JP7285690B2 - Information processing method, information processing device and program - Google Patents

Description

The present disclosure relates to an information processing method, an information processing device, and a program.

Techniques for estimating the position of a sound source (especially the distance between a microphone and the sound source) have been disclosed. For example, Patent Literature 1 discloses a technique of estimating the position of a sound source by triangulation using a plurality of microphones installed at a plurality of mutually different positions. Further, for example, Patent Literature 2 discloses a technique for estimating the position of a sound source by creating a database in advance for each sound collected by a microphone when the sound source is placed at each position in space. By estimating the position of the sound source by such a technique, for example, the sound environment in the space (original sound field) in which the microphones are installed can be reproduced in another place. For example, when holding a teleconference remotely, it is possible to create an environment as if the other party were talking in the same room. In addition, the technology is also useful in, for example, public viewing or online games.

Japanese Patent No. 4926091 Japanese Patent No. 5079761

For example, in the technique disclosed in Patent Document 1, it is necessary to install sound collectors (microphones) at a plurality of positions. In addition, in the technology disclosed in Patent Document 2, the database must be created in advance, and the database is specific to each space. Therefore, it is necessary to create the database for each space. be. In other words, the techniques disclosed in Patent Documents 1 and 2 have a problem that it takes time and effort to estimate the distance between the sound collector and the sound source.

Therefore, an object of the present disclosure is to provide an information processing method and the like that can easily estimate the distance between the sound collector and the sound source.

An information processing method according to an aspect of the present disclosure acquires a sound signal indicating sound collected by a sound collector, calculates the volume of the sound collected by the sound collector from the acquired sound signal, and acquires The type of the sound source of the sound collected by the sound collecting device is identified from the sound signal obtained, and the type of the sound source is associated in advance with a standard volume, which is the volume of the sound from the sound source at a predetermined distance from the sound source. Based on the standard volume corresponding to the identified type and the calculated volume in the database, estimate the distance between the sound collector and the sound source of the sound collected by the sound collector, and output the estimation result. do.

It should be noted that these general or specific aspects may be realized by systems, devices, methods, recording media, or computer programs, and any combination of systems, devices, methods, recording media, and computer programs. may be implemented with

According to the information processing method and the like according to the present disclosure, it is possible to easily estimate the distance between the sound collector and the sound source.

FIG. 1 is a block diagram showing an example of the configuration of an information processing apparatus according to Embodiment 1. FIG. 2 is a flowchart illustrating an example of the operation of the information processing apparatus according to Embodiment 1. FIG. FIG. 3 is a table showing an example of the database. FIG. 4 is a block diagram showing an example of the configuration of an information processing apparatus according to Embodiment 2. As shown in FIG. FIG. 5 is a flow chart showing an example of the operation of the information processing device according to the second embodiment. FIG. 6 is a block diagram showing an example of the configuration of an information processing apparatus according to Embodiment 3. As shown in FIG. FIG. 7 is a block diagram showing an example of the configuration of an information processing apparatus according to Embodiment 4. As shown in FIG. FIG. 8 is a block diagram showing an example of the configuration of an information processing apparatus according to Embodiment 5. As shown in FIG. FIG. 9 is a block diagram showing an example of the configuration of an information processing apparatus according to Embodiment 6. In FIG.

An information processing method according to an aspect of the present disclosure acquires a sound signal indicating sound collected by a sound collector, calculates the volume of the sound collected by the sound collector from the acquired sound signal, and acquires The type of the sound source of the sound collected by the sound collecting device is identified from the sound signal obtained, and the type of the sound source is associated in advance with a standard volume, which is the volume of the sound from the sound source at a predetermined distance from the sound source. Based on the standard volume corresponding to the identified type and the calculated volume in the database, estimate the distance between the sound collector and the sound source of the sound collected by the sound collector, and output the estimation result. do.

According to this, if the standard volume is measured for sounds of various types of sound sources and a database is generated in advance, then the sound collected by the sound collecting device is generated in the space where the distance to the sound source is estimated. The distance between the sound collector and the sound source can be easily estimated simply by calculating the volume and identifying the type of sound source of the sound.

Further, the distance may be estimated based on a predetermined relational expression between the calculated attenuation amount of the volume with respect to the standard volume and the distance.

According to this, since there is a relationship between the volume attenuation amount and the distance, the distance between the sound collector and the sound source can be more easily estimated by using a predetermined relational expression indicating the relationship.

Also, a sound signal indicating an arbitrary sound is used as input data, and the type of the sound source of the arbitrary sound is used as correct data to learn a learning model, and from the acquired sound signal, the sound collected by the sound collector. The type of sound source may be identified.

According to this, it is possible to easily identify the type of the sound source of the sound collected by the sound collector by using the machine-learned learning model.

Further, when the type of the sound source of the sound collected by the sound collecting device cannot be identified from the acquired sound signal, the feature amount of the sound signal is further calculated, and the sound collecting device and the sound collecting device collect the sound. Obtaining distance information indicating the distance from the sound source of the sounded sound, and registering in the database the calculated feature amount and the calculated volume at the distance indicated by the obtained distance information as the standard volume in association with each other. You may

According to this, for a sound source of a type that could not be identified, since the feature amount of the sound of the sound source and the standard volume are stored in a database, the distance between the sound collector and the sound source having the feature amount can be estimated.

Further, when the acquired sound signals indicate sounds from a plurality of sound sources, the sound signals are separated for each sound source, and the volume is calculated, the type is identified, and the distance is calculated for each of the separated sound signals. Estimation and output of the estimation result may be performed.

According to this, the distance to the sound collector can be easily estimated for each of the plurality of sound sources.

Further, it is possible to determine whether the type of the identified sound source outputs audible sound or non-audible sound, and estimate the distance according to the determination result. Specifically, the distance is estimated based on a predetermined relational expression between the attenuation amount of the volume calculated with respect to the standard volume and the distance, and the relational expression is such that the identified type outputs an audible sound. It may be determined in advance for the case of outputting an inaudible sound and the case of outputting an inaudible sound.

According to this, it is possible to estimate the distance between the sound collector and the sound source that outputs audible sound, and the distance between the sound collector and the sound source that outputs inaudible sound.

Further, the sound collector is a microphone array composed of a plurality of microphones, and the direction of the sound source of the sound with respect to the sound collector is determined based on the sound collection time difference of the sound collected by each of the plurality of microphones. can be estimated.

According to this, the accurate position of the sound source can be estimated from the distance between the sound source of the sound collected by the sound collector and the sound collector and the direction of the sound source with respect to the sound collector.

Further, a correspondence relationship between the type of the identified sound source and the estimation result of the distance between the sound collector and the sound source is accumulated each time the distance is estimated, and the accumulated correspondence relationship and the sound collection are stored. Based on the estimation result of the distance between the device and the sound source of the sound collected by the sound collecting device, the accuracy of identifying the type of the sound source is determined, and the determination result is fed back to determine the type of the sound source. May be used for identification.

According to this, it is possible to improve the accuracy of identifying the type of sound source.

An information processing device according to an aspect of the present disclosure includes a sound signal acquisition unit that acquires a sound signal indicating sound collected by a sound collector, and a sound signal that is collected by the sound collector from the acquired sound signal. A calculation unit that calculates the volume, an identification unit that identifies the type of sound source of the sound collected by the sound collector from the acquired sound signal, and the type of the sound source and the sound from the sound source at a predetermined distance from the sound source. Based on the standard volume corresponding to the identified type and the calculated volume in a database in which the standard volume is associated in advance, the sound collector and the sound collector collect sound an estimating unit for estimating the distance to the sound source of the generated sound; and an output unit for outputting the estimation result.

According to this, it is possible to provide an information processing device capable of easily estimating the distance between the sound collector and the sound source.

A program according to an aspect of the present disclosure is a program that causes a computer to execute the above information processing method.

According to this, it is possible to provide a program capable of easily estimating the distance between the sound collector and the sound source.

Hereinafter, embodiments will be specifically described with reference to the drawings.

It should be noted that the embodiments described below are all comprehensive or specific examples. Numerical values, shapes, components, arrangement positions and connection forms of components, steps, order of steps, and the like shown in the following embodiments are examples and are not intended to limit the present disclosure. In addition, among the constituent elements in the following embodiments, constituent elements that are not described in independent claims representing the highest concept will be described as optional constituent elements.

(Embodiment 1)
Embodiment 1 will be described below with reference to FIGS. 1 to 3. FIG.

FIG. 1 is a block diagram showing an example of the configuration of an information processing device 1 according to Embodiment 1. As shown in FIG. FIG. 1 also shows the sound collector 100 and a sound source 200 of a sound generated at a position d [m] away from the sound collector 100 in the space where the sound collector 100 is installed.

The information processing device 1 is a device for estimating the distance between the sound collector 100 and the sound source 200 of the sound collected by the sound collector 100 . The information processing device 1 is a computer provided around the space in which the sound collector 100 is provided, or a server device or the like provided in a place different from the space. Further, the information processing device 1 may be provided integrally with the sound collecting device 100, and may be, for example, a portable device. In other words, the position of the sound collector 100 , that is, the distance between the current position and the sound source 200 may be estimated while carrying the information processing device 1 integrated with the sound collector 100 .

The sound collection device 100 is, for example, one microphone. The sound collector 100 converts the collected sound into a sound signal (electrical signal) and outputs the signal to the information processing device 1 . The sound collector 100 is installed in a space (a room or the like) where the distance to the sound source is estimated. The sound collector 100 and the information processing device 1 are connected by wire or wirelessly.

As shown in FIG. 1 , the information processing device 1 includes a sound signal acquisition section 10 , a calculation section 20 , an identification section 30 , a learning model 31 , an estimation section 40 , a database 41 and an output section 50 . The information processing apparatus 1 is a computer including a processor (microprocessor), a user interface, a communication interface (a communication circuit, etc., not shown), a memory, and the like. The user interface includes, for example, a display such as an LCD (Liquid Crystal Display) or an input device such as a keyboard or touch panel. The memory is ROM, RAM, etc., and can store a control program (computer program) executed by the processor. Note that the information processing apparatus 1 may have one memory, or may have a plurality of memories. A learning model 31 and a database 41, which will be described later, are stored in one or more memories.

The sound signal acquisition unit 10, the calculation unit 20, the identification unit 30, the estimation unit 40, and the output unit 50, which are functional components of the processor, are implemented by the processor operating according to the control program. Further, the processor operates according to the control program to perform processing for controlling the communication interface, the user interface, and the like. Note that the sound signal acquisition unit 10a, the identification unit 30a, the estimation units 40a and 40b, the calculation unit 60, the distance information acquisition unit 70, the registration unit 80, and the determination unit 90 in Embodiments 2 to 5 described later are also controlled by the processor. It is realized by operating according to a program.

The sound signal acquisition unit 10, the calculation unit 20, the identification unit 30, the learning model 31, the estimation unit 40, the database 41, and the output unit 50 will be described with reference to FIG.

FIG. 2 is a flow chart showing an example of the operation of the information processing device 1 according to the first embodiment.

First, the sound signal acquisition unit 10 acquires a sound signal indicating sound collected by the sound collector 100 (step S11). Specifically, when the communication interface included in the information processing device 1 receives a sound signal output (transmitted) from the sound collecting device 100, the sound signal acquisition unit 10 acquires the sound signal. Note that the sound collected by the sound collector 100 is the sound from the sound source 200 .

Next, the calculator 20 calculates the volume of the sound collected by the sound collector 100 from the sound signal acquired by the sound signal acquirer 10 (step S12). Calculation of the volume is generally performed, so a detailed description thereof will be omitted. For example, the volume can be calculated from the amplitude indicated by the sound signal (electrical signal).

Next, the identifying unit 30 identifies the type of the sound source 200 of the sound collected by the sound collecting device 100 from the sound signal obtained by the sound signal obtaining unit 10 (step S13). For example, the identification unit 30 uses a sound signal indicating an arbitrary sound as input data and the type of the sound source of the arbitrary sound as correct data to learn a learning model 31, and the sound signal acquired by the sound signal acquisition unit 10. , the type of the sound source 200 of the sound collected by the sound collector 100 is identified. Note that the sound signal may be converted into a frequency spectrum or the like and used as input data. Also, the learning model 31 is, for example, a neural network or the like. In the following, for example, a case of learning "male voice" as the type of sound source will be described.

First, a number of sound signals known to represent male voices are prepared as arbitrary sounds. Then, the prepared sound signals representing male voices are input to the learning model 31 as input data, and the learning model 31 is trained so that the correct answer is "male voice". As a result, when a sound signal whose type is unknown is input to the learning model 31 that has completed learning, and the type is "male voice", the learning model 31 outputs "male voice" as the correct answer. Become so.

Also, in FIG. 2, the type of sound source 200 is identified after the calculation of volume, but the calculation of volume may be performed after the type of sound source 200 is identified. That is, the order in which steps S12 and S13 are performed may be reversed. Also, the calculation of volume and the identification of the type of sound source 200 may be performed in parallel.

Next, the estimation unit 40 corresponds to the type identified by the identification unit 30 in the database 41 in which the type of sound source and the standard volume, which is the volume of the sound from the sound source at a predetermined distance from the sound source, are associated in advance. The distance between the sound collector 100 and the sound source 200 of the sound collected by the sound collector 100 is estimated based on the standard volume and the volume calculated by the calculator 20 (step S14). Here, the database 41 will be explained using FIG.

FIG. 3 is a table showing an example of the database 41. As shown in FIG.

Generally, the volume of sound is determined to some extent depending on the type of sound source. Therefore, for various types of sound sources, the volume of the sound from the sound source at a predetermined distance from the sound source is measured as the standard volume. This makes it possible to create a database 41 in which various types of sound sources and standard volumes are associated with each other. FIG. 3 shows an example of the standard volume of "male voice", "female voice", "car noise", "vacuum cleaner sound", and "flowing water sound" when the predetermined distance is 1 m. It is shown. It should be noted that since the volume of a human voice differs from person to person, for example, by measuring the standard volume of the voice of a person using the space in which the sound collector 100 is installed, the voice and the standard volume of each person can be measured. may be associated with each other and stored in a database.

Specifically, the estimation unit 40 calculates a predetermined distance between the sound volume attenuation amount calculated by the calculation unit 20 and the sound source of the sound collected by the sound collection device 100 with respect to the standard sound volume. The distance between the sound collector 100 and the sound source 200 is estimated based on the relational expression. An example of the relational expression is shown in Equation 1 below.

In Equation 1, d [m] indicates the distance between the sound collector 100 and the sound source 200, r [m] indicates a predetermined distance, A0 [dB] indicates the standard volume at the predetermined distance, and A [dB] is The volume calculated in the calculator 20 is shown. Also, in Equation 1, the sound source 200 is regarded as a point sound source.

For example, when the type of the sound source 200 of the sound collected by the sound collector 100 identified by the identification unit 30 is “male voice”, the predetermined distance 1 m corresponding to the type “male voice” is stored in the database 41. The standard sound volume is 55 dB. Also, assume that the volume calculated by the calculation unit 20 at this time is 43 dB. In this case, the estimation unit 40 substitutes r=1, A0=55, and A=43 in Equation 1 to estimate the distance between the sound collector 100 and the sound source 200 to be about 4 m.

Then, the output unit 50 outputs the estimation result of the estimation unit 40 (step S15). For example, the output unit 50 outputs the estimation result to a user interface included in the information processing device 1 or a device such as a mobile terminal or a server device that can communicate with the information processing device 1 . Note that the output unit 50 may output not only the distance between the sound collecting device 100 and the sound source 200 but also the type of the sound source as the estimation result.

As described above, if the standard volumes of sounds of various types of sound sources are measured and the database 41 is generated in advance, then the sound collector 100 collects the sound in the space where the distance of the sound source 200 is estimated. The distance between the sound collector 100 and the sound source 200 can be easily estimated simply by calculating the volume of the sound and identifying the type of the sound source 200 of the sound. Specifically, in order to estimate the distance, it is not necessary to install the sound collector 100 at a plurality of positions, and it is not necessary to create a unique database for each space. Therefore, even in a space where the distance between the sound collecting device 100 and the sound source 200 has not been estimated due to the time and effort involved, the distance can be easily estimated, and the distance can be estimated. You can increase the space.

In addition, since there is a relationship between the volume attenuation amount and the distance, using a predetermined relational expression (for example, Equation 1) showing the relationship makes it easier to determine the distance between the sound collecting device 100 and the sound source 200. can be estimated to

Further, by using the machine-learned learning model 31, the type of the sound source 200 of the sound collected by the sound collector 100 can be easily identified.

(Embodiment 2)
Next, Embodiment 2 will be described with reference to FIGS. 4 and 5. FIG.

FIG. 4 is a block diagram showing an example of the configuration of an information processing device 1a according to the second embodiment.

The information processing apparatus 1a according to Embodiment 2 is different from Embodiment 1 in that it includes an identification unit 30a instead of the identification unit 30, and further includes a calculation unit 60, a distance information acquisition unit 70, and a registration unit 80. It differs from the information processing device 1 . Since other points are the same as those in the information processing apparatus 1 according to the first embodiment, description thereof will be omitted, and the description will focus on the above-mentioned different points using FIG.

In the second embodiment, the learning model 31 is not made to learn the type of the sound source 201 of the sound generated in the space where the sound collector 100 is installed, and the database 41 stores the type of the sound source 201 and the corresponding type. Suppose the corresponding standard volume is not included.

FIG. 5 is a flow chart showing an example of the operation of the information processing device 1a according to the second embodiment.

The identifying unit 30a identifies the type of the sound source 201 of the sound collected by the sound collecting device 100 from the sound signal acquired by the sound signal acquiring unit 10, as in step S13 of FIG. At this time, for example, the identification unit 30a determines whether or not the type of the sound source 201 can be identified (step S21). For example, when the learning model 31 learns the type of the sound source 201 of the sound generated in the space where the sound collector 100 is installed, the identification unit 30a can identify the type of the sound source 201, and the type of the sound source 201 is not learned, the identification unit 30a cannot identify the type of the sound source 201. When the identifying unit 30a determines that the type of the sound source 201 can be identified (Yes in step S21), the processes in steps S14 and S15 are performed as in the first embodiment.

When the identification unit 30 a determines that the type of the sound source 201 cannot be identified (No in step S<b>21 ), the calculation unit 60 calculates the feature amount of the sound signal acquired by the sound signal acquisition unit 10 . The feature amount indicates the feature of the sound signal, and is, for example, a frequency spectrum. For example, the feature amount is stored in a memory or the like in which the learning model 31 is stored.

Next, the distance information acquisition unit 70 acquires distance information indicating the distance between the sound collector 100 and the sound source 201 of the sound collected by the sound collector 100 (step S23). For example, in a space in which the sound collector 100 is installed, the user actually measures the distance between the sound collector 100 and the sound source 201, inputs the measurement result via a user interface provided in the information processing device 1, or receives information. The distance information acquisition unit 70 acquires the distance information by inputting via a mobile terminal or the like that can communicate with the processing device 1 .

Then, the registration unit 80 associates the feature amount calculated by the calculation unit 60 with the volume calculated by the calculation unit 20 at the distance indicated by the distance information acquired by the distance information acquisition unit 70 as the standard volume. It is registered in the database 41 (step S24). As described above, the standard volume is the volume of the sound from the sound source at a predetermined distance from the sound source. The volume is the volume calculated by the calculation unit 20 . For example, if the distance indicated by the distance information acquired by the distance information acquisition unit 70 is 2 m and the sound volume calculated by the calculation unit 20 is 49 dB, the feature amount calculated by the calculation unit 60 and A standard sound volume of 49 dB at a predetermined distance of 2 m is associated.

Henceforth, it becomes possible to estimate the distance between the sound source 201 and the sound source of the same type (that is, a sound source that generates sound having an equivalent feature value: called an equivalent sound source) and the sound collector 100 . Specifically, the identification unit 30a confirms that a feature amount equivalent to the feature amount of the sound signal indicating the sound from the equivalent sound source acquired by the sound signal acquisition unit 10 is stored in the memory. It is determined that the type of the sound source 201 of the sound collected by the sound collector 100 can be identified from the sound signal. In other words, even if the learning model 31 has not learned the type of equivalent sound source, it is no longer determined that the type of sound source cannot be identified (that is, No in step S21), and the distance estimation process can be performed.

The estimation unit 40 extracts the feature identified by the identification unit 30 in the database 41 in which the feature amount registered in the registration unit 80 and the standard sound volume at the distance indicated by the distance information obtained by the distance information acquisition unit 70 are associated with each other. The distance between the sound collector 100 and the sound source 201 of the sound collected by the sound collector 100 is estimated based on the standard volume corresponding to the volume and the volume of the sound of the equivalent sound source calculated by the calculation unit 20 . For example, assume that the volume calculated by the calculator 20 is 43 dB. In this case, since the standard sound volume at the predetermined distance of 2 m is 49 dB as described above, the estimation unit 40 substitutes r=2, A0=49, and A=43 in Equation 1 to obtain the sound collector The distance between 100 and the equivalent sound source is estimated to be about 4m.

As described above, for the sound source 201 of the type that could not be identified, the sound feature amount and the standard volume of the sound source 201 are stored in a database. be able to estimate the distance to

The learning model 31 may be updated so as to be able to deal with an unknown sound source 201 by making the learning model 31 learn using the sound signal representing the sound from the sound source 201 as input data and the type of the sound source 201 as correct data. good.

(Embodiment 3)
Next, Embodiment 3 will be described with reference to FIG.

FIG. 6 is a block diagram showing an example of the configuration of an information processing device 1b according to the third embodiment.

The information processing apparatus 1b according to the third embodiment differs from the information processing apparatus 1 according to the first embodiment in that a sound signal acquisition section 10a is provided instead of the sound signal acquisition section 10. FIG. Since other points are the same as those in the information processing apparatus 1 according to the first embodiment, the description is omitted, and the description will focus on the above-mentioned different points.

The sound signal acquisition unit 10 a separates the sound signals for each of the sound sources 202 and 203 when the acquired sound signals indicate sounds from a plurality of sound sources 202 and 203 . Separating a sound signal for each sound source from a sound signal containing sound components from multiple sound sources is a commonly used technique, so detailed description is omitted here, but the separation method is not particularly limited. . Then, for each of the separated sound signals, the calculation unit 20 calculates the volume, the identification unit 30 identifies the type, the estimation unit 40 estimates the distance, and the output unit 50 outputs the estimation result. conduct. That is, the distance d1 is estimated for the sound source 202, and the distance d2 is estimated for the sound source 203.

Note that the information processing device 1b may include a plurality of sets of the calculation unit 20, the identification unit 30, and the estimation unit 40, and the above processes may be performed in parallel on each of the separated sound signals. As a result, the distances d1 and d2 can be estimated in real time for both of the multiple sound sources 202 and 203 .

As described above, when the acquired sound signals indicate sounds from the plurality of sound sources 202 and 203, by separating the sound signals for each of the sound sources 202 and 203, the collected sound for each of the plurality of sound sources 202 and 203 is obtained. The distance to the device 100 can be easily estimated.

(Embodiment 4)
Next, Embodiment 4 will be described with reference to FIG.

FIG. 7 is a block diagram showing an example of the configuration of an information processing device 1c according to the fourth embodiment.

The information processing apparatus 1c according to the fourth embodiment differs from the information processing apparatus 1 according to the first embodiment in that an estimating section 40a is provided instead of the estimating section 40 and a determining section 90 is provided. Since other points are the same as those in the information processing apparatus 1 according to the first embodiment, the description is omitted, and the description will focus on the above-mentioned different points.

The determination unit 90 determines whether the type of the sound source 200 identified by the identification unit 30 outputs audible sound or non-audible sound. For example, the determination unit 90 determines whether the type of the sound source 200 identified by the identification unit 30 outputs audible sound or non-audible sound based on the frequency spectrum of the sound signal acquired by the sound signal acquisition unit 10. can determine what

Then, the estimation unit 40 a estimates the distance between the sound collector 100 and the sound source 200 also according to the determination result of the determination unit 90 . Specifically, in the same way as the estimating unit 40, the estimating unit 40a sets a predetermined relationship between the volume attenuation calculated by the calculating unit 20 with respect to the standard volume and the distance between the sound collecting device 100 and the sound source. The distance between the sound collecting device 100 and the sound source 200 is estimated based on the equation. are predetermined respectively. This is because audible sound and non-audible sound have different volume attenuation amounts depending on the distance. For example, when the type of the sound source 200 outputs audible sound, the relational expression is Equation (1). On the other hand, when the type of the sound source 200 outputs inaudible sound, the relational expression is different from the expression (1). Furthermore, inaudible sounds include ultrasonic waves and infrasound frequencies, and the relational expressions for ultrasonic waves and infrasound sounds are different from each other.

As described above, the distance between the sound collector 100 and the sound source 200 that outputs audible sound and the distance between the sound collector 100 and the sound source 200 that outputs inaudible sound can be estimated.

For example, the distance between the sound source 200 that outputs ultrasonic waves as inaudible sound and the sound collector 100 can be estimated. Specifically, when an abnormality occurs in a device, an ultrasonic wave may be generated from the portion of the device where the abnormality has occurred, making it possible to specify the device in which the abnormality has occurred. Further, since the content of the abnormality and the frequency of the ultrasonic waves may correspond, the content of the abnormality may be output together with the estimation result.

Also, for example, the distance between the sound source 200 that outputs infrasound as inaudible sound and the sound collector 100 can be estimated. Specifically, when a tornado occurs, an ultra-low frequency wave is generated from the tornado, and the distance to the location where the tornado occurred can be estimated. In this case, the space in which the sound collector 100 is installed has a scale of several tens of kilometers, including the outdoors.

(Embodiment 5)
Next, Embodiment 5 will be described with reference to FIG.

FIG. 8 is a block diagram showing an example of the configuration of an information processing device 1d according to the fifth embodiment.

The information processing device 1d according to the fifth embodiment differs from the information processing device 1 according to the first embodiment in that an estimating unit 40b is provided instead of the estimating unit 40. FIG. Further, the fifth embodiment differs from the first embodiment in that a sound collector 100a is installed instead of the sound collector 100 in a space where distance estimation is performed. Since other points are the same as those in the information processing apparatus 1 according to the first embodiment, the description is omitted, and the description will focus on the above-mentioned different points.

In Embodiment 5, the sound collector 100a is a microphone array consisting of a plurality of microphones. The multiple microphones are arranged at different positions in the sound collector 100a. As a result, the sound from the sound source 200 reaches each microphone with a time difference.

As with the estimating unit 40, the estimating unit 40b uses sound signals of sounds collected by any one of the plurality of microphones included in the sound collecting device 100a to determine the relationship between the sound collecting device 100a and the sound source 200. The distance is estimated, and the direction of the sound source 200 of the sound with respect to the sound collector 100a is estimated based on the sound collection time difference of the sound collected by each of the plurality of microphones. The estimating unit 40b can estimate the direction of the sound source 200 with respect to the sound collector 100a from the positional relationship of the plurality of microphones and the time difference (sound collection time difference) in which the sound from the sound source 200 is collected by each microphone.

As described above, the direction of the sound source 200 with respect to the sound collector 100a can also be estimated. The exact position of the sound source 200 can be estimated from the direction with respect to 100a.

(Embodiment 6)
Next, Embodiment 6 will be described with reference to FIG.

FIG. 9 is a block diagram showing an example of the configuration of an information processing device 1e according to the sixth embodiment.

The information processing apparatus 1e according to the sixth embodiment differs from the information processing apparatus 1 according to the first embodiment in that an identification accuracy determination unit 42 is further provided. Further, the sixth embodiment differs from the first embodiment in that an estimating unit 40 c is provided instead of the estimating unit 40 and a database 41 a is provided instead of the database 41 . Since other points are the same as those in the information processing apparatus 1 according to the first embodiment, the description is omitted, and the description will focus on the above-mentioned different points.

The estimating unit 40c accumulates, for example, the database 41a, each time the distance is estimated, the correspondence relationship between the type of sound source of the sound identified by the identifying unit 30 and the estimation result of the distance between the sound collector 100 and the sound source. . Note that the database 41a is the same as the database 41 except that such correspondence relationships are accumulated. Further, for example, a database for storing such correspondence may be provided separately from the database 41 .

For example, when sound is generated from a specific type of sound source, the estimation results may tend to be the same each time the distance is estimated for the sound source. This is because when a specific type of sound source is a sound source that is fixed in a facility or the like, the volume of the sound from the sound source has the same tendency, and the estimation result of the distance estimated from the volume of the sound also has the same tendency. Therefore, for such a specific type of sound source, the actual distance from the sound collecting device 100 can be specified to some extent from the tendency of the distance estimation results.

The identification accuracy determination unit 42 identifies the type of the sound source 200 based on the accumulated correspondence relationship and the estimation result of the distance between the sound collector 100 and the sound source 200 of the sound collected by the sound collector 100. to determine the accuracy of Specifically, if the distance indicated by the estimation result for the type of sound identified by the identification unit 30 is substantially the same as the tendency of the distance corresponding to the type in the correspondence relationship accumulated in the database 41a , the estimation by the estimation unit 40c is performed with high accuracy, that is, it can be determined that the discrimination of the type of sound by the identification unit 30 is performed with high accuracy. On the other hand, if the distance indicated by the estimation result for the type of sound identified by the identifying unit 30 deviates from the tendency of the distance corresponding to the type in the correspondence accumulated in the database 41a, the estimating unit 40c It can be determined that the estimation is not accurately performed, that is, the identification of the sound type by the identification unit 30 is not accurately performed.

Then, the identification accuracy determination unit 42 feeds back the determination result to the identification unit 30, and the identification unit 30 uses the determination result to identify the type of sound source. Such feedback can improve the accuracy of identifying the type of sound source.

(Other embodiments)
Although the information processing apparatus of the present disclosure has been described above based on the embodiments, the present disclosure is not limited to the above embodiments. As long as they do not deviate from the spirit of the present disclosure, modifications that can be made by those skilled in the art to the present embodiment, and forms constructed by combining the components of different embodiments are also included within the scope of the present disclosure. be

For example, in the third to sixth embodiments, each information processing apparatus may have functions corresponding to the identifying section 30a, the calculating section 60, the distance information acquiring section 70, and the registering section 80 in the second embodiment. In other words, in Embodiments 3 to 6, each information processing apparatus may have a function of registering in the databases 41 and 41a the feature amount and the standard volume of the sound source for which the type cannot be identified.

Further, for example, in the second, fourth to sixth embodiments, each information processing apparatus may have a function corresponding to the sound signal acquiring section 10a in the third embodiment. That is, in Embodiments 2, 4 to 6, each information processing device separates a sound signal for each sound source when there are a plurality of sound sources, and calculates volume and volume for each of the separated sound signals. It may have a function of identifying the type, estimating the distance, and outputting the estimation result.

Further, for example, in the second, third, fifth and sixth embodiments, each information processing apparatus may have functions corresponding to the estimation unit 40a and the determination unit 90 in the fourth embodiment. That is, in Embodiments 2, 3, 5 and 6, each information processing device determines whether the type of sound source outputs audible sound or non-audible sound, and may have the function of estimating the distance.

Further, for example, in Embodiments 2 to 4 and 6, the sound collection device 100 may be a sound collection device 100a that is a microphone array composed of a plurality of microphones, and each information processing device may be the It may have a function corresponding to the estimation unit 40b. That is, in Embodiments 2 to 4 and 6, each information processing device may have a function of estimating the direction of the sound source with respect to the sound collector 100a.

Further, for example, in the second to fifth embodiments, each information processing device may have functions corresponding to the identification accuracy determination unit 42 and the estimation unit 40c in the sixth embodiment. That is, in Embodiments 2 to 5, each information processing device may have a function for improving the accuracy of identifying the type of sound source.

Further, for example, in the above embodiments, the identification units 30 and 30a use the learning model 31 to identify the type of sound source, but the learning model 31 may not be used. For example, the frequency spectrum or the like indicated by the sound signal has different characteristics depending on the type of sound source. may

Further, for example, when the information processing device is realized by a server device or the like, the functional components included in the information processing device may be distributed and arranged in a plurality of server devices.

Further, the present disclosure can be realized not only as an information processing apparatus, but also as an information processing method including steps of processing performed by each component constituting the information processing apparatus.

Specifically, as shown in FIG. 2, the information processing method acquires a sound signal indicating the sound collected by the sound collector 100 (step S11), and the sound collector 100 collects the sound from the acquired sound signal. The volume of the sound produced is calculated (step S12), the type of the sound source 200 of the sound collected by the sound collector 100 is identified from the acquired sound signal (step S13), and the type of the sound source and the volume at a predetermined distance from the sound source Based on the standard volume corresponding to the identified type in the database 41 in which the standard volume, which is the volume of the sound from the sound source, is associated in advance, and the calculated volume, the sound collector 100 and the sound collector 100 The distance between the collected sound and the sound source 200 is estimated (step S14), and the estimation result is output (step S15).

Also, for example, those steps may be executed by a computer (computer system). The present disclosure can be realized as a program for causing a computer to execute the steps included in those methods. Furthermore, the present disclosure can be implemented as a non-temporary computer-readable recording medium such as a CD-ROM recording the program.

For example, when the present disclosure is implemented by a program, each step is executed by executing the program using hardware resources such as a computer's CPU, memory, and input/output circuits. That is, each step is executed by the CPU obtaining data from a memory, an input/output circuit, or the like, performing an operation, or outputting the operation result to the memory, an input/output circuit, or the like.

Further, each of the plurality of components included in the information processing apparatus of the above embodiments may be implemented as a dedicated or general-purpose circuit. These components may be implemented as one circuit or as multiple circuits.

Also, the plurality of components included in the information processing apparatus of the above embodiments may be implemented as an LSI (Large Scale Integration), which is an integrated circuit (IC). These components may be made into one chip individually, or may be made into one chip so as to include some or all of them. LSIs are sometimes called system LSIs, super LSIs, or ultra LSIs depending on the degree of integration.

Also, the integrated circuit is not limited to an LSI, and may be realized by a dedicated circuit or a general-purpose processor. A programmable FPGA (Field Programmable Gate Array) or a reconfigurable processor in which connection and setting of circuit cells inside the LSI can be reconfigured may be used.

Furthermore, if a technology for integrating circuits to replace LSIs emerges due to advances in semiconductor technology or another technology derived from it, it is natural that such technology will be used to integrate each component included in the information processing apparatus. good too.

In addition, forms obtained by applying various modifications that a person skilled in the art can think of to the embodiments, and forms realized by arbitrarily combining the components and functions in each embodiment within the scope of the present disclosure are also included in this disclosure.

One aspect of the present disclosure can be used, for example, in a device for identifying the position of a sound source.

1, 1a, 1b, 1c, 1d, 1e Information processing device 10, 10a Sound signal acquisition unit 20 Calculation unit 30, 30a Identification unit 31 Learning model 40, 40a, 40b, 40c Estimation unit 41, 41a Database 42 Identification accuracy determination unit 50 output unit 60 calculation unit 70 distance information acquisition unit 80 registration unit 90 determination unit 100, 100a sound collector 200, 201, 202, 203 sound source

Claims (9)

An information processing method executed by an information processing device,
Acquiring a sound signal indicating the sound collected by the sound collecting device,
calculating the volume of the sound collected by the sound collecting device from the acquired sound signal;
identifying the type of sound source of the sound collected by the sound collecting device from the acquired sound signal;
Based on the standard volume corresponding to the identified type and the calculated volume in a database in which the type of sound source and the standard volume, which is the volume of the sound from the sound source at a predetermined distance from the sound source, are associated in advance. , estimating the distance between the sound collector and the sound source of the sound collected by the sound collector;
output the estimation result,
Information processing methods. estimating the distance based on a predetermined relational expression between the calculated attenuation amount of the volume with respect to the standard volume and the distance;
The information processing method according to claim 1. Using a sound signal representing an arbitrary sound as input data and the type of the sound source of the arbitrary sound as correct data, a learning model is trained, and from the acquired sound signal, the sound source of the sound collected by the sound collecting device. identify the type
The information processing method according to claim 1 or 2. When the type of the sound source of the sound collected by the sound collecting device cannot be identified from the acquired sound signal, further calculating a feature amount of the sound signal,
Acquiring distance information indicating the distance between the sound collector and a sound source of the sound collected by the sound collector;
registering the calculated feature amount and the calculated volume at the distance indicated by the acquired distance information as the standard volume in association with each other in the database;
The information processing method according to any one of claims 1 to 3. if the obtained sound signals represent sounds from different types of sound sources, separating the sound signals for each sound source;
calculating the volume, identifying the type, estimating the distance, and outputting the estimation result for each of the separated sound signals;
The information processing method according to any one of claims 1 to 4. The sound collecting device is a microphone array consisting of a plurality of microphones,
Further, estimating the direction of the sound source of the sound with respect to the sound collector based on the sound collection time difference of the sound collected by each of the plurality of microphones.
The information processing method according to any one of claims 1 to 5 . accumulating the correspondence relationship between the type of the identified sound source and the estimation result of the distance between the sound collector and the sound source each time the distance is estimated;
Determining the accuracy of identifying the type of sound source based on the accumulated correspondence relationship and the estimation result of the distance between the sound collector and the sound source of the sound collected by the sound collector,
The information processing method according to any one of claims 1 to 6 , wherein the determination result is fed back and used for identifying the type of sound source. a sound signal acquisition unit that acquires a sound signal indicating the sound collected by the sound collecting device;
a calculation unit that calculates the volume of the sound collected by the sound collecting device from the acquired sound signal;
an identification unit that identifies the type of sound source of the sound collected by the sound collecting device from the acquired sound signal;
The standard volume corresponding to the identified type and the calculated volume in a database in which the type of sound source and the standard volume, which is the volume of the sound from the sound source at a predetermined distance from the sound source, are associated in advance. an estimating unit for estimating the distance between the sound collector and the sound source of the sound collected by the sound collector, based on
and an output unit that outputs the estimation result,
Information processing equipment. A program that causes a computer to execute the information processing method according to any one of claims 1 to 7 .

Images