JP2010175431A - Device, method and program for estimating sound source direction - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enhance an estimation precision of a sound source direction in a method and a device for estimating a sound source direction, and a program therefor. <P>SOLUTION: This device for estimating a sound source direction includes a microphone array having three microphones disposed at vertexes of a regular triangle, a frequency converter for converting signals received by the respective microphones of the microphone array to signals in a frequency region, an arrival time difference calculating section for calculating out an arrival time difference with respect to each combination of a microphone pair of different microphones, and a sound source direction estimating section that obtains a sound source candidate on the basis of the arrival time difference so as to classify sound source direction candidates. The sound source direction estimating section has a sparse characteristic determining section that determines whether or not a sparse characteristic can be assumed by each frequency bin of the arrival time difference. The sound source direction estimating section obtains a sound source candidate on the basis of the arrival time difference of the frequency bin of which the sparse characteristic can be assumed so as to classify the sound source direction candidate. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a sound source direction estimating apparatus and method for detecting the direction of a speaker used in a videophone or audio conference, and a program thereof.

A sound source direction estimation method used in a conventional audio conference or the like is disclosed in Non-Patent Document 1, for example. As shown in FIG. 13, N (N ≧ 2) different sound source directions _Sn are estimated using a microphone array composed of three microphones 1, 2, and 3 arranged at the vertices of an equilateral triangle as shown in FIG. To do. FIG. 14 shows an example of the functional configuration of the sound source direction estimating apparatus 300 and the operation will be described.

The sound source direction estimation device 300 includes three microphones 1, 2, 3 arranged at the vertices of an equilateral triangle, frequency conversion units 11, 12, 13, arrival time difference calculation units 21, 22, 23, and a sound source direction estimation unit. 150. Signals x _i (n) at time samples n received by the microphones 1, 2 and 3 are input to the frequency converters 11, 12, and 13 in the frequency domain obtained for each frame which is a set of a plurality of time samples. It is converted into a signal X _i (ω, m). Here, m and ω respectively indicate the number of the signal frame subjected to frequency conversion and the frequency of the signal after conversion. The frequency-converted microphone sound reception signal is input to arrival time difference calculation units 21, 22, and 23. The arrival time difference calculation units 21, 22, and 23 perform the calculation of Expression (1) for each of the combinations of three different microphone pairs, and the arrival time differences τ _ij (ω, m) (i, j) in the respective microphone pairs. ≦ 3, i ≠ j) is output. i and j indicate microphone numbers.

The arrival time difference τ _ij is input to the sound source direction estimation unit 150, and the estimated sound source direction θ _n ^ is output. Note that ^ is correct in the figure. FIG. 15 shows an example of the functional configuration of the sound source direction estimation unit 150 and its operation will be described. The sound source direction estimation unit 150 includes a vectorization unit 151, a sound source direction calculation unit 152, and a histogram calculation unit 153. The vectorization unit 151 receives the arrival time differences τ ₁₂ (ω, m), τ ₂₃ (ω, m), and τ ₃₁ (ω, m) output from the arrival time difference calculation units 21, 22, and 23, and uses the expression (2 The arrival time difference vector t (ω, m) shown in FIG. The vectorization unit 151 simply vectorizes the input arrival time difference τ _ij (ω, m).

  The sound source direction calculation unit 32 multiplies the input arrival time difference vector t (ω, m) by the coordinate transformation matrix D given by the equation (4) from the left as in the equation (3), and outputs the output The sound source direction candidate θ ′ (ω, m) is obtained from the one element and the second element by the calculation of Expression (5).

The histogram calculation unit 153 obtains a histogram from the input sound source direction candidate θ ′ (ω, m), and outputs the direction giving the peak of the histogram as the sound source direction estimated value θ _a ^ (a = 1,..., A ′). To do. A ′ is the maximum number of simultaneously generated sound sources given in advance.

  Here, the histogram is calculated by classifying all the sound source direction candidates θ ′ (ω, m) obtained in the respective frequency bins of a plurality of consecutive frames for each predetermined angle width. As the number of frames used for obtaining the histogram, the number of frames corresponding to a length of time that does not move the sound source is selected. For example, if the frame length is 16 ms and the sound source is considered not to move for about 0.5 seconds, the histogram is obtained using the sound source direction candidate θ ′ (ω, m) obtained in each of 30 frames, for example. It is done. The number of sound source direction candidates θ ′ (ω, m) is 3840 (128 × 30) frequencies when the signal sampling frequency is 16 kHz and the frequency conversion method is short-time Fourier transform using, for example, 256 points of data. Equal to the number of bins.

Masao Matsuo, Yusuke Hioka and Nozomu Hamada, “Estimating DOA of multiple speech signals by improved histogram mapping method,” Proceedings of IWAENC2005, pp.129-132.

  In the conventional method, when the sound source signal is a non-stationary signal such as voice and the components are concentrated at a specific frequency, an arbitrary frequency bin at an arbitrary time is only the component of one of the multiple sound sources. The processing is performed under the assumption called sparseness in the time-frequency domain.

[What is sparsity?]
Here, sparsity is a characteristic in which the energy of the signal of interest is concentrated in some areas in a certain area (in many cases, the time frequency area) and is zero in many other areas. This is called signal sparsity.

  However, since the assumption of signal sparsity generally breaks as the number of sound sources increases, the conventional technology cannot estimate the sound source direction with sufficient accuracy. For example, when a speaker located in a different direction speaks at the same time, the estimation accuracy of the sound source direction deteriorates. Further, in an actual environment, sounds other than voice are often generated, and most of these sounds are signals in which sound components spread in a steady and wide frequency, such as the sound of an air conditioner or a fan of a personal computer. Since these sounds cannot be assumed to be sparse, when they are superimposed on the sound of the sound source, the sound source direction estimation accuracy is further deteriorated.

  The present invention has been made in view of this point, and even if a speaker located in a different direction speaks at the same time, a sound source direction estimating apparatus, a method thereof, and a program thereof that can accurately estimate those directions. The purpose is to provide.

  The sound source direction estimating device of the present invention is different from a microphone array composed of three microphones arranged at the vertices of an equilateral triangle, and a frequency conversion unit that converts signals received by the microphones of the microphone array into signals in the frequency domain. An arrival time difference calculating unit that calculates an arrival time difference for each combination of microphone pairs of a microphone, and a sound source direction estimating unit that obtains a sound source candidate from the arrival time difference and classifies the sound source direction candidate. The sound source direction estimation unit includes a sparsity determination unit that determines whether sparsity can be assumed for each frequency bin of arrival time differences, obtains sound source candidates from the arrival time differences of frequency bins that can assume sparsity, Classify direction candidates.

  According to this invention, the sparsity determination unit removes the arrival time difference between frequency bins where the sparsity of the sound source cannot be assumed, and obtains the sound source candidate from the arrival time difference between the frequency bins where the remaining sparsity can be assumed. Therefore, the sound source direction estimating apparatus of the present invention excludes the arrival time difference of frequency bins where both voices are mixed even if speakers at different positions speak at the same time, and based on the arrival time difference consisting of a single sound source, Estimate the direction. Therefore, sound source direction estimation can be performed with high accuracy.

The figure which shows the function structural example of the sound source direction estimation apparatus 100 of this invention. The figure which shows the operation | movement flow of the sound source direction estimation apparatus 100. The figure which shows the function structural example of the sound source direction estimation part 30. FIG. The figure which shows the function structural example of the sparsity determination part. The figure which shows the operation | movement flow of the sparsity determination part. The figure which shows the example of an arrival time difference vector and an arrival time difference orthonormal vector. The figure which shows an example of vector orthogonality P ((theta)) in case there exist multiple sound sources. The figure which shows an example of vector orthogonality P ((theta)) in case there is one sound source. The figure which shows the function structural example of sparsity determination part 34 '. The figure which shows the operation | movement flow of sparse property determination part 34 '. The figure which shows an example of the result of having estimated the sound source direction with the conventional sound source direction estimation apparatus 300. FIG. The figure which shows an example of the result of having estimated the sound source direction with the sound source direction estimation apparatus 100 of this invention. The figure which shows the plane of a microphone array. The figure which shows the function structural example of the conventional sound source direction estimation apparatus 300. FIG. The figure which shows the function structural example of the conventional sound source direction estimation part 150. FIG.

  Embodiments of the present invention will be described below with reference to the drawings. The same components in the drawings are denoted by the same reference numerals, and the description thereof is omitted.

  FIG. 1 shows a functional configuration example of a sound source direction estimating apparatus 100 of the present invention. The sound source direction estimation apparatus 100 includes a microphone array including three microphones, frequency conversion units 11, 12, and 13, arrival time difference calculation units 21, 22, and 23, and a sound source direction estimation unit 30. The sound source direction estimation apparatus 100 differs from the sound source direction estimation apparatus 300 described in the related art only in that the sound source direction estimation unit 30 includes a sparsity determination unit 34 and the processing procedure using the determination result.

The same part as the operation of the sound source direction estimation apparatus 300 of the prior art will be briefly described with reference to the operation flow of FIG. The frequency converters 11, 12, and 13 convert the signals received by the microphones 1, 2, and 3 into signals in the frequency domain (step S11). The arrival time difference calculation units 21, 22, and 23 have arrival time differences τ _ij (ω, m) (τ ₁₂ (ω, m) and τ ₂₃ (ω) for each combination of microphone pairs of different microphones 1,2,3. , M), τ ₃₁ (ω, m)) is calculated (step S21). The sound source direction estimation unit 30 _{obtains a} sound source candidate θ ′ (ω, m) from the arrival time difference τ _ij (ω, m), and classifies the sound source candidate θ ′ (ω, m) (step S30).

The sound source direction estimating apparatus 100 of the present invention includes a sparsity determining unit 34 that determines whether or not the sound source direction estimating unit 30 can assume the sparsity for each frequency bin of the arrival time difference τ _ij (ω, m). And new. The sound source direction estimation unit 30 obtains sound source candidates from the arrival time difference τ _ij (ω, m) of the frequency bins that can be assumed to be sparsity output from the sparsity determination unit 34, and classifies the sound source candidates (step S30). This determination of sparsity is performed for each frame m and for each frequency bin ω. Therefore, even if speakers at different positions speak at the same time, the arrival time difference τ _ij (ω, m) of the frequency bins where the two voices are mixed is excluded, so that the direction of each sound source can be estimated with high accuracy. .

  FIG. 3 shows a functional configuration example of the sound source direction estimation unit 30. The sound source direction estimation unit 30 includes a vectorization unit 151, a sparsity determination unit 34, a sound source direction calculation unit 152 ′, and a histogram calculation unit 153. As is clear from the functional configuration example of the conventional sound source direction estimation apparatus 300 (FIG. 15), the sound source direction estimation unit 30 includes a sparsity determination unit 34 between the vectorization unit 151 and the sound source direction calculation unit 152. And the point that the sound source direction calculating unit 152 ′ calculates the sound source direction with reference to the determination result is different from the conventional sound source direction estimating unit 150.

An example of the functional configuration of the sparsity determination unit 34 of this embodiment is shown in FIG. 4 and its operation flow is shown in FIG. The sparsity determination unit 34 includes an orthogonal matrix calculation unit 35, a vector orthogonality calculation unit 36, and an orthogonality determination unit 38. Orthogonal matrix calculating section 35, the arrival time difference vector t (ω, m) output from the vectorization unit 151 as an input, the arrival time difference vector t (ω, m) orthogonal to the two arrival time differences orthonormal vectors t _⊥1 (Ω, m) and t _⊥2 (ω, m) are output (step S35). This orthonormal vector can be obtained, for example, by Gramschmitt orthonormalization. (Reference: “G. Strang,“ Linear Algebra and Its Applications ”Industrial Books, pages 141-143”)

Arrival time difference orthonormal vector t _{⊥1 (ω,} m) and t _{⊥2 (ω,} m) are input to the vector orthogonality calculating unit 36 calculates the orthogonality with respect to the theoretical value t _e of the arrival time difference vector _(theta) (Step S36). The theoretical value t _e (θ) of the arrival time difference vector is a value that can be calculated by Equation (6).

Here, d is the length of one side of the triangle formed by the microphones 1, 2, and 3 arranged at the vertices of the triangle (see FIG. 13). c is the speed of sound. In this way, t _e (θ) is a theoretical value that can be calculated regardless of the actual measurement value. The theoretical value t _e (θ) of the arrival time difference vector may be sequentially read out as recorded in the recording unit 37 as shown in FIG. 4, or a value recorded in advance in the vector orthogonality calculating unit 36 may be used. It may be used.

Here, the meaning of _obtaining two arrival time difference normal orthogonal vectors t _⊥1 (ω, m) and t _⊥2 (ω, m) orthogonal to the arrival time difference vector t (ω, m) will be described. FIG. 6 shows arrival time difference normal orthogonal vectors t _⊥1 (ω, m) and t _⊥2 (ω, m) with respect to an arbitrary arrival time difference vector t (ω, m). In order to know the direction of this arrival time difference vector t (ω, m), a vector whose direction is known and its arrival time difference orthonormal vector t _⊥1 (ω, m), t _⊥2 (ω, m) What is necessary is just to see whether it is orthogonal. If they are orthogonal, the direction of the arrival time difference vector t (ω, m) is the same as the direction of the known vector.

The vector orthogonality calculation unit 36 calculates the orthogonality _P between the arrival time difference normal orthogonal vectors t _⊥1 (ω, m) and t _⊥2 (ω, m) and the theoretical arrival time difference vector t _e (θ). (Θ) is calculated by equation (7) (step S36).

Expression (7) is obtained by _calculating all directions 0 with respect to arrival time difference orthonormal vectors t _⊥1 (ω, m) and t _⊥2 (ω, m) corresponding to individual arrival time difference vectors t (ω, m). It is calculated for the arrival time difference vector t _e (θ) with a theoretical value of ˜359 degrees. Since the direction of the arrival time difference vector t _e (θ) of the theoretical value calculated by the equation (7) is known, the theoretical value and the arrival time difference normal orthogonal vector t _⊥1 (ω, m), t _⊥2 (ω, When m) is orthogonal to each other, the first and second terms of the denominator of Equation (7) are each 0. Therefore, the orthogonality P (θ) has a large value. On the other hand, when the angle is different from the theoretical value, since the denominator first term and second term of Equation (7) have values of a certain magnitude, the value of the orthogonality P (θ) becomes a small value.

In this way, the arrival time difference normal orthogonal vectors t _⊥1 (ω, m), t _⊥2 (ω, m) orthogonal to the arrival time difference vector t (ω, m) are obtained, and these and the theoretical arrival time difference vector t _{e (theta)} and that evaluates whether orthogonal, arrival time difference vector t (ω, m) is, whether a vector made by a single sound source, could mixed signals of other excitation vector Can be determined.

Specific examples of the orthogonality P (θ) calculated by the equation (7) are shown in FIGS. The horizontal axis represents the arrival direction of the signal [degree], and the vertical axis represents the maximum vector orthogonality maxP (θ). Here, the 0 degree direction is the direction of the microphone 1 viewed from the center of the microphone array when the microphone array is placed on the desk (FIG. 13). FIG. 7 shows the maximum vector orthogonality maxP (θ) at each angle when the angle of the sound source 1 located at an angle of 10 degrees and another sound source 2 is changed from 0 degrees to 360 degrees. . The maximum vector orthogonality maxP (θ) shows a large value of about 32 only when the angles of the sound source 1 and the sound source 2 coincide with each other, and shows a small value of about 12 or less in the other directions.

  FIG. 8 shows the maximum vector orthogonality P (θ) when the angle of the sound source is changed from 0 degrees to 360 degrees when there is only one sound source. The maximum vector orthogonality maxP (θ) in all directions of the signal arrival direction shows the same magnitude (about 32) as the angle of 10 degrees in FIG.

  The orthogonality determination unit 38 determines the orthogonality of the arrival time difference vector t (ω, m) by comparing the orthogonality P (θ) with the threshold Th (step S38). The arrival time difference vector t (ω, m) having high orthogonality is a vector from a sound source at one fixed position, that is, an arrival time difference vector t (ω, m) that can assume sparsity. Conversely, the arrival time difference vector t (ω, m) having a small orthogonality P (θ) cannot assume sparsity.

  Whether or not this sparsity can be assumed is determined by setting the threshold Th to 15, for example, as shown in Expression (8) (step S380).

If the orthogonality P (θ) is larger than Th = 15, the sparsity determination result N _J (ω, m) is 1 (step S382), and if it is smaller, N _J (ω, m) is 0 (step S381). The arrival time difference vector t (ω, m) is updated until the determination on the arrival time difference vector t (ω, m) is completed (Y in step S383) (step S384). Therefore, the sparsity of the arrival time difference vectors t (ω, m) of all frames m and frequency bins ω is determined.

The sound source direction calculation unit 152 ′ refers to the sparsity determination result N _J (ω, m), and shows only the arrival time difference vector t (ω, m) of N _J (ω, m) = 1 in the equation (5). The sound source direction candidate θ ′ (ω, m) is calculated and output to the histogram calculation unit 153. The calculation of the sound source direction candidate θ ′ (ω, m) and the operation of obtaining the histogram by the histogram calculation unit 153 and setting the angle giving the peak value as the sound source direction are the same as in the prior art.

  As described above, the sound source direction estimating apparatus 100 estimates the sound source direction using the arrival time difference vector t (ω, m) of frequency bins that can be assumed to be sparsity, so that speakers at different positions speak at the same time. Even in some cases, the direction of each sound source can be accurately estimated. Note that the sparsity has been determined by a method for obtaining a normalized orthogonal vector with respect to the arrival time difference vector, but the present invention is not limited to this method. Another embodiment of the sparsity determination method will be described next.

The sparsity determination method according to the second embodiment is a method for determining sparsity by evaluating the difference in direction between the arrival time difference vector t (ω, m) and the theoretical arrival time difference vector t _e (θ). FIG. 9 shows a functional configuration example of the sparsity determination unit 34 ′ of the second embodiment. The sparsity determination unit 34 ′ includes an inter-vector distance calculation unit 90 and a vector matching determination unit 91.

The inter-vector distance calculation unit 90 receives the arrival time difference vector t (ω, m) as an input, and obtains the theoretical value t _e (θ) itself of the arrival time difference vector from the normalized actual measurement value normalized by the magnitude of the arrival time difference vector itself. A distance P ′ (θ) that is an absolute value of a value obtained by subtracting a normalized theoretical value normalized by the magnitude of is calculated by Expression (9).

Here, t _e (θ) is the magnitude of the theoretical value of the arrival time difference vector calculated by Expression (6). The theoretical value t _e (θ) of the arrival time difference vector may be sequentially read from the recording unit 37 ′ as shown in FIG. 9, or the value recorded in the inter-vector distance calculation unit 90 may be used. It may be used.

The distance P ′ (θ) is a value that becomes 0 when the direction of the arrival time difference vector t (ω, m) coincides with the direction of the theoretical value t _e (θ) of the arrival time difference vector. Therefore, it is possible to determine whether the arrival time difference vector t (ω, m) is a vector from one sound source or a vector influenced by another sound source according to the magnitude of the value. That is, whether or not the arrival time difference vector t (ω, m) can be assumed as sparsity can be determined based on the magnitude of the distance P ′ (θ).

  In the case of the second embodiment, the magnitude of the distance P ′ (θ) is determined by the vector matching determination unit 91 (step S91). Contrary to the first embodiment, the smaller the value of the distance P ′ (θ) is the arrival time difference vector t (ω, m) that can assume the sparsity. Other processes are the same as those in the first embodiment. In this way, it is possible to determine whether or not the arrival time difference vector t (ω, m) has sparsity.

〔simulation result〕
In order to confirm the effect of the present invention, the sound source direction estimation performance of the conventional sound source direction estimation apparatus 300 and the sound source direction estimation apparatus 100 of the present invention were compared. In the simulation, the sound source is a male positioned in the direction of 10 degrees and a female positioned in the direction of 20 degrees, and white noise without sparseness is superimposed on the voice uttered by both of them at a signal-to-noise ratio of 10 dB. Performed under conditions.

  As a result, the obtained histograms are shown in FIGS. The horizontal axis represents the arrival direction of the signal in [degrees], and the vertical axis represents [frequency]. FIG. 11 is a histogram obtained by the conventional sound source direction estimating apparatus 300. The vertices of the histogram are shifted in the directions of 5 degrees and 15 degrees. FIG. 12 is a histogram obtained by the sound source direction estimating apparatus 100 of the present invention. Two different peaks are correctly generated in the directions of 10 degrees and 20 degrees, and the peaks are conspicuous when compared with FIG. Thus, it was confirmed that the sound source direction estimation accuracy of the sound source direction estimation device 100 of the present invention was high.

  The sound source direction estimation apparatus and method of the present invention described above are not limited to the above-described embodiment, and can be appropriately changed without departing from the spirit of the present invention. For example, the processes described in the above-described apparatus and method are not only executed in time series in the order described, but are also executed in parallel or individually as required by the processing capability of the apparatus that executes the processes. Also good.

  Further, when the processing means in the above apparatus is realized by a computer, the processing contents of functions that each apparatus should have are described by a program. Then, by executing this program on the computer, the processing means in each apparatus is realized on the computer.

The program describing the processing contents can be recorded on a computer-readable recording medium. As the computer-readable recording medium, for example, any recording medium such as a magnetic recording device, an optical disk, a magneto-optical recording medium, and a semiconductor memory may be used. Specifically, for example, as a magnetic recording device, a hard disk device, a flexible disk, a magnetic tape, or the like, and as an optical disk, a DVD (Digital Versatile Disc), a DVD-RAM
(Random Access Memory), CD-ROM (Compact Disc Read Only Memory), CD-R
(Recordable) / RW (ReWritable) or the like can be used as a magneto-optical recording medium, MO (Magneto Optical disc) or the like as a semiconductor memory, and flash memory or the like as a semiconductor memory.

  The program is distributed by selling, transferring, or lending a portable recording medium such as a DVD or CD-ROM in which the program is recorded. Further, the program may be distributed by storing the program in a recording device of a server computer and transferring the program from the server computer to another computer via a network.

  Each means may be configured by executing a predetermined program on a computer, or at least a part of these processing contents may be realized by hardware.

Claims (7)

A microphone array consisting of three microphones arranged at the apex of an equilateral triangle;
A frequency converter that converts signals received by the microphones of the microphone array into signals in the frequency domain;
An arrival time difference calculation unit for calculating an arrival time difference for each combination of microphone pairs of the different microphones;
In a sound source direction estimating device comprising a sound source direction estimating unit that obtains sound source candidates from the arrival time difference and classifies the sound source direction candidates,
The sound source direction estimation unit includes a sparsity determination unit that determines whether sparsity can be assumed for each frequency bin of the arrival time difference, and obtains a sound source candidate from the arrival time difference of the frequency bins that can assume the sparsity, A sound source direction estimating apparatus for classifying the sound source direction candidates. In the sound source direction estimating apparatus according to claim 1,
The sparsity determination unit
An orthogonal matrix calculation unit for calculating two arrival time difference normal orthogonal vectors orthogonal to each frequency bin of the arrival time difference vector by using the arrival time difference vector;
A vector orthogonality calculating unit that calculates the orthogonality of the two arrival time difference normal orthogonal vectors with respect to a theoretical value of the arrival time difference vector, using the two arrival time difference normal orthogonal vectors as inputs;
A sound source direction estimation apparatus comprising: an orthogonality determination unit that compares the orthogonality with a threshold value and determines the sparsity of the arrival time difference vector. In the sound source direction estimating apparatus according to claim 1,
The sparsity determination unit
The absolute value of the value obtained by subtracting the normalized theoretical value normalized by the magnitude of the theoretical value of the arrival time difference vector from the normalized measured value normalized by the magnitude of the arrival time difference vector itself. A distance calculation unit between vectors for calculating a distance as a value for each frequency bin;
A sound source direction estimation apparatus comprising: a vector matching determination unit that determines the sparsity of the arrival time difference vector by comparing the distance with a threshold value. A frequency conversion process in which the frequency conversion unit converts a signal received by each microphone of a microphone array including three microphones into a signal in the frequency domain;
An arrival time difference calculating unit for calculating an arrival time difference for each combination of microphone pairs of the different microphones; and
In a sound source direction estimation method, including a sound source direction estimation unit, wherein a sound source direction estimation unit obtains sound source candidates from the arrival time difference and classifies the sound source direction candidates.
The sound source direction estimation process includes a sparsity determination process in which the sparsity determination unit determines whether or not the sparsity can be assumed for each frequency bin of the arrival time difference. From the arrival time difference of the frequency bins where the sparsity can be assumed. A method of estimating a sound source direction, which is a process of obtaining sound source candidates and classifying the sound source direction candidates. In the sound source direction estimation method according to claim 4,
The sparseness determination process is as follows:
An orthogonal matrix calculation process in which an orthogonal matrix calculation unit calculates two arrival time difference normal orthogonal vectors orthogonal to each frequency bin of the arrival time difference vector with the arrival time difference vector as an input;
A vector orthogonality calculating unit that receives the two arrival time difference normal orthogonal vectors as input and calculates the orthogonality of the two arrival time difference normal orthogonal vectors with respect to the theoretical arrival time difference vector;
A sound source direction estimation method, wherein the orthogonality determination unit includes an orthogonality determination step of determining the sparsity of the arrival time difference vector by comparing the orthogonality with a threshold value. In the sound source direction estimation method according to claim 4,
The sparseness determination process is as follows:
Normalized theoretical value normalized by the magnitude of the theoretical value of the arrival time difference vector from the actual measurement value normalized by the magnitude of the arrival time difference vector itself, with the inter-vector distance calculation unit receiving the arrival time difference vector as an input The inter-vector distance calculation process for calculating the distance that can be expressed by the absolute value of the value obtained by subtracting for each frequency bin,
A sound source direction estimation method, comprising: a vector matching determination step in which the vector matching determination unit determines the sparsity of the arrival time difference vector by comparing the distance with a threshold.   An apparatus program for causing a computer to function as the sound source direction estimating apparatus according to any one of claims 1 to 3.

Images