Database management device and database management method

Abstract

This database management device includes: a feature amount extraction unit that extracts, from inputted data of noise, a set of acoustic feature amounts of the noise; and a database update unit that generates a node α (the size of the node α is 1) labeled with the extracted set of the acoustic feature amounts, and updates a noise database on the basis of the distance between the node α and a node αi (i = 1,..., K; K represents the number of nodes included in the noise database) included in the noise database.

Description

Database management device and database management method

[0001] The present invention relates to database technology, and more particularly to a database for managing noise data.

[0002] As a technique for separating observed acoustic signals into sound sources, there is a sound source separation technique as described in Non-Patent Document 1. In sound source separation techniques, noises that interfere with a target sound are often treated as a single sound. In other words, sound source separation techniques rarely classify noises into several types and handle noises while taking into account the characteristics of each type.

[0003] Shoji Makino (Editor), "Audio Source Separation," Springer, 2018.

[0004] However, if noises are classified into several types based on acoustic features, it is expected that it will be easier to formulate policies for creating noise control measures and noise cancellation technologies. It is also expected that it will be possible to determine whether or not established noise control measures and noise cancellation technologies are applicable to new noises.

[0005] Therefore, an object of the present invention is to provide a noise database technology for managing noise types classified based on acoustic features.

[0006] One aspect of the present invention is a database management device that manages a noise database, which is a database for managing noise data. The noise database is configured as a graph having nodes labeled with one or more sets of acoustic features and edges representing the proximity between the nodes. The device includes: a feature extraction unit that extracts a set of acoustic features of the noise from input noise data; and a database update unit that generates a node α (where the size of node α is 1) labeled with the extracted set of acoustic features and updates the noise database based on the distance between node α and node αi (i = 1, ..., K, K represents the number of nodes included in the noise database).

[0007] According to the present invention, it is possible to manage types of noise classified based on acoustic features.

[0008] It is a block diagram showing the configuration of the database management device 100. It is a flowchart showing the operation of the database management device 100. It is a diagram showing an example of the functional configuration of a computer that realizes each device in the embodiment of the present invention.

[0009] Hereinafter, an embodiment of the present invention will be described in detail. Note that components having the same functions are given the same numbers and redundant explanations will be omitted.

[0010] <Technical Background> In an embodiment of the present invention, a graph is used as the structure of a noise database. Hereinafter, this database will be referred to as the noise database. The noise database is configured as a graph having nodes labeled with one or more sets of acoustic features and edges representing the proximity between the nodes. Here, a set of acoustic features refers to multiple acoustic features. Furthermore, the number of sets of acoustic features labeled on a node will be referred to as the size of the node. Therefore, a node will have a size represented by a natural number greater than or equal to 1.

[0011] Consider the case where a new node α is added to the noise database. Here, node α is a node labeled with a set of acoustic features. In this case, an edge connecting node αi and node α included in the noise database is generated according to the following rules:

[0012] (Rule 1) For each acoustic feature included in the label assigned to a node, if the distance between node αi and node α calculated using that acoustic feature is smaller than or equal to a predetermined value, an edge connecting node αi and node α is generated. Here, the distance between node αi and node α calculated using a certain acoustic feature is, for example, the absolute value of the difference between the representative value of that acoustic feature included in the label of node αi (for example, the value of that acoustic feature included in the set of acoustic features that is recorded first in the noise database among the set of one or more acoustic features included in the label of node αi) and the value of that acoustic feature included in the label of node α.

[0013] Therefore, if N is the number of acoustic features included in the label assigned to the node, the number of edges connecting node αi and node α is 0 to N inclusive.

[0014] Furthermore, if the number of edges connecting node αi and node α is N, node α'i is generated by absorbing node α into node αi according to three rules: (Rule 2-1), (Rule 2-2), and (Rule 2-3).

[0015] (Rule 2-1) The size of node α'i is set to the size of node αi plus 1. By doing so, the size of node α'i represents the number of sets of acoustic features labeled with node α'i.

[0016] (Rule 2-2) The label of node α'i is the union of a set whose elements are one or more sets of acoustic features included in the label of node αi and a set whose element is one set of acoustic features included in the label of node α. In other words, the label of node αi is set to {S1, S2, ..., S M} (where S1(i=1, 2, …, M) is the set of acoustic features included in the label of node αi), and the label of node α is {S} (where S is the set of acoustic features included in the label of node α), then the label of node α'i is {S1, S2, …, S M , S}.

[0017] (Rule 2-3) The edge connecting node α'i and a node other than node αi contained in the noise database is the same as the edge connecting node αi and a node other than node αi contained in the noise database. In other words, the edge connecting node α'i and a node other than node αi contained in the noise database is a continuation of the edge connecting node αi and a node other than node αi contained in the noise database.

[0018] Therefore, the noise database has the following characteristics:

[0019] (1) Since larger nodes correspond to more noise data, it is considered that larger nodes are more likely to be applicable to existing noise control and noise cancellation technologies.

[0020] (2) Although a graph may contain nodes that are not connected to other nodes (i.e., isolated nodes), it is expected that the number of isolated nodes will decrease as the number of nodes increases.

[0021] (3) It is possible to find the number of dominant nodes in the graph and the nodes that have edges between many nodes.

[0022] First Embodiment A database management device 100 manages a database for managing noise data (hereinafter referred to as a noise database). The noise database is configured as a graph having nodes labeled with one or more sets of acoustic features and edges representing the proximity between the nodes.

[0023] The database management device 100 will be described below with reference to FIGS. 1 and 2. FIG. 1 is a block diagram showing the configuration of the database management device 100. FIG. 2 is a flowchart showing the operation of the database management device 100. As shown in FIG. 1, the database management device 100 includes a feature extraction unit 110, a database update unit 120, and a recording unit 190. The recording unit 190 is a component that appropriately records information necessary for the processing of the database management device 100. The recording unit 190 records, for example, a noise database.

[0024] The operation of the database management device 100 will be described with reference to FIG.

[0025] In S110, the feature extraction unit 110 extracts and outputs a set of acoustic features of the noise from the input noise data. The feature extraction unit 110 may extract any acoustic feature as the acoustic feature of the noise data, such as the mean, variance, time rate of change of the mean, time rate of change of the variance, kurtosis of the data distribution, skewness of the data distribution, amplitude envelope, 1 / N octave bundle level, formant frequency, fundamental frequency, and eigenfrequency.

[0026] In S120, the database update unit 120 receives as input the set of acoustic features output in S110, generates a node α (where the size of node α is 1) labeled with the set of acoustic features, and for each node αi (i = 1, ..., K, where K represents the number of nodes included in the noise database) included in the noise database, generates an edge connecting node αi and node α if the distance between node αi and node α, calculated for each acoustic feature included in the label assigned to the node, using the acoustic feature is less than or equal to a predetermined value. If the number of edges connecting node αi and node α matches the number of acoustic features included in the label assigned to the node, the database update unit 120 generates a single node (where the label of the node is the union of a set whose elements are one or more sets of acoustic features included in the label of node αi and a set whose element is one set of acoustic features included in the label of node α, and the size is the value obtained by adding 1 to the size of node αi) to update the noise database. In other words, the database update unit 120 generates a node α (where the size of node α is 1) labeled with the set of acoustic features extracted in S110, and updates the noise database based on the distance between node α and node αi (i = 1, ..., K, K represents the number of nodes included in the noise database) included in the noise database.

[0027] The database management device 100 may manage the noise data after dividing the noise data into several frequency bands. For example, the database management device 100 can use noise data for frequency band 1 and noise data for frequency band 2 to configure and manage a first noise database for managing the noise data for frequency band 1 and a second noise database for managing the noise data for frequency band 2. Managing the first noise database and the second noise database makes it possible to efficiently cross-reference two graph nodes corresponding to one piece of noise data. Cross-referencing between the first noise database and the second noise database makes it possible to interpolate unmeasured data (i.e., data for which no true data exists) in data that has only been roughly measured from data that has been measured in detail.

[0028] According to the embodiment of the present invention, it is possible to manage the types of noise classified based on acoustic features.

[0029] <Additional Notes> The functions performed by the components described herein may be implemented in circuitry or processing circuitry, including general-purpose processors, application-specific processors, integrated circuits, ASICs (Application Specific Integrated Circuits), CPUs (Central Processing Units), conventional circuits, and / or combinations thereof, programmed to perform the described functions. A processor includes transistors and other circuits and is considered to be circuitry or processing circuitry. A processor may also be a programmed processor that executes a program stored in a memory.

[0030] In this specification, a circuitry, unit, or means is hardware that is programmed to realize or performs the described functions, which may be any hardware disclosed herein or any hardware known to be programmed to realize or perform the described functions.

[0031] If the hardware is a processor considered to be a type of circuitry, the circuitry, means, or unit is a combination of the hardware and software used to configure the hardware and / or processor.

[0032] The various processes described above can be implemented by loading a program that executes each step of the above method into the recording unit 2020 of the computer 2000 shown in Figure 3, and operating the control unit 2010, input unit 2030, output unit 2040, display unit 2050, etc.

[0033] The program describing the processing contents can be recorded on a computer-readable recording medium, which may be, for example, a magnetic recording device, an optical disk, a magneto-optical recording medium, a semiconductor memory, or any other suitable recording medium.

[0034] The program may be distributed by, for example, selling, transferring, lending, etc. portable recording media such as DVDs and CD-ROMs on which the program is recorded. Furthermore, the program may be stored in a storage device of a server computer, and then transferred from the server computer to other computers via a network, thereby distributing the program.

[0035] A computer that executes such a program may first temporarily store the program recorded on a portable recording medium or transferred from a server computer in its own storage device. Then, when executing a process, the computer reads the program stored in its storage device and executes the process in accordance with the read program. Alternatively, the computer may read the program directly from a portable recording medium and execute the process in accordance with the program. Furthermore, the computer may execute the process in accordance with the program each time a program is transferred from a server computer to the computer. Alternatively, the server computer may not transfer the program to the computer, but may instead execute the process through a so-called ASP (Application Service Provider) service, which realizes the processing function by issuing an execution instruction and obtaining the results. Furthermore, the server computer may execute the process at the terminal using a so-called SaaS (Software as a Service) service, which allows users to use part of the server computer along with the program. In this embodiment, the program includes information used for processing by an electronic computer that is equivalent to a program (such as data that is not a direct instruction to a computer but has properties that dictate computer processing).

[0036] Furthermore, in this embodiment, the device is configured by executing a predetermined program on a computer, but at least a part of the processing contents may be realized by hardware.

[0037] The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention. Furthermore, the processes described in the above embodiments may not only be executed in chronological order according to the order described, but may also be executed in parallel or individually depending on the processing capacity of the device that executes the processes or as needed.

Claims

1. A database management device that manages a noise database, which is a database for managing noise data, wherein the noise database is configured as a graph having nodes labeled with one or more sets of acoustic features and edges representing the proximity between the nodes, the database management device comprising: a feature extraction unit that extracts a set of acoustic features of the noise from input noise data; and a database update unit that generates a node α (where the size of node α is 1) labeled with the extracted set of acoustic features and updates the noise database based on the distance between node α and node αi (i=1, ..., K, K represents the number of nodes included in the noise database).

2. A database management device according to claim 1, wherein the database update unit generates a node α (where the size of node α is 1) labeled with the extracted set of acoustic features, and generates an edge connecting node αi and node α if the distance between node αi and node α calculated using each acoustic feature included in the label assigned to node αi (i = 1, ..., K, K represents the number of nodes included in the noise database) included in the noise database and node α is smaller than or equal to a predetermined value, and generates a single node (where the label of the node is the union of a set whose elements are one or more sets of acoustic features included in the label of node αi and a set whose element is one set of acoustic features included in the label of node α, and the size is the size of node αi plus 1) to update the noise database.

3. A database management method in which a database management device manages a noise database, which is a database for managing noise data, wherein the noise database is configured as a graph having nodes labeled with one or more sets of acoustic features and edges representing the proximity between the nodes, the database management method comprising: a feature extraction step in which the database management device extracts a set of acoustic features of the noise from input noise data; and a database update step in which the database management device generates a node α (where the size of node α is 1) labeled with the extracted set of acoustic features, and updates the noise database based on the distance between node α and node αi (i = 1, ..., K, K represents the number of nodes included in the noise database).

Images