Wi-fi signal-based copresence estimation method and device

Abstract

This operating method of a user device may include the steps of: obtaining a first Wi-Fi RSSI corresponding to nearby Wi-Fi networks of the user device and a plurality of second Wi-Fi RSSIs corresponding to nearby Wi-Fi networks of each of other devices; inputting the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs to a copresence estimation model to estimate whether the other devices are present in the same space as the user device; generating a feedback data set including the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs on the basis of feedback on a result of the estimation; generating a filtered data set by filtering out a pre-collected first Wi-Fi RSSI and a pre-collected plurality of second Wi-Fi RSSIs according to a preset condition; and updating the copresence estimation model by using the feedback data set and the filtered data set.

Description

Wi-Fi signal-based coexistence estimation method and device

[0001] The present disclosure relates to a method and device for estimating coexistence between devices based on Wi-Fi signals and optimizing an artificial intelligence model for estimating coexistence.

[0002] Technology utilizing Wi-Fi signals is being used to estimate the location of devices in indoor environments. As AI is utilized in various fields, it can also be applied to estimating the location of devices in indoor environments. However, even with AI, inference results may not always be reliable, as they may be inaccurate depending on the user's environment. While user feedback can be used as training data, it is difficult to obtain training data appropriate for the user's environment based solely on user feedback.

[0003] According to one aspect of the present disclosure, a method of operating a user device may be provided. The method may include obtaining a first Wi-Fi RSSI corresponding to nearby Wi-Fi networks of the user device. The method may include obtaining a plurality of second Wi-Fi RSSIs corresponding to nearby Wi-Fi networks of each of other devices. The method may include applying the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs as input data to a copresence estimation model to estimate whether the other devices exist in the same space as the user device. The method may include receiving feedback on the estimation result. The method may include creating a feedback dataset comprising the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs based on the feedback. The method may include obtaining a pre-collected first Wi-Fi RSSI and a plurality of pre-collected second Wi-Fi RSSIs. The method may include a step of creating a filtered dataset by filtering the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs according to preset conditions. The method may include a step of updating the coexistence estimation model using the feedback dataset and the filtered dataset.

[0004] According to one aspect of the present disclosure, a user device may be provided. The user device may include: a communication interface; a memory storing one or more instructions; and one or more processors executing the one or more instructions stored in the memory. The one or more processors may, by executing the one or more instructions, obtain a first Wi-Fi RSSI corresponding to nearby Wi-Fi networks of the user device. The one or more processors may, by executing the one or more instructions, obtain a plurality of second Wi-Fi RSSIs corresponding to nearby Wi-Fi networks of each of other devices. The one or more processors may, by executing the one or more instructions, apply the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs as input data to a copresence estimation model to estimate whether the other devices exist in the same space as the user device. The one or more processors may, by executing the one or more instructions, receive feedback on the estimation result. The one or more processors may, by executing the one or more instructions, create a feedback dataset comprising the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs based on the feedback. The one or more processors may, by executing the one or more instructions, obtain the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs. The one or more processors may, by executing the one or more instructions, create a filtered dataset by filtering the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs according to preset conditions.The one or more processors can update the coexistence estimation model using the feedback dataset and the filtered dataset by executing the one or more instructions.

[0005] According to one aspect of the present disclosure, a computer-readable recording medium having recorded thereon a program for executing any one of the aforementioned and hereinafter described methods related to the operation of a user device may be provided.

[0006] FIG. 1 is a diagram illustrating an operation of a user device estimating a location indoors according to one embodiment of the present disclosure.

[0007] FIG. 2 is a diagram illustrating an operation of personalizing an artificial intelligence model that estimates whether a user device exists in the same space as another device according to one embodiment of the present disclosure.

[0008] FIG. 3 is a diagram for explaining the operation of a coexistence estimation model used by a user device according to one embodiment of the present disclosure.

[0009] FIG. 4A is a diagram illustrating an operation of a user device obtaining feedback on a coexistence estimation result according to one embodiment of the present disclosure.

[0010] FIG. 4b is a diagram illustrating an operation of a user device generating feedback data according to one embodiment of the present disclosure.

[0011] FIG. 5A is a diagram illustrating data collection operations of a user device and another device according to one embodiment of the present disclosure.

[0012] FIG. 5b is a diagram illustrating an operation of a user device generating a filtered dataset according to one embodiment of the present disclosure.

[0013] FIG. 6 is a diagram illustrating a training dataset generated by a user device according to one embodiment of the present disclosure.

[0014] FIG. 7A is a flowchart illustrating a method by which a user device operates when receiving feedback according to one embodiment of the present disclosure.

[0015] Figure 7b is a flowchart for explaining the operation of Figure 7a in more detail.

[0016] FIG. 8 is a diagram illustrating an operation of a user device generating a training dataset according to one embodiment of the present disclosure.

[0017] FIG. 9 is a diagram illustrating an operation of a user device generating a training dataset according to one embodiment of the present disclosure.

[0018] FIG. 10 is a diagram illustrating an operation of a user device using a coexistence estimation model according to one embodiment of the present disclosure.

[0019] FIG. 11 is a block diagram illustrating a configuration of a user device according to one embodiment of the present disclosure.

[0020] Hereinafter, terms used in this specification will be briefly described, and the present disclosure will be described in detail. In this disclosure, the expression “at least one of a, b, or c” can refer to “a,” “b,” “c,” “a and b,” “a and c,” “b and c,” “all of a, b, and c,” or variations thereof.

[0021] The terms used in this disclosure are selected from widely used, common terms, taking into account the functions of the disclosure. However, these terms may vary depending on the intentions of those skilled in the art, precedents, the emergence of new technologies, etc. Furthermore, in certain cases, terms may be arbitrarily selected by the applicant, in which case their meanings will be described in detail in the relevant description. Therefore, the terms used in this disclosure should not be defined simply as names, but rather based on the meanings of the terms and the overall content of the disclosure.

[0022] Singular expressions may include plural expressions unless the context clearly indicates otherwise. Terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art described herein. Furthermore, terms containing ordinal numbers, such as "first" or "second," used herein may be used to describe various components, but such components should not be limited by such terms. Such terms are used solely to distinguish one component from another.

[0023] When a part of the specification is said to "include" a component, unless otherwise specifically stated, this does not exclude other components but rather implies the inclusion of other components. Furthermore, terms such as "part" and "module" used in the specification refer to a unit that processes at least one function or operation, which may be implemented in hardware, software, or a combination of hardware and software.

[0024] Below, with reference to the attached drawings, embodiments of the present disclosure are described in detail so that those skilled in the art can easily practice the present disclosure. However, the present disclosure may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, portions irrelevant to the description have been omitted for clarity of explanation, and similar reference numerals have been used throughout the specification to indicate similar parts.

[0025] The present disclosure will be described below with reference to the attached drawings.

[0026] FIG. 1 is a diagram illustrating an operation of a user device estimating a location indoors according to one embodiment of the present disclosure.

[0027] Referring to FIG. 1, a user device (2000) according to one embodiment can estimate its relative location with respect to a plurality of other devices (100) indoors. For example, the user device (2000) can estimate whether the user device (2000) is located in the same space as each of the other devices (100) or in a different space.

[0028] In one embodiment, "space" may refer to a physically separate three-dimensional environment with a spatial layout (e.g., floor, walls, pillars, ceiling, etc.). For example, assuming a user's home environment, Room A (110), Room B (120), Room C (130), and Living Room (140) may each be referred to as spaces. In the example of FIG. 1, if a smartphone is in the Living Room (140) and a TV is in Room B (120), the smartphone and TV can be viewed as existing in different spaces.

[0029] In one embodiment, the user device (2000) and other devices (100) may be interconnected and interact with each other. For example, the user device (2000) and other devices (100) may be connected to the same network (e.g., home Wi-Fi). Alternatively, the user device (2000) and other devices (100) may be connected to the same server (e.g., cloud server), but are not limited thereto.

[0030] The user device (2000) may include various types of devices. For example, the user device (2000) may be a mobile device such as a tablet PC, a smartphone, etc. The other devices (100) may include various types of devices. The other devices (100) may include, but are not limited to, various home appliances such as a washing machine, a dryer, a vacuum cleaner, a robot vacuum cleaner, an air conditioner, a clothes manager, an air purifier, a dishwasher, etc. In one embodiment, the user may manage and control the other devices (100) using the user device (2000). The user device (2000) may receive a user input and request information from a plurality of other devices (100) or control the plurality of other devices (100) based on the received user input.

[0031] In one embodiment, the user device (2000) can estimate whether other devices (100) exist in the same space as the user device. In this case, the user device (2000) can estimate whether the devices coexist using a coexistence estimation model, which is an artificial intelligence model. The user device (2000) can optimize the coexistence estimation model to suit the user space. To optimize the coexistence estimation model, the user device (2000) can collect and process data to generate training data.

[0032] The specific operations of the user device (2000) for collecting data for training the coexistence estimation model and fine-tuning the coexistence estimation model will be described in more detail through the drawings and descriptions thereof described below.

[0033] FIG. 2 is a diagram illustrating an operation of personalizing an artificial intelligence model that estimates whether a user device exists in the same space as another device according to one embodiment of the present disclosure.

[0034] In operation S210, the user device (2000) may obtain Wi-Fi RSSI corresponding to Wi-Fi networks nearby the user device (2000). In the present disclosure, the Wi-Fi RSSI received by the user device (2000) may be referred to as a first Wi-Fi RSSI.

[0035] In one embodiment, the user device (2000) may obtain an RSSI value indicating the reception strength of a Wi-Fi signal received through a communication circuit (e.g., a Wi-Fi card, etc.) included in a communication interface. The RSSI may be a value expressed in units of dBm.

[0036] In one embodiment, if there are multiple Wi-Fi routers or wireless access points (APs) around the user device (2000), there may be multiple RSSI values. For example, the RSSI may be a sequence of values ​​corresponding to each of the surrounding Wi-Fi networks.

[0037] In operation S220, the user device (2000) may acquire a plurality of Wi-Fi RSSIs corresponding to nearby Wi-Fi networks of each of the other devices. In the present disclosure, a Wi-Fi RSSI received by a device other than the user device (2000) may be referred to as a second Wi-Fi RSSI. In other words, in order to distinguish the Wi-Fi RSSI of the user device (2000) from the Wi-Fi RSSI of the other devices, the Wi-Fi RSSI of the user device (2000) will be referred to as a first Wi-Fi RSSI, and the Wi-Fi RSSI of any other device will be referred to as a second Wi-Fi RSSI, but both the first Wi-Fi RSSI and the second Wi-Fi RSSI essentially mean a Wi-Fi RSSI that the device generally acquires.

[0038] In one embodiment, the number of other devices may be two or more. For example, the user device (2000) may receive the Wi-Fi RSSI of device A from another device, device A, and may receive the Wi-Fi RSSI of device B from another device, device B. In other words, the user device (2000) may obtain a plurality of second Wi-Fi RSSIs corresponding to each of the other devices. Continuing with the example described above, the plurality of second Wi-Fi RSSIs may include the Wi-Fi RSSI of device A and the Wi-Fi RSSI of device B.

[0039] In operation S230, the user device (2000) can estimate whether other devices exist in the same space as the user device (2000) by applying the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs as input data to a coexistence estimation model.

[0040] In one embodiment, the coexistence estimation model may be an artificial intelligence model trained to take as input the Wi-Fi RSSI of two devices and estimate whether the two devices exist in the same space.

[0041] The user device (2000) can input input data consisting of a first Wi-Fi RSSI and a second Wi-Fi RSSI into a coexistence estimation model. The plurality of second Wi-Fi RSSIs are Wi-Fi RSSIs of each of devices other than the user device (2000). For example, the plurality of second Wi-Fi RSSIs may include the Wi-Fi RSSI of device A and the Wi-Fi RSSI of device B. For example, the user device (2000) can input the Wi-Fi RSSI of device A among the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs into the coexistence estimation model to estimate whether device A exists in the same space as the user device (2000). In addition, for example, the user device (2000) can input the Wi-Fi RSSI of device B among the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs into the coexistence estimation model to estimate whether device B exists in the same space as the user device (2000).

[0042] In operation S240, the user device (2000) can receive feedback on the estimation result. The user device (2000) can obtain user input for inputting feedback from the user.

[0043] In one embodiment, the feedback may include positive feedback. Positive feedback may indicate that the estimation result of the coexistence estimation model is accurate. For example, the user device (2000) may estimate that device A exists in the same space as the user device (2000), or may estimate that device A exists in a different space. If the estimation result of the user device (2000) is accurate, the user may input user input indicating positive feedback.

[0044] In one embodiment, the feedback may include negative feedback. Negative feedback may indicate that the estimation result of the coexistence estimation model is inaccurate. For example, the user device (2000) may estimate that device B is present in the same space as the user device (2000), or may estimate that device B is present in a different space. If the estimation result of the user device (2000) is inaccurate, the user may input user input indicating negative feedback.

[0045] In one embodiment, the user device (2000) may request feedback from the user. For example, the user device (2000) may execute a feedback request function that outputs a result of estimating whether the user device (2000) and another device (e.g., device A and / or device B) coexist in the same space, and requests the accuracy / inaccuracy of the estimation result. In response to the feedback request, the user device (2000) may obtain user input for providing feedback on the estimation result of coexistence.

[0046] In one embodiment, the user device (2000) may be a mobile device (e.g., a smartphone, etc.). The other devices may be devices with relatively low mobility (e.g., a TV, etc.). Accordingly, the other devices may have location information (e.g., Room A, Room B, Living Room, etc.) pre-stored. The location information of the other devices may be input by the user. In one embodiment, the feedback may include location information of the user device (2000). The user device (2000) may receive current location information of the user device (2000) (e.g., Room A, Room B, Living Room, etc.). The current location information of the user device (2000) may be used to match the location information of the other devices.

[0047] In operation S250, the user device (2000) may generate a feedback dataset consisting of a first Wi-Fi RSSI and a plurality of second Wi-Fi RSSIs based on the feedback.

[0048] In one embodiment, the user device may create data pairs to be included in the feedback dataset based on the estimation result obtained in operation S230 and the user feedback obtained in operation S240. For example, the user device (2000) may create a positive pair, which is a data pair indicating that another device exists in the same space as the user device (2000). Also, for example, the user device (2000) may create a negative pair, which is a data pair indicating that another device exists in a different space from the user device (2000).

[0049] In other words, the feedback dataset may include multiple data pairs consisting of {first Wi-Fi RSSI, second Wi-Fi RSSI}, and each data pair may be annotated with information indicating whether it is the same space or a different space.

[0050] In operation S260, the user device (2000) can obtain a pre-collected first Wi-Fi RSSI and a plurality of pre-collected second Wi-Fi RSSIs.

[0051] The pre-collected first Wi-Fi RSSI may be collected periodically or aperiodically from the user device (2000). The plurality of pre-collected second Wi-Fi RSSIs may be collected periodically or aperiodically from other devices. For example, the plurality of pre-collected second Wi-Fi RSSIs may include the Wi-Fi RSSI pre-collected from device A and the Wi-Fi RSSI pre-collected from device B. The user device (2000) may receive the pre-collected second Wi-Fi RSSIs from each of the other devices.

[0052] In one embodiment, operation S260 may be performed only when feedback of operation S240 is received. The user device (2000) may perform operation S260 in response to obtaining feedback. Based on obtaining feedback, the user device (2000) may update the coexistence estimation model by performing operations S270 to S280 described below.

[0053] In operation S270, the user device (2000) can generate a filtered dataset by filtering the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs according to preset conditions.

[0054] In one embodiment, the user device (2000) may apply the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs as input data to a coexistence estimation model. The user device (2000) may estimate whether another device was present in the same space as the user device (2000) at the time when the Wi-Fi RSSI was collected using the coexistence estimation model. For example, the user device (2000) may input the Wi-Fi RSSI of device A among the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs into the coexistence estimation model, and estimate whether device A was present in the same space as the user device (2000) at that time. Additionally, for example, the user device (2000) can input the Wi-Fi RSSI of device B among the first Wi-Fi RSSI collected in advance and the plurality of second Wi-Fi RSSIs collected in advance into the coexistence estimation model to estimate whether device B existed in the same space as the user device (2000) at that time.

[0055] In one embodiment, the preset conditions for filtering data by the user device (2000) may be defined based on the number of devices present in the same space as the user device (2000). The user device (2000) may determine whether to filter the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs based on the number of devices estimated to be present in the same space as the user device among other devices. The preset conditions are further described in the description of FIGS. 5A and 5B .

[0056] The user device (2000) can generate a filtered dataset by filtering out bad data and leaving only good data among the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs. The filtered dataset can include multiple data pairs consisting of {first Wi-Fi RSSI, second Wi-Fi RSSI}, and each data pair can be annotated with information indicating whether it is in the same space or a different space.

[0057] In operation S280, the user device (2000) can update the coexistence estimation model using the feedback dataset and the filtered dataset.

[0058] In one embodiment, the feedback dataset and the filtered dataset are comprised of Wi-Fi RSSI collected within the user's space. The user device (2000) can fine-tune the coexistence estimation model using the feedback dataset and the filtered dataset as training datasets. By using the feedback dataset for training, the user device (2000) can increase the accuracy of the coexistence estimation within the user's space. Meanwhile, in the case of feedback data, it is difficult to secure a large amount of data, and securing a sufficient amount of data may require a long time. The user device (2000) can increase the accuracy of the coexistence estimation model while allowing feedback to be reflected early by using the filtered dataset that can supplement the feedback dataset for training.

[0059] In one embodiment, the user device (2000) may assign a higher weight to the feedback dataset than to the filtered dataset. The user device (2000) may use the weights to update the coexistence estimation model, thereby ensuring that the insufficient number of feedback data is given a higher weight.

[0060] FIG. 3 is a diagram for explaining the operation of a coexistence estimation model used by a user device according to one embodiment of the present disclosure.

[0061] In this disclosure, “model” refers to a coexistence estimation model (300) unless otherwise specified. The coexistence estimation model (300) may be an artificial intelligence model trained to receive input data (310) consisting of pairs of data from two different devices, and output a result estimating whether the two different devices exist in the same space or in different spaces.

[0062] Referring to FIG. 3, a user device (2000) can perform an inference operation by inputting input data (310) into a coexistence estimation model (300).

[0063] In one embodiment, the input data (310) may be comprised of a first Wi-Fi RSSI (312) and a second Wi-Fi RSSI (314). The first Wi-Fi RSSI (312) may correspond to Wi-Fi networks nearby the user device (2000), and the second Wi-Fi RSSI (314) may correspond to Wi-Fi networks nearby the other device. In the example of FIG. 3, the first Wi-Fi RSSI (312) represents the strengths of signals received by the user device (2000) from each of ten wireless access points (APs) nearby the user device (2000), and the second Wi-Fi RSSI (314) represents the strengths of signals received by the other device from each of ten wireless access points (APs) nearby the other device. The user device (2000) can input input data (310) consisting of a first Wi-Fi RSSI (312) and a second Wi-Fi RSSI (314) into a coexistence estimation model (3000) to estimate whether the user device (2000) and another device exist in the same space. In this case, each of the Wi-Fi RSSI pairs used for coexistence estimation may be acquired from each device at the same time.

[0064] In one embodiment, the coexistence estimation model (300) may be implemented using a deep neural network architecture and algorithm of artificial intelligence for performing a classification task, or may be implemented through a modification of a known deep neural network architecture and algorithm of artificial intelligence. For example, the coexistence estimation model (300) may be implemented through a deep neural network model including one or more multi-layer perceptrons (MLPs) composed of layers including weights, or may be implemented through a deep neural network model based on a transformer architecture including an attention mechanism, but is not limited thereto.

[0065] In one embodiment, the coexistence estimation model (300) may be a pre-trained model. The coexistence estimation model (300) may be trained using correct data consisting of Wi-Fi RSSI data pairs. The correct data may be annotated with information indicating whether two devices are in the same space. In this case, a Wi-Fi RSSI data pair labeled as "same space" may be referred to as a positive pair, and a Wi-Fi RSSI data pair labeled as "different space" may be referred to as a negative pair.

[0066] The term "pre-trained model" may be referred to by various expressions that represent the same or similar concepts. For example, the term "pre-trained model" may be replaced with expressions such as "base model," "pre-fine-tuned model," or "pre-updated model," but is not limited to the examples described above.

[0067] In one embodiment, the user device (2000) can generate data for updating the coexistence estimation model (300). The user device (2000) can generate a training dataset for personalizing the coexistence estimation model (300) using data collected as the user device operates in the user's space. The user device (2000) can fine-tune the base model using the training dataset. In other words, the user device (2000) can optimize the coexistence estimation model (300) to operate well in the user's space.

[0068] Data for updating the coexistence estimation model (300) may include feedback data generated based on user feedback and filtered data obtained by filtering out high-quality data from previously collected data. The specific operations by which the user device (2000) generates data for updating the coexistence estimation model (300) will be described in detail in the description of the drawings described below.

[0069] Meanwhile, the optimization of the coexistence estimation model (300) is intended to improve coexistence estimation performance in the user's space. Therefore, the updated coexistence estimation model (300) may be referred to as an "optimized model," a "personalized model," a "customized model," a "fine-tuned model," a "updated model," etc., and may also be referred to by various other expressions representing concepts identical or similar to the examples described above.

[0070] FIG. 4A is a diagram illustrating an operation of a user device obtaining feedback on a coexistence estimation result according to one embodiment of the present disclosure.

[0071] In one embodiment, the user device (2000) can estimate whether each of a plurality of other devices exists in the same space as the user device (2000). FIG. 4b illustrates an example of estimating coexistence and generating feedback data for device A (400) among other devices.

[0072] In one embodiment, the user device (2000) may perform a coexistence estimation task to infer whether device A (400) exists in the same space as the user device (2000). The coexistence estimation task may be initiated based on a user input that executes a coexistence estimation function.

[0073] The user device (2000) can input the Wi-Fi RSSI of the user device (2000) and the Wi-Fi RSSI of device A (400) into the coexistence estimation model (410) to estimate whether device A (400) exists in the same space as the user device (2000). The user device (2000) can obtain feedback on the estimation result. The feedback may include information indicating that the estimation result is inaccurate (or information indicating that the estimation result is accurate). In addition, the feedback may include location information of the user device (2000) and / or location information of device A (400).

[0074] Referring to the first example (402), the user device (2000) exists in a different space from device A (400). The user device (2000) may input the first Wi-Fi RSSI A (420), which is the Wi-Fi RSSI of the user device (2000), and the second Wi-Fi RSSI A (430), which is the Wi-Fi RSSI of device A (400), into the coexistence estimation model (410). At this time, the estimation result of the coexistence estimation model (410) may be output as “same space.” Since the user device (2000) and device A (400) exist in different spaces, the estimation result in the first example (402) is an incorrect prediction.

[0075] Referring to the second example (404), the user device (2000) exists in the same space as the device A (400). The user device (2000) inputs the first Wi-Fi RSSI B (422), which is the Wi-Fi RSSI of the user device (2000), and the second Wi-Fi RSSI B (432), which is the Wi-Fi RSSI of the device A (400), into the coexistence estimation model (410), and can obtain the output of “same space” as the estimation result. Since the user device (2000) and device A (400) exist in the same space, the estimation result in the second example (404) is a correct prediction.

[0076] In one embodiment, the user device (2000) can obtain feedback on the estimation result. This feedback can be obtained in various ways. For example, after the estimation result is output, the user device (2000) can receive user input indicating feedback on the estimation result. Furthermore, for example, the user device (2000) can output the estimation result and display a user interface requesting feedback on the estimation result.

[0077] In one embodiment, the feedback may include information indicating that the estimation result is inaccurate and / or location information of the user device (2000). If the feedback includes information indicating that the estimation result is inaccurate, the user device (2000) may configure a negative pair of the Wi-Fi RSSI of a device estimated to be present in the same space as the user device (2000) among a plurality of other devices based on the feedback and the Wi-Fi RSSI of the user device (2000). A negative pair is a label indicating that the two devices are present in different spaces. In addition, if the feedback includes location information of the user device (2000), the user device (2000) may configure a positive pair of the Wi-Fi RSSI of a device whose location information corresponds to the location information of the user device (2000) among a plurality of other devices based on the feedback and the Wi-Fi RSSI of the user device (2000). A positive pair is a label indicating that the two devices are present in the same space. In this case, the user device (2000) can receive location information of multiple other devices from multiple other devices.

[0078] Referring back to the first example (402), the estimation result of the first example (402) is that device A (400) exists in the same space as the user device (2000). Since this is an incorrect prediction, the feedback may be information indicating that the estimation result is inaccurate. In this case, based on the feedback, the user device (2000) may configure a negative pair of the second Wi-Fi RSSI A (430), which is the Wi-Fi RSSI of device A (400), which was estimated to exist in the same space as the user device (2000), and the first Wi-Fi RSSI A (420), which is the Wi-Fi RSSI of the user device (2000). In addition, the user device (2000) may identify a device having location information corresponding to the location information of the user device (2000), among devices other than device A (400), and create a positive pair of the first Wi-Fi RSSI and the second Wi-Fi RSSI.

[0079] In one embodiment, the feedback may include information indicating that the estimation result is accurate. If the feedback includes information indicating that the estimation result is accurate, the user device (2000) may configure the Wi-Fi RSSI of the user device (2000) and the Wi-Fi RSSI of another device as a positive pair based on the feedback. For example, in the second example (404), when feedback on the estimation result is received, the user device (2000) may configure the first Wi-Fi RSSI B (422), which is the Wi-Fi RSSI of the user device (2000), and the second Wi-Fi RSSI B (432), which is the Wi-Fi RSSI of device A (400), as a positive pair.

[0080] FIG. 4b is a diagram illustrating an operation of a user device generating feedback data according to one embodiment of the present disclosure.

[0081] In one embodiment, the user device (2000) may create a feedback dataset (440) comprising feedback data consisting of a first Wi-Fi RSSI and a second Wi-Fi RSSI of another device based on the feedback.

[0082] In one embodiment, the other devices may be multiple. For example, the multiple other devices may include device A and device B. Feedback data A (442) may be obtained based on feedback on the estimation result of coexistence between the user device (2000) and device A. In this case, feedback data A (442) may include {first Wi-Fi RSSI A, second Wi-Fi RSSI A} and feedback information of the user device (2000) and device A. Similarly, feedback B may include {first Wi-Fi RSSI B, second Wi-Fi RSSI B} and feedback information of the user device (2000) and device B.

[0083] The user device (2000) can create a feedback dataset (440) in which each of the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs are paired. Meanwhile, when the user device (2000) estimates coexistence between the user device (2000) and each of the plurality of other devices, feedback may be received for only some of the estimation results for the plurality of other devices. In this case, the user device (2000) may include only data corresponding to the feedback in the feedback dataset (440). For example, feedback on the estimation result of coexistence between the user device (2000) and device A and feedback on the estimation result of coexistence between the user device (2000) and device B may be obtained, whereas feedback on the estimation result of coexistence between the user device (2000) and device C may not be obtained. In this case, the feedback dataset (440) may only include feedback data A (442) and feedback data B (444) corresponding to the devices from which user device (2000) feedback was obtained.

[0084] FIG. 5A is a diagram illustrating data collection operations of a user device and another device according to one embodiment of the present disclosure.

[0085] In explaining FIGS. 5A and 5B, it is assumed that the multiple devices corresponding to the second Wi-Fi RSSI are each located in different spaces. A specific example is as shown in FIG. 5A, where device A is located in room A, device B is located in the living room, device C is located in room B, and device D is located in room C.

[0086] In one embodiment, the user device (2000) may continuously collect a first Wi-Fi RSSI corresponding to Wi-Fi networks surrounding the user device (2000). The user device (2000) may store the first Wi-Fi RSSI periodically or aperiodically. The first Wi-Fi RSSI stored through continuous collection may be referred to as a “pre-collected first Wi-Fi RSSI (510).”

[0087] And, each of the other devices can continuously collect the second Wi-Fi RSSI corresponding to the Wi-Fi networks around each device. The second Wi-Fi RSSI stored through continuous collection may be referred to as “pre-collected second Wi-Fi RSSI (520).” The pre-collected second Wi-Fi RSSI (520) may include the second Wi-Fi RSSI of a plurality of other devices. For example, the pre-collected second Wi-Fi RSSI (520) may include the second Wi-Fi RSSI (522) of device A, the second Wi-Fi RSSI (524) of device B, ..., etc.

[0088] In one embodiment, the user device (2000) can receive a second Wi-Fi RSSI (520) that has been previously collected from other devices. The user device (2000) can match the first Wi-Fi RSSI (510) that has been previously collected and the second Wi-Fi RSSI (520) that has been previously collected, to generate a raw data set (530) that is composed of pairs of {first Wi-Fi RSSI, second Wi-Fi RSSI}. The user device (2000) can filter out bad data from the raw data set (530) based on preset conditions and retain only good data. The operation of the user device (2000) filtering data is further described with reference to FIG. 5B.

[0089] FIG. 5b is a diagram illustrating an operation of a user device generating a filtered dataset according to one embodiment of the present disclosure.

[0090] In one embodiment, the user device (2000) can generate a filtered dataset (540) by filtering a pre-collected first Wi-Fi RSSI (510) and a plurality of pre-collected second Wi-Fi RSSIs (520) according to preset conditions.

[0091] The user device (2000) can perform an inference operation of the coexistence estimation model by applying the pre-collected first Wi-Fi RSSI (510) and the pre-collected plurality of second Wi-Fi RSSIs (520) as input data to the coexistence estimation model. For example, the user device (2000) can estimate whether device A exists in the same space as the user device (2000) by applying the pre-collected first Wi-Fi RSSI (510) and the pre-collected second Wi-Fi RSSI (522) of device A to the coexistence estimation model. And, the user device (2000) can estimate whether device B exists in the same space as the user device (2000) by using the pre-collected first Wi-Fi RSSI (510) and the pre-collected second Wi-Fi RSSI (524) of device B, and can estimate whether device C exists in the same space as the user device (2000) by using the pre-collected first Wi-Fi RSSI (510) and the pre-collected second Wi-Fi RSSI (526) of device C.

[0092] The user device (2000) may filter data based on the estimation results of the coexistence estimation model. The user device (2000) may determine whether to filter the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs based on the number of devices estimated to exist in the same space as the user device (2000) among a plurality of other devices.

[0093] In one embodiment, the preset conditions for determining whether to filter data after the user device (2000) performs coexistence estimation for multiple other devices may be as follows.

[0094] Among other devices, if the number of devices estimated to exist in the same space as the user device (2000) is 1, the user device (2000) may configure the second Wi-Fi RSSI and the first Wi-Fi RSSI of the devices estimated to exist in the same space as the user device (2000) as a positive pair, and configure the second Wi-Fi RSSI and the first Wi-Fi RSSI of the remaining devices as a negative pair and include them in the filtered data set (540).

[0095] Among other devices, if the number of devices estimated to exist in the same space as the user device (2000) is 2 or more, the user device (2000) may classify this as a potential error case and include the first Wi-Fi RSSI and the second Wi-Fi RSSI used for inference in the preliminary data set (560).

[0096] Among other devices, if the number of devices estimated to exist in the same space as the user device (2000) is 0, the user device (2000) can filter the first Wi-Fi RSSI and the second Wi-Fi RSSI used for inference.

[0097] For example, referring to the first example (552) of the estimation result, the estimation result indicates that the user device (2000) exists in the same space as device A, and in a different space from device B and device C. In this case, since devices A, B, and C are each in different spaces, the estimation result data of the first example (552) is reliable good data. That is, it means that the user device (2000) exists in the same space as device A, and in a different space from device B and device C, and the inference of the coexistence estimation model is accurate. Accordingly, the user device (2000) can include good data pairs in the filtered dataset (540) by configuring the first Wi-Fi RSSI and the second Wi-Fi RSSI corresponding to device A as a positive pair, and the first Wi-Fi RSSI and the second Wi-Fi RSSI corresponding to device B, and the first Wi-Fi RSSI and the second Wi-Fi RSSI corresponding to device C as a negative pair.

[0098] Also, for example, referring to the second example (554) of the estimation result, the estimation result indicates that the user device (2000) exists in the same space as devices A and B, and in a different space from device C. In this case, since devices A, B, and C are in different spaces, the estimation result data of the second example (554) includes an error, which indicates a potential error that may occur in the user's actual usage environment. In this case, the user device (2000) may include the first Wi-Fi RSSI and the second Wi-Fi RSSI used for inference in the preliminary dataset (560). In one embodiment, as the user device (2000) updates the coexistence estimation model, the performance of the coexistence estimation model may be improved. Accordingly, at some point after continuous updates, the data included in the preliminary dataset (560) may also be good data that can be used to update the coexistence estimation model. After the coexistence estimation model is updated, the user device (2000) can re-examine the data of the preliminary dataset (560). The user device (2000) can perform a coexistence estimation task using the data of the preliminary dataset (560). In this case, if the number of devices estimated to exist in the same space as the user device (2000) is 1, the user device (2000) can include the first Wi-Fi RSSI and the second Wi-Fi RSSI used for inference in the filtered dataset (540).

[0099] Also, referring to the third example (554) of the estimation result, for example, the estimation result indicates that the user device (2000) exists in a different space from device A, device B, and device C. That is, if there is no device estimated to exist in the same space as in the third example (554), the user device (2000) does not use the first Wi-Fi RSSI and the second Wi-Fi RSSI used for inference to update the coexistence estimation model.

[0100] By generating a filtered dataset (540), the user device (2000) can secure positive pairs of the first Wi-Fi RSSI and the second Wi-Fi RSSI, which can be used for fine-tuning the coexistence estimation model. By updating the coexistence estimation model using the filtered dataset (540), the user device (2000) can improve the performance of the spatial estimation model for the user space.

[0101] Meanwhile, in FIG. 5b, although the number of other devices other than the user device (2000) is exemplified as three, the number of other devices is not limited thereto. In other words, the number of other devices may be two or more. If the number of devices estimated to exist in the same space as the user device (2000) is one among the devices, the user device (2000) can secure a positive pair of the first Wi-Fi RSSI and the second Wi-Fi RSSI.

[0102] FIG. 6 is a diagram illustrating a training dataset generated by a user device according to one embodiment of the present disclosure.

[0103] In one embodiment, the user device (2000) can generate a training dataset (600). The user device (2000) can use the training dataset (600) to fine-tune a coexistence estimation model (610).

[0104] The training dataset (600) may be composed of a filtered dataset (602) and a feedback dataset (604).

[0105] The filtered dataset (602) may include data pairs {first Wi-Fi RSSI, second Wi-Fi RSSI} filtered according to preset conditions from among the pre-collected Wi-Fi RSSI. The data pairs may be positive pairs indicating the same space or negative pairs indicating different spaces. The operation of the user device (2000) generating the filtered dataset (602) has been described above in the description of FIGS. 5A and 5B , and therefore, a repeated description thereof will be omitted.

[0106] The feedback dataset (604) may include {first Wi-Fi RSSI, second Wi-Fi RSSI} data pairs generated based on the feedback. The data pairs may be positive pairs indicating the same space or negative pairs indicating different spaces. The operation of the user device (2000) generating the feedback dataset (604) has been described above with reference to FIGS. 4A and 4B , and therefore, a repeated description thereof will be omitted.

[0107] In order for the user device (2000) to fine-tune the coexistence estimation model (610), a sufficient amount of data is required in the training dataset (600). However, feedback data, which is high-quality data that can be used for training, is collected as the user device (2000) is used, so the amount of data is smaller than that of data that is continuously collected, and it takes a lot of time to collect a sufficient amount of data. In addition to the feedback data, the user device (2000) can quickly reflect the feedback data in the coexistence estimation model (610) by continuously filtering data collected in advance and including it in the training dataset (600).

[0108] In one embodiment, the operation of the user device (2000) to generate the training dataset (600) may be initiated based on feedback received by the user device (2000). For example, after the user uses the user device (2000) to estimate whether another device is present in the same space as the user device, feedback on the estimation result may be received from the user. In this case, the user device (2000) may create a training dataset (600) including the feedback dataset (604) and the filtered dataset (602) in response to receiving the feedback, and update the coexistence estimation model. This is further described with reference to FIGS. 7A and 7B .

[0109] FIG. 7A is a flowchart illustrating a method by which a user device operates when receiving feedback according to one embodiment of the present disclosure.

[0110] In operation S710, the user device (2000) can continuously collect and store Wi-Fi RSSI (first Wi-Fi RSSI). The first Wi-Fi RSSI can be collected periodically or aperiodically when the user device (2000) is in a standby or active state.

[0111] In operation S715, another device (700) can continuously collect and store Wi-Fi RSSI (second Wi-Fi RSSI). There may be one or more other devices (700). The second Wi-Fi RSSI may be collected periodically or aperiodically when the other device (700) is in a standby or active state. If there are multiple other devices (700), each device can continuously collect Wi-Fi RSSI. For example, different devices A and B can each continuously collect second Wi-Fi RSSI indicating surrounding Wi-Fi networks.

[0112] In operation S720, the user device (2000) can identify that feedback is received.

[0113] In one embodiment, operations S210 to S230 of FIG. 2 may be performed before operation S720 is performed. In other words, the user device (2000) may input the first Wi-Fi RSSI of the user device (2000) and the second Wi-Fi RSSI of the other device (700) into a coexistence estimation model to estimate whether the other device (700) exists in the same space as the user device (2000). The user device (2000) may receive a user input for providing feedback on the estimation result. If no feedback is received, the user device (2000) may perform operation S710 again. If feedback is received, the user device (2000) may perform operations S730 and S740.

[0114] In one embodiment, when feedback is received, the user device (2000) may generate a feedback dataset based on the feedback. The feedback dataset may include feedback data consisting of one or more {first Wi-Fi RSSI, second Wi-Fi RSSI}.

[0115] In operation S730, the user device (2000) may request a second Wi-Fi RSSI from another device (700) based on the received feedback. If there are multiple other devices (700), the user device (2000) may request the second Wi-Fi RSSI from each of the multiple other devices (700).

[0116] In operation S735, the other device (700) may identify a data request from the user device (2000). If the data request is not identified, the other device (700) may perform operation S715 again. If the data request is identified, the other device (700) may transmit the pre-collected second Wi-Fi RSSI to the user device (2000) (operation S745). The pre-collected second Wi-Fi RSSI may be the one collected in operation S715.

[0117] In operation S740, the user device (2000) may obtain a pre-collected first Wi-Fi RSSI. The pre-collected first Wi-Fi RSSI may be collected in operation S710.

[0118] In operation S750, the user device (2000) may generate a filtered dataset through filtering based on the inferred results of the coexistence estimation model. The user device (2000) may generate the filtered dataset by filtering the pre-collected first Wi-Fi RSSI and the pre-collected second Wi-Fi RSSI. Specific operations included in operation S750 will be further described with reference to FIG. 7B.

[0119] In operation S760, the user device (2000) may integrate the filtered dataset and the feedback dataset. The user device (2000) may generate a training dataset including the filtered dataset and the feedback dataset.

[0120] In operation S770, the user device (2000) may further update the coexistence estimation model. The user device (2000) may fine-tune the coexistence estimation model using the newly generated training dataset, thereby optimizing the space estimation model to suit the user's space. After updating the coexistence estimation model, the user device (2000) may perform operation S710 again to collect the first Wi-Fi RSSI. If feedback is received again thereafter, the user device (2000) may perform operations S720 to S770 to update the coexistence estimation model again.

[0121] In one embodiment, the user device (2000) may iterate over generating a training dataset and updating the coexistence estimation model each time feedback is received.

[0122] In one embodiment, the user device (2000) may repeatedly generate a training dataset and update the coexistence estimation model based on preset conditions. For example, the user device (2000) may update the coexistence estimation model every time a preset number of N feedbacks are received. However, the preset conditions are not limited thereto, and various conditions triggered by the receipt of feedback may be defined as preset conditions.

[0123] Figure 7b is a flowchart for explaining the operation of Figure 7a in more detail.

[0124] Operations S751 to S756 of FIG. 7b describe detailed operations of operation S750 of FIG. 7a. Accordingly, operations S710 to S740 of FIG. 7a may be performed before operation S751 is performed.

[0125] In operation S751, the user device (2000) creates a data pair consisting of a first Wi-Fi RSSI and one of the second Wi-Fi RSSIs of each of the other devices.

[0126] The user device (2000) may obtain a pre-collected first Wi-Fi RSSI and a pre-collected second Wi-Fi RSSI of other devices, and create a data pair of the first Wi-Fi RSSI and the second Wi-Fi RSSI. For example, the user device (2000) may create a data pair {first Wi-Fi RSSI, second Wi-Fi RSSI A} that matches the Wi-Fi RSSI of the user device (2000) and the Wi-Fi RSSI of device A, and may create a data pair {first Wi-Fi RSSI, second Wi-Fi RSSI B} that matches the Wi-Fi RSSI of the user device (2000) and the Wi-Fi RSSI of device B.

[0127] In operation S752, the user device (2000) may input a data pair into a coexistence estimation model to perform an inference operation. For example, the user device (2000) may input a data pair {first Wi-Fi RSSI, second Wi-Fi RSSI A} into the coexistence estimation model to perform an inference operation to estimate whether device A exists in the same space as the user device (2000).

[0128] In operation S753, the user device (2000) checks whether inference is completed for all data pairs. For example, inference may have been performed for the data pair {first Wi-Fi RSSI, second Wi-Fi RSSI A}, but inference may not have been performed for the data pair {first Wi-Fi RSSI, second Wi-Fi RSSI B}. In this case, the user device (2000) may perform operation S752 again to estimate whether device B exists in the same space as the user device (2000). The user device (2000) may repeat operation S752 until inference operations for all other devices are completed. When inference is completed for all data pairs and coexistence estimation results for all other devices are obtained, the user device (2000) may perform operation S754.

[0129] In operation S754, the user device (2000) can identify whether only one data pair is in the same space as the inference result. If only one data pair is in the same space, the user device (2000) can perform operation S755 to store the data pairs used for inference in a filtered dataset. For example, it may be estimated that device A exists in the same space as the user device (2000) and device B exists in a different space from the user device (2000). In this case, the user device (2000) can include the data pairs used for inference in the filtered dataset. In other cases, the user device (2000) can perform operation S756 to store the data pairs used for inference in a reserve dataset. For example, it may be estimated that both device A and device B exist in the same space as the user device (2000). In this case, the user device (2000) can filter the data pairs used for inference and include them in the reserve dataset. The data pairs included in the preliminary dataset are not directly used to update the current coexistence estimation model, but may be used to update the coexistence estimation model in the future. For example, after the coexistence estimation model of the user device (2000) is updated, the coexistence estimation task can be performed using the data pairs included in the preliminary dataset. In this case, if the number of devices estimated to exist in the same space as the user device (2000) is 1, the user device (2000) can include the data pairs used for inference in the filtered dataset.

[0130] FIG. 8 is a diagram illustrating an operation of a user device generating a training dataset according to one embodiment of the present disclosure.

[0131] In explaining FIG. 8, it is explained as an example that device A exists in room A, device B exists in room C, device C exists in room C, and the user device (2000) is a mobile device.

[0132] If the user device (2000) is a mobile device, the amount of data collected may vary by device. For example, if the user device (2000) is present in room B for a long time, a relatively large number of data pairs of the user device (2000) and device B may be collected. Accordingly, even in the filtered dataset (800) created by the user device (2000) by filtering the first Wi-Fi RSSI of the user device (2000) collected in advance and the second Wi-Fi RSSI collected in advance according to preset conditions, there may be a deviation in the amount of data by device. In order to reduce the data deviation within the training dataset when updating the coexistence estimation model, the user device (2000) may sample data of the filtered dataset (800).

[0133] In one embodiment, the user device (2000) may classify a plurality of second Wi-Fi RSSIs that match the first Wi-Fi RSSI of the user device by device. For example, the user device (2000) may classify the data pairs included in the filtered data set (800) into a data set of device A (810), a data set of device B (820), and a data set of device C (830). As a result of the classification, the data set (810) of device A may include data pairs in which the first Wi-Fi RSSI of the user device (2000) and the second Wi-Fi RSSI of device A are matched, the data set (820) of device B may include data pairs in which the first Wi-Fi RSSI of the user device (2000) and the second Wi-Fi RSSI of device B are matched, and the data set (830) of device C may include data pairs in which the first Wi-Fi RSSI of the user device (2000) and the second Wi-Fi RSSI of device C are matched.

[0134] The user device (2000) may sample data pairs to be included in the filtered dataset based on the classification result. For example, the user device (2000) may sample data based on the median number of data pairs matched for each device. Specifically, if the dataset (810) of device A includes two data pairs, the dataset (820) of device B includes eight data pairs, and the dataset (830) of device C includes one data pair, the user device (2000) may set the median value, 2, as the number of samples. Accordingly, the sampled dataset (812) of device A and the sampled dataset (822) of device B each include two data pairs, and the sampled dataset (832) of device C includes one data pair. Since the dataset (830) of device C has fewer data pairs than the number of samples, all data is used. Meanwhile, the standard by which the user device (2000) samples data is not limited to the median of the number of data pairs. For example, the average value, preset values, etc. may also be used.

[0135] The user device (2000) can reduce the deviation of data included in the filtered data set (800) by sampling data pairs included in the filtered data set (800). The filtered data set (800) can be integrated with a feedback data set to create a training data set for updating a coexistence estimation model.

[0136] FIG. 9 is a diagram illustrating an operation of a user device generating a training dataset according to one embodiment of the present disclosure.

[0137] In one embodiment, the user device (2000) can generate a filtered dataset by filtering the pre-collected first Wi-Fi RSSI and the pre-collected plurality of second Wi-Fi RSSIs according to preset conditions. In this case, unlike the example in FIG. 5B, two or more other devices may exist in the same space. For example, in the example of FIG. 9, it is described as an example that Device A, Device B, and Device C exist in Room A, Device D exists in Room B, and the user device (2000) is a mobile device.

[0138] In one embodiment, the user device (2000) may perform an inference operation (900) of the coexistence estimation model by applying a pre-collected first Wi-Fi RSSI and a plurality of pre-collected second Wi-Fi RSSIs as input data to the coexistence estimation model.

[0139] For example, the user device (2000) can estimate whether device A exists in the same space as the user device (2000) by applying the pre-collected first Wi-Fi RSSI and the pre-collected second Wi-Fi RSSI of device A to a coexistence estimation model. In addition, the user device (2000) can estimate whether device B, device C, and device D exist in the same space as the user device (2000) by using the pre-collected first Wi-Fi RSSI and the pre-collected second Wi-Fi RSSI of each of device B, device C, and device D.

[0140] The user device (2000) may filter data based on the estimation results of the coexistence estimation model. When two or more different devices exist in a space, the user device (2000) may determine whether to filter the first Wi-Fi RSSI collected in advance and the second Wi-Fi RSSI collected in advance based on information about the different devices and the number of devices estimated to exist in the same space as the user device (2000).

[0141] For example, the device information of multiple different devices, Device A, Device B, and Device C, may include information that the devices are in room A.

[0142] Referring to the first example (910) of the estimation result, the estimation result of the first example (910) indicates that the user device (2000) exists in the same space as devices A, B, and C, and in a different space from device D. In this case, since there is information that devices A, B, and C exist in the same space, the estimation result data of the first example (910) is reliable good data. That is, it means that the user device (2000) exists in the same space as devices A, B, and C, and in a different space from device D, and the inference of the coexistence estimation model is accurate. Accordingly, the user device (2000) can include the {first Wi-Fi RSSI, second Wi-Fi RSSI} data pairs used for the inference in the first example (910) in the filtered dataset.

[0143] Referring to the second example (920) of the estimation result, the estimation result of the second example (920) indicates that the user device (2000) exists in a different space from devices A, B, and C, and exists in the same space as device D. In this case, since there is information that devices A, B, and C exist in the same space, the estimation result data of the second example (920) is reliable good data. That is, it means that the user device (2000) exists in a different space from devices A, B, and C, and exists in the same space as device D, and the inference of the coexistence estimation model is accurate. Accordingly, the user device (2000) can include the {first Wi-Fi RSSI, second Wi-Fi RSSI} data pairs used for the inference in the second example (920) in the filtered dataset.

[0144] Referring to the third example (930) of the estimation result, the estimation result of the third example (930) indicates that the user device (2000) exists in the same space as devices A and B, and in a different space from devices C and D. In this case, since there is information that devices A, device B, and device C exist in the same space, the estimation result data of the third example (930) is data that accurately estimates more than half (two out of three). Accordingly, the user device (2000) can include the {first Wi-Fi RSSI, second Wi-Fi RSSI} data pairs used for inference in the third example (930) in the filtered dataset.

[0145] Referring to the fourth example (940) and the fifth example (950) of the estimation results, the estimation result data of the fourth example (940) is data that is accurately estimated less than half (1 out of 3), and the estimation result data of the fifth example (950) is considered unusable data because it is estimated that devices A and D, which exist in different spaces, exist in the same space. Therefore, the data pairs used for the inference of the fourth example (940) and the fifth example (950) may not be included in the filtered dataset.

[0146] In one embodiment, the user device (2000) may assign a weight for sampling to the first Wi-Fi RSSI-second Wi-Fi RSSI based on the estimation result of the coexistence estimation model when the number of devices located in the same space among other devices is two or more.

[0147] For example, referring to the first example (910) of the estimation result, the estimation result in the first example (910) is an accurate estimation result, and there are three positive pairs inferred that the user device (2000) is in the same space as device A, device B, and device C. Accordingly, the user device (2000) can assign a first weight (e.g., 2), which is a relatively high weight, to the {first Wi-Fi RSSI, second Wi-Fi RSSI} data pairs used in the inference in the first example (910).

[0148] For example, referring to the second example (920) of the estimation result, the estimation result in the second example (920) is an accurate estimation result, and there is one positive pair inferred that the user device (2000) is in the same space as the device D. Therefore, the user device (2000) can assign a second weight (e.g., 1) that is a lower weight than the first weight to the {first Wi-Fi RSSI, second Wi-Fi RSSI} data pairs used in the inference in the second example (920).

[0149] For example, referring to the third example (930) of the estimation result, the estimation result in the third example (930) is a result in which more than half (2 out of 3) are accurately estimated. In this case, since the estimation results are not all accurate, the user device (2000) may assign a third weight (e.g., 0.66) that is lower than the first and second weights to the {first Wi-Fi RSSI, second Wi-Fi RSSI} data pairs used for inference in the third example (930).

[0150] In one embodiment, weights may be utilized when the user device (2000) generates a filtered dataset. For example, the user device (2000) may sample data based on the weights. Specifically, the user device (2000) may sort data pairs in order of weights and sample N data pairs with high weights to construct a filtered dataset. Alternatively, the user device (2000) may construct a filtered dataset by sampling data pairs in proportion to the weights.

[0151] FIG. 10 is a diagram illustrating an operation of a user device using a coexistence estimation model according to one embodiment of the present disclosure.

[0152] In one embodiment, the user device (2000) can control multiple other devices. For example, the user device (2000) can interact with other devices within the user's home. The other devices within the home and the user device (2000) may be in communication connection. For example, the user device (2000) and the other devices may be connected to a home network or to a server (e.g., a cloud server).

[0153] In operation S1010, the user device (2000) may execute a home device control function. The user device (2000) may receive a user input for executing a function for controlling a home device. The home device control function may be provided through an application (1010). For example, multiple home devices registered to the application may be controlled through a smart home application.

[0154] In operation S1020, the user device (2000) may display an inference result using a coexistence estimation model. The user device (2000) may estimate a device existing in the same space as the current location of the user device (2000) and display the estimation result on the screen. For example, the user device (2000) may display a device identified as existing in the same space as the user device (2000) and information about the device. For example, if the device identified as currently existing in the same space as the user is device A, and information about device A is registered as being in room A, the user device (2000) may display on the screen a device inferred to be in the same space as the user. The user device (2000) may control the operation of another device based on a user input for selecting another device displayed on the screen. For example, if the other device is a TV, operations such as turning the TV on / off, changing channels, adjusting the volume, and executing an application may be performed through the user device (2000).

[0155] In operation S1030, the user device (2000) may obtain user feedback. In one embodiment, the result inferred using the coexistence estimation model in operation S1020 may be inaccurate. In this case, the user may input feedback to the user device (2000). For example, the user device (2000) may provide a user interface that can evaluate the accuracy / inaccuracy of the inference result. The user device (2000) may store feedback information based on the user input. In one embodiment, the user device (2000) may also receive user input for inputting location information of the device from the user.

[0156] In operation S1040, the user device (2000) may generate a feedback dataset. The operation of the user device (2000) generating a feedback dataset has been described above in the description of the previous drawings, and therefore, a repeated description is omitted for brevity.

[0157] In operation S1045, the user device (2000) may generate a filtered dataset. The filtered dataset may be generated by filtering pre-collected Wi-Fi RSSI data. The operation of the user device (2000) generating the filtered dataset has been described above in the description of the previous drawings, and therefore, a repeated description is omitted for brevity.

[0158] In operation S1050, the user device (2000) may update the coexistence estimation model. The user device (2000) may fine-tune the coexistence estimation model using a training dataset consisting of a feedback dataset and a filtered dataset. The operation of the user device (2000) updating the coexistence estimation model has been described above in the description of the previous drawings, and therefore, a repeated description is omitted for brevity.

[0159] FIG. 11 is a block diagram illustrating a configuration of a user device according to one embodiment of the present disclosure.

[0160] In one embodiment, the user device (2000) may include a communication interface (2100), memory (2200), and a processor (2300).

[0161] The communication interface (2100) may include a communication circuit. The communication interface (2100) may use at least one of data communication methods including, for example, wired LAN, wireless LAN, Wi-Fi, Bluetooth, ZigBee, Wi-Fi Direct (WFD), infrared Data Association (IrDA), Bluetooth Low Energy (BLE), Near Field Communication (NFC), Wireless Broadband Internet (Wibro), World Interoperability for Microwave Access (WiMAX), Shared Wireless Access Protocol (SWAP), Wireless Gigabit Alliances (WiGig), and RF communication. The communication interface (2100) may be implemented to include a plurality of modules (for example, a Wi-Fi module, etc.) for implementing the above-described communication methods.

[0162] The communication interface (2100) can transmit and receive data for performing operations of the user device (2000) with other devices. For example, the user device (2000) can transmit and receive various data (device information, Wi-Fi RSSI information) used by the user device (2000) to interact with other devices with other devices through the communication interface (2100).

[0163] The memory (2200) may store instructions, data structures, and program codes that can be read by the processor (2300). There may be one or more memories (2200). In the disclosed embodiments, operations performed by the processor (2300) may be implemented by executing instructions or codes of a program stored in the memory (2200).

[0164] The memory (2200) may include non-volatile memory such as read-only memory (ROM) (e.g., programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM)), flash memory (e.g., memory card, solid-state drive (SSD)), and analog recording type (e.g., hard disk drive (HDD), magnetic tape, optical disk), and volatile memory such as random-access memory (RAM) (e.g., dynamic random-access memory (DRAM), static random-access memory (SRAM)).

[0165] The memory may store data, instructions, and programs that enable the user device (2000) to operate. For example, the memory (2200) may store a coexistence estimation module (2210), a training data management module (2220), and a device management module (2230).

[0166] The processor (2300) can control the overall operations of the user device (2000). For example, the processor (2300) can control the overall operations of the user device (2000) by executing one or more instructions of a program stored in the memory (2200). There may be one or more processors (2300).

[0167] The processor (2300) may execute a coexistence estimation module (2210) to estimate whether the user device (2000) and another device exist in the same space. The coexistence estimation module (2210) may include a coexistence estimation model and data and codes for operating the coexistence estimation model. Since the operations by which the user device (2000) utilizes the coexistence estimation model have been described above, a repeated description is omitted for brevity.

[0168] The processor (2300) may execute the training data management module (2220) to generate a training dataset. The training data management module may collect the Wi-Fi RSSI of the user device (2000) and the Wi-Fi RSSI of other devices. The training data management module (2220) may generate a feedback dataset based on feedback received from the user. In addition, the training data management module (2220) may continuously filter pre-collected data to generate a filtered dataset. When the filtered dataset is generated, the coexistence estimation module (2210) may be used to obtain inference data of the pre-collected data. Since the operations for generating the training data by the user device (2000) have been described above, a repeated description is omitted for brevity.

[0169] The processor (2300) can manage multiple devices by executing the device management module (2230). The device management module (2230) can transmit and receive data for interacting with the devices. The device management module (2230) can transmit user input inputted to the user device (2000) to another device, transmit control commands for controlling another device to the other device, and receive device information from the other device. Since the operations for the user device (2000) to interact with other devices have been described above, a repeated description is omitted for brevity.

[0170] Meanwhile, the modules stored in the aforementioned memory (2200) are for convenience of explanation and are not necessarily limited thereto. Other modules may be added to implement the aforementioned embodiments, some of the aforementioned modules may be implemented as a single module, or a single model of the aforementioned modules may be implemented by separating them into multiple modules.

[0171] In one embodiment, the one or more processors (2300) may include at least one of a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), an Accelerated Processing Unit (APU), a Many Integrated Core (MIC), a Digital Signal Processor (DSP), and a Neural Processing Unit (NPU). The one or more processors (2300) may be implemented in the form of an integrated system on a chip (SoC) including one or more electronic components. Each of the one or more processors (2300) may also be implemented as separate hardware (H / W).

[0172] When a method according to an embodiment of the present disclosure includes a plurality of operations, the plurality of operations may be performed by one processor (2300) or by a plurality of processors (2300). For example, when a first operation, a second operation, and a third operation are performed by a method according to an embodiment, the first operation, the second operation, and the third operation may all be performed by a first processor, or the first operation and the second operation may be performed by a first processor (e.g., a general-purpose processor) and the third operation may be performed by a second processor (e.g., an AI-only processor). Here, an example of the second processor may be an AI-only processor, and the AI-only processor may perform operations for training / inference of an AI model. However, the embodiments of the present disclosure are not limited thereto.

[0173] One or more processors (2300) according to the present disclosure may be implemented as a single-core processor or as a multi-core processor.

[0174] When a method according to one embodiment of the present disclosure includes multiple operations, the multiple operations may be performed by one core or may be performed by multiple cores included in one or more processors (2300).

[0175] Meanwhile, although not illustrated in FIG. 11, the user device (2000) may further include additional components to perform the operations of the embodiments of the present disclosure. For example, the user device (2000) may further include a display, a camera, a microphone, an input / output interface, and the like.

[0176] The present disclosure proposes a user device and an operating method thereof capable of estimating whether the user device exists in the same space as another device in an environment where the user device interacts with multiple other devices. Furthermore, the present disclosure proposes a method for generating a training dataset for personalizing a coexistence estimation model and utilizing the training dataset to ensure that the coexistence estimation model operates well in the user's environment. The technical problems to be achieved in the present disclosure are not limited to those mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art to which the present invention pertains from the description herein.

[0177] According to one aspect of the present disclosure, a method of operating a user device may be provided. The method may include obtaining a first Wi-Fi RSSI corresponding to Wi-Fi networks nearby the user device.

[0178] The method may include obtaining a plurality of second Wi-Fi RSSIs corresponding to nearby Wi-Fi networks of each of the other devices.

[0179] The method may include a step of applying the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs as input data to a copresence estimation model to estimate whether the other devices exist in the same space as the user device.

[0180] The method may include a step of receiving feedback on the estimation result. The method may include a step of creating a feedback dataset, comprising the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs, based on the feedback.

[0181] The method may include a step of obtaining a pre-collected first Wi-Fi RSSI and a plurality of pre-collected second Wi-Fi RSSIs.

[0182] The method may include a step of creating a filtered dataset by filtering the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs according to preset conditions.

[0183] The method may include a step of updating the coexistence estimation model using the feedback dataset and the filtered dataset.

[0184] The feedback may include information indicating that the estimation result is inaccurate and location information of the user device.

[0185] The step of creating the feedback dataset may include a step of configuring the second Wi-Fi RSSI and the first Wi-Fi RSSI of a device estimated to exist in the same space as the user device among the other devices as a negative pair based on the feedback.

[0186] The step of creating the feedback dataset may include a step of configuring a second Wi-Fi RSSI and the first Wi-Fi RSSI of a device whose location information corresponds to the location information of the user device among the other devices as a positive pair based on the feedback.

[0187] The step of creating the filtered dataset may include the step of applying the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs as input data to the coexistence estimation model.

[0188] The step of creating the filtered dataset may include a step of determining whether to filter the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs based on the number of devices estimated to exist in the same space as the user device among the other devices.

[0189] The step of creating the filtered dataset may include, when the number of devices estimated to exist in the same space as the user device among the other devices is 1, configuring the second Wi-Fi RSSI of the device estimated to exist in the same space as the user device and the first Wi-Fi RSSI as a positive pair, and configuring the second Wi-Fi RSSI and the first Wi-Fi RSSI of the remaining devices as a negative pair.

[0190] The step of creating the filtered dataset may include the step of classifying, by device, the plurality of pre-collected second Wi-Fi RSSIs of the other devices that match the pre-collected first Wi-Fi RSSI of the user device.

[0191] The step of creating the filtered dataset may include a step of sampling a first Wi-Fi RSSI-second Wi-Fi RSSI pair to be included in the filtered dataset based on the classification result.

[0192] The step of creating the above filtered dataset may include a step of assigning a weight for sampling to the first Wi-Fi RSSI-the second Wi-Fi RSSI based on the estimation result of the coexistence estimation model, when the number of devices located in the same space among the other devices is two or more.

[0193] The step of updating the coexistence estimation model may include a step of assigning a higher weight to the feedback dataset than to the filtered dataset.

[0194] The step of updating the coexistence estimation model may include the step of updating the coexistence estimation model using the weights.

[0195] According to one aspect of the present disclosure, a user device may be provided. The user device may include: a communication interface; a memory storing one or more instructions; and one or more processors executing the one or more instructions stored in the memory.

[0196] The one or more processors can obtain a first Wi-Fi RSSI corresponding to nearby Wi-Fi networks of the user device by executing the one or more instructions.

[0197] The one or more processors can obtain a plurality of second Wi-Fi RSSIs corresponding to nearby Wi-Fi networks of each of the other devices by executing the one or more instructions.

[0198] The one or more processors can estimate whether the other devices exist in the same space as the user device by applying the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs as input data to a copresence estimation model by executing the one or more instructions.

[0199] The one or more processors can receive feedback on the estimation result by executing the one or more instructions.

[0200] The one or more processors can create a feedback dataset comprising the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs based on the feedback by executing the one or more instructions.

[0201] The one or more processors can obtain a pre-collected first Wi-Fi RSSI and a plurality of pre-collected second Wi-Fi RSSIs by executing the one or more instructions.

[0202] The one or more processors can create a filtered dataset by filtering the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs according to preset conditions by executing the one or more instructions.

[0203] The one or more processors can update the coexistence estimation model using the feedback dataset and the filtered dataset by executing the one or more instructions.

[0204] The feedback may include information indicating that the estimation result is inaccurate and location information of the user device.

[0205] The one or more processors may, by executing the one or more instructions, configure the second Wi-Fi RSSI and the first Wi-Fi RSSI of a device estimated to exist in the same space as the user device among the other devices as a negative pair based on the feedback.

[0206] The one or more processors can, by executing the one or more instructions, positively pair the second Wi-Fi RSSI and the first Wi-Fi RSSI of a device among the other devices whose location information corresponds to the location information of the user device based on the feedback.

[0207] The one or more processors can apply the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs as input data to the coexistence estimation model by executing the one or more instructions.

[0208] The one or more processors may determine, by executing the one or more instructions, whether to filter the pre-collected first Wi-Fi RSSI and the plurality of pre-collected second Wi-Fi RSSIs based on the number of devices estimated to exist in the same space as the user device among the other devices.

[0209] The one or more processors, by executing the one or more instructions, can configure the second Wi-Fi RSSI of the device estimated to exist in the same space as the user device and the first Wi-Fi RSSI as a positive pair when the number of devices estimated to exist in the same space as the user device among the other devices is 1, and configure the second Wi-Fi RSSI and the first Wi-Fi RSSI of the remaining devices as a positive pair.

[0210] The one or more processors can classify, by device, the plurality of pre-collected second Wi-Fi RSSIs of the other devices that match the pre-collected first Wi-Fi RSSI of the user device by executing the one or more instructions.

[0211] The one or more processors can sample a first Wi-Fi RSSI-second Wi-Fi RSSI pair to be included in the filtered dataset based on the classification result by executing the one or more instructions.

[0212] The one or more processors may, by executing the one or more instructions, assign weights for sampling to the first Wi-Fi RSSI-second Wi-Fi RSSI based on the estimation result of the coexistence estimation model when the number of devices located in the same space among the other devices is two or more.

[0213] The one or more processors may, by executing the one or more instructions, assign a higher weight to the feedback dataset than to the filtered dataset.

[0214] The one or more processors can update the coexistence estimation model using the weights by executing the one or more instructions.

[0215] Meanwhile, embodiments of the present disclosure may also be implemented in the form of a recording medium containing computer-executable instructions, such as program modules, executed by a computer. Computer-readable media may be any available media that can be accessed by a computer, and include both volatile and nonvolatile media, removable and non-removable media. Furthermore, computer-readable media may include computer storage media and communication media. Computer storage media include both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. Communication media may typically include computer-readable instructions, data structures, or other data in a modulated data signal, such as program modules.

[0216] Additionally, a computer-readable storage medium may be provided in the form of a non-transitory storage medium. Here, the term "non-transitory storage medium" simply means a tangible device that does not contain signals (e.g., electromagnetic waves). This term does not distinguish between cases where data is permanently stored in the storage medium and cases where data is temporarily stored. For example, a "non-transitory storage medium" may include a buffer in which data is temporarily stored.

[0217] According to one embodiment, the method according to various embodiments disclosed in the present document may be provided as included in a computer program product. The computer program product may be traded as a product between a seller and a buyer. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., compact disc read-only memory (CD-ROM)), or may be distributed online (e.g., downloaded or uploaded) through an application store or directly between two user devices (e.g., smartphones). In the case of online distribution, at least a portion of the computer program product (e.g., a downloadable app) may be temporarily stored or temporarily generated in a machine-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or an intermediary server.

[0218] The above description of the present disclosure is provided for illustrative purposes only, and those skilled in the art will readily appreciate that modifications to other specific forms can be made without altering the technical spirit or essential features of the present disclosure. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, components described as being single may be implemented in a distributed manner, and similarly, components described as being distributed may be implemented in a combined manner.

[0219] The scope of the present disclosure is indicated by the claims described below rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present disclosure.

Claims

1. In terms of the operation method of the user device, A step of obtaining a first Wi-Fi RSSI corresponding to nearby Wi-Fi networks of the user device; A step of obtaining a plurality of second Wi-Fi RSSIs corresponding to nearby Wi-Fi networks of each of the other devices; A step of applying the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs as input data to a copresence estimation model to estimate whether the other devices exist in the same space as the user device; A step of receiving feedback on the above estimation results; A step of creating a feedback dataset, comprising the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs based on the feedback; A step of acquiring a pre-collected first Wi-Fi RSSI and a plurality of pre-collected second Wi-Fi RSSI; A step of creating a filtered dataset by filtering the pre-collected first Wi-Fi RSSI and the pre-collected plurality of second Wi-Fi RSSI according to preset conditions; and A method comprising the step of updating the coexistence estimation model using the feedback dataset and the filtered dataset.

2. In paragraph 1, A method wherein the feedback includes information indicating that the estimation result is inaccurate and location information of the user device.

3. In paragraph 2, The steps to create the above feedback dataset are: A method comprising the step of configuring a second Wi-Fi RSSI and the first Wi-Fi RSSI of a device estimated to exist in the same space as the user device among the other devices as a negative pair based on the feedback.

4. In paragraph 2, The steps to create the above feedback dataset are: A method comprising: configuring a second Wi-Fi RSSI and the first Wi-Fi RSSI of a device among the other devices, the location information of which corresponds to the location information of the user device, as a positive pair based on the feedback.

5. In paragraph 1, The steps to create the above filtered dataset are: A step of applying the pre-collected first Wi-Fi RSSI and the pre-collected plurality of second Wi-Fi RSSIs as input data to the coexistence estimation model; and A method comprising: determining whether to filter the pre-collected first Wi-Fi RSSI and the pre-collected plurality of second Wi-Fi RSSIs based on the number of devices estimated to exist in the same space as the user device among the other devices.

6. In paragraph 5, The steps to create the above filtered dataset are: A method comprising: when the number of devices estimated to exist in the same space as the user device among the other devices is 1, configuring the second Wi-Fi RSSI and the first Wi-Fi RSSI of the device estimated to exist in the same space as the user device as a positive pair, and configuring the second Wi-Fi RSSI and the first Wi-Fi RSSI of the remaining devices as a negative pair.

7. In paragraph 5, The steps to create the above filtered dataset are: A step of classifying the plurality of pre-collected second Wi-Fi RSSIs of the other devices that match the pre-collected first Wi-Fi RSSI of the user device by device; and A method comprising the step of sampling a first Wi-Fi RSSI-second Wi-Fi RSSI pair to be included in the filtered dataset based on the classification result.

8. On the user device, communication interface; Memory that stores one or more instructions; and comprising one or more processors executing one or more instructions stored in said memory; The one or more processors, by executing the one or more instructions, Obtaining a first Wi-Fi RSSI corresponding to nearby Wi-Fi networks of the user device, Acquire multiple secondary Wi-Fi RSSIs corresponding to nearby Wi-Fi networks of each of the other devices, By applying the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs as input data to a copresence estimation model, it is estimated whether the other devices exist in the same space as the user device, Receive feedback on the above estimation results, Based on the above feedback, a feedback dataset is created, which consists of the first Wi-Fi RSSI and the plurality of second Wi-Fi RSSIs, Obtaining pre-collected first Wi-Fi RSSI and multiple pre-collected second Wi-Fi RSSI, A filtered dataset is created by filtering the pre-collected first Wi-Fi RSSI and the pre-collected plurality of second Wi-Fi RSSIs according to preset conditions, A user device that updates the coexistence estimation model using the feedback dataset and the filtered dataset.

9. In paragraph 8, A user device, wherein the feedback includes information indicating that the estimation result is inaccurate and location information of the user device.

10. In paragraph 9, The one or more processors, by executing the one or more instructions, A user device configured to configure a second Wi-Fi RSSI and the first Wi-Fi RSSI of a device estimated to exist in the same space as the user device among the other devices as a negative pair based on the feedback.

11. In paragraph 9, The one or more processors, by executing the one or more instructions, A user device, wherein based on the feedback, the second Wi-Fi RSSI of a device among the other devices, the location information of which corresponds to the location information of the user device, and the first Wi-Fi RSSI are configured as a positive pair.

12. In paragraph 8, The one or more processors, by executing the one or more instructions, Applying the pre-collected first Wi-Fi RSSI and the pre-collected plurality of second Wi-Fi RSSIs as input data to the coexistence estimation model, A user device that determines whether to filter the pre-collected first Wi-Fi RSSI and the pre-collected plurality of second Wi-Fi RSSIs based on the number of devices estimated to exist in the same space as the user device among the other devices.

13. In paragraph 12, The one or more processors, by executing the one or more instructions, A user device, wherein, if the number of devices estimated to exist in the same space as the user device among the other devices is 1, the second Wi-Fi RSSI of the device estimated to exist in the same space as the user device and the first Wi-Fi RSSI are configured as a positive pair, and the second Wi-Fi RSSI and the first Wi-Fi RSSI of the remaining devices are configured as a negative pair.

14. In paragraph 12, The one or more processors, by executing the one or more instructions, Classify the plurality of pre-collected second Wi-Fi RSSIs of the other devices that match the pre-collected first Wi-Fi RSSI of the user device by device, A user device that samples a first Wi-Fi RSSI-second Wi-Fi RSSI pair to be included in the filtered dataset based on the classification result.

15. A computer-readable recording medium having recorded thereon a program for executing the method of any one of clauses 1 to 7 on a computer.

Images