system

Abstract

Provide a system. A means for receiving audio data; means for preprocessing received audio data; a means for analyzing the pre-processed voice data and using machine learning models to predict emotions and needs; means for transmitting the prediction result to a user terminal; means for notifying a user of the received prediction result; A system including:

Description

[Technical Field]

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

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

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

[0004] Understanding the cries of babies and animals is a significant challenge for parents and owners. They are often unable to identify the cause of the cry, which often delays appropriate response. This invention aims to solve this problem by providing a system that analyzes crying data, predicts the cause and emotion of the cry, and provides feedback to the parent or owner. Specifically, it provides a method for collecting and analyzing crying data and quickly notifying the user of the results. [Means for solving the problem]

[0005] The present invention solves the above problems by the following means. First, a means for receiving voice data is provided. Next, a means for preprocessing the received voice data is provided. This preprocessing includes noise removal and unifying the sampling rate. Furthermore, a means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs is provided. The machine learning model used for this analysis is composed of a convolutional neural network and a recurrent neural network. Next, a means for transmitting the prediction results to a user terminal and a means for notifying the received prediction results to the user are provided. In this way, a system is provided that can analyze the cause of crying and emotions in real time and provide rapid feedback.

[0006] "Audio data" is data that is a digital recording of a human or animal voice.

[0007] The "receiving means" is a device or part of a system that has the function of receiving audio data from the outside.

[0008] The "preprocessing means" refers to a device or function that performs necessary processing on the audio data before analysis, such as noise removal and standardization of the sampling rate.

[0009] The "analyzing means" is a device or function that extracts the characteristics of the preprocessed voice data and performs classification and prediction based on the data.

[0010] A "machine learning model" is an algorithm that learns certain patterns and characteristics based on past data and makes predictions and classifications for new data.

[0011] A "convolutional neural network (CNN)" is a type of deep learning algorithm for recognizing spatial patterns in data, and is a model primarily used for image and audio analysis.

[0012] A "recurrent neural network (RNN)" is a type of deep learning algorithm designed to process time series data, and is primarily a model for learning patterns in sequential data.

[0013] "Prediction results" refer to specific outputs such as emotions and needs based on data analyzed by the machine learning model.

[0014] "User terminal" refers to a device such as a computer or smartphone used by a parent or owner.

[0015] The "notification means" is a device or function for conveying information such as analysis results to the user.

[0016] "System" refers to a set of devices or software that includes the above means and models and operates in cooperation with each other. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8]FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0018] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0019] First, the terms used in the following description will be explained.

[0020] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0021] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0022] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0023] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0024] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0028] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0029] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0030] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0032] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0035] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0036] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0037] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0038] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. Hereinafter, an embodiment of the system of the present invention will be described in detail.

[0039] Data collection and preprocessing

[0040] The server collects a large amount of crying data sets labeled by experts and stores them in a cloud database. The data sets include audio files of crying babies and animals and corresponding labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed by removing noise and standardizing the sampling rate, resulting in data suitable for analysis.

[0041] Training a machine learning model

[0042] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). This model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and can predict emotions and needs based on that information. The resulting model is stored on a cloud server and used for real-time analysis.

[0043] Audio recording and real-time analysis

[0044] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotions and needs indicated by the crying are predicted.

[0045] Sending and notifying feedback

[0046] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent as a push notification to the user's smartphone. The device receives this and immediately notifies the user.

[0047] Specific examples

[0048] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's cry and sends the data to the server. The server analyzes this audio data and predicts that the baby is hungry. The prediction result is returned from the server to the device, which then notifies the user with a message saying, "Your baby may be hungry." The user can check this notification and immediately prepare milk to feed the baby. In this way, a quick and accurate response is possible.

[0049] This invention makes it possible to quickly determine why a baby or animal is crying, making it easier for parents and owners to take appropriate action, which is expected to reduce stress for babies and animals and improve their health and happiness.

[0050] The processing flow will be explained below.

[0051] Step 1:

[0052] The server collects expert-labeled datasets of baby and animal cries and stores them in a cloud database, which includes audio files and corresponding labels such as "hungry," "sleepy," and "anxious."

[0053] Step 2:

[0054] The server preprocesses the collected audio data, including noise removal, unifying the sampling rate, normalizing, etc. The preprocessed data is then converted into a form suitable for analysis.

[0055] Step 3:

[0056] The server uses the preprocessed data to train machine learning models, specifically convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to learn audio features (e.g., pitch, rhythm, decibels).

[0057] Step 4:

[0058] The server stores the trained machine learning model in the cloud, which is then used for real-time analytics.

[0059] Step 5:

[0060] The device has the ability to record the crying sounds of babies and animals in real time. When a crying sound is detected, the device will start recording the audio data.

[0061] Step 6:

[0062] The device transmits the recorded audio data to a cloud server via the Internet.

[0063] Step 7:

[0064] The server then pre-processes the received audio data again, which includes noise reduction and sampling rate unification.

[0065] Step 8:

[0066] The server then uses the pre-processed audio data to analyze it using a trained machine learning model, which predicts the crying as an emotion or need, such as "hunger," "sleepy," or "anxiety."

[0067] Step 9:

[0068] The server sends the analysis results to the user's device, which include a specific alert message.

[0069] Step 10:

[0070] The device will then notify the user of the received prediction results, which will be displayed as a push notification or an alert.

[0071] Step 11:

[0072] The user can check the notification and respond to the baby or animal's needs based on the notification. For example, if the baby or animal is predicted to be hungry, the user can prepare milk and give it to the baby. This process allows parents and owners to quickly understand the cause of the crying and take appropriate action.

[0073] Example 1

[0074] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0075] Conventional voice recognition systems lack the means to analyze the cries of babies and animals and provide fast and accurate feedback on their emotions and needs. Furthermore, due to the lack of proper noise reduction and standardization of sampling rates for each individual cry, analysis results tend to be inaccurate. Furthermore, the lack of a function to notify the user of prediction results in real time makes it difficult to respond quickly. There is a need to address these issues with conventional technology.

[0076] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0077] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data based on noise reduction and unified sampling rate, means for analyzing the preprocessed voice data and using a machine learning model consisting of a convolutional neural network and a recurrent neural network to predict emotions and needs, means for transmitting the prediction results to a user terminal, and means for sending a push notification of the received prediction results to the user. This makes it possible to quickly and accurately analyze the cries of babies and animals and predict emotions and needs based on the analysis, and notify the user of the results in real time.

[0078] "Voice data" is data that represents the voice of a human or animal in digital form.

[0079] A "means for receiving" is a method or apparatus for obtaining audio data from another device or system.

[0080] "Noise reduction" is the process of removing unwanted background sounds and noise from audio data.

[0081] "Sampling rate unification" is the process of making the sampling frequencies of different audio data uniform.

[0082] "Preprocessing means" refers to a method or device that performs necessary processing or conversion on the audio data before analysis.

[0083] "Emotions and needs" refers to the feelings and demands that babies and animals express through their cries.

[0084] A "convolutional neural network" is a type of deep learning model that excels at analyzing data such as images and audio.

[0085] A "recurrent neural network" is a type of deep learning model that excels at analyzing continuous data over time.

[0086] A "machine learning model" is an artificial intelligence algorithm that uses data to learn and perform specific tasks.

[0087] A "means of use" is a method or device that applies a machine learning model to perform analysis or prediction.

[0088] A "user terminal" is a device such as a smartphone, PC, or tablet that a user interacts with directly.

[0089] A "transmitting means" is a method or apparatus for sending data to another device or system.

[0090] "Push notification" is a communication technology that sends specific information to a user's device in real time and displays it instantly.

[0091] A "cloud server" is a server that includes computing resources and storage provided over the Internet, and is a place where data is stored and computational processing is performed.

[0092] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. An embodiment of the system of the present invention will be described in detail below.

[0093] Data collection and preprocessing

[0094] The server collects a large amount of expert-labeled crying data sets from the Internet and stores them in a cloud database. The collected data includes audio files of babies and animals crying and their corresponding labels, such as "hungry," "sleepy," and "anxious." The server then preprocesses the collected audio data using the Python LibROSA library, removing noise and standardizing the sampling rate.

[0095] Training a machine learning model

[0096] The server splits the preprocessed data into a training dataset and a test dataset. Then, it uses TENSORFLOW® to build and train a convolutional neural network (CNN) or a recurrent neural network (RNN). The trained model is stored on the cloud server.

[0097] Audio recording and real-time analysis

[0098] The device (e.g., the user's smartphone or a dedicated device) has the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again using the LibROSA library and analyzes it using a trained machine learning model. This allows the emotions and needs indicated by the cry to be predicted.

[0099] Sending and notifying feedback

[0100] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent to the user's smartphone as a push notification. The device receives this and immediately notifies the user.

[0101] Specific examples

[0102] For example, if your baby starts crying in the middle of the night,

[0103] 1. The device records the baby's cry and sends the data to the server, which then activates the smartphone's recording app and captures the audio.

[0104] 2. The server re-preprocesses the received audio data and analyzes it using the trained machine learning model. It uses LibROSA on the server side to remove noise and input the data into the model.

[0105] 3. The server obtains the prediction result, "I'm hungry," and sends it to the device. The server uses an HTTP POST request to send the prediction result in JSON format.

[0106] 4. The device pushes a message to the user saying, "Your baby might be hungry." The smartphone displays the notification so that the user can check it.

[0107] The system can quickly determine why a baby or animal is crying, allowing parents and owners to take appropriate action, reducing stress and improving their health and happiness.

[0108] Prompt Sentence Examples

[0109] "Please explain how your smartphone automatically analyzes your baby's emotions and needs and notifies you when your baby is crying."

[0110] Based on this example, generative AI models can be used to provide context-sensitive feedback.

[0111] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0112] Step 1: Data collection

[0113] The server collects expertly labeled baby and animal cry datasets from the Internet and stores them in a cloud database. It performs the following operations:

[0114] Input: A crying audio file and its corresponding label (e.g., "hungry" or "sleepy").

[0115] Data Processing: During the data collection process, the data is stored in a database in a suitable format.

[0116] Output: A crying dataset stored in a cloud database.

[0117] Step 2: Data Preprocessing

[0118] The server performs noise reduction and sampling rate unification on the collected audio data. It uses the Python LibROSA library and performs the following operations.

[0119] Input: Audio data stored in a cloud database.

[0120] Data processing: LibROSA was used to remove background noise from the audio files and standardize the sampling rate to 44.1kHz.

[0121] Output: Preprocessed audio data.

[0122] Step 3: Train the model

[0123] The server splits the preprocessed data into training and test datasets and trains the model using a deep learning framework. It performs the following operations:

[0124] Input: Preprocessed audio data.

[0125] Data processing: 80% of the total data will be used as a training dataset and 20% as a test dataset.

[0126] Output: A dataset split into training and testing.

[0127] The server builds and trains models using convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

[0128] Input: Training dataset.

[0129] Data calculation: We use TensorFlow to train the model with 10 epochs.

[0130] Output: A trained machine learning model.

[0131] The server stores the trained model on a cloud server.

[0132] Input: A trained model.

[0133] Data calculation: Save the model using TensorFlow's model.save method.

[0134] Output: A trained model stored on a cloud server.

[0135] Step 4: Record your audio

[0136] The device (e.g., a smartphone) records the cries of babies and animals in real time. It performs the following operations:

[0137] Input: Babies and animals crying.

[0138] Data processing: Capture audio using a recording app.

[0139] Output: Recorded audio data.

[0140] Step 5: Sending audio data

[0141] The device sends the recorded audio data to a cloud server via the Internet. The following operations are performed:

[0142] Input: Recorded audio data.

[0143] Data calculation: Upload audio data to the server via HTTP POST request.

[0144] Output: The audio data sent to the server.

[0145] Step 6: Real-time analysis

[0146] The server then preprocesses the received audio data again and analyzes it using the trained machine learning model.

[0147] Input: Transmitted audio data.

[0148] Data processing: LibROSA is used to remove noise and unify the sampling rate again.

[0149] Output: Audio data in a format suitable for analysis.

[0150] The server uses trained machine learning models to analyze voice data and predict emotions and needs.

[0151] Input: Preprocessed audio data.

[0152] Data calculation: Using machine learning models, predict emotions and needs such as "hungry" or "sleepy."

[0153] Output: Predicted emotion and needs data.

[0154] Step 7: Submit your feedback

[0155] The server sends the analysis results to the user's terminal and performs the following operations.

[0156] Input: Parsed emotion and needs data.

[0157] Data calculation: Prediction results are sent to the terminal in JSON format via an HTTP POST request.

[0158] Output: Analysis results sent to the user's terminal.

[0159] Step 8: User Notification

[0160] The device will then push the received analysis results to the user. The device will perform the following operations:

[0161] Input: Analysis results sent from the server.

[0162] Data calculation: A message is displayed using the smartphone's notification function.

[0163] Output: The message that was sent to the user.

[0164] This step makes it possible to analyze the cries of babies and animals, predict their emotions and needs based on the analysis, and notify the user of the results in real time.

[0165] (Application example 1)

[0166] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0167] Conventional crying analysis systems simply analyze voice data on a local device, resulting in problems with the accuracy and real-time nature of the analysis results. Furthermore, when used in pet shops and veterinary clinics, a quick and appropriate response is required, particularly on-site, so it is necessary to provide more accurate analysis results. This invention aims to improve the accuracy of voice data analysis by utilizing a cloud server and provide users with quick and appropriate feedback.

[0168] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0169] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for transmitting the prediction results to a user terminal, means for notifying the user of the received prediction results, means for transmitting the received voice data to a cloud server and acquiring analysis results at the cloud server, and means for generating notification content based on the analysis results and instructing the user on an appropriate response. This allows highly accurate analysis results to be obtained quickly, enabling the user to immediately take appropriate measures.

[0170] "Audio data" is information that has been digitally recorded as the cries and cries of babies and animals.

[0171] "Preprocessing" is the process of processing audio data, such as removing noise and standardizing the sampling rate, to prepare it in a form suitable for analysis.

[0172] A "machine learning model" is a computational model that includes algorithms for learning from data and making predictions and classifications for new data.

[0173] "Deep learning" is a type of machine learning that uses multi-layered neural networks to analyze data and learn complex patterns.

[0174] A "convolutional neural network (CNN)" is a machine learning model primarily used to analyze image and audio data, and is a network with the ability to capture the spatial structure of data.

[0175] A "recurrent neural network (RNN)" is a machine learning model suitable for analyzing time series data, and is a network that proceeds with learning by taking into account the temporal dependency of the data.

[0176] A "cloud server" is a remote server that stores data and performs calculations over the Internet, and is a computer that provides services in response to requests from a large number of users.

[0177] A "user terminal" is a device used to record voice data, receive analysis results, and receive notifications, and specific examples include smartphones and dedicated devices.

[0178] A "notification" is a message that conveys analysis results and instructions based on those results to the user, and is mainly displayed in the form of a push notification or an alert.

[0179] "Analysis results" are predicted information about emotions and needs obtained after analyzing voice data using a machine learning model.

[0180] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. An embodiment of the system of the present invention will be described in detail below.

[0181] Data collection and preprocessing

[0182] The server collects a large amount of crying data sets labeled by experts and stores them in a cloud database. The data sets include audio files of crying babies and animals and corresponding labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed by removing noise and standardizing the sampling rate, resulting in data suitable for analysis.

[0183] Training a machine learning model

[0184] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). This model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and can predict emotions and needs based on that information. The resulting model is stored on a cloud server and used for real-time analysis.

[0185] Audio recording and real-time analysis

[0186] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotions and needs indicated by the crying are predicted.

[0187] Sending and notifying feedback

[0188] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent as a push notification to the user's smartphone. The device receives this and immediately notifies the user.

[0189] Examples from pet shops and veterinary clinics

[0190] The Pet Care Assistant application analyzes the cries (cries) of pets brought in by customers at pet shops and veterinary clinics. For example, if a dog barks at a pet shop, the device records the sound and sends it to a server on the cloud. The server analyzes the audio data and obtains a prediction result, such as "hungry." As a result, the store clerk receives a message saying, "This dog may be hungry," and can recommend appropriate dog food to the owner.

[0191] Hardware and software used

[0192] Hardware used

[0193] Smartphone: A device that uses the recording function to capture voice data and send it to a server

[0194] Cloud server: Runs the deep learning model and analyzes the voice data

[0195] Software used

[0196] Python: Writing program logic

[0197] pydub: A library for preprocessing audio data

[0198] requests: A library for managing HTTP requests / responses.

[0199] Examples and Prompts

[0200] For example, if a dog in a pet shop is barking for a specific reason, the Pet Care Assistant can analyze the sound and notify the store clerk that "This dog may be hungry," allowing the store to provide prompt and appropriate service to customers.

[0201] Prompt Sentence Examples

[0202] Audio data: Audio data in BASE64 format

[0203] Device ID: Device identifier

[0204] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0205] Step 1:

[0206] The device records surrounding audio in real time.

[0207] Input: Audio signal from the external environment

[0208] Output: Recorded audio data (e.g., WAV file)

[0209] What it does: Uses the microphone on your smartphone or dedicated device to capture the cries of babies and animals.

[0210] Step 2:

[0211] The terminal preprocesses the recorded audio data.

[0212] Input: Recorded audio data

[0213] Output: Preprocessed audio data

[0214] Specific operations: Noise removal and unification of sampling rates are performed on the audio data to prepare it in a form suitable for analysis.

[0215] Step 3:

[0216] The terminal transmits the pre-processed voice data to a server on the cloud via the Internet.

[0217] Input: Preprocessed audio data

[0218] Output: Data transmission result to cloud server (success / failure)

[0219] Specific operation: Generate an HTTP request and send the preprocessed audio data to the cloud server.

[0220] Step 4:

[0221] The server again pre-processes the received audio data.

[0222] Input: Preprocessed audio data

[0223] Output: Re-preprocessed audio data

[0224] Specific operation: Additional noise reduction and format conversion is performed on the server side as needed.

[0225] Step 5:

[0226] The server analyzes the audio data using a trained machine learning model.

[0227] Input: Re-preprocessed audio data

[0228] Output: Prediction of emotions and needs corresponding to crying

[0229] Specific operation: Using convolutional neural networks (CNN) and recurrent neural networks (RNN), voice features are extracted and emotions and needs are predicted.

[0230] Step 6:

[0231] The server transmits the analysis results to the user's terminal.

[0232] Input: Prediction result (e.g., "I'm hungry," "I'm sleepy," etc.)

[0233] Output: Data transmission result to user device (success / failure)

[0234] Specific operation: Generate an HTTP response and send a message containing the prediction result to the user terminal.

[0235] Step 7:

[0236] The terminal notifies the user of the received analysis results.

[0237] Input: A message containing the analysis results

[0238] Output: Notification to the user (e.g. push notification)

[0239] Specific operation: The analysis results are displayed as a push notification using the notification function of the user's smartphone or dedicated device.

[0240] Step 8:

[0241] The user checks the notification and takes appropriate action.

[0242] Input: Notification of analysis results

[0243] Output: User's response (e.g., feeding a baby, feeding a dog, etc.)

[0244] Specific Action: Based on the notification content, the user performs an action according to the request of the baby or pet.

[0245] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0246] The present invention relates to a system that analyzes the cries of babies and animals, recognizes not only their emotions and needs but also the emotions of a user, and provides feedback. An embodiment of the system of the present invention will be described in detail below.

[0247] Data collection and preprocessing

[0248] The server collects a large amount of crying data labeled by experts and stores it in a cloud database. The data set includes audio files of crying babies and animals, along with labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed, including noise removal and standardization of the sampling rate, to obtain data suitable for analysis.

[0249] Training a machine learning model

[0250] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). The model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs based on them. The resulting model is stored on a cloud server and used for real-time analysis.

[0251] User Emotion Recognition

[0252] The device is equipped with an emotion engine that recognizes the user's emotions based on their voice input. This engine analyzes the user's speech data in real time and recognizes their emotional state, such as "fatigue," "stress," or "relief." This is done automatically when the user speaks to the device.

[0253] Audio recording and real-time analysis

[0254] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the crying of a baby or animal in real time. When a cry is detected, the device records the audio data and sends it to a server in the cloud. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotion and needs indicated by the cry are predicted.

[0255] Sending and notifying feedback

[0256] The server then sends the results of the analysis of the baby or animal's cry, along with the user's emotion recognition results, to the user's device. Specifically, the server sends a message containing information about the cause of the cry and advice or suggestions based on the user's emotional state.

[0257] The device will notify the user of these results, which will be displayed as push notifications or alerts, allowing the user to take immediate action.

[0258] Specific examples

[0259] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device recognizes the user's emotions, such as "fatigue" and "stress," in real time. The server analyzes the received voice data and predicts that the user is "hungry." If the server determines that the user is feeling tired, it creates a message containing advice such as "Your baby may be hungry. If possible, ask a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[0260] This embodiment allows for quick understanding of why a baby or animal is crying and allows for a response that takes into account the parent's or owner's own emotional state, which is expected to improve the health and well-being of not only the baby or animal, but also the parent or owner themselves.

[0261] The processing flow will be explained below.

[0262] Step 1:

[0263] The server collects expert-labeled datasets of baby and animal cries and stores them in a cloud database, including audio files and labels such as "hungry," "sleepy," and "anxious."

[0264] Step 2:

[0265] The server preprocesses the collected audio data, which includes noise removal and sampling rate unification, to prepare the data for analysis.

[0266] Step 3:

[0267] The server uses the preprocessed data to train machine learning models, which are based on convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

[0268] Step 4:

[0269] The server stores the trained machine learning model in the cloud, which is then used for subsequent real-time analysis.

[0270] Step 5:

[0271] The device has the ability to record the crying sounds of babies and animals in real time. When a crying sound is detected, the device will start recording the audio data.

[0272] Step 6:

[0273] The device transmits the recorded voice data to a server via the Internet, and simultaneously detects the user's voice input and analyzes it using an emotion engine.

[0274] Step 7:

[0275] The server preprocesses the received audio data, which includes noise reduction and sampling rate unification.

[0276] Step 8:

[0277] The server uses the pre-processed audio data to analyze it with a trained machine learning model, which predicts the cause of the crying (hungry, sleepy, etc.).

[0278] Step 9:

[0279] The server integrates the results of the user's emotion recognition with the results of the crying analysis. For example, if the user feels "tired," that information is also included in the prediction results.

[0280] Step 10:

[0281] The server sends the integrated results to the user's terminal, and the message includes the cause of the crying as well as advice based on the user's condition.

[0282] Step 11:

[0283] The device notifies the user of messages received from the server, which are displayed as push notifications or alerts.

[0284] Step 12:

[0285] The user can check the notification and respond to the baby or animal's needs based on the notification. For example, if the system predicts that the baby is hungry and the user feels tired, it will suggest asking other family members for help. The user can also take action such as feeding the baby if necessary.

[0286] This series of steps enables the system to quickly identify the cause of a baby or animal's cry and provide appropriate advice that takes into account the user's emotional state.

[0287] Example 2

[0288] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0289] Conventional emotion recognition systems are limited in their ability to analyze the cries of babies and animals, and are unable to simultaneously recognize the user's emotional state and provide appropriate feedback. This makes it difficult for parents and owners to respond efficiently and quickly, often resulting in stress and fatigue. Furthermore, analysis results can sometimes be inaccurate, preventing effective countermeasures from being implemented.

[0290] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0291] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for analyzing the user's voice data in real time and recognizing the emotional state, means for recording the crying of a baby or animal in real time and transmitting the voice data to the server, means for transmitting the prediction result to a user terminal, and means for notifying the user of the received prediction result. This makes it possible to quickly and accurately understand the meaning of the crying of a baby or animal and provide appropriate feedback according to the user's emotional state.

[0292] "Means for receiving audio data" refers to the function that allows a user terminal or dedicated device to record the cries of babies or animals and send them to a server.

[0293] "Means for preprocessing received voice data" refers to a function for performing preprocessing such as noise removal and unifying the sampling rate on collected voice data.

[0294] "Means of analyzing pre-processed voice data and using machine learning models to predict emotions and needs" refers to a function that analyzes voice data using deep learning convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to predict emotions and needs.

[0295] "Means for analyzing the user's voice data in real time and recognizing their emotional state" refers to a function that analyzes the voice data spoken by the user in real time and recognizes their emotional state, such as "fatigue" or "stress."

[0296] "Means for recording the cries of babies and animals in real time and transmitting the audio data to a server" refers to a function that enables a user terminal or dedicated device to record the cries of babies and animals in real time and transmit the audio data to a cloud server.

[0297] "Means for transmitting prediction results to a user terminal" refers to a function that enables the server to transmit analysis results and feedback messages to a user terminal.

[0298] "Means for notifying the user of the received prediction results" refers to a function that allows the user's device to notify the user of the prediction results received from the server as a push notification or an alert.

[0299] MODE FOR CARRYING OUT THE INVENTION

[0300] The present invention relates to a system that analyzes the cries of babies and animals, recognizes not only their emotions and needs but also the emotions of the user, and provides feedback. This system performs a series of processes from collecting audio data to analyzing it and providing feedback. An embodiment of the system of the present invention will be described in detail below.

[0301] Data collection and preprocessing

[0302] The server collects a large amount of expert-labeled crying data sets from online repositories and stores them in a cloud database. The collected data includes audio files of crying babies and animals, along with labels such as "hungry," "sleepy," and "anxious." The data is then converted to a uniform sampling rate (e.g., 44.1 kHz) and preprocessed with noise reduction (using a Gaussian filter, etc.).

[0303] Training a machine learning model

[0304] The server then loads the preprocessed audio data into a deep learning environment (e.g., TensorFlow or PyTorch) and trains it using a convolutional neural network (CNN) or recurrent neural network (RNN). This training creates a model that learns the feature vectors of the audio data (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs with high accuracy. After training is complete, the model is stored on a cloud server and made available for real-time analysis.

[0305] User Emotion Recognition

[0306] The device is equipped with an emotion engine that analyzes emotions in real time based on the user's voice input. This engine uses, for example, IBM Watson (registered trademark) Tone Analyzer to analyze the voice data (e.g., "tired" or "irritated") spoken by the user and recognizes the user's emotional state. This real-time recognition is performed in a natural situation where the user is speaking to the device.

[0307] Audio recording and real-time analysis

[0308] The device (e.g., a smartphone or dedicated device) has the ability to record the cries of babies and animals in real time. When a cry is detected, the device automatically starts recording and sends the audio data to a cloud-based server. The server then preprocesses the received audio data and analyzes it using a trained machine learning model. This allows for highly accurate prediction of the emotions and needs indicated by the cry.

[0309] Sending and notifying feedback

[0310] The server combines the results of the crying analysis and the user's emotion recognition and sends them to the user's device. Specific feedback includes information about the cause of the crying and advice or suggestions based on the user's emotional state. For example, "Your baby may be hungry. If possible, ask a family member for help." The device then notifies the user of this feedback via a push notification or alert, encouraging them to take prompt action.

[0311] Specific examples

[0312] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device analyzes the user's voice and recognizes emotions such as "fatigue" and "stress" in real time. The server analyzes the received crying data and predicts that the baby is "hungry." If the server determines that the user is tired, it creates a message containing advice such as "Your baby may be hungry. If possible, try asking a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[0313] Prompt Sentence Examples

[0314] 1. "Analyze an audio file of a crying baby and predict its emotions and needs. The user's emotional state is 'Tired'."

[0315] 2. "Analyze a dog's cry and tell me its cause. The user's emotion is 'stress'."

[0316] This system can quickly and accurately identify the emotions and needs expressed by the cries of babies and animals, and provide appropriate feedback based on the user's emotional state, which is expected to improve the health and happiness of not only babies and animals, but also their parents and owners.

[0317] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0318] Step 1: Collect data

[0319] The server collects expert-labeled crying datasets from online repositories. The input includes crying audio data and corresponding emotion and need labels. The output is stored in a cloud database.

[0320] Step 2: Preprocessing the data

[0321] The server performs preprocessing on the collected audio data. The raw audio data collected in step 1 is used as input. Specific preprocessing operations include noise removal (applying a Gaussian filter) and unifying the sampling rate (e.g., changing to 44.1 kHz). The preprocessed audio data is obtained as output.

[0322] Step 3: Train the machine learning model

[0323] The server uses the preprocessed audio data to train a machine learning model in a deep learning environment (such as TensorFlow or PyTorch). The preprocessed audio data and its associated label information are used as input. Specifically, a convolutional neural network (CNN) or recurrent neural network (RNN) is used to learn feature vectors (e.g., pitch, rhythm, decibels). The output is a trained model.

[0324] Step 4: Recognizing user emotions

[0325] The device collects the user's voice data in real time and analyzes their emotions. The user's spoken voice data is used as input. Specifically, it uses an emotion engine (for example, IBM Watson Tone Analyzer) to recognize emotional states such as "fatigue" and "stress" in real time. The output is the recognition result of the user's emotional state.

[0326] Step 5: Record your audio

[0327] The device records the cries of babies and animals in real time. The cries of babies and animals are used as input. Specifically, the device's microphone is used to record the cries and save them as audio files. The recorded audio data is obtained as output.

[0328] Step 6: Real-time analysis

[0329] The device sends the recorded crying data to a server on the cloud. The audio data recorded in step 5 is used as input. The server receives this audio data, preprocesses it again, and then analyzes it using a trained machine learning model. Specific operations include preprocessing (noise removal, unifying the sampling rate) and feature extraction (pitch, rhythm, decibel analysis). The analysis results are obtained as output.

[0330] Step 7: Feedback Generation

[0331] The server generates a feedback message based on the analysis results and the user's emotion recognition results. The analysis results from step 6 and the emotion recognition results from step 4 are used as input. The specific operation is to combine information about the cause of the crying with advice or suggestions based on the user's emotional state. The output is a feedback message.

[0332] Step 8: Notification

[0333] The server sends the generated feedback message to the user device. The feedback message generated in step 7 is used as input. The device receives this message and displays it to the user as a push notification or an alert. Specific actions include playing a notification sound or displaying a message on the screen. The output is that the user can confirm the feedback message.

[0334] In this way, the system can analyze the cries of babies and animals and provide appropriate feedback that takes into account the user's emotional state.

[0335] (Application example 2)

[0336] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0337] Conventional systems for analyzing the cries of babies and animals are limited to analyzing only the emotions and needs of the cry and providing feedback based on that analysis, and lack consideration of the user's emotions and state. Furthermore, feedback after analyzing the cry is not linked to the user's emotional state, making it difficult for the user to respond appropriately and promptly. The present invention aims to solve these problems by providing more appropriate feedback by taking the user's emotional state into account, enabling a prompt response.

[0338] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0339] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for recognizing a user's voice input and analyzing the user's emotions, means for transmitting advice or suggestions based on the prediction results and the user's emotion analysis results to a user terminal, and means for notifying the user of the received prediction results or advice. This makes it possible to provide appropriate feedback that takes into account not only the analysis of the crying of a baby or animal but also the user's emotional state.

[0340] The "means for receiving audio data" refers to a device or module that has the function of capturing audio data, including the cries of babies and animals, in real time and incorporating it into the system.

[0341] The "means for preprocessing received voice data" refers to a device or module that has the function of processing collected voice data, such as noise removal and standardizing the sampling rate, and converting it into a format suitable for analysis.

[0342] "Means of using machine learning models to analyze preprocessed voice data and predict emotions and needs" refers to algorithms and systems that use deep learning techniques to analyze and predict emotions and needs from preprocessed voice data.

[0343] "Means for recognizing user voice input and analyzing the user's emotions" refers to a device or module that has the function of analyzing the user's voice in real time and determining the user's emotional state, such as "fatigue," "stress," or "relief."

[0344] "Means for sending advice and suggestions based on prediction results and user emotion analysis results to a user terminal" refers to a device or system that has the function of generating advice and suggestions that take into account the user's emotional state based on the analysis results and sending them to the user's smartphone or dedicated device.

[0345] The "means for notifying the user of the received prediction results and advice" refers to a device or application that has a notification function for appropriately notifying the user of the analysis results and advice sent to the terminal.

[0346] Data collection and preprocessing

[0347] The server collects a large amount of crying data labeled by experts and stores it in a cloud database. The data set includes audio files of crying babies and animals, along with associated labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed, including noise removal and standardization of the sampling rate, to obtain data suitable for analysis.

[0348] Training a machine learning model

[0349] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). The model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs based on them. The resulting model is stored on a cloud server and used for real-time analysis.

[0350] User Emotion Recognition

[0351] The device is equipped with an emotion engine that recognizes the user's emotions based on their voice input. This engine analyzes the user's speech data in real time and recognizes emotional states such as "fatigue," "stress," and "relief." This is done automatically when the user speaks to the device.

[0352] Audio recording and real-time analysis

[0353] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the crying of a baby or animal in real time. When a cry is detected, the device records the audio data and sends it to a server in the cloud. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotion and needs indicated by the cry are predicted.

[0354] Sending and notifying feedback

[0355] The server sends the results of analyzing the baby or animal's cry, along with the user's emotion recognition results, to the user's device. Specifically, information about the cause of the cry and advice or suggestions based on the user's emotional state are sent as a message. The device then notifies the user of these results. The notifications are displayed as push notifications or alerts, helping the user respond immediately.

[0356] Specific examples

[0357] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device recognizes the user's emotions, such as "fatigue" or "stress," in real time. The server analyzes the received voice data and predicts that the user is "hungry." If the server determines that the user is feeling tired, it creates a message containing advice such as, "Your baby may be hungry. If possible, ask a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[0358] Prompt Sentence Examples

[0359] We will imagine a situation where a baby or animal starts crying in the middle of the night, record an audio file of the crying, and analyze it. We will also record the user's voice and analyze their emotions. Based on these two results, we will design a system that notifies the user of appropriate advice.

[0360] This embodiment allows for quick understanding of why a baby or animal is crying and allows for a response that takes into account the parent's or owner's own emotional state, which is expected to improve the health and well-being of not only the baby or animal, but also the parent or owner themselves.

[0361] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0362] Step 1:

[0363] The device records the cries of babies and animals in real time. The input is environmental audio data, and the output is a recorded audio data file. Specifically, the device's microphone captures the audio and saves it as a digital audio file.

[0364] Step 2:

[0365] The terminal sends the recorded audio data file to the server. The input is the audio data file recorded in step 1, and the output is the audio data sent to the server. Specifically, the audio data file is uploaded to the server via an Internet connection.

[0366] Step 3:

[0367] The server preprocesses the received audio data. The input is the audio data sent to the server, and the output is the preprocessed audio data. Specifically, the server removes noise, unifies the sampling rate, and converts the data into a format suitable for analysis.

[0368] Step 4:

[0369] The server analyzes the preprocessed audio data using a machine learning model. The input is the preprocessed audio data, and the output is the predicted emotion and needs of the crying voice. Specifically, it uses a convolutional neural network (CNN) and a recurrent neural network (RNN) to analyze the audio features and classify the emotion and needs.

[0370] Step 5:

[0371] The device recognizes the user's voice input in real time and analyzes the user's emotions. The input is the voice data spoken by the user, and the output is the analysis result of the user's emotional state. Specifically, the emotion engine installed in the device analyzes the voice and determines the user's emotional state, such as "fatigue," "stress," or "relief."

[0372] Step 6:

[0373] The server integrates the prediction results and the user's emotion analysis results based on the analysis results to generate advice and suggestions. The input is the prediction results of the crying emotion and needs and the analysis results of the user's emotional state, and the output is a message containing advice and suggestions. Specifically, the server uses a generative AI model to create appropriate advice and suggestions.

[0374] Step 7:

[0375] The server sends the generated advice and suggestion messages to the device. The input is the message containing the advice or suggestion, and the output is the message sent to the device. Specifically, the message is delivered to the device via the cloud server.

[0376] Step 8:

[0377] The device notifies the user of the received advice or suggestion message. The input is the advice or suggestion message sent to the device, and the output is the notification displayed to the user. Specifically, the device displays the message to the user using a push notification or a pop-up alert.

[0378] Prompt Sentence Examples

[0379] We will imagine a situation where a baby or animal starts crying in the middle of the night, record an audio file of the crying, and analyze it. We will also record the user's voice and analyze their emotions. Based on these two results, we will design a system that notifies the user of appropriate advice.

[0380] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0381] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0382] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0383] [Second embodiment]

[0384] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0385] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0386] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0387] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0388] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0389] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0390] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0391] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0392] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0393] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0394] In the smart glasses 214, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0395] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0396] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. Hereinafter, an embodiment of the system of the present invention will be described in detail.

[0397] Data collection and preprocessing

[0398] The server collects a large amount of crying data sets labeled by experts and stores them in a cloud database. The data sets include audio files of crying babies and animals and corresponding labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed by removing noise and standardizing the sampling rate, resulting in data suitable for analysis.

[0399] Training a machine learning model

[0400] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). This model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and can predict emotions and needs based on that information. The resulting model is stored on a cloud server and used for real-time analysis.

[0401] Audio recording and real-time analysis

[0402] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotions and needs indicated by the crying are predicted.

[0403] Sending and notifying feedback

[0404] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent as a push notification to the user's smartphone. The device receives this and immediately notifies the user.

[0405] Specific examples

[0406] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's cry and sends the data to the server. The server analyzes this audio data and predicts that the baby is hungry. The prediction result is returned from the server to the device, which then notifies the user with a message saying, "Your baby may be hungry." The user can check this notification and immediately prepare milk to feed the baby. In this way, a quick and accurate response is possible.

[0407] This invention makes it possible to quickly determine why a baby or animal is crying, making it easier for parents and owners to take appropriate action, which is expected to reduce stress for babies and animals and improve their health and happiness.

[0408] The processing flow will be explained below.

[0409] Step 1:

[0410] The server collects expert-labeled datasets of baby and animal cries and stores them in a cloud database, which includes audio files and corresponding labels such as "hungry," "sleepy," and "anxious."

[0411] Step 2:

[0412] The server preprocesses the collected audio data, including noise removal, unifying the sampling rate, normalizing, etc. The preprocessed data is then converted into a form suitable for analysis.

[0413] Step 3:

[0414] The server uses the preprocessed data to train machine learning models, specifically convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to learn audio features (e.g., pitch, rhythm, decibels).

[0415] Step 4:

[0416] The server stores the trained machine learning model in the cloud, which is then used for real-time analytics.

[0417] Step 5:

[0418] The device has the ability to record the crying sounds of babies and animals in real time. When a crying sound is detected, the device will start recording the audio data.

[0419] Step 6:

[0420] The device transmits the recorded audio data to a cloud server via the Internet.

[0421] Step 7:

[0422] The server then pre-processes the received audio data again, which includes noise reduction and sampling rate unification.

[0423] Step 8:

[0424] The server then uses the pre-processed audio data to analyze it using a trained machine learning model, which predicts the crying as an emotion or need, such as "hunger," "sleepy," or "anxiety."

[0425] Step 9:

[0426] The server sends the analysis results to the user's device, which include a specific alert message.

[0427] Step 10:

[0428] The device will then notify the user of the received prediction results, which will be displayed as a push notification or an alert.

[0429] Step 11:

[0430] The user can check the notification and respond to the baby or animal's needs based on the notification. For example, if the baby or animal is predicted to be hungry, the user can prepare milk and give it to the baby. This process allows parents and owners to quickly understand the cause of the crying and take appropriate action.

[0431] Example 1

[0432] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0433] Conventional voice recognition systems lack the means to analyze the cries of babies and animals and provide fast and accurate feedback on their emotions and needs. Furthermore, due to the lack of proper noise reduction and standardization of sampling rates for each individual cry, analysis results tend to be inaccurate. Furthermore, the lack of a function to notify the user of prediction results in real time makes it difficult to respond quickly. There is a need to address these issues with conventional technology.

[0434] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0435] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data based on noise reduction and unified sampling rate, means for analyzing the preprocessed voice data and using a machine learning model consisting of a convolutional neural network and a recurrent neural network to predict emotions and needs, means for transmitting the prediction results to a user terminal, and means for sending a push notification of the received prediction results to the user. This makes it possible to quickly and accurately analyze the cries of babies and animals and predict emotions and needs based on the analysis, and notify the user of the results in real time.

[0436] "Voice data" is data that represents the voice of a human or animal in digital form.

[0437] A "means for receiving" is a method or apparatus for obtaining audio data from another device or system.

[0438] "Noise reduction" is the process of removing unwanted background sounds and noise from audio data.

[0439] "Sampling rate unification" is the process of making the sampling frequencies of different audio data uniform.

[0440] "Preprocessing means" refers to a method or device that performs necessary processing or conversion on the audio data before analysis.

[0441] "Emotions and needs" refers to the feelings and demands that babies and animals express through their cries.

[0442] A "convolutional neural network" is a type of deep learning model that excels at analyzing data such as images and audio.

[0443] A "recurrent neural network" is a type of deep learning model that excels at analyzing continuous data over time.

[0444] A "machine learning model" is an artificial intelligence algorithm that uses data to learn and perform specific tasks.

[0445] A "means of use" is a method or device that applies a machine learning model to perform analysis or prediction.

[0446] A "user terminal" is a device such as a smartphone, PC, or tablet that a user interacts with directly.

[0447] A "transmitting means" is a method or apparatus for sending data to another device or system.

[0448] "Push notification" is a communication technology that sends specific information to a user's device in real time and displays it instantly.

[0449] A "cloud server" is a server that includes computing resources and storage provided over the Internet, and is a place where data is stored and computational processing is performed.

[0450] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. An embodiment of the system of the present invention will be described in detail below.

[0451] Data collection and preprocessing

[0452] The server collects a large amount of expert-labeled crying data sets from the Internet and stores them in a cloud database. The collected data includes audio files of babies and animals crying and their corresponding labels, such as "hungry," "sleepy," and "anxious." The server then preprocesses the collected audio data using the Python LibROSA library, removing noise and standardizing the sampling rate.

[0453] Training a machine learning model

[0454] The server splits the preprocessed data into a training dataset and a test dataset, then uses TensorFlow to build and train a convolutional neural network (CNN) or recurrent neural network (RNN). The trained model is then stored on the cloud server.

[0455] Audio recording and real-time analysis

[0456] The device (e.g., the user's smartphone or a dedicated device) has the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again using the LibROSA library and analyzes it using a trained machine learning model. This allows the emotions and needs indicated by the cry to be predicted.

[0457] Sending and notifying feedback

[0458] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent to the user's smartphone as a push notification. The device receives this and immediately notifies the user.

[0459] Specific examples

[0460] For example, if your baby starts crying in the middle of the night,

[0461] 1. The device records the baby's cry and sends the data to the server, which then activates the smartphone's recording app and captures the audio.

[0462] 2. The server re-preprocesses the received audio data and analyzes it using the trained machine learning model. It uses LibROSA on the server side to remove noise and input the data into the model.

[0463] 3. The server obtains the prediction result, "I'm hungry," and sends it to the device. The server uses an HTTP POST request to send the prediction result in JSON format.

[0464] 4. The device pushes a message to the user saying, "Your baby might be hungry." The smartphone displays the notification so that the user can check it.

[0465] The system can quickly determine why a baby or animal is crying, allowing parents and owners to take appropriate action, reducing stress and improving their health and happiness.

[0466] Prompt Sentence Examples

[0467] "Please explain how your smartphone automatically analyzes your baby's emotions and needs and notifies you when your baby is crying."

[0468] Based on this example, generative AI models can be used to provide context-sensitive feedback.

[0469] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0470] Step 1: Data collection

[0471] The server collects expertly labeled baby and animal cry datasets from the Internet and stores them in a cloud database. It performs the following operations:

[0472] Input: A crying audio file and its corresponding label (e.g., "hungry" or "sleepy").

[0473] Data Processing: During the data collection process, the data is stored in a database in a suitable format.

[0474] Output: A crying dataset stored in a cloud database.

[0475] Step 2: Data Preprocessing

[0476] The server performs noise reduction and sampling rate unification on the collected audio data. It uses the Python LibROSA library and performs the following operations.

[0477] Input: Audio data stored in a cloud database.

[0478] Data processing: LibROSA was used to remove background noise from the audio files and standardize the sampling rate to 44.1kHz.

[0479] Output: Preprocessed audio data.

[0480] Step 3: Train the model

[0481] The server splits the preprocessed data into training and test datasets and trains the model using a deep learning framework. It performs the following operations:

[0482] Input: Preprocessed audio data.

[0483] Data processing: 80% of the total data will be used as a training dataset and 20% as a test dataset.

[0484] Output: A dataset split into training and testing.

[0485] The server builds and trains models using convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

[0486] Input: Training dataset.

[0487] Data calculation: We use TensorFlow to train the model with 10 epochs.

[0488] Output: A trained machine learning model.

[0489] The server stores the trained model on a cloud server.

[0490] Input: A trained model.

[0491] Data calculation: Save the model using TensorFlow's model.save method.

[0492] Output: A trained model stored on a cloud server.

[0493] Step 4: Record your audio

[0494] The device (e.g., a smartphone) records the cries of babies and animals in real time. It performs the following operations:

[0495] Input: Babies and animals crying.

[0496] Data processing: Capture audio using a recording app.

[0497] Output: Recorded audio data.

[0498] Step 5: Sending audio data

[0499] The device sends the recorded audio data to a cloud server via the Internet. The following operations are performed:

[0500] Input: Recorded audio data.

[0501] Data calculation: Upload audio data to the server via HTTP POST request.

[0502] Output: The audio data sent to the server.

[0503] Step 6: Real-time analysis

[0504] The server then preprocesses the received audio data again and analyzes it using the trained machine learning model.

[0505] Input: Transmitted audio data.

[0506] Data processing: LibROSA is used to remove noise and unify the sampling rate again.

[0507] Output: Audio data in a format suitable for analysis.

[0508] The server uses trained machine learning models to analyze voice data and predict emotions and needs.

[0509] Input: Preprocessed audio data.

[0510] Data calculation: Using machine learning models, predict emotions and needs such as "hungry" or "sleepy."

[0511] Output: Predicted emotion and needs data.

[0512] Step 7: Submit your feedback

[0513] The server sends the analysis results to the user's terminal and performs the following operations.

[0514] Input: Parsed emotion and needs data.

[0515] Data calculation: Prediction results are sent to the terminal in JSON format via an HTTP POST request.

[0516] Output: Analysis results sent to the user's terminal.

[0517] Step 8: User Notification

[0518] The device will then push the received analysis results to the user. The device will perform the following operations:

[0519] Input: Analysis results sent from the server.

[0520] Data calculation: A message is displayed using the smartphone's notification function.

[0521] Output: The message that was sent to the user.

[0522] This step makes it possible to analyze the cries of babies and animals, predict their emotions and needs based on the analysis, and notify the user of the results in real time.

[0523] (Application example 1)

[0524] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0525] Conventional crying analysis systems simply analyze voice data on a local device, resulting in problems with the accuracy and real-time nature of the analysis results. Furthermore, when used in pet shops and veterinary clinics, a quick and appropriate response is required, particularly on-site, so it is necessary to provide more accurate analysis results. This invention aims to improve the accuracy of voice data analysis by utilizing a cloud server and provide users with quick and appropriate feedback.

[0526] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0527] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for transmitting the prediction results to a user terminal, means for notifying the user of the received prediction results, means for transmitting the received voice data to a cloud server and acquiring analysis results at the cloud server, and means for generating notification content based on the analysis results and instructing the user on an appropriate response. This allows highly accurate analysis results to be obtained quickly, enabling the user to immediately take appropriate measures.

[0528] "Audio data" is information that has been digitally recorded as the cries and cries of babies and animals.

[0529] "Preprocessing" is the process of processing audio data, such as removing noise and standardizing the sampling rate, to prepare it in a form suitable for analysis.

[0530] A "machine learning model" is a computational model that includes algorithms for learning from data and making predictions and classifications for new data.

[0531] "Deep learning" is a type of machine learning that uses multi-layered neural networks to analyze data and learn complex patterns.

[0532] A "convolutional neural network (CNN)" is a machine learning model primarily used to analyze image and audio data, and is a network with the ability to capture the spatial structure of data.

[0533] A "recurrent neural network (RNN)" is a machine learning model suitable for analyzing time series data, and is a network that proceeds with learning by taking into account the temporal dependency of the data.

[0534] A "cloud server" is a remote server that stores data and performs calculations over the Internet, and is a computer that provides services in response to requests from a large number of users.

[0535] A "user terminal" is a device used to record voice data, receive analysis results, and receive notifications, and specific examples include smartphones and dedicated devices.

[0536] A "notification" is a message that conveys analysis results and instructions based on those results to the user, and is mainly displayed in the form of a push notification or an alert.

[0537] "Analysis results" are predicted information about emotions and needs obtained after analyzing voice data using a machine learning model.

[0538] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. An embodiment of the system of the present invention will be described in detail below.

[0539] Data collection and preprocessing

[0540] The server collects a large amount of crying data sets labeled by experts and stores them in a cloud database. The data sets include audio files of crying babies and animals and corresponding labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed by removing noise and standardizing the sampling rate, resulting in data suitable for analysis.

[0541] Training a machine learning model

[0542] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). This model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and can predict emotions and needs based on that information. The resulting model is stored on a cloud server and used for real-time analysis.

[0543] Audio recording and real-time analysis

[0544] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotions and needs indicated by the crying are predicted.

[0545] Sending and notifying feedback

[0546] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent as a push notification to the user's smartphone. The device receives this and immediately notifies the user.

[0547] Examples from pet shops and veterinary clinics

[0548] The Pet Care Assistant application analyzes the cries (cries) of pets brought in by customers at pet shops and veterinary clinics. For example, if a dog barks at a pet shop, the device records the sound and sends it to a server on the cloud. The server analyzes the audio data and obtains a prediction result, such as "hungry." As a result, the store clerk receives a message saying, "This dog may be hungry," and can recommend appropriate dog food to the owner.

[0549] Hardware and software used

[0550] Hardware used

[0551] Smartphone: A device that uses the recording function to capture voice data and send it to a server

[0552] Cloud server: Runs the deep learning model and analyzes the voice data

[0553] Software used

[0554] Python: Writing program logic

[0555] pydub: A library for preprocessing audio data

[0556] requests: A library for managing HTTP requests / responses.

[0557] Examples and Prompts

[0558] For example, if a dog in a pet shop is barking for a specific reason, the Pet Care Assistant can analyze the sound and notify the store clerk that "This dog may be hungry," allowing the store to provide prompt and appropriate service to customers.

[0559] Prompt Sentence Examples

[0560] Audio data: Audio data in BASE64 format

[0561] Device ID: Device identifier

[0562] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0563] Step 1:

[0564] The device records surrounding audio in real time.

[0565] Input: Audio signal from the external environment

[0566] Output: Recorded audio data (e.g., WAV file)

[0567] What it does: Uses the microphone on your smartphone or dedicated device to capture the cries of babies and animals.

[0568] Step 2:

[0569] The terminal preprocesses the recorded audio data.

[0570] Input: Recorded audio data

[0571] Output: Preprocessed audio data

[0572] Specific operations: Noise removal and unification of sampling rates are performed on the audio data to prepare it in a form suitable for analysis.

[0573] Step 3:

[0574] The terminal transmits the pre-processed voice data to a server on the cloud via the Internet.

[0575] Input: Preprocessed audio data

[0576] Output: Data transmission result to cloud server (success / failure)

[0577] Specific operation: Generate an HTTP request and send the preprocessed audio data to the cloud server.

[0578] Step 4:

[0579] The server again pre-processes the received audio data.

[0580] Input: Preprocessed audio data

[0581] Output: Re-preprocessed audio data

[0582] Specific operation: Additional noise reduction and format conversion is performed on the server side as needed.

[0583] Step 5:

[0584] The server analyzes the audio data using a trained machine learning model.

[0585] Input: Re-preprocessed audio data

[0586] Output: Prediction of emotions and needs corresponding to crying

[0587] Specific operation: Using convolutional neural networks (CNN) and recurrent neural networks (RNN), voice features are extracted and emotions and needs are predicted.

[0588] Step 6:

[0589] The server transmits the analysis results to the user's terminal.

[0590] Input: Prediction result (e.g., "I'm hungry," "I'm sleepy," etc.)

[0591] Output: Data transmission result to user device (success / failure)

[0592] Specific operation: Generate an HTTP response and send a message containing the prediction result to the user terminal.

[0593] Step 7:

[0594] The terminal notifies the user of the received analysis results.

[0595] Input: A message containing the analysis results

[0596] Output: Notification to the user (e.g. push notification)

[0597] Specific operation: The analysis results are displayed as a push notification using the notification function of the user's smartphone or dedicated device.

[0598] Step 8:

[0599] The user checks the notification and takes appropriate action.

[0600] Input: Notification of analysis results

[0601] Output: User's response (e.g., feeding a baby, feeding a dog, etc.)

[0602] Specific Action: Based on the notification content, the user performs an action according to the request of the baby or pet.

[0603] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0604] The present invention relates to a system that analyzes the cries of babies and animals, recognizes not only their emotions and needs but also the emotions of a user, and provides feedback. An embodiment of the system of the present invention will be described in detail below.

[0605] Data collection and preprocessing

[0606] The server collects a large amount of crying data labeled by experts and stores it in a cloud database. The data set includes audio files of crying babies and animals, along with labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed, including noise removal and standardization of the sampling rate, to obtain data suitable for analysis.

[0607] Training a machine learning model

[0608] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). The model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs based on them. The resulting model is stored on a cloud server and used for real-time analysis.

[0609] User Emotion Recognition

[0610] The device is equipped with an emotion engine that recognizes the user's emotions based on their voice input. This engine analyzes the user's speech data in real time and recognizes their emotional state, such as "fatigue," "stress," or "relief." This is done automatically when the user speaks to the device.

[0611] Audio recording and real-time analysis

[0612] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the crying of a baby or animal in real time. When a cry is detected, the device records the audio data and sends it to a server in the cloud. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotion and needs indicated by the cry are predicted.

[0613] Sending and notifying feedback

[0614] The server then sends the results of the analysis of the baby or animal's cry, along with the user's emotion recognition results, to the user's device. Specifically, the server sends a message containing information about the cause of the cry and advice or suggestions based on the user's emotional state.

[0615] The device will notify the user of these results, which will be displayed as push notifications or alerts, allowing the user to take immediate action.

[0616] Specific examples

[0617] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device recognizes the user's emotions, such as "fatigue" and "stress," in real time. The server analyzes the received voice data and predicts that the user is "hungry." If the server determines that the user is feeling tired, it creates a message containing advice such as "Your baby may be hungry. If possible, ask a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[0618] This embodiment allows for quick understanding of why a baby or animal is crying and allows for a response that takes into account the parent's or owner's own emotional state, which is expected to improve the health and well-being of not only the baby or animal, but also the parent or owner themselves.

[0619] The processing flow will be explained below.

[0620] Step 1:

[0621] The server collects expert-labeled datasets of baby and animal cries and stores them in a cloud database, including audio files and labels such as "hungry," "sleepy," and "anxious."

[0622] Step 2:

[0623] The server preprocesses the collected audio data, which includes noise removal and sampling rate unification, to prepare the data for analysis.

[0624] Step 3:

[0625] The server uses the preprocessed data to train machine learning models, which are based on convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

[0626] Step 4:

[0627] The server stores the trained machine learning model in the cloud, which is then used for subsequent real-time analysis.

[0628] Step 5:

[0629] The device has the ability to record the crying sounds of babies and animals in real time. When a crying sound is detected, the device will start recording the audio data.

[0630] Step 6:

[0631] The device transmits the recorded voice data to a server via the Internet, and simultaneously detects the user's voice input and analyzes it using an emotion engine.

[0632] Step 7:

[0633] The server preprocesses the received audio data, which includes noise reduction and sampling rate unification.

[0634] Step 8:

[0635] The server uses the pre-processed audio data to analyze it with a trained machine learning model, which predicts the cause of the crying (e.g., hunger, sleepiness, etc.).

[0636] Step 9:

[0637] The server integrates the results of the user's emotion recognition with the results of the crying analysis. For example, if the user feels "tired," that information is also included in the prediction results.

[0638] Step 10:

[0639] The server sends the integrated results to the user's terminal, and the message includes the cause of the crying as well as advice based on the user's condition.

[0640] Step 11:

[0641] The device notifies the user of messages received from the server, which are displayed as push notifications or alerts.

[0642] Step 12:

[0643] The user can check the notification and respond to the baby or animal's needs based on the notification. For example, if the system predicts that the baby is hungry and the user feels tired, it will suggest asking other family members for help. The user can also take action such as feeding the baby if necessary.

[0644] This series of steps enables the system to quickly identify the cause of a baby or animal's cry and provide appropriate advice that takes into account the user's emotional state.

[0645] Example 2

[0646] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0647] Conventional emotion recognition systems are limited in their ability to analyze the cries of babies and animals, and are unable to simultaneously recognize the user's emotional state and provide appropriate feedback. This makes it difficult for parents and owners to respond efficiently and quickly, often resulting in stress and fatigue. Furthermore, analysis results can sometimes be inaccurate, preventing effective countermeasures from being implemented.

[0648] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0649] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for analyzing the user's voice data in real time and recognizing the emotional state, means for recording the crying of a baby or animal in real time and transmitting the voice data to the server, means for transmitting the prediction result to a user terminal, and means for notifying the user of the received prediction result. This makes it possible to quickly and accurately understand the meaning of the crying of a baby or animal and provide appropriate feedback according to the user's emotional state.

[0650] "Means for receiving audio data" refers to the function that allows a user terminal or dedicated device to record the cries of babies or animals and send them to a server.

[0651] "Means for preprocessing received voice data" refers to a function for performing preprocessing such as noise removal and unifying the sampling rate on collected voice data.

[0652] "Means of analyzing pre-processed voice data and using machine learning models to predict emotions and needs" refers to a function that analyzes voice data using deep learning convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to predict emotions and needs.

[0653] "Means for analyzing the user's voice data in real time and recognizing their emotional state" refers to a function that analyzes the voice data spoken by the user in real time and recognizes their emotional state, such as "fatigue" or "stress."

[0654] "Means for recording the cries of babies and animals in real time and transmitting the audio data to a server" refers to a function that enables a user terminal or dedicated device to record the cries of babies and animals in real time and transmit the audio data to a cloud server.

[0655] "Means for transmitting prediction results to a user terminal" refers to a function that enables the server to transmit analysis results and feedback messages to a user terminal.

[0656] "Means for notifying the user of the received prediction results" refers to a function that allows the user's device to notify the user of the prediction results received from the server as a push notification or an alert.

[0657] MODE FOR CARRYING OUT THE INVENTION

[0658] The present invention relates to a system that analyzes the cries of babies and animals, recognizes not only their emotions and needs but also the emotions of the user, and provides feedback. This system performs a series of processes from collecting audio data to analyzing it and providing feedback. An embodiment of the system of the present invention will be described in detail below.

[0659] Data collection and preprocessing

[0660] The server collects a large amount of expert-labeled crying data sets from online repositories and stores them in a cloud database. The collected data includes audio files of crying babies and animals, along with labels such as "hungry," "sleepy," and "anxious." The data is then converted to a uniform sampling rate (e.g., 44.1 kHz) and preprocessed with noise reduction (using a Gaussian filter, etc.).

[0661] Training a machine learning model

[0662] The server then loads the preprocessed audio data into a deep learning environment (e.g., TensorFlow or PyTorch) and trains it using a convolutional neural network (CNN) or recurrent neural network (RNN). This training creates a model that learns the feature vectors of the audio data (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs with high accuracy. After training is complete, the model is stored on a cloud server and made available for real-time analysis.

[0663] User Emotion Recognition

[0664] The device is equipped with an emotion engine that analyzes emotions in real time based on the user's voice input. This engine uses, for example, IBM Watson Tone Analyzer to analyze the voice data of the user's speech (e.g., "I'm tired," "I'm irritated," etc.) and recognizes the user's emotional state. This real-time recognition is performed in a natural situation where the user is speaking to the device.

[0665] Audio recording and real-time analysis

[0666] The device (e.g., a smartphone or dedicated device) has the ability to record the cries of babies and animals in real time. When a cry is detected, the device automatically starts recording and sends the audio data to a cloud-based server. The server then preprocesses the received audio data and analyzes it using a trained machine learning model. This allows for highly accurate prediction of the emotions and needs indicated by the cry.

[0667] Sending and notifying feedback

[0668] The server combines the results of the crying analysis and the user's emotion recognition and sends them to the user's device. Specific feedback includes information about the cause of the crying and advice or suggestions based on the user's emotional state. For example, "Your baby may be hungry. If possible, ask a family member for help." The device then notifies the user of this feedback via a push notification or alert, encouraging them to take prompt action.

[0669] Specific examples

[0670] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device analyzes the user's voice and recognizes emotions such as "fatigue" and "stress" in real time. The server analyzes the received crying data and predicts that the baby is "hungry." If the server determines that the user is tired, it creates a message containing advice such as "Your baby may be hungry. If possible, try asking a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[0671] Prompt Sentence Examples

[0672] 1. "Analyze an audio file of a crying baby and predict its emotions and needs. The user's emotional state is 'Tired'."

[0673] 2. "Analyze a dog's cry and tell me its cause. The user's emotion is 'stress'."

[0674] This system can quickly and accurately identify the emotions and needs expressed by the cries of babies and animals, and provide appropriate feedback based on the user's emotional state, which is expected to improve the health and happiness of not only babies and animals, but also their parents and owners.

[0675] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0676] Step 1: Collect data

[0677] The server collects expert-labeled crying datasets from online repositories. The input includes crying audio data and corresponding emotion and need labels. The output is stored in a cloud database.

[0678] Step 2: Preprocessing the data

[0679] The server performs preprocessing on the collected audio data. The raw audio data collected in step 1 is used as input. Specific preprocessing operations include noise removal (applying a Gaussian filter) and unifying the sampling rate (e.g., changing to 44.1 kHz). The preprocessed audio data is obtained as output.

[0680] Step 3: Train the machine learning model

[0681] The server uses the preprocessed audio data to train a machine learning model in a deep learning environment (such as TensorFlow or PyTorch). The preprocessed audio data and its associated label information are used as input. Specifically, a convolutional neural network (CNN) or recurrent neural network (RNN) is used to learn feature vectors (e.g., pitch, rhythm, decibels). The output is a trained model.

[0682] Step 4: Recognizing user emotions

[0683] The device collects the user's voice data in real time and analyzes their emotions. The user's spoken voice data is used as input. Specifically, it uses an emotion engine (for example, IBM Watson Tone Analyzer) to recognize emotional states such as "fatigue" and "stress" in real time. The output is the recognition result of the user's emotional state.

[0684] Step 5: Record your audio

[0685] The device records the cries of babies and animals in real time. The cries of babies and animals are used as input. Specifically, the device's microphone is used to record the cries and save them as audio files. The recorded audio data is obtained as output.

[0686] Step 6: Real-time analysis

[0687] The device sends the recorded crying data to a server on the cloud. The audio data recorded in step 5 is used as input. The server receives this audio data, preprocesses it again, and then analyzes it using a trained machine learning model. Specific operations include preprocessing (noise removal, unifying the sampling rate) and feature extraction (pitch, rhythm, decibel analysis). The analysis results are obtained as output.

[0688] Step 7: Feedback Generation

[0689] The server generates a feedback message based on the analysis results and the user's emotion recognition results. The analysis results from step 6 and the emotion recognition results from step 4 are used as input. The specific operation is to combine information about the cause of the crying with advice or suggestions based on the user's emotional state. The output is a feedback message.

[0690] Step 8: Notification

[0691] The server sends the generated feedback message to the user device. The feedback message generated in step 7 is used as input. The device receives this message and displays it to the user as a push notification or an alert. Specific actions include playing a notification sound or displaying a message on the screen. The output is that the user can confirm the feedback message.

[0692] In this way, the system can analyze the cries of babies and animals and provide appropriate feedback that takes into account the user's emotional state.

[0693] (Application example 2)

[0694] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0695] Conventional systems for analyzing the cries of babies and animals are limited to analyzing only the emotions and needs of the cry and providing feedback based on that analysis, and lack consideration of the user's emotions and state. Furthermore, feedback after analyzing the cry is not linked to the user's emotional state, making it difficult for the user to respond appropriately and promptly. The present invention aims to solve these problems by providing more appropriate feedback by taking the user's emotional state into account, enabling a prompt response.

[0696] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0697] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for recognizing a user's voice input and analyzing the user's emotions, means for transmitting advice or suggestions based on the prediction results and the user's emotion analysis results to a user terminal, and means for notifying the user of the received prediction results or advice. This makes it possible to provide appropriate feedback that takes into account not only the analysis of the crying of a baby or animal but also the user's emotional state.

[0698] The "means for receiving audio data" refers to a device or module that has the function of capturing audio data, including the cries of babies and animals, in real time and incorporating it into the system.

[0699] The "means for preprocessing received voice data" refers to a device or module that has the function of processing collected voice data, such as noise removal and standardizing the sampling rate, and converting it into a format suitable for analysis.

[0700] "Means of using machine learning models to analyze preprocessed voice data and predict emotions and needs" refers to algorithms and systems that use deep learning techniques to analyze and predict emotions and needs from preprocessed voice data.

[0701] "Means for recognizing user voice input and analyzing the user's emotions" refers to a device or module that has the function of analyzing the user's voice in real time and determining the user's emotional state, such as "fatigue," "stress," or "relief."

[0702] "Means for sending advice and suggestions based on prediction results and user emotion analysis results to a user terminal" refers to a device or system that has the function of generating advice and suggestions that take into account the user's emotional state based on the analysis results and sending them to the user's smartphone or dedicated device.

[0703] The "means for notifying the user of the received prediction results and advice" refers to a device or application that has a notification function for appropriately notifying the user of the analysis results and advice sent to the terminal.

[0704] Data collection and preprocessing

[0705] The server collects a large amount of crying data labeled by experts and stores it in a cloud database. The data set includes audio files of crying babies and animals, along with associated labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed, including noise removal and standardization of the sampling rate, to obtain data suitable for analysis.

[0706] Training a machine learning model

[0707] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). The model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs based on them. The resulting model is stored on a cloud server and used for real-time analysis.

[0708] User Emotion Recognition

[0709] The device is equipped with an emotion engine that recognizes the user's emotions based on their voice input. This engine analyzes the user's speech data in real time and recognizes emotional states such as "fatigue," "stress," and "relief." This is done automatically when the user speaks to the device.

[0710] Audio recording and real-time analysis

[0711] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the crying of a baby or animal in real time. When a cry is detected, the device records the audio data and sends it to a server in the cloud. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotion and needs indicated by the cry are predicted.

[0712] Sending and notifying feedback

[0713] The server sends the results of analyzing the baby or animal's cry, along with the user's emotion recognition results, to the user's device. Specifically, information about the cause of the cry and advice or suggestions based on the user's emotional state are sent as a message. The device then notifies the user of these results. The notifications are displayed as push notifications or alerts, helping the user respond immediately.

[0714] Specific examples

[0715] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device recognizes the user's emotions, such as "fatigue" or "stress," in real time. The server analyzes the received voice data and predicts that the user is "hungry." If the server determines that the user is feeling tired, it creates a message containing advice such as, "Your baby may be hungry. If possible, ask a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[0716] Prompt Sentence Examples

[0717] We will imagine a situation where a baby or animal starts crying in the middle of the night, record an audio file of the crying, and analyze it. We will also record the user's voice and analyze their emotions. Based on these two results, we will design a system that notifies the user of appropriate advice.

[0718] This embodiment allows for quick understanding of why a baby or animal is crying and allows for a response that takes into account the parent's or owner's own emotional state, which is expected to improve the health and well-being of not only the baby or animal, but also the parent or owner themselves.

[0719] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0720] Step 1:

[0721] The device records the cries of babies and animals in real time. The input is environmental audio data, and the output is a recorded audio data file. Specifically, the device's microphone captures the audio and saves it as a digital audio file.

[0722] Step 2:

[0723] The terminal sends the recorded audio data file to the server. The input is the audio data file recorded in step 1, and the output is the audio data sent to the server. Specifically, the audio data file is uploaded to the server via an Internet connection.

[0724] Step 3:

[0725] The server preprocesses the received audio data. The input is the audio data sent to the server, and the output is the preprocessed audio data. Specifically, the server removes noise, unifies the sampling rate, and converts the data into a format suitable for analysis.

[0726] Step 4:

[0727] The server analyzes the preprocessed audio data using a machine learning model. The input is the preprocessed audio data, and the output is the predicted emotion and needs of the crying voice. Specifically, it uses a convolutional neural network (CNN) and a recurrent neural network (RNN) to analyze the audio features and classify the emotion and needs.

[0728] Step 5:

[0729] The device recognizes the user's voice input in real time and analyzes the user's emotions. The input is the voice data spoken by the user, and the output is the analysis result of the user's emotional state. Specifically, the emotion engine installed in the device analyzes the voice and determines the user's emotional state, such as "fatigue," "stress," or "relief."

[0730] Step 6:

[0731] The server integrates the prediction results and the user's emotion analysis results based on the analysis results to generate advice and suggestions. The input is the prediction results of the crying emotion and needs and the analysis results of the user's emotional state, and the output is a message containing advice and suggestions. Specifically, the server uses a generative AI model to create appropriate advice and suggestions.

[0732] Step 7:

[0733] The server sends the generated advice and suggestion messages to the device. The input is the message containing the advice or suggestion, and the output is the message sent to the device. Specifically, the message is delivered to the device via the cloud server.

[0734] Step 8:

[0735] The device notifies the user of the received advice or suggestion message. The input is the advice or suggestion message sent to the device, and the output is the notification displayed to the user. Specifically, the device displays the message to the user using a push notification or a pop-up alert.

[0736] Prompt Sentence Examples

[0737] We will imagine a situation where a baby or animal starts crying in the middle of the night, record an audio file of the crying, and analyze it. We will also record the user's voice and analyze their emotions. Based on these two results, we will design a system that notifies the user of appropriate advice.

[0738] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0739] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0740] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0741] [Third embodiment]

[0742] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0743] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0744] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0745] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0746] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0747] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0748] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0749] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0750] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0751] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0752] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0753] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0754] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. Hereinafter, an embodiment of the system of the present invention will be described in detail.

[0755] Data collection and preprocessing

[0756] The server collects a large amount of crying data sets labeled by experts and stores them in a cloud database. The data sets include audio files of crying babies and animals and corresponding labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed by removing noise and standardizing the sampling rate, resulting in data suitable for analysis.

[0757] Training a machine learning model

[0758] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). This model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and can predict emotions and needs based on that information. The resulting model is stored on a cloud server and used for real-time analysis.

[0759] Audio recording and real-time analysis

[0760] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotions and needs indicated by the crying are predicted.

[0761] Sending and notifying feedback

[0762] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent as a push notification to the user's smartphone. The device receives this and immediately notifies the user.

[0763] Specific examples

[0764] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's cry and sends the data to the server. The server analyzes this audio data and predicts that the baby is hungry. The prediction result is returned from the server to the device, which then notifies the user with a message saying, "Your baby may be hungry." The user can check this notification and immediately prepare milk to feed the baby. In this way, a quick and accurate response is possible.

[0765] This invention makes it possible to quickly determine why a baby or animal is crying, making it easier for parents and owners to take appropriate action, which is expected to reduce stress for babies and animals and improve their health and happiness.

[0766] The processing flow will be explained below.

[0767] Step 1:

[0768] The server collects expert-labeled datasets of baby and animal cries and stores them in a cloud database, which includes audio files and corresponding labels such as "hungry," "sleepy," and "anxious."

[0769] Step 2:

[0770] The server preprocesses the collected audio data, including noise removal, unifying the sampling rate, normalizing, etc. The preprocessed data is then converted into a form suitable for analysis.

[0771] Step 3:

[0772] The server uses the preprocessed data to train machine learning models, specifically convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to learn audio features (e.g., pitch, rhythm, decibels).

[0773] Step 4:

[0774] The server stores the trained machine learning model in the cloud, which is then used for real-time analytics.

[0775] Step 5:

[0776] The device has the ability to record the crying sounds of babies and animals in real time. When a crying sound is detected, the device will start recording the audio data.

[0777] Step 6:

[0778] The device transmits the recorded audio data to a cloud server via the Internet.

[0779] Step 7:

[0780] The server then pre-processes the received audio data again, which includes noise reduction and sampling rate unification.

[0781] Step 8:

[0782] The server then uses the pre-processed audio data to analyze it using a trained machine learning model, which predicts the crying as an emotion or need, such as "hunger," "sleepy," or "anxiety."

[0783] Step 9:

[0784] The server sends the analysis results to the user's device, which include a specific alert message.

[0785] Step 10:

[0786] The device will then notify the user of the received prediction results, which will be displayed as a push notification or an alert.

[0787] Step 11:

[0788] The user can check the notification and respond to the baby or animal's needs based on the notification. For example, if the baby or animal is predicted to be hungry, the user can prepare milk and give it to the baby. This process allows parents and owners to quickly understand the cause of the crying and take appropriate action.

[0789] Example 1

[0790] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0791] Conventional voice recognition systems lack the means to analyze the cries of babies and animals and provide fast and accurate feedback on their emotions and needs. Furthermore, due to the lack of proper noise reduction and standardization of sampling rates for each individual cry, analysis results tend to be inaccurate. Furthermore, the lack of a function to notify the user of prediction results in real time makes it difficult to respond quickly. There is a need to address these issues with conventional technology.

[0792] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0793] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data based on noise reduction and unified sampling rate, means for analyzing the preprocessed voice data and using a machine learning model consisting of a convolutional neural network and a recurrent neural network to predict emotions and needs, means for transmitting the prediction results to a user terminal, and means for sending a push notification of the received prediction results to the user. This makes it possible to quickly and accurately analyze the cries of babies and animals and predict emotions and needs based on the analysis, and notify the user of the results in real time.

[0794] "Voice data" is data that represents the voice of a human or animal in digital form.

[0795] A "means for receiving" is a method or apparatus for obtaining audio data from another device or system.

[0796] "Noise reduction" is the process of removing unwanted background sounds and noise from audio data.

[0797] "Sampling rate unification" is the process of making the sampling frequencies of different audio data uniform.

[0798] "Preprocessing means" refers to a method or device that performs necessary processing or conversion on the audio data before analysis.

[0799] "Emotions and needs" refers to the feelings and demands that babies and animals express through their cries.

[0800] A "convolutional neural network" is a type of deep learning model that excels at analyzing data such as images and audio.

[0801] A "recurrent neural network" is a type of deep learning model that excels at analyzing continuous data over time.

[0802] A "machine learning model" is an artificial intelligence algorithm that uses data to learn and perform specific tasks.

[0803] A "means of use" is a method or device that applies a machine learning model to perform analysis or prediction.

[0804] A "user terminal" is a device such as a smartphone, PC, or tablet that a user interacts with directly.

[0805] A "transmitting means" is a method or apparatus for sending data to another device or system.

[0806] "Push notification" is a communication technology that sends specific information to a user's device in real time and displays it instantly.

[0807] A "cloud server" is a server that includes computing resources and storage provided over the Internet, and is a place where data is stored and computational processing is performed.

[0808] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. An embodiment of the system of the present invention will be described in detail below.

[0809] Data collection and preprocessing

[0810] The server collects a large amount of expert-labeled crying data sets from the Internet and stores them in a cloud database. The collected data includes audio files of babies and animals crying and their corresponding labels, such as "hungry," "sleepy," and "anxious." The server then preprocesses the collected audio data using the Python LibROSA library, removing noise and standardizing the sampling rate.

[0811] Training a machine learning model

[0812] The server splits the preprocessed data into a training dataset and a test dataset, then uses TensorFlow to build and train a convolutional neural network (CNN) or recurrent neural network (RNN). The trained model is then stored on the cloud server.

[0813] Audio recording and real-time analysis

[0814] The device (e.g., the user's smartphone or a dedicated device) has the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again using the LibROSA library and analyzes it using a trained machine learning model. This allows the emotions and needs indicated by the cry to be predicted.

[0815] Sending and notifying feedback

[0816] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent to the user's smartphone as a push notification. The device receives this and immediately notifies the user.

[0817] Specific examples

[0818] For example, if your baby starts crying in the middle of the night,

[0819] 1. The device records the baby's cry and sends the data to the server, which then activates the smartphone's recording app and captures the audio.

[0820] 2. The server re-preprocesses the received audio data and analyzes it using the trained machine learning model. It uses LibROSA on the server side to remove noise and input the data into the model.

[0821] 3. The server obtains the prediction result, "I'm hungry," and sends it to the device. The server uses an HTTP POST request to send the prediction result in JSON format.

[0822] 4. The device pushes a message to the user saying, "Your baby might be hungry." The smartphone displays the notification so that the user can check it.

[0823] The system can quickly determine why a baby or animal is crying, allowing parents and owners to take appropriate action, reducing stress and improving their health and happiness.

[0824] Prompt Sentence Examples

[0825] "Please explain how your smartphone automatically analyzes your baby's emotions and needs and notifies you when your baby is crying."

[0826] Based on this example, generative AI models can be used to provide context-sensitive feedback.

[0827] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0828] Step 1: Data collection

[0829] The server collects expertly labeled baby and animal cry datasets from the Internet and stores them in a cloud database. It performs the following operations:

[0830] Input: A crying audio file and its corresponding label (e.g., "hungry" or "sleepy").

[0831] Data Processing: During the data collection process, the data is stored in a database in a suitable format.

[0832] Output: A crying dataset stored in a cloud database.

[0833] Step 2: Data Preprocessing

[0834] The server performs noise reduction and sampling rate unification on the collected audio data. It uses the Python LibROSA library and performs the following operations.

[0835] Input: Audio data stored in a cloud database.

[0836] Data processing: LibROSA was used to remove background noise from the audio files and standardize the sampling rate to 44.1kHz.

[0837] Output: Preprocessed audio data.

[0838] Step 3: Train the model

[0839] The server splits the preprocessed data into training and test datasets and trains the model using a deep learning framework. It performs the following operations:

[0840] Input: Preprocessed audio data.

[0841] Data processing: 80% of the total data will be used as a training dataset and 20% as a test dataset.

[0842] Output: A dataset split into training and testing.

[0843] The server builds and trains models using convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

[0844] Input: Training dataset.

[0845] Data calculation: We use TensorFlow to train the model with 10 epochs.

[0846] Output: A trained machine learning model.

[0847] The server stores the trained model on a cloud server.

[0848] Input: A trained model.

[0849] Data calculation: Save the model using TensorFlow's model.save method.

[0850] Output: A trained model stored on a cloud server.

[0851] Step 4: Record your audio

[0852] The device (e.g., a smartphone) records the cries of babies and animals in real time. It performs the following operations:

[0853] Input: Babies and animals crying.

[0854] Data processing: Capture audio using a recording app.

[0855] Output: Recorded audio data.

[0856] Step 5: Sending audio data

[0857] The device sends the recorded audio data to a cloud server via the Internet. The following operations are performed:

[0858] Input: Recorded audio data.

[0859] Data calculation: Upload audio data to the server via HTTP POST request.

[0860] Output: The audio data sent to the server.

[0861] Step 6: Real-time analysis

[0862] The server then preprocesses the received audio data again and analyzes it using the trained machine learning model.

[0863] Input: Transmitted audio data.

[0864] Data processing: LibROSA is used to remove noise and unify the sampling rate again.

[0865] Output: Audio data in a format suitable for analysis.

[0866] The server uses trained machine learning models to analyze voice data and predict emotions and needs.

[0867] Input: Preprocessed audio data.

[0868] Data calculation: Using machine learning models, predict emotions and needs such as "hungry" or "sleepy."

[0869] Output: Predicted emotion and needs data.

[0870] Step 7: Submit your feedback

[0871] The server sends the analysis results to the user's terminal and performs the following operations.

[0872] Input: Parsed emotion and needs data.

[0873] Data calculation: Prediction results are sent to the terminal in JSON format via an HTTP POST request.

[0874] Output: Analysis results sent to the user's terminal.

[0875] Step 8: User Notification

[0876] The device will then push the received analysis results to the user. The device will perform the following operations:

[0877] Input: Analysis results sent from the server.

[0878] Data calculation: A message is displayed using the smartphone's notification function.

[0879] Output: The message that was sent to the user.

[0880] This step makes it possible to analyze the cries of babies and animals, predict their emotions and needs based on the analysis, and notify the user of the results in real time.

[0881] (Application example 1)

[0882] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0883] Conventional crying analysis systems simply analyze voice data on a local device, resulting in problems with the accuracy and real-time nature of the analysis results. Furthermore, when used in pet shops and veterinary clinics, a quick and appropriate response is required, particularly on-site, so it is necessary to provide more accurate analysis results. This invention aims to improve the accuracy of voice data analysis by utilizing a cloud server and provide users with quick and appropriate feedback.

[0884] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0885] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for transmitting the prediction results to a user terminal, means for notifying the user of the received prediction results, means for transmitting the received voice data to a cloud server and acquiring analysis results at the cloud server, and means for generating notification content based on the analysis results and instructing the user on an appropriate response. This allows highly accurate analysis results to be obtained quickly, enabling the user to immediately take appropriate measures.

[0886] "Audio data" is information that has been digitally recorded as the cries and cries of babies and animals.

[0887] "Preprocessing" is the process of processing audio data, such as removing noise and standardizing the sampling rate, to prepare it in a form suitable for analysis.

[0888] A "machine learning model" is a computational model that includes algorithms for learning from data and making predictions and classifications for new data.

[0889] "Deep learning" is a type of machine learning that uses multi-layered neural networks to analyze data and learn complex patterns.

[0890] A "convolutional neural network (CNN)" is a machine learning model primarily used to analyze image and audio data, and is a network with the ability to capture the spatial structure of data.

[0891] A "recurrent neural network (RNN)" is a machine learning model suitable for analyzing time series data, and is a network that proceeds with learning by taking into account the temporal dependency of the data.

[0892] A "cloud server" is a remote server that stores data and performs calculations over the Internet, and is a computer that provides services in response to requests from a large number of users.

[0893] A "user terminal" is a device used to record voice data, receive analysis results, and receive notifications, and specific examples include smartphones and dedicated devices.

[0894] A "notification" is a message that conveys analysis results and instructions based on those results to the user, and is mainly displayed in the form of a push notification or an alert.

[0895] "Analysis results" are predicted information about emotions and needs obtained after analyzing voice data using a machine learning model.

[0896] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. An embodiment of the system of the present invention will be described in detail below.

[0897] Data collection and preprocessing

[0898] The server collects a large amount of crying data sets labeled by experts and stores them in a cloud database. The data sets include audio files of crying babies and animals and corresponding labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed by removing noise and standardizing the sampling rate, resulting in data suitable for analysis.

[0899] Training a machine learning model

[0900] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). This model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and can predict emotions and needs based on that information. The resulting model is stored on a cloud server and used for real-time analysis.

[0901] Audio recording and real-time analysis

[0902] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotions and needs indicated by the crying are predicted.

[0903] Sending and notifying feedback

[0904] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent as a push notification to the user's smartphone. The device receives this and immediately notifies the user.

[0905] Examples from pet shops and veterinary clinics

[0906] The Pet Care Assistant application analyzes the cries (cries) of pets brought in by customers at pet shops and veterinary clinics. For example, if a dog barks at a pet shop, the device records the sound and sends it to a server on the cloud. The server analyzes the audio data and obtains a prediction result, such as "hungry." As a result, the store clerk receives a message saying, "This dog may be hungry," and can recommend appropriate dog food to the owner.

[0907] Hardware and software used

[0908] Hardware used

[0909] Smartphone: A device that uses the recording function to capture voice data and send it to a server

[0910] Cloud server: Runs the deep learning model and analyzes the voice data

[0911] Software used

[0912] Python: Writing program logic

[0913] pydub: A library for preprocessing audio data

[0914] requests: A library for managing HTTP requests / responses.

[0915] Examples and Prompts

[0916] For example, if a dog in a pet shop is barking for a specific reason, the Pet Care Assistant can analyze the sound and notify the store clerk that "This dog may be hungry," allowing the store to provide prompt and appropriate service to customers.

[0917] Prompt Sentence Examples

[0918] Audio data: Audio data in BASE64 format

[0919] Device ID: Device identifier

[0920] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0921] Step 1:

[0922] The device records surrounding audio in real time.

[0923] Input: Audio signal from the external environment

[0924] Output: Recorded audio data (e.g., WAV file)

[0925] What it does: Uses the microphone on your smartphone or dedicated device to capture the cries of babies and animals.

[0926] Step 2:

[0927] The terminal preprocesses the recorded audio data.

[0928] Input: Recorded audio data

[0929] Output: Preprocessed audio data

[0930] Specific operations: Noise removal and unification of sampling rates are performed on the audio data to prepare it in a form suitable for analysis.

[0931] Step 3:

[0932] The terminal transmits the pre-processed voice data to a server on the cloud via the Internet.

[0933] Input: Preprocessed audio data

[0934] Output: Data transmission result to cloud server (success / failure)

[0935] Specific operation: Generate an HTTP request and send the preprocessed audio data to the cloud server.

[0936] Step 4:

[0937] The server again pre-processes the received audio data.

[0938] Input: Preprocessed audio data

[0939] Output: Re-preprocessed audio data

[0940] Specific operation: Additional noise reduction and format conversion is performed on the server side as needed.

[0941] Step 5:

[0942] The server analyzes the audio data using a trained machine learning model.

[0943] Input: Re-preprocessed audio data

[0944] Output: Prediction of emotions and needs corresponding to crying

[0945] Specific operation: Using convolutional neural networks (CNN) and recurrent neural networks (RNN), voice features are extracted and emotions and needs are predicted.

[0946] Step 6:

[0947] The server transmits the analysis results to the user's terminal.

[0948] Input: Prediction result (e.g., "I'm hungry," "I'm sleepy," etc.)

[0949] Output: Data transmission result to user device (success / failure)

[0950] Specific operation: Generate an HTTP response and send a message containing the prediction result to the user terminal.

[0951] Step 7:

[0952] The terminal notifies the user of the received analysis results.

[0953] Input: A message containing the analysis results

[0954] Output: Notification to the user (e.g. push notification)

[0955] Specific operation: The analysis results are displayed as a push notification using the notification function of the user's smartphone or dedicated device.

[0956] Step 8:

[0957] The user checks the notification and takes appropriate action.

[0958] Input: Notification of analysis results

[0959] Output: User's response (e.g., feeding a baby, feeding a dog, etc.)

[0960] Specific Action: Based on the notification content, the user performs an action according to the request of the baby or pet.

[0961] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0962] The present invention relates to a system that analyzes the cries of babies and animals, recognizes not only their emotions and needs but also the emotions of a user, and provides feedback. An embodiment of the system of the present invention will be described in detail below.

[0963] Data collection and preprocessing

[0964] The server collects a large amount of crying data labeled by experts and stores it in a cloud database. The data set includes audio files of crying babies and animals, along with labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed, including noise removal and standardization of the sampling rate, to obtain data suitable for analysis.

[0965] Training a machine learning model

[0966] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). The model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs based on them. The resulting model is stored on a cloud server and used for real-time analysis.

[0967] User Emotion Recognition

[0968] The device is equipped with an emotion engine that recognizes the user's emotions based on their voice input. This engine analyzes the user's speech data in real time and recognizes their emotional state, such as "fatigue," "stress," or "relief." This is done automatically when the user speaks to the device.

[0969] Audio recording and real-time analysis

[0970] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the crying of a baby or animal in real time. When a cry is detected, the device records the audio data and sends it to a server in the cloud. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotion and needs indicated by the cry are predicted.

[0971] Sending and notifying feedback

[0972] The server then sends the results of the analysis of the baby or animal's cry, along with the user's emotion recognition results, to the user's device. Specifically, the server sends a message containing information about the cause of the cry and advice or suggestions based on the user's emotional state.

[0973] The device will notify the user of these results, which will be displayed as push notifications or alerts, allowing the user to take immediate action.

[0974] Specific examples

[0975] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device recognizes the user's emotions, such as "fatigue" and "stress," in real time. The server analyzes the received voice data and predicts that the user is "hungry." If the server determines that the user is feeling tired, it creates a message containing advice such as "Your baby may be hungry. If possible, ask a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[0976] This embodiment allows for quick understanding of why a baby or animal is crying and allows for a response that takes into account the parent's or owner's own emotional state, which is expected to improve the health and well-being of not only the baby or animal, but also the parent or owner themselves.

[0977] The processing flow will be explained below.

[0978] Step 1:

[0979] The server collects expert-labeled datasets of baby and animal cries and stores them in a cloud database, including audio files and labels such as "hungry," "sleepy," and "anxious."

[0980] Step 2:

[0981] The server preprocesses the collected audio data, which includes noise removal and sampling rate unification, to prepare the data for analysis.

[0982] Step 3:

[0983] The server uses the preprocessed data to train machine learning models, which are based on convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

[0984] Step 4:

[0985] The server stores the trained machine learning model in the cloud, which is then used for subsequent real-time analysis.

[0986] Step 5:

[0987] The device has the ability to record the crying sounds of babies and animals in real time. When a crying sound is detected, the device will start recording the audio data.

[0988] Step 6:

[0989] The device transmits the recorded voice data to a server via the Internet, and simultaneously detects the user's voice input and analyzes it using an emotion engine.

[0990] Step 7:

[0991] The server preprocesses the received audio data, which includes noise reduction and sampling rate unification.

[0992] Step 8:

[0993] The server uses the pre-processed audio data to analyze it with a trained machine learning model, which predicts the cause of the crying (hungry, sleepy, etc.).

[0994] Step 9:

[0995] The server integrates the results of the user's emotion recognition with the results of the crying analysis. For example, if the user feels "tired," that information is also included in the prediction results.

[0996] Step 10:

[0997] The server sends the integrated results to the user's terminal, and the message includes the cause of the crying as well as advice based on the user's condition.

[0998] Step 11:

[0999] The device notifies the user of messages received from the server, which are displayed as push notifications or alerts.

[1000] Step 12:

[1001] The user can check the notification and respond to the baby or animal's needs based on the notification. For example, if the system predicts that the baby is hungry and the user feels tired, it will suggest asking other family members for help. The user can also take action such as feeding the baby if necessary.

[1002] This series of steps enables the system to quickly identify the cause of a baby or animal's cry and provide appropriate advice that takes into account the user's emotional state.

[1003] Example 2

[1004] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1005] Conventional emotion recognition systems are limited in their ability to analyze the cries of babies and animals, and are unable to simultaneously recognize the user's emotional state and provide appropriate feedback. This makes it difficult for parents and owners to respond efficiently and quickly, often resulting in stress and fatigue. Furthermore, analysis results can sometimes be inaccurate, preventing effective countermeasures from being implemented.

[1006] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1007] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for analyzing the user's voice data in real time and recognizing the emotional state, means for recording the crying of a baby or animal in real time and transmitting the voice data to the server, means for transmitting the prediction result to a user terminal, and means for notifying the user of the received prediction result. This makes it possible to quickly and accurately understand the meaning of the crying of a baby or animal and provide appropriate feedback according to the user's emotional state.

[1008] "Means for receiving audio data" refers to the function that allows a user terminal or dedicated device to record the cries of babies or animals and send them to a server.

[1009] "Means for preprocessing received voice data" refers to a function for performing preprocessing such as noise removal and unifying the sampling rate on collected voice data.

[1010] "Means of analyzing pre-processed voice data and using machine learning models to predict emotions and needs" refers to a function that analyzes voice data using deep learning convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to predict emotions and needs.

[1011] "Means for analyzing the user's voice data in real time and recognizing their emotional state" refers to a function that analyzes the voice data spoken by the user in real time and recognizes their emotional state, such as "fatigue" or "stress."

[1012] "Means for recording the cries of babies and animals in real time and transmitting the audio data to a server" refers to a function that enables a user terminal or dedicated device to record the cries of babies and animals in real time and transmit the audio data to a cloud server.

[1013] "Means for transmitting prediction results to a user terminal" refers to a function that enables the server to transmit analysis results and feedback messages to a user terminal.

[1014] "Means for notifying the user of the received prediction results" refers to a function that allows the user's device to notify the user of the prediction results received from the server as a push notification or an alert.

[1015] MODE FOR CARRYING OUT THE INVENTION

[1016] The present invention relates to a system that analyzes the cries of babies and animals, recognizes not only their emotions and needs but also the emotions of the user, and provides feedback. This system performs a series of processes from collecting audio data to analyzing it and providing feedback. An embodiment of the system of the present invention will be described in detail below.

[1017] Data collection and preprocessing

[1018] The server collects a large amount of expert-labeled crying data sets from online repositories and stores them in a cloud database. The collected data includes audio files of crying babies and animals, along with labels such as "hungry," "sleepy," and "anxious." The data is then converted to a uniform sampling rate (e.g., 44.1 kHz) and preprocessed with noise reduction (using a Gaussian filter, etc.).

[1019] Training a machine learning model

[1020] The server then loads the preprocessed audio data into a deep learning environment (e.g., TensorFlow or PyTorch) and trains it using a convolutional neural network (CNN) or recurrent neural network (RNN). This training creates a model that learns the feature vectors of the audio data (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs with high accuracy. After training is complete, the model is stored on a cloud server and made available for real-time analysis.

[1021] User Emotion Recognition

[1022] The device is equipped with an emotion engine that analyzes emotions in real time based on the user's voice input. This engine uses, for example, IBM Watson Tone Analyzer to analyze the voice data of the user's speech (e.g., "I'm tired," "I'm irritated," etc.) and recognizes the user's emotional state. This real-time recognition is performed in a natural situation where the user is speaking to the device.

[1023] Audio recording and real-time analysis

[1024] The device (e.g., a smartphone or dedicated device) has the ability to record the cries of babies and animals in real time. When a cry is detected, the device automatically starts recording and sends the audio data to a cloud-based server. The server then preprocesses the received audio data and analyzes it using a trained machine learning model. This allows for highly accurate prediction of the emotions and needs indicated by the cry.

[1025] Sending and notifying feedback

[1026] The server combines the results of the crying analysis and the user's emotion recognition and sends them to the user's device. Specific feedback includes information about the cause of the crying and advice or suggestions based on the user's emotional state. For example, "Your baby may be hungry. If possible, ask a family member for help." The device then notifies the user of this feedback via a push notification or alert, encouraging them to take prompt action.

[1027] Specific examples

[1028] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device analyzes the user's voice and recognizes emotions such as "fatigue" and "stress" in real time. The server analyzes the received crying data and predicts that the baby is "hungry." If the server determines that the user is tired, it creates a message containing advice such as "Your baby may be hungry. If possible, try asking a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[1029] Prompt Sentence Examples

[1030] 1. "Analyze an audio file of a crying baby and predict its emotions and needs. The user's emotional state is 'Tired'."

[1031] 2. "Analyze a dog's cry and tell me its cause. The user's emotion is 'stress'."

[1032] This system can quickly and accurately identify the emotions and needs expressed by the cries of babies and animals, and provide appropriate feedback based on the user's emotional state, which is expected to improve the health and happiness of not only babies and animals, but also their parents and owners.

[1033] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1034] Step 1: Collect data

[1035] The server collects expert-labeled crying datasets from online repositories. The input includes crying audio data and corresponding emotion and need labels. The output is stored in a cloud database.

[1036] Step 2: Preprocessing the data

[1037] The server performs preprocessing on the collected audio data. The raw audio data collected in step 1 is used as input. Specific preprocessing operations include noise removal (applying a Gaussian filter) and unifying the sampling rate (e.g., changing to 44.1 kHz). The preprocessed audio data is obtained as output.

[1038] Step 3: Train the machine learning model

[1039] The server uses the preprocessed audio data to train a machine learning model in a deep learning environment (such as TensorFlow or PyTorch). The preprocessed audio data and its associated label information are used as input. Specifically, a convolutional neural network (CNN) or recurrent neural network (RNN) is used to learn feature vectors (e.g., pitch, rhythm, decibels). The output is a trained model.

[1040] Step 4: Recognizing user emotions

[1041] The device collects the user's voice data in real time and analyzes their emotions. The user's spoken voice data is used as input. Specifically, it uses an emotion engine (for example, IBM Watson Tone Analyzer) to recognize emotional states such as "fatigue" and "stress" in real time. The output is the recognition result of the user's emotional state.

[1042] Step 5: Record your audio

[1043] The device records the cries of babies and animals in real time. The cries of babies and animals are used as input. Specifically, the device's microphone is used to record the cries and save them as audio files. The recorded audio data is obtained as output.

[1044] Step 6: Real-time analysis

[1045] The device sends the recorded crying data to a server on the cloud. The audio data recorded in step 5 is used as input. The server receives this audio data, preprocesses it again, and then analyzes it using a trained machine learning model. Specific operations include preprocessing (noise removal, unifying the sampling rate) and feature extraction (pitch, rhythm, decibel analysis). The analysis results are obtained as output.

[1046] Step 7: Feedback Generation

[1047] The server generates a feedback message based on the analysis results and the user's emotion recognition results. The analysis results from step 6 and the emotion recognition results from step 4 are used as input. The specific operation is to combine information about the cause of the crying with advice or suggestions based on the user's emotional state. The output is a feedback message.

[1048] Step 8: Notification

[1049] The server sends the generated feedback message to the user device. The feedback message generated in step 7 is used as input. The device receives this message and displays it to the user as a push notification or an alert. Specific actions include playing a notification sound or displaying a message on the screen. The output is that the user can confirm the feedback message.

[1050] In this way, the system can analyze the cries of babies and animals and provide appropriate feedback that takes into account the user's emotional state.

[1051] (Application example 2)

[1052] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1053] Conventional systems for analyzing the cries of babies and animals are limited to analyzing only the emotions and needs of the cry and providing feedback based on that analysis, and lack consideration of the user's emotions and state. Furthermore, feedback after analyzing the cry is not linked to the user's emotional state, making it difficult for the user to respond appropriately and promptly. The present invention aims to solve these problems by providing more appropriate feedback by taking the user's emotional state into account, enabling a prompt response.

[1054] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1055] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for recognizing a user's voice input and analyzing the user's emotions, means for transmitting advice or suggestions based on the prediction results and the user's emotion analysis results to a user terminal, and means for notifying the user of the received prediction results or advice. This makes it possible to provide appropriate feedback that takes into account not only the analysis of the crying of a baby or animal but also the user's emotional state.

[1056] The "means for receiving audio data" refers to a device or module that has the function of capturing audio data, including the cries of babies and animals, in real time and incorporating it into the system.

[1057] The "means for preprocessing received voice data" refers to a device or module that has the function of processing collected voice data, such as noise removal and standardizing the sampling rate, and converting it into a format suitable for analysis.

[1058] "Means of using machine learning models to analyze preprocessed voice data and predict emotions and needs" refers to algorithms and systems that use deep learning techniques to analyze and predict emotions and needs from preprocessed voice data.

[1059] "Means for recognizing user voice input and analyzing the user's emotions" refers to a device or module that has the function of analyzing the user's voice in real time and determining the user's emotional state, such as "fatigue," "stress," or "relief."

[1060] "Means for sending advice and suggestions based on prediction results and user emotion analysis results to a user terminal" refers to a device or system that has the function of generating advice and suggestions that take into account the user's emotional state based on the analysis results and sending them to the user's smartphone or dedicated device.

[1061] The "means for notifying the user of the received prediction results and advice" refers to a device or application that has a notification function for appropriately notifying the user of the analysis results and advice sent to the terminal.

[1062] Data collection and preprocessing

[1063] The server collects a large amount of crying data labeled by experts and stores it in a cloud database. The data set includes audio files of crying babies and animals, along with associated labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed, including noise removal and standardization of the sampling rate, to obtain data suitable for analysis.

[1064] Training a machine learning model

[1065] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). The model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs based on them. The resulting model is stored on a cloud server and used for real-time analysis.

[1066] User Emotion Recognition

[1067] The device is equipped with an emotion engine that recognizes the user's emotions based on their voice input. This engine analyzes the user's speech data in real time and recognizes emotional states such as "fatigue," "stress," and "relief." This is done automatically when the user speaks to the device.

[1068] Audio recording and real-time analysis

[1069] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the crying of a baby or animal in real time. When a cry is detected, the device records the audio data and sends it to a server in the cloud. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotion and needs indicated by the cry are predicted.

[1070] Sending and notifying feedback

[1071] The server sends the results of analyzing the baby or animal's cry, along with the user's emotion recognition results, to the user's device. Specifically, information about the cause of the cry and advice or suggestions based on the user's emotional state are sent as a message. The device then notifies the user of these results. The notifications are displayed as push notifications or alerts, helping the user respond immediately.

[1072] Specific examples

[1073] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device recognizes the user's emotions, such as "fatigue" or "stress," in real time. The server analyzes the received voice data and predicts that the user is "hungry." If the server determines that the user is feeling tired, it creates a message containing advice such as, "Your baby may be hungry. If possible, ask a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[1074] Prompt Sentence Examples

[1075] We will imagine a situation where a baby or animal starts crying in the middle of the night, record an audio file of the crying, and analyze it. We will also record the user's voice and analyze their emotions. Based on these two results, we will design a system that notifies the user of appropriate advice.

[1076] This embodiment allows for quick understanding of why a baby or animal is crying and allows for a response that takes into account the parent's or owner's own emotional state, which is expected to improve the health and well-being of not only the baby or animal, but also the parent or owner themselves.

[1077] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1078] Step 1:

[1079] The device records the cries of babies and animals in real time. The input is environmental audio data, and the output is a recorded audio data file. Specifically, the device's microphone captures the audio and saves it as a digital audio file.

[1080] Step 2:

[1081] The terminal sends the recorded audio data file to the server. The input is the audio data file recorded in step 1, and the output is the audio data sent to the server. Specifically, the audio data file is uploaded to the server via an Internet connection.

[1082] Step 3:

[1083] The server preprocesses the received audio data. The input is the audio data sent to the server, and the output is the preprocessed audio data. Specifically, the server removes noise, unifies the sampling rate, and converts the data into a format suitable for analysis.

[1084] Step 4:

[1085] The server analyzes the preprocessed audio data using a machine learning model. The input is the preprocessed audio data, and the output is the predicted emotion and needs of the crying voice. Specifically, it uses a convolutional neural network (CNN) and a recurrent neural network (RNN) to analyze the audio features and classify the emotion and needs.

[1086] Step 5:

[1087] The device recognizes the user's voice input in real time and analyzes the user's emotions. The input is the voice data spoken by the user, and the output is the analysis result of the user's emotional state. Specifically, the emotion engine installed in the device analyzes the voice and determines the user's emotional state, such as "fatigue," "stress," or "relief."

[1088] Step 6:

[1089] The server integrates the prediction results and the user's emotion analysis results based on the analysis results to generate advice and suggestions. The input is the prediction results of the crying emotion and needs and the analysis results of the user's emotional state, and the output is a message containing advice and suggestions. Specifically, the server uses a generative AI model to create appropriate advice and suggestions.

[1090] Step 7:

[1091] The server sends the generated advice and suggestion messages to the device. The input is the message containing the advice or suggestion, and the output is the message sent to the device. Specifically, the message is delivered to the device via the cloud server.

[1092] Step 8:

[1093] The device notifies the user of the received advice or suggestion message. The input is the advice or suggestion message sent to the device, and the output is the notification displayed to the user. Specifically, the device displays the message to the user using a push notification or a pop-up alert.

[1094] Prompt Sentence Examples

[1095] We will imagine a situation where a baby or animal starts crying in the middle of the night, record an audio file of the crying, and analyze it. We will also record the user's voice and analyze their emotions. Based on these two results, we will design a system that notifies the user of appropriate advice.

[1096] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1097] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1098] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1099] [Fourth embodiment]

[1100] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1101] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1102] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1103] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1104] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1105] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1106] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1107] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1108] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1109] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1110] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1111] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1112] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1113] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. Hereinafter, an embodiment of the system of the present invention will be described in detail.

[1114] Data collection and preprocessing

[1115] The server collects a large amount of crying data sets labeled by experts and stores them in a cloud database. The data sets include audio files of crying babies and animals and corresponding labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed by removing noise and standardizing the sampling rate, resulting in data suitable for analysis.

[1116] Training a machine learning model

[1117] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). This model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and can predict emotions and needs based on that information. The resulting model is stored on a cloud server and used for real-time analysis.

[1118] Audio recording and real-time analysis

[1119] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotions and needs indicated by the crying are predicted.

[1120] Sending and notifying feedback

[1121] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent as a push notification to the user's smartphone. The device receives this and immediately notifies the user.

[1122] Specific examples

[1123] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's cry and sends the data to the server. The server analyzes this audio data and predicts that the baby is hungry. The prediction result is returned from the server to the device, which then notifies the user with a message saying, "Your baby may be hungry." The user can check this notification and immediately prepare milk to feed the baby. In this way, a quick and accurate response is possible.

[1124] This invention makes it possible to quickly determine why a baby or animal is crying, making it easier for parents and owners to take appropriate action, which is expected to reduce stress for babies and animals and improve their health and happiness.

[1125] The processing flow will be explained below.

[1126] Step 1:

[1127] The server collects expert-labeled datasets of baby and animal cries and stores them in a cloud database, which includes audio files and corresponding labels such as "hungry," "sleepy," and "anxious."

[1128] Step 2:

[1129] The server preprocesses the collected audio data, including noise removal, unifying the sampling rate, normalizing, etc. The preprocessed data is then converted into a form suitable for analysis.

[1130] Step 3:

[1131] The server uses the preprocessed data to train machine learning models, specifically convolutional neural networks (CNNs) and recurrent neural networks (RNNs), to learn audio features (e.g., pitch, rhythm, decibels).

[1132] Step 4:

[1133] The server stores the trained machine learning model in the cloud, which is then used for real-time analytics.

[1134] Step 5:

[1135] The device has the ability to record the crying sounds of babies and animals in real time. When a crying sound is detected, the device will start recording the audio data.

[1136] Step 6:

[1137] The device transmits the recorded audio data to a cloud server via the Internet.

[1138] Step 7:

[1139] The server then pre-processes the received audio data again, which includes noise reduction and sampling rate unification.

[1140] Step 8:

[1141] The server then uses the pre-processed audio data to analyze it using a trained machine learning model, which predicts the crying as an emotion or need, such as "hunger," "sleepy," or "anxiety."

[1142] Step 9:

[1143] The server sends the analysis results to the user's device, which include a specific alert message.

[1144] Step 10:

[1145] The device will then notify the user of the received prediction results, which will be displayed as a push notification or an alert.

[1146] Step 11:

[1147] The user can check the notification and respond to the baby or animal's needs based on the notification. For example, if the baby or animal is predicted to be hungry, the user can prepare milk and give it to the baby. This process allows parents and owners to quickly understand the cause of the crying and take appropriate action.

[1148] Example 1

[1149] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1150] Conventional voice recognition systems lack the means to analyze the cries of babies and animals and provide fast and accurate feedback on their emotions and needs. Furthermore, due to the lack of proper noise reduction and standardization of sampling rates for each individual cry, analysis results tend to be inaccurate. Furthermore, the lack of a function to notify the user of prediction results in real time makes it difficult to respond quickly. There is a need to address these issues with conventional technology.

[1151] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1152] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data based on noise reduction and unified sampling rate, means for analyzing the preprocessed voice data and using a machine learning model consisting of a convolutional neural network and a recurrent neural network to predict emotions and needs, means for transmitting the prediction results to a user terminal, and means for sending a push notification of the received prediction results to the user. This makes it possible to quickly and accurately analyze the cries of babies and animals and predict emotions and needs based on the analysis, and notify the user of the results in real time.

[1153] "Voice data" is data that represents the voice of a human or animal in digital form.

[1154] A "means for receiving" is a method or apparatus for obtaining audio data from another device or system.

[1155] "Noise reduction" is the process of removing unwanted background sounds and noise from audio data.

[1156] "Sampling rate unification" is the process of making the sampling frequencies of different audio data uniform.

[1157] "Preprocessing means" refers to a method or device that performs necessary processing or conversion on the audio data before analysis.

[1158] "Emotions and needs" refers to the feelings and demands that babies and animals express through their cries.

[1159] A "convolutional neural network" is a type of deep learning model that excels at analyzing data such as images and audio.

[1160] A "recurrent neural network" is a type of deep learning model that excels at analyzing continuous data over time.

[1161] A "machine learning model" is an artificial intelligence algorithm that uses data to learn and perform specific tasks.

[1162] A "means of use" is a method or device that applies a machine learning model to perform analysis or prediction.

[1163] A "user terminal" is a device such as a smartphone, PC, or tablet that a user interacts with directly.

[1164] A "transmitting means" is a method or apparatus for sending data to another device or system.

[1165] "Push notification" is a communication technology that sends specific information to a user's device in real time and displays it instantly.

[1166] A "cloud server" is a server that includes computing resources and storage provided over the Internet, and is a place where data is stored and computational processing is performed.

[1167] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. An embodiment of the system of the present invention will be described in detail below.

[1168] Data collection and preprocessing

[1169] The server collects a large amount of expert-labeled crying data sets from the Internet and stores them in a cloud database. The collected data includes audio files of babies and animals crying and their corresponding labels, such as "hungry," "sleepy," and "anxious." The server then preprocesses the collected audio data using the Python LibROSA library, removing noise and standardizing the sampling rate.

[1170] Training a machine learning model

[1171] The server splits the preprocessed data into a training dataset and a test dataset, then uses TensorFlow to build and train a convolutional neural network (CNN) or recurrent neural network (RNN). The trained model is then stored on the cloud server.

[1172] Audio recording and real-time analysis

[1173] The device (e.g., the user's smartphone or a dedicated device) has the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again using the LibROSA library and analyzes it using a trained machine learning model. This allows the emotions and needs indicated by the cry to be predicted.

[1174] Sending and notifying feedback

[1175] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent to the user's smartphone as a push notification. The device receives this and immediately notifies the user.

[1176] Specific examples

[1177] For example, if your baby starts crying in the middle of the night,

[1178] 1. The device records the baby's cry and sends the data to the server, which then activates the smartphone's recording app and captures the audio.

[1179] 2. The server re-preprocesses the received audio data and analyzes it using the trained machine learning model. It uses LibROSA on the server side to remove noise and input the data into the model.

[1180] 3. The server obtains the prediction result, "I'm hungry," and sends it to the device. It uses an HTTP POST request to send the prediction result in JSON format.

[1181] 4. The device pushes a message to the user saying, "Your baby might be hungry." The smartphone displays the notification so that the user can check it.

[1182] The system can quickly determine why a baby or animal is crying, allowing parents and owners to take appropriate action, reducing stress and improving their health and happiness.

[1183] Prompt Sentence Examples

[1184] "Please explain how your smartphone automatically analyzes your baby's emotions and needs and notifies you when your baby is crying."

[1185] Based on this example, generative AI models can be used to provide context-sensitive feedback.

[1186] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1187] Step 1: Data collection

[1188] The server collects expertly labeled baby and animal cry datasets from the Internet and stores them in a cloud database. It performs the following operations:

[1189] Input: A crying audio file and its corresponding label (e.g., "hungry" or "sleepy").

[1190] Data Processing: During the data collection process, the data is stored in a database in a suitable format.

[1191] Output: A crying dataset stored in a cloud database.

[1192] Step 2: Data Preprocessing

[1193] The server performs noise reduction and sampling rate unification on the collected audio data. It uses the Python LibROSA library and performs the following operations.

[1194] Input: Audio data stored in a cloud database.

[1195] Data processing: LibROSA was used to remove background noise from the audio files and standardize the sampling rate to 44.1kHz.

[1196] Output: Preprocessed audio data.

[1197] Step 3: Train the model

[1198] The server splits the preprocessed data into training and test datasets and trains the model using a deep learning framework. It performs the following operations:

[1199] Input: Preprocessed audio data.

[1200] Data processing: 80% of the total data will be used as a training dataset and 20% as a test dataset.

[1201] Output: A dataset split into training and testing.

[1202] The server builds and trains models using convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

[1203] Input: Training dataset.

[1204] Data calculation: We use TensorFlow to train the model with 10 epochs.

[1205] Output: A trained machine learning model.

[1206] The server stores the trained model on a cloud server.

[1207] Input: A trained model.

[1208] Data calculation: Save the model using TensorFlow's model.save method.

[1209] Output: A trained model stored on a cloud server.

[1210] Step 4: Record your audio

[1211] The device (e.g., a smartphone) records the cries of babies and animals in real time. It performs the following operations:

[1212] Input: Babies and animals crying.

[1213] Data processing: Capture audio using a recording app.

[1214] Output: Recorded audio data.

[1215] Step 5: Sending audio data

[1216] The device sends the recorded audio data to a cloud server via the Internet. The following operations are performed:

[1217] Input: Recorded audio data.

[1218] Data calculation: Upload audio data to the server via HTTP POST request.

[1219] Output: The audio data sent to the server.

[1220] Step 6: Real-time analysis

[1221] The server then preprocesses the received audio data again and analyzes it using the trained machine learning model.

[1222] Input: Transmitted audio data.

[1223] Data processing: LibROSA is used to remove noise and unify the sampling rate again.

[1224] Output: Audio data in a format suitable for analysis.

[1225] The server uses trained machine learning models to analyze voice data and predict emotions and needs.

[1226] Input: Preprocessed audio data.

[1227] Data calculation: Using machine learning models, predict emotions and needs such as "hungry" or "sleepy."

[1228] Output: Predicted emotion and needs data.

[1229] Step 7: Submit your feedback

[1230] The server sends the analysis results to the user's terminal and performs the following operations.

[1231] Input: Parsed emotion and needs data.

[1232] Data calculation: Prediction results are sent to the terminal in JSON format via an HTTP POST request.

[1233] Output: Analysis results sent to the user's terminal.

[1234] Step 8: User Notification

[1235] The device will then push the received analysis results to the user. The device will perform the following operations:

[1236] Input: Analysis results sent from the server.

[1237] Data calculation: A message is displayed using the smartphone's notification function.

[1238] Output: The message that was sent to the user.

[1239] This step makes it possible to analyze the cries of babies and animals, predict their emotions and needs based on the analysis, and notify the user of the results in real time.

[1240] (Application example 1)

[1241] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1242] Conventional crying analysis systems simply analyze voice data on a local device, resulting in problems with the accuracy and real-time nature of the analysis results. Furthermore, when used in pet shops and veterinary clinics, a quick and appropriate response is required, particularly on-site, so it is necessary to provide more accurate analysis results. This invention aims to improve the accuracy of voice data analysis by utilizing a cloud server and provide users with quick and appropriate feedback.

[1243] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1244] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for transmitting the prediction results to a user terminal, means for notifying the user of the received prediction results, means for transmitting the received voice data to a cloud server and acquiring analysis results at the cloud server, and means for generating notification content based on the analysis results and instructing the user on an appropriate response. This allows highly accurate analysis results to be obtained quickly, enabling the user to immediately take appropriate measures.

[1245] "Audio data" is information that has been digitally recorded as the cries and cries of babies and animals.

[1246] "Preprocessing" is the process of processing audio data, such as removing noise and standardizing the sampling rate, to prepare it in a form suitable for analysis.

[1247] A "machine learning model" is a computational model that includes algorithms for learning from data and making predictions and classifications for new data.

[1248] "Deep learning" is a type of machine learning that uses multi-layered neural networks to analyze data and learn complex patterns.

[1249] A "convolutional neural network (CNN)" is a machine learning model primarily used to analyze image and audio data, and is a network with the ability to capture the spatial structure of data.

[1250] A "recurrent neural network (RNN)" is a machine learning model suitable for analyzing time series data, and is a network that proceeds with learning by taking into account the temporal dependency of the data.

[1251] A "cloud server" is a remote server that stores data and performs calculations over the Internet, and is a computer that provides services in response to requests from a large number of users.

[1252] A "user terminal" is a device used to record voice data, receive analysis results, and receive notifications, and specific examples include smartphones and dedicated devices.

[1253] A "notification" is a message that conveys analysis results and instructions based on those results to the user, and is mainly displayed in the form of a push notification or an alert.

[1254] "Analysis results" are predicted information about emotions and needs obtained after analyzing voice data using a machine learning model.

[1255] The present invention relates to a system that analyzes the cries of babies and animals and provides feedback on their emotions and needs to their parents or owners. An embodiment of the system of the present invention will be described in detail below.

[1256] Data collection and preprocessing

[1257] The server collects a large amount of crying data sets labeled by experts and stores them in a cloud database. The data sets include audio files of crying babies and animals and corresponding labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed by removing noise and standardizing the sampling rate, resulting in data suitable for analysis.

[1258] Training a machine learning model

[1259] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). This model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and can predict emotions and needs based on that information. The resulting model is stored on a cloud server and used for real-time analysis.

[1260] Audio recording and real-time analysis

[1261] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the cries of babies and animals in real time. The recorded audio data is then sent over the Internet to a cloud server. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotions and needs indicated by the crying are predicted.

[1262] Sending and notifying feedback

[1263] The server sends the analysis results to the user's device. Specifically, a message containing the emotion and needs of the baby or animal's cry (e.g., "hungry" or "sleepy") is sent as a push notification to the user's smartphone. The device receives this and immediately notifies the user.

[1264] Examples from pet shops and veterinary clinics

[1265] The Pet Care Assistant application analyzes the cries (cries) of pets brought in by customers at pet shops and veterinary clinics. For example, if a dog barks at a pet shop, the device records the sound and sends it to a server on the cloud. The server analyzes the audio data and obtains a prediction result, such as "hungry." As a result, the store clerk receives a message saying, "This dog may be hungry," and can recommend appropriate dog food to the owner.

[1266] Hardware and software used

[1267] Hardware used

[1268] Smartphone: A device that uses the recording function to capture voice data and send it to a server

[1269] Cloud server: Runs the deep learning model and analyzes the voice data

[1270] Software used

[1271] Python: Writing program logic

[1272] pydub: A library for preprocessing audio data

[1273] requests: A library for managing HTTP requests / responses.

[1274] Examples and Prompts

[1275] For example, if a dog in a pet shop is barking for a specific reason, the Pet Care Assistant can analyze the sound and notify the store clerk that "This dog may be hungry," allowing the store to provide prompt and appropriate service to customers.

[1276] Prompt Sentence Examples

[1277] Audio data: Audio data in BASE64 format

[1278] Device ID: Device identifier

[1279] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1280] Step 1:

[1281] The device records surrounding audio in real time.

[1282] Input: Audio signal from the external environment

[1283] Output: Recorded audio data (e.g., WAV file)

[1284] What it does: Uses the microphone on your smartphone or dedicated device to capture the cries of babies and animals.

[1285] Step 2:

[1286] The terminal preprocesses the recorded audio data.

[1287] Input: Recorded audio data

[1288] Output: Preprocessed audio data

[1289] Specific operations: Noise removal and unification of sampling rates are performed on the audio data to prepare it in a form suitable for analysis.

[1290] Step 3:

[1291] The terminal transmits the pre-processed voice data to a server on the cloud via the Internet.

[1292] Input: Preprocessed audio data

[1293] Output: Data transmission result to cloud server (success / failure)

[1294] Specific operation: Generate an HTTP request and send the preprocessed audio data to the cloud server.

[1295] Step 4:

[1296] The server again pre-processes the received audio data.

[1297] Input: Preprocessed audio data

[1298] Output: Re-preprocessed audio data

[1299] Specific operation: Additional noise reduction and format conversion is performed on the server side as needed.

[1300] Step 5:

[1301] The server analyzes the audio data using a trained machine learning model.

[1302] Input: Re-preprocessed audio data

[1303] Output: Prediction of emotions and needs corresponding to crying

[1304] Specific operation: Using convolutional neural networks (CNN) and recurrent neural networks (RNN), voice features are extracted and emotions and needs are predicted.

[1305] Step 6:

[1306] The server transmits the analysis results to the user's terminal.

[1307] Input: Prediction result (e.g., "I'm hungry," "I'm sleepy," etc.)

[1308] Output: Data transmission result to user device (success / failure)

[1309] Specific operation: Generate an HTTP response and send a message containing the prediction result to the user terminal.

[1310] Step 7:

[1311] The terminal notifies the user of the received analysis results.

[1312] Input: A message containing the analysis results

[1313] Output: Notification to the user (e.g. push notification)

[1314] Specific operation: The analysis results are displayed as a push notification using the notification function of the user's smartphone or dedicated device.

[1315] Step 8:

[1316] The user checks the notification and takes appropriate action.

[1317] Input: Notification of analysis results

[1318] Output: User's response (e.g., feeding a baby, feeding a dog, etc.)

[1319] Specific Action: Based on the notification content, the user performs an action according to the request of the baby or pet.

[1320] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1321] The present invention relates to a system that analyzes the cries of babies and animals, recognizes not only their emotions and needs but also the emotions of a user, and provides feedback. An embodiment of the system of the present invention will be described in detail below.

[1322] Data collection and preprocessing

[1323] The server collects a large amount of crying data labeled by experts and stores it in a cloud database. The data set includes audio files of crying babies and animals, along with labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed, including noise removal and standardization of the sampling rate, to obtain data suitable for analysis.

[1324] Training a machine learning model

[1325] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). The model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs based on them. The resulting model is stored on a cloud server and used for real-time analysis.

[1326] User Emotion Recognition

[1327] The device is equipped with an emotion engine that recognizes the user's emotions based on their voice input. This engine analyzes the user's speech data in real time and recognizes their emotional state, such as "fatigue," "stress," or "relief." This is done automatically when the user speaks to the device.

[1328] Audio recording and real-time analysis

[1329] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the crying of a baby or animal in real time. When a cry is detected, the device records the audio data and sends it to a server in the cloud. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotion and needs indicated by the cry are predicted.

[1330] Sending and notifying feedback

[1331] The server then sends the results of the analysis of the baby or animal's cry, along with the user's emotion recognition results, to the user's device. Specifically, the server sends a message containing information about the cause of the cry and advice or suggestions based on the user's emotional state.

[1332] The device will notify the user of these results, which will be displayed as push notifications or alerts, allowing the user to take immediate action.

[1333] Specific examples

[1334] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device recognizes the user's emotions, such as "fatigue" and "stress," in real time. The server analyzes the received voice data and predicts that the user is "hungry." If the server determines that the user is feeling tired, it creates a message containing advice such as "Your baby may be hungry. If possible, ask a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[1335] This embodiment allows for quick understanding of why a baby or animal is crying and allows for a response that takes into account the parent's or owner's own emotional state, which is expected to improve the health and well-being of not only the baby or animal, but also the parent or owner themselves.

[1336] The processing flow will be explained below.

[1337] Step 1:

[1338] The server collects expert-labeled datasets of baby and animal cries and stores them in a cloud database, including audio files and labels such as "hungry," "sleepy," and "anxious."

[1339] Step 2:

[1340] The server preprocesses the collected audio data, which includes noise removal and sampling rate unification, to prepare the data for analysis.

[1341] Step 3:

[1342] The server uses the preprocessed data to train machine learning models, which are based on convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

[1343] Step 4:

[1344] The server stores the trained machine learning model in the cloud, which is then used for subsequent real-time analysis.

[1345] Step 5:

[1346] The device has the ability to record the crying sounds of babies and animals in real time. When a crying sound is detected, the device will start recording the audio data.

[1347] Step 6:

[1348] The device transmits the recorded voice data to a server via the Internet, and simultaneously detects the user's voice input and analyzes it using an emotion engine.

[1349] Step 7:

[1350] The server preprocesses the received audio data, which includes noise reduction and sampling rate unification.

[1351] Step 8:

[1352] The server uses the pre-processed audio data to analyze it with a trained machine learning model, which predicts the cause of the crying (hungry, sleepy, etc.).

[1353] Step 9:

[1354] The server integrates the results of the user's emotion recognition with the results of the crying analysis. For example, if the user feels "tired," that information is also included in the prediction results.

[1355] Step 10:

[1356] The server sends the integrated results to the user's terminal, and the message includes the cause of the crying as well as advice based on the user's condition.

[1357] Step 11:

[1358] The device notifies the user of messages received from the server, which are displayed as push notifications or alerts.

[1359] Step 12:

[1360] The user can check the notification and respond to the baby or animal's needs based on the notification. For example, if the system predicts that the baby is hungry and the user feels tired, it will suggest asking other family members for help. The user can also take action such as feeding the baby if necessary.

[1361] This series of steps enables the system to quickly identify the cause of a baby or animal's cry and provide appropriate advice that takes into account the user's emotional state.

[1362] Example 2

[1363] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1364] Conventional emotion recognition systems are limited in their ability to analyze the cries of babies and animals, and are unable to simultaneously recognize the user's emotional state and provide appropriate feedback. This makes it difficult for parents and owners to respond efficiently and quickly, often resulting in stress and fatigue. Furthermore, analysis results can sometimes be inaccurate, preventing effective countermeasures from being implemented.

[1365] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1366] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for analyzing the user's voice data in real time and recognizing the emotional state, means for recording the crying of a baby or animal in real time and transmitting the voice data to the server, means for transmitting the prediction result to a user terminal, and means for notifying the user of the received prediction result. This makes it possible to quickly and accurately understand the meaning of the crying of a baby or animal and provide appropriate feedback according to the user's emotional state.

[1367] "Means for receiving audio data" refers to the function that allows a user terminal or dedicated device to record the cries of babies or animals and send them to a server.

[1368] "Means for preprocessing received voice data" refers to a function for performing preprocessing such as noise removal and unifying the sampling rate on collected voice data.

[1369] "Means of analyzing pre-processed voice data and using machine learning models to predict emotions and needs" refers to a function that analyzes voice data using deep learning convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to predict emotions and needs.

[1370] "Means for analyzing the user's voice data in real time and recognizing their emotional state" refers to a function that analyzes the voice data spoken by the user in real time and recognizes their emotional state, such as "fatigue" or "stress."

[1371] "Means for recording the cries of babies and animals in real time and transmitting the audio data to a server" refers to a function that enables a user terminal or dedicated device to record the cries of babies and animals in real time and transmit the audio data to a cloud server.

[1372] "Means for transmitting prediction results to a user terminal" refers to a function that enables the server to transmit analysis results and feedback messages to a user terminal.

[1373] "Means for notifying the user of the received prediction results" refers to a function that allows the user's device to notify the user of the prediction results received from the server as a push notification or an alert.

[1374] MODE FOR CARRYING OUT THE INVENTION

[1375] The present invention relates to a system that analyzes the cries of babies and animals, recognizes not only their emotions and needs but also the emotions of the user, and provides feedback. This system performs a series of processes from collecting audio data to analyzing it and providing feedback. An embodiment of the system of the present invention will be described in detail below.

[1376] Data collection and preprocessing

[1377] The server collects a large amount of expert-labeled crying data sets from online repositories and stores them in a cloud database. The collected data includes audio files of crying babies and animals, along with labels such as "hungry," "sleepy," and "anxious." The data is then converted to a uniform sampling rate (e.g., 44.1 kHz) and preprocessed with noise reduction (using a Gaussian filter, etc.).

[1378] Training a machine learning model

[1379] The server then loads the preprocessed audio data into a deep learning environment (e.g., TensorFlow or PyTorch) and trains it using a convolutional neural network (CNN) or recurrent neural network (RNN). This training creates a model that learns the feature vectors of the audio data (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs with high accuracy. After training is complete, the model is stored on a cloud server and made available for real-time analysis.

[1380] User Emotion Recognition

[1381] The device is equipped with an emotion engine that analyzes emotions in real time based on the user's voice input. This engine uses, for example, IBM Watson Tone Analyzer to analyze the voice data of the user's speech (e.g., "I'm tired," "I'm irritated," etc.) and recognizes the user's emotional state. This real-time recognition is performed in a natural situation where the user is speaking to the device.

[1382] Audio recording and real-time analysis

[1383] The device (e.g., a smartphone or dedicated device) has the ability to record the cries of babies and animals in real time. When a cry is detected, the device automatically starts recording and sends the audio data to a cloud-based server. The server then preprocesses the received audio data and analyzes it using a trained machine learning model. This allows for highly accurate prediction of the emotions and needs indicated by the cry.

[1384] Sending and notifying feedback

[1385] The server combines the results of the crying analysis and the user's emotion recognition and sends them to the user's device. Specific feedback includes information about the cause of the crying and advice or suggestions based on the user's emotional state. For example, "Your baby may be hungry. If possible, ask a family member for help." The device then notifies the user of this feedback via a push notification or alert, encouraging them to take prompt action.

[1386] Specific examples

[1387] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device analyzes the user's voice and recognizes emotions such as "fatigue" and "stress" in real time. The server analyzes the received crying data and predicts that the baby is "hungry." If the server determines that the user is tired, it creates a message containing advice such as "Your baby may be hungry. If possible, try asking a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[1388] Prompt Sentence Examples

[1389] 1. "Analyze an audio file of a crying baby and predict its emotions and needs. The user's emotional state is 'Tired'."

[1390] 2. "Analyze a dog's cry and tell me its cause. The user's emotion is 'stress'."

[1391] This system can quickly and accurately identify the emotions and needs expressed by the cries of babies and animals, and provide appropriate feedback based on the user's emotional state, which is expected to improve the health and happiness of not only babies and animals, but also their parents and owners.

[1392] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1393] Step 1: Collect data

[1394] The server collects expert-labeled crying datasets from online repositories. The input includes crying audio data and corresponding emotion and need labels. The output is stored in a cloud database.

[1395] Step 2: Preprocessing the data

[1396] The server performs preprocessing on the collected audio data. The raw audio data collected in step 1 is used as input. Specific preprocessing operations include noise removal (applying a Gaussian filter) and unifying the sampling rate (e.g., changing to 44.1 kHz). The preprocessed audio data is obtained as output.

[1397] Step 3: Train the machine learning model

[1398] The server uses the preprocessed audio data to train a machine learning model in a deep learning environment (such as TensorFlow or PyTorch). The preprocessed audio data and its associated label information are used as input. Specifically, a convolutional neural network (CNN) or recurrent neural network (RNN) is used to learn feature vectors (e.g., pitch, rhythm, decibels). The output is a trained model.

[1399] Step 4: Recognizing user emotions

[1400] The device collects the user's voice data in real time and analyzes their emotions. The user's spoken voice data is used as input. Specifically, it uses an emotion engine (for example, IBM Watson Tone Analyzer) to recognize emotional states such as "fatigue" and "stress" in real time. The output is the recognition result of the user's emotional state.

[1401] Step 5: Record your audio

[1402] The device records the cries of babies and animals in real time. The cries of babies and animals are used as input. Specifically, the device's microphone is used to record the cries and save them as audio files. The recorded audio data is obtained as output.

[1403] Step 6: Real-time analysis

[1404] The device sends the recorded crying data to a server on the cloud. The audio data recorded in step 5 is used as input. The server receives this audio data, preprocesses it again, and then analyzes it using a trained machine learning model. Specific operations include preprocessing (noise removal, unifying the sampling rate) and feature extraction (pitch, rhythm, decibel analysis). The analysis results are obtained as output.

[1405] Step 7: Feedback Generation

[1406] The server generates a feedback message based on the analysis results and the user's emotion recognition results. The analysis results from step 6 and the emotion recognition results from step 4 are used as input. The specific operation is to combine information about the cause of the crying with advice or suggestions based on the user's emotional state. The output is a feedback message.

[1407] Step 8: Notification

[1408] The server sends the generated feedback message to the user device. The feedback message generated in step 7 is used as input. The device receives this message and displays it to the user as a push notification or an alert. Specific actions include playing a notification sound or displaying a message on the screen. The output is that the user can confirm the feedback message.

[1409] In this way, the system can analyze the cries of babies and animals and provide appropriate feedback that takes into account the user's emotional state.

[1410] (Application example 2)

[1411] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1412] Conventional systems for analyzing the cries of babies and animals are limited to analyzing only the emotions and needs of the cry and providing feedback based on that analysis, and lack consideration of the user's emotions and state. Furthermore, feedback after analyzing the cry is not linked to the user's emotional state, making it difficult for the user to respond appropriately and promptly. The present invention aims to solve these problems by providing more appropriate feedback by taking the user's emotional state into account, enabling a prompt response.

[1413] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1414] In this invention, the server includes means for receiving voice data, means for preprocessing the received voice data, means for analyzing the preprocessed voice data and using a machine learning model to predict emotions and needs, means for recognizing a user's voice input and analyzing the user's emotions, means for transmitting advice or suggestions based on the prediction results and the user's emotion analysis results to a user terminal, and means for notifying the user of the received prediction results or advice. This makes it possible to provide appropriate feedback that takes into account not only the analysis of the crying of a baby or animal but also the user's emotional state.

[1415] The "means for receiving audio data" refers to a device or module that has the function of capturing audio data, including the cries of babies and animals, in real time and incorporating it into the system.

[1416] The "means for preprocessing received voice data" refers to a device or module that has the function of processing collected voice data, such as noise removal and standardizing the sampling rate, and converting it into a format suitable for analysis.

[1417] "Means of using machine learning models to analyze preprocessed voice data and predict emotions and needs" refers to algorithms and systems that use deep learning techniques to analyze and predict emotions and needs from preprocessed voice data.

[1418] "Means for recognizing user voice input and analyzing the user's emotions" refers to a device or module that has the function of analyzing the user's voice in real time and determining the user's emotional state, such as "fatigue," "stress," or "relief."

[1419] "Means for sending advice and suggestions based on prediction results and user emotion analysis results to a user terminal" refers to a device or system that has the function of generating advice and suggestions that take into account the user's emotional state based on the analysis results and sending them to the user's smartphone or dedicated device.

[1420] The "means for notifying the user of the received prediction results and advice" refers to a device or application that has a notification function for appropriately notifying the user of the analysis results and advice sent to the terminal.

[1421] Data collection and preprocessing

[1422] The server collects a large amount of crying data labeled by experts and stores it in a cloud database. The data set includes audio files of crying babies and animals, along with associated labels such as "hungry," "sleepy," and "anxious." The collected data is first preprocessed, including noise removal and standardization of the sampling rate, to obtain data suitable for analysis.

[1423] Training a machine learning model

[1424] The server uses the preprocessed data to train a deep learning convolutional neural network (CNN) or recurrent neural network (RNN). The model learns the audio characteristics of crying (e.g., pitch, rhythm, decibels, etc.) and predicts emotions and needs based on them. The resulting model is stored on a cloud server and used for real-time analysis.

[1425] User Emotion Recognition

[1426] The device is equipped with an emotion engine that recognizes the user's emotions based on their voice input. This engine analyzes the user's speech data in real time and recognizes emotional states such as "fatigue," "stress," and "relief." This is done automatically when the user speaks to the device.

[1427] Audio recording and real-time analysis

[1428] The device (e.g., a user's smartphone or a dedicated device) is equipped with the ability to record the crying of a baby or animal in real time. When a cry is detected, the device records the audio data and sends it to a server in the cloud. The server then preprocesses the received audio data again and analyzes it using a trained machine learning model. As a result of the analysis, the emotion and needs indicated by the cry are predicted.

[1429] Sending and notifying feedback

[1430] The server sends the results of analyzing the baby or animal's cry, along with the user's emotion recognition results, to the user's device. Specifically, information about the cause of the cry and advice or suggestions based on the user's emotional state are sent as a message. The device then notifies the user of these results. The notifications are displayed as push notifications or alerts, helping the user respond immediately.

[1431] Specific examples

[1432] For example, suppose a baby starts crying in the middle of the night. In this case, the device records the baby's crying and sends the data to the server. At the same time, the device recognizes the user's emotions, such as "fatigue" or "stress," in real time. The server analyzes the received voice data and predicts that the user is "hungry." If the server determines that the user is feeling tired, it creates a message containing advice such as, "Your baby may be hungry. If possible, ask a family member for help." This message is sent to the device, which then notifies the user. The user can check the notification and take the necessary action promptly.

[1433] Prompt Sentence Examples

[1434] We will imagine a situation where a baby or animal starts crying in the middle of the night, record an audio file of the crying, and analyze it. We will also record the user's voice and analyze their emotions. Based on these two results, we will design a system that notifies the user of appropriate advice.

[1435] This embodiment allows for quick understanding of why a baby or animal is crying and allows for a response that takes into account the parent's or owner's own emotional state, which is expected to improve the health and well-being of not only the baby or animal, but also the parent or owner themselves.

[1436] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1437] Step 1:

[1438] The device records the cries of babies and animals in real time. The input is environmental audio data, and the output is a recorded audio data file. Specifically, the device's microphone captures the audio and saves it as a digital audio file.

[1439] Step 2:

[1440] The terminal sends the recorded audio data file to the server. The input is the audio data file recorded in step 1, and the output is the audio data sent to the server. Specifically, the audio data file is uploaded to the server via an Internet connection.

[1441] Step 3:

[1442] The server preprocesses the received audio data. The input is the audio data sent to the server, and the output is the preprocessed audio data. Specifically, the server removes noise, unifies the sampling rate, and converts the data into a format suitable for analysis.

[1443] Step 4:

[1444] The server analyzes the preprocessed audio data using a machine learning model. The input is the preprocessed audio data, and the output is the predicted emotion and needs of the crying voice. Specifically, it uses a convolutional neural network (CNN) and a recurrent neural network (RNN) to analyze the audio features and classify the emotion and needs.

[1445] Step 5:

[1446] The device recognizes the user's voice input in real time and analyzes the user's emotions. The input is the voice data spoken by the user, and the output is the analysis result of the user's emotional state. Specifically, the emotion engine installed in the device analyzes the voice and determines the user's emotional state, such as "fatigue," "stress," or "relief."

[1447] Step 6:

[1448] The server integrates the prediction results and the user's emotion analysis results based on the analysis results to generate advice and suggestions. The input is the prediction results of the crying emotion and needs and the analysis results of the user's emotional state, and the output is a message containing advice and suggestions. Specifically, the server uses a generative AI model to create appropriate advice and suggestions.

[1449] Step 7:

[1450] The server sends the generated advice and suggestion messages to the device. The input is the message containing the advice or suggestion, and the output is the message sent to the device. Specifically, the message is delivered to the device via the cloud server.

[1451] Step 8:

[1452] The device notifies the user of the received advice or suggestion message. The input is the advice or suggestion message sent to the device, and the output is the notification displayed to the user. Specifically, the device displays the message to the user using a push notification or a pop-up alert.

[1453] Prompt Sentence Examples

[1454] We will imagine a situation where a baby or animal starts crying in the middle of the night, record an audio file of the crying, and analyze it. We will also record the user's voice and analyze their emotions. Based on these two results, we will design a system that notifies the user of appropriate advice.

[1455] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1456] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1457] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1458] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1459] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1460] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1461] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1462] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1463] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1464] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1465] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1466] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1467] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1468] 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.

[1469] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1470] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1471] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1472] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1473] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1474] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1475] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1476] The following is further disclosed regarding the above embodiment.

[1477] (Claim 1)

[1478] means for receiving audio data;

[1479] means for preprocessing received audio data;

[1480] a means for analyzing the pre-processed voice data and using machine learning models to predict emotions and needs;

[1481] means for transmitting the prediction result to a user terminal;

[1482] means for notifying a user of the received prediction result;

[1483] A system including:

[1484] (Claim 2)

[1485] 10. The system of claim 1, further comprising means for preprocessing the audio data based on noise reduction and sampling rate unification.

[1486] (Claim 3)

[1487] The system according to claim 1, characterized in that the machine learning model used for the analysis is composed of a convolutional neural network and a recurrent neural network.

[1488] "Example 1"

[1489] (Claim 1)

[1490] means for receiving audio data;

[1491] means for preprocessing the received audio data based on noise reduction and sampling rate unification;

[1492] A means for analyzing the preprocessed voice data and using a machine learning model consisting of a convolutional neural network and a recurrent neural network to predict emotions and needs;

[1493] means for transmitting the prediction result to a user terminal;

[1494] a means for sending a push notification of the received prediction result to a user;

[1495] A system including:

[1496] (Claim 2)

[1497] 10. The system of claim 1, further comprising: means for splitting data into a training data set and a test data set; and means for storing the trained model on a cloud server.

[1498] (Claim 3)

[1499] 2. The system according to claim 1, further comprising means for recording voice data from a user terminal in real time and transmitting the voice data to a cloud server via the Internet.

[1500] "Application Example 1"

[1501] (Claim 1)

[1502] means for receiving audio data;

[1503] means for preprocessing received audio data;

[1504] a means for analyzing the pre-processed voice data and using machine learning models to predict emotions and needs;

[1505] means for transmitting the prediction result to a user terminal;

[1506] means for notifying a user of the received prediction result;

[1507] A means for transmitting the received voice data to a cloud server and acquiring an analysis result from the cloud server;

[1508] A means for generating notification content based on the analysis results and instructing the user to take appropriate action;

[1509] A system including:

[1510] (Claim 2)

[1511] 10. The system of claim 1, further comprising means for preprocessing the audio data based on noise reduction and sampling rate unification.

[1512] (Claim 3)

[1513] The system according to claim 1, characterized in that the machine learning model used for the analysis is composed of a convolutional neural network and a recurrent neural network.

[1514] "Example 2: Combining Emotion Engines"

[1515] (Claim 1)

[1516] means for receiving audio data;

[1517] means for preprocessing received audio data;

[1518] a means for analyzing the pre-processed voice data and using machine learning models to predict emotions and needs;

[1519] means for analyzing the user's voice data in real time to recognize the user's emotional state;

[1520] A means to record the cries of babies and animals in real time and send the audio data to a server.

[1521] means for transmitting the prediction result to a user terminal;

[1522] means for notifying a user of the received prediction result;

[1523] A system including:

[1524] (Claim 2)

[1525] 10. The system of claim 1, further comprising means for preprocessing the audio data based on noise reduction and sampling rate unification.

[1526] (Claim 3)

[1527] The system according to claim 1, characterized in that the machine learning model used for the analysis is composed of a convolutional neural network and a recurrent neural network.

[1528] "Application example 2 when combining emotion engines"

[1529] (Claim 1)

[1530] means for receiving audio data;

[1531] means for preprocessing received audio data;

[1532] a means for analyzing the pre-processed voice data and using machine learning models to predict emotions and needs;

[1533] means for recognizing a user's voice input and analyzing the user's emotions;

[1534] means for transmitting advice and suggestions based on the prediction results and the user's emotion analysis results to the user terminal;

[1535] a means for notifying the user of the received prediction results and advice;

[1536] A system including:

[1537] (Claim 2)

[1538] 10. The system of claim 1, further comprising means for preprocessing the audio data based on noise reduction and sampling rate unification.

[1539] (Claim 3)

[1540] The system according to claim 1, characterized in that the machine learning model used for the analysis is composed of a convolutional neural network and a recurrent neural network. [Explanation of symbols]

[1541] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving audio data; means for preprocessing received audio data; a means for analyzing the pre-processed voice data and using machine learning models to predict emotions and needs; means for transmitting the prediction result to a user terminal; means for notifying a user of the received prediction result; A system including:

2. 2. The system of claim 1, further comprising means for preprocessing the audio data based on noise reduction and sampling rate unification.

3. The system according to claim 1, wherein the machine learning model used for the analysis is composed of a convolutional neural network and a recurrent neural network.

Images