JP6045511B2 - Acoustic signal detection system, acoustic signal detection method, acoustic signal detection server, acoustic signal detection apparatus, and acoustic signal detection program - Google Patents

Description

  The present invention relates to an acoustic signal detection system, an acoustic signal detection method, an acoustic signal detection server, an acoustic signal detection device, and an acoustic signal detection program for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated.

  In recent years, a technique for detecting an acoustic signal having a harmonic structure such as a specific person's voice or musical instrument sound from an acoustic signal in which a plurality of types of acoustic signals are mixed is known (for example, Patent Document 1). In the speech recognition apparatus disclosed in Patent Document 1, a source signal is separated from a source signal (sound, image, radio wave, etc.) in a real environment in which background noise exists and is convoluted and mixed by blind source signal separation processing. The target acoustic signal is detected (for example, Patent Document 1).

JP 2003-78423 A

  However, when performing sound source separation as disclosed in Patent Document 1, in order to separate a noise-mixed recording into a signal and noise, complicated arithmetic processing such as blind source signal separation processing is necessary. In source signal separation processing, information such as the positional relationship between the sound source and the microphone is unknown, and it is necessary to take into account audio delays, etc., so the frequency characteristics can be adjusted by filtering as preprocessing or inverse filtering as postprocessing. Perform processing such as restoration. Therefore, in the conventional sound source separation, the computation processing becomes complicated, the processing load increases, it takes time to detect the target specific acoustic signal, and the load due to real-time processing becomes excessive, so mobile There is a problem that it is not suitable for portable devices such as terminals.

  Therefore, the present invention solves the above-described problems, and can reduce the processing load even under noise and detect a target specific sound signal in real time, and trigger the specific sound signal. It is an object to provide an acoustic signal detection system, an acoustic signal detection method, an acoustic signal detection server, an acoustic signal detection device, and an acoustic signal detection program that can provide a variety of services and provide a variety of services. And

In order to solve the above-described problems, the present invention provides an acoustic signal detection system that detects a specific acoustic signal included in an evaluation acoustic signal to be evaluated, acquires the evaluation acoustic signal, and acquires the acquired evaluation acoustic signal. For a signal, the presence / absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the comparison destination template in which the distribution of the acoustic signal and its temporal change are binarized for each scale and recorded in time series is obtained. Based on the evaluation target conversion unit to be generated and the specific acoustic signal , the presence or absence of the acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are binary for each scale. Sound source template generation unit for generating sound source templates recorded in time series, sound source template acquisition unit for acquiring sound source templates, comparison destination template and sound source A template comparison unit that compares the template and calculates a matching rate of the binarized acoustic signal distribution; and a matching rate output unit that outputs a comparison result signal according to the matching rate calculated by the template comparison unit. Based on an event database that stores a template identifier that identifies a sound source template generated by the sound source template generation unit and an event identifier that identifies an event to be generated, and a comparison result signal output from the matching rate output unit, An event generation management unit that compares an event database with respect to a template identifier of a sound source template related to the comparison result signal, selects an event to be generated, and generates the selected event; When generating a sound source template for each scale Detecting a definitive signal strength, extracts the presence or absence of the acoustic signal detected sound signal based on whether exceeds a predetermined threshold value, said threshold value, depending on the type of event to be the generation It is characterized by being set to a different value .

Another invention is an acoustic signal detection method for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated,
(1) Acquire an evaluation acoustic signal, extract the presence or absence of an acoustic signal in a predetermined frequency band based on the scale average rate for the acquired evaluation acoustic signal, and determine the distribution and temporal change of the acoustic signal for each scale. Processing to convert the evaluation target to generate a comparison target template binarized and recorded in time series,
(2) For a specific acoustic signal, the presence or absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are binarized for each scale and recorded in time series. A template comparison process that generates and obtains a sound source template, and compares the comparison target template with the sound source template to calculate a matching rate of the binarized acoustic signal distribution;
(3) relevance ratio output processing for outputting a comparison result signal corresponding to the relevance ratio calculated by the template comparison processing ;
(4) a database control process for storing a template identifier for identifying a sound source template generated in the template comparison process and an event identifier for identifying an event to be generated in association with each other in an event database;
(5) Based on the comparison result signal output from the matching rate output process, the event database is checked for the template identifier of the sound source template related to the comparison result signal, the event to be generated is selected, and the selected event is generated. Event occurrence management processing and <br/>
In the template comparison process, when generating a sound source template for a specific sound signal, the signal intensity in each scale is detected, and the presence or absence of the sound signal is determined based on whether or not the detected sound signal exceeds a predetermined threshold value. Extract and
The threshold value is set to a different value depending on the type of event to be generated .

According to the present invention as described above, the specific acoustic signal is obtained by comparing the templates generated by extracting the presence / absence of the signal in the frequency band based on the scale equal temperament, so that complicated calculation processing is not required. Matching processing with a sound source can be performed, the load due to real-time processing can be reduced by high-speed computation by downsizing data, and mounting on a portable device is possible. Further, since the presence / absence of an acoustic signal is detected by a binarized template, detection under noise is possible. As a result, it is possible to provide a service triggered by specific sound detection, such as distributing coupons or points when a theme song broadcast in a store is detected. Diversification and enhancement can be achieved.
When generating the sound source template, it is preferable to detect the presence or absence of a sound of each scale after removing the harmonic component.
Further, at the time of comparison with the sound source template, the presence / absence of each tone may be detected without removing the harmonic component and including the harmonic component.

  Here, “acquisition of evaluation sound signal” includes not only acquiring sound directly as an electrical signal using an acoustic device such as a microphone, but also acquiring it as a sound signal or sound file by wireless / wired communication. It is. On the other hand, the “sound source template” is generated using any device of the system. For example, the user may generate the sound source template on the user terminal, or download the generated sound source template from the server. Also good. In addition, the “scale average rate” includes various scales in addition to the 12 scales composed of semitones. Furthermore, “binarization of the presence / absence of an acoustic signal” means detecting only whether there is a sound of each scale, for example, determining whether the acoustic signal exceeds a predetermined threshold value. The technique to do is mentioned.

  Here, an event is a process that occurs on condition that a certain requirement is satisfied, such as exception processing in a program or branch processing, and the system and service provided by the process generated by this event are also used here. It is included in the “event”. For example, when a specific sound source is detected, various information processing services such as message output by voice or text display, image / video playback, application activation, and the like are also included in the event.

  In this case, an event triggered by the detection of the specific sound in the information processing terminal or server can be generated by associating the sound source template used for detection of the specific sound with an arbitrary event. In addition, it is possible to collect information on which user has detected what kind of sound and where, to investigate realistic user trends, and to improve marketing by building big data. .

  In this case, by generating a sound source template by changing the threshold value, different events can be generated depending on the type of the sound source template even when the same specific sound is detected. As a result, for example, even if a family enters the same store at the same time and detects the same specific sound, it is possible to provide services suitable for user identification such as age and gender, such as generating different events for adults and children. .

  Furthermore, another invention is an acoustic signal detection server and an acoustic signal detection device for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated, which can be used in the acoustic signal detection system and the acoustic signal detection method. is there.

Specifically, the acoustic signal detection server of the present invention is an acoustic signal detection server that detects a specific acoustic signal included in an evaluation acoustic signal to be evaluated, acquires the evaluation acoustic signal through a communication network, and About the acquired evaluation sound signal, the presence or absence of the sound signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the sound signal and its temporal change are binarized for each scale and recorded in time series. The comparison target template for generating the comparison target template and the presence or absence of the acoustic signal in a predetermined frequency band for the specific acoustic signal are extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are A sound source template acquisition unit that generates and acquires sound source templates that are binarized and recorded in time series, a comparison destination template, and a sound source template A template comparison unit that calculates the matching rate of the binarized distribution of the acoustic signal, and a matching rate output unit that sends a comparison result signal corresponding to the matching rate calculated by the template comparison unit to the communication network An event database that associates and stores a template identifier that identifies a sound source template generated by the sound source template generation unit and an event identifier that identifies an event to be generated, and a comparison result signal output from the matching rate output unit And an event occurrence management unit that collates the event database for the template identifier of the sound source template related to the comparison result signal, selects an event to be generated, and generates the selected event ,
When generating the sound source template for the specific sound signal, the sound source template generation unit detects a signal intensity at each scale and determines whether the detected sound signal exceeds a predetermined threshold value. The threshold value is set to a different value depending on the type of event to be generated .

According to the present invention, the specific sound signal is obtained by comparing the templates generated by extracting the presence / absence of the signal in the frequency band based on the scale equal temperament. And real-time processing is possible by high-speed calculation by downsizing the data. Further, since the presence / absence of an acoustic signal is detected by a binarized template, detection under noise is possible. As a result, it is possible to provide a service triggered by specific sound detection, such as distributing coupons or points when a theme song broadcast in a store is detected. Diversification and enhancement can be achieved. In particular, in the present invention, the server arranged on the communication network generates a comparison destination template from the evaluation acoustic signal and calculates the matching rate of the template, thereby reducing the processing burden on the terminal owned by the user. In addition, the memory capacity can be effectively used.
When generating the sound source template, it is preferable to detect the presence or absence of a sound of each scale after removing the harmonic component.
Further, at the time of comparison with the sound source template, the presence / absence of each tone may be detected without removing the harmonic component and including the harmonic component.

  In this case, an event triggered by the detection of the specific sound in the server can be generated by associating the sound source template used for the detection of the specific sound with an arbitrary event. In addition, it is possible to collect information on which user has detected what kind of sound and where, to investigate realistic user trends, and to improve marketing by building big data. .

  In this case, by generating a sound source template by changing the threshold value, different events can be generated depending on the type of the sound source template even when the same specific sound is detected. As a result, for example, even if a family enters the same store at the same time and detects the same specific sound, it is possible to provide services suitable for user identification such as age and gender, such as generating different events for adults and children. .

On the other hand, the invention of the acoustic signal detection device is an acoustic signal detection device that detects a specific acoustic signal included in an evaluation acoustic signal to be evaluated, and acquires a sound and outputs it as an evaluation acoustic signal Then, for the evaluation acoustic signal output from the acoustic acquisition unit, the presence or absence of the acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are binarized for each scale. And a target conversion unit that generates a comparison target template recorded in time series, and based on the specific acoustic signal, the presence or absence of the acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and A sound source template generation unit that generates a sound source template that is binarized for each scale and recorded in time series, and a sound source template that acquires the sound source template The acquisition unit, the template comparison unit that compares the comparison target template and the sound source template to calculate the matching rate of the binarized acoustic signal distribution, and the comparison result according to the matching rate calculated by the template comparison unit A matching rate output unit that outputs a signal, a template identifier that identifies a sound source template generated by the sound source template generation unit, an event database that stores an event identifier that identifies an event to be generated, and a matching rate output unit Based on the output comparison result signal, the event database is checked for the template identifier of the sound source template related to the comparison result signal, an event to be generated is selected, and an event occurrence management unit that generates the selected event is provided. The sound source template generator When generating the source template, detects the signal intensity at each scale, extracts the presence or absence of a sound signal based on the detected acoustic signals on whether exceeds a predetermined threshold value, said threshold value, generates It is set to a different value depending on the type of event to be performed .

According to the present invention, the specific sound signal is obtained by comparing the templates generated by extracting the presence / absence of the signal in the frequency band based on the scale equal temperament. And real-time processing is possible by high-speed calculation by downsizing the data. Further, since the presence / absence of an acoustic signal is detected by a binarized template, detection under noise is possible. As a result, it is possible to provide a service triggered by specific sound detection, such as distributing coupons or points when a theme song broadcast in a store is detected. Diversification and enhancement can be achieved. In particular, in the present invention, since the acoustic signal detection device itself takes the lead in detecting the specific acoustic signal and calculating the comparison result, it does not depend on the server device or the like arranged on the communication network, Further real-time processing is possible by the stand-alone operation in the acoustic signal detection device.
When generating the sound source template, it is preferable to detect the presence or absence of a sound of each scale after removing the harmonic component.
Further, at the time of comparison with the sound source template, the presence / absence of each tone may be detected without removing the harmonic component and including the harmonic component.

  In this case, an event triggered by the detection of the specific sound in the information processing terminal can be generated by associating the sound source template used for detection of the specific sound with an arbitrary event. In addition, it is possible to collect information on which user has detected what kind of sound and where, to investigate realistic user trends, and to improve marketing by building big data. .

  In this case, by generating a sound source template by changing the threshold value, different events can be generated depending on the type of the sound source template even when the same specific sound is detected. As a result, for example, even if a family enters the same store at the same time and detects the same specific sound, it is possible to provide services suitable for user identification such as age and gender, such as generating different events for adults and children. .

  The system, method, server, and apparatus according to the present invention described above can be realized by executing a program written in a predetermined language on a computer.

That is, the present invention is an acoustic signal detection program for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated by a computer,
(1) An evaluation acoustic signal is acquired by a computer, and for the acquired evaluation acoustic signal, the presence or absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are An evaluation target conversion step for generating a comparison target template binarized for each scale and recorded in time series,
(2) For a specific acoustic signal, the presence or absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are binarized for each scale and recorded in time series. A template comparison step of generating and obtaining a sound source template, comparing the comparison target template with the sound source template, and calculating a matching rate of the binarized acoustic signal distribution;
(3) a precision ratio output step of outputting a comparison result signal corresponding to the precision ratio calculated by the template comparison step ;
(4) a database control step for associating a template identifier for identifying the sound source template generated in the template comparison step with an event identifier for identifying the event to be generated in an event database;
(5) Based on the comparison result signal output in the precision ratio output step, the event database is checked for a template identifier of a sound source template related to the comparison result signal, an event to be generated is selected, and the selected event is selected. to execute an event generating management step of generating a processing including <br/>,
In the template comparison step, when the sound source template is generated for the specific sound signal, the signal intensity in each scale is detected, and the sound signal is determined based on whether the detected sound signal exceeds a predetermined threshold value. The threshold value is set to a different value depending on the type of event to be generated .

  By installing such a program on a computer or IC chip such as a user terminal or a Web server and executing it on the CPU, an acoustic signal detection system, an acoustic signal detection method, an acoustic signal detection server having the above-described functions, And an acoustic signal detection apparatus can be constructed | assembled easily. This program can be distributed through, for example, a communication line, and can be transferred as a package application that operates on a stand-alone computer by being recorded on a recording medium readable by a general-purpose computer. Specifically, the recording medium can be recorded on various recording media such as a RAM card in addition to a magnetic recording medium such as a flexible disk or a cassette tape, or an optical disk such as a CD-ROM or DVD-ROM. And according to the computer-readable recording medium which recorded this program, the above-mentioned acoustic signal detection system, acoustic signal detection method, acoustic signal detection server, and acoustic signal detection device are used using a general-purpose computer or a dedicated computer. The program can be easily implemented, and the program can be easily stored, transported, and installed.

  As described above, according to the present invention, a target specific acoustic signal can be detected in real time while reducing the processing load even under noise. And by detecting specific sound detection in real time, it is possible to provide a service triggered by the sound, and to diversify and enhance the service.

It is a conceptual diagram which shows the whole structure of the acoustic signal detection system which concerns on 1st Embodiment. It is a block diagram which shows the internal structure of the management server which concerns on 1st Embodiment. It is a block diagram which shows the internal structure of the user terminal which concerns on 1st Embodiment, and a sound reproduction means. It is explanatory drawing which shows the sound source template and comparison destination template which concern on 1st Embodiment. It is explanatory drawing which shows the part position where the sound source template which concerns on 1st Embodiment is produced. It is explanatory drawing which shows the double buffering process which concerns on 1st Embodiment. It is explanatory drawing which shows the template comparison method which concerns on 1st Embodiment. (A) And (b) is explanatory drawing which shows the display information displayed according to the relevance rate which concerns on 1st Embodiment. (A)-(c) is explanatory drawing which shows the display information displayed according to the relevance rate which concerns on this embodiment. (A) And (b) is explanatory drawing which shows the display information displayed according to the relevance rate which concerns on 1st Embodiment. It is a flowchart figure which shows the acoustic signal detection method which concerns on 1st Embodiment. It is a conceptual diagram which shows the whole structure of the acoustic signal detection system which concerns on 2nd Embodiment. It is a block diagram which shows the internal structure of the management server which concerns on 2nd Embodiment. It is a block diagram which shows the internal structure of the user terminal and sound reproduction means which concern on 2nd Embodiment. It is a flowchart figure which shows the acoustic signal detection method which concerns on 2nd Embodiment.

[First Embodiment]
(Outline of acoustic signal detection system)
Exemplary embodiments of a system according to the present invention will be described below in detail with reference to the accompanying drawings. FIG. 1 is a conceptual diagram showing an overall configuration of a system according to the first embodiment.

  As shown in FIG. 1, the acoustic signal detection system according to the present embodiment is an acoustic signal detection system that detects a specific acoustic signal included in an evaluation acoustic signal to be evaluated, and connects communication lines to each other. The management server 2 arranged on the communication network 5 constructed as described above, the radio base station 3, and the user terminal 1 capable of radio communication through the radio base station 3 are provided.

  The communication network 5 is an IP network using the communication protocol TCP / IP, and various communication lines (public lines such as telephone lines, ISDN lines, ADSL lines, optical lines, private lines, wireless communication networks) are mutually connected. It is a distributed communication network constructed by connecting. This IP network includes a LAN such as an intranet (company network) or a home network based on 10BASE-T, 100BASE-TX, or the like.

  The wireless base station 3 is a device that is connected to the communication network 5 through a gateway device (not shown), establishes a wireless communication connection with the user terminal 1, and provides a telephone call and data communication by the user terminal 1.

  The user terminal 1 is an arithmetic processing unit including a CPU, and is a portable mobile computer having a wireless communication function, a PDA (Personal Digital Assistance), a mobile phone, a smartphone, a tablet PC, or the like. This user terminal 1 communicates wirelessly with a relay point such as a base station, and can receive a communication service such as a call or data communication while moving. Examples of the communication method of the user terminal 1 include an FDMA method, a TDMA method, a CDMA method, a W-CDMA, a PHS (Personal Handyphone System) method, and the like.

  The user terminal 1 is equipped with functions such as a digital camera function, an application software execution function, or a GPS function, and also includes a microphone for acquiring an acoustic signal from the outside. Functions as an acoustic signal detection device that acquires a sound signal or a sound file.

  The management server 2 is a general server device that is arranged on the communication network 5 and manages services for users. In a document system such as WWW (World Wide Web), an HTML (Hyper Text Markup Language) file or an image file is used. , A server computer that transmits information such as music files or software having such functions, stores information such as HTML documents and images, and communicates over the Internet in response to requests from client software such as a Web browser These pieces of information are transmitted through the network 5.

  In particular, the management server 2 generates, for the user terminal 1, a sound source template D1 for detecting a specific sound signal and event information D2 that is information relating to an event to be generated in association with the sound source template D1. These pieces of information are distributed to the user terminal 1.

  Here, the “specific sound signal” means a signal containing only sound to be detected by the system, for example, in an original theme song, a station, or a train that is flown into a store in a supermarket or a home electronics store. The name of the station or the departure melody, one phrase of the artist's music, etc. are included.

  In addition, the “sound source template D1” refers to the extraction of the presence or absence of an acoustic signal in a predetermined frequency band based on the scale average rate for a specific acoustic signal, and the distribution of the acoustic signal and its temporal change are binary for each scale. Information recorded in time series (STT (short-time Fourier transform) data). The “sound source template” is generated using any device of the system. For example, the user may generate the sound source template on the user terminal, or the generated sound source template may be downloaded from the server. .

  In addition, the “scale average rate” includes various scales in addition to the 12 scales composed of semitones. “Binary binarization of the presence / absence of an acoustic signal” means detecting only whether there is a sound of each scale, for example, a method of determining by whether an acoustic signal exceeds a predetermined threshold Is mentioned.

  In addition, the event occurs when the user moves to a predetermined area A and detects a specific sound source. Various events such as message output by voice or text display, image / video playback, application startup, etc. Information processing services. Data or programs related to these various information processing services are distributed as event information D2. The information D2 to be distributed includes, for example, coupon information, ticket information, present information, transfer guidance information, notification information for notifying that it is a destination, and the like.

  The area A is a place where a specific sound is output, such as a shopping mall or a concert venue, and sound (theme song or sound effect) is output from the sound reproduction means 4 such as a speaker in the area A. It has become so. Then, by listening to the sound with the microphone of the user terminal 1 and acquiring it as sound data, an event is generated according to the presence or absence of the sound source template D1 included in the sound.

(Internal structure of each device)
Next, the internal structure of each device constituting the acoustic signal detection system described above will be described. FIG. 2 is a block diagram showing an internal configuration of the management server 2 according to the first embodiment, and FIG. 3 is a block diagram showing an internal configuration of the user terminal 1 according to the first embodiment. 4A to 4C are explanatory diagrams showing the sound source template D1 and the comparison destination template according to the first embodiment. Here, the horizontal axis in FIG. 4 indicates the passage of time, and the vertical axis in FIG. 4 indicates the frequency. FIG. 5 is an explanatory diagram showing a part position where the sound source template D1 is created. The “module” used in the description refers to a functional unit that is configured by hardware such as an apparatus or a device, software having the function, or a combination thereof, and achieves a predetermined operation. .

(1) Management server 2
The management server 2 can be composed of a plurality of server groups such as a Web server and a database server in addition to a single server device. In the present embodiment, the communication server 21, the control unit 22, and various databases 23 to 25.

  The communication unit 21 is a communication interface that transmits and receives data to and from the user terminal through the communication network 5. The databases 23 to 25 are a database group that accumulates various types of information related to the system. In the present embodiment, the user information database 23 that holds information about the user who owns the user terminal 1, the event database 24, and the sound source template And a database 25.

  In the user information database 23, in addition to the identification number for identifying the user, the telephone number of the user terminal and the e-mail address, the attribute information such as the user's age, sex, address, and preference information, and the history information on which acoustic detection has been performed Are recorded in association with each other.

  The sound source template database 25 is a storage device that stores the sound source template D1 generated for the specific sound signal, and the sound source template D1 generated by the sound source template generation unit 222 of the control unit 22 identifies the sound source template D1. It is stored in association with the template identifier.

  The event database 24 is an apparatus for accumulating event information D2 generated for a user, and an event identifier for identifying the event is added. In the present embodiment, the event database 24 stores the event identifier of the event to be generated in association with the template identifier.

  The event to be generated may be, for example, an event common to all users, or may be changed according to user attributes. That is, even if the sound source template D1 is the same, for example, if the user is a woman, an event identifier relating to cosmetics, food, etc. is associated. For example, if the user's gender is male, an event relating to sports such as golf or home appliances Can be associated with each other.

  Also, for example, as shown in FIG. 5, parts at different positions are cut out from one phrase of the same music, generated as sound source templates D11 to D13, and sound source templates at different positions are distributed according to user attributes. Thus, for example, even when a family enters the same store at the same time, it is possible to change event information to be generated by the user, such as generating different events for adults and children.

  At this time, the part to be cut may overlap with the sound source templates D11 and D12 or may not overlap with other templates like the sound source template D13. Furthermore, an event can be selected based on the user's age and preference information, and a service suitable for identifying the user such as age and sex can be provided.

  As described above, in order to change an event to be generated for each user, attribute information (for example, female) is associated with event information (for example, information on food) in the event database 24, and this attribute A template identifier (for example, a theme song of a makeup company) corresponding to the information is associated with the event identifier.

  The control unit 22 includes a processor such as a CPU or DSP (Digital Signal Processor), a memory, hardware such as other electronic circuits, software such as a program having the function, or a combination thereof. It is a module, and various function modules are virtually constructed by appropriately reading and executing a program, and various processes for operation control of each unit and user operations are performed by the constructed function modules.

  The control unit 22 includes a sound source data acquisition unit 221, a sound source template generation unit 222, an event generation unit 223, a distribution management unit 224, and a user registration unit 225 as functions for detecting sound.

  The user registration unit 225 is a module that accepts user registration for receiving this system. When the user registration unit 225 acquires various types of user information, the user registration unit 225 assigns an identifier for identifying the user and records the information in the user information database 23.

  The sound source data acquisition unit 221 is a module that acquires a specific acoustic signal to be evaluated. For example, the sound source data acquisition unit 221 acquires an audio signal or an audio file from a service providing side that issues a coupon.

  The sound source template generation unit 222 is a module that generates the sound source template D1 based on the specific sound signal acquired by the sound source data acquisition unit 221. Specifically, the sound source data acquisition unit 221 generates a scale table by rounding the sound frequency band to the 12 scale average rate. As a method for creating a scale table, the front and rear fm * 2 ^ (± 1/12) from the center frequency fm are used as boundary values. Further, a maximum of N volume (Amplitude) excluding harmonic components is extracted from values exceeding a certain threshold. Note that N is an integer of 1 or more. As for this harmonic component, when creating a sound source template, the presence or absence of the sound of each scale is detected after removing the harmonic component, and when comparing with the template, the harmonic component is not removed. The presence / absence of each tone is detected while including the harmonic component. As a result, it is possible to reduce the effect of harmonic components that are harmonics of the scale to be detected, and to create a more appropriate sound source template. Accuracy can be increased.

  Then, by performing this processing with each window width (number of blocks), as shown in FIG. 4A, the presence or absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal A sound source template D1 having a predetermined length in which the temporal change is binarized for each scale and recorded in time series is created. The number of feature points extracted in each window width (number of blocks) may be the same or different. When extracting different numbers, all the values exceeding the threshold value excluding the harmonic component are acquired.

  In the present embodiment, the sound source template generation unit 222 detects the signal intensity in each scale when generating the sound source template D1 for the specific sound signal, and whether the detected sound signal exceeds a predetermined threshold value. The presence or absence of an acoustic signal may be extracted based on whether or not it is present. This threshold value is set to a different value depending on the type of event to be generated.

  In the present embodiment, it is assumed that a template is prepared with a unit of fundamental tone +1/2 Freq and a frequency excluding the speech area. Furthermore, it is assumed here that the machine learning function is used to learn the frequency domain to be extracted, the number, and the window width. As the machine learning function, an artificial neural network, a genetic algorithm, supervised learning such as reinforcement learning, and the like can be used. For this purpose, learning data is stored in the server.

  The sound source template D1 generated in this way is added with a template identifier and stored in the sound source template database 25. As shown in FIG. 5, the sound source template generation unit 222 can also create sound source templates D11 to D13 from different parts even for the same music piece.

  The event generation unit 223 is a module that generates event information to be provided to the user, and stores it in the event database 24 in association with an event identifier that identifies the generated event information D2.

  The distribution management unit 224 is a module that manages control for distributing various types of data on the communication network, and includes a sound source template distribution unit 224a. The sound source template distribution unit 224a is a module that distributes the sound source template D1 generated by the sound source template generation unit 222 to the user terminal 1 based on access from the user terminal.

  At this time, the distribution management unit 224 registers the template identifier of the sound source template D1 distributed by the sound source template distribution unit 224a and the event identifier of the event to be generated in the event database 24 in association with each other. The distribution management unit 224 may associate the template identifier with the event identifier of the event to be generated in advance, and distribute the sound source template D1 and the event information D2 to the user terminal 1 together.

  The distribution management unit 224 has a function of referring to attribute information of the accessed user and selecting event information related to the attribute information. According to this function, for example, as shown in FIG. 5, when sound source templates D11 to D13 of different parts are generated from one phrase, the sound source templates D11 to D13 of each part are selected, Different events can be selected for.

(2) User terminal and sound reproduction means Next, the internal structure of the user terminal 1 and the sound reproduction means 4 will be described. FIG. 6 is an explanatory diagram illustrating a double buffering process according to the present embodiment, and FIG. 7 is an explanatory diagram illustrating a template comparison method according to the present embodiment. Moreover, FIGS. 8-10 is explanatory drawing which shows the display information displayed according to the relevance rate which concerns on this embodiment.

  As shown in FIG. 3, the sound reproducing means 4 is a means for reproducing sound, a control unit 41 that reads out and reproduces an audio signal or an audio file, and a speaker 42 that outputs the audio signal or audio file as sound to the outside. And buffers 43a and 43b for temporarily storing audio signals and audio files. The control unit 41 is an electronic circuit for music reproduction such as an audio amplifier, reads out audio signals and audio files recorded on various recording media, amplifies the audio signals and audio files with an amplifier and the like, and supplies them to the speaker 42. input. The speaker 42 is a device that converts an audio signal and an audio file into physical vibrations and outputs them to the outside.

  When the specific sound is reproduced, the control unit 41 performs double buffering processing using the buffer 43a and the buffer 43b as shown in FIG. In the present embodiment, the same raw waveform data is held in the two buffers 43a and 43b, and the raw waveform data is input to the playback device of the control unit 41 to output sound alternately, thereby reproducing the data. Is always present and the sound is output without interruption.

  At this time, as shown in FIG. 6, the control unit 41 associates the event so as to generate an event when one of the buffers 43 a becomes empty after starting reproduction, and when an event occurs, The next data is prepared in the buffer, and then playback is started. At this time, by controlling so that the sound data remains in the other buffer 43b, the reproduced sound is reproduced without interruption. In the present embodiment, both buffers 43a and 43b are set at the beginning, and the buffers 43a and 43b are alternately used from the next time.

  On the other hand, as shown in FIG. 3, the user terminal 1 includes a wireless interface 11, an application execution unit 17, an output interface 13, an input interface 12, a memory 15, a sound acquisition unit 16, and a sound source template reception unit 14. And.

  The wireless interface 11 includes a wireless communication function based on a mobile communication protocol for performing a call or data communication, and a wireless communication function based on a data communication protocol such as a wireless LAN. This wireless interface can be realized by a wireless LAN adapter or the like in a mobile computer or PDA. The wireless interface 11 includes a short-distance communication interface such as infrared communication and Bluetooth (registered trademark) in addition to a wireless LAN such as WiFi.

  The output interface 13 is a device that outputs video and sound, such as a display and a speaker. In particular, the output interface 13 includes a display unit 13a such as a liquid crystal display, and a GUI (Graphical User Interface) constructed by an application is displayed on the display unit 13a.

  The input interface 12 is a device for inputting user operations such as operation buttons and a touch panel. The input interface 12 includes a microphone 12a that is output from the outside and acquires an evaluation acoustic signal to be evaluated. The evaluation sound signal acquired by the microphone 12 a is converted into data and input to the sound acquisition unit 16. This “evaluation acoustic signal” is an acoustic signal that is a mixture of various types of externally output sound, and a signal that contains a specific acoustic signal within the noise, or a signal that contains only the noise without the specific acoustic signal. Is also included. In addition, “acquisition of an evaluation acoustic signal” includes a case where sound is directly acquired as an electric signal using an acoustic device such as a microphone, and is acquired as an audio signal or an audio file by wireless / wired communication. The evaluation acoustic signal is a signal in which a specific acoustic signal and a noise acoustic signal cross each other.

  The memory 15 is a storage device that stores an OS (Operating System), programs for various applications, other data, and the like. The memory 15 includes a sound source template D1 distributed from the management server 2 and a sound source. Event information D2 associated with the template D1 is stored in the event database 151. These data are associated based on the template identifier and the event identifier. In the memory 15, the acoustic signal detection program according to the present invention is downloaded from the management server 2 and stored.

  The sound source template receiving unit 14 is a module that receives the sound source template D1 and event information D2 related to the event to be generated based on the sound source template D1 through the communication network 5, and the sound source template receiving unit 14 includes the sound source template D1. As a sound source template acquisition unit. The acquired sound source template D1 and event information D2 are accumulated in the event database 151.

  The sound acquisition unit 16 is a module that acquires sound input from the microphone 12a and outputs the sound as an evaluation sound signal or a specific sound signal. The sound acquisition unit 16 also performs a double buffering process as a sound capturing process, and records the same evaluation sound signal that is input to the two buffers at different times to prevent sound omission. .

  Then, the signal acquired as the evaluation sound signal is input to the evaluation target conversion unit 171, and the signal acquired as the specific sound signal is input to the sound source template generation unit 175. This input destination is determined by starting the sound signal program and selecting either the creation menu or the template comparison menu for the sound source template D1.

  The application execution unit 17 is a module that executes applications such as a general OS, browser software, electronic mail, and image display software, and is usually realized by a CPU or the like. By executing the acoustic signal detection program downloaded from the management server 2 by the application execution unit 17, the evaluation object conversion unit 171, the template comparison unit 172, the matching rate output unit 173, and the event occurrence management unit 174 are executed on the CPU. Then, functional modules of the reception management unit 176 and the sound source template generation unit 175 are virtually constructed.

  The evaluation target conversion unit 171 extracts the presence / absence of an acoustic signal in a predetermined frequency band for the evaluation acoustic signal acquired from the acoustic acquisition unit 16 based on the scale average rate, and the distribution of the acoustic signal and its temporal change are This module generates a comparison destination template D3 binarized for each musical scale and recorded in time series.

  More specifically, the evaluation target conversion unit 171 creates a scale table by rounding the frequency band of the sound to the 12 scale average rate, and the sound volume within a value exceeding a certain threshold is selected from the scale table. (Amplitude) A maximum of M is extracted and generated as a comparison destination template D3. In this evaluation destination template, it is assumed that N <M. Moreover, the comparison destination template D3 shall contain a harmonic component.

  Then, the same processing is performed for each window width (number of blocks) to create a sound source template D1. At this time, since the evaluation target conversion unit 171 performs real-time processing, unlike the sound source template generation, the same number is assumed. By such processing, a comparison destination template D3 as shown in FIG. 4C is generated.

  The template comparison unit 172 is a module that compares the comparison destination template D3 and the sound source template D1 and calculates the matching rate of the binarized acoustic signal distribution. Specifically, the template comparison unit 172 extracts the sound source template D1 shown in FIG. 4B from the event database 151, and compares the comparison target template D3 as shown in FIG. 4C with one window width (block). The precision is calculated by making a comparison while shifting the number.

  Here, the processing by the evaluation object conversion unit 171 and the template comparison unit 172 is executed in real time by having the input comparison destination template D3 in the FIFO (First In First Out) method. Specifically, as illustrated in FIG. 7, the evaluation acoustic signal input to the evaluation target conversion unit 171 is subjected to double buffering processing and is subjected to templating processing in block width units. Then, the data having the same template width as the previously analyzed sound source template D1 is held in the FIFO method, and the comparison destination template D3 to be held is compared with the sound source template D1, and the matching rate output unit 173 performs the comparison. Calculate the precision. It should be noted that the comparison destination template D3 is generated in real time while the precision ratio output unit 173 calculates the precision ratio in comparison with the sound source template D1.

  The matching rate output unit 173 is a module that outputs a comparison result signal corresponding to the matching rate calculated by the template comparison unit 172. In the present embodiment, the comparison result may be divided depending on whether or not the matching rate is a perfect match, or the matching rate may be divided into a plurality of stages, and the comparison result may be changed according to the ratio of the matching rate. . In this way, by changing the comparison result signal in accordance with the ratio of the relevance ratio, it is possible to execute a sound radar function that notifies that a specific acoustic signal that triggers the occurrence of an event is nearby.

Specifically, when the sound radar function is executed, the precision for sound radar is calculated. The calculation of the matching rate is as follows: [M_a] is the template matching rate for only the frequency band ignoring time, [M_t] is the template matching rate considering time, and [α] is the weighting coefficient.
(Expression 1) Precision for sound radar = M_a + α * M_t
Is calculated by

  The event occurrence management unit 174 collates the event database 151 with respect to the template identifier of the sound source template D1 related to the comparison result signal based on the comparison result signal output from the matching rate output unit 173, and selects an event to be generated. This module generates selected events. The event occurrence management unit 174 has a matching rate display unit 174a that controls to display a graphic corresponding to the matching rate based on the comparison result signal output from the matching rate output unit 173.

  For example, when the relevance ratio display unit 174a determines only whether the relevance ratios are completely coincident in the relevance ratio output section 173, when the comparison result signal indicating successful detection is acquired, as shown in FIG. When a comparison result signal indicating that the matching rate is not satisfied and detection failure is obtained is displayed on the screen of the display unit 13a, a character as shown in FIG. 8B is displayed on the screen of the display unit 13a.

  On the other hand, when the sound radar precision is calculated by the sound radar function, a comparison result signal corresponding to the precision is input, and the event occurrence management unit 174 changes the event according to the comparison result. Also change the display on the screen.

  When this process is executed, a plurality of sound source templates D1 in which the presence or absence of an acoustic signal is changed according to a predetermined threshold value are used, and the comparison result signal is stepwise according to the strength of the signal intensity. Change. Specifically, when the relevance ratio is low, the relevance ratio display unit 174a displays a character with a facial expression “the specific sound signal cannot be heard” as shown in FIG. In addition, when the matching rate is a certain degree, the character is changed to a character having an expression of “the specific sound is nearby” as shown in FIG. Then, if the relevance ratio is completely coincident, a character with a facial expression of “event can be generated” is displayed as shown in FIG. 9C, and then event information is displayed. In addition, when the accuracy rate up to the occurrence of the event has not been reached, the intensity and input direction of the input sound signal are detected, and the eye size and mouth movement are indicated so as to indicate the distance and direction to the specific sound. It may be changed.

  As a method for detecting the input direction, for example, the total volume of the matching sound in the comparison result with the template is calculated as the total number of matching blocks in the template, and the transition of the number of blocks is recorded as a history. There is a method of comparing the calculation result and the total of the blocks of the current calculation result. In this method, the number of matching blocks of the previous time and this time are compared, and in the comparison, if the number of blocks increases, it is determined that they are approaching, and if they are decreasing, it is determined that they are far away. As another method, the directionality can be detected more clearly by making the microphone device for acquiring the sound directional, and by using multiple microphones and beam forming, The sound source direction may be specified by the strength of the sound. When the terminal is provided with position information acquisition means such as a GPS, a compass, and a gyro, the input direction may be specified from the moving direction of the terminal and the intensity change of the acoustic signal.

  By executing such a sound radar function, it is possible to notify and guide the user whether the sound source template D1 that provides the service is near the user or whether the sound source template D1 is approaching the location.

  In addition, the matching rate display unit 174a classifies the generated comparison target template D3 by dividing the presence / absence of the acoustic signal by the scale average rate as shown in FIGS. You may display with a shape. In this display information, one circle represents one octave, and a higher frequency (high octave region) is displayed on the outer circle. Even in this case, the display may be changed so as to indicate the distance or direction to the specific sound.

  The sound source template generation unit 175 is a module that generates the sound source template D1 on the user terminal 1 side based on the specific sound signal. The generation of the sound source template D1 is generated by a process similar to that of the sound source template generation unit 222 in the management server 2, and a process of extracting the presence / absence of an acoustic signal according to a threshold value can also be executed.

  The reception management unit 176 is a module that associates the template identifier of the sound source template D1 received by the sound source template reception unit 14 with the information D2 related to the event to be generated based on the template identifier, and registers the information in the event database 151.

(Acoustic signal detection method)
Next, the acoustic signal detection method of the present invention will be described. FIG. 11 is a flowchart showing an outline of the operation of the acoustic signal detection system. In this embodiment, it is assumed that user registration is completed in the management server 2.

  In the acoustic signal detection method for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated, first, the event generation unit 223 generates event information D2 related to an event generated for the user (S101). This event information D2 is stored in the event database 24 in association with an event identifier for identifying the event.

  Next, the sound source template generation unit 222 acquires a specific sound signal (S102), and performs sound source template generation processing for generating a sound source template D1 based on the specific sound signal (S103). At this time, in the generation process of the sound source template D1, when the sound source template D1 is generated for the specific sound signal, the signal intensity in each scale is detected, and whether or not the detected sound signal exceeds a predetermined threshold value. Based on this, the presence / absence of an acoustic signal may be extracted. Note that the threshold value is set to a different value depending on the type of event to be generated.

  The generated sound source template D1 is added with a template identifier for identifying the sound source template D1 and stored in the sound source template database 25. At this time, the distribution management unit 224 associates a template identifier for identifying the sound source template D1 with an event identifier for identifying an event to be generated, and performs database control processing for storing the event identifier in the event database 24 (S103). At this time, the distribution management unit 224 refers to the user information database 23 and selects event information D2 based on user attribute information (gender, age, or preference information) possessed by the user terminal 1 that is the distribution destination. You may control to do.

  Thereafter, the sound source template distribution unit 224a distributes the sound source template D1 through the communication network 5 when the management server 2 takes the initiative or acquires a signal requesting the sound source template D1 from the user terminal 1. Is executed (S105). At this time, the distribution management unit 224 may also distribute the event information D2 to be generated based on the sound source template D1, and if the event to be generated is changed, the new event information D2 is sent to the user. Delivered to the terminal 1. When the new event information D2 is distributed, the distribution management unit 224 updates and registers the event to be generated in the event database 24 in association with the template identifier.

  In the user terminal 1, when the sound source template receiving unit 14 receives the sound source template D1 and event information D2 to be generated based on the sound source template D1 (S106), the sound source template D1 is identified by the control of the reception management unit 176. The template identifier to be associated with the event identifier for identifying the event to be generated is stored in the event database 151 (S107). Thereafter, the acoustic signal detection program is activated, and external sound is read from the microphone 12a (S109) while a predetermined sound is being output from the speaker 42 of the sound reproducing means 4 (S108).

  The sound acquisition unit 16 of the user terminal 1 acquires this sound as an evaluation sound signal, and inputs the acquired evaluation sound signal to the evaluation target conversion unit 171. The evaluation target conversion unit 171 executes a process of converting the evaluation target to generate the comparison target template D3 based on the evaluation acoustic signal (S110).

  When the comparison destination template D3 is generated by the FIFO method, the comparison destination template D3 is input to the template comparison unit 172. The template comparison unit 172 compares the comparison destination template D3 with the sound source template D1, and executes a template comparison process for calculating the matching rate of the binarized acoustic signal distribution (S111).

  The precision calculated by the template comparison process is input to the precision ratio output unit 173, and the precision ratio output unit 173 executes a precision ratio output process of outputting a comparison result signal corresponding to the precision ratio (S112). The comparison result signal output by the precision ratio output process is input to the event occurrence management unit 174.

  The event occurrence management unit 174 determines whether or not an event has occurred based on the output comparison result signal (S113). The presence / absence of this event is determined based on, for example, whether the comparison result signals are completely identical for both templates or whether the degree of coincidence is a predetermined threshold value or more. Note that an event to be generated may be selected according to the degree of coincidence, and thereby, a message to be displayed and an application to be activated can be switched depending on whether the target sound source is far or near.

  An event here is a process that occurs on the condition that certain requirements are satisfied, such as exception processing and branch processing in the program, and the system or service provided by the process generated by this event Is also included in the “event” mentioned here. For example, when a specific sound source is detected, various information processing services such as message output by voice or text display, image / video playback, application activation, and the like are also included in the event.

  If it is determined in step S113 that no event will occur ("N" in S113), the processing from steps S108 to S112 is repeated without generating an event. On the other hand, if it is determined that an event will occur ("Y" in S113), the event database 24 is checked for the template identifier of the sound source template D1 related to the comparison result signal, and an event to be generated is selected and selected. The event occurrence management process for generating the event is executed (S114). At this time, the matching ratio display unit 174a displays a graphic corresponding to the matching ratio based on the comparison result signal output by the matching ratio output process (S115).

  Thereafter, the process from step S108 to step S115 is repeated until the specific sound detection process is selected until the specific sound detection process is selected by event occurrence or user operation ("N" in S116). ("Y" in S116), the process ends.

(Acoustic signal detection program)
The above-described acoustic signal detection system, acoustic signal detection device, acoustic signal detection server, and object control method according to the first embodiment can be realized by executing a program described in a predetermined language on a computer. That is, by installing this program on a computer such as a user terminal or a Web server or an IC chip and executing it on the CPU 2, a system having the above-described functions can be easily constructed. This program can be distributed through a communication line, for example, and can be transferred as a package application that operates on a stand-alone computer.

  Such a program can be recorded on a recording medium that can be read by a personal computer. Using a general-purpose computer or a dedicated computer, the above-described acoustic signal detection system, acoustic signal detection device, acoustic signal detection server, In addition, the object control method can be implemented, and the program can be easily stored, transported, and installed.

(Action / Effect)
According to the present embodiment, the specific acoustic signal is obtained by comparing the templates generated by extracting the presence / absence of the signal in the frequency band based on the scale equal temperament, so that complicated calculation processing is not required. Matching processing with a sound source can be performed, and real-time processing can be performed by high-speed computation by downsizing data. Further, since the presence / absence of an acoustic signal is detected by a binarized template, detection under noise is possible. As a result, it is possible to provide a service triggered by specific sound detection, such as distributing coupons or points when a theme song broadcast in a store is detected. Diversification and enhancement can be achieved.

  Moreover, in this embodiment, the event which triggered the detection of the specific sound in the user terminal 1 can be generated by associating the sound source template D1 used for detection of the specific sound with an arbitrary event. In addition, it is possible to collect information on which user has detected what kind of sound and where, to investigate realistic user trends, and to improve marketing by building big data. . In particular, in the present embodiment, the user terminal itself takes the lead in detecting the specific acoustic signal and calculating the comparison result, so that it is more reliable without depending on the server device or the like arranged on the communication network 5. Real-time processing is possible.

  Further, according to the present embodiment, the sound source template generation unit 222 detects the signal intensity in each scale when generating the sound source template D1, and sets the detected acoustic signal to a different value according to the type of event. Since the presence / absence of an acoustic signal is extracted based on whether or not the threshold value is exceeded, the same specific sound is detected by creating the sound source template D1 by changing the threshold value. Even if it exists, a different event can be generated according to the kind of sound source template D1. As a result, for example, even if a family enters the same store at the same time and detects the same specific sound, it is possible to provide services suitable for user identification such as age and gender, such as generating different events for adults and children. .

  Further, according to the present embodiment, the template identifier of the sound source template D1 distributed by the sound source template distribution unit 224a is registered in the event database 24 in association with the event to be generated. The sound source template D1 can be distributed via a Web page, e-mail, etc., and an event can be associated with each sound source template D1, and the service on the Web and the service at the actual store can be closely linked. The service can be refined.

  In the present embodiment, the precision ratio display unit 174a displays a graphic corresponding to the precision ratio based on the comparison result signal output from the precision ratio output section 173, so that the precision ratio is one of the events. By displaying the above, the operator can visually recognize the precision and can visually confirm the distance to the sound source. As a result, for example, a mechanism like a sound source radar that indicates the distance and direction to a specific sound can be constructed, and even in a noisy environment, the user can be guided to the position of the sound source. It is possible to provide various services with high game characteristics using.

[Second Embodiment]
In the above-described embodiment, the acoustic signal detection program, the sound source template D1, the event information D2, and the like are held in the user terminal, and the process of detecting the acoustic signal in the stand-alone format is executed. The processing is not limited, and these processes may be executed in a server arranged on the communication network 5. FIG. 12 is a conceptual diagram showing an overall configuration of an acoustic signal detection system according to the second embodiment. Note that in the second embodiment, the same components as those in the first embodiment described above are denoted by the same reference numerals, and the functions and the like are the same unless otherwise specified, and the description thereof is omitted.

(Outline of acoustic signal detection system)
As shown in FIG. 11, the acoustic signal detection system according to the present embodiment performs wireless communication through server groups 6 and 7, a wireless base station 3, and a wireless base station 3 that are arranged on a communication network 5. The user terminal 1A capable of

  The user terminal 1A is a portable information processing terminal having an arithmetic processing function and a wireless communication function including a CPU similar to that in the first embodiment described above, and wirelessly communicates with a relay point such as a base station. Communication services such as calls and data communication can be received while moving. The user terminal 1 </ b> A includes an execution function of application software and the like, and includes a microphone that acquires an acoustic signal from the outside. The microphone acquires external sound as an audio signal or an audio file.

  The server groups 6 and 7 are general server devices having different processing functions. In the present embodiment, a plurality of template comparison servers 7 (7a to 7n) that compare both templates, and a plurality of control management servers 6 (between the user terminal 1A and the template comparison server 7) that perform load distribution. 6a-6n). The control management server 6 also has a function of generating an event for the user terminal according to the comparison result acquired from the template comparison server 7.

  The control management server 6 (6a to 6n) and the template comparison server 7 (7a to 7n) are configured by a plurality of server devices so as to perform template comparison and event generation processing according to the type of service to be provided. It is assumed that

  In this embodiment, the user terminal 1A acquires the evaluation acoustic signal from the outside, and transmits the evaluation acoustic signal in its data format as it is to the communication network 5, so that the control management server 6 and the template comparison server 7 A specific acoustic signal detection process is executed, and the comparison result is returned to the user terminal 1A to generate an event.

(Internal structure of each device)
Next, the internal structure of each device constituting the acoustic signal detection system described above will be described. FIG. 12 is a block diagram showing an internal configuration of a user terminal 1A according to the second embodiment, and FIG. 13 is a block diagram showing an internal configuration of a control management server and a template comparison server according to the second embodiment.

(1) User terminal First, the internal structure of the user terminal 1A will be described. As illustrated in FIG. 12, the user terminal 1 </ b> A includes a wireless interface 11, an application execution unit 18, an output interface 13, an input interface 12, a memory 15, and a sound acquisition unit 16.

  Similarly to the above, the sound acquisition unit 16 is a module that acquires the sound input from the microphone 12a and outputs the sound as an evaluation sound signal or a specific sound signal. In the present embodiment, these signals are transmitted to the acoustic transmission unit 181. The application execution unit 18 is a module that executes an application such as a general OS, browser software, electronic mail, and image display software, and is also realized by a CPU or the like in this embodiment. By executing the acoustic signal detection program downloaded from the management server 2 by the application execution unit 18, the acoustic transmission unit 181 and the event execution unit 182 are configured in the present embodiment.

  The acoustic transmission unit 181 is a module that transmits the evaluation acoustic signal acquired from the acoustic acquisition unit 16 or the specific acoustic signal to the control management server 6 together with the user identifier for identifying the user terminal 1A by converting the packet into an IP packet. . It is assumed that the control management server 6 (6a to 6n) serving as a distribution destination is a server device that has undergone user registration.

  The event execution unit 182 is a module that acquires event information D2 selected based on the comparison result sent from the control management server 6 and executes a predetermined event according to the event information D2. The event executed in this embodiment also includes various information processing services such as message output by voice or text display, image / video playback, application activation, etc. Event information D2 to be acquired is coupon information Ticket information, present information, transfer guidance information, and notification information for notifying that the destination is the destination. And the event execution part 182 displays the display information regarding an event on the display part 13a using this event information D2.

(2) Internal Configuration of Control Management Server and Template Comparison Server Next, the internal configuration of the control management server 6 and the template comparison server 7 will be described. In the present embodiment, the control management server 6 has a user management function, a sound source template creation function, a comparison destination template creation function, and an event occurrence management function, and a user registration unit 62 and an evaluation target conversion unit 63. An event generation management unit 64, an event generation unit 65, a sound source template generation unit 68, a distribution management unit 69, a communication unit 61, a user information database 67, and an event database 66.

  The user registration unit 62 is a module that accepts user registration for receiving this system. When the user registration unit 62 acquires various types of user information, the user registration unit 62 assigns an identifier for identifying the user and records the information in the user information database 67. In the user information database 67, in addition to the identifier for identifying the user, the telephone number and mail address of the user terminal, attribute information such as the user's age, gender, address, and preference information, and history information on which acoustic detection has been performed Are recorded in association with each other.

  The evaluation target conversion unit 63 extracts the presence / absence of an acoustic signal in a predetermined frequency band from the evaluation acoustic signal transmitted from the user terminal 1A based on the scale average rate, and the distribution of the acoustic signal and its temporal change are the scales. This is a module that generates a comparison destination template D3 binarized and recorded in time series. Since this specific process is the same as that of the first embodiment, a description thereof will be omitted.

  The event generation unit 65 is a module that generates event information D2 to be provided to the user, and stores it in the event database 66 in association with an event identifier that identifies the generated event information D2.

  The sound source template generation unit 68 is a module that generates a sound source template D1 based on, for example, a specific sound signal acquired from a service provider that issues a coupon, and a template identifier is added to the generated sound source template D1. Also in the present embodiment, the sound source template generation unit 68 detects the signal intensity in each scale when generating the sound source template D1, and based on whether the detected acoustic signal exceeds a predetermined threshold value. The presence or absence of an acoustic signal may be extracted. The threshold value is set to a different value depending on the type of event to be generated.

  The distribution management unit 69 is a module that manages control for distributing various types of data on the communication network, and includes a sound source template distribution unit 69a. The sound source template distribution unit 69a is a module that distributes the sound source template generated by the sound source template generation unit 68 to the template comparison server 7 having the template comparison unit 74 through the communication network 5. In this embodiment, the template identifier is assigned to the sound source template distribution unit 69a. Send with added status.

  In addition, the distribution management unit 69 stores the template identifier of the sound source template D1 distributed by the sound source template distribution unit 69a in association with the event identifier of the event to be generated in the event database 66. In this embodiment, the control management server 6 generates and stores the sound source template D1, but the present invention is not limited to this, and the template comparison server 7 generates the sound source template D1. May be accumulated.

  The event database 66 is a device for accumulating event information D2 to be generated for the user, and is stored with an event identifier for identifying the event, and a template identifier for identifying the sound source template D1 and an event to be generated. Is stored in association with an event identifier for identifying the event ID.

  The event occurrence management unit 64 should collate the event database 66 with respect to the template identifier of the sound source template D1 related to the comparison result signal based on the comparison result signal output from the matching rate output unit 75 of the template comparison server 7, and generate the event database 66. This module selects an event and notifies the user terminal 1A of information related to the selected event.

  The event occurrence management unit 64 receives a control signal for displaying a graphic corresponding to the matching rate based on the comparison result signal output from the matching rate output unit of the template comparison server 7 via the communication network 5. It has a matching rate display control unit 64a for sending to 1A. The communication unit 61 is a communication interface that transmits and receives data between the user terminal 1 </ b> A and the template comparison server 7 through the communication network 5.

  Next, the template comparison server 7 will be described. In this embodiment, the template comparison server 7 has a template comparison function and a precision ratio output function, and a sound source template acquisition part 72, a template comparison part 74, a precision ratio output part 75, and a sound source template database. 73 and a communication unit 71.

  The sound source template acquisition unit 72 is a module for acquiring the sound source template generated in the control management server 6 and stores it in the sound source template database 73 based on the associated template identifier. The communication unit 71 is a communication interface that transmits and receives data to and from the control management server 6 through the communication network 5.

  The sound source template database 73 is a storage device that stores the sound source template D1 generated for the specific sound signal, and stores a template identifier that identifies the sound source template D1 in association with the sound source template D1.

  The template comparison unit 74 is a module that compares the comparison target template D3 and the sound source template D1 and calculates the matching ratio of the binarized acoustic signal distribution. The matching rate output unit 75 is a module that sends a comparison result signal corresponding to the matching rate calculated by the template comparison unit 74 to the control management server 6 via the communication network 5. In addition, since the specific process of the template comparison part 74 and the precision ratio output part 75 is the same as that of 1st Embodiment, the description is abbreviate | omitted. Each process including the real-time process and the double buffering process described in the first embodiment is also executed in the present embodiment, and the description thereof is omitted.

(Acoustic signal detection method)
Next, an acoustic signal detection method according to the second embodiment will be described. FIG. 15 is a flowchart showing an outline of the operation of the acoustic signal detection system. Also in this embodiment, it is assumed that user registration is completed in the control management server 6.

  In the acoustic signal detection method according to the present embodiment, first, the event generation unit 65 generates event information D2 related to an event generated for the user (S201). This event information D2 is stored in the event database 66 in association with an event identifier for identifying the event.

  Next, the sound source template generation unit 68 acquires a specific sound signal (S202), and performs sound source template generation processing for generating a sound source template D1 based on the specific sound signal (S203). At this time, in the generation process of the sound source template D1, when the sound source template D1 is generated for the specific sound signal, the signal intensity in each scale is detected, and whether or not the detected sound signal exceeds a predetermined threshold value. Based on this, the presence / absence of an acoustic signal may be extracted. Note that the threshold value is set to a different value depending on the type of event to be generated. Thereafter, the generated sound source template D1 is added with a template identifier for identifying the sound source template D1, and distributed to the template comparison server 7 by the sound source template distribution unit 69a (S204).

  At this time, the distribution management unit 69 performs a database control process in which the template identifier of the sound source template D1 distributed by the sound source template distribution unit 69a and the event identifier of the event to be generated are associated and stored in the event database 66 (S205). . In the template comparison server 7, the sound source template acquisition unit 72 acquires the transmitted sound source template D1 (S206) and stores it in the sound source template database 73 based on the associated template identifier (S207).

  After that, when a predetermined sound is being output from the speaker 42 of the sound reproducing means 4 (S208), when the acoustic signal detection program is started, external sound is read from the microphone 12a. The sound acquisition unit 16 of the user terminal 1A acquires this sound as an evaluation sound signal, and the acquired evaluation sound signal is transmitted to the control management server 6 via the sound transmission unit 181 (S209). The evaluation target conversion unit 63 of the control management server 6 acquires this evaluation acoustic signal (S210), and executes the process of converting the evaluation target to generate the comparison destination template D3 (S211). Note that a user identifier is added to the evaluation acoustic signal distributed from the user terminal 1 so that it can be determined from which user terminal 1A it is sent, and mutual communication is performed based on this user identifier. ing.

  When the comparison destination template D3 is generated, the comparison destination template D3 is distributed to the template comparison server 7 via the communication network 5 (S212). When the template comparison server 7 acquires the comparison destination template D3 (S213), the data is input to the template comparison unit 74. The template comparison unit 74 compares the comparison destination template D3 with the sound source template D1, and executes a template comparison process for calculating the matching rate of the binarized acoustic signal distribution (S214).

  The precision calculated by the template comparison process is input to the precision ratio output unit 75, and the precision ratio output unit 75 executes a precision ratio output process for outputting a comparison result signal corresponding to the precision ratio (S215). The comparison result signal output by the precision ratio output process is transmitted to the control management server 6 via the communication network 5.

  In the control management server 6, the result signal is input to the event occurrence management unit 64, and the event occurrence management unit 64 determines whether or not an event has occurred based on the output comparison result signal (S216). The presence / absence of this event is selected according to the event to be generated, for example, whether the comparison result signal is a perfect match between the two templates or whether it is equal to or greater than a predetermined threshold value. If it is determined that no event occurs ("N" in S216), the processing from step S208 to S215 is repeated without generating an event.

  On the other hand, if it is determined that an event will occur ("Y" in S216), the event database 66 is checked for the template identifier of the sound source template D1 related to the comparison result signal, and an event to be generated is selected and selected. The event occurrence management process for generating the event is executed (S217). Specifically, the event information D2 is distributed to the user terminal 1A based on the user identifier. At this time, the precision ratio display control unit 64a also transmits a display control signal for performing graphic display control according to the precision ratio based on the comparison result signal output by the precision ratio output process.

  Next, the event execution unit 182 of the user terminal 1A displays a graphic corresponding to the comparison result on the display unit 13a based on the event information D2 and the display control signal (S218). Thereafter, the process from step S208 to step S220 is repeated until the specific sound detection process is selected until an event occurs or a specific sound detection process is selected by a user operation ("N" in S219). (“Y” in S219) ends.

(Acoustic signal detection program)
The acoustic signal detection system, the acoustic signal detection device, the acoustic signal detection server, and the object control method according to the second embodiment described above can also be realized by executing a program described in a predetermined language on a computer. . That is, by installing this program on a computer such as a user terminal or a Web server or an IC chip and executing it on the CPU, a system having the above-described functions can be easily constructed. This program can be distributed through a communication line, for example, and can be transferred as a package application that operates on a stand-alone computer.

  Such a program can be recorded on a recording medium that can be read by a personal computer. Using a general-purpose computer or a dedicated computer, the above-described acoustic signal detection system, acoustic signal detection device, acoustic signal detection server, In addition, the object control method can be implemented, and the program can be easily stored, transported, and installed.

(Action / Effect)
In this embodiment as well, the specific sound signal is obtained by comparing the templates generated by extracting the presence / absence of the signal in the frequency band based on the scale equal temperament, so that the sound source is not required for complicated calculation processing. And real-time processing is possible by high-speed calculation by downsizing the data. Further, since the presence / absence of an acoustic signal is detected by a binarized template, detection under noise is possible. As a result, it is possible to provide a service triggered by specific sound detection, such as distributing coupons or points when a theme song broadcast in a store is detected. Diversification and enhancement can be achieved.

  In the present embodiment, an event triggered by detection of specific sound in the user terminal 1A can be generated by associating the sound source template D1 used for detection of specific sound with an arbitrary event. In addition, it is possible to collect information on which user has detected what kind of sound and where, to investigate realistic user trends, and to improve marketing by building big data. . In particular, in the present embodiment, in the servers 6 and 7 arranged on the communication network 5, the evaluation acoustic signal is used as a comparison destination template, or both templates are compared to calculate the matching rate. It is possible to reduce the processing load on the user terminal 1A possessed by the user and to effectively use the memory capacity.

  Further, according to the present embodiment, the sound source template generation unit 68 detects the signal intensity in each scale when generating the sound source template D1, and sets the detected acoustic signal to a different value depending on the type of event. Since the presence / absence of an acoustic signal is extracted based on whether or not the threshold value is exceeded, the same specific sound is detected by creating the sound source template D1 by changing the threshold value. Even if it exists, a different event can be generated according to the kind of sound source template D1. As a result, for example, even if a family enters the same store at the same time and detects the same specific sound, it is possible to provide services suitable for user identification such as age and gender, such as generating different events for adults and children. .

  Further, according to the present embodiment, the template identifier of the sound source template D1 distributed by the sound source template distribution unit 69a is registered in the event database 66 in association with the event to be generated. The sound source template D1 can be distributed via a Web page, e-mail, etc., and an event can be associated with each sound source template D1, and the service on the Web and the service at the actual store can be closely linked. The service can be refined.

  Further, in the present embodiment, the relevance ratio display control unit 64a performs control to display a graphic according to the relevance ratio based on the comparison result signal output from the relevance ratio output section 75. As a result, by displaying the relevance ratio, the operator can visually recognize the relevance ratio and visually confirm the distance to the sound source. As a result, for example, a mechanism like a sound source radar that indicates the distance and direction to a specific sound can be constructed, and even in a noisy environment, the user can be guided to the position of the sound source. It is possible to provide various services with high game characteristics using.

[Various services]
Next, each service applying the acoustic signal detection system and method according to the above-described embodiment will be described below. Note that the following services can be implemented in the first embodiment and the second embodiment described above.

(Coupon delivery service)
First, the case where the acoustic signal detection system and method according to each embodiment described above are applied to a coupon distribution service will be described as an example. When the user terminal 1 recognizes a specific sound output from a store speaker when a user visits an actual store such as a supermarket or a home appliance mass retailer, the coupon distribution service uses the display unit 13a of the user terminal 1 or 1A. A discount service is provided by displaying coupons and point information on the screen.

  More specifically, in the coupon distribution service, the “sound source template” uses a theme song at the same time as a store or a sound effect output during a time sale. The event to be generated for the sound source template D1 may be, for example, store point gift, discount service, or time sale information.

  At this time, for example, by changing the event that occurs according to the user's attributes, for example, even if the same theme song broadcast in a shopping mall is played, women can receive cosmetics and groceries. For example, a coupon related to a toy or the like can be generated as an event for a child, and a coupon related to golf or a home appliance can be generated as an event for a male. In this case, the same part portion may be used as the sound source template D1, and the associated event information may be changed according to the user attribute. Different part portions may be generated as the sound source templates D11 to D13, and each sound source template may be generated. Different event information may be associated with each other.

(Line quality judgment service)
Next, the case where the acoustic signal detection system and method of each embodiment as described above is applied to a line quality determination service will be described as an example. This line quality determination service outputs a specific sound around the telephone at the end of the telephone opening work, and determines the line quality based on whether or not echo (resonance) has occurred in the user terminals 1 and A1. . In this case, the user terminals 1 and 1A are owned by the opening contractor. Further, when this line quality determination is performed, the evaluation formula is 10-level evaluation using the following evaluation formula.

Evaluation formula = Normalized (Sum (M_n [i] × Amp [i])) range: [0, 10]
Here, “M_n” is a matching rate, and “Amp” indicates the average of the volume (Amplitude) at that time (template width). In the above equation, normalization may be performed as needed.

  Further, in this service, as shown in FIGS. 11 (a) and 11 (b), the display unit 13a changes the facial expression of the character in accordance with the state of echo so that the opening contractor can easily recognize the determination result. . Specifically, the image shown in FIG. 11A is displayed when no echo is generated, and the image shown in FIG. 11B is displayed when an echo is generated. In this embodiment, this evaluation is based on 10 levels, but the number is not limited.

  When this line quality determination is performed with live voice, a sound source template is created in real time, and evaluation is performed using the same expression as the above-described evaluation expression. At this time, it is possible to cope with this by appropriately changing the length of the voice to be buffered with respect to the length of the raw voice as the sound source.

  According to such an embodiment, by using the present invention for the line quality service, it is possible to easily investigate whether or not an echo is generated at the time of telephone opening construction.

(Route guidance service)
Next, a case where the acoustic signal detection system and method according to each embodiment described above is applied to a route guidance service will be described as an example. This route guidance service provides guidance to a predetermined destination such as a target platform, a ticket gate, or a toilet based on a specific sound output from a speaker at a station, for example.

  The “sound source template” used in this service includes, for example, voices that call the station name continuously, sound effects that are played at each station (departure music box, etc.), and sound effects that are used independently at other stations. . The destination to be guided includes, for example, a predetermined platform, a ticket gate, and a toilet location. In this service, the relationship between the type of sound and signal intensity and the destination is stored as map information and associated with event information.

  This map information may be accumulated in the user terminal 1 or in the management server 2 or the control management server 6. When the management server 2 or the control management server 6 holds the map information, the update is easy. In addition, when making it hold | maintain in the user terminal 1, the newest map information shall be delivered to the user terminal 1 sequentially.

  And when using for this service, the sound radar function mentioned above is used and it is made to guide whether the place which should go next is near or far according to a relevance rate. Thereby, it can be made easy to guide to the destination. When the present system and method are used for the route guidance service in this way, for example, even at stations such as Shinjuku Station, Shibuya Station, and Tokyo Station that are complicated to transfer, and large stations that have been visited for the first time, they are output from that location. Based on the sound, it can be appropriately guided to the destination.

(Sound authentication service)
Next, a case where the acoustic signal detection system and method of each embodiment as described above is applied to an authentication service will be described as an example.

  This authentication service distributes the sound source template D1 to the user as a ticket and uses the sound source template D1 for authentication. The “sound source template” used in this service is, for example, the company's theme song when visiting the company, the support song of the team when watching sports, and the music of the artist as the sound source template when watching a concert. Become. An event that occurs is a ticket issue.

  Here, by playing music including a specific sound signal at the entrance or the like, an event occurs only on the user terminal 1 that has downloaded the sound source template D1, and a ticket can be displayed on the screen. Can be allowed.

  Further, when this service is implemented as a concert ticket, different part portions may be generated as sound source templates D11 to D13, and event information (ticket contents) having different contents may be associated with each sound source template D1.

  Thereby, for example, since a different event can be generated for each user, for example, seat information of a concert venue can be distributed to the user. Similarly to the above route guidance service, by using the sound radar function and map information in the venue, it is also possible to guide the way to the seats and facilities (toilet, shop, etc.) in the venue.

  In addition, by associating a gift-winning event with the sound source templates D11 to D13 generated from a predetermined part, when an artist plays the part part during a concert, a service is provided to notify a predetermined user of the gift-winning it can.

  When the artist's music is distributed as the sound source template D1, the music can be used as an incoming melody. This is beneficial for the user because the favorite artist's music can be acquired as an incoming melody during normal use. In this case, the association of the event identifier is canceled after use as a concert ticket. Further, the sound source template D1 itself is subjected to processing such as copy guard to enhance security.

(Override prevention service)
Next, a case where the acoustic signal detection system and method according to each of the above-described embodiments is applied to a road overpass prevention service will be described as an example.

  The overpass prevention service, for example, notifies the user on the train that he has arrived at the target station. The “sound source template” used in this service includes a station name, a sound at the time of departure, a sound when the door is opened, and an announcement sound in the car. At this time, when different sounds are used at each station, for example, a sound in front of one station may be recognized to notify that the next station is the destination. Further, when recognizing the station name, the sound source template D1 is created from the sound as the intonation music, not the voice recognition.

  In this case, for example, when riding on a train or bus, it is possible to recognize the sound output at the destination and notify the user that the user has arrived at the destination.

[Example of change]
The description of each embodiment described above is an example of the present invention. For this reason, the present invention is not limited to the above-described embodiment, and various modifications can be made according to the design and the like as long as they do not depart from the technical idea of the present invention. Below, the modification of this invention is demonstrated.

(Modification 1)
In the template comparison described above, a method of calculating the matching rate by comparing templates in which binarized acoustic signals are distributed was used. However, the present invention is not limited to this, and various comparison methods are used. Can be used. For example, difference data from the maximum volume may be used in the sound source template D1. Specifically, first, difference data from the maximum volume is calculated and stored in each block of the sound source template D1. At this time, the difference data is normalized and stored. Next, template matching is performed. Note that a cross-correlation coefficient is calculated in template matching.

  Then, in the region with the highest matching rate, the volume (Amplitude) difference data in the same frequency region as the template is extracted, and the volume (Amplitude) difference data is normalized. Next, the sound source template and the comparison destination template are compared. At this time, a portion (threshold value) where the difference data is greatly different is processed as noise. The template matching rate and the volume (Amplitude) difference matching rate are added and compared with a threshold value to determine the comparison result.

  Moreover, you may provide the function which adjusts the parameter of a template automatically with respect to embodiment mentioned above. In other words, by having this automatic adjustment function, the optimum parameter can be automatically and dynamically changed in an environment where the S / N ratio, which is the ratio of the signal amount (signal) and the noise amount (noise), is different. For example, it is possible to adapt optimal parameters for different noise levels in the outside environment and in the home environment. In addition, optimal parameter adjustment can be performed dynamically in real time.

  Hereinafter, the automatic parameter adjustment will be described. Here, a case where function learning is used and a case where dynamic parameters are used will be described.

-Parameter adjustment using machine learning First, after creating a template, prepare learning data. As the learning data, a mixture of a template sound source and noise and an arbitrary sound source that does not contain a template sound source are prepared. Here, a plurality of data with different noise levels are prepared as data mixed with sound.

  First, using the supervised machine learning function, the window width (block size), the number of template volumes (Amplitude), and the number of sound sources (Amplitude) of the sound sources to be compared are learned. At this time, if the learning does not converge for the high and low noise levels, the learning result is held individually. If machine learning does not converge for all learning data, the learning data is classified by the SN ratio, and learning is performed for each classification.

  In this way, the matching rate is held for all the learning data using the data learned for each classification. This data may be used for dynamic parameter adjustment. When learning converges for all teacher signals, the learning result is applied to all learned sound source templates. The generated learning data is held in either the user terminal or the server.

  The stored data includes a sound source template, an SN ratio, an average sound volume (Amplitude), a block size, the number of sound volumes (Amplitude) of templates used for comparison, and the number of sound volumes (Amplitude) of sound sources to be compared.

  When the learning function is used as described above, a single template may have a plurality of attribute information. The machine learning to be applied can use various functions such as an artificial neural network, a genetic algorithm, supervised learning such as subject learning.

  Next, application of dynamic parameters will be described. When applying the dynamic parameter, first, after starting the application, the intensity of sound coming from the microphone is set to have an average value for a certain fixed time (t_amp) (step 1). Then, parameter candidates are extracted from the average value of the volume (Amplitude), which is the attribute information of the template (step 2), and the matching rate is calculated for all the parameters that are candidates. Then, the matching rate and the threshold value are compared (step 3).

  Here, when it is determined that the matching is performed exceeding the threshold (Yes in Step 3), the matching rate is calculated using the parameters that are not candidates (Step 4). All parameters that are not candidates may be used, or may be one or may be extracted at random (step 5).

  If the relevance ratio using a parameter that is not a candidate is not far from the retained relevance ratio, it is determined that the relevance is satisfied (step 6). On the other hand, if the relevance ratios using parameters that are not candidates are far from each other, it is determined that they are not suitable (step 7).

  On the other hand, when it is determined that the threshold value has not been exceeded and it has not been matched (No in Step 3), parameter candidates are extracted again (Step 8). The parameter candidates may be based on the S / N ratio or may be performed for all parameters. In this parameter candidate extraction, the sound source determined to have been matched in Step 4 is held, and the learned parameter that performs the closest behavior is selected (Step 9).

  Then, based on the selected parameter, a matching rate is calculated (step 10), and the process is repeated until it is determined that it does not match as in step 7 above. After that, if it is determined that it is not suitable as in step 7, the process of step 8 is performed.

(Modification 2)
In each of the embodiments described above, the sound source template D1 generated from the specific sound signal is stored in the user terminal or server, and the evaluation sound signal (comparison target template D3) output from the sound reproducing means 4 is read and compared. The configuration. However, the present invention is not limited to this, and for example, sound data composed of a sound source template may be output from the speakers of the user terminals 1 and 1A, read by an external device, and compared. That is, in the present modification example, the user terminals 1 and 1A function as the sound reproducing unit 4, and the external device functions as the user terminals 1 and 1A.

  In this case, a template that is held by an external device and is a comparison destination is also a template generated from the specific acoustic signal. Event information is accumulated on the server side and associated with the identifier of the template.

  Then, the sound source template to be output is acquired from the server and held in the memory of the user terminal 1, and is output so as to be readable to an external device, whereby both templates match in the external device. If they match, an event associated with the template identifier can be generated.

  By using such a configuration according to the second modification, a concert ticket authentication service can be executed. That is, in this authentication service, a template generated from a specific acoustic signal is distributed as a ticket to the user terminals 1 and 1A. On the other hand, an authentication device equipped with a microphone capable of reading an acoustic signal is installed at the entrance of the venue. In this authentication apparatus, a template to be compared may be stored in the internal memory, and if a communication process is possible, a template to be compared may be stored in a server arranged on the communication network 5. Good.

  Then, with the user terminal 1 approaching the authentication device installed at the entrance, the sound based on the template is output from the speaker, so that whether or not the template matches the template stored on the authentication side. Authentication processing is performed by judging the above.

    By using such an authentication service, the frequency is cut when generating the sound source template at this time, so the sound is different from the original sound. But security can be improved. In addition, since it is not static data such as a bar code or QR code (registered trademark) but dynamic data, it is difficult to be true and the security level is high.

  In this modification, sound based on the template is output from the speaker of the user terminal 1 or 1A. For example, as shown in FIGS. 10A and 10B, the display based on the template is displayed on the display unit 13a. While displaying the screen, the authentication apparatus may be provided with an imaging unit such as a camera, and the authentication process may be performed using an image. In this case, since the process can be performed without sound, the authentication process can be executed even in a situation where no sound can be output to the outside.

(Other changes)
In the second embodiment, each function related to acoustic signal detection is distributed and arranged in a plurality of server devices. However, the present invention is not limited to this, and the server device as a single unit, A function may be provided, or a larger number of server apparatuses specialized for the function of each module may be arranged in a distributed manner so that each apparatus cooperates by communication.

A ... Area D1, D11-D13 ... Sound source template D2 ... Event information D3 ... Comparison template 1,1A ... User terminal 2 ... Management server 3 ... Radio base station 4 ... Sound reproduction means 5 ... Communication network 6 ... Control management server 7 ... Template comparison server 11 ... Wireless interface 12 ... Input interface 12a ... Microphone 13 ... Output interface 13a ... Display unit 14 ... Sound source template receiving unit 15 ... Memory 16 ... Sound acquisition unit 17, 18 ... Application execution unit 21 ... Communication unit 22 ... Control unit 23 ... User information database 24 ... Event database 25 ... Sound source template database 41 ... Control unit 42 ... Speakers 43a, 43b ... Buffer 61 ... Communication unit 62 ... User registration unit 63 ... Evaluation target conversion unit 64 ... A Vent occurrence management unit 64a ... relevance ratio display control unit 65 ... event generation unit 66 ... event database 67 ... user information database 68 ... sound source template generation unit 69 ... distribution management unit 69a ... sound source template distribution unit 71 ... communication unit 72 ... sound source template Acquisition unit 73 ... sound source template database 74 ... template comparison unit 75 ... precision ratio output part 151 ... event database 171 ... evaluation object conversion part 172 ... template comparison part 173 ... precision ratio output part 174 ... event occurrence management part 174a ... precision ratio display Unit 175 ... sound source template generation unit 176 ... reception management unit 181 ... sound transmission unit 182 ... event execution unit 221 ... sound source data acquisition unit 222 ... sound source template generation unit 223 ... event generation unit 224 ... distribution management unit 224a ... Plate delivery unit 225 ... user registration unit

Claims (15)

An acoustic signal detection system for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated,
The evaluation acoustic signal is acquired, and for the acquired evaluation acoustic signal, the presence or absence of the acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are obtained for each scale. An evaluation target conversion unit that generates a comparison target template that is quantified and recorded in time series, and
Based on the specific acoustic signal , the presence or absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are binarized for each scale and recorded in time series. A sound source template generator for generating a sound source template,
A sound source template acquisition unit for acquiring the sound source template;
A template comparison unit that compares the comparison target template with the sound source template and calculates a binarized distribution ratio of the acoustic signal;
A precision output unit that outputs a comparison result signal corresponding to the precision calculated by the template comparison unit ;
An event database for storing a template identifier for identifying a sound source template generated by the sound source template generation unit and an event identifier for identifying an event to be generated;
Based on the comparison result signal output from the precision ratio output unit, the event database is checked for the template identifier of the sound source template related to the comparison result signal, the event to be generated is selected, and the event that generates the selected event An outbreak management section and <br/>
When generating the sound source template for the specific sound signal, the sound source template generation unit detects a signal intensity at each scale and determines whether the detected sound signal exceeds a predetermined threshold value. extract the presence or absence of,
The acoustic signal detection system , wherein the threshold value is set to a different value depending on a type of the event to be generated .   The acoustic signal detection system according to claim 1, wherein the sound source template generation unit detects the presence or absence of a sound of each scale after removing a harmonic component when generating the sound source template.   The said template comparison part detects the presence or absence of each gradation sound, without including a harmonic component, at the time of a comparison with the said sound source template, without including a harmonic component. The described acoustic signal detection system. An acoustic signal detection server for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated,
Through the communication network, the evaluation acoustic signal is acquired, and for the acquired evaluation acoustic signal, the presence or absence of the acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are An evaluation target conversion unit that generates a comparison template that is binarized for each scale and recorded in time series;
Based on the specific acoustic signal, the presence or absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are binarized for each scale and recorded in time series. A sound source template generator for generating a sound source template,
A sound source template acquisition unit for acquiring the sound source template;
A template comparison unit that compares the comparison target template with the sound source template and calculates a binarized distribution ratio of the acoustic signal;
A matching rate output unit that sends a comparison result signal according to the matching rate calculated by the template comparison unit to the communication network ;
An event database for storing a template identifier for identifying a sound source template generated by the sound source template generation unit and an event identifier for identifying an event to be generated;
Based on the comparison result signal output from the precision ratio output unit, the event database is checked for the template identifier of the sound source template related to the comparison result signal, the event to be generated is selected, and the event that generates the selected event An outbreak management section and <br/>
When generating the sound source template for the specific sound signal, the sound source template generation unit detects a signal intensity at each scale and determines whether the detected sound signal exceeds a predetermined threshold value. extract the presence or absence of,
The acoustic signal detection server , wherein the threshold value is set to a different value depending on a type of the event to be generated . The acoustic signal detection server according to claim 4 , wherein the sound source template generation unit detects the presence or absence of a sound of each scale after removing a harmonic component when generating the sound source template. 6. The template comparison unit according to claim 4 or 5 , wherein, when comparing with the sound source template, the presence / absence of each tone is detected without removing the harmonic component and including the harmonic component. The acoustic signal detection server described. An acoustic signal detection device for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated,
A sound acquisition unit that acquires sound and outputs the sound as the evaluation sound signal;
For the evaluation acoustic signal output from the acoustic acquisition unit, the presence or absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are binarized for each scale. An evaluation target conversion unit that generates a comparison destination template recorded in time series, and
Based on the specific acoustic signal, the presence or absence of an acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are binarized for each scale and recorded in time series. A sound source template generator for generating a sound source template,
A sound source template acquisition unit for acquiring the sound source template;
A template comparison unit that compares the comparison target template with the sound source template and calculates a binarized distribution ratio of the acoustic signal;
A precision output unit that outputs a comparison result signal corresponding to the precision calculated by the template comparison unit ;
An event database for storing a template identifier for identifying a sound source template generated by the sound source template generation unit and an event identifier for identifying an event to be generated;
Based on the comparison result signal output from the precision ratio output unit, the event database is checked for the template identifier of the sound source template related to the comparison result signal, the event to be generated is selected, and the event that generates the selected event An outbreak management section and <br/>
When generating the sound source template for the specific sound signal, the sound source template generation unit detects a signal intensity at each scale and determines whether the detected sound signal exceeds a predetermined threshold value. extract the presence or absence of,
The acoustic signal detection apparatus , wherein the threshold value is set to a different value depending on a type of the event to be generated . The acoustic signal detection device according to claim 7 , wherein the sound source template generation unit detects the presence or absence of a sound of each scale after removing a harmonic component when generating the sound source template. 9. The template comparison unit according to claim 7 or 8 , wherein, when comparing with the sound source template, the presence / absence of each tone is detected without removing the harmonic component and including the harmonic component. The acoustic signal detection device described. An acoustic signal detection program for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated by a computer, the computer,
The evaluation acoustic signal is acquired, and for the acquired evaluation acoustic signal, the presence or absence of the acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are obtained for each scale. An evaluation target conversion step for generating a comparison target template that has been quantified and recorded in time series, and
For the specific sound signal, the presence or absence of the sound signal in a predetermined frequency band is extracted based on the scale average rate, and the sound signal distribution and its temporal change are binarized for each scale and recorded in time series A template comparison step of generating and obtaining a template, comparing the comparison target template with the sound source template, and calculating a matching ratio of the binarized distribution of the acoustic signal;
A precision ratio output step of outputting a comparison result signal corresponding to the precision ratio calculated by the template comparison step ;
A database control step of associating a template identifier for identifying the sound source template generated in the template comparison step with an event identifier for identifying an event to be generated in an event database;
Based on the comparison result signal output in the precision ratio output step, the event database is checked with respect to the template identifier of the sound source template related to the comparison result signal, an event to be generated is selected, and an event that generates the selected event An outbreak management step and a process including <br/>
In the template comparison step, when the sound source template is generated for the specific sound signal, the signal intensity in each scale is detected, and the sound signal is determined based on whether the detected sound signal exceeds a predetermined threshold value. extract the presence or absence of,
The acoustic signal detection program , wherein the threshold value is set to a different value depending on a type of the event to be generated . 11. The acoustic signal detection program according to claim 10 , wherein in the template comparison step, the presence or absence of a sound of each scale is detected after the harmonic component is removed at the time of generating the sound source template. In the template comparison step, in the time of comparison between the sound source template, without removing the harmonic components, to claim 10 or 11, characterized in that detecting the presence or absence of each floor articulatory while including harmonic components The described acoustic signal detection program. An acoustic signal detection method for detecting a specific acoustic signal included in an evaluation acoustic signal to be evaluated,
The evaluation acoustic signal is acquired, and for the acquired evaluation acoustic signal, the presence or absence of the acoustic signal in a predetermined frequency band is extracted based on the scale average rate, and the distribution of the acoustic signal and its temporal change are obtained for each scale. A process for converting the evaluation target to generate a comparison target template that is recorded in chronological order,
For the specific sound signal, the presence or absence of the sound signal in a predetermined frequency band is extracted based on the scale average rate, and the sound signal distribution and its temporal change are binarized for each scale and recorded in time series A template comparison process that generates and obtains a template, compares the comparison target template with the sound source template, and calculates a matching ratio of the binarized distribution of the acoustic signal;
Relevance ratio output processing for outputting a comparison result signal corresponding to the relevance ratio calculated by the template comparison processing ;
A database control process for associating a template identifier for identifying a sound source template generated in the template comparison process with an event identifier for identifying an event to be generated in an event database;
Based on the comparison result signal output from the precision ratio output process, the event database is checked with respect to the template identifier of the sound source template related to the comparison result signal, the event to be generated is selected, and the selected event is generated Including outbreak management processing and <br/>
In the template comparison process, when the sound source template is generated for the specific sound signal, the signal intensity in each scale is detected, and the sound signal is determined based on whether the detected sound signal exceeds a predetermined threshold value. extract the presence or absence of,
The acoustic signal detection method , wherein the threshold value is set to a different value depending on a type of the event to be generated . The acoustic signal detection method according to claim 13 , wherein in the template comparison process, at the time of generating a sound source template, the presence or absence of a sound of each scale is detected after removing harmonic components. 15. The template comparison process according to claim 13 or 14 , wherein, when comparing with the sound source template, the presence or absence of each tone is detected while the harmonic component is included without removing the harmonic component. The acoustic signal detection method as described.