Information processing system, information processing method, and program

Abstract

One aspect of the present invention provides an information processing system comprising at least one processor. The processor is configured to execute the following steps by reading a program: a first acquisition step in which channel state information transmitted from a wireless communication device is acquired; a specification step in which the channel state information is input to a first processing model, and the first processing model outputs action information relating to an action of a subject; and an evaluation step in which the action information is input to a second processing model, and the second processing model outputs an evaluation relating to the action of the subject corresponding to the action information. The first processing model is a trained model that has been trained so as to be able to accept channel state information as an input and output action information of a subject corresponding to the channel state information, and the second processing model is a processing model that can accept the input of action information and output an evaluation corresponding to the action information.

Description

Information Processing System, Information Processing Method, and Program

[0001] The present invention relates to an information processing system, an information processing method, and a program.

[0002] Patent Document 1 is a document related to a method executed in a wireless LAN (Wireless Local Area Network) system. In this Patent Document 1, when sensing measurements started by a non-access point station (STA) that is not an access point (AP) are executed, a new signal transmission / reception procedure between STAs is proposed. In Patent Document 1, a sensing measurement procedure executed by an AP by a non-AP STA transmitting a sensing start frame to the AP is proposed. In Patent Document 1, a procedure in which an AP requests a responder STA to transmit an NDP frame by a non-AP STA transmitting a sensing start frame to the AP is proposed. In Patent Document 1, a method for configuring frames transmitted and received in the above procedure is proposed.

[0003] Japanese Patent Translation Publication No. 2024-500889

[0004] However, there is still room for improvement in the technology of sensing using a wireless communication system.

[0005] According to one aspect of the present invention, there is provided an information processing system including at least one processor, the processor being configured to execute the following steps by reading a program. In a first acquisition step, channel state information transmitted from a wireless communication device is acquired. In a specific step, the channel state information is input into a first processing model, and in an evaluation step, operation information regarding the operation of a subject is output from the first processing model. Then, in the evaluation step, the operation information is input into a second processing model, and an evaluation regarding the operation of the subject corresponding to the operation information is output from the second processing model. Here, the first processing model is a learned model that has been learned to be able to input channel state information and output operation information of a subject corresponding to the channel state information, and the second processing model is a processing model that can input operation information and output an evaluation corresponding to the operation information.

[0006] According to one aspect of the present invention, it is possible to provide an information processing system, etc., that can provide more useful sensing using a wireless communication system.

[0007] This is a diagram showing the configuration of the information processing system 1. This is a block diagram showing the hardware configuration of the server 2. This is a block diagram showing the hardware configuration of the information processing device 3. This is a block diagram showing the hardware configuration of the wireless communication device 4. This is a flowchart showing an overview of the processing performed by the information processing system 1. This shows an example of a Wi-Fi signal when the door between room A and room B is closed. This shows an example of a Wi-Fi signal when the door between room A and room B is open. This is a floor plan 9 showing an example of the installation location of the wireless communication device 4. This is an activity diagram showing an example of the flow of processing performed by the information processing system 1. This is a diagram showing an example of the flow of information by the information processing system 1. This is a diagram showing an example of the relationship between identified small action information and recognized actions.

[0008] Embodiments of the present invention will be described below with reference to the drawings. The various features shown in the embodiments below can be combined with each other.

[0009] Incidentally, the program for realizing the software appearing in one embodiment may be provided as a computer-readable non-transitor-readable medium, or it may be provided so that it can be downloaded from an external server, or it may be provided so that the program is launched on an external computer and its functions are realized on a client terminal (so-called cloud computing).

[0010] Furthermore, in various information processing according to one embodiment, an input and an output corresponding to the input can be realized. Here, as long as an output is obtained as a result of the input, the form of the information referenced in such information processing (hereinafter referred to as "reference information") is not limited. The reference information may be, for example, rule-based information such as a database, a lookup table, or a predetermined function (including a decision formula such as a regression equation constructed by a statistical method), or a pre-trained model that has learned the correlation between input and output in advance, or a large-scale language model that can output a desired result by inputting a prompt.

[0011] Furthermore, in one embodiment, "part" may include, for example, hardware resources implemented by a circuit in a broad sense, and the information processing of software that can be specifically realized by these hardware resources. Also, in one embodiment, various types of information are handled, and this information can be represented, for example, by the physical values ​​of signal values ​​representing voltage and current, the high or low values ​​of signal values ​​as a set of binary bits composed of 0s or 1s, or by quantum superposition (so-called qubits), and communication and calculations can be performed on a circuit in a broad sense.

[0012] Furthermore, a circuit in a broad sense is a circuit realized by combining at least an appropriate combination of circuits, circuits, processors, and memory. The processor may be a general-purpose processor or a dedicated circuit. In other words, this includes application-specific integrated circuits (ASICs), programmable logic devices (for example, simple programmable logic devices (SPLDs), complex programmable logic devices (CPLDs), and field programmable gate arrays (FPGAs)), etc.

[0013] 1. Hardware Configuration This section describes the hardware configuration.

[0014] <Information Processing System 1> Figure 1 is a configuration diagram representing information processing system 1. Information processing system 1 comprises a server 2, an information processing device 3, and a wireless communication device 4. The server 2, the information processing device 3, and the wireless communication device 4 are configured to communicate with each other via a telecommunications line (network). The wireless communication device 4 may comprise a wireless communication device 4a and a wireless communication device 4b. The wireless communication device 4a is configured to communicate with the server 2 and the information processing device 3 via a telecommunications line (network), and the wireless communication device 4b may be configured to communicate with the wireless communication device 4a. Multiple wireless communication devices 4b may be provided in relation to the wireless communication device 4a.

[0015] Here, the system exemplified in Information Processing System 1 consists of one or more devices or components. Therefore, it should be noted that Information Processing System 1 includes either Server 2 alone, or Server 2, Information Processing Device 3, and Wireless Communication Device 4. Information Processing Device 3 may also have the functionality of Wireless Communication Device 4. More specifically, Information Processing System 1 may include elements selected from the group consisting of Server 2, Information Processing Device 3, and Wireless Communication Device 4. Unselected elements may not be included in Information Processing System 1, but may be electrically connected to the selected elements as external elements. These components will be described below.

[0016] <Server 2> Figure 2 is a block diagram showing the hardware configuration of Server 2. Server 2 comprises a communication unit 21, a storage unit 22, and a control unit 23, and these components are electrically connected within Server 2 via a communication bus 20. Each component will be described further.

[0017] The communication unit 21 preferably uses wired communication methods such as USB, IEEE 1394, Thunderbolt®, and wired LAN network communication, but may also include wireless LAN network communication, mobile communication such as 3G / LTE / 5G, Bluetooth® communication, Zigbee, THREAD®, etc., as needed. In other words, it is more preferable to implement it as a collection of these multiple communication methods. That is, the server 2 may communicate various information from the outside via the communication unit 21 and the network.

[0018] The storage unit 22 stores various types of information as defined above. This can be implemented, for example, as a storage device such as a solid-state drive (SSD) that stores various programs related to the server 2 executed by the control unit 23, or as memory such as random access memory (RAM) that stores temporarily necessary information (arguments, arrays, etc.) related to program calculations. The storage unit 22 stores various programs and variables related to the server 2 executed by the control unit 23.

[0019] The control unit 23 performs processing and control of the overall operation related to the server 2. The control unit 23 is, for example, a Central Processing Unit (CPU) (not shown). The control unit 23 realizes various functions related to the server 2 by reading predetermined programs stored in the memory unit 22. That is, information processing by software stored in the memory unit 22 is concretely realized by the control unit 23, which is an example of hardware, so that each step related to each function described later can be executed. These will be described in more detail in the next section. Note that the control unit 23 is not limited to being a single unit, and may be implemented with multiple control units 23 for each function, or a combination thereof.

[0020] <Information Processing Device 3> Figure 3 is a block diagram showing the hardware configuration of the information processing device 3. The information processing device 3 comprises a communication unit 31, a storage unit 32, a control unit 33, a display unit 34, an input unit 35, and an output unit 36, and these components are electrically connected within the information processing device 3 via a communication bus 30. Each component will be described further. The descriptions of the communication unit 31, the storage unit 32, and the control unit 33 are the same as the descriptions of each part in the server 2, so they will be omitted.

[0021] The communication unit 31 can receive channel status information from the wireless communication device 4. Specifically, the communication unit 31 can receive channel status information from the wireless communication device via wireless LAN network communication, mobile communication such as 3G / LTE / 5G, Bluetooth® communication, Zigbee, THREAD®, and other wireless communication methods. The communication unit 31 can also communicate with the wireless communication device 4 via wired communication methods. Furthermore, the communication unit 31 can communicate with the server 2, etc., via wired communication methods such as USB, IEEE 1394, Thunderbolt®, and wired LAN network communication, or via wireless communication methods such as wireless LAN network communication, mobile communication such as 3G / LTE / 5G, Bluetooth® communication, Zigbee, THREAD®, and other wireless communication methods.

[0022] The display unit 34 may be included in the housing of the information processing device 3 or it may be an external component. The display unit 34 displays a graphical user interface (GUI) screen that can be operated by the user. Preferably, this is done by using different display devices such as a CRT display, liquid crystal display, organic EL display, and plasma display, depending on the type of information processing device 3.

[0023] The input unit 35 may be included in the housing of the information processing device 3 or it may be externally attached. The input unit 35 may be composed of, for example, a sound collector such as a microphone, which collects external sounds and outputs an audio signal indicating the collected sound. The audio signal is transferred as a command signal to the control unit 33 via the communication bus 30, and the control unit 33 can perform predetermined controls and calculations as needed. Furthermore, the information that the input unit 35 accepts is not limited to the above-mentioned voice, etc. Specifically, the input unit 35 may be configured to accept user input via a touch panel, switch buttons, mouse, QWERTY keyboard, etc., in conjunction with the display unit 34.

[0024] The output unit 36 ​​may be included in the housing of the information processing device 3 or it may be external. For example, the output unit 36 ​​may be composed of a speaker or the like and output sound or signal sound generated by the information processing system 1. Alternatively, for example, the output unit 36 ​​may be composed of a light-emitting means and output light based on information generated by the information processing system 1. For example, the output unit 36 ​​may output information generated by the information processing system 1 from an external speaker or light-emitting means. Alternatively, for example, the output unit 36 ​​may cooperate with a voice assistant-compatible smart home appliance and output information generated by the information processing system 1 from the smart home appliance as voice or the like.

[0025] As the information processing device 3, a smartphone, tablet terminal, personal computer, wearable device, IoT device, etc., can be used. An IoT device is a device that can communicate via a network.

[0026] <Wireless Communication Device 4> Figure 4 is a block diagram showing the hardware configuration of wireless communication device 4. Wireless communication device 4 comprises a communication unit 41, a storage unit 42, a control unit 43, an input unit 44, and an output unit 45, and these components are electrically connected within the wireless communication device 4 via a communication bus 40. Each component will be described further. The communication unit 41, storage unit 42, and control unit 43 will be described only in parts that differ from the descriptions of each part in server 2. The descriptions of input unit 44 and output unit 45 are the same as the descriptions of each part in information processing device 3 and will therefore be omitted.

[0027] The communication unit 41 can send and receive various types of information via the network. The wireless communication device 4 can communicate with the server 2 and the information processing device 3 using wired communication means such as USB, IEEE 1394, Thunderbolt (registered trademark), wired LAN network communication, or wireless communication means such as wireless LAN network communication, 3G / LTE / 5G mobile communication, Bluetooth (registered trademark) communication, Zigbee, and THREAD (registered trademark). If there are multiple wireless communication devices 4, the multiple wireless communication devices 4 may communicate with each other. For communication between multiple wireless communication devices 4, short-range wireless communication such as RFID and NFC may be used.

[0028] The communication unit 41 can send and receive information using wireless communication technology. The communication unit 41 includes, for example, a wireless LAN module. A wireless LAN is a type of local area network that provides network connectivity between devices within a certain range. Wireless LANs do not require physical wiring and can send and receive information via wireless communication. It is preferable that the communication unit 41 uses a wireless LAN module, which is a wireless communication technology. A wireless LAN module is an integrated hardware component for providing wireless LAN functionality. The wireless LAN module controls the sending and receiving of wireless signals and performs processing in accordance with the Wi-Fi standard (IEEE 802.11) to communicate with the network. The wireless LAN module has a built-in transceiver and operates in conjunction with the antenna.

[0029] A transceiver is configured to transmit and receive radio signals. It comprises a transmitter that transmits radio signals and a receiver that receives reflected radio signals with high sensitivity. The transmitter can control the intensity and pattern of the transmitted signal. The transmitter should ideally be able to transmit a stable signal in a specific frequency band. The receiver may include a high-precision analog-to-digital converter for detailed analysis of changes in the received signal. That is, the transmitter converts digital data into a radio signal and transmits it from the antenna, and the receiver converts the received radio signal back into digital data.

[0030] An antenna is a component that transmits and receives wireless signals. An omnidirectional antenna capable of receiving signals equally from all directions may be used as the antenna. A wireless LAN chip may use either an external antenna or an internal antenna.

[0031] The wireless LAN module should preferably be a MIMO (Multiple Input Multiple Output) compatible module. By using multiple antennas, it is possible to receive changes in amplitude and phase of wireless LAN signals that have passed through multiple different paths and analyze the detailed channel state. By using multiple antennas, it is also possible to understand the effects of signal reflection and obstacles. By using multiple antennas, it is possible to send and receive signals from different angles and positions. By using multiple antennas, it is possible to acquire channel state information such as signal phase, amplitude, and delay. It is even more preferable to use a Wi-Fi module based on the IEEE 802.11bf standard for wireless LAN. IEEE 802.11bf is a standard for sensing using Wi-Fi. IEEE 802.11bf defines a technology for detecting environmental changes and movements using Wi-Fi signals.

[0032] The communication unit 41 may also have a function for building a mesh network with the information processing device 3 and other wireless communication devices 4. Specifically, the communication unit 41 includes a mesh network compatible module and transfers information received from the information processing device 3 and other wireless communication devices 4 to another device. The mesh network compatible module is, for example, a communication module compatible with a Wi-Fi i mesh network, a communication module compatible with the Zigbee protocol, or a communication module compatible with the Thread protocol. The communication unit 41 may also include a dual-protocol communication module that supports both wireless LAN modules and mesh network compatible modules.

[0033] The storage unit 42 stores various types of information as defined above. This can be implemented, for example, as a storage device such as a solid-state drive (SSD) that stores various programs related to the wireless communication device 4 executed by the control unit 43, or as a memory such as a random access memory (RAM) that stores temporarily necessary information (arguments, arrays, etc.) related to program calculations. The storage unit 42 stores various programs and variables related to the wireless communication device 4 executed by the control unit 43. The storage unit 42 can also store signals received via the communication unit 41 (i.e., received signals) and signals transmitted via the communication unit 41 (i.e., transmitted signals). Furthermore, the storage unit 42 can temporarily or permanently store channel status information of wireless LAN signals and data within the mesh network.

[0034] The control unit 43 performs processing and control of the overall operation related to the wireless communication device 4. For example, the control unit 43 can receive signals via the communication unit 41, process the received signals, generate transmission signals, and execute control for signal transmission. The control unit 43 may perform data framing, error checking, retransmission control, encryption, etc., in accordance with the 802.11 standard. The control unit 43 is, for example, a digital signal processor (DSP) for analyzing changes in amplitude, phase, etc., of the received signal in real time. The DSP processes channel status information (CSI) and received signal strength index (RSSI), etc. The control unit 43 is, for example, a central processing unit (CPU). The CPU further analyzes the data processed by the DSP. The control unit 43 realizes various functions related to the wireless communication device 4 by reading predetermined programs stored in the storage unit 42. In other words, the information processing performed by the software stored in the memory unit 42 is concretely realized by the control unit 43, which is an example of hardware, to manage data collection from sensors, data preprocessing, storage to local storage, transmission to the cloud, etc.

[0035] In the information processing system 1, there may be only one wireless communication device 4, but it is preferable to use multiple wireless communication devices 4. When using multiple wireless communication devices 4, the information processing system 1 may include multiple wireless communication devices 4a and multiple wireless communication devices 4b.

[0036] The wireless communication device 4a may be the master unit and the wireless communication device 4b may be the slave unit. The wireless communication device 4a, as the master unit, may be configured to communicate with the server 2 and the information processing device 3 via a telecommunications line (network), and the wireless communication device 4b, as the slave unit, may be configured to communicate with the wireless communication device 4a.

[0037] The wireless communication device 4 can be any device capable of communication, such as a router, smartphone, smart speaker, wearable device, or IoT device. IoT devices include smart home appliances, smart lighting, smart doors, smart locks, smart sensors, smart cameras, and smart plugs. Smart home appliances are refrigerators, washing machines, air conditioners, etc., that are capable of communication. The wireless communication device 4 may also be equipped with motion sensors, temperature sensors, humidity sensors, infrared sensors, weight sensors, etc.

[0038] Using multiple wireless communication devices 4 offers numerous advantages, including improved spatial resolution, enhanced measurement accuracy, wider coverage, increased redundancy and reliability, and data collection from different perspectives. When multiple devices are positioned in different locations, the environment can be observed from more viewpoints, improving spatial resolution. This allows for more precise detection of object position and movement. Integrating measurement results from different devices averages out errors, resulting in more reliable data. Using multiple devices expands the coverage range, enabling sensing of larger areas. Having multiple wireless communication devices 4 improves the overall system reliability because even if one device fails, other devices can continue to provide data. Furthermore, increased data redundancy improves the accuracy of sensing results. Devices positioned in different locations can collect data from different angles and directions, providing more comprehensive information about the environment and objects.

[0039] It is preferable to use the 2.4 GHz, 5 GHz, and 6 GHz bands for wireless communication. This configuration allows wireless signals to penetrate obstacles, thus reducing the risk of blind spots, which can be a problem when using cameras or infrared sensors.

[0040] 2. Functional Configuration This section describes the functional configuration of this embodiment. The control unit 23 or control unit 43 is configured to perform, for example, the following steps. The following steps are optional.

[0041] The control unit 23 is configured to be able to acquire information from the wireless communication device 4 or other devices as a first acquisition step. Also, the control unit 23 can be configured to acquire various information by reading out various information stored in a storage area that is at least part of the storage unit 22 and writing the read information into a work area that is at least part of the storage unit 22. The storage area is, for example, an area implemented as a storage device such as an SSD in the storage unit 22. The work area is, for example, an area implemented as a memory such as a RAM. The control unit 23 is configured to be able to acquire channel state information transmitted from the wireless communication device 4 as a first acquisition step. Further, the control unit 23 may be configured to be able to acquire the converted channel state information from the wireless communication device 4 as a first acquisition step.

[0042] The control unit 23 is configured to be able to acquire information from the information processing device 3 or other devices as a second acquisition step. The control unit 23 may be configured to be able to acquire various information regarding the subject as a second acquisition step. It may be configured to be able to acquire the health information of the subject.

[0043] The control unit 23 specifies operation information based on the channel state information acquired in the first acquisition step as a specific step. The control unit 23 may be configured to be able to specify operation information by inputting it into the first processing model and causing the first processing model to output operation information regarding the operation of the subject. As a specific step, the control unit 23 can numerically convert the change in radio waves calculated from the channel state information into a specific movement, organize the numerically converted specific movement, and specify the operation information.

[0044] The first processing model is a learned model that has been learned to be able to take the channel state information as an input and output the operation information of the subject corresponding to the channel state information. The first processing model may be a learning model that has learned a combination of learning channel state information and learning operation information associated with the learning channel state information as teacher data.

[0045] The first processing model is preferably a processing model that excels at processing time series data with long-term dependencies. Examples of processing models that excel at processing time series data with long-term dependencies include LSTM (Long Short-Term Memory), GRU (Gated Recurrent Unit), Transformer, and Bidirectional RNN. The first processing model is preferably a processing model that performs calculations by combining an LSTM layer and an attention layer.

[0046] The LSTM layer is a type of recurrent neural network (RNN) designed to learn long-term dependencies. The LSTM layer has a gate mechanism (input gate, forget gate, output gate) that allows it to appropriately retain or forget past information. The attention layer learns the importance of each element in the sequence and processes the important information with emphasis. Attention dynamically calculates which parts of the information are important at each point in time and reflects this in its output. This allows for focusing on important parts even in long sequences. For example, when channel state information is input to the first processing model, it is passed to the LSTM layer, which learns the long-term dependencies of the channel state information. The output of the LSTM is input to the next attention layer as a sequence of hidden states of the channel state information. The attention layer focuses on specific parts of the sequence data and generates a context vector by taking a weighted average based on the importance at each point in time. The context vector is input to a fully connected layer, and behavior information may be output as the classification result.

[0047] The control unit 23 inputs the channel state information into, for example, a "sitting" motion identification model, a "standing up" motion identification model, and a "standing" motion identification model. The "sitting" motion identification model is configured to be able to identify "sitting" motion information, the "standing up" motion identification model is configured to be able to identify "standing up" motion information, and the "standing" motion identification model is configured to be able to identify "standing" motion information. Thus, it is preferable that the motion recognition model uses a learning model that has learned, as teacher data, a combination of learning channel state information and motion information for each learning motion associated with the learning channel state information.

[0048] As a specific step, the control unit 23 may input the channel state information into each of a plurality of small motion identification models and output small motion information from each of the plurality of small motion identification models. The first processing model may include a plurality of small motion identification models. The plurality of small motion identification models are learned models that take the channel state information as an input and can output small motion information corresponding to the channel state information, and the small motion information may be information regarding a part of the motion of the subject.

[0049] As a specific step, the control unit 23 may input the plurality of small motion information output from the plurality of small motion identification models into the motion recognition model and output motion information. The first processing model may include a plurality of small motion identification models and the motion recognition model. When the small motion information, which is a numerically converted specific motion, is input, the motion recognition model can estimate a complex motion by combining the small motion information and output the motion information. The motion recognition model may estimate the meaning of the complex motion by means of the transition and comparison of the front and back data.

[0050] As an evaluation step, the control unit 23 inputs the motion information into the second processing model and outputs an evaluation regarding the motion of the subject corresponding to the motion information from the second processing model. As an evaluation step, the control unit 23 may also input the plurality of small motion information into the second processing model and output an evaluation. When the first processing model includes a plurality of small motion identification models, as an evaluation step, the control unit 23 may input the plurality of small motion information output from the plurality of small motion identification models into the second processing model and output an evaluation.

[0051] The second processing model is a processing model that takes motion information as input and outputs an evaluation corresponding to that motion information. In the second processing model, for example, the activity level of the subject can be calculated based on motion information, and the progression of dementia or the level of care required can be predicted from the activity level. Furthermore, in the second processing model, for example, the progression of dementia or the level of care required can be predicted from the movement patterns based on motion information.

[0052] The second processing model may be a processing model capable of outputting an evaluation of the subject's physical condition corresponding to the motion information.

[0053] The second processing model may be a processing model capable of outputting an evaluation of the subject's physical information corresponding to behavioral information associated with a predetermined disease. If the second processing model is a model that outputs an evaluation of the subject's physical information corresponding to behavioral information associated with a specific disease, the evaluation includes the detection of precursors related to the predetermined disease.

[0054] The control unit 23 is configured to generate various types of information as a generation step. The control unit 23 can generate suggestions regarding actions recommended for the subject based on the evaluation as a generation step. In generating suggestions regarding actions recommended for the subject based on the evaluation as a generation step, the control unit 23 may refer to a prediction system based on medical knowledge. The prediction system may store information on the relationship between movement information and evaluation results based on movement information and dementia or the level of care required, or it may store information on movement information and evaluation results based on movement information and movements related to disease.

[0055] The control unit 23 may generate proposals using an artificial intelligence module as a generation step. The artificial intelligence module is configured to receive input from the control unit 23 and return the instructed output. The artificial intelligence module has a trained model constructed by a learning method such as supervised learning, unsupervised learning, or self-supervised learning. Supervised learning performs machine learning using training data. Training data consists of pairs of input data and output data (correct answer data) for learning. Furthermore, the language model may not only be one trained for a specific task, but also a general-purpose model that can be used universally for a wide range of tasks.

[0056] The artificial intelligence module may be an AI (Artificial Intelligence) equipped with transformers such as GPT (Generative Pretrained Transformer, including GPT-1, GPT-2, GPT-3, and GPT-4), BERT (Bidirectional Encoder Representations from Transformers), BART (Bidirectional and Auto-regressive Transformer), and language models such as recurrent neural networks (RNN).

[0057] The artificial intelligence module includes a general-purpose natural language processing learning model, such as a Large Language Model (LLM), which has learned from a vast amount of data. Such a general-purpose learning model includes language models that can handle various tasks without fine-tuning using methods such as One-shot Learning and Few-shot Learning. Furthermore, the general-purpose learning model can also handle various tasks using Zero-shot Learning. The artificial intelligence used in the control unit 23 may be separate learning models or a common general-purpose learning model.

[0058] More typically, an artificial intelligence module may have a large-scale language model. That is, the artificial intelligence module may generate proposals based on evaluation using the large-scale language model. A large-scale language model is a deep learning model that pre-trains on a vast amount of data about a language model that models human spoken language based on its occurrence probability. However, the models that an artificial intelligence module may have are not limited to the above.

[0059] The control unit 23 presents various information as a presentation step. This information can be presented to the user via the display unit 34 or other devices. For example, the control unit 23 is configured to present a suggestion as a presentation step. The control unit 23 may also cause the suggestion to be displayed on the information processing device 3 as a presentation step. In such a case, for example, the control unit 23 may control the display unit 34 to display visual information such as a screen, still images or moving images, icons, messages, etc. The control unit 23 may also present the suggestion by voice as a presentation step.

[0060] The control unit 23 manages various types of information as part of its management steps. The control unit 23 may be configured to manage acquired channel status information as part of its management steps. The control unit 23 may be configured to manage acquired operation information as part of its management steps. The control unit 23 may be configured to manage acquired subject information as part of its management steps. The control unit 23 may be configured to manage acquired evaluations as part of its management steps. The control unit 23 may manage operation information in association with evaluations as part of its management steps. In addition, some of the management steps may be performed by the control unit 33 or the control unit 43.

[0061] As a display control step, the control unit 23 controls the display unit 34 to display visual information such as images including screens, still images, or moving images, icons, and messages. The control unit 23 may generate only rendering information for displaying visual information on the display unit 34. As a display control step, the control unit 23 may display the visual information in a manner that is visible to the user. As a display control step, the control unit 23 may display the progress of the evaluation in a manner that is visible to the user. As a display control step, the control unit 23 may display visual information generated based on operation information in a manner that is visible.

[0062] The control unit 43 is configured to convert channel state information under predetermined conditions as a conversion step. The control unit 43 can perform data compression, dimensionality reduction, extraction using an amplified autocorrelation function, sampling, etc., on the channel state information as a conversion step. Note that some of the processing in the conversion step may be configured to be performed by the control unit 23.

[0063] The control unit 43 may compress the channel status information data as a conversion step. By compressing the data, data redundancy can be reduced and the data size can be reduced. The control unit 43 may also compress the data by deleting some of the data information as a conversion step.

[0064] The control unit 43 may perform dimensionality reduction of channel state information as a transformation step. By reducing the number of data features, the data size can be reduced. The control unit 43 may reduce the dimensionality of the principal components by linearly transforming the data in a direction that maximizes the data variance as a transformation step. Alternatively, the control unit 43 may perform data dimensionality reduction by selecting only the important features from the original data as a transformation step. Furthermore, the control unit 43 may retain only the important information in the process of encoding (compressing) and decoding (reconstructing) the data by learning a low-dimensional representation of the data using a neural network as a transformation step.

[0065] The control unit 43 may, as a conversion step, extract the characteristics of dynamic objects from the channel state information using an amplified autocorrelation function (A-ACF). An autocorrelation function is a function that measures the correlation between different points in time for a given time-series dataset. This allows evaluation of how closely the data is related over time. An amplified autocorrelation function is a modified autocorrelation function that emphasizes specific dynamic characteristics and periodicity, and can be used to highlight and detect the characteristics of specific dynamic objects. In particular, in the analysis of dynamic objects, an amplified autocorrelation function can highlight information related to their movement and changes. The control unit 43 may, as a conversion step, extract the characteristics of dynamic objects from the channel state information using an amplified autocorrelation function (A-ACF) that reflects only the characteristics of dynamic objects, and then perform dimensionality reduction. By extracting the characteristics of dynamic objects using an amplified autocorrelation function (A-ACF) as a preprocessing step for dimensionality reduction, dimensionality reduction can be performed while preserving the periodicity and patterns of the data.

[0066] The control unit 43 may perform sampling as a conversion step, selecting some data points from the channel state information. By sampling, the data capacity can be reduced. The control unit 43 may also reduce the data capacity as a conversion step by selecting data at regular intervals along the time axis. Alternatively, the control unit 43 may, as a conversion step, decimate the data in the less variable parts and retain the data in the important variable parts.

[0067] <AI Agent> Each step performed by the control unit 23 may be performed by a so-called AI agent (autonomous agent) that autonomously performs tasks such as data collection and conversation necessary for executing tasks or goals input by the user, using natural language processing, processing models constructed by machine learning, general-purpose large-scale language models, etc. The conversation (instructions in natural language) between the user and the AI ​​agent is conducted, for example, in the form of voice or text chat. The AI ​​agent is included, for example, in an artificial intelligence module.

[0068] An "AI agent" is a model that, upon input of a goal (objective, purpose, etc.) such as "I want to create XX" or a task such as "output XX," breaks down the processes necessary to reach the goal or achieve the task into subtasks, actions, etc., and performs necessary data collection and analysis, program generation and execution, etc. The AI ​​agent autonomously selects and executes tasks, actions, etc. without requiring user intervention (operation input). Furthermore, the AI ​​agent may autonomously plan and execute, and evaluate the execution results itself, thereby autonomously learning to achieve the goal. For example, the AI ​​agent may autonomously update itself based on the execution results of subtasks (e.g., collected information, results of information analysis, etc.). The AI ​​agent may include a generative AI (including a large-scale language model) that has been trained to perform these processes.

[0069] The AI ​​agent has functions such as generating tasks in response to user input, including instructions, using a processing model such as a generation AI; collecting information from the storage unit 22 of the information processing system 1 or from external systems; and analyzing the collected information using a processing model such as a generation AI.

[0070] The AI ​​agent may generate a prompt to execute a subtask and input this prompt into an external large-scale language model to obtain the execution result of the subtask from the large-scale language model, or it may execute the generated prompt itself. Alternatively, the AI ​​agent may create input information to cause an external processing model to execute a specific process, and input this input information into the processing model to obtain the execution result of the subtask from the processing model.

[0071] Specifically, an AI agent may perform a task in the following manner, for example. First, the AI ​​agent receives specific instructions from the user, including a goal or task. Next, the AI ​​agent divides the goal or task into multiple subtasks or actions necessary to achieve the goal or task, and generates subtasks or actions. After dividing the goal or task, the AI ​​agent collects information necessary to perform the multiple subtasks and actions. For example, the AI ​​agent collects information by accessing databases, the internet, web searches, or API (Application Programming Interface) integration with external services. After collecting the information, the AI ​​agent, for example, uses external tools to perform the multiple subtasks and actions based on the collected information, for example, by generating and executing code, and outputs the task corresponding to the goal or the execution result of the input task.

[0072] Furthermore, the AI ​​agent may verify or evaluate the results of a subtask or action itself, and if it determines that a review is necessary, it may modify or improve the subtask or action and re-execute it. For example, after generating code to perform a specific subtask, the AI ​​agent may verify the code itself and provide feedback to modify the code.

[0073] The control unit 23 may read prompt templates to be supplied to the AI ​​agent, predetermined output constraint instructions, and user attribute information (age range, considerations, etc., regarding the subject or related parties) from the storage unit 22. These are used to generate prompts, which will be described later. Prompts are generated by at least one of (i) a rule-based system (template + insert variables) or (ii) automatic generation based on the output features of a second processing model. In case (i), the template is indexed by evaluation category, minor action pattern, and user attributes. In case (ii), the artificial intelligence module may be used as a prompt generator in the first pass and the final proposal may be generated in the second pass.

[0074] The control unit 23 may pass the operation information obtained by the first processing model to the AI ​​agent, and the AI ​​agent may format the output of the evaluation system (indicators and supporting features) shown in Figure 10 and the subject information into a prompt context. Based on the context, the AI ​​agent may perform a predetermined determination process regarding whether or not to output a prompt, and if it is determined that the output is inappropriate, it may suppress the output of the prompt and switch to generating static suggestions based on a template.

[0075] The control unit 23 may select generation using an AI agent in the generation step and have the AI ​​agent configure and supply prompts. The prompts configured by the AI ​​agent may include at least evaluation results (score and supporting features), related sub-action sequences (see Figure 11), subject information (minimum necessary), output format specification (conversation text / card / voice script), predetermined output constraint instructions (prohibited words / avoidance instructions, etc.), and personalization instructions such as tone and vocabulary level.

[0076] When using an artificial intelligence module, the control unit 23 may instruct the AI ​​agent to incorporate system instructions to be given to the artificial intelligence module. The AI ​​agent may add instructions such as "Do not make medical judgments" or "In case of emergency, guide to the designated counter," and format the output text to include supporting elements (measurement window used, representative small action, time range), and may instruct it to regenerate as needed.

[0077] In the presentation step, the control unit 23 may have the AI ​​agent perform the final formatting of the presentation content and channel selection. The AI ​​agent dynamically adjusts the endings of sentences, polite / informal style, font size, inclusion of pictograms, speech speed during speech synthesis, intonation, etc., according to the attributes of the subject or related parties. If an output unit 36 ​​or 45 such as a smart speaker is used, the voice script and visual recognition cards are associated, and a version with higher explanatory density may be delivered via a separate channel for family members and medical personnel.

[0078] The control unit 33 of the information processing device 3 may cause the AI ​​agent to perform some processing. That is, the control unit 33 may cause the AI ​​agent to perform at least some of the processing, such as configuring a prompt, inputting the prompt, verifying and formatting the output, and selecting a presentation channel. Here, the AI ​​agent that operates under the instructions of the control unit 33 may be either (a) a program module stored in and executed on the information processing device 3, or (b) an external agent service that operates for the information processing device 3 based on instructions from the information processing device 3. The control unit 33 may receive information about the output content from the user, send instructions to the AI ​​agent according to the received information, and obtain output from the AI ​​agent according to the received information. The control unit 33 may cause the AI ​​agent to generate a prompt based on the received information and predetermined reference information (prompt template, output format specification, tone and vocabulary level, etc.). The control unit 33 may input the generated prompt to an artificial intelligence module and cause the artificial intelligence module to execute processing. Subsequently, the control unit 33 may cause the output unit 36 ​​to display the generation results received from the artificial intelligence module, and may also execute control functions within the device (notifications, reminders, etc.) as necessary. The control unit 33 may also transmit the obtained generation results or a summary thereof to the server 2 as synchronization information (logs).

[0079] 3. Flow of Information Processing This section describes the flow of the information processing method executed by the information processing system 1. As shown below, the information processing method includes each step executed by the information processing system. The information processing program of this embodiment causes the computer to execute each step of the information processing system. The order of processing can be changed as appropriate, multiple processes may be executed simultaneously, and some processes may be omitted.

[0080] 3.1 Overview Figure 5 is a flowchart showing an overview of the processing performed by the information processing system 1. In this processing, first, the control unit 23 acquires channel status information transmitted from the wireless communication device as a first acquisition step (step S001). Next, as a selection step, the control unit 23 inputs the channel status information into a first processing model and causes the first processing model to output operation information related to the subject's operation (step S002). Next, as an evaluation step, the control unit 23 inputs the operation information into a second processing model and causes the second processing model to output an evaluation of the subject's operation corresponding to the operation information (step S003).

[0081] In summary, according to one embodiment, the information processing system comprises at least one processor. The processor is configured to execute the following steps by reading a program. The control unit 23 acquires channel status information transmitted from the wireless communication device as a first acquisition step. The control unit 23 inputs the channel status information into a first processing model as a identification step and causes the first processing model to output operation information related to the subject's operation. The control unit 23 inputs the operation information into a second processing model as an evaluation step and causes the second processing model to output an evaluation of the subject's operation corresponding to the operation information. Here, the first processing model is a trained model that takes channel status information as input and is trained to output subject operation information corresponding to the channel status information, and the second processing model is a processing model that takes operation information as input and is trained to output an evaluation corresponding to the operation information. In this configuration, the operation of the subject can be identified based on the channel status information transmitted from the wireless communication device 4, and an evaluation corresponding to the subject's operation can be output.

[0082] 3.2 Specific Example Below, using Figures 6 and 7, we will explain an example of how the wireless communication device 4 collects channel status information. Wireless communication device 4a is installed in room A, and wireless communication device 4b is installed in room B. There is a door between room A and room B. Wireless communication device 4b is transmitting a Wi-Fi signal. Figure 6 shows an example of the Wi-Fi signal when the door between room A and room B is closed. Figure 7 shows an example of the Wi-Fi signal when the door between room A and room B is open. The Wi-Fi signal transmitted from wireless communication device 4b experiences reflection, refraction, and scattering as it passes through the surrounding environment, so the strength and phase of the Wi-Fi signal change when the door is opened or closed.

[0083] Channel State Information (CSI) is information about the state of a communication channel that the receiver transmits to the transmitter in wireless communication. Channel State Information provides detailed information about the characteristics of the communication path from the transmitter to the receiver. It includes information about changes in Wi-Fi signal strength and phase.

[0084] Wireless communication device 4a receives a Wi-Fi signal transmitted by wireless communication device 4b. The Wi-Fi signal is transmitted by dividing it into multiple subcarriers using Orthogonal Frequency-Division Multiplexing (OFDM) technology. Multiple subcarriers refer to multiple narrow bandwidths obtained by dividing a wide bandwidth. The received signal pattern changes due to changes in the surrounding environment. In Wi-Fi communication (especially standards such as IEEE 802.11ac and 802.11ax), tens to hundreds of subcarriers are used.

[0085] Wi-Fi sensing is a sensing technology that uses Wi-Fi signals to detect changes and movements in the environment. By analyzing channel status information from multiple subcarriers, it becomes possible to track human movement, recognize gestures, detect objects, and monitor indoor environments. Wireless communication sensing is a sensing technology that utilizes wireless communication signals. Detailed motion detection of people and objects in remote indoor environments, which was previously achieved using cameras and wearables, can now be achieved simply by installing wireless communication devices, offering high accuracy while also ensuring privacy. Wireless communication sensing is expected to have applications in areas such as monitoring the elderly and children, supporting operations in elderly care facilities, security services for companies and homes, health management services, and services for monitoring the usage status of conference rooms and accommodations. Wireless communication sensing is also expected to be used in toilets, construction sites, educational settings such as nurseries and schools, entertainment, drone control, vehicle abandonment detection, energy saving, and home automation.

[0086] Wi-Fi signals have the advantage of penetrating walls and obstacles, functioning effectively in both visible and obstructed environments, and maintaining performance even in low-light conditions regardless of the time of day. Furthermore, Wi-Fi signals offer a higher level of privacy protection compared to video camera recording. Compared to radar technology, Wi-Fi devices are significantly more cost-effective and are commonly integrated into various commercially available products, making them easily usable in a wide range of environments. For these reasons, Wi-Fi-based motion detection can be used in security, smart home technology, health monitoring applications, and more.

[0087] The details of the above information processing will be explained below as an example, using Figures 8 to 9.

[0088] Figure 8 shows a floor plan 9, which is an example of the installation location of the wireless communication device 4. Floor plan 9 is the layout of the room in which the wireless communication device 4 is installed. Floor plan 9 includes one bedroom and one toilet. In floor plan 9, as an example, wireless communication device 4a is installed in the bedroom, wireless communication device 4b1 is installed next to the bed in the bedroom, and wireless communication device 4b2 is installed in the toilet. Wireless communication devices 4a, 4b1, and 4b2 each form an elliptical sensing area. In floor plan 9, by installing wireless communication devices 4a and 4b1 in the lower left area and wireless communication device 4b2 in the upper right area, all rooms in floor plan 9 can be sensed.

[0089] Figure 9 is an activity diagram showing an example of the processing flow performed by the information processing system 1. This example flow may be encompassed within the scope defined in the overview described above. The following explanation will follow each activity in this activity diagram. Note that the information processing may include any exception handling not shown. Exception handling includes interrupting the information processing or omitting each process. The selections or inputs made in the information processing may be based on user operation or may be performed automatically without user operation.

[0090] The control unit 43 of the wireless communication device 4a acquires channel status information transmitted by wireless communication devices 4a and 4b (Activity A101).

[0091] Next, the control unit 43 may perform a conversion step to convert the channel status information under predetermined conditions (activity A102). In the wireless communication device 4a, the control unit 43 performs a conversion process on the channel status information as a conversion step to improve the data format and quality of the channel status information. In addition, in the wireless communication device 4a, the control unit 43 performs a conversion process to compress the channel status information data as a conversion step to reduce the load on the server 2.

[0092] The control unit 43 preferably performs data compression or dimensionality reduction processing on the channel state information under predetermined conditions as a conversion step. The control unit 43 preferably extracts useful features from the channel state information using an autocorrelation function as a conversion step. The control unit 43 preferably extracts useful features from the channel state information by extracting an amplified autocorrelation function that reflects only the characteristics of dynamic objects as a conversion step. In this manner, the influence of environmental factors can be reduced due to scattering / reflection from static objects such as furniture and building structures. Therefore, when using the system in a new environment, the time-consuming data collection and retraining can be reduced.

[0093] Next, the control unit 43 of the wireless communication device 4a transmits channel status information to the control unit 23 of the server 2 (activity A103). The control unit 43 may also transmit the converted channel status information to the control unit 23 of the server 2.

[0094] Next, as a first acquisition step, the control unit 23 acquires channel status information transmitted from the wireless communication device 4a via the network (activity A104). If the channel status information has been converted in activity A102, the control unit 23 can acquire the converted channel status information as a first acquisition step.

[0095] As a specific step, the control unit 23 inputs channel status information to the first processing model. The first processing model may be a model that identifies multiple minor operations (activity A105).

[0096] The first processing model is a pre-trained model that takes channel state information as input and is trained to output subject behavior information corresponding to the channel state information.

[0097] The motion information is information about the subject's actions corresponding to the channel state information, identified based on the channel state information. The subject is a person or an animal. The subject's actions may include, for example, information about the subject's actions such as "sitting," "standing up," "walking," or "falling." The subject's motion information may also include, for example, whether or not there was movement, the magnitude of the movement, the location where the movement occurred, the rate of breathing, and specific movements.

[0098] Incidentally, when each operation information is broken down into more detailed operations, it may include small operations that are common to various operations. Therefore, as a specific step, the control unit 23 may input channel state information to multiple small operation identification models and output multiple small operation information. As a specific step, the control unit 23 may identify the information of small operations obtained by breaking down the operation information, and then combine the identified small operation information to recognize the operation information. The first processing model may include a small operation identification model that identifies small operation information and an operation recognition model that recognizes operations from combinations of small operation information, etc. With this configuration, a wider variety of operations can be identified compared to when operations are identified using only an operation identification model.

[0099] Figure 10 shows an example of the information flow by the information processing system 1. The control unit 23 inputs channel state information to the A operation identification model, the B operation identification model, the C operation identification model, and the D operation identification model. The A operation identification model is a trained model that has been trained to output A operation information, the B operation identification model outputs B operation information, the C operation identification model outputs C operation information, and the D operation identification model outputs D operation information. Subsequently, the control unit 23 can input the identified sub-operation information, A to D operation information, into the operation recognition model. The control unit 23 can acquire the operation information output from the operation recognition model. The control unit 23 can obtain evaluation results by inputting the output operation information into the evaluation system.

[0100] Figure 11 shows an example of the relationship between identified small movement information and the recognized movement. If the identified small movement is a repetition of movement A, movement B, and movement A, the recognized movement will be "continuing to sit". On the other hand, if the identified small movement is a sequence of movement A, movement B, movement C, and movement D, the recognized movement will be "standing up". Movements A and B are common to both continuing to sit and standing up. By dividing the movement into small movements in this way, it becomes possible to capture more partial characteristics of the movement and improve detection accuracy.

[0101] In other words, the first processing model may include multiple sub-motion identification models. The multiple sub-motion identification models are trained models that take channel state information as input and are capable of outputting sub-motion information corresponding to the channel state information, and the sub-motion information may be information about a part of the subject's movement. As a identification step, the control unit 23 may input the channel state information to each of the multiple sub-motion identification models and have each of the multiple sub-motion identification models output sub-motion information. The movement information may be, for example, standing, sitting, getting up, walking, etc., or it may be fine movements (sub-movements) obtained by breaking down larger movements such as standing, sitting, getting up, walking, etc.

[0102] Next, the control unit 23 acquires operation information related to the subject's movements output from the first processing model. If the first processing model is a set of multiple sub-movement identification models, the control unit 23 may acquire multiple sub-movement information outputs from the sub-movement identification models (Activity A106).

[0103] Next, the control unit 23 may, as a specific step, input multiple small motion information to the motion recognition model (activity A107). The motion recognition model is a processing model that takes multiple small motion information as input and outputs motion information based on the multiple small motion information. When the motion recognition model receives multiple small motion information as input, it outputs motion information. The motion recognition model is preferably a model that can recognize motion information by combining multiple small motion information. Since common small motion information can be used to identify multiple motion information, it becomes possible to identify more complex motions with fewer motion identification models. For example, the motion recognition model may identify more advanced and complex motions by combining small motion information as parallel simultaneous processing or time-series serial processing.

[0104] In activity A107, the control unit 23 may further combine the motion information output from the motion recognition model in a time series. For example, the control unit 23 may recognize that the subject performed a "standing up" action followed by a "walking action". The control unit 23 may also store the motion information identified by the motion recognition model in association with subject information. When storing the information in association with subject information, the control unit 23 may store the time the motion information occurred, the duration of the motion information, etc., along with the identified motion information.

[0105] Next, the control unit 23 can acquire motion information related to the subject's movements output from the motion recognition model (Activity A108).

[0106] Next, the control unit 23 may acquire information about the subject via the information processing device 3 (Activity A109).

[0107] Next, the control unit 23 inputs the operation information to the second processing model as an evaluation step. The control unit 23 may input multiple small operation information to the second processing model as an evaluation step. The second processing model may be, for example, an evaluation system that takes operation information as input and outputs an evaluation corresponding to the operation information. The control unit 23 may input subject information about the subject along with the operation information to the evaluation system (activity A110).

[0108] As an evaluation step, the control unit 23 aggregates and analyzes the operation information based on the time of occurrence of the operation information, the duration of the operation information, the number and frequency of occurrences of the operation information during a given period, and the total number of occurrences during a given period. Based on the results of the analysis of the operation information, the control unit 23 may evaluate the presence or absence of activity, the amount of activity, lifestyle patterns, frequency of going out, etc. The control unit 23 may perform the evaluation using an evaluation system capable of performing these evaluations.

[0109] The evaluation may involve quantifying the living conditions and health status of elderly individuals.

[0110] In activity A110, the control unit 23 may input a processing model capable of outputting an evaluation of the subject's physical condition corresponding to the motion information, as a second processing model. In this configuration, the control unit 23 can obtain an evaluation of the subject's physical condition. For example, the control unit 23 can calculate the subject's activity level based on the motion information and predict the progression of dementia or the level of care required from the activity level. Alternatively, the control unit 23 can predict the progression of dementia or the level of care required from the movement patterns based on the motion information.

[0111] Next, the control unit 23 may acquire an evaluation output from an evaluation system, which is an example of a second processing model. The evaluation is, for example, an evaluation of the subject's actions corresponding to the action information (Activity A111).

[0112] Furthermore, in activity A110, the control unit 23 may input, as a second processing model, a processing model capable of outputting an evaluation of the subject's physical information corresponding to movement information associated with a predetermined disease. The evaluation may include the detection of signs related to the predetermined disease. In this manner, the control unit 23 can obtain an evaluation of the subject's physical information corresponding to movement information associated with a predetermined disease. The control unit 23 may, for example, detect signs of Parkinson's disease by identifying the presence or absence of minor movement information specific to Parkinson's disease.

[0113] Next, the control unit 23 can store and manage the output evaluation in a database in association with the subject information (Activity A112).

[0114] Next, the control unit 23 can generate suggestions for actions recommended to the subject based on the evaluation as a generation step (Activity A113). The control unit 23 may input the evaluation into the suggestion system and generate suggestions recommended to the subject based on the evaluation. The control unit 23 may generate suggestions for the subject based on the acquired evaluation results, such as whether or not the subject is active, the amount of activity, lifestyle patterns, and frequency of going out. The control unit 23 may also generate suggestions based on evaluation results such as predictions of the progression of dementia or predictions of the level of care required.

[0115] In the generation step, when generating suggestions regarding actions recommended for the subject based on the evaluation, the control unit 23 may use a large-scale language model to generate the suggestions.

[0116] In addition, in Activity A109, the control unit 23 may acquire the subject's health information as a second acquisition step, and the control unit 23 may generate a proposal based on the evaluation and health information as a presentation step. In this manner, the control unit 23 can generate a proposal that is appropriate to the subject's health condition.

[0117] Next, as a presentation step, the control unit 23 transmits the generated proposal to the information processing device 3 via the network (activity A114). The information processing device 3 can then present the received proposal.

[0118] The control unit 23 may present the proposal by voice via the output unit 36 ​​of the information processing device 3. Alternatively, the control unit 23 may present the proposal by voice via the output unit 45 of the wireless communication device 4. By engaging in voice dialogue with the subject, activities aimed at maintaining and improving the subject's quality of life (QOL) can be promoted.

[0119] Here, as a presentation step, the control unit 23 may present the proposal to an information processing device 3 owned by a user other than the subject. For example, the control unit 23 may present the proposal to an information processing device 3 owned by a family member of the subject. Alternatively, for example, the control unit 23 may present the proposal to an information processing device 3 owned by a medical professional related to the subject.

[0120] Next, the control unit 23 may receive feedback regarding the presented evaluation and proposal via the output unit 36 ​​of the information processing device 3. The control unit 23 may receive the feedback via the information processing device 3 owned by the subject, via the information processing device 3 owned by the subject's family, or via the information processing device 3 of a medical professional related to the subject. For example, if the control unit 23 presents an evaluation result in activity A114 indicating that the subject is attempting to stand up, the information processing device 3 may receive feedback on whether the evaluation result was correct and transmit it to the control unit 23.

[0121] Next, the control unit 23 can update the first processing model or the second processing model based on the received feedback. In this configuration, the first processing model can output more accurate operational information regarding the subject's movements, and the second processing model can output evaluations corresponding to the operational information more accurately.

[0122] In the above embodiment, the control unit 33 may perform some or all of the processing performed by the control unit 23. That is, each step performed in the above embodiment may be performed by the control unit 33 alone, or it may be performed jointly by the control unit 23 and the control unit 33. Furthermore, for processing that uses an AI agent, either the agent on the control unit 23 side or the agent on the control unit 33 side may be used. For example, in activity A114, when the control unit 33 of the information processing device 3 outputs the evaluation or suggestion received from the server 2 to the information processing device 3, the control unit 33 may sequentially execute the process of generating the suggestion on its side. The following is an example of a typical flow on the control unit 33 side. The control unit 33 has the application on the information processing device 3 present the content of the evaluation or suggestion, requests confirmation and supplementary input from the user, and accepts the input. The control unit 33 generates the suggestion on its side by inserting the user input and the evaluation into the template and reference information (output format specification, tone and vocabulary level, etc.) held in the storage unit 32 (or storage unit 22 if necessary). The control unit 33 checks the generated content for compliance with a predetermined format, length, presence or absence of elements, and consistency, and formats it as necessary. The control unit 33 presents the generated content on the output unit 36 ​​and executes control within the information processing device 3 (notification, schedule registration, external device linkage, etc.) as necessary. The control unit 33 may also send necessary elements of the generated content or its summary as feedback information to the server 2 for processing of activity A115.

[0123] 4. Modifications Furthermore, the following embodiments may be adopted. The above embodiments of information processing are merely examples. The present invention is not limited thereto and can be modified as appropriate without departing from the technical idea of ​​the invention.

[0124] The control unit 23 may, as a management step, manage the output evaluations in chronological order. The control unit 23 may also, as a display control step, display the progress of the evaluations in a way that is easily visible to the user. For example, the control unit 23 may display to the information processing device 3 the presence or absence of activity, the amount of activity, lifestyle patterns, frequency of going out, etc., in chronological order. The control unit 23 may also display to the information processing device 3 the time of occurrence of the operation information, the duration of the operation information, the number of occurrences and frequency of occurrence of the operation information within an arbitrary period, the total number of occurrences within an arbitrary period, etc.

[0125] The control unit 23 may, as a management step, manage the operation information in association with the evaluation. The control unit 23 may also, as a display control step, display visual information generated based on the operation information in a visible manner. For example, the control unit 23 may display the time of occurrence of the operation information, the duration of the operation information, the number of occurrences and frequency of occurrences of the operation information during an arbitrary period, and the total number of occurrences during an arbitrary period in a graph or table on the information processing device 3.

[0126] In activity A106, the control unit 23 may input multiple small movement information into an evaluation system, which is an example of a second processing model, as an evaluation step. In this configuration, an evaluation focusing on specific small movements of the subject can be performed. For example, the control unit 23 can perform an evaluation based on small movements specific to Parkinson's disease, etc.

[0127] The overall configuration shown in Figure 1 is just one example and is not limited to it. For example, Server 2 may be distributed across two or more devices, or it may be replaced by a cloud computing system. Also, all processing may be performed on Server 2, or all processing may be performed on Information Processing Unit 3. An application may be installed on Information Processing Unit 3, and Information Processing Unit 3 and Server 2 may work together to execute the processing described above.

[0128] Server 2 may be on-premises or in a cloud environment. In the case of a cloud-based Server 2, for example, it may provide the above functions and processing in the form of SaaS (Software as a Service) or cloud computing.

[0129] In the above embodiment, server 2 performed various storage and control functions, but instead of server 2, multiple external devices may be used. That is, various information and programs may be distributed and stored across multiple external devices using blockchain technology or the like.

[0130] The product may be provided in any of the following embodiments.

[0131] (1) An information processing system comprising at least one processor, wherein the processor is configured to perform the following steps by reading a program, wherein in a first acquisition step, channel state information transmitted from a wireless communication device is acquired; in a identification step, the channel state information is input to a first processing model and operation information relating to the operation of a subject is output from the first processing model; and in an evaluation step, the operation information is input to a second processing model and an evaluation relating to the operation of the subject corresponding to the operation information is output from the second processing model, wherein the first processing model is a trained model that takes the channel state information as input and is trained to output the operation information of the subject corresponding to the channel state information; and the second processing model is a processing model that takes the operation information as input and is trained to output the evaluation corresponding to the operation information.

[0132] (2) In the information processing system described in (1) above, the first processing model includes a plurality of small motion identification models, each of the plurality of small motion identification models is a trained model that takes the channel state information as input and is trained to output small motion information corresponding to the channel state information, the small motion information is information relating to a part of the subject's movement, in the identification step, the channel state information is input to each of the plurality of small motion identification models and the small motion information is output from each of the plurality of small motion identification models, and in the evaluation step, a plurality of the small motion pieces of information are input to the second processing model and the evaluation is output.

[0133] (3) In the information processing system described in (1) or (2) above, the identification step includes a plurality of small motion identification models and a motion recognition model, the identification step inputs the channel state information to the plurality of small motion identification models and causes them to output a plurality of small motion information, inputs the plurality of small motion information to the motion recognition model and causes it to output the motion information, each of the plurality of small motion identification models is a trained model that takes the channel state information as input and is trained to be able to output small motion information corresponding to the channel state information, the small motion information is information relating to a part of the subject's movement, and the motion recognition model is a processing model that takes a plurality of the small motion information as input and is able to output the motion information based on the plurality of the small motion information.

[0134] (4) In the information processing system described in any one of (1) to (3) above, the second processing model is a processing model capable of outputting the evaluation of the physical state of the subject corresponding to the operation information.

[0135] (5) In the information processing system described in any one of (1) to (4) above, the second processing model is a processing model capable of outputting the evaluation of the subject's physical information corresponding to the movement information associated with a predetermined disease, and the evaluation includes the detection of signs related to the predetermined disease.

[0136] (6) An information processing system according to any one of (1) to (5) above, wherein in the generation step, a proposal relating to an action recommended for the subject based on the evaluation is generated, and in the presentation step, the proposal is presented.

[0137] (7) The information processing system described in (6) above, wherein in the second acquisition step, health information of the subject is acquired, and in the presentation step, the proposal generated based on the evaluation and the health information is presented.

[0138] (8) In the information processing system described in (7) above, the system presents the proposal by voice in the presentation step.

[0139] (9) In the information processing system described in (7) or (8) above, the system provides, in the presentation step, the proposal to a user other than the subject.

[0140] (10) An information processing system as described in any one of (1) to (9) above, wherein the management step further manages the output evaluations in chronological order, and the display control step further displays the progress of the evaluations in a way that is visible to the user.

[0141] (11) In the information processing system described in (10) above, the management step manages the operation information in association with the evaluation, and the display control step displays the visual information generated based on the operation information in a visible manner.

[0142] (12) An information processing system according to any one of (1) to (11) above, wherein in the conversion step, the channel state information is converted under predetermined conditions, and in the first acquisition step, the converted channel state information is acquired.

[0143] (13) An information processing method comprising each step performed by the information processing system described in any one of (1) to (12) above.

[0144] (14) A program that causes a computer to perform each step of the information processing system described in any one of (1) to (12) above.

[0145] Finally, various embodiments of the present invention have been described, but these are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents.

[0146] 1: Information processing system, 2: Server, 20: Communication bus, 21: Communication unit, 22: Storage unit, 23: Control unit, 3: Information processing device, 30: Communication bus, 31: Communication unit, 32: Storage unit, 33: Control unit, 34: Display unit, 35: Input unit, 36: Output unit, 4: Wireless communication device, 4a: Wireless communication device, 4b1: Wireless communication device, 4b2: Wireless communication device, 40: Communication bus, 41: Communication unit, 42: Storage unit, 43: Control unit, 44: Input unit, 45: Output unit, 9: Floor plan, A: Room, B: Room

Claims

1. An information processing system comprising at least one processor, wherein the processor is configured to perform the following steps by reading a program: in a first acquisition step, channel state information transmitted from a wireless communication device is acquired; in a identification step, the channel state information is input to a first processing model and operation information relating to the operation of a subject is output from the first processing model; and in an evaluation step, the operation information is input to a second processing model and an evaluation relating to the operation of the subject corresponding to the operation information is output from the second processing model, wherein the first processing model is a trained model that takes the channel state information as input and is trained to output the operation information of the subject corresponding to the channel state information; and the second processing model is a processing model that takes the operation information as input and is trained to output the evaluation corresponding to the operation information.

2. The information processing system according to claim 1, wherein the first processing model includes a plurality of sub-motion identification models, each of the plurality of sub-motion identification models is a trained model that takes the channel state information as input and is trained to output sub-motion information corresponding to the channel state information, the sub-motion information is information relating to a part of the operation of the subject, in the identification step, the channel state information is input to each of the plurality of sub-motion identification models and the sub-motion information is output from each of the plurality of sub-motion identification models, and in the evaluation step, a plurality of the sub-motion information is input to the second processing model and the evaluation is output.

3. The information processing system according to claim 1, wherein the identification step includes a plurality of small motion identification models and a motion recognition model, the identification step inputs the channel state information to the plurality of small motion identification models and causes them to output a plurality of small motion information, the motion recognition model inputs the plurality of small motion information and causes it to output motion information, each of the plurality of small motion identification models is a trained model that takes the channel state information as input and is trained to be able to output small motion information corresponding to the channel state information, the small motion information is information relating to a part of the operation of the subject, and the motion recognition model is a processing model that takes the plurality of small motion information as input and is able to output motion information based on the plurality of small motion information.

4. The information processing system according to claim 1, wherein the second processing model is a processing model capable of outputting the evaluation of the physical state of the subject corresponding to the operation information.

5. The information processing system according to claim 1, wherein the second processing model is a processing model capable of outputting the evaluation of the subject's physical information corresponding to the movement information associated with a predetermined disease, and the evaluation includes the detection of signs related to the predetermined disease.

6. An information processing system according to claim 1, wherein the generation step further generates suggestions regarding actions recommended for the subject based on the evaluation, and the presentation step further presents the suggestions.

7. An information processing system according to claim 6, wherein in the second acquisition step, health information of the subject is acquired, and in the presentation step, the proposal generated based on the evaluation and the health information is presented.

8. The information processing system according to claim 7, wherein in the presentation step, the proposal is presented by voice.

9. The information processing system according to claim 7, wherein the presentation step includes presenting the proposal to a user other than the subject.

10. An information processing system according to claim 1, wherein the management step further manages the output evaluations in chronological order, and the display control step further displays the progress of the evaluations in a way that is visible to the user.

11. An information processing system according to claim 10, wherein the management step manages the operation information in association with the evaluation, and the display control step displays the visual information generated based on the operation information in a visible manner.

12. An information processing system according to claim 1, wherein in the conversion step, the channel state information is converted under predetermined conditions, and in the first acquisition step, the converted channel state information is acquired.

13. An information processing method comprising each step performed by the information processing system described in any one of claims 1 to 12.

14. A program for causing a computer to perform each step of the information processing system described in any one of claims 1 to 12.

Images