Sound environment analysis system, sound environment analysis method, and program

Abstract

This system provides an acoustic environment analysis system that can analyze the acoustic environment of a space where various sounds can occur, based on objective criteria regardless of the user. [Solution] The sound environment analysis system 1 comprises a sound data acquisition unit 11, a separation unit 13, an identification unit 14, and an unpleasant sound identification unit 15. The sound data acquisition unit 11 acquires sound data in the space. The separation unit 13 separates the sound data into one or more partial sound data for each sound source in the space. The identification unit 14 identifies the type of each of the one or more partial sound data and assigns a label to it. The unpleasant sound identification unit 15 identifies which label in the sound data is an unpleasant sound. The unpleasant sound identification unit 15 identifies the sound of the label as an unpleasant sound if the sound pressure of the partial sound data to which a label has been assigned is equal to or greater than a first threshold.

Description

Technical Field

[0001] The present disclosure relates to a sound environment analysis system, a sound environment analysis method, and a program, and particularly to a sound environment analysis system, a sound environment analysis method, and a program for analyzing sounds generated in a space.

Background Art

[0002] In a space, it has been proposed to provide a comfortable sound environment. For example, the sound selection and processing device of Patent Document 1 identifies and selectively removes sounds that cause discomfort from mixed sounds.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, in the sound selection and processing device of Patent Document 1, it is necessary to obtain from the user which part of the mixed sound causes discomfort.

[0005] An object of the present disclosure is to provide a sound environment analysis system, a sound environment analysis method, and a program that can analyze a sound environment based on objective criteria without depending on a user for a space where various sounds can occur.

Means for Solving the Problems

[0006] A sound environment analysis system according to one aspect of the present disclosure comprises a sound data acquisition unit, a separation unit, an identification unit, and an unpleasant sound identification unit. The sound data acquisition unit acquires sound data in a space. The separation unit separates the sound data into one or more partial sound data for each sound source in the space. The identification unit identifies each of the one or more partial sound data and assigns a label to it. The unpleasant sound identification unit identifies which of the sound data with the label is an unpleasant sound. The unpleasant sound identification unit identifies the sound of the label as an unpleasant sound if the sound pressure of the partial sound data to which the label is assigned is equal to or greater than a first threshold.

[0007] A sound environment analysis method according to one aspect of the present disclosure is performed by one or more processors. The sound environment analysis method includes a sound data acquisition step, a separation step, an identification step, and an unpleasant sound identification step. In the sound data acquisition step, sound data in a space is acquired. In the separation step, the sound data is separated into one or more partial sound data for each sound source in the space. In the identification step, each of the one or more partial sound data is identified and a label is assigned to it. In the unpleasant sound identification step, it is determined that which of the sound data with the label is an unpleasant sound. In the unpleasant sound identification step, if the sound pressure of the partial sound data to which the label is assigned is equal to or greater than a first threshold, the sound of the label is determined to be an unpleasant sound.

[0008] A program according to one aspect of this disclosure causes one or more processors to execute the sound environment analysis method. [Effects of the Invention]

[0009] According to one aspect of this disclosure, the sound environment analysis system, sound environment analysis method, and program make it possible to analyze the sound environment of a space where various sounds may occur, based on objective criteria, regardless of the user. [Brief explanation of the drawing]

[0010] [Figure 1]Figure 1 is a block diagram of the sound environment analysis system according to Embodiment 1. [Figure 2] Figure 2 is a floor plan of the space where the above-mentioned sound environment analysis system is installed. [Figure 3] Figure 3 is a flowchart showing the operation of the sound environment analysis system described above. [Figure 4] Figure 4 is a flowchart showing the operation of the sound environment analysis system according to Embodiment 2. [Figure 5] Figure 5 is a flowchart showing the operation of the sound environment system according to Embodiment 3. [Modes for carrying out the invention]

[0011] The sound environment analysis system, sound environment analysis method, and program according to Embodiments 1 to 3 will be described in detail below with reference to the drawings. However, the figures described in the embodiments below are schematic diagrams, and the ratios of the size and thickness of each component do not necessarily reflect the actual dimensional ratios. Furthermore, the configurations described in the embodiments below are merely examples of this disclosure. This disclosure is not limited to the embodiments below, and various modifications are possible depending on the design, etc., as long as the effects of this disclosure can be achieved.

[0012] (Embodiment 1) (1) Overview The sound environment analysis system 1 (see Figure 1) is a system that analyzes the sound generation conditions occurring in space SP1 (see Figure 2).

[0013] Space SP1 is a space that is divided into a single section, such as an office, shop, factory, warehouse, building, school, welfare facility, or hospital.

[0014] (2) Composition As shown in Figure 1, the sound environment analysis system 1 comprises a sound data acquisition unit 11, a separation unit 13, an identification unit 14, and an unpleasant sound identification unit 15. The sound environment analysis system 1 includes a sound environment analysis device 10 and a sensor SN1. The sound environment analysis device 10 comprises a sound data acquisition unit 11, a storage unit 12, a separation unit 13, an identification unit 14, and an unpleasant sound identification unit 15.

[0015] Sensor SN1 is a sensor that acquires sound data from space SP1. Sensor SN1 includes multiple (two in Figure 1) sound sensor arrays SA1. As shown in Figure 2, the multiple (three in Figure 2) sound sensor arrays SA1 are arranged in space SP1, separated from each other. Here, it is preferable that the multiple sound sensor arrays SA1 are positioned so that any location in space SP1 is visible from at least two of the sound sensor arrays SA1. "A location in space SP1 is visible from the sound sensor arrays SA1" means that there are no sound-blocking structures such as walls on the line segment connecting the location in space SP1 and the sound sensor arrays SA1. That is, in Figure 2, it is preferable that at least two of the three line segments connecting any location in space SP1 to each of the multiple sound sensor arrays SA1a, SA1b, and SA1c do not intersect with sound-blocking structures such as walls.

[0016] Each of the multiple sound sensor arrays SA1 includes multiple (two in Figure 1) sound sensors SE1. Each of the multiple sound sensors SE1 is an element that converts sound into an electronic signal, such as a microphone. In the sound sensor array SA1, the multiple sound sensors SE1 differ from each other in at least one of their installation position and directivity. The sound sensor array SA1 includes, for example, four sound sensors SE1 arranged in an X shape in a plan view.

[0017] The sound data acquisition unit 11 acquires sound data within the space SP1. The sound data acquisition unit 11 acquires data from each of the multiple sound sensor arrays SA1 included in the sensor SN1, for example. The sound data acquisition unit 11 performs A / D conversion on each of the electrical signals output by the multiple sound sensors SE1. The sound data acquisition unit 11 stores the sound data for each sound sensor SE1 in the storage unit 12.

[0018] The memory unit 12 stores the position information of each of the plurality of sound sensor arrays SA1. The memory unit 12 stores, for example, the installation position of each of the plurality of sound sensor arrays SA1 and the orientation of each of the plurality of sound sensors SE1 in each sound sensor array SA1. Here, the position information of each of the plurality of sound sensor arrays SA1 is, for example, three-dimensional information indicating a position in the space SP1. Note that the position information of each of the plurality of sound sensor arrays SA1 may be two-dimensional information when viewing the space SP1 in a plan view and does not include height information.

[0019] In addition, the memory unit 12 stores the sound data output by the sound data acquisition unit 11.

[0020] The separation unit 13 separates the sound data into one or more partial sound data for each sound source in the space SP1. The separation unit 13 calculates, for example, the sound source direction and the frequency spectrum of the sound from the sound source for each of the plurality of sound sensor arrays SA1. Then, the separation unit 13 estimates the position of the sound source based on the position of each of the plurality of sound sensor arrays SA1 and separates the sound data into partial sound data for each sound source.

[0021] The identification unit 14 discriminates the type of one or more partial sound data and assigns a label. The identification unit 14 assigns one label to each of the one or more partial sound data, for example, using a learned model generated by supervised machine learning. When there are a plurality of partial sound data, the identification unit 14 assigns one label to each of the plurality of partial sound data. Note that the same one label may be assigned to two or more partial sound data. The label is the type of sound that can become noise when generated in the space SP1. The label includes, for example, "key type sound", "footstep sound", "printer operation sound", "male voice", "female voice". Note that the method by which the identification unit 14 assigns a label is not limited to using a learned model, and the type of sound may be discriminated based on the intensity distribution for each frequency, the loudness of the sound, the time change of the loudness, etc., which are the characteristics of the partial sound data.

[0022] The unpleasant sound identification unit 15 identifies which label in the sound data is an unpleasant sound. The unpleasant sound identification unit 15 calculates the sound pressure for each label, for example. The sound pressure for each label is, for example, the maximum sound pressure of one or more subtone data to which the label is assigned. For example, the maximum sound pressure for "key type sound" is the maximum sound pressure of one or more subtone data to which the label "key type sound" is assigned. Note that the sound pressure for each label may also be the average or median value of the sound pressure of one or more subtone data to which the label is assigned, or other representative values, but in that case, the first threshold and second threshold described later must also be set to match those representative values.

[0023] The unpleasant sound identification unit 15 identifies a labeled sound as unpleasant if the sound pressure of the labeled partial sound data is equal to or greater than a first threshold. For example, if the maximum sound pressure of the "key-typing sound" is equal to or greater than the first threshold, the unpleasant sound identification unit 15 identifies the "key-typing sound" as unpleasant, regardless of the magnitude of the sound pressure of other labeled sounds. The first threshold may be a uniform value regardless of the label, or it may differ for each label. If the first threshold differs for each label, it becomes possible to more appropriately identify unpleasant sounds even if the lower limit of sound pressure that people perceive as unpleasant differs depending on the type of sound. The unpleasant sound identification unit 15 may identify multiple labels as unpleasant sounds.

[0024] Furthermore, the unpleasant sound identification unit 15 identifies a sound with a label as unpleasant if the ratio of the sound pressure of the label to the sound pressure of other labels exceeds a certain threshold. In other words, the unpleasant sound identification unit 15 identifies a sound with a specific label as unpleasant if the ratio of the sound pressure of a partial sound data to which that specific label is assigned to is greater than or equal to a second threshold for sound data that is not assigned that specific label. For example, if the ratio of the sound pressure of "footsteps" to the sound pressure of other sounds is greater than or equal to a second threshold, "footsteps" is identified as an unpleasant sound. The second threshold may be a uniform value regardless of the label, or it may differ for each label. If the second threshold differs for each label, for example, by setting a smaller second threshold for labels of sounds that are less likely to be buried by other sounds, it becomes possible to identify unpleasant sounds more appropriately. This allows the unpleasant sound identification unit 15 to identify only unpleasant sounds that are not masked by other sounds, by setting, for example, the sound pressure at which an unpleasant sound is perceived regardless of other sounds as the first threshold, and the sound pressure ratio at which an unpleasant sound is perceived when its sound pressure is higher than that of other sounds as the second threshold. The unpleasant sound identification unit 15 may identify multiple labels as unpleasant sounds.

[0025] Furthermore, if the unpleasant sound identification unit 15 finds that the partial sound data to which a label has been assigned does not meet any of the above-mentioned "conditions for identifying an unpleasant sound," it determines that the sound with the label is not an unpleasant sound. Here, "the sound with the label is not an unpleasant sound" simply means that it does not match the conditions that the sound environment analysis system 1 determines to be an unpleasant sound, and does not mean that the sound environment analysis system 1 has definitively determined that the sound is not unpleasant for any user. The unpleasant sound identification unit 15 may also acquire information indicating whether or not a label is an unpleasant sound for which it has determined to be not an unpleasant sound. For example, the unpleasant sound identification unit 15 may provide the user with the partial sound data corresponding to the label that it has determined to be not an unpleasant sound, and obtain a response from the user indicating whether or not that sound is unpleasant to them. The information indicating whether or not a sound is unpleasant may be information obtained from multiple users, such as survey results. It is preferable that the acquisition of information indicating whether or not a sound is unpleasant is performed separately from the operation of the sound environment analysis system 1, as described later.

[0026] The sound environment analysis system 1 includes a computer system. The computer system mainly consists of a processor and memory as hardware. The functions of the sound environment analysis system 1 in this disclosure are realized by the processor executing a program recorded in the memory of the computer system. The program may be pre-recorded in the memory of the computer system, provided via a telecommunication line, or provided on a non-temporary recording medium such as a memory card, optical disk, or hard disk drive that can be read by the computer system. The processor of the computer system consists of one or more electronic circuits including semiconductor integrated circuits (ICs) or large-scale integrated circuits (LSIs). The integrated circuits such as ICs or LSIs referred to here are named differently depending on the degree of integration, and include integrated circuits called system LSIs, VLSIs (Very Large Scale Integration), or ULSIs (Ultra Large Scale Integration). Furthermore, FPGAs (Field-Programmable Gate Arrays) that are programmed after the manufacture of LSIs, or logic devices that allow for the reconfiguration of junction relationships or circuit compartments within LSIs, can also be used as processors. Multiple electronic circuits may be integrated onto a single chip or distributed across multiple chips. Multiple chips may be integrated onto a single device or distributed across multiple devices. The computer system referred to here includes a microcontroller having one or more processors and one or more memories. Therefore, the microcontroller also consists of one or more electronic circuits, including semiconductor integrated circuits or large-scale integrated circuits.

[0027] Furthermore, it is not essential for the sound environment analysis system 1 to have multiple functions integrated into a single housing; the components of the sound environment analysis system 1 may be distributed across multiple housings. For example, the sound data acquisition unit 11, storage unit 12, separation unit 13, identification unit 14, and unpleasant sound identification unit 15 do not all need to be components of the sound environment analysis device 10; at least one component may be a component of a different device. Moreover, at least some of the functions of the sound environment analysis system 1, for example, some of the functions of the identification unit 14, may be implemented by the cloud (cloud computing), etc.

[0028] Conversely, at least some of the functions of the sound environment analysis system 1, which are dispersed across multiple components in the embodiment, may be integrated into a single housing.

[0029] (3) Operation Figure 3 is a flowchart showing the operation of the sound environment analysis system 1 according to Embodiment 1.

[0030] The sound data acquisition unit 11 of the sound environment analysis system 1 acquires sound data (step ST1). The sound data acquisition unit 11 of the sound environment analysis system 1 acquires sound data from multiple sound sensors SE1 of multiple sound sensor array SA1 and stores the sound data in the storage unit 12.

[0031] Next, the separation unit 13 of the sound environment analysis system 1 performs sound source separation (step ST2). The separation unit 13 estimates the direction of the sound source for each of the multiple sound sensor arrays SA1, for example. Then, the separation unit 13 estimates the sound from the same sound source for two or more of the multiple sound sensor arrays SA1, for example, and generates the position of each of the multiple sound sources and partial sound data which is the sound from each of the multiple sound sources.

[0032] Next, the identification unit 14 of the sound environment analysis system 1 identifies the type of each of the one or more partial sound data and assigns a label to it (step ST3). For each of the one or more partial sound data, the identification unit 14 estimates the type of sound based on its characteristics and assigns a label indicating the type of sound. The identification unit 14 assigns a label to each of the one or more partial sound data, for example, by using a trained model. Alternatively, the identification unit 14 may assign a label by analyzing the characteristics of each of the one or more partial sound data.

[0033] Next, the unpleasant sound identification unit 15 of the sound environment analysis system 1 calculates the sound pressure for each label (step ST4). The unpleasant sound identification unit 15 calculates the sound pressure for each of one or more partial sound data, and calculates a representative sound pressure value for each label.

[0034] Next, the unpleasant sound identification unit 15 of the sound environment analysis system 1 determines whether the sound pressure is above a threshold (step ST51). The unpleasant sound identification unit 15 selects one of the one or more labels assigned in step ST3 and determines whether the sound pressure of the selected label is above a first threshold. If the sound pressure of the selected label is above the first threshold (Yes in step ST51), the unpleasant sound identification unit 15 determines that the sound of that label is an unpleasant sound (step ST52).

[0035] On the other hand, if the sound pressure of the selected label is not equal to or greater than the first threshold (No in step ST51), the unpleasant sound identification unit 15 calculates the sound pressure ratio (step ST53). The unpleasant sound identification unit 15 calculates the sound pressure ratio by dividing the sound pressure of the selected label by the sound pressure of all other labels. The sound pressure of all other labels is, for example, the maximum value among the sound pressures of the other labels. For example, if "footsteps" is selected from multiple labels such as "key typing sound," "footsteps," "printer operation sound," "male voice," and "female voice," the sound pressure of all other labels is the maximum value among the sound pressure of "key typing sound," "printer operation sound," "male voice," and "female voice." The unpleasant sound identification unit 15 determines whether the sound pressure ratio of the selected label is equal to or greater than the threshold (step ST54).

[0036] The unpleasant sound identification unit 15 determines that the sound of the selected label is unpleasant if the sound pressure ratio of that label is equal to or greater than the second threshold (Yes in step ST54) (step ST52). On the other hand, the unpleasant sound identification unit 15 determines that the sound of the selected label is not unpleasant if the sound pressure ratio of that label is not equal to or greater than the second threshold (No in step ST54) (step ST55).

[0037] If the sound environment analysis system 1 determines that a selected label is an unpleasant sound (step ST52) or not an unpleasant sound (step ST55), it determines whether there are any unprocessed labels (step ST6). If, for the labels that the identification unit 14 assigned to each of the one or more partial sound data in step ST3, there are labels that have not been determined to be either unpleasant or not unpleasant (Yes in step ST6), the sound environment analysis system 1 repeats steps ST51 to ST55. If, for all the labels that the identification unit 14 assigned to each of the one or more partial sound data in step ST3, the sound environment analysis system 1 is determined to be either unpleasant or not unpleasant (No in step ST6), it proceeds to step ST7.

[0038] The sound environment analysis system 1 waits until a predetermined interval has elapsed, and once the predetermined interval has elapsed, it returns to step ST1 to acquire and analyze the next sound data (step ST7). The predetermined interval is, for example, 5 seconds.

[0039] (4) Effects The sound environment analysis system 1 according to Embodiment 1 comprises a sound data acquisition unit 11, a separation unit 13, an identification unit 14, and an unpleasant sound identification unit 15. The sound data acquisition unit 11 acquires sound data within space SP1. The separation unit 13 separates the sound data into one or more partial sound data for each sound source within space SP1. The identification unit 14 identifies the type of each of the one or more partial sound data and assigns a label to it. The unpleasant sound identification unit 15 identifies which of the sound data labels is an unpleasant sound. The unpleasant sound identification unit 15 identifies a labeled sound as an unpleasant sound if the sound pressure of the partial sound data to which a label has been assigned is equal to or greater than a first threshold. With the above configuration, the sound environment analysis system 1 according to Embodiment 1 makes it possible to analyze the sound environment of space SP1, where various sounds may occur, based on objective criteria regardless of the user. Specifically, the sound environment analysis system 1 makes it possible to analyze the sound environment based on sound pressure for each identified sound by identifying the type of sound.

[0040] Furthermore, in the sound environment analysis system 1 according to Embodiment 1, the unpleasant sound identification unit 15 further identifies a labeled sound as unpleasant if the ratio of the sound pressure of one or more partially sound data to a partially sound data to an unlabeled partially sound data is equal to or greater than a second threshold. In this manner, the sound environment analysis system 1 according to Embodiment 1 makes it possible to identify whether or not a partially sound data to which a specific label has been assigned is unpleasant, while also taking into account whether or not it is masked by partially sound data to which other labels have been assigned.

[0041] Furthermore, in the sound environment analysis system 1 according to Embodiment 1, the unpleasant sound identification unit 15 may also acquire information indicating whether or not a sound is unpleasant for labels that have been assigned to one or more partial data points but have not been identified as unpleasant. In this embodiment, the sound environment analysis system 1 according to Embodiment 1 can identify whether or not a sound is unpleasant even for sounds that the unpleasant sound identification unit 15 has not been able to identify as unpleasant.

[0042] Furthermore, the sound environment analysis method according to Embodiment 1 is executed by one or more processors. The sound environment analysis method includes a sound data acquisition step ST1, a separation step ST2, an identification step ST3, and unpleasant sound identification steps ST4 to ST55. In the sound data acquisition step ST1, sound data within space SP1 is acquired. In the separation step ST2, the sound data is separated into one or more partial sound data for each sound source within space SP1. In the identification step ST3, each of the one or more partial sound data is identified and a label is assigned to it. In the unpleasant sound identification steps ST4 to ST55, it is identified that the sound of the label is an unpleasant sound if the sound pressure of the partial sound data to which a label has been assigned is equal to or greater than a first threshold. Furthermore, the program according to Embodiment 1 causes one or more processors to execute the sound environment analysis method according to Embodiment 1. In the above embodiment, the sound environment analysis method and program according to Embodiment 1 make it possible to analyze the sound environment of a space SP1 where various sounds may occur, based on objective criteria regardless of the user. Specifically, the sound environment analysis method and program make it possible to analyze the sound environment based on sound pressure for each identified sound by identifying the type of sound.

[0043] (Embodiment 2) (1) Composition In the sound environment analysis system 1 according to Embodiment 2, one additional condition is added to the sound environment analysis system 1 according to Embodiment 1, which causes the unpleasant sound identification unit 15 to identify the sound of the label as an unpleasant sound.

[0044] The unpleasant sound identification unit 15 identifies a sound with a specific label as an unpleasant sound if the partial sound data to which a specific label has been assigned contains a component whose frequency is outside a predetermined range and whose intensity is equal to or greater than an intensity threshold.

[0045] The unpleasant sound identification unit 15 obtains a frequency spectrum by, for example, performing a Fast Fourier Transform (FFT) on the partial sound data. The unpleasant sound identification unit 15 identifies the sound with the label assigned to the partial sound data as an unpleasant sound if there is a peak in the frequency spectrum of the partial sound data at a frequency that is not included in a predetermined range and whose intensity is above a threshold.

[0046] The predetermined frequency range is, for example, a range of frequencies in the sound to which the label is assigned that are less likely to cause discomfort to a person. Alternatively, the predetermined frequency range is, for example, a range of frequencies typically included in the sound to which the label is assigned. For example, if the label is "male voice" or "female voice," the predetermined frequency range includes frequencies included in normal speech but does not include frequencies specific to shouting, yelling, etc. Similarly, if the label is "printer operation sound," the predetermined frequency range includes frequencies included in normal printer operation but does not include frequencies of abnormal noises caused by malfunctions. The intensity threshold is the lower limit of sound intensity at which discomfort is induced by a sound whose frequency falls outside the predetermined range.

[0047] In other words, the unpleasant sound identification unit 15 identifies a sound as unpleasant if it contains sounds of a specific type that do not normally include, or if it contains sounds of a frequency that is particularly likely to cause discomfort.

[0048] (2) Operation Figure 4 is a flowchart showing the operation of the sound environment analysis system 1 according to Embodiment 2. Note that the same steps as those used in the sound environment analysis system 1 according to Embodiment 1 are given the same step numbers, and detailed explanations are omitted.

[0049] The sound data acquisition unit 11 of the sound environment analysis system 1 acquires sound data (step ST1).

[0050] Next, the separation unit 13 of the sound environment analysis system 1 performs sound source separation (step ST2).

[0051] Next, the identification unit 14 of the sound environment analysis system 1 identifies and labels each of the one or more partial sound data (step ST3).

[0052] Next, the unpleasant sound identification unit 15 of the sound environment analysis system 1 calculates the sound pressure for each label (step ST4).

[0053] Next, the unpleasant sound identification unit 15 of the sound environment analysis system 1 determines whether the sound pressure is above a threshold (step ST51). The unpleasant sound identification unit 15 selects one of the one or more labels assigned in step ST3 and determines whether the sound pressure of the selected label is above a first threshold. If the sound pressure of the selected label is above the first threshold (Yes in step ST51), the unpleasant sound identification unit 15 determines that the sound of that label is an unpleasant sound (step ST52).

[0054] On the other hand, if the sound pressure of the selected label is not equal to or greater than the first threshold (No in step ST51), the unpleasant sound identification unit 15 calculates the sound pressure ratio (step ST53). The unpleasant sound identification unit 15 divides the sound pressure of the selected label by the sound pressure of all other labels to calculate the sound pressure ratio. The unpleasant sound identification unit 15 determines whether the sound pressure ratio of the selected label is equal to or greater than the threshold (step ST54).

[0055] The unpleasant sound identification unit 15 determines that the sound of the selected label is unpleasant if the sound pressure ratio of the selected label is greater than or equal to the second threshold (Yes in step ST54) (step ST52). On the other hand, if the sound pressure ratio of the selected label is not greater than or equal to the second threshold (No in step ST54), the unpleasant sound identification unit 15 determines whether the frequency of the partial sound data of the selected label is within a predetermined range (step ST56).

[0056] The unpleasant sound identification unit 15 determines that the sound of the selected label is unpleasant if the frequency of the partial sound data of the selected label is outside a predetermined range and the intensity is equal to or greater than the intensity threshold (No in step ST56) (step ST52). On the other hand, the unpleasant sound identification unit 15 determines that the sound of the selected label is not unpleasant if the frequency of the partial sound data of the selected label is within a predetermined range, or if the intensity of the component whose frequency is outside the predetermined range is less than the intensity threshold (Yes in step ST56) (step ST55).

[0057] If the sound environment analysis system 1 determines that a selected label is an unpleasant sound (step ST52) or not an unpleasant sound (step ST55), it determines whether there are any unprocessed labels (step ST6). If, for the labels that the identification unit 14 assigned to each of the one or more partial sound data in step ST3, there are labels that have not been determined to be either unpleasant or not unpleasant (Yes in step ST6), the sound environment analysis system 1 repeats steps ST51 to ST55. If, for all the labels that the identification unit 14 assigned to each of the one or more partial sound data in step ST3, the sound environment analysis system 1 is determined to be either unpleasant or not unpleasant (No in step ST6), it proceeds to step ST7.

[0058] The sound environment analysis system 1 waits until a predetermined interval has elapsed, and once the predetermined interval has elapsed, it returns to step ST1 to acquire and analyze the next sound data (step ST7). The predetermined interval is, for example, 5 seconds. Here, the predetermined interval in step ST7 is not limited to 5 seconds and may be any time. Alternatively, the sound environment analysis system 1 may skip step ST7 and return to step ST1 once step ST6 is completed. In this case, the sound environment analysis system 1 repeats step ST1 in cycles of a time length necessary for data separation and identification. For example, if the time required from step ST1 to step ST7 is less than 5 seconds, the sound environment analysis system 1 may proceed to the next step ST1. It goes without saying that if the time required from step ST1 to step ST7 is longer than 5 seconds, the sound environment analysis system 1 may extend its return to step ST1 until that time is completed.

[0059] (3) Effects In the sound environment analysis system 1 according to Embodiment 2, the unpleasant sound identification unit 15 further identifies a labeled sound as unpleasant if, among one or more partial sound data, a partial sound data to which a label has been assigned contains a component whose frequency is outside a predetermined range and whose intensity is equal to or greater than an intensity threshold. In this manner, the sound environment analysis system 1 according to Embodiment 2 can identify a sound as unpleasant if, with respect to a particular label, it contains a sound with a frequency that is not normally included, or a sound with a frequency that is particularly likely to cause discomfort.

[0060] (Embodiment 3) (1) Composition In the sound environment analysis system 1 according to Embodiment 3, one additional condition is added for the unpleasant sound identification unit 15 to identify the sound of the label as an unpleasant sound, compared to the sound environment analysis system 1 according to Embodiment 2.

[0061] The unpleasant sound identification unit 15 identifies a labeled sound as unpleasant if the labeled partial sound data does not have periodicity.

[0062] The unpleasant sound identification unit 15, for example, performs a Fast Fourier Transform on the partial sound data to obtain a frequency spectrum. The unpleasant sound identification unit 15 performs the Fast Fourier Transform multiple times while moving the time window targeted for the Fast Fourier Transform, and determines whether the peak frequency changes periodically. If the peak frequency changes periodically, such as with 1 / f fluctuation, the unpleasant sound identification unit 15 identifies the sound with the label assigned to that partial sound data as not being an unpleasant sound.

[0063] (2) Operation Figure 5 is a flowchart showing the operation of the sound environment analysis system 1 according to Embodiment 3. Note that the same steps as those used for the sound environment analysis system 1 according to Embodiment 2 are given the same step numbers, and detailed explanations are omitted.

[0064] The sound data acquisition unit 11 of the sound environment analysis system 1 acquires sound data (step ST1).

[0065] Next, the separation unit 13 of the sound environment analysis system 1 performs sound source separation (step ST2).

[0066] Next, the identification unit 14 of the sound environment analysis system 1 identifies and labels each of the one or more partial sound data (step ST3).

[0067] Next, the unpleasant sound identification unit 15 of the sound environment analysis system 1 calculates the sound pressure for each label (step ST4).

[0068] Next, the unpleasant sound identification unit 15 of the sound environment analysis system 1 determines whether the sound pressure is above a threshold (step ST51). The unpleasant sound identification unit 15 selects one of the one or more labels assigned in step ST3 and determines whether the sound pressure of the selected label is above a first threshold. If the sound pressure of the selected label is above the first threshold (Yes in step ST51), the unpleasant sound identification unit 15 determines that the sound of that label is an unpleasant sound (step ST52).

[0069] On the other hand, if the sound pressure of the selected label is not equal to or greater than the first threshold (No in step ST51), the unpleasant sound identification unit 15 calculates the sound pressure ratio (step ST53). The unpleasant sound identification unit 15 divides the sound pressure of the selected label by the sound pressure of all other labels to calculate the sound pressure ratio. The unpleasant sound identification unit 15 determines whether the sound pressure ratio of the selected label is equal to or greater than the threshold (step ST54).

[0070] The unpleasant sound identification unit 15 determines that the sound of the selected label is unpleasant if the sound pressure ratio of the selected label is greater than or equal to the second threshold (Yes in step ST54) (step ST52). On the other hand, if the sound pressure ratio of the selected label is not greater than or equal to the second threshold (No in step ST54), the unpleasant sound identification unit 15 determines whether the frequency of the partial sound data of the selected label is within a predetermined range (step ST56).

[0071] The unpleasant sound identification unit 15 determines that the sound of the selected label is unpleasant if the frequency of the partial sound data of the selected label is outside a predetermined range and the intensity is greater than or equal to the intensity threshold (No in step ST56) (step ST52). On the other hand, if the frequency of the partial sound data of the selected label is within a predetermined range, or if the intensity of the component whose frequency is outside the predetermined range is less than the intensity threshold (Yes in step ST56), the unpleasant sound identification unit 15 determines whether or not the partial sound data to which the label is assigned is periodic (step ST57).

[0072] The unpleasant sound identification unit 15 determines that the sound of the selected label is unpleasant if the partial tone data of that label does not have periodicity (Yes in step ST57) (step ST52). On the other hand, the unpleasant sound identification unit 15 determines that the sound of the selected label is not unpleasant if the partial tone data of that label has periodicity (No in step ST57) (step ST55).

[0073] If the sound environment analysis system 1 determines that a selected label is an unpleasant sound (step ST52) or not an unpleasant sound (step ST55), it determines whether there are any unprocessed labels (step ST6). If, for the labels that the identification unit 14 assigned to each of the one or more partial sound data in step ST3, there are labels that have not been determined to be either unpleasant or not unpleasant (Yes in step ST6), the sound environment analysis system 1 repeats steps ST51 to ST55. If, for all the labels that the identification unit 14 assigned to each of the one or more partial sound data in step ST3, the sound environment analysis system 1 is determined to be either unpleasant or not unpleasant (No in step ST6), it proceeds to step ST7.

[0074] The sound environment analysis system 1 waits until a predetermined interval has elapsed, and once the predetermined interval has elapsed, it returns to step ST1 to acquire and analyze the next sound data (step ST7). The predetermined interval is, for example, 5 seconds.

[0075] (3) Effects The unpleasant sound identification unit 15 according to Embodiment 3 further identifies a labeled sound as unpleasant if, among one or more partial sound data, the partial sound data to which a label has been assigned does not have periodicity. With this configuration, the sound environment analysis system 1 according to Embodiment 3 can avoid identifying a sound with a specific label as unpleasant if it is a periodic sound such as 1 / f fluctuation.

[0076] (Other modifications according to the embodiment) (1) In the sound environment analysis system 1 of Embodiments 1 to 3, both step ST52 and steps ST53 and ST54 are performed as the operation to determine whether or not a sound is unpleasant. However, in the sound environment analysis system 1, it is also possible to determine whether or not a sound is unpleasant using only either sound pressure or sound pressure ratio.

[0077] Similarly, in the sound environment analysis system 1 of Embodiment 3, it is not necessary to determine whether the frequency is within a predetermined range.

[0078] (2) The sound environment analysis system 1 of Embodiments 1 to 3 includes a plurality of sound sensor arrays SA1, but the sound environment analysis system 1 may acquire sound data from other devices or systems.

[0079] (Appearance) The sound environment analysis system (1) according to the first embodiment comprises a sound data acquisition unit (11), a separation unit (13), an identification unit (14), and an unpleasant sound identification unit (15). The sound data acquisition unit (11) acquires sound data within a space (SP1). The separation unit (13) separates the sound data into one or more partial sound data for each sound source within the space (SP1). The identification unit (14) identifies the type of each of the one or more partial sound data and assigns a label to it. The unpleasant sound identification unit (15) identifies which label of the sound data is an unpleasant sound. The unpleasant sound identification unit (15) identifies the sound of the label as an unpleasant sound if the sound pressure of the partial sound data to which a label has been assigned is equal to or greater than a first threshold.

[0080] According to the sound environment analysis system (1) described above, it becomes possible to analyze the sound environment of a space (SP1) where various sounds may occur, based on objective criteria regardless of the user.

[0081] In the sound environment analysis system (1) according to the second embodiment, in the first embodiment, the unpleasant sound identification unit (15) further identifies a labeled sound as unpleasant if the ratio of the sound pressure of a labeled partial sound data to the sound pressure of an unlabeled partial sound data is equal to or greater than a second threshold.

[0082] According to the sound environment analysis system (1) described above, it is possible to determine whether or not a partial sound data to which a specific label has been assigned is an unpleasant sound, while also taking into account whether or not it is masked by partial sound data to which other labels have been assigned.

[0083] In the sound environment analysis system (1) according to the third embodiment, in the first or second embodiment, the unpleasant sound identification unit further identifies a labeled sound as unpleasant if one or more partial sound data that has been labeled contains a component whose frequency is outside a predetermined range and whose intensity is equal to or greater than an intensity threshold.

[0084] According to the sound environment analysis system (1) described above, it is possible to identify a sound as unpleasant if it contains sounds of frequencies not normally present in a sound of a particular label, or if it contains sounds of frequencies that are particularly likely to cause discomfort.

[0085] In the sound environment analysis system (1) according to the fourth embodiment, in any of the first to third embodiments, the unpleasant sound identification unit (15) further identifies the sound of the label as an unpleasant sound if the labeled partial sound data among one or more partial sound data does not have periodicity.

[0086] According to the sound environment analysis system (1) described above, it is possible to avoid classifying sounds with a specific label as unpleasant if they are periodic sounds such as 1 / f fluctuations.

[0087] In the fifth embodiment of the sound environment analysis system (1), in any of the first to fourth embodiments, the unpleasant sound identification unit (15) further acquires information indicating whether or not a sound is unpleasant for labels that have been assigned to one or more partial data points but have not been identified as unpleasant sounds.

[0088] According to the sound environment analysis system (1) in the above embodiment, it becomes possible to determine whether or not a sound is unpleasant even for sounds that the unpleasant sound identification unit (15) cannot determine whether or not it is an unpleasant sound.

[0089] The sixth embodiment of the sound environment analysis method is performed by one or more processors. The sound environment analysis method includes a sound data acquisition step (ST1), a separation step (ST2), an identification step (ST3), and unpleasant sound identification steps (ST4 to ST57). In the sound data acquisition step (ST1), sound data in the space is acquired. In the separation step (ST2), the sound data is separated into one or more partial sound data for each sound source in the space (SP1). In the identification step (ST3), each of the one or more partial sound data is identified and a label is assigned to it. In the unpleasant sound identification steps (ST4 to ST57), the label is identified as an unpleasant sound if the sound pressure of the partial sound data to which a label has been assigned is equal to or greater than a first threshold.

[0090] According to the sound environment analysis method described above, it becomes possible to analyze the sound environment of a space (SP1) where various sounds may occur, based on objective criteria regardless of the user.

[0091] The program according to the seventh embodiment causes one or more processors to execute the sound environment analysis method according to the sixth embodiment.

[0092] According to the program described above, it becomes possible to analyze the sound environment of a space (SP1) where various sounds may occur, based on objective criteria, regardless of the user. [Explanation of symbols]

[0093] 1. Sound Environment Analysis System 11. Sound data acquisition unit 13 Separation part 14 Identification Unit 15 Unpleasant sound identification section SP1 Space ST1 Step (Audio Data Acquisition Step) ST2 Step (Separation Step) ST3 Step (Identification Step) ST4, ST51~ST57 Steps (Unpleasant Sound Identification Steps)

Claims

1. A sound data acquisition unit that acquires sound data within a space, A separation unit that separates the aforementioned sound data into one or more partial sound data for each sound source in the space, An identification unit that identifies and labels each of the one or more partial sound data mentioned above, An unpleasant sound identification unit that identifies which of the aforementioned sound data's labels is an unpleasant sound, Equipped with, The unpleasant sound identification unit identifies the sound of the label as an unpleasant sound when the sound pressure of the partial sound data to which the label is assigned is equal to or greater than a first threshold among the one or more partial sound data. Sound environment analysis system.

2. The unpleasant sound identification unit further identifies the sound with the label as an unpleasant sound if the ratio of the sound pressure of the partial sound data to which the label has been assigned to it, among the one or more partial sound data, to the sound pressure of the partial sound data to which the label has not been assigned is equal to or greater than a second threshold. The sound environment analysis system according to claim 1.

3. The unpleasant sound identification unit further identifies the sound of the label as unpleasant if, among the one or more partial sound data, the partial sound data to which the label is assigned contains a component whose frequency is outside a predetermined range and whose intensity is equal to or greater than an intensity threshold. The sound environment analysis system according to claim 1 or 2.

4. The unpleasant sound identification unit further identifies the sound of the label as an unpleasant sound if the partial sound data to which the label is assigned does not have periodicity among the one or more partial sound data. The sound environment analysis system according to claim 1 or 2.

5. The unpleasant sound identification unit further acquires information indicating whether or not an unpleasant sound is an unpleasant sound, for labels among the labels assigned to the one or more partial data that have not been identified as unpleasant sounds. The sound environment analysis system according to claim 1 or 2.

6. A sound environment analysis method performed by one or more processors, A sound data acquisition step to acquire sound data within a space, A separation step of separating the sound data into one or more partial sound data for each sound source in the space, An identification step of identifying each of the one or more partial sound data and assigning a label to them, An unpleasant sound identification step to identify which of the aforementioned sound data's labels is an unpleasant sound, Includes, In the unpleasant sound identification step, if the sound pressure of the partial sound data to which the label is assigned is equal to or greater than a first threshold among the one or more partial sound data, the sound of the label is identified as an unpleasant sound. Sound environment analysis method.

7. A program that causes one or more processors to execute the sound environment analysis method according to claim 6.

Images