for complementing the detection results of abnormal behavior or criminal activity based on the detected emotional state. This enables more accurate detection of abnormal behavior that takes the user's emotional state into consideration and immediate response.

[1559] A "security camera" is a device installed for security purposes that captures video data in real time.

[1560] "Video data" is data that includes visual information captured by a security camera.

[1561] "Real-time analysis" refers to the process of analyzing video data immediately after it is received.

[1562] "Abnormal behavior" refers to behavior that deviates from normal patterns of behavior and may be potentially dangerous or criminal.

[1563] "Criminal activity" refers to any activity that violates the law and threatens security or public safety.

[1564] "Multimodal AI" is AI that integrates multiple data modalities (e.g., video, audio, text) to analyze and make decisions.

[1565] "Processing means" refers to a combination of software and hardware for analyzing received video data and detecting abnormal behavior or criminal activity.

[1566] An "emotion engine" is a system for analyzing a user's emotional state, and is a technology that analyzes based on facial expressions, voice, body movements, etc.

[1567] A "reporting means" is a system that has the function of sending notifications to administrators or security personnel when abnormal behavior or criminal activity is detected.

[1568] The "alert level" is an indicator of the degree of alertness that is set based on detected abnormal behavior and emotional state.

[1569] "Past data" refers to a collection of video data and emotion data that has been previously collected and analyzed.

[1570] "Predicting future crises and crimes" is the process of predicting potential future dangers and criminal acts based on past data.

[1571] The system of the present invention receives video data from security cameras, analyzes the data in real time to detect abnormal behavior or criminal activity, and recognizes the user's emotions and adjusts the alert level, thereby achieving rapid and highly accurate security measures.

[1572] System configuration

[1573] The system consists of the following main elements:

[1574] 1. Device:

[1575] Security camera: A device that captures video data in real time.

[1576] Edge device: Equipment that performs initial data processing.

[1577] 2. Server:

[1578] Processing server: Includes multimodal artificial intelligence (AI) that analyzes received video data and detects abnormal behavior or criminal activity.

[1579] Emotion engine: An engine that analyzes the user's emotional state.

[1580] Notification Server: A server that sends notifications to security personnel and administrators based on detection results.

[1581] 3. User:

[1582] Administrators and security personnel who use the system.

[1583] What the program does

[1584] The server analyzes video data received from security cameras and edge devices to detect abnormal behavior and criminal activity. Video data is captured and analyzed in real time using software such as OpenCV. Multimodal artificial intelligence (e.g., YourAnomalyDetectionLibrary) identifies patterns of abnormal behavior and detects anomalous behavior.

[1585] At the same time, an emotion engine (such as YourEmotionRecognitionLibrary) is used to analyze the user's emotional state. Emotional states are recognized from data such as facial expressions, voice, and body movements, and indicators such as stress, anxiety, and anger are derived. As a result, the detection results of abnormal behavior and criminal acts are supplemented, making it possible to set more accurate alert levels.

[1586] The notification server generates alerts based on detected abnormal behaviors and emotional states and sends real-time notifications to security personnel and administrators, ensuring a prompt response.

[1587] Specific examples

[1588] When the system of the present invention is implemented in a commercial facility, security cameras capture video data within the facility in real time, which is then transmitted to a central server via an edge device. A multimodal AI on the central server analyzes the video data and detects abnormal behavior (e.g., multiple people entering the facility at the same time). At the same time, an emotion engine analyzes the emotional states of surrounding users (e.g., anxiety or fear) and detects the emotional states associated with the occurrence of abnormal behavior. Based on the results, a notification system sends an alert to the facility manager, who can immediately dispatch security guards.

[1589] Prompt Sentence Examples

[1590] Design a system that analyzes video data in real time to detect abnormal and criminal behavior, and also recognizes user emotions. Include the ability to generate and send instant notifications to security personnel based on the detection results. Demonstrate specific application examples in various locations (e.g., commercial facilities, public transportation, etc.).

[1591] As a result, the present invention realizes security measures that take into account the emotional state of the user, enabling highly accurate detection of abnormal behavior and rapid response.

[1592] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1593] Step 1:

[1594] Security camera video data capture

[1595] The device captures video data in real time using a security camera. The input is a continuous video stream from the security camera, and the output is the captured video data. This video data is sent to the edge device for initial processing.

[1596] Step 2:

[1597] Pre-processing and transfer of video data

[1598] The edge device of the terminal temporarily processes the captured video data and sends it to a local server. The input is the captured video data, and the output is the processed video data. The video data is compressed and encrypted before being transferred to the central server.

[1599] Step 3:

[1600] Video data decompression and analysis

[1601] The server decompresses the video data received on the central server and sends it to a multimodal artificial intelligence (AI) analysis module. The input is compressed and encrypted video data, and the output is video data ready for analysis. The AI ​​analysis module analyzes the video data in real time to detect abnormal behavior and criminal activity.

[1602] Step 4:

[1603] Detecting abnormal and criminal behavior

[1604] The server's AI analysis module analyzes the received video data and detects abnormal behavior such as a person falling or multiple people entering at the same time, as well as criminal activity. The input is the decompressed video data, and the output is the detection results.

[1605] Step 5:

[1606] Emotional Data Analysis

[1607] The server's emotion engine analyzes the user's facial expressions, voice, and body movements based on the video data to recognize the user's emotional state (e.g., anxiety, fear). The input is video data, and the output is emotional state data.

[1608] Step 6:

[1609] Integrated analysis of abnormal behavior and emotional states

[1610] The server integrates the abnormal behavior detection results from the AI ​​analysis module and the emotional state data from the emotion engine, and sets an alert level based on the occurrence of abnormal behavior and the associated emotional state. The input is the abnormal behavior detection results and emotional state data, and the output is the integrated analysis results.

[1611] Step 7:

[1612] Alert level notification

[1613] The server's notification system generates alerts based on the integrated analysis results and sends them to security personnel and administrators in real time. The input is the integrated analysis results and the output is an alert notification, enabling users to respond quickly.

[1614] Specifically, in the case of a commercial facility, for example, video data captured in step 1 is processed in step 2 to detect abnormal behavior or emotional states during specific times or locations, and an alert is sent to the administrator in step 7. Upon receiving this alert, the facility's security staff will quickly head to the scene.

[1615] The following prompts are available for the generative AI model:

[1616] "Design a system that analyzes video data in real time to detect abnormal and criminal behavior, and also recognizes user emotions. Include a function to instantly generate and send notifications to security personnel based on the detection results. Please also provide specific application examples in various locations (e.g., commercial facilities, public transportation, etc.)."

[1617] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1618] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1619] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1620] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1621] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1622] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1623] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1624] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1625] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1626] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1627] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1628] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1629] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1630] 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.

[1631] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1632] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1633] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1634] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1635] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1636] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1637] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1638] The following is further disclosed regarding the above embodiment.

[1639] (Claim 1)

[1640] means for receiving video data from a security camera;

[1641] a processing means including a multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity;

[1642] A system that includes a notification mechanism to report abnormal or criminal behavior if it detects it.

[1643] (Claim 2)

[1644] 10. The system of claim 1, further comprising means for sending a notification to security personnel or management upon detection of abnormal behavior or criminal activity.

[1645] (Claim 3)

[1646] 2. The system according to claim 1, further comprising means for learning past data from the received video data and predicting future crises and crimes.

[1647] (Claim 4)

[1648] 10. The system of claim 1, further comprising means for communicating with the security cameras and for performing system maintenance and subscription management.

[1649] (Claim 5)

[1650] 4. The system according to claim 3, further comprising means for predicting dangers at specific times or locations based on the learned data and notifying the user of the results.

[1651] (Claim 6)

[1652] 10. The system of claim 1, further comprising means for temporarily processing the security camera data on an edge device or a local server and then transmitting the data to the central server.

[1653] (Claim 7)

[1654] 5. The system according to claim 4, further comprising means for performing periodic self-diagnosis as part of system maintenance management, and notifying a maintenance team if an abnormality is detected.

[1655] "Example 1"

[1656] (Claim 1)

[1657] means for receiving video data from a security camera;

[1658] means including an edge device for temporarily processing, compressing, and encrypting the received video data;

[1659] means for transmitting the compressed and encrypted data to a central server;

[1660] a means including a multimodal artificial intelligence for analyzing received video data in real time to detect abnormal behavior or criminal activity;

[1661] a notification system that notifies the user when abnormal or criminal activity is detected;

[1662] A means for storing past video data and using it to predict future crises;

[1663] A means for managing subscription contract information and performing system maintenance and management;

[1664] A system including:

[1665] (Claim 2)

[1666] 10. The system of claim 1, further comprising means for sending a notification to security personnel or management upon detection of abnormal behavior or criminal activity.

[1667] (Claim 3)

[1668] 2. The system according to claim 1, further comprising means for learning past data from the received video data and predicting future crises and crimes.

[1669] "Application Example 1"

[1670] (Claim 1)

[1671] means for receiving video data from a security camera;

[1672] a processing means including a multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity;

[1673] A notification means for reporting abnormal behavior or criminal activity when such behavior is detected;

[1674] A means to predict future crises by learning from past data and notify of risks at specific times and locations.

[1675] a means of sending alerts to a smartphone;

[1676] A system including:

[1677] (Claim 2)

[1678] 10. The system of claim 1, further comprising means for sending a notification to security personnel or management upon detection of abnormal behavior or criminal activity.

[1679] (Claim 3)

[1680] 2. The system according to claim 1, further comprising means for learning past data from the received video data and predicting future crises and crimes.

[1681] "Example 2: Combining Emotion Engines"

[1682] (Claim 1)

[1683] means for receiving video data from a security camera;

[1684] a processing means including a multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity;

[1685] a means for analyzing the emotional state of a user using an emotion engine and complementing the abnormal behavior detection result;

[1686] A notification means for reporting abnormal behavior or criminal activity when such behavior is detected;

[1687] A means of collecting and learning from past video data and emotion data,

[1688] A system that includes a means of predicting future crises and criminal activity.

[1689] (Claim 2)

[1690] 10. The system of claim 1, further comprising means for sending a notification to security personnel or management upon detection of abnormal behavior or criminal activity.

[1691] (Claim 3)

[1692] 10. The system according to claim 1, further comprising means for periodically collecting the received video data and emotion data, and predicting future crises and crimes by learning from past data.

[1693] "Application example 2 when combining emotion engines"

[1694] (Claim 1)

[1695] means for receiving video data from a security camera;

[1696] a processing means including a multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity;

[1697] means including an emotion engine for analyzing video data of a user to recognize the user's emotional state;

[1698] means for complementing the detection of abnormal behavior or criminal activity based on the detected emotional state;

[1699] A system that includes a notification mechanism to report abnormal or criminal behavior if it detects it.

[1700] (Claim 2)

[1701] 10. The system of claim 1, further comprising means for adjusting an alert level based on detection of abnormal or criminal behavior and the user's emotional state, and sending a notification to security personnel or administrators.

[1702] (Claim 3)

[1703] 2. The system according to claim 1, further comprising means for learning past data from the received video data and emotion data and predicting future crises and crimes. [Explanation of symbols]

[1704] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving video data from a security camera; a processing means including a multimodal artificial intelligence that analyzes the received video data in real time and detects abnormal behavior or criminal activity; A system that includes a notification mechanism to report abnormal or criminal behavior if it detects it.

2. 10. The system of claim 1, further comprising means for sending a notification to security personnel or management upon detection of abnormal or criminal activity.

3. 2. The system according to claim 1, further comprising means for learning past data from the received video data and predicting future crises and crimes.

4. 10. The system of claim 1, further comprising means for communicating with the security cameras and for performing system maintenance and subscription management.

5. 4. The system according to claim 3, further comprising means for predicting danger in a specific time period or location based on the learned data and notifying the result.

6. The system of claim 1 , further comprising means for temporarily processing the security camera data in an edge device or a local server and then transmitting the data to the central server.

7. 5. The system according to claim 4, further comprising means for performing periodic self-diagnosis as part of system maintenance management, and notifying a maintenance team if an abnormality is detected.

Images