JP2020088672A - Listener authentication system - Google Patents

Abstract

To provide a technique capable of authenticating, in real time, a listener who is listening to the radio.SOLUTION: A listener authentication system is composed of: a server 10; and a portable phone 20 that can communicate with the server 10 and is operated by a listener of a radio program. The listener raises the portable phone 20 over a radio 30 at hand to capture sound of an on-air radio program output from its speaker. Listening sound feature data is extracted from a listening sound signal of the radio captured by the portable phone 20 and is transmitted together with listening time information to the server 10. The server 10 extracts broadcast sound feature data from a broadcast sound signal of the radio program and compares listening sound feature data transmitted from the portable phone 20 of the listener with time information in the respective data. The server 10 includes authentication means 106 for authenticating that the listener has listened to the radio program on the condition that agreement is recognized between the sound feature data and the time information.SELECTED DRAWING: Figure 1

Description

The present invention relates to a technique for grasping a listening situation of a radio program.

It has been reported that advertising revenue exceeds 80% of the total in the broadcasting business such as television and radio. Therefore, it can be said that the most important issue for the broadcaster in terms of management is to acquire more viewers (listeners) and increase the chances of viewing the program. For example, in the case of radio, many people listen to it on car radios or portable radios, and in order to calculate accurate listening ratios, currently, the diary format is used to ask the surveyee to write the program actually listened to from the radio on a sheet. Surveys are the mainstream.

This kind of diary-style hearing rate survey is very complicated for listeners (subjects to be surveyed), and it also takes time to obtain a final tabulation result. Moreover, since the accuracy of the survey largely depends on the accuracy of the response of the listener, there is a high possibility that a human error in the response is included. In relation to such a problem, various techniques have been proposed for automatically identifying a broadcasting station, a broadcasting program, contents, etc. by analyzing a radio sound during broadcasting.

For example, the investigation device disclosed in Patent Document 1 sequentially scans the selection frequency of the reference radio with the sound output from the speaker of the car radio and the sound reproduced originally by the reference radio installed in the vehicle. By comparison, the listening station is specified. The measurement data in which the listening station is specified for each vehicle is transmitted to the survey center by wireless communication, and the listening rate is calculated in real time. In the survey device mounted on each vehicle, the frequency spectrum signals of radio voices are compared with each other. Regarding comparison or identification of broadcast program contents, for example, Patent Document 2 suggests that samples, hashes, and fingerprints selected from audio or video signals can be used as representative values for identification.

Further, for example, according to the car radio listening situation investigation method of Patent Document 3, the feature quantity extracted from the received signal of the car radio is transmitted to the investigation center, and the characteristics of the radio broadcast master data recorded in the database on the investigation center side. It is possible to investigate the listening situation of each broadcasting station by comparing the amount with the feature amount extracted by the car radio of each vehicle. According to Patent Document 3, the detection signal selected by the radio is divided into a plurality of pass band-specific signals by a digital filter, and the power vector obtained by logarithmically converting the average amplitude of each pass band-specific signal is the characteristic amount. It has been suggested that

JP, 7-327017, A Japanese Patent Publication No. 2017-535214 JP, 11-275032, A

The above-mentioned conventional techniques judge the listening situation by which radio station or radio program the operating radio is selecting, and the listener surely listens to the broadcast program. It does not prove anything. Conversely, if a mechanism for authenticating listeners who are listening to radio programs in real time can be provided, not only accurate calculation of the listening ratio but also the possibility of expanding interactive program production with listener participation, for example, will be expanded. Can be expected.

The present invention has been made in view of these circumstances, and an object thereof is to provide a technique capable of authenticating a listener listening to a radio in real time.

In order to solve the above-mentioned problems, the present invention provides a demodulation means for demodulating a sound signal (this is called a broadcast sound signal) from a broadcast radio signal of a radio program being broadcast, and a listener holding the radio from a speaker of the radio. A microphone that captures the sound of the output radio program and converts it into a sound signal (referred to as a listening sound signal), a comparing unit that compares the broadcast sound signal and the listening sound signal, and the broadcast sound by the comparing unit. A listener authentication system, comprising: an authenticator that authenticates that the listener has listened to the radio program based on the fact that the signal and the listening sound signal are in agreement.

The listener authentication system includes a first feature extraction unit that extracts sound feature data (referred to as broadcast sound feature data) from the broadcast sound signal by a predetermined method, and a sound feature data (from the listening sound signal by the predetermined method). It is preferable to further include a second feature extracting means for extracting this (referred to as listening sound feature data), and the comparing means to compare the broadcast sound feature data and the listening sound feature data.

Further, it is preferable that the listener authentication system further includes a time stamp unit that associates time information when the listener listens to the radio program with the listening sound signal.

In addition, the listener authentication system is determined that the broadcast sound feature data and the listening sound feature data match, and the broadcast time and the listening sound feature data of the broadcast sound signal at a time position where these sound feature data match. It is preferable that the authentication unit authenticates the listener when it is determined that the listening times shown are the same.

Further, in the listener authentication system, it is preferable that the predetermined method is a method of analyzing a frequency spectrum of sound.

Further, in the listener authentication system, it is preferable that each sound feature data of the broadcast sound signal and the listening sound signal can be expressed as a spectrogram indicating a temporal change of a power spectrum in a frequency domain of sound.

Further, in the listener authentication system, each sound feature data of the broadcast sound signal and the listening sound signal is based on array data of a vector including a pair of peak frequencies of a power spectrum in a sound frequency domain and their distances as elements. It is preferably one.

Further, in the listener authentication system, it is preferable that each sound feature data of the broadcast sound signal and the listening sound signal is data obtained by hashing the array data of the vector.

Further, in the listener authentication system, a server includes at least the first feature extracting means and the comparing means, and an information terminal which is operated by a listener and is accessible to the server includes at least the microphone and the second feature extracting means. It is preferable to provide.

Further, in the listener authentication system, a server includes at least the first feature extraction unit, and an information terminal accessible by the server operated by a listener includes at least the microphone, the second feature extraction unit, and the comparison unit. It may be provided.

Further, it is preferable that the listener authentication system further includes a database that records the program identification information of the radio program and the information of the listener who is authenticated to listen to the radio program in association with each other.

Further, in the listener authentication system, it is preferable that a listening authentication history of a radio program listened to by the listener is searchably recorded in the database based on identification information of the listener.

The present invention also includes a step of demodulating a broadcast sound signal from a broadcast radio wave signal of a radio program being broadcast, and a listener holding the information terminal over the radio so that the sound of the radio program actually output from the speaker of the radio. Based on the fact that the matching between the broadcast sound signal and the listening sound signal is recognized and the step of comparing the broadcast sound signal and the listening sound signal with each other is recognized. And a step of authenticating that the user has listened to the radio program.

Further, the present invention is an application program installed in an information terminal, which realizes at least the function of the second feature extraction means by a calculation process performed by a processor of the information terminal.

Further, the present invention is an application program installed in an information terminal, the application program realizing at least the functions of the second feature extraction means and the comparison means by arithmetic processing performed by a processor of the information terminal. Good.

According to the present invention, a listener who is listening to a predetermined radio program can be authenticated in real time. By using this system, the listening rate of a radio program can be calculated quickly and accurately without the listener taking time and effort. Also, listener authentication can be applied to various purposes such as applying for presents during radio programs and participating in events.

1 is a conceptual diagram illustrating a listener authentication system according to an exemplary embodiment of the present invention. It is a functional block diagram of the listener authentication system of FIG. It is a figure which illustrates the spectrogram obtained by spectrum-analyzing a certain voice. It is the figure which plotted the peak of a power spectrum based on the spectrogram of FIG. It is a figure for demonstrating the vector arrangement obtained by connecting the pair of peaks by the peak spectrogram of FIG. 6 is a flowchart illustrating a comparative authentication process of broadcast sound and listening sound. It is a figure which illustrates a broadcast sound and a listening sound which mutually correspond in a certain time position. It is a flow chart which illustrates the procedure which performs listener authentication using an automatic voice response device (IVR).

An embodiment of a listener authentication system according to the present invention will be described below with reference to the drawings. FIG. 1 is a conceptual diagram for explaining a listener authentication system according to an embodiment of the present invention. The listener authentication system is configured as a system in which a server 10, which is a cloud computer for providing this service, and a mobile phone 20, which is an information terminal operated by a listener of a radio 30, are communicatively connected. .. The listener's mobile phone 20 is connected to the server 10 via the Internet using its Wi-Fi connection function, for example.

In the cloud, a radio receiver 12 can be installed for acquiring the sound of a radio program that is the target of listener authentication (this is referred to as "broadcast sound" in this specification). Further, the cloud may include a recording device (not shown) that constantly records a radio program broadcast by each radio broadcasting station in the area. Then, the server 10 may reproduce the sound signal of the broadcast sound for listener authentication from the recording device. Further, the server 10 may obtain broadcast sound through digital audio stream data distributed from a radio broadcasting station. Further, the cloud may be provided with a database 11 for recording sound feature data of broadcast sounds, account registration information of listeners, listening authentication history information of radio programs listened to by listeners, and the like.

In order to authenticate the listener of the radio in real time, the present listener authentication system includes the above-described sound of the broadcast sound and the radio sound captured by the listener by holding the mobile phone 20 over the radio 30 at hand (this is referred to as “in this specification”). If there is a match between them, it is verified that the listener is listening to the radio program.
Here, in the "voice" of the radio, in addition to the voice in the normal sense spoken by the announcer, all the sounds from the radio such as commercials, music and sound effects are targeted. In the case of a music piece, a part of the performance is often repeated, and thus it may be excluded from real-time authentication targets.

A configuration example of the listener authentication system will be described in more detail. FIG. 2 is a functional block diagram of a listener authentication system according to one embodiment. In FIG. 2, the server 10 has a tuner 121 that tunes in a broadcast wave of a radio program to be authenticated and a wave detector (demodulates) a sound signal from the selected broadcast wave signal in order to obtain the above-described broadcast sound. And a demodulator 122. The tuner 121 and the detector (demodulator) 122 may be configured by the radio receiver 12 described above.

The server 10 also includes an A/D converter 101, a feature extraction unit 102, and a recording unit 103. The sound signal of the radio program to be authenticated is digitally converted by the A/D converter 101 at a sampling frequency of, for example, several tens kHz, and then the sound feature data is extracted by the feature extraction means 102. It is preferable that the feature extraction unit 102 extract the sound feature data by spectrally analyzing the time change of the frequency and the phase of the sound signal. By extracting and comparing the sound feature data from the radio voice, the transaction of the data is reduced, and the influence of the volume difference of the listening sound, noise, etc. is eliminated, and the comparison processing can be performed with high accuracy.

Here, an example of an analysis method in which the feature extraction unit 102 extracts sound feature data from radio sound will be specifically described. The feature extracting means 102 first applies a short time Fourier transform (STFT) to the digital sound signal to obtain a spectrogram showing the time change of the power spectrum in the frequency domain of the sound. The shift time of the window function (frame) can be, for example, about tens of milliseconds to hundreds of milliseconds. Here, FIG. 3 illustrates a spectrogram obtained by spectrally analyzing a certain voice. In FIG. 3, time is plotted on the horizontal axis and frequency is plotted on the vertical axis, and the power of the spectrum is shown in shades (gray scale).

From this spectrogram, in order to remove noise and further downsize the processed data, a peak (local maximum value) of the power spectrum may be obtained and the peak may be plotted in the spectrogram (see FIG. 4). For example, by applying a high-pass filter to the amplitude and performing logarithmic conversion, it is possible to obtain the peak of the power spectrum in which the high amplitude is emphasized.

The spectrogram showing the time change of the frequency of the sound as illustrated in FIG. 3 and FIG. 4 is a so-called fingerprint showing the feature of the voice, and the feature is imaged. Therefore, it is possible to compare the broadcast sound and the listening sound with their respective spectrograms.

Further, the feature extracting means 102 may extract the array data of vectors including the pair of peak frequencies and their distances as elements from the above-mentioned spectrogram as the sound feature data in the present invention. Specifically, as shown in FIG. 5, and arbitrarily selected anchor point A _{f (W)} at a certain time W, the anchor point A _{f (W)} a plurality of peaks B _f1 in a predetermined area as viewed from An array MB _f (W) of vectors connecting (W), B _f2 (W), B _f3 (W),... Is obtained.

MB _f (W)=[P _f (W), ΔP _fn (W), ΔT _fn (W)] (1)

Here, the vector element P _f (W) is the frequency of the anchor point A _f (W), and ΔP _fn (W) is the n-th peak B _fn (W) within the predetermined area and the anchor point A _f (W). The frequency difference, ΔT _fn (W), is the time difference between the nth peak B _fn (W) and the anchor point A _f (W).

A plurality of anchor points are selected at the time W, but each has a peak at a position closest to each frequency (for example, f=1 kHz, 2 kHz, 3 kHz,...) Arbitrarily determined in advance in the audible region. The points may be used (for example, A _{1 kHz} (W), A _{2 kHz} (W), A _{3 kHz} (W),... ).

Then, the feature extraction means 102 causes the vector arrays MB _f (1), MB _f (2), and MB _f (3) obtained from the spectrogram every predetermined time W (W=1, 2, 3,... ). ),... Can be used as sound feature data. The interval of the predetermined time W is preferably the window function shift time at the time of the above short-time Fourier transform, but may be longer than that.
Further, it is preferable that the feature extracting unit 102 hash the vector array MB _f (W) of each time W into fixed-length data in order to facilitate comparison between the sound feature data.

Returning to FIG. 2, in the recording unit 103 (or the database 11), the sound feature data of the broadcast sound (hereinafter, referred to as “broadcast sound feature data”) extracted by the feature extracting unit 102 is recorded at the recording time of the broadcast sound. It is recorded together with the information of (broadcast time).

The mobile phone 20 operated by the listener includes a microphone 201, an A/D converter 202, a feature extraction unit 203, and a transmission unit 206. The listener holds the mobile phone 20 over the radio 30 while listening to a radio program requiring authentication on the radio 30, and outputs the sound actually output from the speaker of the radio 30 to the microphone 201 of the mobile phone 20 for several seconds. take in.

The microphone 201 converts the received listening sound into a sound signal (listening sound signal). The A/D converter 202 digitally converts the listening sound signal at a sampling frequency of, for example, several tens of kHz. The feature extraction unit 203 extracts sound feature data (hereinafter referred to as “listening sound feature data”) from the listening sound signal by the same analysis method as the above-described feature extraction unit 102 of the server 10.

The mobile phone 20 also includes time stamp means 204 and transmission means 206. The time stamp means 204 periodically obtains accurate time information from the base station of the communication carrier and listens to the time information (while listening to the radio sound while the listener holds the mobile phone 20 over the radio 30). Time stamp information) is associated with the listening sound feature data. Then, the transmitting unit 206 transmits the listening sound feature data, the listening time information, and the listener ID to the server 10 in succession to the extraction processing by the feature extracting unit 203.

The server 10 also includes a receiving unit 104, a comparing unit 105, and an authenticating unit 106. The receiving unit 105 receives the listening sound characteristic data transmitted from the mobile phone 20. The comparing means 105 compares the broadcast sound signal (broadcast sound feature data in this embodiment) recorded in the recording means 103 and the listening sound signal (listening sound feature data in this embodiment) transmitted from the mobile phone 20. Compare.

As shown in the flowchart of FIG. 6, the comparison unit 105 determines whether or not the listening time indicated by the listening sound characteristic data is in the authentication target time zone of the target radio program (step S1). When the listening time is in the authentication target time zone (step S1; YES), the comparing unit 105 sets the listening time to the broadcast sound characteristic data of the broadcast sound broadcast in the authentication target time zone of the radio program. The listening sound characteristic data is swept on the time axis (step S2). As shown in FIG. 7, the comparing means 105 determines that the similarity index when comparing the broadcast sound feature data and the listening sound feature data of the radio program broadcast during the authentication time period is equal to or greater than a predetermined threshold value. Then, it is determined that these sound feature data match (step S3; YES).

The comparison means 105 further determines whether or not the difference between the broadcast time of the broadcast sound signal and the listening time indicated by the listening sound feature data at a time position where the sound feature data match each other is equal to or less than a certain threshold value. Yes (step S4).

Finally, when the broadcast sound feature data and the listening sound feature data match and the times (time positions) of them also match (step S4; YES), the authentication means 106 causes the listener to play the radio program. It is verified that the user has certainly listened (step S5).
In addition, when the listening time of the listening sound is outside the authentication target time period (step S1; NO), when the broadcast sound feature data and the listening sound feature data do not match (step S3; NO), and if the broadcast sound and the listening sound However, if the times (time positions) do not match (step S4; NO), the authentication means 106 does not authenticate the listener (step S6).

In the listener authentication system of this embodiment, the following patterns are examples of the timing at which the listener captures the listening sound.
(1) Optional time zone:
When the authentication target time period extends over the entire radio program, the listener may capture the listening sound at any time during the time period during which the radio program is being broadcast.
(2) Fixed time period:
The radio program may be fixed in a specific time zone, for example, 5 minutes immediately after the start. By fixing the authentication target time period to a few minutes after broadcasting, the accuracy of the collected listening rate can be improved. Further, for example, if authentication target time zones are provided before and after a commercial, the number of listening to the commercial can be expected to be further increased.
(3) Designated time zone:
During the live broadcast of the radio program, the announcer may specify, for example, "I will perform listener authentication for 5 minutes from now. Hold your mobile phone over the radio." In this case, since the listener cannot predict the authentication target time zone, it is possible to keep the listener's interest in the radio program for a long time.
Further, for example, in the case of applying for a gift lottery, the authentication target time zone may be limited to only the time zone in which the guidance is broadcast.

Further, the authenticating means 106 may determine the authentication not only by the coincidence of the voices during one time period during the broadcast of the program but also when the coincidence is determined by a combination of a plurality of times. For example, only if the same listener is checked for the second time during the time period specified in the program after determining the matching of the listening sound once every five minutes immediately after the start of the radio program, The authentication of may be confirmed.
Alternatively, when you give a lottery notice of a present on a radio program, you can authenticate the first listener to qualify for the application, and during the program, you can also say "I will give a lottery for presents. Hold your mobile phone over the radio. Please be acknowledged in the second time zone when it was announced that you can participate in the present lottery.

In the cloud database 11, account registration information of each listener who uses this system, listening authentication history information of each listener, and the like are recorded. The listener can register information such as listener ID (or telephone number), login password, area, sex, age, affiliation, and hobby in the database 11 as account registration information. Further, the listening authentication history information is associated with the program identification information of the radio program and the listener ID that has been authenticated as listening to the radio program. For example, the radio program may be divided into blocks for every 15 minutes or corners, and the listener authentication history may be recorded for each block.

In the embodiment of FIG. 2, the A/D converter 202, the feature extraction means 203 and the time stamp means 204 are calculated by the microprocessor of the mobile phone 20 according to a dedicated application program (app) installed from the server 10, for example. It is realized by executing processing. In addition to the authentication function described above, this application program can provide a new account registration of a listener who uses the service of this system, login/logout to the server 10, and a human interface function when making an authentication request. it can.

In addition, listener authentication can be performed using an automatic voice response device (IVR) even with a so-called feature phone or desk phone in which the application cannot be installed. An example of the procedure is shown in FIG. In the case of this example, the listener can use the telephone number of the mobile phone 20 as an ID for personal identification, and the listener information associated with the ID can be registered in the database 11 on the server 10 side in advance. Good.

In the above-described embodiment of the present invention, in order to reduce the data transaction on the server 10 side, the sound feature data is extracted from the radio sound and compared. However, radio voices may be directly compared, or voice amplitudes, frequencies, and spectrograms or fingerprints that are sound characteristic data may be appropriately combined and compared.

Further, in order to further reduce the concentration of access to the server 10 and the processing load, the following modifications and measures may be taken.
(1) For example, the voice comparison for listener authentication may be performed by the process on the client side (mobile phone 20) instead of the server 10 side.
(2) Listening sound data transmitted from the mobile phones 10 of a large number of listeners is temporarily buffered in a storage device on the cloud, and the comparison process on the server 10 side is serialized and sequentially performed. May be.
(3) The processing scale may be expanded by clustering the server 10 side.

With the system configuration described above, the server 10 can authenticate the listener listening to the target radio program in real time. Since this system directly picks up the sound of the actual radio and compares and authenticates the data of the listening time in real time, it is possible to obtain extremely high accuracy in listening verification. Further, for the listener, it suffices to simply "hold" the mobile phone 20 over the radio 30, and the burden can be significantly reduced compared to the conventional diary survey.

In addition, the server 10 can calculate the listening rate of the radio program immediately and in detail by statistically collecting the number of authenticated listeners. Further, for example, by referring to the account registration information of the database 11, it is possible to determine the listening rate by region, sex, or age of the listener. Further, when the mobile phone 10 can acquire more accurate position information by the position information system (GPS or the like), it is possible to add more accurate listening position information by transmitting the information to the server 10. it can. The radio broadcasting station side can analyze such detailed listener listening information and lead to more attractive program production and acquisition of advertising sponsors.

Further, for example, if the authentication is performed before and after the commercial, the effect of increasing the number of listening to the commercial can be expected.

In addition, application qualifications such as present application and participation in events can be limited to only authenticated listeners. From the server 10, listener information including the ID of the authenticated listener and information including the authenticated program and broadcast time are transmitted to the radio broadcasting station. By requesting the action of "hovering" on the radio in real time, it is possible to create a sense of unity between the radio station and the listeners. On the listener's side, application for an event can be done only by "over-the-head authentication", which can save the trouble. Furthermore, the listener can appeal to the radio station or the sponsor that he/she is actually listening to the radio program in an objective manner.

Further, in the database 11, the listening authentication history of the radio program listened to by the listener is recorded in a searchable manner based on the listener ID. Further, in this system, time stamp information indicating the time of authentication is also recorded in the listening authentication history information. Based on the listening history of the program content and the like retrieved from the time stamp information, a recommended advertisement or the like that matches the taste of the listener can be broadcast.

10 server 11 database 12 radio receiver 20 mobile phone (information terminal)
30 radio 101 A/D converter 102 feature extraction means 103 recording means 104 receiving means 105 comparing means 106 authenticating means 121 tuner 122 detector (demodulator)
201 Microphone 202 A/D converter 203 Feature extraction means 204 Time stamp means 206 Transmission means

Claims (15)

Demodulation means for demodulating a sound signal (this is called a broadcast sound signal) from a broadcast radio signal of a radio program being broadcast,
A microphone that captures the sound of the radio program output from a speaker of the radio by a listener over the radio, and converts the sound signal into a sound signal (referred to as a listening sound signal).
Comparing means for comparing the broadcast sound signal and the listening sound signal,
A listener authentication system, comprising: an authentication unit that authenticates that the listener has listened to the radio program based on the matching between the broadcast sound signal and the listening sound signal recognized by the comparison unit. First feature extraction means for extracting sound feature data (referred to as broadcast sound feature data) from the broadcast sound signal by a predetermined method; and sound feature data (this is a sound feature) from the listening sound signal by the predetermined method. Second feature extracting means for extracting
The listener authentication system according to claim 1, wherein the comparison unit compares the broadcast sound feature data and the listening sound feature data. The listener authentication system according to claim 2, further comprising time stamp means for associating time information when a listener listens to a radio program with the listening sound feature data. It is determined that the broadcast sound feature data and the listening sound feature data match, and it is determined that the broadcast time of the broadcast sound signal and the listening time indicated by the listening sound feature data at the time position where these sound feature data match. 4. The listener authentication system according to claim 3, wherein the authentication unit authenticates the listener when being performed. The listener authentication system according to claim 2, wherein the predetermined method is a method of analyzing a frequency spectrum of sound. The listener authentication system according to claim 5, wherein each sound feature data of the broadcast sound signal and the listening sound signal can be expressed as a spectrogram indicating a temporal change of a power spectrum in a frequency domain of sound. 7. The sound feature data of each of the broadcast sound signal and the listening sound signal is based on array data of a vector including pairs of peak frequencies of a power spectrum in a frequency domain of sound and their distances as elements. Listener authentication system described in. The listener authentication system according to claim 7, wherein each sound feature data of the broadcast sound signal and the listening sound signal is data obtained by hashing the array data of the vector. A server comprises at least the first feature extraction means and the comparison means,
The listener authentication system according to any one of claims 2 to 8, wherein an information terminal operated by a listener and capable of accessing the server includes at least the microphone and the second feature extraction means. The server comprises at least the first feature extraction means,
The listener authentication system according to any one of claims 2 to 8, wherein an information terminal that is operated by a listener and that can access the server includes at least the microphone, the second feature extraction unit, and the comparison unit. 9. The listener authentication system according to claim 1, further comprising a database that records the program identification information of the radio program and the information of the listener authenticated to listen to the radio program in association with each other. The listener authentication system according to claim 11, wherein in the database, a listening authentication history of a radio program listened to by the listener is recorded in a searchable manner based on identification information of the listener. A step of demodulating a broadcast sound signal from a broadcast radio signal of a radio program being broadcast,
A listener holding the information terminal over the radio, capturing the sound of the radio program actually output from the speaker of the radio, and converting the sound into a listening sound signal;
Comparing the broadcast sound signal and the listening sound signal;
Listener authentication method, including the step of authenticating that the listener has listened to the radio program based on the agreement between the broadcast sound signal and the listening sound signal. An application program installed in an information terminal, which realizes at least the function of the second feature extraction means according to any one of claims 2 to 8 by arithmetic processing performed by a processor of the information terminal. An application program installed in an information terminal, wherein at least the functions of the second feature extraction means and the comparison means according to any one of claims 2 to 8 are realized by arithmetic processing performed by a processor of the information terminal. Application program.