JP2016121957A - Target sound section determination device, target sound section determination method, and target sound section determination program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a target sound section determination device capable of correctly determining a section that a target sound has when an allophone including only one or a small number of frequency components is the target sound.SOLUTION: A target sound section determination device has: directional forming means 12, 13 of forming a plurality of directional signals having different predetermined directions including dead angles from input sound signals S1(n), S2(n); coherence coefficient calculation means 14 of obtaining a coherence coefficient using the plurality of generated directional signals; coherence coefficient feature quantity calculation means 15 of regarding the coherence coefficient as a time change signal and obtaining a coherence coefficient feature quantity representing the frequency and magnitude of change in inclination direction of a signal waveform thereof; and target sound section determination means 16 of determining whether a section of an input sound signal is the target sound section based upon a range of the coherence coefficient and a range of the coherence coefficient feature quantity.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a target sound segment determination device, a target sound segment determination method, and a target sound segment determination program, and is applied to, for example, a case where a segment of a target sound including only one or a small number of frequency components is determined from an input sound signal. It is possible.

  Patent Document 1 has already proposed a technique for detecting a section (target sound section) including a target sound coming from the front based on the arrival direction of an input sound signal.

  By the way, in recent years, the sound source separation processing has come to be used for processing of sound other than sound such as sound generated by a machine for inspection of mechanical equipment. The sound emitted from the machine (mechanical sound) includes, for example, only one or a small number of frequency components (a signal including one frequency component is a sine wave signal having that frequency, and in this specification, In other words, it is a steady sound in which energy is concentrated on a specific frequency component.

  In the sound source separation process, a technique capable of accurately detecting a target sound section including a special sound such as a mechanical sound has been demanded.

JP 2013-1206026 A JP 2014-106337 A

  However, for example, when a sine wave abnormal noise comes from the front, the detection method of the target speech section described in Patent Document 1 is not the target sound even though it arrives from the front. There is a risk of erroneously determining it as a sound (hereinafter referred to as a non-target sound).

  In the method of detecting the target speech section of Patent Document 1, the power and the degree of correlation in each frequency component of the two processed signals that have given specific directivity to the signals captured by the two microphones are reflected. The coherence coefficient coef (f, K) for each frequency bin shown in Equation (3) of Patent Document 1 and the coherence COH (K) shown in Equation (4) are used (f is an index representing a frequency bin) , K is an index representing the input frame). The cause of misjudgment of sinusoidal abnormal noise is that the components included in the input sound signal are concentrated at a specific frequency, so the coherence coefficient coef (f, K) in the frequency bin of the frequency of the sine wave has a large value, This is because other frequency bins have a minute value, so that the coherence COH (K) obtained by averaging the coherence coefficient coef in the entire band becomes a small value and does not reach the determination threshold Θ of the target sound section.

  Therefore, there is a need for a target sound section determination device, a target sound section determination method, and a target sound section determination program that can correctly determine a section of the target sound when an abnormal sound including only one or a small number of frequency components is the target sound. Yes.

  A first aspect of the present invention is a target sound section determination device that determines a target sound section from a section of an input sound signal using a sound including one or a small number of frequency components as a target sound, and (1) an input sound signal Directivity forming means for forming a plurality of directional signals having a directional characteristic having a dead angle in a predetermined azimuth by performing a delay subtraction process, wherein the directional signals are different from each other in a predetermined azimuth having a blind angle; (2) Coherence coefficient calculation means for obtaining a coherence coefficient using a plurality of formed directional signals, (3) The number of times the obtained coherence coefficient is regarded as a time-varying signal, and the inclination direction of the signal waveform changes. A coherence coefficient feature quantity calculating means for obtaining a coherence coefficient feature quantity representing the magnitude; and (4) an input sound signal based on the coherence coefficient range and the coherence coefficient feature quantity range. Section is characterized by having a target sound period determining means for determining whether a target sound period.

  The second aspect of the present invention is a target sound section determination method for determining a target sound section from a section of an input sound signal using a sound including one or a small number of frequency components as a target sound, and (1) directivity forming means However, by performing a delay subtraction process on the input sound signal, a plurality of directivity signals having a directivity characteristic having a blind spot in a predetermined direction and a plurality of directivity signals having different blind headings are formed. (2) The coherence coefficient calculation means acquires a coherence coefficient using the formed directional signals, and (3) the coherence coefficient feature quantity calculation means regards the obtained coherence coefficient as a time-varying signal, A coherence coefficient feature amount representing the number of times and the magnitude of the change in the inclination direction of the signal waveform is acquired; and (4) the target sound section determination means determines the range of the coherence coefficient and the coherence coefficient feature amount. Based on the Nji, the section of the input sound signal and judging whether the target sound period.

  A third aspect of the present invention is a target sound section determination program applied to determine a target sound section from a section of an input sound signal using a sound including one or a small number of frequency components as a target sound. (1) By performing a delay subtraction process on the input sound signal, a plurality of directivity signals having a directivity characteristic having a dead angle in a predetermined direction and a plurality of directivity signals having different blind directions are formed. Directivity forming means, (2) coherence coefficient calculating means for obtaining a coherence coefficient using a plurality of formed directivity signals, and (3) taking the obtained coherence coefficient as a time-varying signal, A coherence coefficient feature quantity calculating means for obtaining a coherence coefficient feature quantity representing the number of times the inclination direction changes and its magnitude; and (4) the range of the coherence coefficient and the coherence coefficient characteristic. Based on the amount of range, the section of the input sound signal, characterized in that function as target sound period determining means for determining whether a target sound period.

  According to the present invention, when an abnormal sound including only one or a small number of frequency components is a target sound, a target sound section determination device, a target sound section determination method, and a target sound section determination program that can correctly determine a section of the target sound. Can be realized.

It is a block diagram which shows the structure of the target sound area determination apparatus which concerns on 1st Embodiment. It is explanatory drawing which shows the value of the coherence coefficient for every frequency component (frequency bin) in case the frequency component contained in the noise is one (when the noise is a sine wave noise). It is a block diagram which shows the detailed structure of the determination part which concerns on 1st Embodiment. It is a flowchart which shows operation | movement of the determination part which concerns on 1st Embodiment.

(A) First Embodiment Hereinafter, a first embodiment of a target sound section determination device, a target sound section determination method, and a target sound section determination program according to the present invention will be described with reference to the drawings.

  The target sound section determination apparatus according to the first embodiment uses an abnormal sound including only one or a small number of frequency components as a target sound, and determines a section of the target sound.

(A-1) Parameter to be applied and reason for application In the target sound section determination device according to the first embodiment, in addition to the coherence coefficient coef (f, K) used in the technique described in Patent Document 1, the coherence coefficient The modGI value modGI (f, K) shown in the equation (1) for coef (f, K) is applied. In equation (1), the coherence coefficient coef (f, K) is represented by s (K), and the modGI value modGI (f, K) is represented by modGI.

  The modGI value will be briefly described (refer to Patent Document 2 for details). modGI means a modified gradient index (hereinafter referred to as GI).

  For the GI before amendment, refer to the reference “Naofumi Aoki,” A Band Extension Technology for Narrow-Band Telephony Speech Based on Full Wave Rectification., IEICE Co. 729-731, 2010 ".

  GI is an index for measuring the number of times and the magnitude of the change in the inclination direction of the signal waveform. The GI is obtained by dividing the sum of absolute difference values of successive samples when the tilt direction is changed by the square root of the power of the frame. Therefore, the GI is likely to increase as the number of changes in inclination within one frame increases, and also increases as the amount of change when the inclination changes increases. From such a property, it can also be said that GI is directly connected to the amount of high frequency components included in the input waveform.

  However, since GI uses as a calculation element a variable ΔΨ (n) that takes only two values of 0 or 2 and has many jumps with large values in time series, the value increases or decreases irregularly. There is a characteristic ("value goes wild").

  In view of the fact that modGI has the property that the value of GI is rampant (has a jump with a large value), instead of GI, it has a high correlation with GI, and a new feature with stable changes that suppresses a large jump in value It is proposed as a quantity. modGI is the “power of the second-order difference of the calculation target signal” normalized by the “power of the calculation target signal” with respect to an arbitrary signal (a coherence coefficient in this application) of the feature quantity calculation target (this is a constant multiple). Are also included).

  Since modGI has a high correlation with GI, it functions as a stable index that measures the number and magnitude of changes in the slope direction of the signal waveform, and reflects the amount of high-frequency components contained in the input waveform. Function.

  Here, the behavior of the modGI value when a sine wave noise comes from the front is examined. When the modGI value modGI (f, K) is calculated for each frequency component with respect to the coherence coefficient coef (f, K), the coherence coefficient coef (f, K) is a substantially constant value (this value) in the frequency component of the frequency of the sine wave. Although the value is large), the modGI value modGI (f, K) is a minute value because the characteristic is close to a DC signal. In addition, since there is no input at a frequency component other than the frequency of the sine wave, the coherence coefficient coef (f, K) is very small and has a substantially constant value. The modGI value modGI (f, K) is a minute value.

  When the behavior of the coherence coefficient coef (f, K) and the modGI value modGI (f, K) is compared between the frequency component included in the incoming sine wave abnormal noise and the other frequency components, (a1) sine wave abnormal noise is compared. The coherence coefficient coef (f, K) is a large steady value and the modGI value modGI (f, K) is a small value. (A2) The coherence is a frequency component not included in the sinusoidal abnormal noise. It can be seen that there is a difference that the coefficient coef (f, K) is an extremely small steady value and the modGI value modGI (f, K) is a minute value.

  The target sound section determination device according to the first embodiment utilizes the behaviors of (a1) and (a2), and attempts to accurately determine an abnormal sound coming from the front.

(A-2) Configuration of First Embodiment FIG. 1 is a block diagram showing a configuration of a target sound segment determination device 10 according to the first embodiment.

  The target sound segment determination device 10 of the first embodiment may be constructed by connecting various hardware components, and some components (for example, microphone, analog / digital). The conversion unit (the part excluding the A / D conversion unit) may be constructed so as to realize the function by applying an execution configuration of a program such as a CPU, a ROM, and a RAM. Regardless of which construction method is applied, the functional detailed configuration of the target sound segment determination device 10 is the configuration shown in FIG.

  In FIG. 1, the target sound section determination device 10 according to the first embodiment includes a pair of microphones m1, m2, an FFT unit 11, a first directivity forming unit 12, a second directivity forming unit 13, and a coherence coefficient. It has the calculation part 14, the modGI calculation part 15, and the determination part 16.

  The pair of microphones m1 and m2 are arranged apart from each other by a predetermined distance (or an arbitrary distance), and each captures surrounding sounds. Each of the microphones m1 and m2 is omnidirectional (or has a very gentle directivity in the front direction). The acoustic signals (input sound signals) captured by the microphones m1 and m2 are converted into digital signals s1 (n) and s2 (n) via a corresponding A / D converter (not shown), and are given to the FFT unit 11. It is done. Note that n is an index indicating the input order of samples, and is expressed as a positive integer. In the text, it is assumed that the smaller n is the older input sample, and the larger n is the newer input sample.

The FFT unit 11 receives input sound signal sequences s1 (n) and s2 (n) from the microphones m1 and m2, and performs fast Fourier transform (or discrete Fourier transform) on the input sound signals s1 and s2. Thereby, the input sound signals s1 and s2 are expressed in the frequency domain. In performing the Fast Fourier Transform, analysis frames FRAME1 (K) and FRAME2 (K) composed of predetermined N samples are configured and applied from the input sound signals s1 (n) and s2 (n). An example of constructing the analysis frame FRAME1 (K) from the input sound signal s1 (n) is shown in the following equation (2), and the analysis frame FRAME2 (K) is the same.

  K is an index indicating the order of frames and is expressed by a positive integer. In the text, it is assumed that the smaller the K, the older the analysis frame, and the larger, the newer the analysis frame. In the following description, it is assumed that the index representing the latest analysis frame to be analyzed is K unless otherwise specified.

The FFT unit 11 converts the frequency domain signals X1 (f, K) and X2 (f, K) into the frequency domain signals X1 (f, K) by performing a fast Fourier transform process for each analysis frame. And X2 (f, K) are given to the first directivity forming unit 12 and the second directivity forming unit 13. Note that f is an index representing a frequency. X1 (f, K) is not a single value, but is composed of spectral components of multiple frequencies f1 to fm, as shown in equation (3). The same applies to X2 (f, K) and later-described B1 (f, K) and B2 (f, K).
X1 (f, K)
= {(F1, K), (f2, K), ..., (fm, K)} (3)

The first directivity forming unit 12 forms a signal B1 (f, K) having strong directivity in a specific direction from the two frequency domain signals X1 (f, K) and X2 (f, K), The directivity forming unit 12 generates a signal B2 (f, K) having strong directivity in a specific direction (different from the above-described specific direction) from the two frequency domain signals X1 (f, K) and X2 (f, K). To form. As a method for forming the signals B1 (f, K) and B2 (f, K) having strong directivity in a specific direction, an existing method can be applied. For example, the directivity is strong in the right direction by applying the equation (4). B2 (f, K) having strong directivity in the left direction can be formed by applying B1 (f, K) and (5). In the equations (4) and (5), the frame index K is omitted because it is not involved in the calculation.

The coherence coefficient calculation unit 14 obtains a coherence coefficient coef (f, K) by performing the calculation shown in the equation (6) based on the two directivity signals B1 (f) and B2 (f) described above. is there. Note that B2 (f) ^* in the equation (6) is a conjugate complex number of B2 (f).

  Conceptually speaking, the coherence coefficient coef (f, K) represents a correlation between certain frequency components of a signal arriving from the right and a signal arriving from the left. Therefore, the case where the coherence coefficient coef (f, K) is small is a case where the correlation between the frequency components of the two directivity signals B1 and B2 is small, and conversely, the case where the coherence coefficient coef (f, K) is large. In other words, it can be said that the correlation is large. If the correlation is small, the arrival direction of the frequency component in the input sound signal is greatly deviated to the right or left, or there is a clear regularity that makes it difficult for noise-like correlation to appear even if there is no deviation. This is the case with few components. Therefore, when the value of the coherence coefficient coef (f, K) is large, it can be said that there is no deviation in the arrival direction, and that component in the input sound signal comes from the front.

  However, in the case of abnormal noise, the frequency component included is 1 or a small number, so when the abnormal noise comes from the front, the coherence coefficient coef (f, K) for the included frequency component. Only the value of increases. FIG. 2 shows the value (solid line) of the coherence coefficient coef (f, K) for each frequency component (frequency bin) when the frequency component included in the noise is one (when the noise is a sinusoidal noise). ). In FIG. 2, for reference, the value of the coherence coefficient coef (f, K) for each frequency component (frequency bin) in the audio signal is indicated by a broken line.

  The coherence coefficient calculation unit 14 gives the obtained coherence coefficient coef (f, K) to the modGI calculation unit 15 and the determination unit 16.

  The modGI calculation unit 15 calculates a modGI value modGI (f, K) for the coherence coefficient coef (f, K) for each frequency component, and gives the obtained modGI value modGI (f, K) to the determination unit 16 It is. As the calculation formula of the modGI value modGI (f, K), the above-described formula (1) is applied, and the coherence coefficient coef (f, K) is substituted into the calculation target signal s (K) of the formula (1) to obtain the modGI value. modGI (f, K) is calculated. The expression (1) is the same calculation expression as the expression (13) in Patent Document 2, but the modGI value obtained by applying the expressions (5) and (10) to (12) described in Patent Document 2. modGI (f, K) may be calculated.

  FIG. 3 is a block diagram illustrating a detailed configuration of the determination unit 16 according to the first embodiment.

  In FIG. 3, the determination unit 16 includes an input signal reception unit 21, a coherence coefficient range calculation unit 22, a modGI range calculation unit 23, a target sound section determination unit 24, and a determination result transmission unit 25.

  The determination unit 16 determines whether or not the input sound signal is the target sound (abnormal sound) based on the coherence coefficient coef (f, K) and the modGI value modGI (f, K).

  The input signal receiving unit 21 receives the coherence coefficient coef (f, K) from the coherence coefficient calculation unit 14 and the modGI value modGI (f, K) from the modGI calculation unit 15.

  The coherence coefficient range calculation unit 22 searches for a minimum value and a maximum value in a sequence of coef (f, K) using a known search algorithm, and calculates a difference range_coef between the two.

  The modGI range calculation unit 23 searches for a minimum value and a maximum value in a series of modGI (f, K) using a known search algorithm, and calculates a difference range_modGI between the two.

  The target sound section determination unit 24 uses the range_coef calculated by the coherence coefficient range calculation unit 22 and the range_modGI calculated by the modGI range calculation unit 23 to determine whether or not the input sound signal is the target sound (abnormal sound). Judgment.

  The determination result transmission unit 25 transmits the determination result res (K) by the target sound section determination unit 24 to a signal processing unit (not shown). The determination result res (K) is used in, for example, a signal processing unit (for example, a voice switch processing unit).

(A-3) Operation of the First Embodiment Next, the operation of the target sound section determination device 10 according to the first embodiment will be described in the order of the overall operation and the operation of the determination unit 16 with reference to the drawings. .

  Signals s1 (n) and s2 (n) input from the pair of microphones m1 and m2 are respectively converted from time domain to frequency domain signals X1 (f, K) and X2 (f, K) by the FFT unit 11. Thereafter, directivity signals B1 (f, K) and B2 (f, K) having a blind spot in a predetermined direction are generated by the first and second directivity forming units 12 and 13, respectively. Then, in the coherence coefficient calculation unit 14, the directivity signals B1 (f, K) and B2 (f, K) are applied, the calculation of equation (6) is executed, and the coherence coefficient coef (f, K) is calculated. Then, it is given to the modGI calculation unit 15 and the determination unit 16. In the mod GI calculation unit 15, the mod GI value mod GI (f, K) for the coherence coefficient coef (f, K) is calculated according to, for example, the equation (1) and is given to the determination unit 16.

  FIG. 4 is a flowchart illustrating the operation of the determination unit 16 according to the first embodiment. The operation illustrated in FIG. 4 is repeatedly performed every time the processing target frame K is switched to a new frame.

  The input signal receiving unit 21 receives the coherence coefficient coef (f, K) from the coherence coefficient calculating unit 14 and the modGI value modGI (f, K) from the modGI calculating unit 15 (S101).

  The coherence coefficient range calculation unit 22 searches for the maximum value MAX (coef (f, K)) and the minimum value MIN (coef (f, K)) among the coherence coefficients coef (f, K) of all frequency components. Then, a difference range_coef (K) between the maximum value MAX (coef (f, K)) obtained by the search and the minimum value MIN (coef (f, K)) is calculated (step S102).

  The modGI range calculation unit 23 searches the maximum value MAX (modGI (f, K)) and the minimum value MIN (modGI (f, K)) among the modGI values modGI of all frequency components, and is obtained by the search. The difference range_modGI (K) between the maximum value MAX (modGI (f, K)) and the minimum value MIN (modGI (f, K)) is calculated (step S103).

  The target sound section determination unit 24 compares the calculated difference range_coef (K) with the threshold value Ψ (step S104). Similarly, the target sound section determination unit 24 compares the calculated difference range_modGI (K) with the threshold Φ. Here, the threshold Ψ is determined in advance by simulation or the like, and is selected to be a value capable of separating the difference range_coef (K) related to the target sound (abnormal sound) and the difference range_coef (K) related to the non-target sound. The The same applies to the threshold Φ.

  When the calculated difference range_coef (K) is equal to or larger than the threshold Ψ and the calculated difference range_modGI (K) is smaller than the threshold Φ, the current processing target frame K is a frame in the front abnormal noise section. Is determined. On the other hand, when the above condition is not satisfied, it is determined that the current processing target frame K is a frame in the non-target sound section.

  The target sound section determination unit 24 stores “1” in the determination result res (K) for the frame (K) determined to be a frame in the front abnormal sound section in the process of step S104 described above. (Step S105).

  For the frame (K) determined to be a frame within the non-target sound section in the process of step S104 described above, the target sound section determination unit 24 stores “0” in the determination result res (K). (Step S106). Note that the value (“1” or “0”) to be substituted for the determination result res (K) in the processing of step S105 and step S106 is an example, and an arbitrary value may be stored depending on the application to be used. .

  The target sound section determination unit 24 increments the parameter K (step S107). Then, the new frame becomes the processing target frame K, and the above-described operation is repeated.

  The determination result res (K) is transmitted by the determination result transmitting unit 25 to a processing unit (not shown) (voice switch processing unit or the like).

(A-4) Effect of First Embodiment According to the first embodiment, even when an abnormal sound including only one or a small number of frequency components is the target sound, it can be determined as the target sound. . As a result, it can be used for abnormal sound detection / inspection for components other than speech.

(B) Other Embodiments In addition to the above-described embodiments, the following modified embodiments can be exemplified.

  (B-1) Although the case where modGI is applied has been described in the first embodiment, since the GI before correction is also an index for measuring the number and magnitude of changes in the inclination direction of the signal waveform, GI may be applied instead of mod GI in the first embodiment.

  (B-2) In the first embodiment, the processing performed with the frequency domain signal may be performed with the time domain signal if possible.

  (B-3) The present invention is characterized by the configuration after obtaining the coherence coefficient, and the configuration before that is not necessarily limited to that of the first embodiment. For example, a signal of a microphone array having three or more microphones may be processed to obtain a coherence coefficient, and then modGI (or GI) may be calculated to detect a frequency component peculiar to the target sound.

  DESCRIPTION OF SYMBOLS 10 ... Target sound area determination apparatus, m1, m2 ... Microphone, 11 ... FFT part, 12 ... 1st directivity formation part, 13 ... 2nd directivity formation part, 14 ... Coherence coefficient calculation part, 15 ... modGI calculation , 16 ... determination unit, 21 ... input signal reception unit, 22 ... coherence coefficient range calculation unit, 23 ... modGI range calculation unit, 24 ... target sound section determination unit, 25 ... determination result transmission unit.

Claims (5)

A target sound section determination device for determining a target sound section from a section of an input sound signal using a sound including one or a small number of frequency components as a target sound,
Directivity that forms a plurality of directional signals with different azimuths with a dead angle by applying a delayed subtraction process to the input sound signal to give a directional characteristic having a blind spot in a predetermined direction. Forming means;
A coherence coefficient calculating means for obtaining a coherence coefficient using a plurality of formed directional signals;
A coherence coefficient feature amount calculating means for capturing the obtained coherence coefficient as a time-varying signal and obtaining a coherence coefficient feature amount representing the number of times and the magnitude of the change in the inclination direction of the signal waveform;
A target sound comprising: target sound section determination means for determining whether or not an input sound signal section is a target sound section based on the coherence coefficient range and the coherence coefficient feature amount range. Section determination device.   2. The object according to claim 1, wherein the coherence coefficient feature amount calculating unit calculates a value obtained by normalizing the power of the second-order difference of the coherence coefficient with the power of the coherence coefficient as a coherence coefficient feature amount. Sound segment determination device.   The target sound section determination means determines that the current input sound signal section is the target sound section when the coherence coefficient range is equal to or greater than the first threshold and the coherence coefficient feature amount range is smaller than the second threshold. In other cases, it is determined as a non-target sound section, and the target sound section determination device according to claim 1 or 2. A target sound section determination method for determining a target sound section from a section of an input sound signal using a sound including one or a small number of frequency components as a target sound,
The directivity forming means performs a delay subtraction process on the input sound signal to provide a plurality of directivity signals having a directivity characteristic having a blind spot in a predetermined direction, and a plurality of directivities having different predetermined directions having a blind spot. Form a signal,
A coherence coefficient calculation means acquires a coherence coefficient using a plurality of formed directional signals,
The coherence coefficient feature amount calculation means regards the obtained coherence coefficient as a time-varying signal, acquires a coherence coefficient feature amount representing the number of times and the magnitude of the change in the inclination direction of the signal waveform,
A target sound section, wherein the target sound section determination means determines whether the section of the input sound signal is a target sound section based on the coherence coefficient range and the coherence coefficient feature amount range. Judgment method. A target sound section determination program applied to determine a target sound section from a section of an input sound signal, with a sound including one or a small number of frequency components as a target sound,
Computer
Directivity that forms a plurality of directional signals with different azimuths with a dead angle by applying a delayed subtraction process to the input sound signal to give a directional characteristic having a blind spot in a predetermined direction. Forming means;
A coherence coefficient calculating means for obtaining a coherence coefficient using a plurality of formed directional signals;
A coherence coefficient feature amount calculating means for capturing the obtained coherence coefficient as a time-varying signal and obtaining a coherence coefficient feature amount representing the number of times and the magnitude of the change in the inclination direction of the signal waveform;
Based on the range of the coherence coefficient and the range of the coherence coefficient feature value, the input sound signal is made to function as a target sound section determination unit that determines whether or not a section of the input sound signal is a target sound section. Target sound section judgment program.

Images