system

Abstract

We provide the system. [Solution] A device for collecting audio data, A device for extracting features from the aforementioned audio data, A device that receives the extracted feature data, A device equipped with an algorithm that analyzes the aforementioned characteristic data and identifies abnormal sounds, A device that notifies an external organization of a warning when an abnormal sound is detected, A device that provides controls for managing privacy, A system that includes this.

Description

Technical Field

[0005] ,

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

Background Art

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

Summary of the Invention

Problems to be Solved by the Invention

[0004] Due to recent technological developments, it has become possible to collect voice data using smartphones and IoT devices. However, a system for detecting abnormal sounds in real time using such voice data and taking appropriate actions is not yet mature. In particular, improving the detection accuracy of abnormal sounds and the rapid reporting process after detection are issues. Also, in the introduction of such a system, a data management mechanism that takes into account the privacy of users is important.

Means for Solving the Problems

[0005] The present invention solves the above problems by providing a system that includes a device for collecting audio data, a device for extracting features from the audio data, a device for receiving the extracted feature data, a device equipped with an algorithm for analyzing the feature data and identifying abnormal sounds, a device for notifying an external organization of a warning when an abnormal sound is detected, and a device for performing controls to manage privacy. This system improves the accuracy of abnormal sound detection, enables rapid notification, and appropriately protects the user's privacy.

[0006] "Audio data" refers to information recorded in digital format from ambient sounds.

[0007] An "abnormal sound" is a sound that differs from normal ambient noise and indicates an event requiring warning.

[0008] An "apparatus" is a machine or system configured to perform a specific function or role.

[0009] "Feature extraction" is the act of extracting specific patterns or properties necessary for discrimination from collected data.

[0010] An "algorithm" is a set of procedures defined to perform a computational task in order to achieve a specific objective.

[0011] "To notify" is the act of conveying specific information or warnings to another device or person.

[0012] "Privacy" is the right of an individual to have their personal information protected from being disclosed to others without their permission.

[0013] "To control" means to adjust a particular process or system to make it work as intended. [Brief explanation of the drawing]

[0014] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which a plurality of emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Mode for Carrying Out the Invention

[0015] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

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

[0017] In the following embodiments, the labeled processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0018] In the following embodiments, the labeled RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0019] In the following embodiments, the labeled storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0020] In the following embodiments, the labeled communication I / F (Interface) is an interface including a communication processor and an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark), and the like.

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0022] [First Embodiment]

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

[0024] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0025] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0027] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

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

[0031] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0034] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0035] This invention is a system for detecting abnormal sounds in real time using audio data, and is effectively implemented by coordinating information between a terminal with specific functions, a server, and a user. The terminal is embedded in the user's smartphone or IoT device and plays the role of constantly collecting ambient audio data. The terminal has the function of reducing noise from the collected audio data and extracting features necessary for detecting abnormal sounds. As a result, the collected data is transmitted to the server in a format suitable for later analysis.

[0036] The server receives feature data transmitted from the terminal and analyzes the data in real time using advanced algorithms. The generative AI model within the server compares the features of the audio data with known abnormal sound patterns and instantly determines the presence or absence of abnormal sounds. This analysis allows for the immediate identification of sounds that suggest danger, such as explosions, breaking glass, and screams.

[0037] If an unusual sound is detected, the server immediately notifies the relevant emergency services, providing a report that includes the location of the incident and details of the audio. It also sends an alert to nearby users' devices where the unusual sound was detected, urging them to evacuate quickly or move to a safe location. Users can manage notification and privacy settings through the application, controlling data collection and managing past data.

[0038] As a concrete example, consider a scenario where glass breaks at night in a residential area. When a terminal detects this sound, characteristic data is sent to the server. If it is determined to be an abnormal sound, the police are immediately notified, and a warning is sent to nearby users. This system configuration allows for the rapid identification and response to dangerous sounds that might otherwise be overlooked.

[0039] The following describes the processing flow.

[0040] Step 1:

[0041] The device uses its built-in microphone to collect ambient sound in real time. This audio data is temporarily stored in a buffer and prepared for the next processing step.

[0042] Step 2:

[0043] The device applies a noise reduction filter to the stored audio data, clearing the audio signal. This enhances important audio features.

[0044] Step 3:

[0045] The device extracts features from the filtered audio data. These features are necessary information for detecting abnormal sounds, such as sound intensity and frequency spectrum.

[0046] Step 4:

[0047] The terminal packets the extracted features and sends them to the server via wireless communication along with the terminal's location information.

[0048] Step 5:

[0049] The server receives feature data transmitted from the terminal and stores it in a database. This data is then analyzed by an abnormal sound detection algorithm.

[0050] Step 6:

[0051] The server inputs the received features into a generating AI model and performs analysis to detect abnormal sounds. The model compares the similarity to existing abnormal sound patterns to determine the presence or absence of abnormal sounds.

[0052] Step 7:

[0053] If an unusual sound is detected, the server automatically notifies the relevant emergency services and generates and sends a report along with the location information of the incident site.

[0054] Step 8:

[0055] Based on the detection information, the server sends an alarm notification to nearby users' terminals and provides safety guidelines.

[0056] Step 9:

[0057] Users can receive notifications and, if necessary, adjust their privacy settings and choose how their data is handled within the application to manage their personal information.

[0058] (Example 1)

[0059] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0060] The problem that this invention aims to solve is to provide a system that can detect abnormal sounds in real time using audio information and provide appropriate notifications. Conventional technologies have limitations in the accuracy and responsiveness of abnormal sound detection, and also have shortcomings in terms of privacy management and notification to nearby users.

[0061] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0062] In this invention, the server includes means for analyzing audio information and using a generated AI model to identify abnormal sounds, means for notifying an external organization of an alarm when an abnormal sound is detected, and means for transmitting the abnormal sound detection result as a warning to the devices of nearby users. This enables real-time, highly accurate detection of abnormal sounds and a quick and appropriate response.

[0063] "Audio information" refers to sound data acquired from the surrounding environment, and this includes everyday sounds as well as specific abnormal sounds.

[0064] "Feature information" refers to specific acoustic elements extracted from audio information, and is fundamental data for identifying abnormal sounds.

[0065] "Analysis means" refers to devices or algorithms that use a generative AI model to analyze feature information and perform the process of identifying abnormal sounds based on this analysis.

[0066] "Alarm" refers to notification information sent to external organizations or nearby users to alert them to the occurrence of an unusual sound.

[0067] A "remote agency" refers to an organization or institution to which notifications are sent in order to take appropriate action in response to abnormal sounds detected from audio information.

[0068] "Privacy management" refers to a set of settings and processes that control how users' voice data is handled and ensure its use within permitted limits.

[0069] A "generative AI model" is an artificial intelligence technology used for analyzing audio information. It refers to a model that learns sound patterns and is used to identify abnormal sounds.

[0070] "User device" refers to a device owned by the user that is used to receive alarms and facilitate response actions.

[0071] This invention is a system that analyzes audio information in real time, quickly detects abnormal sounds, and provides relevant notifications. The terminal is implemented as a smartphone or IoT device and collects audio information from external sources using a built-in microphone. Audio processing software (e.g., audio editing software) that performs noise reduction and signal enhancement is utilized on the terminal.

[0072] The terminal that collects audio information is responsible for extracting feature information and sending it to the server. The server analyzes the received feature information using a generative AI model (for example, a model using TENSORFLOW® or PyTorch) to identify abnormal sounds. The generative AI model recognizes the pattern of the audio and determines the abnormality by comparing it with known abnormal sound types.

[0073] If an abnormal sound is detected, the server generates an alarm and notifies remote agencies. The notification includes the precise location and details of the situation where the abnormal sound originated, enabling a rapid response. It also sends an alarm to nearby users' devices, urging them to take immediate evacuation action.

[0074] Users can freely customize their privacy settings and notification methods through the application. This allows for the management of voice information tailored to each individual user.

[0075] As a concrete example, if an unusual sound, such as glass breaking, occurs in a residential area late at night, the terminal detects the sound and sends its characteristic information to a server. When the generating AI model identifies the unusual sound, it automatically notifies the police, and evacuation guidance messages are sent to nearby residents.

[0076] Examples of prompt messages include specific scenarios such as "notification when abnormal sounds like breaking glass or screams are detected." This prompt enables the system to immediately recognize and respond to abnormal sounds.

[0077] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0078] Step 1:

[0079] The device collects ambient sounds. At this stage, the device acquires audio information from its built-in microphone and records this data in temporary memory. The audio information is the input data and is stored as an audio waveform on the device.

[0080] Step 2:

[0081] The terminal performs audio processing to reduce noise from the collected audio information. Here, digital signal processing is used for noise filtering, resulting in the output of clear audio data. During this process, the original audio information is input, and the noise-free audio is output.

[0082] Step 3:

[0083] The device extracts features from the noise-reduced audio data. This process calculates audio features (e.g., Mel-frequency cepstrum coefficients and spectral information). The input is the filtered audio data, and the output is the digitized feature information.

[0084] Step 4:

[0085] The terminal sends the extracted feature information to the server. The feature information is transferred to the server using a secure protocol. In this step, the input is the audio feature information, and the output is the completion of the data transfer to the server.

[0086] Step 5:

[0087] The server inputs the received feature information into a generative AI model. The generative AI model analyzes the audio pattern and compares it to known abnormal sounds. The input here is feature information sent from the terminal, and the output is the result of the determination regarding the presence or absence of abnormal sounds.

[0088] Step 6:

[0089] The server generates an alarm and notifies a remote agency when an abnormal sound is detected. The alarm includes location information and audio details to enable a rapid response. The input is the judgment result of the generating AI model, and the output is the transmission of the alarm notification.

[0090] Step 7:

[0091] The server sends an alarm to nearby users' devices, prompting them to respond quickly. The input is information about the detected abnormal sound, and the output is an alarm message to the user.

[0092] (Application Example 1)

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

[0094] Traditionally, abnormal sound detection systems have been inaccurate, making real-time notification difficult and hindering the rapid assurance of user safety. Furthermore, privacy protection measures were inadequate.

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

[0096] In this invention, the server includes means for a computation method that analyzes sound information and identifies abnormal sounds, means for managing user privacy, and means for constantly monitoring sound information with a receiving device and providing emergency notifications when a specific abnormal sound is detected. This enables highly accurate detection of abnormal sounds and real-time notification, improving user safety and privacy protection.

[0097] "Audio information" refers to sound data collected from the surrounding environment and is material that is analyzed for a specific purpose.

[0098] "Features" are important attributes and patterns extracted from sound information, and are data used to identify abnormal sounds.

[0099] A "receiving device" is a device used to collect or receive sound information or characteristic information.

[0100] "Calculation method" refers to a series of computational processes that analyze sound information and characteristic information to identify abnormal sounds.

[0101] An "external organization" refers to an external institution or group that shares information when an abnormal sound is detected.

[0102] "Privacy" refers to the right of users to have their personal information protected regarding the collection and use of sound information.

[0103] "Management" refers to actions taken to control and properly operate the processes related to the collection, analysis, and notification of sound information.

[0104] "Surveillance" refers to the activity of continuously observing sound information and capturing that information when specific conditions occur.

[0105] An "emergency notification" is a warning or information dissemination that is immediately sent to external organizations or users when an abnormal sound is detected.

[0106] The system that realizes this invention first uses a terminal to collect sound information. The terminal continuously collects ambient sounds from the surroundings using the built-in microphone of a smartphone. The collected sound information is filtered on the terminal using the "librosa" library to reduce noise and clarify the signal. In addition, features are extracted from the sound information during this process and preprocessed for abnormal sound detection.

[0107] Next, the server receives feature information sent from the terminal. On the server, a TensorFlow model is used to perform advanced analysis and analyze the sound information in real time. This generative AI model has the ability to compare features in the sound information with known abnormal sound patterns and determine the presence or absence of abnormal sounds. If an abnormal sound is detected, the server immediately notifies the external organization and also sends an alarm to the user's terminal.

[0108] Users receive notifications and respond promptly. Within the application, notification settings and privacy management are available, and users have control over their past data. This ensures user safety while protecting their privacy.

[0109] For example, if the sound of breaking glass occurs at night, the terminal quickly detects the sound and sends it to the server. The server analyzes the sound in real time, and if it recognizes an abnormal sound, it sends a warning to nearby users and also notifies emergency services. This process allows for a quick and effective response.

[0110] An example of a prompt to input into a generative AI model is, "Detect abnormal sounds from this audio data. Could the sound you are currently hearing indicate danger?"

[0111] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0112] Step 1:

[0113] The device collects ambient sounds in real time. It uses its built-in microphone to acquire sound information and receives it as input data. This sound information becomes the raw material for the next processing step.

[0114] Step 2:

[0115] The terminal filters the audio information using the "librosa" library. This filtering reduces noise and enhances the signal. The extracted audio feature data becomes the input for the next analysis step.

[0116] Step 3:

[0117] The terminal sends feature data to the server. The server receives this feature data and begins processing it as input information for advanced analysis.

[0118] Step 4:

[0119] The server analyzes the received feature data using a TensorFlow model. The generative AI model matches the data with abnormal sound patterns to identify abnormal sounds. If an abnormal sound is detected, the result becomes the output information for the next notification step.

[0120] Step 5:

[0121] The server notifies external organizations and users' terminals based on the detection results. This notification includes detailed information about the abnormal sound, along with any necessary warnings.

[0122] Step 6:

[0123] The user receives a notification on their device and checks the situation through the application. The user is required to take appropriate action based on the prompt. "An example of a prompt sentence to input into the generating AI model is: 'Detect abnormal sounds from this audio data. Could the sound you are hearing now indicate danger?'"

[0124] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0125] This invention combines a system that detects abnormal sounds based on audio data with an emotion engine that recognizes user emotions. By integrating audio data collection, feature extraction, abnormal sound detection, warning notification, emotion recognition, and privacy management, this system enables more personalized responses.

[0126] The device continuously records the user's ambient sounds and extracts features that indicate abnormal sounds from the collected audio data. Filtering reduces noise and improves the accuracy of abnormal sound detection. Furthermore, the device is equipped with an emotion engine that analyzes the user's emotional state from their voice tone and intonation. This emotion data is transmitted to the server along with the abnormal sound detection information.

[0127] The server receives data sent from the terminal and uses a generative AI model to identify abnormal sounds, as well as analyze the user's emotional data. Based on this, the server adjusts the priority and content of emergency notifications, taking into account the user's emotional state. For example, if the user is showing signs of anxiety, the server can activate the warning system more quickly and send encouraging and supportive messages appropriate to their emotional state.

[0128] Users can manage their privacy through the system's settings screen and customize how their emotional data is handled and what kind of notifications they receive. This provides a secure environment for using the emotional recognition system.

[0129] As a concrete example, consider a scenario where audio of someone shouting in an office is recorded. The terminal digitizes the audio and sends it to a server, where the emotion engine recognizes it as anger. The server assesses the urgency of the situation, sends notifications tailored to the user and the appropriate response agency, and the user receives appropriate notifications to help them calm down. In this way, dynamic responses tailored to individual situations become possible.

[0130] The following describes the processing flow.

[0131] Step 1:

[0132] The device collects ambient sound in real time using its built-in microphone. The collected audio data is stored in a buffer, and noise reduction filtering is applied to highlight important audio features.

[0133] Step 2:

[0134] The device analyzes the noise-reduced audio data to extract features for identifying abnormal sounds. Furthermore, it utilizes an integrated emotion engine to analyze emotions based on the user's voice tone and intonation.

[0135] Step 3:

[0136] The terminal packets the extracted abnormal sound features and emotion data, and sends them to the server along with the terminal's location information.

[0137] Step 4:

[0138] The server receives feature data and sentiment data transmitted from the terminal and stores it in a dedicated database. A generative AI model instantly identifies abnormal sounds and determines whether action is required.

[0139] Step 5:

[0140] The server analyzes the received emotional data and takes the user's emotional state into account when setting notification priorities and customizing notification content. If the user is experiencing high levels of anxiety, it is determined that a rapid response is necessary.

[0141] Step 6:

[0142] When an abnormal sound is detected, the server automatically sends a notification to the relevant emergency services, including the necessary information (type of abnormal sound, location information, and emotional information).

[0143] Step 7:

[0144] Based on abnormal sound detection and emotion analysis results, the server sends a notification to the user's device containing appropriate safety guidelines. This notification is individually customized based on the user's emotional state.

[0145] Step 8:

[0146] Based on the notifications they receive, users can further adjust their privacy settings and how emotional data is handled within the system, allowing them to use the system with peace of mind.

[0147] (Example 2)

[0148] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0149] Conventional abnormal sound detection systems were limited to detecting abnormal sounds based on audio data and lacked the ability to respond dynamically according to the user's psychological state. This could result in users receiving notifications at inappropriate times or excessive warnings, potentially undermining the system's effectiveness. Furthermore, insufficient privacy management meant users could not use the system with peace of mind.

[0150] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0151] In this invention, the server includes means for analyzing voice information, means for analyzing the user's emotional state, and control means for adjusting notification content based on the emotional state. This enables the provision of more personalized notifications that respond to the user's emotions and enhances privacy management.

[0152] "Audio information" refers to sounds collected from the environment and converted into data that can be processed in a digital format.

[0153] "Attributes" are pieces of information extracted from audio data that describe the characteristics and patterns of sound.

[0154] "Equipment" refers to hardware devices used for collecting, processing, and analyzing audio information.

[0155] A "processing unit" is a computing unit that analyzes audio information and attributes and makes judgments based on specific conditions.

[0156] "Warning information" refers to the content of warnings and notifications sent to external organizations or users when an anomaly is detected.

[0157] An "external organization" is an institution or group outside the system that is responsible for receiving notifications and responding in the event of an anomaly.

[0158] "Privacy" is a concept that protects users' personal information and sensitive data from being illegally acquired or used.

[0159] "Control means" refers to the processes and mechanisms that adjust and manage the operation of a system.

[0160] "Emotional state" refers to information that indicates the user's psychological state, representing the type and degree of emotion inferred from their voice and behavior.

[0161] This invention is a system that performs anomaly detection and emotion analysis using voice information and provides personalized notifications to users. Specific embodiments are described below.

[0162] Collection and processing of audio information

[0163] The device continuously records ambient sounds through a high-performance microphone, collecting them as audio information. This process involves specific filtering to reduce noise and enhance the signal from the collected audio information. The Python audio processing library librosa is used for this process. A digital signal processor (DSP) performs spectral analysis of the audio, extracting sound characteristics as attributes.

[0164] Emotion analysis

[0165] The device is equipped with an emotion engine that analyzes the user's emotional state from their voice. This uses a speech emotion recognition model based on TensorFlow. It analyzes features such as pitch, intonation, and rhythm of the voice to understand the user's emotional state.

[0166] Data transmission and analysis

[0167] The terminal sends extracted voice attribute information and the user's emotional state to the server. The server receives this data, analyzes the anomaly using a generative AI model, and further evaluates the user's emotional state in detail. Based on the type of anomaly and the emotional state, the server individually adjusts the priority and content of notifications.

[0168] Notification generation

[0169] The server sends notifications directly to users based on the analysis results. For example, if a user expresses anxiety, the system is required to respond quickly and immediately provides a warning along with a message of encouragement.

[0170] As a concrete example, consider the occurrence of unusual noises in an office environment. If a terminal detects an unusual noise, such as someone shouting, the sound is recorded, processed, and sent to a server. The server analyzes the situation and sends psychologically appropriate notifications to users in the office, thereby reducing anxiety and encouraging immediate action.

[0171] An example of a prompt using a generative AI model is: "Please explain how to extract abnormal sounds from audio data and analyze emotional states. Also, please provide step-by-step instructions on how to send emergency notifications based on the results." This prompt provides users and developers with guidance to accurately understand and effectively operate the system.

[0172] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0173] Step 1:

[0174] The device continuously acquires ambient sounds using a high-performance microphone. The input is ambient sound, and the output is digital audio data. This audio data is temporarily stored in local storage. During this process, the device filters out audio noise and enhances the signal to generate clean audio data suitable for analysis.

[0175] Step 2:

[0176] The terminal extracts attributes from saved audio data. The input is filtered audio data, and the output is attribute information representing the spectral characteristics of the sound. Specifically, the terminal uses the Python librosa library to split the audio waveform and perform spectral analysis. This analysis yields the frequency components and intensity of the sound as numerical vectors.

[0177] Step 3:

[0178] The device performs analysis using an emotion engine based on attribute information. The input is the spectral features of the speech, and the output is the user's emotional state. Specifically, the device runs an emotion recognition model using TensorFlow to classify emotions from the pitch and intonation of the speech. This allows the device to accurately grasp the user's emotional state.

[0179] Step 4:

[0180] The terminal sends extracted attribute information and emotional state to the server. The input consists of attribute information and emotional state, which are sent to the server as communication data. The terminal encrypts the data during this transmission to ensure security.

[0181] Step 5:

[0182] The server analyzes the received attribute information and emotional state. The input is data sent from the terminal, and the output is identification information to identify anomalies and notification content based on emotion. The server uses a generative AI model to determine if there are anomalies in the voice and further considers the user's emotional state to evaluate whether action is necessary.

[0183] Step 6:

[0184] The server generates and sends notifications to users based on the analysis results. Input consists of anomaly identification information and notification content based on sentiment, while output is a message including warnings and encouragement to the user. The server individually adjusts the priority and content of notifications to ensure they are delivered at the most appropriate time in specific situations.

[0185] (Application Example 2)

[0186] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0187] In modern society, there is a need to detect abnormal sounds occurring in an individual's living environment and to provide appropriate responses tailored to the user's emotional state. However, conventional systems are limited to analyzing only audio data, making it difficult to provide personalized responses that take into account the user's psychological state. Furthermore, balancing privacy security with a rapid warning notification system is also a challenge.

[0188] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0189] In this invention, the server includes means for detecting abnormal sounds from audio data, means for recognizing and analyzing the user's emotional state, and means for adjusting the priority and content of notifications based on the emotional data. This makes it possible to suggest individual alerts and relaxation methods tailored to each user's psychological state.

[0190] "Audio data" refers to sound that has been recorded and stored in digital format.

[0191] "Feature extraction" refers to extracting patterns and information useful for identification from audio data.

[0192] "Abnormal sounds" refer to sounds with specific frequencies or amplitudes that deviate from normal ambient sounds.

[0193] An "algorithm" is a set of steps or calculation methods designed to solve a specific problem.

[0194] A "warning" is a notification intended to draw attention to a danger or problem.

[0195] "Emotional state" refers to information that indicates a user's psychological or emotional response.

[0196] "Notification priority" refers to the criteria used to determine the order in which notifications are sent, based on the importance and urgency of the information being transmitted.

[0197] "Privacy" is a concept that protects personal information and prevents its unauthorized use or disclosure.

[0198] This invention is an automated security system that integrates the detection of abnormal sounds with the recognition of user emotions. The system mainly consists of three components: a terminal, a server, and a user.

[0199] The device has a built-in audio sensor for collecting ambient sounds, allowing for continuous recording of audio data. The collected audio data is preprocessed, and specific features are extracted. Python's LibROSA library is used for extracting audio features, performing noise reduction and signal enhancement. This improves the accuracy of detecting abnormal sounds.

[0200] The server receives audio feature data transmitted from the terminal and uses a generative AI model to identify abnormal sounds. TensorFlow is used for detecting abnormal sounds, and the system works with the terminal to quickly issue a warning when an abnormal sound is detected. At the same time, the server receives emotion data from the terminal and analyzes it using the Google® Cloud Speech-to-Text API to recognize the user's emotional state.

[0201] Through this system, users can monitor their emotional state and the detection status of unusual sounds in real time. For example, if a siren suddenly sounds in a quiet living room, the device records the sound and sends it to the server. The server immediately identifies the sound as suspicious and sends a warning to the user. If the system determines that the user is in an anxious state, it also provides guidance on relaxation methods.

[0202] As an example of a prompt message, you can send instructions to the server in the format of, "Is there anything unusual about this audio data? Also, what emotional state is this user in?"

[0203] This invention enables quick and appropriate responses tailored to individual situations, allowing users to enjoy a safe and secure environment.

[0204] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0205] Step 1:

[0206] The device continuously collects ambient sound using an audio sensor. The collected audio data is filtered using LibROSA for noise reduction and signal enhancement, and its characteristics are extracted. The input is raw audio data, and the output is filtered characteristic data.

[0207] Step 2:

[0208] The terminal sends the extracted audio feature data to the server. The server receives this data and uses a generative AI model to identify abnormal sounds. The input is feature data, and the output is the result of determining whether or not an abnormal sound was detected. Specifically, it identifies the presence or absence of an abnormal sound and prepares a warning if necessary.

[0209] Step 3:

[0210] The server analyzes the user's emotional state along with the abnormal sound detection results. Based on the emotional data received from the terminal, it uses the Google Cloud Speech-to-Text API to analyze the user's voice tone and intonation to identify their emotions. The input is audio data and an emotion analysis model, and the output is the emotion detection result.

[0211] Step 4:

[0212] The server adjusts the priority and content of notifications based on the detected abnormal sound and the user's emotional state. For example, if an abnormal sound is detected and the user is determined to be in an anxious state, a warning is immediately issued, and relaxation advice is provided. The input is the detection result, and the output is the adjusted notification message.

[0213] Step 5:

[0214] Users receive warnings and advice sent by the device and consider how to address them. They can request additional information or actions using prompts. Specifically, if they want more details about a suspicious sound, they can prompt the device to re-analyze it.

[0215] In this way, this system provides a mechanism that simultaneously supports the safety of the voice environment and the psychological stability of the user.

[0216] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0217] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0218] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0219] [Second Embodiment]

[0220] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0221] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0222] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0223] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0224] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0225] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0226] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0227] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0228] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0229] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0230] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0231] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0232] This invention is a system for detecting abnormal sounds in real time using audio data, and is effectively implemented by coordinating information between a terminal with specific functions, a server, and a user. The terminal is embedded in the user's smartphone or IoT device and plays the role of constantly collecting ambient audio data. The terminal has the function of reducing noise from the collected audio data and extracting features necessary for detecting abnormal sounds. As a result, the collected data is transmitted to the server in a format suitable for later analysis.

[0233] The server receives feature data transmitted from the terminal and analyzes the data in real time using advanced algorithms. The generative AI model within the server compares the features of the audio data with known abnormal sound patterns and instantly determines the presence or absence of abnormal sounds. This analysis allows for the immediate identification of sounds that suggest danger, such as explosions, breaking glass, and screams.

[0234] If an unusual sound is detected, the server immediately notifies the relevant emergency services, providing a report that includes the location of the incident and details of the audio. It also sends an alert to nearby users' devices where the unusual sound was detected, urging them to evacuate quickly or move to a safe location. Users can manage notification and privacy settings through the application, controlling data collection and managing past data.

[0235] As a concrete example, consider a scenario where glass breaks at night in a residential area. When a terminal detects this sound, characteristic data is sent to the server. If it is determined to be an abnormal sound, the police are immediately notified, and a warning is sent to nearby users. This system configuration allows for the rapid identification and response to dangerous sounds that might otherwise be overlooked.

[0236] The following describes the processing flow.

[0237] Step 1:

[0238] The device uses its built-in microphone to collect ambient sound in real time. This audio data is temporarily stored in a buffer and prepared for the next processing step.

[0239] Step 2:

[0240] The device applies a noise reduction filter to the stored audio data, clearing the audio signal. This enhances important audio features.

[0241] Step 3:

[0242] The device extracts features from the filtered audio data. These features are necessary information for detecting abnormal sounds, such as sound intensity and frequency spectrum.

[0243] Step 4:

[0244] The terminal packets the extracted features and sends them to the server via wireless communication along with the terminal's location information.

[0245] Step 5:

[0246] The server receives feature data transmitted from the terminal and stores it in a database. This data is then analyzed by an abnormal sound detection algorithm.

[0247] Step 6:

[0248] The server inputs the received features into a generating AI model and performs analysis to detect abnormal sounds. The model compares the similarity to existing abnormal sound patterns to determine the presence or absence of abnormal sounds.

[0249] Step 7:

[0250] If an unusual sound is detected, the server automatically notifies the relevant emergency services and generates and sends a report along with the location information of the incident site.

[0251] Step 8:

[0252] Based on the detection information, the server sends an alarm notification to nearby users' terminals and provides safety guidelines.

[0253] Step 9:

[0254] Users can receive notifications and, if necessary, adjust their privacy settings and choose how their data is handled within the application to manage their personal information.

[0255] (Example 1)

[0256] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0257] The problem that this invention aims to solve is to provide a system that can detect abnormal sounds in real time using audio information and provide appropriate notifications. Conventional technologies have limitations in the accuracy and responsiveness of abnormal sound detection, and also have shortcomings in terms of privacy management and notification to nearby users.

[0258] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0259] In this invention, the server includes means for analyzing audio information and using a generated AI model to identify abnormal sounds, means for notifying an external organization of an alarm when an abnormal sound is detected, and means for transmitting the abnormal sound detection result as a warning to the devices of nearby users. This enables real-time, highly accurate detection of abnormal sounds and a quick and appropriate response.

[0260] "Audio information" refers to sound data acquired from the surrounding environment, and this includes everyday sounds as well as specific abnormal sounds.

[0261] "Feature information" refers to specific acoustic elements extracted from audio information, and is fundamental data for identifying abnormal sounds.

[0262] "Analysis means" refers to devices or algorithms that use a generative AI model to analyze feature information and perform the process of identifying abnormal sounds based on this analysis.

[0263] "Alarm" refers to notification information sent to external organizations or nearby users to alert them to the occurrence of an unusual sound.

[0264] A "remote agency" refers to an organization or institution to which notifications are sent in order to take appropriate action in response to abnormal sounds detected from audio information.

[0265] "Privacy management" refers to a set of settings and processes that control how users' voice data is handled and ensure its use within permitted limits.

[0266] A "generative AI model" is an artificial intelligence technology used for analyzing audio information. It refers to a model that learns sound patterns and is used to identify abnormal sounds.

[0267] "User device" refers to a device owned by the user that is used to receive alarms and facilitate response actions.

[0268] This invention is a system that analyzes audio information in real time, quickly detects abnormal sounds, and provides relevant notifications. The terminal is implemented as a smartphone or IoT device and collects audio information from external sources using a built-in microphone. Audio processing software (e.g., audio editing software) that performs noise reduction and signal enhancement is utilized on the terminal.

[0269] The terminal that collects audio information is responsible for extracting feature information and sending it to the server. The server analyzes the received feature information using a generative AI model (for example, a model using TensorFlow or PyTorch) to identify abnormal sounds. The generative AI model recognizes the patterns in the audio and determines the abnormality by comparing them with known abnormal sound types.

[0270] If an abnormal sound is detected, the server generates an alarm and notifies remote agencies. The notification includes the precise location and details of the situation where the abnormal sound originated, enabling a rapid response. It also sends an alarm to nearby users' devices, urging them to take immediate evacuation action.

[0271] Users can freely customize their privacy settings and notification methods through the application. This allows for the management of voice information tailored to each individual user.

[0272] As a concrete example, if an unusual sound, such as glass breaking, occurs in a residential area late at night, the terminal detects the sound and sends its characteristic information to a server. When the generating AI model identifies the unusual sound, it automatically notifies the police, and evacuation guidance messages are sent to nearby residents.

[0273] Examples of prompt messages include specific scenarios such as "notification when abnormal sounds like breaking glass or screams are detected." This prompt enables the system to immediately recognize and respond to abnormal sounds.

[0274] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0275] Step 1:

[0276] The device collects ambient sounds. At this stage, the device acquires audio information from its built-in microphone and records this data in temporary memory. The audio information is the input data and is stored as an audio waveform on the device.

[0277] Step 2:

[0278] The terminal performs audio processing to reduce noise from the collected audio information. Here, digital signal processing is used for noise filtering, resulting in the output of clear audio data. During this process, the original audio information is input, and the noise-free audio is output.

[0279] Step 3:

[0280] The terminal extracts the features of the voice data with reduced noise. In this process, voice feature quantities (e.g., Mel-frequency cepstral coefficients and spectral information) are calculated. The input is the voice data after filtering, and the output is the digitized feature information.

[0281] Step 4:

[0282] The terminal transmits the extracted feature information to the server. The feature information is transferred to the server using a secure protocol. In this step, the input is the voice feature information, and the output is the completion of the data transfer to the server.

[0283] Step 5:

[0284] The server inputs the received feature information into the generative AI model. The generative AI model analyzes the voice pattern and compares it with known abnormal sounds. The input here is the feature information sent from the terminal, and the output is the determination result regarding the presence or absence of abnormal sounds.

[0285] Step 6:

[0286] When the server detects an abnormal sound, it creates an alarm and notifies a remote agency. The alarm includes location information and voice details, enabling a prompt response. The input is the determination result of the generative AI model, and the output is the transmission of the alarm notification.

[0287] Step 7:

[0288] The server transmits an alarm to the devices of neighboring users. This encourages users to be able to respond promptly. The input is the detected abnormal sound information, and the output is an alarm message to the users.

[0289] (Application Example 1)

[0290] Next, Application Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0291] Traditionally, abnormal sound detection systems have been inaccurate, making real-time notification difficult and hindering the rapid assurance of user safety. Furthermore, privacy protection measures were inadequate.

[0292] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0293] In this invention, the server includes means for a computation method that analyzes sound information and identifies abnormal sounds, means for managing user privacy, and means for constantly monitoring sound information with a receiving device and providing emergency notifications when a specific abnormal sound is detected. This enables highly accurate detection of abnormal sounds and real-time notification, improving user safety and privacy protection.

[0294] "Audio information" refers to sound data collected from the surrounding environment and is material that is analyzed for a specific purpose.

[0295] "Features" are important attributes and patterns extracted from sound information, and are data used to identify abnormal sounds.

[0296] A "receiving device" is a device used to collect or receive sound information or characteristic information.

[0297] "Calculation method" refers to a series of computational processes that analyze sound information and characteristic information to identify abnormal sounds.

[0298] An "external organization" refers to an external institution or group that shares information when an abnormal sound is detected.

[0299] "Privacy" refers to the right of users to have their personal information protected regarding the collection and use of sound information.

[0300] "Management" refers to actions taken to control and properly operate the processes related to the collection, analysis, and notification of sound information.

[0301] "Surveillance" refers to the activity of continuously observing sound information and capturing that information when specific conditions occur.

[0302] An "emergency notification" is a warning or information dissemination that is immediately sent to external organizations or users when an abnormal sound is detected.

[0303] The system that realizes this invention first uses a terminal to collect sound information. The terminal continuously collects ambient sounds from the surroundings using the built-in microphone of a smartphone. The collected sound information is filtered on the terminal using the "librosa" library to reduce noise and clarify the signal. In addition, features are extracted from the sound information during this process and preprocessed for abnormal sound detection.

[0304] Next, the server receives feature information sent from the terminal. On the server, a TensorFlow model is used to perform advanced analysis and analyze the sound information in real time. This generative AI model has the ability to compare features in the sound information with known abnormal sound patterns and determine the presence or absence of abnormal sounds. If an abnormal sound is detected, the server immediately notifies the external organization and also sends an alarm to the user's terminal.

[0305] Users receive notifications and respond promptly. Within the application, notification settings and privacy management are available, and users have control over their past data. This ensures user safety while protecting their privacy.

[0306] For example, if the sound of breaking glass occurs at night, the terminal quickly detects the sound and sends it to the server. The server analyzes the sound in real time, and if it recognizes an abnormal sound, it sends a warning to nearby users and also notifies emergency services. This process allows for a quick and effective response.

[0307] As an example of a prompt sentence to be input into the generative AI model, "Please detect abnormal sounds from this audio data. Is there a possibility that the sound I am hearing now represents danger?" can be considered.

[0308] The flow of the specific process in Application Example 1 will be described using FIG. 12.

[0309] Step 1:

[0310] The terminal collects the ambient environmental sounds in real time. It uses the built-in microphone to obtain sound information and receives it as input data. This sound information serves as the material for the following processes.

[0311] Step 2:

[0312] The terminal performs filtering processing on the sound information using the "librosa" library. The filtering reduces noise and emphasizes the signal. The sound feature data extracted here becomes the input for the next analysis step.

[0313] Step 3:

[0314] The terminal sends the feature data to the server. The server receives this feature data and starts processing it as input information for advanced analysis.

[0315] Step 4:

[0316] The server analyzes the received feature data using the "TensorFlow" model. The generative AI model performs matching with abnormal sound patterns and discriminates abnormal sounds. If an abnormal sound is detected, the result becomes the output information for the next notification step.

[0317] Step 5:

[0318] Based on the detection result, the server notifies external organizations and the user's terminal. Here, together with the detailed information of the abnormal sound, the necessary warnings are sent.

[0319] Step 6:

[0320] The user receives a notification on their device and checks the situation through the application. The user is required to take appropriate action based on the prompt. "An example of a prompt sentence to input into the generating AI model is: 'Detect abnormal sounds from this audio data. Could the sound you are hearing now indicate danger?'"

[0321] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0322] This invention combines a system that detects abnormal sounds based on audio data with an emotion engine that recognizes user emotions. By integrating audio data collection, feature extraction, abnormal sound detection, warning notification, emotion recognition, and privacy management, this system enables more personalized responses.

[0323] The device continuously records the user's ambient sounds and extracts features that indicate abnormal sounds from the collected audio data. Filtering reduces noise and improves the accuracy of abnormal sound detection. Furthermore, the device is equipped with an emotion engine that analyzes the user's emotional state from their voice tone and intonation. This emotion data is transmitted to the server along with the abnormal sound detection information.

[0324] The server receives data sent from the terminal and uses a generative AI model to identify abnormal sounds, as well as analyze the user's emotional data. Based on this, the server adjusts the priority and content of emergency notifications, taking into account the user's emotional state. For example, if the user is showing signs of anxiety, the server can activate the warning system more quickly and send encouraging and supportive messages appropriate to their emotional state.

[0325] Users can manage their privacy through the system's settings screen and customize how their emotional data is handled and what kind of notifications they receive. This provides a secure environment for using the emotional recognition system.

[0326] As a concrete example, consider a scenario where audio of someone shouting in an office is recorded. The terminal digitizes the audio and sends it to a server, where the emotion engine recognizes it as anger. The server assesses the urgency of the situation, sends notifications tailored to the user and the appropriate response agency, and the user receives appropriate notifications to help them calm down. In this way, dynamic responses tailored to individual situations become possible.

[0327] The following describes the processing flow.

[0328] Step 1:

[0329] The device collects ambient sound in real time using its built-in microphone. The collected audio data is stored in a buffer, and noise reduction filtering is applied to highlight important audio features.

[0330] Step 2:

[0331] The device analyzes the noise-reduced audio data to extract features for identifying abnormal sounds. Furthermore, it utilizes an integrated emotion engine to analyze emotions based on the user's voice tone and intonation.

[0332] Step 3:

[0333] The terminal packets the extracted abnormal sound features and emotion data, and sends them to the server along with the terminal's location information.

[0334] Step 4:

[0335] The server receives feature data and sentiment data transmitted from the terminal and stores it in a dedicated database. A generative AI model instantly identifies abnormal sounds and determines whether action is required.

[0336] Step 5:

[0337] The server analyzes the received emotional data and takes the user's emotional state into account when setting notification priorities and customizing notification content. If the user is experiencing high levels of anxiety, it is determined that a rapid response is necessary.

[0338] Step 6:

[0339] When an abnormal sound is detected, the server automatically sends a notification to the relevant emergency services, including the necessary information (type of abnormal sound, location information, and emotional information).

[0340] Step 7:

[0341] Based on abnormal sound detection and emotion analysis results, the server sends a notification to the user's device containing appropriate safety guidelines. This notification is individually customized based on the user's emotional state.

[0342] Step 8:

[0343] Based on the notifications they receive, users can further adjust their privacy settings and the handling of emotional data within the system, allowing them to use the system with peace of mind.

[0344] (Example 2)

[0345] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0346] Conventional abnormal sound detection systems were limited to detecting abnormal sounds based on audio data and lacked the ability to respond dynamically according to the user's psychological state. This could result in users receiving notifications at inappropriate times or excessive warnings, potentially undermining the system's effectiveness. Furthermore, insufficient privacy management meant users could not use the system with peace of mind.

[0347] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0348] In this invention, the server includes means for analyzing voice information, means for analyzing the user's emotional state, and control means for adjusting notification content based on the emotional state. This enables the provision of more personalized notifications that respond to the user's emotions and enhances privacy management.

[0349] "Audio information" refers to sounds collected from the environment and converted into data that can be processed in a digital format.

[0350] "Attributes" are pieces of information extracted from audio data that describe the characteristics and patterns of sound.

[0351] "Equipment" refers to hardware devices used for collecting, processing, and analyzing audio information.

[0352] A "processing unit" is a computing unit that analyzes audio information and attributes and makes classifications based on specific conditions.

[0353] "Warning information" refers to the content of warnings and notifications sent to external organizations or users when an anomaly is detected.

[0354] An "external organization" is an institution or group outside the system that is responsible for receiving notifications and responding in the event of an anomaly.

[0355] "Privacy" is a concept that protects users' personal information and sensitive data from being illegally acquired or used.

[0356] "Control means" refers to the processes and mechanisms that adjust and manage the operation of a system.

[0357] "Emotional state" refers to information that indicates the user's psychological state, representing the type and degree of emotion inferred from their voice and behavior.

[0358] This invention is a system that performs anomaly detection and emotion analysis using voice information and provides personalized notifications to users. Specific embodiments are described below.

[0359] Collection and processing of audio information

[0360] The device continuously records ambient sounds through a high-performance microphone, collecting them as audio information. This process involves specific filtering to reduce noise and enhance the signal from the collected audio information. The Python audio processing library librosa is used for this process. A digital signal processor (DSP) performs spectral analysis of the audio, extracting sound characteristics as attributes.

[0361] Emotion analysis

[0362] The device is equipped with an emotion engine that analyzes the user's emotional state from their voice. This uses a speech emotion recognition model based on TensorFlow. It analyzes features such as pitch, intonation, and rhythm of the voice to understand the user's emotional state.

[0363] Data transmission and analysis

[0364] The device sends extracted voice attribute information and the user's emotional state to the server. The server receives this data, analyzes the anomaly using a generative AI model, and further evaluates the user's emotional state in detail. Based on the type of anomaly and the emotional state, it individually adjusts the priority and content of notifications.

[0365] Notification generation

[0366] The server sends notifications directly to users based on the analysis results. For example, if a user expresses anxiety, the system is required to respond quickly and immediately provides a warning along with a message of encouragement.

[0367] As a concrete example, consider the occurrence of unusual noises in an office environment. If a terminal detects an unusual noise, such as someone shouting, the sound is recorded, processed, and sent to a server. The server analyzes the situation and sends psychologically appropriate notifications to users in the office, thereby reducing anxiety and encouraging immediate action.

[0368] An example of a prompt using a generative AI model is: "Please explain how to extract abnormal sounds from audio data and analyze the emotional state. Also, please explain step-by-step how to send an emergency notification based on the results." This prompt provides users and developers with guidance to accurately understand and effectively operate the system.

[0369] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0370] Step 1:

[0371] The device continuously acquires ambient sounds using a high-performance microphone. The input is ambient sound, and the output is digital audio data. This audio data is temporarily stored in local storage. During this process, the device filters out audio noise and enhances the signal to generate clean audio data suitable for analysis.

[0372] Step 2:

[0373] The terminal extracts attributes from saved audio data. The input is filtered audio data, and the output is attribute information representing the spectral characteristics of the sound. Specifically, the terminal uses the Python librosa library to split the audio waveform and perform spectral analysis. This analysis yields the frequency components and intensity of the sound as numerical vectors.

[0374] Step 3:

[0375] The device performs analysis using an emotion engine based on attribute information. The input is the spectral features of the speech, and the output is the user's emotional state. Specifically, the device runs an emotion recognition model using TensorFlow to classify emotions from the pitch and intonation of the speech. This allows the device to accurately grasp the user's emotional state.

[0376] Step 4:

[0377] The terminal sends extracted attribute information and emotional state to the server. The input consists of attribute information and emotional state, which are sent to the server as communication data. The terminal encrypts the data during this transmission to ensure security.

[0378] Step 5:

[0379] The server analyzes the received attribute information and emotional state. The input is data sent from the terminal, and the output is identification information to identify anomalies and notification content based on emotion. The server uses a generative AI model to determine if there are anomalies in the voice and further considers the user's emotional state to evaluate whether action is necessary.

[0380] Step 6:

[0381] The server generates and sends notifications to users based on the analysis results. Input consists of anomaly identification information and notification content based on sentiment, while output is a message including warnings and encouragement to the user. The server individually adjusts the priority and content of notifications to ensure they are delivered at the most appropriate time in specific situations.

[0382] (Application Example 2)

[0383] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0384] In modern society, there is a need to detect abnormal sounds occurring in an individual's living environment and to provide appropriate responses tailored to the user's emotional state. However, conventional systems are limited to analyzing only audio data, making it difficult to provide personalized responses that take into account the user's psychological state. Furthermore, balancing privacy security with a rapid warning notification system is also a challenge.

[0385] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0386] In this invention, the server includes means for detecting abnormal sounds from audio data, means for recognizing and analyzing the user's emotional state, and means for adjusting the priority and content of notifications based on the emotional data. This makes it possible to suggest individual alerts and relaxation methods tailored to each user's psychological state.

[0387] "Audio data" refers to sound that has been recorded and stored in digital format.

[0388] "Feature extraction" refers to extracting patterns and information useful for identification from audio data.

[0389] "Abnormal sounds" refer to sounds with specific frequencies or amplitudes that deviate from normal ambient sounds.

[0390] An "algorithm" is a set of steps or calculation methods designed to solve a specific problem.

[0391] A "warning" is a notification intended to draw attention to a danger or problem.

[0392] "Emotional state" refers to information that indicates a user's psychological or emotional response.

[0393] "Notification priority" refers to the criteria used to determine the order in which notifications are sent, based on the importance and urgency of the information being transmitted.

[0394] "Privacy" is a concept that protects personal information and prevents its unauthorized use or disclosure.

[0395] This invention is an automated security system that integrates the detection of abnormal sounds with the recognition of user emotions. The system mainly consists of three components: a terminal, a server, and a user.

[0396] The device has a built-in audio sensor for collecting ambient sounds, allowing for continuous recording of audio data. The collected audio data is preprocessed, and specific features are extracted. Python's LibROSA library is used for extracting audio features, performing noise reduction and signal enhancement. This improves the accuracy of detecting abnormal sounds.

[0397] The server receives audio feature data transmitted from the terminal and uses a generative AI model to identify abnormal sounds. TensorFlow is used for abnormal sound detection, and the system works with the terminal to quickly issue a warning when abnormal sounds are detected. At the same time, the server receives emotion data from the terminal and analyzes it using the Google Cloud Speech-to-Text API to recognize the user's emotional state.

[0398] Through this system, users can monitor their emotional state and the detection status of unusual sounds in real time. For example, if a siren suddenly sounds in a quiet living room, the device records the sound and sends it to the server. The server immediately identifies the sound as suspicious and sends a warning to the user. If the system determines that the user is in an anxious state, it also provides guidance on relaxation methods.

[0399] As an example of a prompt message, you can send instructions to the server in the format of, "Is there anything unusual about this audio data? Also, what emotional state is this user in?"

[0400] This invention enables quick and appropriate responses tailored to individual situations, allowing users to enjoy a safe and secure environment.

[0401] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0402] Step 1:

[0403] The device continuously collects ambient sound using an audio sensor. The collected audio data is filtered using LibROSA for noise reduction and signal enhancement, and its characteristics are extracted. The input is raw audio data, and the output is filtered characteristic data.

[0404] Step 2:

[0405] The terminal sends the extracted audio feature data to the server. The server receives this data and uses a generative AI model to identify abnormal sounds. The input is feature data, and the output is the result of determining whether or not an abnormal sound was detected. Specifically, it identifies the presence or absence of an abnormal sound and prepares a warning if necessary.

[0406] Step 3:

[0407] The server analyzes the user's emotional state along with the abnormal sound detection results. Based on the emotional data received from the terminal, it uses the Google Cloud Speech-to-Text API to analyze the user's voice tone and intonation to identify their emotions. The input is audio data and an emotion analysis model, and the output is the emotion detection result.

[0408] Step 4:

[0409] The server adjusts the priority and content of notifications based on the detected abnormal sound and the user's emotional state. For example, if an abnormal sound is detected and the user is determined to be in an anxious state, a warning is immediately issued, and relaxation advice is provided. The input is the detection result, and the output is the adjusted notification message.

[0410] Step 5:

[0411] Users receive warnings and advice sent by the device and consider how to address them. They can request additional information or actions using prompts. Specifically, if they want more details about a suspicious sound, they can prompt the device to re-analyze it.

[0412] In this way, this system provides a mechanism that simultaneously supports the safety of the voice environment and the psychological stability of the user.

[0413] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0414] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0415] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0416] [Third Embodiment]

[0417] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0418] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0419] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0420] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0421] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0422] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0423] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0424] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0425] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0426] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0427] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0428] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0429] This invention is a system for detecting abnormal sounds in real time using audio data, and is effectively implemented by coordinating information between a terminal with specific functions, a server, and a user. The terminal is embedded in the user's smartphone or IoT device and plays the role of constantly collecting ambient audio data. The terminal has the function of reducing noise from the collected audio data and extracting features necessary for detecting abnormal sounds. As a result, the collected data is transmitted to the server in a format suitable for later analysis.

[0430] The server receives feature data transmitted from the terminal and analyzes the data in real time using advanced algorithms. The generative AI model within the server compares the features of the audio data with known abnormal sound patterns and instantly determines the presence or absence of abnormal sounds. This analysis allows for the immediate identification of sounds that suggest danger, such as explosions, breaking glass, and screams.

[0431] If an unusual sound is detected, the server immediately notifies the relevant emergency services, providing a report that includes the location of the incident and details of the audio. It also sends an alert to nearby users' devices where the unusual sound was detected, urging them to evacuate quickly or move to a safe location. Users can manage notification and privacy settings through the application, controlling data collection and managing past data.

[0432] As a concrete example, consider a scenario where glass breaks at night in a residential area. When a terminal detects this sound, characteristic data is sent to the server. If it is determined to be an abnormal sound, the police are immediately notified, and a warning is sent to nearby users. This system configuration allows for the rapid identification and response to dangerous sounds that might otherwise be overlooked.

[0433] The following describes the processing flow.

[0434] Step 1:

[0435] The device uses its built-in microphone to collect ambient sound in real time. This audio data is temporarily stored in a buffer and prepared for the next processing step.

[0436] Step 2:

[0437] The device applies a noise reduction filter to the stored audio data, clearing the audio signal. This enhances important audio features.

[0438] Step 3:

[0439] The device extracts features from the filtered audio data. These features are necessary information for detecting abnormal sounds, such as sound intensity and frequency spectrum.

[0440] Step 4:

[0441] The terminal packets the extracted features and sends them to the server via wireless communication along with the terminal's location information.

[0442] Step 5:

[0443] The server receives feature data transmitted from the terminal and stores it in a database. This data is then analyzed by an abnormal sound detection algorithm.

[0444] Step 6:

[0445] The server inputs the received features into a generating AI model and performs analysis to detect abnormal sounds. The model compares the similarity to existing abnormal sound patterns to determine the presence or absence of abnormal sounds.

[0446] Step 7:

[0447] If an unusual sound is detected, the server automatically notifies the relevant emergency services and generates and sends a report along with the location information of the incident site.

[0448] Step 8:

[0449] Based on the detection information, the server sends an alarm notification to nearby users' terminals and provides safety guidelines.

[0450] Step 9:

[0451] Users can receive notifications and, if necessary, adjust their privacy settings and choose how their data is handled within the application to manage their personal information.

[0452] (Example 1)

[0453] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0454] The problem that this invention aims to solve is to provide a system that can detect abnormal sounds in real time using audio information and provide appropriate notifications. Conventional technologies have limitations in the accuracy and responsiveness of abnormal sound detection, and also have shortcomings in terms of privacy management and notification to nearby users.

[0455] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0456] In this invention, the server includes means for analyzing audio information and using a generated AI model to identify abnormal sounds, means for notifying an external organization of an alarm when an abnormal sound is detected, and means for transmitting the abnormal sound detection result as a warning to the devices of nearby users. This enables real-time, highly accurate detection of abnormal sounds and a quick and appropriate response.

[0457] "Audio information" refers to sound data acquired from the surrounding environment, and this includes everyday sounds as well as specific abnormal sounds.

[0458] "Feature information" refers to specific acoustic elements extracted from audio information, and is fundamental data for identifying abnormal sounds.

[0459] "Analysis means" refers to devices or algorithms that use a generative AI model to analyze feature information and perform the process of identifying abnormal sounds based on this analysis.

[0460] "Alarm" refers to notification information sent to external organizations or nearby users to alert them to the occurrence of an unusual sound.

[0461] A "remote agency" refers to an organization or institution to which notifications are sent in order to take appropriate action in response to abnormal sounds detected from audio information.

[0462] "Privacy management" refers to a set of settings and processes that control how users' voice data is handled and ensure its use within permitted limits.

[0463] A "generative AI model" is an artificial intelligence technology used for analyzing audio information. It refers to a model that learns sound patterns and is used to identify abnormal sounds.

[0464] "User device" refers to a device owned by the user that is used to receive alarms and facilitate response actions.

[0465] This invention is a system that analyzes audio information in real time, quickly detects abnormal sounds, and provides relevant notifications. The terminal is implemented as a smartphone or IoT device and collects audio information from external sources using a built-in microphone. Audio processing software (e.g., audio editing software) that performs noise reduction and signal enhancement is utilized on the terminal.

[0466] The terminal that collects audio information is responsible for extracting feature information and sending it to the server. The server analyzes the received feature information using a generative AI model (for example, a model using TensorFlow or PyTorch) to identify abnormal sounds. The generative AI model recognizes the patterns in the audio and determines the abnormality by comparing them with known abnormal sound types.

[0467] If an abnormal sound is detected, the server generates an alarm and notifies remote agencies. The notification includes the precise location and details of the situation where the abnormal sound originated, enabling a rapid response. It also sends an alarm to nearby users' devices, urging them to take immediate evacuation action.

[0468] Users can freely customize their privacy settings and notification methods through the application. This allows for the management of voice information tailored to each individual user.

[0469] As a concrete example, if an unusual sound, such as glass breaking, occurs in a residential area late at night, the terminal detects the sound and sends its characteristic information to a server. When the generating AI model identifies the unusual sound, it automatically notifies the police, and evacuation guidance messages are sent to nearby residents.

[0470] Examples of prompt messages include specific scenarios such as "notification when abnormal sounds like breaking glass or screams are detected." This prompt enables the system to immediately recognize and respond to abnormal sounds.

[0471] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0472] Step 1:

[0473] The device collects ambient sounds. At this stage, the device acquires audio information from its built-in microphone and records this data in temporary memory. The audio information is the input data and is stored as an audio waveform on the device.

[0474] Step 2:

[0475] The terminal performs audio processing to reduce noise from the collected audio information. Here, digital signal processing is used for noise filtering, resulting in the output of clear audio data. During this process, the original audio information is input, and the noise-free audio is output.

[0476] Step 3:

[0477] The device extracts features from the noise-reduced audio data. This process calculates audio features (e.g., Mel-frequency cepstrum coefficients and spectral information). The input is the filtered audio data, and the output is the digitized feature information.

[0478] Step 4:

[0479] The terminal sends the extracted feature information to the server. The feature information is transferred to the server using a secure protocol. In this step, the input is the audio feature information, and the output is the completion of the data transfer to the server.

[0480] Step 5:

[0481] The server inputs the received feature information into a generative AI model. The generative AI model analyzes the audio pattern and compares it to known abnormal sounds. The input here is feature information sent from the terminal, and the output is the result of the determination regarding the presence or absence of abnormal sounds.

[0482] Step 6:

[0483] The server generates an alarm and notifies a remote agency when an abnormal sound is detected. The alarm includes location information and audio details to enable a rapid response. The input is the judgment result of the generating AI model, and the output is the transmission of the alarm notification.

[0484] Step 7:

[0485] The server sends an alarm to nearby users' devices, prompting them to respond quickly. The input is information about the detected abnormal sound, and the output is an alarm message to the user.

[0486] (Application Example 1)

[0487] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0488] Traditionally, abnormal sound detection systems have been inaccurate, making real-time notification difficult and hindering the rapid assurance of user safety. Furthermore, privacy protection measures were inadequate.

[0489] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0490] In this invention, the server includes means for a computation method that analyzes sound information and identifies abnormal sounds, means for managing user privacy, and means for constantly monitoring sound information with a receiving device and providing emergency notifications when a specific abnormal sound is detected. This enables highly accurate detection of abnormal sounds and real-time notification, improving user safety and privacy protection.

[0491] "Audio information" refers to sound data collected from the surrounding environment and is material that is analyzed for a specific purpose.

[0492] "Features" are important attributes and patterns extracted from sound information, and are data used to identify abnormal sounds.

[0493] A "receiving device" is a device used to collect or receive sound information or characteristic information.

[0494] "Calculation method" refers to a series of computational processes that analyze sound information and characteristic information to identify abnormal sounds.

[0495] An "external organization" refers to an external institution or group that shares information when an abnormal sound is detected.

[0496] "Privacy" refers to the right of users to have their personal information protected regarding the collection and use of sound information.

[0497] "Management" refers to actions taken to control and properly operate the processes related to the collection, analysis, and notification of sound information.

[0498] "Surveillance" refers to the activity of continuously observing sound information and capturing that information when specific conditions occur.

[0499] An "emergency notification" is a warning or information dissemination that is immediately sent to external organizations or users when an abnormal sound is detected.

[0500] The system that realizes this invention first uses a terminal to collect sound information. The terminal continuously collects ambient sounds from the surroundings using the built-in microphone of a smartphone. The collected sound information is filtered on the terminal using the "librosa" library to reduce noise and clarify the signal. In addition, features are extracted from the sound information during this process and preprocessed for abnormal sound detection.

[0501] Next, the server receives feature information sent from the terminal. On the server, a TensorFlow model is used to perform advanced analysis and analyze the sound information in real time. This generative AI model has the ability to compare features in the sound information with known abnormal sound patterns and determine the presence or absence of abnormal sounds. If an abnormal sound is detected, the server immediately notifies the external organization and also sends an alarm to the user's terminal.

[0502] Users receive notifications and respond promptly. Within the application, notification settings and privacy management are available, and users have control over their past data. This ensures user safety while protecting their privacy.

[0503] For example, if the sound of breaking glass occurs at night, the terminal quickly detects the sound and sends it to the server. The server analyzes the sound in real time, and if it recognizes an abnormal sound, it sends a warning to nearby users and also notifies emergency services. This process allows for a quick and effective response.

[0504] An example of a prompt to input into a generative AI model is, "Detect abnormal sounds from this audio data. Could the sound you are currently hearing indicate danger?"

[0505] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0506] Step 1:

[0507] The device collects ambient sounds in real time. It uses its built-in microphone to acquire sound information and receives it as input data. This sound information becomes the raw material for the next processing step.

[0508] Step 2:

[0509] The terminal filters the audio information using the "librosa" library. This filtering reduces noise and enhances the signal. The extracted audio feature data becomes the input for the next analysis step.

[0510] Step 3:

[0511] The terminal sends feature data to the server. The server receives this feature data and begins processing it as input information for advanced analysis.

[0512] Step 4:

[0513] The server analyzes the received feature data using a TensorFlow model. The generative AI model matches the data with abnormal sound patterns to identify abnormal sounds. If an abnormal sound is detected, the result becomes the output information for the next notification step.

[0514] Step 5:

[0515] Based on the detection results, the server sends notifications to external organizations and user terminals. This includes detailed information about the abnormal sound, along with any necessary warnings.

[0516] Step 6:

[0517] The user receives a notification on their device and checks the situation through the application. The user is required to take appropriate action based on the prompt. "An example of a prompt sentence to input into the generating AI model is: 'Detect abnormal sounds from this audio data. Could the sound you are hearing now indicate danger?'"

[0518] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0519] This invention combines a system that detects abnormal sounds based on audio data with an emotion engine that recognizes user emotions. By integrating audio data collection, feature extraction, abnormal sound detection, warning notification, emotion recognition, and privacy management, this system enables more personalized responses.

[0520] The device continuously records the user's ambient sounds and extracts features that indicate abnormal sounds from the collected audio data. Filtering reduces noise and improves the accuracy of abnormal sound detection. Furthermore, the device is equipped with an emotion engine that analyzes the user's emotional state from their voice tone and intonation. This emotion data is transmitted to the server along with the abnormal sound detection information.

[0521] The server receives data sent from the terminal and uses a generative AI model to identify abnormal sounds, as well as analyze the user's emotional data. Based on this, the server adjusts the priority and content of emergency notifications, taking into account the user's emotional state. For example, if the user is showing signs of anxiety, the server can activate the warning system more quickly and send encouraging and supportive messages appropriate to their emotional state.

[0522] Users can manage their privacy through the system's settings screen and customize how their emotional data is handled and what kind of notifications they receive. This provides a secure environment for using the emotional recognition system.

[0523] As a concrete example, consider a scenario where audio of someone shouting in an office is recorded. The terminal digitizes the audio and sends it to a server, where the emotion engine recognizes it as anger. The server assesses the urgency of the situation, sends notifications tailored to the user and the appropriate response agency, and the user receives appropriate notifications to help them calm down. In this way, dynamic responses tailored to individual situations become possible.

[0524] The following describes the processing flow.

[0525] Step 1:

[0526] The device collects ambient sound in real time using its built-in microphone. The collected audio data is stored in a buffer, and noise reduction filtering is applied to highlight important audio features.

[0527] Step 2:

[0528] The device analyzes the noise-reduced audio data to extract features for identifying abnormal sounds. Furthermore, it utilizes an integrated emotion engine to analyze emotions based on the user's voice tone and intonation.

[0529] Step 3:

[0530] The terminal packets the extracted abnormal sound features and emotion data, and sends them to the server along with the terminal's location information.

[0531] Step 4:

[0532] The server receives feature data and sentiment data transmitted from the terminal and stores it in a dedicated database. A generative AI model instantly identifies abnormal sounds and determines whether action is required.

[0533] Step 5:

[0534] The server analyzes the received emotional data and takes the user's emotional state into account when setting notification priorities and customizing notification content. If the user is experiencing high levels of anxiety, it is determined that a rapid response is necessary.

[0535] Step 6:

[0536] When an abnormal sound is detected, the server automatically sends a notification to the relevant emergency services, including the necessary information (type of abnormal sound, location information, and emotional information).

[0537] Step 7:

[0538] Based on abnormal sound detection and emotion analysis results, the server sends a notification to the user's device containing appropriate safety guidelines. This notification is individually customized based on the user's emotional state.

[0539] Step 8:

[0540] Based on the notifications they receive, users can further adjust their privacy settings and the handling of emotional data within the system, allowing them to use the system with peace of mind.

[0541] (Example 2)

[0542] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0543] Conventional abnormal sound detection systems were limited to detecting abnormal sounds based on audio data and lacked the ability to respond dynamically according to the user's psychological state. This could result in users receiving notifications at inappropriate times or excessive warnings, potentially undermining the system's effectiveness. Furthermore, insufficient privacy management meant users could not use the system with peace of mind.

[0544] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0545] In this invention, the server includes means for analyzing voice information, means for analyzing the user's emotional state, and control means for adjusting notification content based on the emotional state. This enables the provision of more personalized notifications that respond to the user's emotions and enhances privacy management.

[0546] "Audio information" refers to sounds collected from the environment and converted into data that can be processed in a digital format.

[0547] "Attributes" are pieces of information extracted from audio data that describe the characteristics and patterns of sound.

[0548] "Equipment" refers to hardware devices used for collecting, processing, and analyzing audio information.

[0549] A "processing unit" is a computing unit that analyzes audio information and attributes and makes classifications based on specific conditions.

[0550] "Warning information" refers to the content of warnings and notifications sent to external organizations or users when an anomaly is detected.

[0551] An "external organization" is an institution or group outside the system that is responsible for receiving notifications and responding in the event of an anomaly.

[0552] "Privacy" is a concept that protects users' personal information and sensitive data from being illegally acquired or used.

[0553] "Control means" refers to the processes and mechanisms that adjust and manage the operation of a system.

[0554] "Emotional state" refers to information that indicates the user's psychological state, representing the type and degree of emotion inferred from their voice and behavior.

[0555] This invention is a system that performs anomaly detection and emotion analysis using voice information and provides personalized notifications to users. Specific embodiments are described below.

[0556] Collection and processing of audio information

[0557] The device continuously records ambient sounds through a high-performance microphone, collecting them as audio information. This process involves specific filtering to reduce noise and enhance the signal from the collected audio information. The Python audio processing library librosa is used for this process. A digital signal processor (DSP) performs spectral analysis of the audio, extracting sound characteristics as attributes.

[0558] Emotion analysis

[0559] The device is equipped with an emotion engine that analyzes the user's emotional state from their voice. This uses a speech emotion recognition model based on TensorFlow. It analyzes features such as pitch, intonation, and rhythm of the voice to understand the user's emotional state.

[0560] Data transmission and analysis

[0561] The device sends extracted voice attribute information and the user's emotional state to the server. The server receives this data, analyzes the anomaly using a generative AI model, and further evaluates the user's emotional state in detail. Based on the type of anomaly and the emotional state, it individually adjusts the priority and content of notifications.

[0562] Notification generation

[0563] The server sends notifications directly to users based on the analysis results. For example, if a user expresses anxiety, the system is required to respond quickly and immediately provides a warning along with a message of encouragement.

[0564] As a concrete example, consider the occurrence of unusual noises in an office environment. If a terminal detects an unusual noise, such as someone shouting, the sound is recorded, processed, and sent to a server. The server analyzes the situation and sends psychologically appropriate notifications to users in the office, thereby reducing anxiety and encouraging immediate action.

[0565] An example of a prompt using a generative AI model is: "Please explain how to extract abnormal sounds from audio data and analyze the emotional state. Also, please explain step-by-step how to send an emergency notification based on the results." This prompt provides users and developers with guidance to accurately understand and effectively operate the system.

[0566] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0567] Step 1:

[0568] The device continuously acquires ambient sounds using a high-performance microphone. The input is ambient sound, and the output is digital audio data. This audio data is temporarily stored in local storage. During this process, the device filters out audio noise and enhances the signal to generate clean audio data suitable for analysis.

[0569] Step 2:

[0570] The terminal extracts attributes from saved audio data. The input is filtered audio data, and the output is attribute information representing the spectral characteristics of the sound. Specifically, the terminal uses the Python librosa library to split the audio waveform and perform spectral analysis. This analysis yields the frequency components and intensity of the sound as numerical vectors.

[0571] Step 3:

[0572] The device performs analysis using an emotion engine based on attribute information. The input is the spectral features of the speech, and the output is the user's emotional state. Specifically, the device runs an emotion recognition model using TensorFlow to classify emotions from the pitch and intonation of the speech. This allows the device to accurately grasp the user's emotional state.

[0573] Step 4:

[0574] The terminal sends extracted attribute information and emotional state to the server. The input consists of attribute information and emotional state, which are sent to the server as communication data. The terminal encrypts the data during this transmission to ensure security.

[0575] Step 5:

[0576] The server analyzes the received attribute information and emotional state. The input is data sent from the terminal, and the output is identification information to identify anomalies and notification content based on emotion. The server uses a generative AI model to determine if there are anomalies in the voice and further considers the user's emotional state to evaluate whether action is necessary.

[0577] Step 6:

[0578] The server generates and sends notifications to users based on the analysis results. Input consists of anomaly identification information and notification content based on sentiment, while output is a message including warnings and encouragement to the user. The server individually adjusts the priority and content of notifications to ensure they are delivered at the most appropriate time in specific situations.

[0579] (Application Example 2)

[0580] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0581] In modern society, there is a need to detect abnormal sounds occurring in an individual's living environment and to provide appropriate responses tailored to the user's emotional state. However, conventional systems are limited to analyzing only audio data, making it difficult to provide personalized responses that take into account the user's psychological state. Furthermore, balancing privacy security with a rapid warning notification system is also a challenge.

[0582] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0583] In this invention, the server includes means for detecting abnormal sounds from audio data, means for recognizing and analyzing the user's emotional state, and means for adjusting the priority and content of notifications based on the emotional data. This makes it possible to suggest individual alerts and relaxation methods tailored to each user's psychological state.

[0584] "Audio data" refers to sound that has been recorded and stored in digital format.

[0585] "Feature extraction" refers to extracting patterns and information useful for identification from audio data.

[0586] "Abnormal sounds" refer to sounds with specific frequencies or amplitudes that deviate from normal ambient sounds.

[0587] An "algorithm" is a set of steps or calculation methods designed to solve a specific problem.

[0588] A "warning" is a notification intended to draw attention to a danger or problem.

[0589] "Emotional state" refers to information that indicates a user's psychological or emotional response.

[0590] "Notification priority" refers to the criteria used to determine the order in which notifications are sent, based on the importance and urgency of the information being transmitted.

[0591] "Privacy" is a concept that protects personal information and prevents its unauthorized use or disclosure.

[0592] This invention is an automated security system that integrates the detection of abnormal sounds with the recognition of user emotions. The system mainly consists of three components: a terminal, a server, and a user.

[0593] The device has a built-in audio sensor for collecting ambient sounds, allowing for continuous recording of audio data. The collected audio data is preprocessed, and specific features are extracted. Python's LibROSA library is used for extracting audio features, performing noise reduction and signal enhancement. This improves the accuracy of detecting abnormal sounds.

[0594] The server receives audio feature data transmitted from the terminal and uses a generative AI model to identify abnormal sounds. TensorFlow is used for abnormal sound detection, and the system works with the terminal to quickly issue a warning when abnormal sounds are detected. At the same time, the server receives emotion data from the terminal and analyzes it using the Google Cloud Speech-to-Text API to recognize the user's emotional state.

[0595] Through this system, users can monitor their emotional state and the detection status of unusual sounds in real time. For example, if a siren suddenly sounds in a quiet living room, the device records the sound and sends it to the server. The server immediately identifies the sound as suspicious and sends a warning to the user. If the system determines that the user is in an anxious state, it also provides guidance on relaxation methods.

[0596] As an example of a prompt message, you can send instructions to the server in the format of, "Is there anything unusual about this audio data? Also, what emotional state is this user in?"

[0597] This invention enables quick and appropriate responses tailored to individual situations, allowing users to enjoy a safe and secure environment.

[0598] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0599] Step 1:

[0600] The device continuously collects ambient sound using an audio sensor. The collected audio data is filtered using LibROSA for noise reduction and signal enhancement, and its characteristics are extracted. The input is raw audio data, and the output is filtered characteristic data.

[0601] Step 2:

[0602] The terminal sends the extracted audio feature data to the server. The server receives this data and uses a generative AI model to identify abnormal sounds. The input is feature data, and the output is the result of determining whether or not an abnormal sound was detected. Specifically, it identifies the presence or absence of an abnormal sound and prepares a warning if necessary.

[0603] Step 3:

[0604] The server analyzes the user's emotional state along with the abnormal sound detection results. Based on the emotional data received from the terminal, it uses the Google Cloud Speech-to-Text API to analyze the user's voice tone and intonation to identify their emotions. The input is audio data and an emotion analysis model, and the output is the emotion detection result.

[0605] Step 4:

[0606] The server adjusts the priority and content of notifications based on the detected abnormal sound and the user's emotional state. For example, if an abnormal sound is detected and the user is determined to be in an anxious state, a warning is immediately issued, and relaxation advice is provided. The input is the detection result, and the output is the adjusted notification message.

[0607] Step 5:

[0608] Users receive warnings and advice sent by the device and consider how to address them. They can request additional information or actions using prompts. Specifically, if they want more details about a suspicious sound, they can prompt the device to re-analyze it.

[0609] In this way, this system provides a mechanism that simultaneously supports the safety of the voice environment and the psychological stability of the user.

[0610] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0611] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0612] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0613] [Fourth Embodiment]

[0614] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0615] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0616] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0617] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0618] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0619] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0620] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0621] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors in the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0622] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0623] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0624] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0625] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0626] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0627] This invention is a system for detecting abnormal sounds in real time using audio data, and is effectively implemented by coordinating information between a terminal with specific functions, a server, and a user. The terminal is embedded in the user's smartphone or IoT device and plays the role of constantly collecting ambient audio data. The terminal has the function of reducing noise from the collected audio data and extracting features necessary for detecting abnormal sounds. As a result, the collected data is transmitted to the server in a format suitable for later analysis.

[0628] The server receives feature data transmitted from the terminal and analyzes the data in real time using advanced algorithms. The generative AI model within the server compares the features of the audio data with known abnormal sound patterns and instantly determines the presence or absence of abnormal sounds. This analysis allows for the immediate identification of sounds that suggest danger, such as explosions, breaking glass, and screams.

[0629] If an unusual sound is detected, the server immediately notifies the relevant emergency services, providing a report that includes the location of the incident and details of the audio. It also sends an alert to nearby users' devices where the unusual sound was detected, urging them to evacuate quickly or move to a safe location. Users can manage notification and privacy settings through the application, controlling data collection and managing past data.

[0630] As a concrete example, consider a scenario where glass breaks at night in a residential area. When a terminal detects this sound, characteristic data is sent to the server. If it is determined to be an abnormal sound, the police are immediately notified, and a warning is sent to nearby users. This system configuration allows for the rapid identification and response to dangerous sounds that might otherwise be overlooked.

[0631] The following describes the processing flow.

[0632] Step 1:

[0633] The device uses its built-in microphone to collect ambient sound in real time. This audio data is temporarily stored in a buffer and prepared for the next processing step.

[0634] Step 2:

[0635] The device applies a noise reduction filter to the stored audio data, clearing the audio signal. This enhances important audio features.

[0636] Step 3:

[0637] The device extracts features from the filtered audio data. These features are necessary information for detecting abnormal sounds, such as sound intensity and frequency spectrum.

[0638] Step 4:

[0639] The terminal packets the extracted features and sends them to the server via wireless communication along with the terminal's location information.

[0640] Step 5:

[0641] The server receives feature data transmitted from the terminal and stores it in a database. This data is then analyzed by an abnormal sound detection algorithm.

[0642] Step 6:

[0643] The server inputs the received features into a generating AI model and performs analysis to detect abnormal sounds. The model compares the similarity to existing abnormal sound patterns to determine the presence or absence of abnormal sounds.

[0644] Step 7:

[0645] If an unusual sound is detected, the server automatically notifies the relevant emergency services and generates and sends a report along with the location information of the incident site.

[0646] Step 8:

[0647] Based on the detection information, the server sends an alarm notification to nearby users' terminals and provides safety guidelines.

[0648] Step 9:

[0649] Users can receive notifications and, if necessary, adjust their privacy settings and choose how their data is handled within the application to manage their personal information.

[0650] (Example 1)

[0651] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0652] The problem that this invention aims to solve is to provide a system that can detect abnormal sounds in real time using audio information and provide appropriate notifications. Conventional technologies have limitations in the accuracy and responsiveness of abnormal sound detection, and also have shortcomings in terms of privacy management and notification to nearby users.

[0653] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0654] In this invention, the server includes means for analyzing audio information and using a generated AI model to identify abnormal sounds, means for notifying an external organization of an alarm when an abnormal sound is detected, and means for transmitting the abnormal sound detection result as a warning to the devices of nearby users. This enables real-time, highly accurate detection of abnormal sounds and a quick and appropriate response.

[0655] "Audio information" refers to sound data acquired from the surrounding environment, and this includes everyday sounds as well as specific abnormal sounds.

[0656] "Feature information" refers to specific acoustic elements extracted from audio information, and is fundamental data for identifying abnormal sounds.

[0657] "Analysis means" refers to devices or algorithms that use a generative AI model to analyze feature information and perform the process of identifying abnormal sounds based on this analysis.

[0658] "Alarm" refers to notification information sent to external organizations or nearby users to alert them to the occurrence of an unusual sound.

[0659] A "remote agency" refers to an organization or institution to which notifications are sent in order to take appropriate action in response to abnormal sounds detected from audio information.

[0660] "Privacy management" refers to a set of settings and processes that control how users' voice data is handled and ensure its use within permitted limits.

[0661] A "generative AI model" is an artificial intelligence technology used for analyzing audio information. It refers to a model that learns sound patterns and is used to identify abnormal sounds.

[0662] "User device" refers to a device owned by the user that is used to receive alarms and facilitate response actions.

[0663] This invention is a system that analyzes audio information in real time, quickly detects abnormal sounds, and provides relevant notifications. The terminal is implemented as a smartphone or IoT device and collects audio information from external sources using a built-in microphone. Audio processing software (e.g., audio editing software) that performs noise reduction and signal enhancement is utilized on the terminal.

[0664] The terminal that collects audio information is responsible for extracting feature information and sending it to the server. The server analyzes the received feature information using a generative AI model (for example, a model using TensorFlow or PyTorch) to identify abnormal sounds. The generative AI model recognizes the patterns in the audio and determines the abnormality by comparing them with known abnormal sound types.

[0665] If an abnormal sound is detected, the server generates an alarm and notifies remote agencies. The notification includes the precise location and details of the situation where the abnormal sound originated, enabling a rapid response. It also sends an alarm to nearby users' devices, urging them to take immediate evacuation action.

[0666] Users can freely customize their privacy settings and notification methods through the application. This allows for the management of voice information tailored to each individual user.

[0667] As a concrete example, if an unusual sound, such as glass breaking, occurs in a residential area late at night, the terminal detects the sound and sends its characteristic information to a server. When the generating AI model identifies the unusual sound, it automatically notifies the police, and evacuation guidance messages are sent to nearby residents.

[0668] Examples of prompt messages include specific scenarios such as "notification when abnormal sounds like breaking glass or screams are detected." This prompt enables the system to immediately recognize and respond to abnormal sounds.

[0669] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0670] Step 1:

[0671] The device collects ambient sounds. At this stage, the device acquires audio information from its built-in microphone and records this data in temporary memory. The audio information is the input data and is stored as an audio waveform on the device.

[0672] Step 2:

[0673] The terminal performs audio processing to reduce noise from the collected audio information. Here, digital signal processing is used for noise filtering, resulting in the output of clear audio data. During this process, the original audio information is input, and the noise-free audio is output.

[0674] Step 3:

[0675] The device extracts features from the noise-reduced audio data. This process calculates audio features (e.g., Mel-frequency cepstrum coefficients and spectral information). The input is the filtered audio data, and the output is the digitized feature information.

[0676] Step 4:

[0677] The terminal sends the extracted feature information to the server. The feature information is transferred to the server using a secure protocol. In this step, the input is the audio feature information, and the output is the completion of the data transfer to the server.

[0678] Step 5:

[0679] The server inputs the received feature information into a generative AI model. The generative AI model analyzes the audio pattern and compares it to known abnormal sounds. The input here is feature information sent from the terminal, and the output is the result of the determination regarding the presence or absence of abnormal sounds.

[0680] Step 6:

[0681] The server generates an alarm and notifies a remote agency when an abnormal sound is detected. The alarm includes location information and audio details to enable a rapid response. The input is the judgment result of the generating AI model, and the output is the transmission of the alarm notification.

[0682] Step 7:

[0683] The server sends an alarm to nearby users' devices, prompting them to respond quickly. The input is information about the detected abnormal sound, and the output is an alarm message to the user.

[0684] (Application Example 1)

[0685] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0686] Traditionally, abnormal sound detection systems have been inaccurate, making real-time notification difficult and hindering the rapid assurance of user safety. Furthermore, privacy protection measures were inadequate.

[0687] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0688] In this invention, the server includes means for a computation method that analyzes sound information and identifies abnormal sounds, means for managing user privacy, and means for constantly monitoring sound information with a receiving device and providing emergency notifications when a specific abnormal sound is detected. This enables highly accurate detection of abnormal sounds and real-time notification, improving user safety and privacy protection.

[0689] "Audio information" refers to sound data collected from the surrounding environment and is material that is analyzed for a specific purpose.

[0690] "Features" are important attributes and patterns extracted from sound information, and are data used to identify abnormal sounds.

[0691] A "receiving device" is a device used to collect or receive sound information or characteristic information.

[0692] "Calculation method" refers to a series of computational processes that analyze sound information and characteristic information to identify abnormal sounds.

[0693] An "external organization" refers to an external institution or group that shares information when an abnormal sound is detected.

[0694] "Privacy" refers to the right of users to have their personal information protected regarding the collection and use of sound information.

[0695] "Management" refers to actions taken to control and properly operate the processes related to the collection, analysis, and notification of sound information.

[0696] "Surveillance" refers to the activity of continuously observing sound information and capturing that information when specific conditions occur.

[0697] An "emergency notification" is a warning or information dissemination that is immediately sent to external organizations or users when an abnormal sound is detected.

[0698] The system that realizes this invention first uses a terminal to collect sound information. The terminal continuously collects ambient sounds from the surroundings using the built-in microphone of a smartphone. The collected sound information is filtered on the terminal using the "librosa" library to reduce noise and clarify the signal. In addition, features are extracted from the sound information during this process and preprocessed for abnormal sound detection.

[0699] Next, the server receives feature information sent from the terminal. On the server, a TensorFlow model is used to perform advanced analysis and analyze the sound information in real time. This generative AI model has the ability to compare features in the sound information with known abnormal sound patterns and determine the presence or absence of abnormal sounds. If an abnormal sound is detected, the server immediately notifies the external organization and also sends an alarm to the user's terminal.

[0700] Users receive notifications and respond promptly. Within the application, notification settings and privacy management are available, and users have control over their past data. This ensures user safety while protecting their privacy.

[0701] For example, if the sound of breaking glass occurs at night, the terminal quickly detects the sound and sends it to the server. The server analyzes the sound in real time, and if it recognizes an abnormal sound, it sends a warning to nearby users and also notifies emergency services. This process allows for a quick and effective response.

[0702] An example of a prompt to input into a generative AI model is, "Detect abnormal sounds from this audio data. Could the sound you are currently hearing indicate danger?"

[0703] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0704] Step 1:

[0705] The device collects ambient sounds in real time. It uses its built-in microphone to acquire sound information and receives it as input data. This sound information becomes the raw material for the next processing step.

[0706] Step 2:

[0707] The terminal filters the audio information using the "librosa" library. This filtering reduces noise and enhances the signal. The extracted audio feature data becomes the input for the next analysis step.

[0708] Step 3:

[0709] The terminal sends feature data to the server. The server receives this feature data and begins processing it as input information for advanced analysis.

[0710] Step 4:

[0711] The server analyzes the received feature data using a TensorFlow model. The generative AI model matches the data with abnormal sound patterns to identify abnormal sounds. If an abnormal sound is detected, the result becomes the output information for the next notification step.

[0712] Step 5:

[0713] Based on the detection results, the server sends notifications to external organizations and user terminals. This includes detailed information about the abnormal sound, along with any necessary warnings.

[0714] Step 6:

[0715] The user receives a notification on their device and checks the situation through the application. The user is required to take appropriate action based on the prompt. "An example of a prompt sentence to input into the generating AI model is: 'Detect abnormal sounds from this audio data. Could the sound you are hearing now indicate danger?'"

[0716] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0717] This invention combines a system that detects abnormal sounds based on audio data with an emotion engine that recognizes user emotions. By integrating audio data collection, feature extraction, abnormal sound detection, warning notification, emotion recognition, and privacy management, this system enables more personalized responses.

[0718] The device continuously records the user's ambient sounds and extracts features that indicate abnormal sounds from the collected audio data. Filtering reduces noise and improves the accuracy of abnormal sound detection. Furthermore, the device is equipped with an emotion engine that analyzes the user's emotional state from their voice tone and intonation. This emotion data is transmitted to the server along with the abnormal sound detection information.

[0719] The server receives data sent from the terminal and uses a generative AI model to identify abnormal sounds, as well as analyze the user's emotional data. Based on this, the server adjusts the priority and content of emergency notifications, taking into account the user's emotional state. For example, if the user is showing signs of anxiety, the server can activate the warning system more quickly and send encouraging and supportive messages appropriate to their emotional state.

[0720] Users can manage their privacy through the system's settings screen and customize how their emotional data is handled and what kind of notifications they receive. This provides a secure environment for using the emotional recognition system.

[0721] As a concrete example, consider a scenario where audio of someone shouting in an office is recorded. The terminal digitizes the audio and sends it to a server, where the emotion engine recognizes it as anger. The server assesses the urgency of the situation, sends notifications tailored to the user and the appropriate response agency, and the user receives appropriate notifications to help them calm down. In this way, dynamic responses tailored to individual situations become possible.

[0722] The following describes the processing flow.

[0723] Step 1:

[0724] The device collects ambient sound in real time using its built-in microphone. The collected audio data is stored in a buffer, and noise reduction filtering is applied to highlight important audio features.

[0725] Step 2:

[0726] The device analyzes the noise-reduced audio data to extract features for identifying abnormal sounds. Furthermore, it utilizes an integrated emotion engine to analyze emotions based on the user's voice tone and intonation.

[0727] Step 3:

[0728] The terminal packets the extracted abnormal sound features and emotion data, and sends them to the server along with the terminal's location information.

[0729] Step 4:

[0730] The server receives feature data and sentiment data transmitted from the terminal and stores it in a dedicated database. A generative AI model instantly identifies abnormal sounds and determines whether action is required.

[0731] Step 5:

[0732] The server analyzes the received emotional data and takes the user's emotional state into account when setting notification priorities and customizing notification content. If the user is experiencing high levels of anxiety, it is determined that a rapid response is necessary.

[0733] Step 6:

[0734] When an abnormal sound is detected, the server automatically sends a notification to the relevant emergency services, including the necessary information (type of abnormal sound, location information, and emotional information).

[0735] Step 7:

[0736] Based on abnormal sound detection and emotion analysis results, the server sends a notification to the user's device containing appropriate safety guidelines. This notification is individually customized based on the user's emotional state.

[0737] Step 8:

[0738] Based on the notifications they receive, users can further adjust their privacy settings and the handling of emotional data within the system, allowing them to use the system with peace of mind.

[0739] (Example 2)

[0740] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0741] Conventional abnormal sound detection systems were limited to detecting abnormal sounds based on audio data and lacked the ability to respond dynamically according to the user's psychological state. This could result in users receiving notifications at inappropriate times or excessive warnings, potentially undermining the system's effectiveness. Furthermore, insufficient privacy management meant users could not use the system with peace of mind.

[0742] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0743] In this invention, the server includes means for analyzing voice information, means for analyzing the user's emotional state, and control means for adjusting notification content based on the emotional state. This enables the provision of more personalized notifications that respond to the user's emotions and enhances privacy management.

[0744] "Audio information" refers to sounds collected from the environment and converted into data that can be processed in a digital format.

[0745] "Attributes" are pieces of information extracted from audio data that describe the characteristics and patterns of sound.

[0746] "Equipment" refers to hardware devices used for collecting, processing, and analyzing audio information.

[0747] A "processing unit" is a computing unit that analyzes audio information and attributes and makes classifications based on specific conditions.

[0748] "Warning information" refers to the content of warnings and notifications sent to external organizations or users when an anomaly is detected.

[0749] An "external organization" is an institution or group outside the system that is responsible for receiving notifications and responding in the event of an anomaly.

[0750] "Privacy" is a concept that protects users' personal information and sensitive data from being illegally acquired or used.

[0751] "Control means" refers to the processes and mechanisms that adjust and manage the operation of a system.

[0752] "Emotional state" refers to information that indicates the user's psychological state, representing the type and degree of emotion inferred from their voice and behavior.

[0753] This invention is a system that performs anomaly detection and emotion analysis using voice information and provides personalized notifications to users. Specific embodiments are described below.

[0754] Collection and processing of audio information

[0755] The device continuously records ambient sounds through a high-performance microphone, collecting them as audio information. This process involves specific filtering to reduce noise and enhance the signal from the collected audio information. The Python audio processing library librosa is used for this process. A digital signal processor (DSP) performs spectral analysis of the audio, extracting sound characteristics as attributes.

[0756] Emotion analysis

[0757] The device is equipped with an emotion engine that analyzes the user's emotional state from their voice. This uses a speech emotion recognition model based on TensorFlow. It analyzes features such as pitch, intonation, and rhythm of the voice to understand the user's emotional state.

[0758] Data transmission and analysis

[0759] The device sends extracted voice attribute information and the user's emotional state to the server. The server receives this data, analyzes the anomaly using a generative AI model, and further evaluates the user's emotional state in detail. Based on the type of anomaly and the emotional state, it individually adjusts the priority and content of notifications.

[0760] Notification generation

[0761] The server sends notifications directly to users based on the analysis results. For example, if a user expresses anxiety, the system is required to respond quickly and immediately provides a warning along with a message of encouragement.

[0762] As a concrete example, consider the occurrence of unusual noises in an office environment. If a terminal detects an unusual noise, such as someone shouting, the sound is recorded, processed, and sent to a server. The server analyzes the situation and sends psychologically appropriate notifications to users in the office, thereby reducing anxiety and encouraging immediate action.

[0763] An example of a prompt using a generative AI model is: "Please explain how to extract abnormal sounds from audio data and analyze the emotional state. Also, please explain step-by-step how to send an emergency notification based on the results." This prompt provides users and developers with guidance to accurately understand and effectively operate the system.

[0764] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0765] Step 1:

[0766] The device continuously acquires ambient sounds using a high-performance microphone. The input is ambient sound, and the output is digital audio data. This audio data is temporarily stored in local storage. During this process, the device filters out audio noise and enhances the signal to generate clean audio data suitable for analysis.

[0767] Step 2:

[0768] The terminal extracts attributes from saved audio data. The input is filtered audio data, and the output is attribute information representing the spectral characteristics of the sound. Specifically, the terminal uses the Python librosa library to split the audio waveform and perform spectral analysis. This analysis yields the frequency components and intensity of the sound as numerical vectors.

[0769] Step 3:

[0770] The device performs analysis using an emotion engine based on attribute information. The input is the spectral features of the speech, and the output is the user's emotional state. Specifically, the device runs an emotion recognition model using TensorFlow to classify emotions from the pitch and intonation of the speech. This allows the device to accurately grasp the user's emotional state.

[0771] Step 4:

[0772] The terminal sends extracted attribute information and emotional state to the server. The input consists of attribute information and emotional state, which are sent to the server as communication data. The terminal encrypts the data during this transmission to ensure security.

[0773] Step 5:

[0774] The server analyzes the received attribute information and emotional state. The input is data sent from the terminal, and the output is identification information to identify anomalies and notification content based on emotion. The server uses a generative AI model to determine if there are anomalies in the voice and further considers the user's emotional state to evaluate whether action is necessary.

[0775] Step 6:

[0776] The server generates and sends notifications to users based on the analysis results. Input consists of anomaly identification information and notification content based on sentiment, while output is a message including warnings and encouragement to the user. The server individually adjusts the priority and content of notifications to ensure they are delivered at the most appropriate time in specific situations.

[0777] (Application Example 2)

[0778] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0779] In modern society, there is a need to detect abnormal sounds occurring in an individual's living environment and to provide appropriate responses tailored to the user's emotional state. However, conventional systems are limited to analyzing only audio data, making it difficult to provide personalized responses that take into account the user's psychological state. Furthermore, balancing privacy security with a rapid warning notification system is also a challenge.

[0780] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0781] In this invention, the server includes means for detecting abnormal sounds from audio data, means for recognizing and analyzing the user's emotional state, and means for adjusting the priority and content of notifications based on the emotional data. This makes it possible to suggest individual alerts and relaxation methods tailored to each user's psychological state.

[0782] "Audio data" refers to sound that has been recorded and stored in digital format.

[0783] "Feature extraction" refers to extracting patterns and information useful for identification from audio data.

[0784] "Abnormal sounds" refer to sounds with specific frequencies or amplitudes that deviate from normal ambient sounds.

[0785] An "algorithm" is a set of steps or calculation methods designed to solve a specific problem.

[0786] A "warning" is a notification intended to draw attention to a danger or problem.

[0787] "Emotional state" refers to information that indicates a user's psychological or emotional response.

[0788] "Notification priority" refers to the criteria used to determine the order in which notifications are sent, based on the importance and urgency of the information being transmitted.

[0789] "Privacy" is a concept that protects personal information and prevents its unauthorized use or disclosure.

[0790] This invention is an automated security system that integrates the detection of abnormal sounds with the recognition of user emotions. The system mainly consists of three components: a terminal, a server, and a user.

[0791] The device has a built-in audio sensor for collecting ambient sounds, allowing for continuous recording of audio data. The collected audio data is preprocessed, and specific features are extracted. Python's LibROSA library is used for extracting audio features, performing noise reduction and signal enhancement. This improves the accuracy of detecting abnormal sounds.

[0792] The server receives audio feature data transmitted from the terminal and uses a generative AI model to identify abnormal sounds. TensorFlow is used for abnormal sound detection, and the system works with the terminal to quickly issue a warning when abnormal sounds are detected. At the same time, the server receives emotion data from the terminal and analyzes it using the Google Cloud Speech-to-Text API to recognize the user's emotional state.

[0793] Through this system, users can monitor their emotional state and the detection status of unusual sounds in real time. For example, if a siren suddenly sounds in a quiet living room, the device records the sound and sends it to the server. The server immediately identifies the sound as suspicious and sends a warning to the user. If the system determines that the user is in an anxious state, it also provides guidance on relaxation methods.

[0794] As an example of a prompt message, you can send instructions to the server in the format of, "Is there anything unusual about this audio data? Also, what emotional state is this user in?"

[0795] This invention enables quick and appropriate responses tailored to individual situations, allowing users to enjoy a safe and secure environment.

[0796] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0797] Step 1:

[0798] The device continuously collects ambient sound using an audio sensor. The collected audio data is filtered using LibROSA for noise reduction and signal enhancement, and its characteristics are extracted. The input is raw audio data, and the output is filtered characteristic data.

[0799] Step 2:

[0800] The terminal sends the extracted audio feature data to the server. The server receives this data and uses a generative AI model to identify abnormal sounds. The input is feature data, and the output is the result of determining whether or not an abnormal sound was detected. Specifically, it identifies the presence or absence of an abnormal sound and prepares a warning if necessary.

[0801] Step 3:

[0802] The server analyzes the user's emotional state along with the abnormal sound detection results. Based on the emotional data received from the terminal, it uses the Google Cloud Speech-to-Text API to analyze the user's voice tone and intonation to identify their emotions. The input is audio data and an emotion analysis model, and the output is the emotion detection result.

[0803] Step 4:

[0804] The server adjusts the priority and content of notifications based on the detected abnormal sound and the user's emotional state. For example, if an abnormal sound is detected and the user is determined to be in an anxious state, a warning is immediately issued, and relaxation advice is provided. The input is the detection result, and the output is the adjusted notification message.

[0805] Step 5:

[0806] Users receive warnings and advice sent by the device and consider how to address them. They can request additional information or actions using prompts. Specifically, if they want more details about a suspicious sound, they can prompt the device to re-analyze it.

[0807] In this way, this system provides a mechanism that simultaneously supports the safety of the voice environment and the psychological stability of the user.

[0808] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0809] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0810] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0811] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0812] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0813] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0814] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0815] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0816] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0817] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0818] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0819] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0820] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0821] 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.

[0822] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0823] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0824] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0825] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0826] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0827] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0828] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0829] The following is further disclosed regarding the embodiments described above.

[0830] (Claim 1)

[0831] A device for collecting audio data,

[0832] A device for extracting features from the aforementioned audio data,

[0833] A device that receives the extracted feature data,

[0834] A device equipped with an algorithm that analyzes the aforementioned characteristic data and identifies abnormal sounds,

[0835] A device that notifies an external organization of a warning when an abnormal sound is detected,

[0836] A device that provides controls for managing privacy,

[0837] A system that includes this.

[0838] (Claim 2)

[0839] The system according to claim 1, which performs filtering processing to reduce noise and enhance the signal when collecting audio data.

[0840] (Claim 3)

[0841] The system according to claim 1, which performs control to notify the user of safety guidelines based on the detection result of an abnormal sound.

[0842] "Example 1"

[0843] (Claim 1)

[0844] Means for collecting audio information,

[0845] A means for extracting features from the aforementioned audio information,

[0846] A means for receiving the extracted feature information,

[0847] An analysis means for analyzing the aforementioned characteristic information and identifying abnormal sounds,

[0848] A means of notifying a remote agency of an alarm when an abnormal sound is detected,

[0849] Means to change settings to manage privacy,

[0850] A method of using a generative AI model to identify abnormal sounds in real time,

[0851] A means for transmitting the detection result of an abnormal sound as a warning to the devices of nearby users,

[0852] A system that includes this.

[0853] (Claim 2)

[0854] The system according to claim 1, which performs audio processing to reduce noise and enhance the signal when collecting audio information.

[0855] (Claim 3)

[0856] The system according to claim 1, which performs control to notify users of evacuation guidelines based on the detection of abnormal sounds.

[0857] "Application Example 1"

[0858] (Claim 1)

[0859] Means for collecting sound information,

[0860] A means for extracting features from the aforementioned sound information,

[0861] A means for receiving the extracted feature information,

[0862] A means comprising a calculation method for analyzing the aforementioned characteristic information and identifying abnormal sounds,

[0863] A means of sending a notification to an external organization when an abnormal sound is detected,

[0864] Means of managing and manipulating user privacy,

[0865] A means of continuously monitoring sound information using a receiving device and issuing an emergency notification when a specific abnormal sound is detected,

[0866] A system that includes this.

[0867] (Claim 2)

[0868] The system according to claim 1, which performs filtering processing to reduce noise and enhance signals when collecting sound information.

[0869] (Claim 3)

[0870] The system according to claim 1, which performs control to notify the user of safety guidelines based on the detection result of an abnormal sound.

[0871] "Example 2 of combining an emotion engine"

[0872] (Claim 1)

[0873] A device that collects voice information,

[0874] A device for extracting attributes from the aforementioned audio information,

[0875] A device that receives the extracted attribute information,

[0876] A processing device that analyzes the attribute information and determines abnormalities,

[0877] A device that notifies an external organization of warning information when an abnormality is detected,

[0878] Devices that provide controls for managing privacy,

[0879] A device that analyzes the emotional state of the user,

[0880] A control means that adjusts the notification content based on the aforementioned emotional state,

[0881] A system that includes this.

[0882] (Claim 2)

[0883] The system according to claim 1, which performs processing to reduce noise and enhance the signal when collecting audio information.

[0884] (Claim 3)

[0885] The system according to claim 1, which provides appropriate information to the user based on the detection result of an anomaly and the user's emotional state.

[0886] "Application example 2 when combining with an emotional engine"

[0887] (Claim 1)

[0888] Means of collecting audio data,

[0889] A means for extracting features from the aforementioned audio data,

[0890] A means for receiving the extracted feature data,

[0891] A means comprising an algorithm for analyzing the aforementioned characteristic data and identifying abnormal sounds,

[0892] A means of notifying an external organization of a warning when an abnormal sound is detected,

[0893] A means for recognizing and analyzing the emotional state of a user,

[0894] A means of adjusting the priority and content of notifications based on sentiment data,

[0895] Means of implementing controls to manage privacy,

[0896] A system that includes this.

[0897] (Claim 2)

[0898] The system according to claim 1, which performs filtering processing to reduce noise and enhance the signal when collecting audio data.

[0899] (Claim 3)

[0900] The system according to claim 1, which controls the system to notify the user of safety guidelines and relaxation suggestions based on the detection results of abnormal sounds and emotion analysis. [Explanation of Symbols]

[0901] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A device for collecting audio data, A device for extracting features from the aforementioned audio data, A device that receives the extracted feature data, A device equipped with an algorithm that analyzes the aforementioned characteristic data and identifies abnormal sounds, A device that notifies an external organization of a warning when an abnormal sound is detected, A device that provides controls for managing privacy, A system that includes this.

2. The system according to claim 1, which performs filtering processing to reduce noise and enhance the signal when collecting audio data.

3. The system according to claim 1, which performs control to notify the user of safety guidelines based on the detection result of an abnormal sound.

Images