JP2017026475A - Discrimination device, discrimination method, program, model generation device, and model generation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To discriminate the movement of a moving body from a signal including various kinds of disturbance with higher accuracy.SOLUTION: A discrimination device includes: a feature amount calculation section for calculating feature amount of the movement of a moving body on the basis of a sensor signal indicating the movement of the moving body detected by a sensor; an evaluation value calculation section for calculating an evaluation value corresponding to the movement of the moving body according to the feature amount of the movement of the moving body calculated by the feature amount calculation section using a recurrence learning model generated on the basis of a plurality of pieces of data in which the feature amount of a prescribed movement and an evaluation value corresponding to the prescribed movement are made as a pair; and an adjustment section for adjusting a threshold compared with the evaluation value corresponding to the movement of the moving body to discriminate the movement of the moving body.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a discrimination device, a discrimination method, a program, a model generation device, and a model generation method.

  In recent years, a technique for discriminating the movement of a moving body based on a signal derived from the movement of a moving body such as a person detected by a non-contact sensor has been developed. For example, Patent Document 1 below discloses an invention relating to a control circuit including at least one sensitivity correction unit that performs sensor sensitivity correction in processing of a signal detected by a sensor. The sensitivity correction means adjusts the amplification factor of the amplifier circuit in the control circuit, suppresses a frequency component of relatively fast movement using a first-order lag filter, or uses a second-order lag filter. And the function of suppressing the frequency component of relatively slow movement. In such a configuration, the sensitivity for determining the presence or absence of a person can be adjusted by switching the sensitivity correction means.

JP 2010-230492 A

  According to the technique disclosed in Patent Document 1 described above, by adjusting the sensitivity according to the movement of the moving body and the environment of the moving body, it is possible to improve the accuracy of determining the presence or absence of a person. However, in the environment where the sensor is installed, the sensor can detect not only complicated human movements but also various disturbances (for example, electrical noise and mechanical movements). The signal detected by the sensor can include a plurality of disturbances. For this reason, there is a limit to discriminating the motion of a moving object in an environment that can include various disturbances by simply switching the sensitivity correction means disclosed in Patent Document 1 and adjusting the sensitivity.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is a novel and capable of discriminating the movement of a moving object with higher accuracy from a signal including various disturbances. An object of the present invention is to provide an improved discrimination device.

  In order to solve the above-described problem, according to an aspect of the present invention, a feature amount calculation unit that calculates a feature amount of the moving body based on a sensor signal indicating the motion of the moving body detected by the sensor, and a predetermined amount The motion of the moving object calculated by the feature amount calculation unit using a regression learning model generated based on a plurality of data that is a set of the feature amount of the motion and the evaluation value corresponding to the predetermined motion An evaluation value calculation unit that calculates an evaluation value corresponding to the motion of the moving object, and a threshold value to be compared with the evaluation value corresponding to the motion of the moving object for discrimination of the motion of the moving object There is provided a determination device including an adjustment unit for adjustment.

  The regression learning model is a regression learning model generated by support vector regression, and the evaluation value calculation unit includes a support vector constituting the regression learning model generated by the support vector regression, and a motion of the moving object. An evaluation value corresponding to the motion of the moving object may be calculated using the feature amount.

  The feature amount calculation unit may calculate the feature amount based on a waveform of the sensor signal.

  The feature amount calculation unit may calculate the amplitude of the sensor signal as the feature amount.

  The feature amount calculation unit may calculate the frequency of the waveform of the sensor signal in a predetermined section as the feature amount.

  The evaluation value corresponding to the predetermined motion is a plurality of values generated based on a plurality of data including the feature amount of the predetermined motion and a reference evaluation value set in advance corresponding to the predetermined motion. May be determined according to an error between a plurality of discrimination results corresponding to the predetermined movement discriminated using the reference regression learning model.

  The adjustment unit may adjust the threshold value stepwise.

  The predetermined movement may include a movement of a living body, and the determination of the movement of the moving body may include determination of whether or not the movement of the moving body is the movement of the living body.

  The sensor may be a Doppler sensor.

  In order to solve the above-described problem, according to another aspect of the present invention, a step of calculating a feature quantity of the moving object based on a sensor signal indicating the movement of the moving object detected by the sensor, Using the regression learning model generated on the basis of a plurality of data in which the feature value of the motion and the evaluation value corresponding to the predetermined motion are paired, the moving object according to the feature value of the motion of the moving object And a step of adjusting a threshold value to be compared with the evaluation value corresponding to the movement of the moving object for the determination of the movement of the moving object. The

  In order to solve the above-described problem, according to another aspect of the present invention, the computer calculates a feature amount of the motion of the moving object based on a sensor signal indicating the motion of the moving object detected by the sensor. Calculated by the feature amount calculation unit using a regression learning model generated based on a plurality of data including a feature amount calculation unit, a feature amount of the predetermined motion, and an evaluation value corresponding to the predetermined motion An evaluation value calculation unit that calculates an evaluation value corresponding to the motion of the moving object in accordance with the feature amount of the motion of the moving object, and the evaluation value corresponding to the motion of the moving object for discrimination of the motion of the moving object A program for causing the program to function as an adjustment unit that adjusts a threshold value to be compared is provided.

  In order to solve the above-described problem, according to another aspect of the present invention, a feature amount for calculating a feature amount of the predetermined motion based on a sensor signal indicating the predetermined motion by the moving object detected by the sensor. A model generation apparatus comprising: a calculation unit; and a model generation unit that generates a regression learning model based on a plurality of pieces of data, each of which includes a feature amount of the predetermined motion and an evaluation value corresponding to the predetermined motion. Provided.

  The model generation device includes a plurality of regression learning models generated based on a plurality of data in which the feature quantity of the predetermined movement and a reference evaluation value set in advance corresponding to the predetermined movement are paired. A regression learning model evaluation value determining unit that determines an evaluation value corresponding to the predetermined movement according to an error between a plurality of determination results corresponding to the predetermined movement determined by using the predetermined movement may further be provided.

  In order to solve the above problem, according to another aspect of the present invention, a step of calculating a feature amount of the predetermined movement based on a sensor signal indicating the predetermined movement by the moving object detected by the sensor; And a step of generating a regression learning model based on a plurality of data in which the feature quantity of the predetermined motion and the evaluation value corresponding to the predetermined motion are paired.

  As described above, according to the present invention, it is possible to determine the motion of a moving object with higher accuracy from a signal including various disturbances.

It is a figure showing the outline of the discriminating system concerning one embodiment of the present invention. It is a block diagram which shows the structural example of the model production | generation apparatus which concerns on the 1st Embodiment of this invention. It is a figure showing an example of a data list stored in signal DB concerning the embodiment. It is a figure for demonstrating the number of zero crossings of the waveform of signal data. FIG. 4 is a diagram illustrating an example of an evaluation value allocation processing flow by an evaluation value generation unit for each data set stored in the data list illustrated in FIG. 3. It is a figure which shows an example of the evaluation list | wrist stored in evaluation value DB which concerns on the embodiment. It is a flowchart which shows the operation example of the model production | generation apparatus which concerns on the same embodiment. It is a block diagram showing an example of composition of a discriminating device concerning the embodiment. It is a flowchart which shows the operation example of the discrimination | determination apparatus concerning the embodiment. It is a block diagram which shows the structural example of the model production | generation apparatus which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows the structural example of the database which stores the various data used in the evaluation value generation part and evaluation value generation part which concern on the embodiment. It is a figure showing an example of data list 1 (there is no disturbance) stored in feature-value DB concerning the embodiment. It is a figure which shows an example of the data list 2 (with disturbance) stored in feature-value DB which concerns on the embodiment. It is a figure which shows an example of the reference evaluation value list | wrist 1 (no disturbance) stored in reference evaluation value DB which concerns on the embodiment. It is a figure showing an example of reference evaluation value list 2 (with disturbance) stored in reference evaluation value DB concerning the embodiment. It is a figure for demonstrating the evaluation value determination process by the evaluation value determination part which concerns on the same embodiment. It is a graph which shows distribution of the unattended rate calculated by each reference regression learning model. It is a flowchart which shows the operation example of the evaluation value generation part which concerns on the same embodiment. It is a flowchart which shows the processing flow of the reference model production | generation part which concerns on the same embodiment. It is a flowchart which shows the processing flow of the discrimination verification part which concerns on the same embodiment. It is a flowchart which shows the processing flow of the evaluation value determination part which concerns on the same embodiment. It is a block diagram which shows the hardware structural example of the information processing apparatus which concerns on embodiment of this invention.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

<1. Overview of discrimination system>
The discriminating system 1 according to an embodiment of the present invention applies a regression learning model generated based on a predetermined motion by the model generating device 10 in advance to the discriminating device 20 and discriminates the motion of the moving object detected by the sensor 2. This is a system for discriminating by the device 20. Hereinafter, an outline of the discrimination system 1 will be described.

  FIG. 1 is a diagram showing an outline of a discrimination system 1 according to an embodiment of the present invention. The configuration of the discrimination system 1 varies depending on the processing to be performed. For example, referring to FIG. 1, when performing model generation processing for generating a regression learning model, the discrimination system 1 has a configuration including a sensor 2 and a model generation device 10. On the other hand, when performing the determination process of the motion of the moving object, the determination system 1 has a configuration including the sensor 2 and the determination device 20. Note that the sensor 2 and the model generation device 10 may integrally form a model generation device, and the sensor 2 and the determination device 20 may integrally form a determination device.

(Sensor)
As shown in FIG. 1, the sensor 2 is a non-contact sensor installed on the ceiling of a room. For example, the sensor 2 emits a radiation wave such as light, electromagnetic wave, or sound wave toward a room that is a detection area, and is moved by a moving object (for example, a person P, an air conditioner AC, or a curtain CU in FIG. 1). It may be a radio wave sensor that receives reflected reflected waves. The radio wave sensor may be a Doppler sensor, for example. At this time, the frequency of the reflected wave changes from the frequency of the radiated wave due to the Doppler effect caused by the motion of the object such as vibration. The sensor 2 generates a signal having a difference frequency between the frequency of the radiated wave and the frequency of the reflected wave. The sensor 2 may be mounted with, for example, a quadrature detection method. In this case, the sensor 2 may generate two types of signals, a cosine wave component (I component) and a sine wave component (Q component). The sensor 2 outputs the generated signal to either the model generation device 10 or the determination device 20.

  The sensor 2 can be installed at any position as long as the motion of the moving object can be detected. In the example shown in FIG. 1, the sensor 2 is configured by integrating the transmission unit for the radiated wave and the reception unit for the reflected wave, but the sensor 2 has a configuration in which the transmission unit and the reception unit are separated. May be realized. Further, the radiation wave radiated from the sensor 2 may be a wave in an arbitrary frequency band as long as the Doppler effect can be generated by the vibration of the object. For example, it is preferable that the radiated wave is a wave in a frequency band in which human movement can be detected. The environment in which the sensor 2 is installed in the model generation process and the determination process may be different. For example, in the model generation process, since it is necessary to detect a signal of a single motion of a moving object, the sensor 2 may be provided in a room that can block disturbance, such as a soundproof room. On the other hand, in the determination process, the installation location of the sensor 2 is not particularly limited as long as it is a location where it is required to detect the movement of a moving object such as a person. For example, the sensor 2 may be installed not only in a room but also in a section partitioned by a partition or a fence. In addition, the sensor 2 is installed inside a structure or body of an exercise facility, a medical facility, a nursing facility, a nursing facility, an attraction facility, a criminal facility, a house, a hotel, a condominium, a building, a car, a railway vehicle, a ship, or the like. Also good. Also, the type of moving object is not limited to the person, furniture, or home appliance as described above.

  In the present embodiment, the sensor 2 is preferably a radio wave sensor (Doppler sensor). For example, when the sensor 2 is an infrared sensor, it is possible to detect the movement of the moving object, but it is difficult to detect the moving object when the moving object is stationary. On the other hand, when the sensor 2 is a radio wave sensor, the vibration of the trunk due to respiration of the moving body can be detected even when the moving body is stationary. Therefore, it is easy to detect the state where the moving body is stationary.

(Model generator)
The model generation apparatus 10 is an apparatus used for a model generation process of a predetermined moving body motion. The model generation device 10 may be realized by one or a plurality of information processing devices on a network, for example. More specifically, the model generation device 10 may be realized by a server, a PC (Personal Computer), or the like. The model generation device 10 acquires a signal derived from the movement of a predetermined moving body from the sensor 2 and generates an evaluation value corresponding to the movement of the predetermined moving body. And the model production | generation apparatus 10 produces | generates a regression learning model based on the signal of the motion of each predetermined moving body, and the evaluation value corresponding to the motion of each moving body. The generated regression learning model is used in the discrimination device 20.

  Here, in the present specification, the movement of the moving body includes not only the movement of the person but also movements other than the person, such as movement of home appliances, vibration of the ground, and vibration of furniture due to wind. Further, the movement of the person may be further subdivided. For example, a movement when sleeping or a movement while sitting may be included in the movement of the person. The evaluation value is a value used for determining the movement of the moving object, and is associated with the movement of each moving object.

(Determination device)
The discriminating device 20 is a device used for discriminating the movement of a specific moving object. The determination device 20 may be realized by one or a plurality of information processing devices on a network, for example. More specifically, the determination device 20 may be realized by a server, a PC, or the like. The discriminating device 20 discriminates the motion of the moving object by using a signal derived from the motion of the moving object detected by the sensor 2 and the regression learning model generated by the model generating apparatus 10. For example, the determination device 20 may determine the movement of the moving object by comparing the evaluation value obtained by using the regression learning model with respect to the movement of the moving object detected by the sensor 2 and an adjustable threshold value. Good. The determination device 20 can transmit the determination result to an external device or the like by wire or wireless via a communication unit (not shown), for example. As a result, for example, the external device can perform processing using the determination result. By using the determination result, for example, it is possible to grasp the presence or absence of the person P. Note that the processing using the discrimination result can be performed not in the external device as described above but in the discrimination device 20 main body.

  Conventionally, the development of a technique for determining the presence or absence of a person has been underway. For example, according to the invention disclosed in Japanese Patent Application Laid-Open No. 2010-230492, the sensitivity for determining the presence or absence of a person can be adjusted, but specific disturbances such as temperature changes and the speed of movement of moving objects can be adjusted. Sensitivity could be adjusted only for signals containing these factors. For this reason, it has been difficult to adjust the sensitivity for improving the discrimination accuracy for complex signals including various disturbances.

  In the determination system 1 of the present invention, the model generation device 10 generates a regression learning model based on a plurality of predetermined scenarios, and the determination device 20 determines the motion of a moving object using the regression learning model. In such a configuration, by applying the motion of the moving object to the regression learning model, the similarity (evaluation value) between the motion of the moving object and a predetermined scenario that is a component of the regression learning model is calculated. That is, by applying the regression learning model to a signal including a plurality of disturbances, it is possible to represent what kind of motion is similar as a continuous evaluation value. Therefore, for example, by setting a threshold value for discrimination within a range of values that can be taken as an evaluation value as a sensitivity adjustment element, it is possible to determine whether or not the motion of a moving object is a predetermined motion by comparing with a threshold value. it can. Since the threshold value can be set continuously if the evaluation value is a continuous value, the sensitivity relating to the determination of the predetermined movement can be finely adjusted. Therefore, it is possible to adjust the sensitivity even for a signal in which a plurality of disturbances that cannot be conventionally adjusted for sensitivity can be mixed. Therefore, it is possible to further improve the determination accuracy of the motion of the moving object.

<2. First Embodiment>
Hereinafter, the model generation device 10 and the determination device 20 according to the first embodiment of the present invention will be described.

[2.1. Model generator]
(1) Configuration FIG. 2 is a block diagram illustrating a configuration example of the model generation device 10 according to the first embodiment of the present invention. Referring to FIG. 2, the model generation apparatus 10 includes a signal acquisition unit 110, a signal DB (database) 111, a filter unit 120, a feature amount calculation unit 130, an evaluation value generation unit 140, an evaluation value DB 141, a model generation unit 150, and A regression learning model DB 151 is provided. The model generation device 10 includes a plurality of sets each including a feature space formed from a feature vector calculated from a signal derived from a known predetermined motion detected by the sensor 2 and an evaluation value corresponding to the predetermined motion. A regression learning model is generated from the data. The generated regression learning model is used in the discrimination device 20. Hereinafter, the model generation device 10 according to the present embodiment will be described.

(Signal acquisition unit)
The signal acquisition unit 110 acquires signal data output from the sensor 2 via a communication unit or an input unit (not shown). In the present embodiment, the sensor 2 is a Doppler sensor. The signal data output from the sensor 2 is data composed of a time-series digital signal, and has two components of I (n) which is an I component and Q (n) which is a Q component. The signal data is collected by the sensor 2 for a specified time. The specified time may be, for example, 10 minutes. Further, the sampling frequency of the signal data is not particularly limited, but is preferably 100 Hz or more in order to determine the movement of the person. The signal acquisition unit 110 acquires a signal that has detected a predetermined movement as signal data, and outputs the acquired signal data to the signal DB 111.

(Signal DB)
The signal DB 111 is a database that stores the signal data acquired from the signal acquisition unit 110. The database has a plurality of data lists, and the data list is composed of a plurality of data sets. FIG. 3 is a diagram illustrating an example of a data list stored in the signal DB 111. Referring to FIG. 3, the data list 1000 includes data files (A1.dat, B2.dat,..., C6.dat) of signal data derived from a predetermined movement acquired from the signal acquisition unit 110. , Attributes, and scenarios, stored as a data set. The ID is an index assigned for each acquired signal data. The attribute is an item indicating the type of motion of the moving object, and a plurality of attributes can be set in one data list. In the data list 1000 shown in FIG. 3, for example, three types of attributes “unmanned (without disturbance)”, “unmanned (with disturbance)”, and “manned” are set. Here, “unmanned (no disturbance)” means that no movement by a moving object including a person is detected. “Unmanned (with disturbance)” means that there is no person, but the movement of another moving object can be detected. “Manned” means a state in which only movement by a person can be detected.

  The scenario is an item indicating a specific moving object set in advance in each attribute and the movement of the moving object. The predetermined movement set as the scenario is a movement selected in advance by the user of the model generation apparatus 10 or the like. In the data list 1000 shown in FIG. 3, “air conditioner” and “curtain” are set as scenarios for the attribute “unmanned (with disturbance)”. This means that the signal data acquired as the attribute of “unmanned (with disturbance)” is signal data derived from vibrations of the air conditioner and the curtain. Similarly, in the attribute “manned”, “sit”, “sleep”, and “walk” are set as scenarios. This means that the signal data acquired as the “manned” scenario is signal data derived from a person's sitting movement, sleeping movement, and walking movement.

  As described above, the signal DB 111 stores a data list including a plurality of data sets including an ID, an attribute, a scenario, and signal data corresponding to the scenario. Note that these data sets and data lists may be added, modified, or deleted as appropriate by the user of the model generation apparatus 10 via an input unit or a communication unit (not shown). The generated data list is output to the filter unit 120.

(Filter part)
The filter unit 120 performs a filtering process on the signal data included in the data list acquired from the signal DB 111. The signal data included in the data list includes a lot of high-frequency component noise that is unrelated to the movement of the moving object. Therefore, the filter unit 120 removes or reduces these noises by performing LPF (Low Pass Filter) processing.

First, the filter unit 120 acquires signal data I _m (n) and Q _m (n) of ID = m stored in the data list (n = 1 to N). Here, N indicates the number of samplings included in the signal data.

Here, I _m (n) and Q _m (n) may be subjected to LPF processing by a second-order IIR (Infinite Impulse Response) filter. The cutoff frequency in the LPF process may be 50 Hz, for example. By such LPF processing, high-frequency noise components included in signal data can be removed or reduced. Here, the signal data after the LPF processing by the IIR filter is performed are I _m ′ (n) and Q _m ′ (n), respectively.

The LPF processing by the IIR filter is performed by repeatedly calculating the following formulas (1) and (2) from n = 1 to N in order. Here, x (n) is an input value and y (n) is an output value. For example, in the following formula (1), I _m (n) and Q _m (n) are substituted for x (n). In the following formula (2), I _m ′ (n) and Q _m ′ (n) are output as y (n). K, a ₀ , a ₁ , a ₂ , b ₁ , and b ₂ are coefficients set at the design stage according to the frequency band.

  Note that A (0) = A (−1) = 0.

After I _m ′ (n) and Q _m ′ (n) are output, the value of A (n−1) is assigned to A (n−2), and A (n−1) is assigned to A (n−1). ) Value is substituted. By performing this substitution process from n = 1 to N, the LPF process is completed. The data list including the signal data for which the LPF processing has been completed is output to the feature amount calculation unit 130.

  In addition, although the filter part 120 concerning this embodiment used the IIR filter as LPF processing, this invention is not limited to this example. For example, the filter unit 120 performs LPF processing by converting the signal to the frequency domain by Fourier transform, removing the frequency in a frequency band other than the predetermined frequency, and then converting the signal to the time domain again by inverse Fourier transform. Also good. The filter unit 120 may perform LPF processing using a known technique other than the above-described technique.

(Feature amount calculation unit)
The feature amount calculation unit 130 calculates the feature amount of the signal data after LPF processing stored in the data list acquired from the filter unit 120. Here, the feature amount is a feature amount calculated based on the waveform of the signal data. For example, the feature amount includes a feature amount calculated from the amplitude of the waveform of the signal data, the frequency component of the waveform, or the like. In the present embodiment, the feature amount calculation unit 130 calculates the amplitude feature amount and the frequency feature amount of the waveform of the signal data as the feature amount.

In the present embodiment, the amplitude feature amount is the value of the amplitude at time n of the signal data after the LPF processing. The amplitude feature amount V _m (n) is calculated by the following equation (3). Further, the amplitude feature amount V _m (n) may be a statistic such as an average value or a maximum value of amplitude values of each sampling data in a unit period starting from the time n.

The frequency feature amount may be, for example, the number of zero crossings of the signal data waveform in the unit period. FIG. 4 is a diagram for explaining the number of zero crossings of the waveform of the signal data. As indicated by a circle in FIG. 4 crossing, zero crossing number, the waveform of the signal data I _m '(n) or Q _m' (n) is the time axis in the unit period T _a and (amplitude = zero value) It is the number of times to do. Specifically, the zero crossing number, among the L sampling data in the unit period T _a for the time n and the top is a number satisfying the following formulas (4) and (5). Unit interval T _a may be, for example, one second, in which case, L represents 100 (i.e., sampling frequency = 100 Hz of the time-series digital signal data outputted from the sensor 2) may be used.

In the above formula unit period _{T a} for the time n and the top (4) and (5) the number of i satisfying each _f 1 m When (n) and _f 2m (n), the zero crossing number _f m (n) Is f _1m (n) + f _2m (n).

In addition to the amplitude feature value V _m (n) and the number of zero crossings f _m (n), the feature value calculation unit 130 further generates a waveform of the signal data I _m ′ (n) or Q _m ′ (n). The feature amount based on the may be calculated. For example, the feature amount calculation unit 130 calculates the number of local maximum or minimum points of a waveform in a unit period, a statistic such as variation or average of local maximum or minimum points in a plurality of unit periods, or a power spectrum in a frequency domain. Alternatively, it may be calculated as a feature amount. By increasing the types of feature amounts, the features of the waveform of the signal data can be extracted more finely.

Feature amount calculation unit 130, first, the L-number of sampling data included from the signal data to the unit interval T _a and J pairs extracted, respectively calculates the feature quantity for each sampling data unit sections T _a. That is, C × J feature quantities are extracted from one signal data (C is the number of feature quantity types). Next, the feature amount calculation unit 130 calculates feature amounts for all signal data included in the acquired data list. Then, the feature amount calculation unit 130 generates a feature vector having the calculated feature amount as a component. The feature vector constitutes a vector of F _m (j) = [V _m (j), f _m (j)]. Here, j indicates the oldest sampling time among the L sampling data as described above. The feature amount calculation unit 130 performs the above processing on all signal data stored in the data list, and generates a feature vector F _m (j) for each signal data. Then, the feature amount calculation unit 130 outputs the generated feature vector F _m (j) to the model generation unit 150.

(Evaluation value decision part)
The evaluation value generation unit 140 generates an evaluation value corresponding to each scenario in the data list stored in the signal DB 111. Here, the evaluation value is a parameter used for discriminating a movement by a specific moving object. For example, when an evaluation value is set in the range of 0 to 1, for a scenario that is a movement by a specific moving object (here, a person's movement), 1 is set as the evaluation value and a scenario that is not a person's movement A value other than 1 can be set as the evaluation value. Here, for example, 0 may be set as an evaluation value for a scenario that is clearly not a person's movement, but an intermediate value between 0 and 1 may be set as an evaluation value for a scenario that is confusing with a person's movement. . Moreover, even if the attributes are the same, different evaluation values may be set for each scenario. In this way, by setting the evaluation value in correspondence with the scenario, adjustment for discriminating a motion that is confused with a motion by a specific moving object such as a person is facilitated in later discrimination processing.

  For example, the evaluation value generation unit 140 may allocate a predetermined evaluation value according to the attribute of the data list stored in the signal DB 111. FIG. 5 is a diagram showing an example of an evaluation value allocation processing flow by the evaluation value generation unit 140 for each data set stored in the data list 1000 shown in FIG. Referring to FIG. 5, first, the evaluation value generation unit 140 acquires a data list stored in the signal DB 111 (S101). The evaluation value generation unit 140 reads the attributes of the data list for all scenarios, and sets evaluation values according to the read attributes.

The evaluation value generation unit 140 determines whether or not the read attribute is “unmanned (no disturbance)” (S103). When the read attribute is “unattended (no disturbance)” (S103 / YES), the evaluation value generation unit 140 sets the evaluation value for the data set to 0.0 (S105). Next, the evaluation value generation unit 140 determines whether or not the read attribute is “unmanned (with disturbance)” (S107). When the read attribute is “unattended (with disturbance)” (S107 / YES), the evaluation value generation unit 140 sets the evaluation value for the data set as 0.5 (S109). When the read attribute is neither “unmanned (without disturbance)” nor “unmanned (with disturbance)” (“manned” in the example shown in FIG. 3) (S107 / NO), the evaluation value generation unit 140 receives the data The evaluation value for the set is set to 1.0 (S101). Based on the above processing flow, the evaluation value generation unit 140 can allocate an evaluation value for each data set. When there are M scenarios, the value set in step S105, S109, or S111 is substituted into the evaluation value Y _m (j) corresponding to the mth scenario.

  In the present embodiment, the evaluation value is assigned to each scenario by the evaluation value generation unit 140, but the present invention is not limited to such an example. For example, the evaluation value generation unit 140 may assign an evaluation value value input by an input unit (not shown) to each scenario. Thereby, the user of the model production | generation apparatus 10 etc. can set the value of the evaluation value with respect to each scenario freely. Moreover, the range of the evaluation value is not limited to the range (0.0 to 1.0) shown above, and can be set freely.

  After assigning the evaluation value to each scenario, the evaluation value generating unit 140 links the ID and evaluation value of each scenario, and outputs the result to the evaluation value DB 141 as an evaluation value list.

(Evaluation value DB)
The evaluation value DB 141 is a database that stores the evaluation value list acquired from the evaluation value generation unit 140. FIG. 6 is a diagram illustrating an example of an evaluation list stored in the evaluation value DB 141. Referring to FIG. 6, in the evaluation list 2000 stored in the evaluation value DB 141, evaluation values are stored as a data set together with IDs, attributes, and scenarios. The meanings of the ID, attribute, and scenario are the same as those of each element included in the data list stored in the signal DB 111, and thus description thereof is omitted. The evaluation value list stored in the evaluation value DB 141 is output to the model generation unit 150. Note that the value of the evaluation value stored in each data set may be set, added, modified, or deleted by a user of the model generation device 10 or the like via an input unit or a communication unit (not shown). In that case, the model generation device 10 does not necessarily include the evaluation value generation unit 140.

(Model generator)
The model generation unit 150 generates a regression learning model based on a plurality of pieces of data including the feature vector output from the feature amount calculation unit 130 and the evaluation value included in the evaluation value list acquired from the evaluation value DB 141. To do. Specifically, first, the model generation unit 150 generates a feature space composed of feature vectors F ₁ (j) to F _M (j) corresponding to J pieces of extracted data and M scenarios. A regression learning model is generated based on the data of the evaluation values Y ₁ (j) to Y _M (j) corresponding to the scenario. Since the evaluation value is assigned by the evaluation value generation unit 140 for each corresponding scenario, the evaluation value Y _m (j) corresponding to the same scenario (that is, signal data) is the same regardless of the value of j. is there.

  Here, in the present embodiment, the regression learning model can be generated using, for example, SVR (Support Vector Regression). The SVR can be realized based on the contents disclosed in, for example, SMOLA, Alex J. and Bernhard SCHOELKOPF, “A Tutorial on Support Vector Regression”, 1998. The regression learning model is not limited to the above example, and may be generated by a known regression analysis method such as a kernel method or multiple regression analysis, for example.

  The regression learning model generated by SVR is composed of a support vector matrix s composed of K elements, a weight vector c thereof, and a threshold value b. Specifically, the support vector matrix s is defined by a matrix represented by the following formula (6), and the weight vector c is defined by a vector represented by the following formula (7). The threshold value b is a value that is uniquely determined. Further, the number K of elements of the support vector matrix s is uniquely determined in the process of calculation by SVR.

  Note that a regression learning model can be generated by mapping data composed of a feature space and an evaluation value to a nonlinear space using a kernel trick. In the present embodiment, a Gaussian kernel function represented by the following formula (8) may be used as a kernel trick. In this case, the coefficient σ that defines the spread of the Gaussian function is an element constituting the regression learning model.

  The model generation unit 150 outputs, to the regression learning model DB 151, a regression learning model generated from the data configured by the feature space corresponding to the M scenarios and the evaluation values by using regression analysis such as SVR.

(Regression learning model DB)
The regression learning model DB 151 is a database that stores the regression learning model output from the model generation unit 150. The regression learning model DB 151 stores, for example, the support vector matrix s, the weight vector c, the threshold value b, the coefficient σ of the Gaussian kernel function, and the like of the regression learning model generated by the above equations (6) to (8). Is done. The regression learning model DB 151 may store one or a plurality of regression learning models. For example, a regression learning model corresponding to the type of moving object or the type of movement of the moving object determined by the determination device 20 described later may be stored in the regression learning model DB 151.

  Data related to the regression learning model stored in the regression learning model DB 151 is duplicated and stored in the regression learning model DB 231 of the discrimination device 20. Thereby, the discriminating device 20 can discriminate the motion of the moving object using the regression learning model generated by the model generating device 10.

(2) Operation FIG. 7 is a flowchart showing an operation example of the model generation apparatus 10 according to the first embodiment of the present invention. First, the signal acquisition unit 110 acquires a signal derived from the predetermined movement detected by the sensor 2 as signal data (S201). The signal acquisition unit 110 outputs a plurality of signal data derived from a plurality of predetermined movements to the signal DB 111. The signal DB 111 stores a plurality of signal data as a data list together with IDs, attributes, and parameters, and outputs them to the filter unit 120 and the evaluation value generation unit 140 (S203).

  Next, the filter unit 120 performs LPF processing on each signal data included in the data list (S205). Each signal data subjected to the LPF process is output to the feature amount calculation unit 130. The feature amount calculation unit 130 calculates a feature amount for each signal data after the LPF processing (S207). Next, the feature quantity calculation unit 130 generates a feature vector including the calculated one or more feature quantities (S209). Data including the generated feature vector is output to the model generation unit 150.

  Next, the model generation unit 150 generates a feature space composed of feature vectors corresponding to M scenarios (S211). On the other hand, the evaluation value generation unit 140 determines an evaluation value for each signal data based on the attributes included in the evaluation value list (S213). If an evaluation value is determined in advance for each signal data, step S211 may be omitted. Data including the determined evaluation value is output to the model generation unit 150.

  And the model production | generation part 150 produces | generates a regression learning model based on the data which made the produced | generated feature space and evaluation value a group (S215). Thereafter, the model generation unit 150 outputs the generated regression learning model to the regression learning model DB 151 (S217).

  The model generation device 10 according to the present embodiment has been described above. Next, the determination device 20 according to the present embodiment will be described.

[2.2. Discriminator]
(1) Configuration FIG. 8 is a block diagram illustrating a configuration example of the determination device 20 according to the first embodiment of the present invention. Referring to FIG. 8, the determination device 20 includes a signal acquisition unit 210, a signal DB 211, a filter unit 220, a feature amount calculation unit 230, an evaluation value calculation unit 240, a regression learning model DB 241, a determination unit 250, and an adjustment unit 260. .

(Signal acquisition unit)
The signal acquisition unit 210 acquires signal data output from the sensor 2 via a communication unit or an input unit (not shown). When performing the discrimination process on the acquired signal in real time, the signal acquisition unit 210 outputs the acquired signal data to the filter unit 220. In addition, when performing a determination process on the acquired signal data afterwards, the signal acquisition unit 210 may output the acquired signal data to the signal DB 211.

(Signal DB)
The signal DB 211 is a database that sequentially stores the signal data acquired by the signal acquisition unit 210. When the determination processing is performed on the stored signal data, the signal data stored in the signal DB 211 is output to the filter unit 220.

(Filter part)
The filter unit 220 performs a filtering process on the signal data acquired from the signal acquisition unit 210 or the signal DB 211. The filter process performed by the filter unit 220 may be an LPF process, like the filter unit 120 of the model generation device 10. In this case, the LPF process may be performed by an IIR filter. Since the IIR filter is the same as the IIR filter described above, the description thereof is omitted. The filter unit 220 outputs the signal data subjected to the LPF process to the feature amount calculation unit 230.

(Feature amount calculation unit)
The feature amount calculation unit 230 calculates the feature amount of the signal data acquired from the filter unit 220. The feature amount calculated by the feature amount calculation unit 230 is the same as the feature amount calculated by the feature amount calculation unit 130 of the model generation device 10. For example, the feature amount calculation unit 230 calculates the amplitude feature amount and the frequency feature amount of the waveform of the signal data as the feature amount. Since the specific calculation method of these feature values is the same as the calculation method by the feature value calculation unit 130 of the model generation device 10, description thereof will be omitted.

  The feature amount calculation unit 230 generates a feature vector having the calculated feature amount as a component. Specifically, the feature quantity calculation unit 230 obtains a two-dimensional feature vector F (k) = [V (k), f (k)] from the signal data I (k) or Q (k) at time k. Is generated. V (k) is an amplitude feature amount, and f (k) is the number of zero crossings in the unit period. The unit period is the same as the unit period used when the feature amount calculation unit 130 of the model generation device 10 calculates the number of zero crossings. The feature amount calculation unit 230 outputs the generated feature vector F (k) to the evaluation value calculation unit 240.

(Evaluation value calculator)
The evaluation value calculation unit 240 uses the feature vector F (k) acquired from the feature amount calculation unit 230 and one regression learning model stored in the regression learning model DB 241 to detect the motion of the moving object detected by the sensor 2. An evaluation value corresponding to is calculated. The regression learning model stored in the regression learning model DB 241 is a regression learning model generated by the model generation device 10. The evaluation value calculation unit 240 calculates, for example, the value of the output value z (k) obtained by substituting the feature vector F (k) for the discriminant function represented by the following equation (9) as the evaluation value. A function g (x _{i, k} , σ) shown in the following formula (9) is a kernel function shown in the above formula (8).

Here, the value of x _{i, k} is given by the following equation (10). Note that V _i and f _i are values of the i component of the support vector matrix s expressed by the above equation (6).

  The output value z (k) calculated by the evaluation value calculation unit 240 is output to the determination unit 250 as an evaluation value.

(Regression learning model DB)
The regression learning model DB 241 stores at least one regression learning model calculated by the model generation device 10. The regression learning model DB 241 stores a part or all of the regression learning model stored in the regression learning model DB 151 of the model generation device 10. The regression learning model stored in the regression learning model DB 241 is output to the evaluation value calculation unit 240. The type of regression learning model output to the evaluation value calculation unit 240 is freely selected by the user of the discrimination device 20 or the like. For example, a regression learning model suitable for the environment in which the determination device 20 is used, the structure of the building, the material, and the like may be selected. The regression learning model stored in the regression learning model DB 241 can be appropriately added, changed, or deleted via an input unit or a communication unit (not shown).

(Determination part)
The determination unit 250 determines the movement of the moving object detected by the sensor 2 based on the evaluation value (output value) z (k) calculated from the evaluation value calculation unit 240. In the present embodiment, the determination unit 250 determines whether the motion of the moving object detected by the sensor 2 is a person's motion based on the evaluation value z (k). More specifically, the determination unit 250 compares the evaluation value z (k) with the sensitivity threshold Th, and determines whether the motion of the moving object is a person's motion based on the magnitude relationship. Here, if the determination result value indicating whether or not the movement is a person is d, the value of the determination result value d is determined based on the following comparison expression. First, when z (k) ≧ Th, d = 1. When d = 1, it is determined that the motion of the moving object is that of a person. On the other hand, if z (k) <Th, d = 0. When d = 0, it is determined that the motion of the moving object is not the motion of the person.

  The determination unit 250 outputs the value of the determination result value d to an external device or a storage unit (not shown) included in the determination device 20. Further, the determination unit 250 may output the evaluation value z (k) instead of the determination result value d or together with the determination result value d. For example, when the appropriate value of the sensitivity threshold Th adjusted by the adjusting unit 260 described later is calibrated in the calibration of the determination device 20, in order to verify the actual evaluation value z (k) and the motion of the moving object. The value of the evaluation value z (k) can be output.

(Adjustment part)
The adjustment unit 260 adjusts the sensitivity threshold Th that is compared with the evaluation value corresponding to the movement of the moving object in order for the determination unit 250 to determine the movement of the moving object. The adjustment unit 260 adjusts the sensitivity threshold Th via an input unit or a communication unit (not shown). The evaluation value z (k) that is the target of the sensitivity threshold Th is a continuous value obtained by the discriminant function represented by the above equation (9). The discriminant function is a continuous function including a support vector matrix s, a weight vector c, and a threshold value b included in the regression learning model. That is, the regression learning model constitutes a continuous function generated based on each scenario. Therefore, for example, even if the signal detected by the sensor 2 is a signal in which a plurality of disturbances are mixed, the one-dimensional evaluation value reflecting the characteristics of the signal corresponding to each scenario is obtained by the discriminant function. Is output. Therefore, the adjustment unit 260 can adjust the sensitivity to the determination of the movement of the moving object in an environment in which a plurality of disturbances exist only by setting the sensitivity threshold Th to an arbitrary value.

  The adjustment unit 260 sets an arbitrary value as the sensitivity threshold Th within a range of values that can be taken by the evaluation value z (k) output by the discriminant function. The value of the sensitivity threshold Th may change continuously, but may change discretely. For example, the adjustment unit 260 may adjust the value of the sensitivity threshold Th in a stepwise manner. As a result, the user using the determination device 20 can easily adjust the value of the sensitivity threshold Th. As the initial value of the sensitivity threshold Th, a predetermined value may be set in advance, or a value corresponding to the environment in which the determination device 20 is used may be set. Further, the adjustment of the value of the sensitivity threshold Th may be manually set by a user using the discriminating device 20, but by using a statistical method such as logistic regression analysis or linear discriminant analysis, the optimum sensitivity threshold Th can be set. A value may be set. The set sensitivity threshold value Th is output to the determination unit 250.

(2) Operation FIG. 9 is a flowchart showing an operation example of the determination apparatus 20 according to the first embodiment of the present invention. First, the signal acquisition unit 210 acquires a signal derived from the motion of the moving object detected by the sensor 2 as signal data (S301). The signal acquisition unit 210 outputs the acquired signal data to the signal DB 211 or the filter unit 220.

  Next, the filter unit 220 performs LPF processing on the acquired signal data (S303). The signal data subjected to the LPF process is output to the feature amount calculation unit 230. The feature amount calculation unit 230 calculates a feature amount for the signal data after the LPF processing (S305). Next, the feature amount calculation unit 230 generates a feature vector including the calculated one or more feature amounts (S307). Data including the generated feature vector is output to the evaluation value calculation unit 240.

  The evaluation value calculation unit 240 calculates an evaluation value corresponding to the motion of the moving object using the acquired feature vector and one regression learning model stored in the regression learning model DB 241 (S309). Next, the determination unit 250 determines the movement of the moving object by comparing the evaluation value with the sensitivity threshold adjusted by the adjustment unit 260 (S311). Then, the determination unit 250 outputs the determination result to an external device or a storage unit (S313).

[2.3. effect]
Up to this point, the first embodiment of the present invention has been described with reference to FIGS. According to the first embodiment of the present invention, the model generation device 10 uses SVR or the like based on a plurality of data including a plurality of predetermined motion feature quantities and evaluation values corresponding to the predetermined motion. Generate a regression learning model. Then, the determination device 20 determines the motion of the moving object detected by the sensor 2 using the regression learning model generated by the model generation device 10. The regression learning model constitutes a discriminant function that is a continuous function, and the discriminator 20 calculates a one-dimensional evaluation value corresponding to the motion of the moving object using the discriminant function. Then, the determination device 20 determines the movement of the moving object by comparing the evaluation value with a sensitivity threshold that can be adjusted to an arbitrary value. In such a configuration, by using a regression learning model in which a plurality of predetermined movements are reflected, even when various disturbances accompanying the movement of the moving object detected by the sensor 2 are included in the signal, the determination device 20 Can set the sensitivity threshold freely without depending on the type of the disturbance. Thereby, it is possible to finely adjust the sensitivity related to the determination of the movement of the moving object. Therefore, for example, it is possible to perform sensitivity adjustment for discrimination even for signals including disturbances that have not been conventionally discriminated, such as disturbances that are not targeted by the sensitivity correction means, or a mixture of a plurality of disturbances. . Therefore, it is possible to further improve the determination accuracy of the motion of the moving object.

<3. Second Embodiment>
Subsequently, a second embodiment of the present invention will be described. In the present embodiment, the evaluation value generation unit 140 included in the model generation device 10 has a function of appropriately determining an evaluation value corresponding to each scenario. In the first embodiment of the present invention, a predetermined value is assigned in advance for each scenario, but this predetermined value is a value determined by subjectivity or experience of the user of the model generation device 10 or the like. . Therefore, in generating a regression learning model, these evaluation values do not necessarily generate an optimal model. Therefore, the evaluation value generation unit 140 according to the present embodiment prepares a plurality of data lists including signal data derived from a plurality of predetermined movements, generates a plurality of reference regression learning models for each data list, Using the reference regression learning model, discrimination processing is performed for all data lists used to generate the reference regression learning model. Then, the evaluation value generation unit 140 compares the discrimination results obtained using each reference regression learning model, and determines an appropriate evaluation value based on the comparison results. As a result, the evaluation value for the predetermined movement can be mechanically determined without subjecting the user or the like. Therefore, the quality of the regression learning model by the model generation device 10 is improved, and the accuracy of the discrimination result by the discrimination device 20 can be further improved. Hereinafter, the model generation device 10 and the evaluation value generation unit 140 according to the second embodiment will be described.

[3.1. Model generator]
(1) Configuration FIG. 10 is a block diagram illustrating a configuration example of the model generation device 10 according to the second embodiment of the present invention. Referring to FIG. 10, the model generation apparatus 10 includes a signal acquisition unit 110, a signal DB 111, a filter unit 120, a feature amount calculation unit 130, a feature amount DB 131, an evaluation value generation unit 140, an evaluation value DB 141, a model generation unit 150, and A regression learning model DB 151 is provided. The signal acquisition unit 110, the signal DB 111, the filter unit 120, the feature amount calculation unit 130, the evaluation value DB 141, the model generation unit 150, and the regression learning model DB 151 according to the present embodiment are the same as the components according to the first embodiment. Since it has the same function, description is abbreviate | omitted. The feature amount DB 131 is a database that stores data including the feature amount of the signal data calculated by the feature amount calculation unit 130 in association with the data set stored in the signal DB 111. The feature amount DB 131 will be described later in detail. Hereinafter, the configuration and the like of the evaluation value generation unit 140 will be described.

[3.2. Evaluation value generator]
(1) Configuration FIG. 11 is a block diagram illustrating a configuration example of a database storing various data used in the evaluation value generation unit 140 and the evaluation value generation unit 140. Referring to FIG. 11, the evaluation value generation unit 140 includes a reference model generation unit 410, a discrimination verification unit 420, and an evaluation value determination unit 430.

(Reference model generator)
The reference model generation unit 410 corresponds to each of the feature amounts (feature vectors) included in the plurality of data lists stored in the feature amount DB 131 and each of the plurality of data lists stored in the reference evaluation value DB 151. A reference regression learning model is generated using the reference evaluation values included in the reference evaluation value list. The reference regression learning model is a regression learning model that is generated using a plurality of data sets including a combination of a feature amount included in each of a plurality of data lists and a reference evaluation value. In this embodiment, two types of data lists (data list 1 (without disturbance) and data list 2 (with disturbance)) are prepared. Therefore, two types of models are generated for the reference regression learning model: a model corresponding to the data list 1 and a model corresponding to the data list 2. In the present embodiment, two types of reference regression learning models are generated. However, the present invention is not limited to this example, and three or more types of data lists are used, and reference regressions corresponding to the data lists are used. A learning model may be generated.

  Each of the plurality of data lists has a different configuration. FIGS. 12 and 13 are diagrams showing examples of data list 1 (without disturbance) 3000 and data list 2 (with disturbance) 4000 stored in the feature amount DB 131. Referring to FIG. 12 and FIG. 13, each data list includes an ID, an attribute, a scenario, a signal data file including signal data, and a data set including a feature data file including feature data. Note that the feature amount data included in the feature amount data file includes a feature amount of a predetermined motion calculated by the feature amount calculation unit 130 in advance. Here, the data list 1 includes data sets whose attributes are “unmanned (no disturbance)” and “manned”. On the other hand, the data list 2 includes data sets whose attributes are “unmanned (no disturbance)”, “unmanned (with disturbance)”, and “manned”. The data list 1 and the data list 2 include data sets having the same attribute and scenario, but the signal data included in each data set may not necessarily be the same. A plurality of signal data may be included for the scenario having the same content.

  FIG. 14 and FIG. 15 are diagrams showing the configurations of reference evaluation value list 1 (without disturbance) 5000 and reference evaluation value list 2 (with disturbance) 6000 stored in reference evaluation value DB 151. Referring to FIGS. 14 and 15, each data list is constituted by a data set including an ID, an attribute, a scenario, and a reference evaluation value. Similarly to the data list 1 and the data list 2 described above, the reference evaluation value list 1 does not include the data set having the attribute “unmanned (with disturbance)”, while the reference evaluation value list 2 has the attribute “ A data set of “Unattended (with disturbance)” is included. In these reference evaluation value lists, the reference evaluation value of the data set having the attribute “unmanned” is set to 0, and the reference evaluation value of the data set having the attribute “manned” is set to 1. The data list 1 corresponds to the reference evaluation value list 1, and the data list 2 corresponds to the reference evaluation value list 2.

  The data list 1 (and the reference evaluation value list 1) does not include a data set of the attribute “unmanned (with disturbance)”. Therefore, when the reference regression learning model (referred to as reference regression learning model 1) generated using the data list 1 and the reference evaluation value list 1 is applied to the discrimination process, the data list 2 and the reference evaluation value list 2 are Compared with the case where the reference regression learning model (referred to as reference regression learning model 2) generated using the above is applied to the discrimination process, it is considered that the probability of discriminating a motion other than a person as a person's motion is increased. On the other hand, in the reference evaluation value list 2, the reference evaluation value of the attribute “unmanned (with disturbance)” is set to zero. Therefore, when the reference regression learning model 2 is applied to the discrimination process, the probability of discriminating a person's movement as a movement other than a person is higher than when the reference regression learning model 1 is applied to the discrimination process. It is done. Therefore, the evaluation value generation unit 140 according to the present embodiment determines evaluation values corresponding to all scenarios according to the difference in the discrimination results between the reference regression learning models. This makes it possible to generate a regression learning model that can improve discrimination accuracy.

  The reference model generation unit 410 generates a reference regression learning model 1 and a reference regression learning model 2, and stores these reference regression learning models in the reference model DB 161.

(Discriminant verification unit)
The discriminating / verifying unit 420 stores the reference regression learning model stored in the reference model DB 161 for the feature amounts included in all the data lists used for generating the reference regression learning model stored in the feature amount DB 131. Each of these is subjected to a discrimination process, the discrimination result for each reference regression learning model is verified, and an evaluation value corresponding to each scenario is determined based on the verification result. Specifically, the discrimination verification unit 420 first acquires feature amounts from all feature amount data included in the data list 1 and the data list 2. And the discrimination | determination verification part 420 discriminate | determines the presence or absence (here presence or absence of a person) of a specific moving body about each feature-value using the reference regression learning model 1 and the reference regression learning model 2. . The discrimination verification unit 420 compares the difference between the reference regression learning models with respect to the discrimination result (unmanned rate) of the presence or absence of a person. Then, the discrimination verification unit 420 determines an evaluation value corresponding to each scenario based on the difference. Hereinafter, processing in the discrimination verification unit 420 will be described in order.

  First, the discrimination verification unit 420 performs discrimination processing on each feature amount included in all data lists used for generating the reference regression learning model using each of the reference regression learning models. The discrimination processing by the discrimination verification unit 420 is the same processing as the processing by the evaluation value calculation unit 240 and the discrimination unit 250 of the discrimination device 20 according to the first embodiment. For example, it is assumed that one reference regression learning model generated by the reference model generation unit 410 includes the following support vector matrix s ′, weight vector c ′, and threshold value b ′. Specifically, the support vector matrix s ′ is defined by a matrix represented by the following formula (11), and the weight vector c ′ is defined by a vector represented by the following formula (12). The threshold value b 'is a value that is uniquely determined.

  Similarly to the first embodiment, it is also possible to generate a regression learning model by mapping data composed of a feature space and an evaluation value to a nonlinear space using a kernel trick. In the present embodiment, a Gaussian kernel represented by the following formula (13) may be used as a kernel trick.

Here, the number of scenarios included in all data lists is M, and the feature vector corresponding to the m-th scenario is F _m (k). The discriminating / verifying unit 420 calculates the value of the output value z _m (k) obtained by substituting the feature vector F _m (k) into the discriminant function represented by the following formula (14) as an evaluation value.

However, the values of x _{i, k} are given by the following equation (15). Here, V _i and f _i are values of the i component of the support vector matrix s expressed by the above equation (11).

Next, the discrimination verifying unit 420 discriminates whether or not the mth scenario is a human body movement based on the evaluation value z _m (k) obtained by the above equation (14). More specifically, the discrimination verifying unit 420 discriminates whether or not the m-th scenario is a human motion based on whether or not the evaluation value z _m (k) is 0 or more. Here, if the determination result value indicating whether a human body motion and d _m, based on the following comparison expression, the discrimination result value the value of d _m are determined. First, when z _m (k) ≧ 0, d _m = 1. When d _m = 1, it is determined that the m-th scenario is a person's movement. On the other hand, when z _m (k) <0, d _m = 0. When d _m = 0, it is determined that the m-th scenario is not a person's movement.

Determination verification section 420 calculates a d _m for all scenarios included in all data lists. Then, the discrimination verifying unit 420 calculates the unmanned rate eval [%] according to the following equation (16).

  The discrimination verification unit 420 performs the discrimination process and the unmanned rate (eval) calculation process for each of the plurality of reference regression learning models generated by the reference model generation unit 410. In the present embodiment, the discrimination verification unit 420 performs discrimination processing for all scenarios using each of the reference regression learning model 1 and the reference regression learning model 2, and performs unmanned rate calculation processing based on the discrimination result. To do. Then, the discrimination verification unit 420 outputs the calculation result of the unmanned rate to the evaluation value determination unit 430.

(Evaluation value decision part)
The evaluation value determination unit 430 determines an evaluation value used for the generation process of the regression learning model based on the determination result acquired from the determination verification unit 420. More specifically, the evaluation value determination unit 430 determines an evaluation value based on the difference of the unattended rate for each reference regression learning model acquired from the discrimination verification unit 420. Hereinafter, the evaluation value determination process by the evaluation value determination unit 430 will be described with reference to FIG.

  FIG. 16 is a diagram for describing the evaluation value determination process by the evaluation value determination unit 430. Referring to FIG. 16, the data list 7000 includes an ID corresponding to a scenario, a reference evaluation value, an unmanned rate in the reference regression learning model 1, an unattended rate in the reference regression learning model 2, a difference in unattended rate, and an evaluation value. It is included. The reference evaluation value is a temporary evaluation value stored in advance in the reference evaluation value DB 151, and the scenario with the attribute “unmanned” is set to 0, and the scenario with the attribute “manned” is set to 1. The unattended rate in the reference regression learning model 1 and the unattended rate in the reference regression learning model 2 are unattended rates calculated using each reference regression learning model. For example, in ID 1001 (attribute “unmanned (no disturbance)”, scenario “none”), the unmanned rate is 100% in both cases. In this case, the evaluation value determination unit 430 determines the value (1) given as the reference evaluation value in advance as the evaluation value. On the other hand, in the ID 1004 (attribute “manned”, scenario “sit”), the unattended rate in the reference regression learning model 1 shows 50%, but the unattended rate in the reference regression learning model 2 shows 90%. Yes. Thus, depending on the scenario, the calculated unmanned rate varies depending on the reference regression learning model. For this reason, it is necessary to determine an evaluation value so as not to output a result different from an assumed determination result for a scenario in which the determination is objectively difficult.

  FIG. 17 is a graph showing the unattended rate distribution calculated by each reference regression learning model. In the graph of FIG. 17, ○ is an unattended rate of a scenario including the attribute “unmanned” included in the data list 1, □ is an unattended rate of a scenario including the attribute “manned” included in the data list 1, and × is in the data list 2 The unattended rate of the scenario including the attribute “Unmanned” included, and Δ are the unattended rate of the scenario including the attribute “Manned” included in the data list 2. As shown in FIG. 17, ◯ and Δ are small in the difference in unattended rate by the reference regression learning model, but there are cases in which the difference in unattended rate is large by the reference regression learning model, such as □ and ×. Therefore, the evaluation value determination unit 430 determines the evaluation value as 0.5 as an intermediate value in order to suppress the influence due to the difference in the reference regression learning model when the difference in the unattended rate increases. On the other hand, when the difference in the unattended rate is small, it is considered that the influence due to the difference in the reference regression learning model is small. Therefore, the value given as the reference evaluation value is determined as it is as the evaluation value. In this way, it is possible to generate a regression learning model that excludes subjectivity by flexibly changing the evaluation value according to the difference in the unattended rate.

Referring to FIG. 16 again, the evaluation value determination unit 430 determines evaluation value values for all scenarios based on the difference in the unattended rate. For example, the evaluation value determination unit 430 determines the evaluation value as 0.5 when the difference in the unattended rate for the same scenario between the reference regression learning model 1 and the reference regression learning model 2 is equal to or greater than _a predetermined threshold Th _a. To do. In the present embodiment, the predetermined threshold value Th _a is 20%. When the difference in the unattended rate is 20% or more, the evaluation value determination unit 430 determines the evaluation value as 0.5 for the scenario. On the other hand, when the unmanned rate difference is less than 20%, the evaluation value determining unit 430 determines the value given as the reference evaluation value as the evaluation value.

Note that the value of the predetermined threshold Th _a can be freely changed in order to adjust the accuracy of the regression learning model. In the present embodiment, the three evaluation values are set by the predetermined threshold value Th _a , but the number of the predetermined threshold values may be plural. Moreover, although 0.5 was set as the intermediate value of the evaluation values when the difference in the unattended rate is large, the value as the intermediate value is not particularly limited. For example, when there are a plurality of predetermined thresholds as described above, various values can be set as intermediate values according to the number of steps set by the thresholds. Further, for example, the value of the evaluation value determined by the evaluation value determination unit 430 may be determined based on the calculated unmanned rate eval value, a statistic, or the like. For example, the value of the evaluation value corresponding to one scenario may be a value corresponding to the average value of the unmanned rate calculated using a plurality of reference regression learning models.

  The evaluation value generation unit 140 outputs an evaluation value list including the value of the evaluation value determined by the evaluation value determination unit 430 to the evaluation value DB 141.

(2) Operation FIG. 18 is a flowchart showing a processing flow of the evaluation value generation unit 140 according to the second embodiment of the present invention. Referring to FIG. 18, the evaluation value generation unit 140 first generates a reference regression learning model corresponding to each data list by the reference model generation unit 410 (S400). Next, the evaluation value generation unit 140 performs discrimination processing for all scenarios using each reference regression learning model by the discrimination verification unit 420, and calculates an unmanned rate (S500). Then, the evaluation value generation unit 140 compares the unmanned rate between the reference regression learning models by the evaluation value determination unit 430, and determines the evaluation value of each scenario according to the comparison result (S600). Hereinafter, the above three processes will be described in order.

  FIG. 19 is a flowchart showing a process flow of the reference model generation unit 410 according to the second embodiment of the present invention. Referring to FIG. 19, the reference model generation unit 410 first acquires one data list from the feature amount DB 131 (S401). Next, the reference model generation unit 410 extracts feature amount data including feature vectors corresponding to all scenarios included in the data list (S403). Further, the reference model generation unit 410 acquires a reference evaluation value list corresponding to the data list from the reference evaluation value DB 151 (S405). Then, the reference model generation unit 410 generates a reference regression learning model based on data that is a combination of the feature space including the plurality of feature vectors and the reference evaluation value included in the reference evaluation value list (S407). The reference model generation unit 410 performs the above-described processing on a plurality of data lists (and a reference evaluation value list), and generates a plurality of reference regression learning models.

  FIG. 20 is a flowchart showing a processing flow of the discrimination verifying unit 420 according to the second embodiment of the present invention. Referring to FIG. 20, the discrimination verification unit 420 first extracts feature vectors of each scenario from all data lists stored in the feature amount DB 131 (S501). Next, the discrimination verification unit 420 calculates an evaluation value by substituting the extracted feature vector into a discriminant function constituted by one reference regression learning model, and outputs a discrimination result (S503). Then, the discrimination verification unit 420 calculates the unmanned rate using discrimination results corresponding to all scenarios (S505). The discrimination verification unit 420 uses the reference regression learning model 1 and the reference regression learning model 2 to calculate the unattended rate in each model for all scenarios.

FIG. 21 is a flowchart showing a process flow of the evaluation value determination unit 430 according to the second embodiment of the present invention. Referring to FIG. 21, the evaluation value determining unit 430 first acquires a reference evaluation value corresponding to one scenario from the reference evaluation value DB 151 (S601). Next, the evaluation value determination unit 430 acquires the unmanned rate in each reference regression learning model calculated by the discrimination verification unit 420 (S603). The evaluation value determination unit 430 calculates a difference (diff) between the acquired unmanned rates (S605), and compares the difference between the unattended rates diff and a predetermined threshold value Th _a (S607). When the undifference rate difference diff is equal to or greater than the threshold Th _a (S607 / YES), the evaluation value determination unit 430 determines that the evaluation value corresponding to the scenario is 0.5 (S609). On the other hand, when the undifference rate difference diff is less than the threshold value Th _a , the evaluation value determination unit 430 determines the reference evaluation value corresponding to the scenario as an evaluation value (S611). The evaluation value determination unit 430 performs the above-described evaluation value determination process for all scenarios.

[3.3. effect]
Up to this point, the second embodiment of the present invention has been described with reference to FIGS. According to the second embodiment of the present invention, the evaluation value generation unit 140 generates a plurality of reference regression learning models each of which includes a plurality of different data lists and reference evaluation value lists, and uses these reference regression learning models. By using this, it is determined whether or not a person exists in the scenarios included in the plurality of data lists. Then, the evaluation value generation unit 140 determines an evaluation value corresponding to each scenario based on the difference in the discrimination result (unmanned rate) between the respective reference regression learning models. In such a configuration, the evaluation value corresponding to each scenario can be mechanically determined by determining the evaluation value used for generating the regression learning model using the regression analysis method. As a result, for example, the evaluation value corresponding to a scenario in which it is difficult to determine whether a person is moving, such as the attribute “unmanned (with disturbance)” or the scenario “sit”, excludes the subjectivity of the user, etc. Can be determined. Therefore, the model generation apparatus 10 according to the present embodiment can generate a regression learning model using an evaluation value list including more appropriate evaluation values. Therefore, it is possible to further improve the determination accuracy of whether or not the moving body is a specific movement.

<4. Hardware configuration example>
The discrimination system 1 according to each embodiment of the present invention has been described above. Next, the hardware configuration of the information processing apparatus according to the embodiment of the present disclosure will be described with reference to FIG. FIG. 22 is a block diagram illustrating a hardware configuration example of the information processing apparatus 900 according to the embodiment of the present disclosure. The illustrated information processing apparatus 900 can realize, for example, the model generation apparatus 10 and the determination apparatus 20 in the above-described embodiment.

  Referring to FIG. 22, the information processing apparatus 900 includes a CPU (Central Processing Unit) 901, a ROM (Read Only Memory) 902, a RAM (Random Access Memory) 903, and a host bus 904. The information processing apparatus 900 includes a bridge 905, an external bus 906, an interface 907, an input device 908, an output device 909, a storage device 910, a drive 911, and a network interface 912.

  The CPU 901 functions as an arithmetic processing device and a control device, and controls the overall operation in the information processing device 900 according to various programs. Further, the CPU 901 may be a microprocessor. Note that the CPU 901 controls the overall operation of the functions including the respective functional units in the information processing apparatus 900 or a part thereof. For example, the CPU 901 stores a program, calculation parameters, and the like used by the CPU 901 in the ROM 902. The RAM 903 temporarily stores programs used in the execution of the CPU 901, parameters that change as appropriate during the execution, and the like. These are connected to each other by a host bus 904 including a CPU bus.

  The host bus 904 is connected to an external bus 906 such as a PCI (Peripheral Component Interconnect / Interface) bus via a bridge 905. Note that the host bus 904, the bridge 905, and the external bus 906 are not necessarily configured separately, and these functions may be mounted on one bus.

  The input device 908 includes input means for inputting information such as a mouse, keyboard, touch panel, button, microphone, switch, and lever, and an input control circuit that generates an input signal based on the input by the user and outputs the input signal to the CPU 901. Etc. A user of the information processing apparatus 900 can input various data and instruct a processing operation to the information processing apparatus 900 by operating the input device 908. The input device 908 realizes the function of an input unit (not shown).

  The output device 909 includes display devices such as a CRT display device, a liquid crystal display (LCD) device, an OLED device, and a lamp. Furthermore, the output device 909 includes an audio output device such as a speaker and headphones. The output device 909 outputs the played content, for example. Specifically, the display device displays various information such as reproduced video data as text or images. On the other hand, the voice output device converts reproduced voice data, text data displayed on the display device, and the like into voice and outputs the voice. The output device 909 realizes the function of a display unit (not shown).

  The storage apparatus 910 is an apparatus for storing data in the information processing apparatus 900 according to the embodiment of the present invention. The storage device 910 may include a storage medium, a recording device that records data in the storage medium, a reading device that reads data from the storage medium, a deletion device that deletes data stored in the storage medium, and the like. The storage device is configured by, for example, an HDD (Hard Disc Drive) or an SSD (Solid State Drive). The storage device 910 stores programs executed by the CPU 901 and various data. The storage device 910 stores each DB provided in each information processing device such as the signal DB 111. Note that the storage device 910 realizes a function of a storage unit (not shown).

  The drive 911 is a storage medium reader / writer, and is built in or externally attached to the information processing apparatus 900. The drive 911 reads information recorded on a mounted removable storage medium 96 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, and outputs the information to the RAM 903. The drive 911 can also write information into the removable storage medium 96.

  The network interface 912 is a communication interface configured by a communication device or the like for connecting to another device, for example. The network interface 912 may be a wireless LAN (Local Area Network) compatible communication device, an LTE (Long Term Evolution) compatible communication device, or a Bluetooth communication device. The network interface 912 may be a wire communication device that performs wired communication. The network interface 912 implements the function of a communication unit (not shown).

  Heretofore, an example of the hardware configuration of the information processing apparatus 900 has been shown. Each component described above may be configured using a general-purpose member, or may be configured by hardware specialized for the function of each component. Such a configuration can be appropriately changed according to the technical level at the time of implementation.

<5. Summary>
The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field to which the present invention pertains can come up with various changes or modifications within the scope of the technical idea described in the claims. Of course, it is understood that these also belong to the technical scope of the present invention.

  For example, in the above embodiment, it is assumed that a person exists in an area where the sensor 2 can detect, but the present invention is not limited to such an example. For example, the sensor may be installed so that the sensor can detect the person only when the person is in a position such as a chair or a bed. Thereby, since it becomes possible to limit the area used as a discrimination | determination object, the discrimination | determination precision of a moving body's motion improves. In addition, the determination device may be used in combination with the result of the presence or absence of a person in another information processing device (for example, a device that measures the amount or posture of a person). In such a configuration, processing by another information processing apparatus can be performed only when a person is present in a specific area. Accordingly, since the influence of a disturbance that can occur is reduced, the measurement accuracy of the person's momentum, posture, and the like can be improved.

  Further, the determination device may include each function of the model generation device. In such a configuration, the discriminating apparatus not only performs discrimination processing using the regression learning model stored in the regression learning model DB, but also generates a regression learning model using signal data obtained from the outside. it can. Thereby, it is possible to generate not only a general regression learning model but also a regression learning model specialized depending on the place or scene to which the discrimination device is applied. Therefore, it is possible to further improve the discrimination accuracy of the motion of the moving object to be discriminated.

  In the above-described embodiment, an example in which a person's movement is a discrimination target has been described. However, the present invention is not limited to such an example. For example, the present invention can be applied to movements of various moving objects such as movements of other living bodies, machines, and cranks. In this case, for example, when the sensor is a radio wave type sensor, it is possible to extract a signal expressing the motion of any moving object by appropriately setting the wavelength of the radiated wave according to the displacement of the signal waveform to be discriminated. Is possible.

  In addition, each step in the processing of each device and each functional unit in the present specification does not necessarily have to be processed in time series in the order described as a flowchart. For example, each step in the processing of each device and each functional unit may be processed in an order different from the order described as the flowchart, or may be processed in parallel.

  A computer program for causing hardware such as a CPU, a ROM, and a RAM built in the model generation device 10 and the determination device 20 to exhibit functions equivalent to the configurations of the model generation device 10 and the determination device 20 described above. Can be created. A storage medium storing the computer program is also provided.

DESCRIPTION OF SYMBOLS 1 Discriminating system 2 Sensor 10 Model generator 20 Discriminator 110, 210 Signal acquisition part 120, 220 Filter part 130, 230 Feature amount calculation part 140 Evaluation value generation part 150 Model generation part 240 Evaluation value calculation part 250 Discrimination part 260 Adjustment part 410 Reference Model Generation Unit 420 Discrimination Verification Unit 430 Evaluation Value Determination Unit

Claims (14)

Based on a sensor signal indicating the motion of the moving object detected by the sensor, a feature value calculating unit for calculating a feature value of the motion of the moving object;
Using the regression learning model generated based on a plurality of data in which a feature amount of a predetermined motion and an evaluation value corresponding to the predetermined motion are paired, the moving object calculated by the feature amount calculation unit An evaluation value calculating unit that calculates an evaluation value corresponding to the movement of the moving object according to a feature amount of the movement;
An adjustment unit that adjusts a threshold value to be compared with the evaluation value corresponding to the movement of the moving object in order to determine the movement of the moving object;
A discrimination device comprising: The regression learning model is a regression learning model generated by support vector regression,
The evaluation value calculation unit calculates an evaluation value corresponding to the motion of the moving object using a support vector constituting the regression learning model generated by the support vector regression and a feature amount of the motion of the moving object. The discrimination device according to claim 1.   The determination device according to claim 1, wherein the feature amount calculation unit calculates the feature amount based on a waveform of the sensor signal.   The determination device according to claim 3, wherein the feature amount calculation unit calculates the magnitude of the amplitude of the sensor signal as the feature amount.   The discrimination device according to claim 3 or 4, wherein the feature amount calculation unit calculates a frequency of a waveform of the sensor signal in a predetermined section as the feature amount.   The evaluation value corresponding to the predetermined movement is a plurality of data generated based on a plurality of data including sets of the feature amount of the predetermined movement and an evaluation value set in advance corresponding to the predetermined movement. The determination apparatus according to claim 1, wherein the determination apparatus is determined according to an error between a plurality of determination results corresponding to the predetermined movement determined using a regression learning model.   The determination device according to claim 1, wherein the adjustment unit adjusts the threshold value in a stepwise manner. The predetermined movement includes a biological movement,
The determination device according to any one of claims 1 to 7, wherein the determination of the movement of the moving object includes determination whether or not the movement of the moving object is the movement of the living body.   The discrimination device according to claim 1, wherein the sensor is a Doppler sensor. Calculating a feature quantity of the motion of the moving object based on a sensor signal indicating the motion of the moving object detected by the sensor;
Using a regression learning model generated on the basis of a plurality of data that is a set of a predetermined motion feature amount and an evaluation value corresponding to the predetermined motion, according to the motion feature amount of the moving object, Calculating an evaluation value corresponding to the movement of the moving object;
Adjusting a threshold to be compared with the evaluation value corresponding to the movement of the moving object for the determination of the movement of the moving object;
Method including Computer
Based on a sensor signal indicating the motion of the moving object detected by the sensor, a feature value calculating unit for calculating a feature value of the motion of the moving object;
Using the regression learning model generated based on a plurality of data in which a feature amount of a predetermined motion and an evaluation value corresponding to the predetermined motion are paired, the moving object calculated by the feature amount calculation unit An evaluation value calculating unit that calculates an evaluation value corresponding to the movement of the moving object according to a feature amount of the movement;
An adjustment unit that adjusts a threshold value to be compared with the evaluation value corresponding to the movement of the moving object in order to determine the movement of the moving object;
Program to function as. A feature amount calculation unit that calculates a feature amount of the predetermined movement based on a sensor signal indicating the predetermined movement of the moving object detected by the sensor;
A model generation unit that generates a regression learning model based on a plurality of data in which the feature quantity of the predetermined movement and an evaluation value corresponding to the predetermined movement are paired;
A model generation device comprising:   The determination is made using a plurality of regression learning models generated based on a plurality of data in which the feature quantity of the predetermined movement and a reference evaluation value set in advance corresponding to the predetermined movement are set. The model generation device according to claim 12, further comprising a regression learning model evaluation value determination unit that determines an evaluation value corresponding to the predetermined movement according to an error between a plurality of determination results corresponding to the predetermined movement. . Calculating a feature quantity of the predetermined movement based on a sensor signal indicating the predetermined movement by the moving object detected by the sensor;
Generating a regression learning model based on a plurality of data sets of the predetermined motion feature amount and an evaluation value corresponding to the predetermined motion;
A model generation method including

Images