JP5337189B2 - Reflector arrangement determination method, apparatus, and program for filter design - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide technology for determining the placement of a reflection object which reflects voices in designing a filter adopted for voice based information. <P>SOLUTION: In cases where a filter adopted for voice based information is designed on the basis of a prescribed evaluation function by using a spatial correlation matrix expressed by transfer characteristics in plural directions in space, the transfer characteristics each are expressed by the sum of the transfer characteristic of a direct sound and each transfer characteristic of one reflection sound reflecting at a reflection object, and the evaluation function is such a function that the more a voice in at least a target direction is emphasized, the smaller the value it takes. A storage unit has stored therein candidates for the placement of the reflection object for a microphone array or a speaker array. A placement determination unit calculates the values of the evaluation function using a spatial correlation matrix expressed by transfer characteristics determined by the candidate, and determines a candidate for placement corresponding to the smallest of those values as the placement of the reflection object. <P>COPYRIGHT: (C)2013,JPO&amp;INPIT

Description

  The present invention relates to a technique for determining the arrangement of reflectors that reflect sound in the design of a filter applied to information based on sound.

  For example, considering a case where a subject is zoomed in with a moving image shooting apparatus (video camera or camcorder) equipped with a microphone, it is preferable for moving image shooting that the sound from only the vicinity of the subject is enhanced in conjunction with the zoom in shooting. Such a technique (speech enhancement technique) for enhancing a narrow range of speech including a desired direction (target direction) has been researched and developed conventionally. Note that the relationship between the direction around the microphone and the sensitivity of the microphone is called directivity. The sharper the directivity in a certain direction, the more the sound in a narrow range including the direction is emphasized. The voice can be suppressed. In this specification, “speech” is not limited to a voice uttered by a person, but refers to a general “sound” such as a musical sound or an environmental noise as well as a voice of a person or an animal.

  As a voice enhancement technique by selectively collecting reflected sounds, for example, there is a multi-beam forming method (see Non-Patent Document 1). The multi-beam forming method is a voice enhancement technology that collects individual sounds such as direct sound and reflected sound, and can pick up the sound in the target direction with a high SN ratio, which is better in the wireless field than in the voice field. It has been studied.

Hereinafter, processing contents of the multi-beam forming method in the frequency domain will be described. Prior to explanation, symbols are defined. Let the frequency index be ω and the frame number index be k. A frequency domain representation of an analog signal received by M microphones is expressed as X ^→ (ω, k) = [X ₁ (ω, k), ..., X _M (ω, k)] ^T , with direction θ _s The direction of arrival of the direct sound from the desired sound source is θ _s1 , and the direction of arrival of the reflected sound is θ _s2 ,..., Θ _sR . T represents transposition and R-1 is the total number of reflected sounds. _Let W ^→ (ω, θ _sr ) be a filter that enhances the voice in the direction θ _sr . Here, r is an integer satisfying 1 ≦ r ≦ R.

In the multi-beam forming method, it is assumed that the arrival direction and arrival time of the direct sound and the reflected sound are known. In other words, the number of objects such as walls, floors, and reflectors that can clearly predict sound reflection is equal to R-1. The reflected sound number R-1 is often set to a relatively small value of 3 or 4. This is based on the fact that a high correlation is recognized between the direct sound and the low-order reflected sound. Since the multi-beam forming method is a method in which each sound is individually emphasized and synchronously added, the output signal Y (ω, k, θ _s ) is given by Equation (1). H represents Hermitian transpose.

A delay synthesis method will be described as a design method of the filter W ^→ (ω, θ _sr ). Assuming that direct sound or reflected sound arrives as a plane wave, the filter W ^→ (ω, θ _sr ) is given by equation (2). h ^→ (ω, θ _sr ) = [h ₁ (ω, θ _sr ),..., h _M (ω, θ _sr )] ^T is a propagation vector of speech coming from the direction θ _sr .

When linear microphone array assuming that the plane wave (M number of microphones microphone array are arranged in a straight line) _{^{arrives, h → (ω, θ sr}} ) elements h _m (ω, θ _sr) that constitute the formula It is given by (3). m is an integer satisfying 1 ≦ m ≦ M. c represents the speed of sound, and u represents the distance between adjacent microphones. j is an imaginary unit. τ (θ _sr ) represents a time delay with respect to the direct sound of the reflected sound coming from the direction θ _sr .

Finally, by converting the output signal Y (ω, k, θ _s ) to the time domain, a signal in which the sound of the sound source in the target direction θ _s is enhanced is obtained.

J.L.Flanagan, A.C.Surendran, E.E.Jan, "Spatially selective sound capture for speech and audio processing," Speech Communication, Volume 13, Issue 1-2, pp.207-222, October 1993.

In the multi-beam forming method, it is assumed that the arrival direction and arrival time of the direct sound and the reflected sound are known. Further, when designing a filter W ^→ (ω, θ _sr ) that emphasizes the voice from a certain direction θ _sr, only the voice in the direction θ _sr is considered alone as expressed by the equation (2). It was.

  However, as will be described later in detail in the embodiment of the present invention, it may be preferable to consider a sound in a certain direction as a mixed sound of a direct sound and a reflected sound at the filter design stage. Alternatively, there is a case where it is required to determine an appropriate arrangement of the reflector that reflects the sound in relation to the speaker array.

  Therefore, an object of the present invention is to provide a technique for determining the arrangement of a reflector that reflects sound with respect to a microphone array or a speaker array in the design of a filter applied to information based on sound.

  A filter applied to information based on speech is designed based on a predetermined evaluation function using a spatial correlation matrix represented by transfer characteristics in a plurality of directions in space, and each transfer characteristic Is expressed as the sum of the transfer characteristics of the direct sound and the transfer characteristics of one reflected sound reflected by the reflector, and the evaluation function is a function that takes at least a value so that the voice in the target direction is emphasized. . The storage unit stores information (hereinafter referred to as arrangement information) indicating the arrangement relationship of the reflector with respect to the microphone array or the speaker array, and the arrangement determination unit sets each candidate of the reflector based on the arrangement information as the candidate. The value of the evaluation function is obtained using the spatial correlation matrix expressed by the transfer characteristic specified on the basis, and the candidate corresponding to the smallest one among the values is determined as the arrangement of the reflectors.

  According to the present invention, in the design of a filter applied to information based on sound, it is possible to determine the arrangement of a reflector that reflects sound with respect to the microphone array or the speaker array.

The figure which shows the function structure of the reflector arrangement | positioning determination apparatus which concerns on embodiment. The figure which shows the function structure of the audio processing apparatus of the application form 1. FIG. The figure which shows the process sequence of the audio | voice processing method of the application form 1. FIG. The figure which shows the function structure of the audio processing apparatus of the application form 2. FIG. The figure which shows the process sequence of the audio | voice processing method of the application form 2. FIG. The figure which shows the function structure of the audio processing apparatus of the application form 3. FIG. The figure which shows the process sequence of the audio | voice processing method of the application form 3. The figure which shows the function structure of the audio processing apparatus of the application form 4. FIG. The figure which shows the process sequence of the audio | voice processing method of the application form 4. The figure which shows the positional relationship etc. of a microphone array and a reflecting plate (the 1). FIG. 2 is a diagram illustrating a positional relationship between a microphone array and a reflecting plate (part 2); The figure which shows the positional relationship etc. of a speaker array and a reflecting plate (the 1). The figure (the 2) which shows the positional relationship etc. of a speaker array and a reflecting plate.

An embodiment of the present invention will be described with reference to FIG. In general, in the present invention, when a sound in a certain direction is considered as a mixed sound of a direct sound and a reflected sound, information based on the sound (in the example of the embodiment, the sound signal is converted into the frequency domain). In the design of a filter applied to a frequency domain signal), it is a technique for determining the arrangement of a reflector that reflects sound with respect to a microphone array or a speaker array, and does not affect the filter design concept itself. Therefore, there is no particular limitation as a filter design technique to which the present invention is applied. The design concept of the filter is a statistical optimization rule. For example, for the difference (estimation error) between the output obtained by applying the filter to the input sample sequence and the desired response (estimation error), the mean square value of the estimation error, the estimation error The expected value of the absolute value of, the expected value of the third or higher power related to the absolute value of the estimation error, etc. can be listed as evaluation functions, and this evaluation function is minimized (maximized depending on the evaluation function and its expression). To design the filter. Here, in order to make the explanation consistent, the evaluation function is a function that outputs a value having a smaller absolute value so that at least the sound in the target direction viewed from the microphone array or the speaker array is emphasized. The reason for "at least ... the target direction" is that not only the target direction but also the distance from the microphone array or the speaker array to the sound source in any design method, as will be described later in <Introduction of distance>. This is because it is possible to design the filter in consideration. Here, there are three types of filter design methods: the minimum variance distortion response method (MVDR method), the filter design method based on the SNR maximization criterion, and the filter design method based on power inversion. Is illustrated. Refer to Reference Document 1 for the minimum variance distortionless response method, and Reference Document 2 for the filter design method based on the S / N ratio maximization criterion and the filter design method based on power inversion.
(Reference 1) Toshiro Oga, Yoshio Yamazaki, Yutaka Kaneda, "Acoustic System and Digital Processing", The Institute of Electronics, Information and Communication Engineers, 1995, pp.203-209
(Reference 2) Nobuyoshi Kikuma, “Adaptive Antenna Technology”, 1st Edition, Ohm Corporation, 2003, pp.35-90

The reflector arrangement determining apparatus 100 according to the embodiment of the present invention is generally present as an entity constituting, for example, the audio processing apparatuses 1, 2, 3, and 4 to be described later, rather than being independently present alone. is there. Furthermore, the reflector arrangement determining device 100 is not an entity that constitutes the speech processing devices 1, 2, 3, and 4 so as to be easily separable from the speech processing devices 1, 2, 3, and 4, but the speech processing device 1 2, 3, 4 itself can be said to have been evaluated unilaterally with a focus on some functions. In short, the reflector arrangement determining device 100 is generally the sound processing device 1, 2, 3, 4 itself. Specifically, the reflector arrangement determining apparatus 100 can be realized by mounting the function of the reflector arrangement determining apparatus 100 in a central processing unit or a dedicated LSI.
However, the reflector arrangement determining apparatus 100 exists as a single independent entity, and the entities constituting the speech processing apparatuses 1, 2, 3, and 4 can be easily separated from the speech processing apparatuses 1, 2, 3, and 4. It is not intended to exclude that. For example, if the object is to determine the arrangement of the reflector itself, there is no obstacle to realizing the reflector arrangement determining apparatus 100 as a single independent entity.
Here, it is assumed that the audio processing devices 1, 2, 3, and 4 are realized by a computer such as a dedicated machine configured by dedicated hardware or a general-purpose machine such as a personal computer, and the reflector arrangement determination is performed as a single independent entity. The same applies to the case where the device 100 is realized.

<1> Filter design method based on minimum variance distortion-free response method Prior to explanation, symbols are defined again. The index of the discrete frequency is ω (there is a relationship of ω = 2πf between the frequency f and the angular frequency ω, so the index ω of the discrete frequency may be identified with the angular frequency ω. The frequency index "is also simply referred to as" frequency "), and the frame number index is k. The frequency domain representation of the kth frame of the analog signal received by M microphones is expressed as X ^→ (ω, k) = [X ₁ (ω, k), ..., X _M (ω, k)] ^T , microphone array Let W ^→ (ω, θ _s ) be a filter that emphasizes the frequency domain representation of the speech in the target direction θ _s with the frequency ω as viewed from the center of the. M is an integer of 2 or more. T represents transposition. At this time, a frequency domain signal (hereinafter referred to as an output signal) Y (ω, k, θ _s ) in which the frequency domain representation of the voice in the target direction θ _s is emphasized by the frequency ω is given by Expression (4). H represents Hermitian transpose.

The “center of the microphone array” can be arbitrarily determined, but generally, the geometric center of the arrangement of the M microphones is the “center of the microphone array”. For example, in the case of a linear microphone array, the microphones at both ends The middle point of the microphone is the “center of the microphone array”. For example, in the case of a planar microphone array arranged in a square matrix of m × m (m ² = M), the positions where the diagonal lines of the microphones at the four corners intersect are “microphone array”. The center of

Filter W ^→ (ω, θ _s) if due to the minimum variance distortionless response method as a design method of the filter W ^→ (ω, θ _s) under the constraint condition of the equation (6), the spatial correlation matrix Q (omega ) audio (hereinafter the direction other than the target direction theta _s using a power of the "target direction theta _s other way voice" is also referred to as "noise") are designed to minimize the frequency omega (formula (Refer to (5)). a ^→ (ω, θ _s ) = [a ₁ (ω, θ _s ), ..., a _M (ω, θ _s )] ^T is the sound source when it is assumed that there is a sound source in the direction θ _s This is a transfer characteristic at a frequency ω between M microphones. In other words, a ^→ (ω, θ _s ) = [a ₁ (ω, θ _s ), ..., a _M (ω, θ _s )] ^T is the direction θ _s to each microphone included in the microphone array. It is a transfer characteristic at the frequency ω of sound.

It is known that the filter W ^→ (ω, θ _s ), which is the optimum solution of the equation (5), is given by the equation (7) (see Reference 1).

From equation (5), it can be seen that the noise power depends on the structure of the spatial correlation matrix Q (ω). Therefore, the structure of the spatial correlation matrix Q (ω) will be described. Let {1, 2,..., P-1} be a set to which the noise arrival direction index p belongs. It is assumed that the index s in the target direction θ _s does not belong to the set {1, 2,..., P-1}. Assuming that P-1 noises come from any direction, the spatial correlation matrix Q (ω) is given by equation (8a). From the viewpoint of creating a filter that functions sufficiently even in the presence of a lot of noise, P is preferably a somewhat large value, and is assumed to be an integer of about M. Here, from the viewpoint of easily explaining the principle of the invention, the target direction θ _s is described as if it were a specific direction (therefore, directions other than the target direction θ _s are set as “noise” directions). Actually, the target direction θ _s is an arbitrary direction that can be a target of speech enhancement, and a plurality of directions are generally assumed as directions that can be the target direction θ _s . From this point of view, the distinction between the target direction θ _s and the noise direction is almost subjective, and there are P number of directions that can be assumed as voice arrival directions regardless of whether the target sound or noise is distinguished. It is more accurate to determine different directions in advance and understand that one of the P directions selected is the target direction and the other directions are noise directions. Therefore, if the union of the set {1, 2, ..., P-1} and the set {s} is Φ, the spatial correlation matrix Q (ω) is included in multiple directions that are assumed as the voice arrival directions. transfer characteristics a to each microphone of the speech in each direction theta _phi to _{^{→ (ω, θ φ) =}} [a 1 (ω, θ φ), ..., a M (ω, θ φ)] T (φ∈Φ) Is a spatial correlation matrix expressed by Equation (8b). Note that | Φ | = P. | Φ | represents the number of elements of the set Φ.

Each microphone in the microphone array has two types of sound waves, a direct sound from a sound source and a reflected sound that is reflected from the sound source by a reflector (here it is assumed to be a plane wave for convenience of explanation, but a spherical wave May be present). Let the number of reflected sounds be Ξ. Ξ is a predetermined integer of 1 or more. At this time, the transfer characteristic a ^→ (ω, θ) = [a ₁ (ω, θ),..., A _M (ω, θ)] ^T directly transmits the sound in the direction that can be the target of speech enhancement to the microphone array. The sum of the direct sound transmission characteristics and the transmission characteristics of one or more reflected sounds that are reflected by the reflector and reach the microphone array, specifically, the direct sound and the ξth (1 ≦ ξ ≦ Ξ) If the arrival time difference from the reflected sound is τ _ξ (θ), and α _ξ (1 ≦ ξ ≦ Ξ) is a coefficient for considering the attenuation of the sound due to the reflection, the direct sound This can be expressed as the sum of the steering vector and the steering vector of several reflected sounds in which the sound attenuation due to reflection and the arrival time difference with respect to the direct sound are corrected. h ^→ _d (ω, θ) = [h _d1 (ω, θ),…, h _dM (ω, θ)] ^T is the steering vector of the direct sound in the direction θ, h ^→ _rξ (ω, θ) = [ h _r1ξ (ω, θ),..., h _rMξ (ω, θ)] ^T represents the steering vector of the reflected sound corresponding to the direct sound in the direction θ. The steering vector is a complex vector in which the phase response characteristics at the frequency ω of each microphone with respect to the reference point are arranged for sound waves in the direction θ as viewed from the center of the microphone array. α _ξ (1 ≦ ξ ≦ Ξ) is usually α _ξ ≦ 1 (1 ≦ ξ ≦ Ξ). For each reflected sound, if the number of reflections from the sound source to the microphone is one, α _ξ (1 ≦ ξ ≦ Ξ) represents the reflectance of the sound of the object reflected by the ξth reflected sound. You can think that it is.

Since it is desired to provide one or more reflected sounds to a microphone array composed of M microphones, it is preferable that one or more reflectors exist. From this point of view, assuming that there is a sound source in the target direction, the positional relationship between the sound source, the microphone array, and one or more reflectors is such that the sound from the sound source is reflected by at least one reflector. Each reflector is preferably arranged to reach the array. Each reflector has a two-dimensional shape (for example, a flat plate) or a three-dimensional shape (for example, a parabolic shape). Moreover, it is preferable that the size of each reflector is equal to or larger than the microphone array (about 1 to 2 times). In order to effectively use the reflected sound, the reflectance α _ξ (1 ≦ ξ ≦ Ξ) of each reflector is at least greater than 0, and more specifically, the amplitude of the reflected sound that reaches the microphone array is a direct sound. For example, each reflector is a rigid solid. Even if the reflector is a movable object (for example, a reflector movably combined with the support on which the microphone array is installed), it cannot be moved (for example, fixed to the support on which the microphone array is installed). Reflection plate). The determination of the arrangement relationship of the reflecting objects will be described later.

Assuming that speech arrives at the linear microphone array as a plane wave, the m-th element h _dm (ω, θ) constituting the direct sound steering vector h ^→ _d (ω, θ) is given by, for example, the following equation (10a). m is an integer satisfying 1 ≦ m ≦ M. c represents the speed of sound, and u represents the distance between adjacent microphones. j is an imaginary unit. The reference point is half the total length of the linear microphone array (the center of the linear microphone array). The direction θ is defined as an angle formed by the direct sound arrival direction and the arrangement direction of the microphones included in the linear microphone array as viewed from the center of the linear microphone array (see FIGS. 10 and 11). There are various ways of expressing the steering vector. For example, if the reference point is the position of the microphone at one end of the linear microphone array, the mth element constituting the direct sound steering vector h ^→ _d (ω, θ). h _dm (ω, θ) is given by, for example, equation (10b). In the following description, it is assumed that the m-th element h _dm (ω, θ) constituting the direct sound steering vector h ^→ _d (ω, θ) is given by equation (10a).

Reflection sound steering vector h ^→ _r (ω, θ) = [h _r1 (ω, θ),…, h _rM (ω, θ)] The m-th element of ^T Similarly (refer to Formula (10a)), it is expressed by Formula (11a). The function Ψ (θ) outputs the arrival direction of the reflected sound. When the direct sound steering vector is expressed by equation (10b), the reflected sound steering vector h ^→ _r (ω, θ) = [h _r1 (ω, θ),..., H _rM (ω, θ) ] The m-th element of ^T is expressed by Expression (11b). In general, the _ξth (1 ≦ ξ ≦ Ξ) steering vector h ^→ _rξ (ω, θ) = [h _r1ξ (ω, θ),…, h _rMξ (ω, θ)] The mth element of ^T Is expressed by equation (11c) or equation (11d). The function Ψ _ξ (θ) outputs the arrival direction of the ξth (1 ≦ ξ ≦ Ξ) reflected sound.

Now, the arrival time difference τ _ξ (θ) and the function ψ _ξ (θ) are determined by the arrangement relationship of the reflector with respect to the microphone array. When the arrival time difference τ _ξ (θ) and the function Ψ _ξ (θ) are determined, the direct sound steering vector h ^→ _d (ω, θ) and the reflected sound steering vector h ^→ _rξ (ω, θ) are determined. When the direct sound steering vector h ^→ _d (ω, θ) and the reflected sound steering vector h ^→ _rξ (ω, θ) are determined, the transfer characteristic a ^→ (ω, θ) is determined. When the transfer characteristic a ^→ (ω, θ) is determined, the spatial correlation matrix Q (ω) is determined. As described above, the noise power depends on the structure of the spatial correlation matrix Q (ω). Therefore, it is important to determine the positional relationship of the reflector with respect to the microphone array. Here, as a specific example, the positional relationship of the reflector with respect to the microphone array is specified based on the angle to the microphone array (the angle between the microphone array direction and the reflector included in the linear microphone array) and the distance from the center of the microphone array. (See FIGS. 10, 11, 12, and 13).

Hereinafter, from the viewpoint of specific explanation, it is assumed that Ξ = 1, the number of reflections of reflected sound is one, and there is one reflector at a position away from the center of the microphone array. The reflector is a thick rigid plate. Hereinafter, the reflector is referred to as a reflector. In this case, since Ξ = 1, the subscript representing this is omitted, and the expression (9a) can be expressed as the expression (9b).

  The storage unit 101 of the reflector arrangement determining apparatus stores information representing the arrangement relationship of the reflector with respect to the microphone array as data (in some embodiments, as described later, “the arrangement relationship of the reflector with respect to the speaker array is changed. The information is “represented”, but here, the case of a microphone array will be described as a representative). An example of information representing the arrangement relationship of the reflectors with respect to the microphone array is a set of predetermined candidates relating to the arrangement of the reflectors with respect to the microphone array, and this set is U. The candidates included in the set U are, for example, that the number of reflector angle candidates with respect to the microphone array is J and the number of distance candidates from the center of the microphone array to the reflector is K. The total number of candidates that are represented by combinations with candidates and included in the set U is J × K (hereinafter abbreviated as JK).

As another example of the information indicating the arrangement relationship of the reflector with respect to the microphone array, information indicating a function may be used. For example, a reflector function candidate C _{Angle, j} = j × Δθ [j = 1, 2,..., J] with respect to the microphone array and a candidate C _distance of the distance from the center of the microphone array to the reflector _{, K} = k × ΔL [k = 1, 2,..., K] may be stored in the storage unit 101 of the reflector arrangement determination device as information representing the arrangement relationship of the reflector with respect to the microphone array. . Here, Δθ is a predetermined angle, and ΔL is a predetermined length. Here, a discrete function that gives candidates for angles and distances at equal intervals is illustrated, but it is of course possible to be a discrete function that gives candidates for angles and distances at non-uniform intervals, or a continuous function (continuous In the case of a function, for example, input values may be set discretely).

The arrangement determining unit 110 of the reflecting object arrangement determining apparatus 100 is configured to calculate JK candidates acquired from the storage unit 101 or JK candidates obtained by the arrangement determining unit 110 according to a function stored in the storage unit 101. for each (candidate index and n), calculates the power (evaluation function) p _n according to formula (12). Ω is a set of frequencies ω. The spatial correlation matrix Q _n (ω) is a spatial correlation matrix based on the arrangement relationship of the reflectors corresponding to the candidate index n (see Expression (8a) or Expression (8b)), and the filter W _n ^→ (ω, θ _s ) Is a filter based on the arrangement relationship of the reflectors corresponding to the candidate index n (see Expression (7)).

If the intended direction theta _s is one according to equation (12), but if the target direction is plural, arrangement determination unit 110 calculates the power (evaluation function) p _n according to formula (13). A plurality of target directions are θ _s1 ,..., Θ _sA . However, the total number | {θ _s1 ,..., Θ _sA } | = A does not exceed P. This processing is not a viewpoint of designing a filter that realizes good speech enhancement by specializing in any one of these target directions, and any one of these target directions. This is based on the viewpoint of designing a filter that realizes good speech enhancement in a balanced manner.

Next, the arrangement determining unit 110 searches for the minimum power among the powers p ₁ ,..., P _JK corresponding to the JK candidates. For example, if the minimum power is _pg , the “angle of the reflector with respect to the microphone array and the distance from the center of the microphone array to the reflector” specified by the index g is the optimum reflector arrangement condition for the microphone array. As determined.

<2> Filter design method based on S / N ratio maximization criteria
In the filter design method based on the SN ratio maximization criterion, the filter W ^→ (ω, θ _s ) is determined based on the criterion for maximizing the SN ratio (SNR) in the target direction θ _s . The spatial correlation matrix of the audio object direction θ _s R _ss (ω), the spatial correlation matrix of the direction other than the target direction theta _s voice and R _nn (ω). At this time, the SNR that is an evaluation function is expressed by Expression (14). R _ss (ω) is expressed by equation (15), and R _nn (ω) is expressed by equation (16). Transfer characteristic a ^→ (ω, θ _s ) = [a ₁ (ω, θ _s ),..., A _M (ω, θ _s )] ^T is expressed by equation (9a) (exactly, equation (9a ) Is θ _s ).

The filter W ^→ (ω, θ _s ) that maximizes the SNR in Expression (14) can be obtained by setting the gradient related to the filter W ^→ (ω, θ _s ) to zero, that is, Expression (17).

Thus, the filter W ^→ (ω, θ _s ) that maximizes the SNR in Expression (14) is given by Expression (18).

Equation (18) includes an inverse matrix of the spatial correlation matrix R _nn (ω) of speech in a direction other than the target direction θ _s , and the inverse matrix of R _nn (ω) is used as the speech in the target direction θ _s . It is known that it may be replaced with an inverse matrix of the spatial correlation matrix R _xx (ω) of the entire input including speech in directions other than the target direction θ _s . Note that R _xx (ω) = R _ss (ω) + R _nn (ω) = Q (ω). That is, the filter W ^→ (ω, θ _s ) that maximizes the SNR in equation (14) may be obtained by equation (19).

  The storage unit 101 of the reflector arrangement determining apparatus stores information representing the arrangement relationship of the reflector with respect to the microphone array as data (in some embodiments, as described later, “the arrangement relationship of the reflector with respect to the speaker array is changed. The information is “represented”, but here, the case of a microphone array will be described as a representative). An example of information representing the arrangement relationship of the reflectors with respect to the microphone array is a set of predetermined candidates relating to the arrangement of the reflectors with respect to the microphone array, and this set is U. The candidates included in the set U are, for example, that the number of reflector angle candidates with respect to the microphone array is J and the number of distance candidates from the center of the microphone array to the reflector is K. The total number of candidates that are represented by combinations with candidates and are included in the set U is JK.

As another example of the information indicating the arrangement relationship of the reflector with respect to the microphone array, information indicating a function may be used. For example, a reflector function candidate C _{Angle, j} = j × Δθ [j = 1, 2,..., J] with respect to the microphone array and a candidate C _distance of the distance from the center of the microphone array to the reflector _{, K} = k × ΔL [k = 1, 2,..., K] may be stored in the storage unit 101 of the reflector arrangement determination device as information representing the arrangement relationship of the reflector with respect to the microphone array. . Here, Δθ is a predetermined angle, and ΔL is a predetermined length. Here, a discrete function that gives candidates for angles and distances at equal intervals is illustrated, but it is of course possible to be a discrete function that gives candidates for angles and distances at non-uniform intervals, or a continuous function (continuous In the case of a function, for example, input values may be set discretely).

The arrangement determining unit 110 of the reflecting object arrangement determining apparatus 100 is configured to calculate JK candidates acquired from the storage unit 101 or JK candidates obtained by the arrangement determining unit 110 according to a function stored in the storage unit 101. for each (candidate index and n), calculates the SN ratio according to equation (20) (evaluation function) p _n. Ω is a set of frequencies ω. Spatial correlation matrices R _ss ⁽ⁿ⁾ (ω) and R _nn ⁽ⁿ⁾ (ω) are spatial correlation matrices based on the arrangement relationship of the reflectors corresponding to the candidate index n (Expression (15), Expression (16)). The filter W _n ^→ (ω, θ _s ) is a filter based on the arrangement relationship of the reflectors corresponding to the candidate index n (see formula (18) or formula (19)).

While if the target direction theta _s is one according to equation (20), if the target direction is plural, arrangement determination unit 110 calculates the SN ratio (evaluation function) p _n according to formula (21). A plurality of target directions are θ _s1 ,..., Θ _sA . However, the total number | {θ _s1 ,..., Θ _sA } | = A does not exceed P. This processing is not a viewpoint of designing a filter that realizes good speech enhancement by specializing in any one of these target directions, and any one of these target directions. This is based on the viewpoint of designing a filter that realizes good speech enhancement in a balanced manner.

Next, the arrangement determining unit 110 searches for the minimum SN ratio among the SN ratios p ₁ ,..., P _JK corresponding to the JK candidates. For example, if the minimum signal-to-noise ratio is _pg , the “angle of the reflector with respect to the microphone array and the distance from the center of the microphone array to the reflector” specified by the index g is the optimum arrangement of the reflector with respect to the microphone array. Determined as a condition.

<3> Filter design method based on power inversion In the filter design method based on power inversion, the filter W ^→ (ω, θ _s ) is determined. Here, as an example, it is assumed that the filter coefficient for the first microphone among the M microphones is fixed. In this design method, the filter W ^→ (ω, θ _s ) is omnidirectional (all the possible directions of speech arrival) using the spatial correlation matrix R _xx (ω) under the constraint of equation (23). It is designed to minimize the power of the voice in the direction (see equation (22)). Transfer characteristic a ^→ (ω, θ _s ) = [a ₁ (ω, θ _s ),..., A _M (ω, θ _s )] ^T is expressed by equation (9a) (exactly, equation (9a ) Is θ _s ). Note that R _xx (ω) = Q (ω).

It is known that the filter W ^→ (ω, θ _s ), which is the optimal solution of the equation (22), is given by the equation (24) (see Reference 2).

  The storage unit 101 of the reflector arrangement determining apparatus stores information representing the arrangement relationship of the reflector with respect to the microphone array as data (in some embodiments, as described later, “the arrangement relationship of the reflector with respect to the speaker array is changed. The information is “represented”, but here, the case of a microphone array will be described as a representative). An example of information representing the arrangement relationship of the reflectors with respect to the microphone array is a set of predetermined candidates relating to the arrangement of the reflectors with respect to the microphone array, and this set is U. The candidates included in the set U are, for example, that the number of reflector angle candidates with respect to the microphone array is J and the number of distance candidates from the center of the microphone array to the reflector is K. The total number of candidates that are represented by combinations with candidates and are included in the set U is JK.

As another example of the information indicating the arrangement relationship of the reflector with respect to the microphone array, information indicating a function may be used. For example, a reflector function candidate C _{Angle, j} = j × Δθ [j = 1, 2,..., J] with respect to the microphone array and a candidate C _distance of the distance from the center of the microphone array to the reflector _{, K} = k × ΔL [k = 1, 2,..., K] may be stored in the storage unit 101 of the reflector arrangement determination device as information representing the arrangement relationship of the reflector with respect to the microphone array. . Here, Δθ is a predetermined angle, and ΔL is a predetermined length. Here, a discrete function that gives candidates for angles and distances at equal intervals is illustrated, but it is of course possible to be a discrete function that gives candidates for angles and distances at non-uniform intervals, or a continuous function (continuous In the case of a function, for example, input values may be set discretely).

The arrangement determining unit 110 of the reflecting object arrangement determining apparatus 100 is configured to calculate JK candidates acquired from the storage unit 101 or JK candidates obtained by the arrangement determining unit 110 according to a function stored in the storage unit 101. for each (candidate index and n), calculates the power (evaluation function) p _n according to formula (25). Ω is a set of frequencies ω. The spatial correlation matrix Q _n (ω) = R _xx (ω) is a spatial correlation matrix based on the arrangement relationship of the reflectors corresponding to the candidate index n (see Expression (8a) or Expression (8b)), and the filter W _n ^→ (ω, θ _s ) is a filter based on the arrangement relationship of the reflectors corresponding to the candidate index n (see Expression (24)).

If the intended direction theta _s is one according to equation (25), but if the target direction is plural, arrangement determination unit 110 calculates the power (evaluation function) p _n according to formula (26). A plurality of target directions are θ _s1 ,..., Θ _sA . However, the total number | {θ _s1 ,..., Θ _sA } | = A does not exceed P. This processing is not a viewpoint of designing a filter that realizes good speech enhancement by specializing in any one of these target directions, and any one of these target directions. This is based on the viewpoint of designing a filter that realizes good speech enhancement in a balanced manner.

Next, the arrangement determining unit 110 searches for the minimum power among the powers p ₁ ,..., P _JK corresponding to the JK candidates. For example, if the minimum power is _pg , the “angle of the reflector with respect to the microphone array and the distance from the center of the microphone array to the reflector” specified by the index g is the optimum reflector arrangement condition for the microphone array. As determined.

<Introduction of distance>
In the above description, in each design method, only the target direction is considered, but the distance to the sound source (in the case of audio reproduction using a speaker array as will be described later) is also considered. It is also possible to design a filter. In this case, in each design method, and the distance from the center of the microphone array expressed as D (in particular describes the distance to the target direction and D _h), the above equation is modified as follows.

<1> In the case of the filter design method using the minimum variance distortion-free response method (4):

Formula (5), Formula (6):

Formula (7):

Expression (8a), Expression (8b): A set to which the index z of the arrival distance of noise belongs is {1, 2,..., Z-1}. It is assumed that the index h of the target distance D _h does not belong to the set {1, 2,..., Z-1}. Further, if the union of the set {1, 2,..., Z-1} and the set {h} is Γ, | Γ | = Z. | Γ | represents the number of elements of the set Γ.

Formula (9a), Formula (9b):

Formula (10a), Formula (10b): However, it is an example in case a sound wave arrives as a spherical wave. m is an integer satisfying 1 ≦ m ≦ M. c represents the speed of sound. j is an imaginary unit. In an appropriately set spatial coordinate system, v ^→ _{θ, D} ^(d) represents the position vector of the position (θ, D), and u ^→ _m represents the position vector of the m-th microphone. The symbols ‖ and ‖ represent the norm. f (‖v ^→ _{θ, D} ^(d) -u ^→ _m ‖) is a function representing the attenuation of the sound wave distance. For example _{^{f (‖v → θ, D (}} d) -u → m ||) = 1 / ‖v ^→ _θ, a _D ^(d) -u ^→ _m ‖ (see equation after substitution (10b)).

Formula (11c), Formula (11d): However, it is an example in case a sound wave arrives as a spherical wave. m is an integer satisfying 1 ≦ m ≦ M. c represents the speed of sound. j is an imaginary unit. In the above spatial coordinate system, v ^→ _{θ, D} ^(ξ) is the position vector where the position (θ, D) is moved to the mirror image object on the reflecting surface of the ξth reflector, u ^→ _m is the mth Represents a microphone position vector. The symbols ‖ and ‖ represent the norm. f (‖v ^→ _{θ, D} ^(ξ) -u ^→ _m ‖) is a function that represents the distance attenuation of the sound wave. For example _{^{f (‖v → θ, D (}} ξ) -u → m ||) = 1 / ‖v ^→ _θ, a ^{_{D (ξ)}} -u ^→ _m ‖ (see equation after substitution (11d)).

Formula (12):

Formula (13): When there are a plurality of target distances, the plurality of target distances are set as D _h1 ,..., Θ _hB . The total number | {D _h1 ,..., Θ _hB } | = B. However, the total number thereof | {D _h1 ,..., Θ _hB } | = B does not exceed Z.

Formula (14), Formula (15), Formula (16):

Formula (17):

Formula (18):

Formula (19):

Formula (20):

Formula (21):

Formula (22), Formula (23):

Formula (24):

Formula (25):

Formula (26):

  In the above description, in any of the design methods, it is assumed that sound is collected by a microphone array. However, the same argument holds even when sound is reproduced by a speaker array. Note that “dual sound” is defined in order to consider the reflected sound in the case of audio reproduction. (1) Voice radiated from a speaker array, and (2) voice that satisfies the condition that the voice is reflected by a reflector and the reflected sound travels in the target direction is called “dual sound”. (See FIGS. 12 and 13). In the above description on the assumption that sound is collected by a microphone array, the microphone array may be read as a speaker array, noise as a leaked sound, and reflected sound as a dual sound.

Hereinafter, application modes of the present invention will be described. The outline of the application form is as follows.
Application form 1:
With respect to the sound collected by the microphone array, the sound in a desired direction is emphasized in a narrow direction.
Application form 2:
With respect to the sound picked up by the microphone array, the sound in a desired direction and distance is emphasized in a narrow direction.
Application form 3:
Sound is reproduced in a narrow direction in a desired direction with a speaker array.
Application form 4:
A speaker array is used to spot-reproduce sound in a narrow direction in a desired direction and distance.

<< Applicable form 1 >>
The functional configuration and processing flow of application form 1 are shown in FIGS. The speech processing apparatus 1 according to the application form 1 includes an AD conversion unit 210, a frame generation unit 220, a frequency domain conversion unit 230, a filter application unit 240, a time domain conversion unit 250, a filter design unit 260, and a storage unit 290.

  First, the position of the reflector is determined by the above-described embodiment of the present invention. Subsequently, the following processing continues (see also FIGS. 10 and 11).

Step S1
In advance, the filter design unit 260 calculates a filter W ^→ (ω, θ _i ) for each frequency for each discrete direction that can be a target of speech enhancement. The total number of I discrete directions that may be subject to speech enhancement (I is 1 or more predetermined integer, satisfying the I ≦ P) _{^{When, W → (ω, θ 1}} ), ..., W → (ω, θ _i ),..., W ^→ (ω, θ _I ) (1 ≦ i ≦ I, ω∈Ω; i is an integer, Ω is a set of frequencies ω) is calculated in advance. For this purpose, transfer characteristic a ^→ (ω, θ _i ) = [a ₁ (ω, θ _i ), ..., a _M (ω, θ _i )] ^T (1 ≦ i ≦ I, ω∈Ω) Although it is necessary to find out, this is the arrangement of the microphones in the microphone array, the positional relationship of the reflectors such as the reflector with respect to the microphone array (this has already been determined), the direct sound and the ξth (1 ≦ ξ ≦ Ξ ) Can be concretely calculated by the equation (9a) based on the environmental information such as the arrival time difference from the reflected sound and the reflectance of the sound of the reflector (exactly, θ in the equation (9a) is θ _i Is). The number of the reflected sounds is set to an integer satisfying 1 ≦ Ξ, but according to the above-described embodiment, Ξ = 1, and since one reflector 300 is installed in the vicinity of the microphone array, the transfer characteristic a ^→ ( (ω, θ _i ) can be specifically calculated by equation (9b) (more precisely, θ in equation (9b) is θ _i ). For the calculation of the steering vector, for example, Expression (10a), Expression (10b), Expression (11a), Expression (11b), Expression (11c), and Expression (11d) can be used. In addition, you may use the transfer characteristic obtained by actual measurement in a real environment, for example, without depending on Formula (9a) and Formula (9b). Then, using the transfer characteristic a ^→ (ω, θ _i ), for example, W ^→ (ω, θ _i ) (1) according to any one of the equations (7), (18), (19), and (24). ≦ i ≦ I) is obtained. In addition, when using Formula (7), Formula (19), or Formula (24), the spatial correlation matrix Q (ω) (or R _xx (ω)) can be calculated by Formula (8b). When using equation (18), the spatial correlation matrix R _nn (ω) can be calculated by equation (16). I × | Ω | filters W ^→ (ω, θ _i ) (1 ≦ i ≦ I, ω∈Ω) are stored in the storage unit 290. | Ω | represents the number of elements of the set Ω.

Step S2
Sound is picked up using M microphones 200-1,..., 200-M constituting the microphone array. M is an integer of 2 or more.

There is no limit to the way the M microphones are arranged. However, by arranging M microphones two-dimensionally or three-dimensionally, there is an advantage that there is no uncertainty in the direction of voice enhancement. In other words, when M microphones are arranged in a straight line in the horizontal direction, for example, it is impossible to distinguish between voices coming from the front direction and voices coming from directly above. Can be prevented. In addition, in order to take a wide range of directions that can be set as the sound collection direction, the directivity of each microphone is a directivity capable of collecting sound with a certain sound pressure in a direction that can be the target direction θ _s that is the sound collection direction. It is better to have Therefore, a microphone having a relatively gentle directivity such as an omnidirectional microphone or a unidirectional microphone is preferable.

Step S3
The AD converter 210 converts an analog signal (sound collected signal) collected by the M microphones 200-1,..., 200-M into a digital signal x ^→ (t) = [x ₁ (t),. _M (t)] Convert to ^T. t represents a discrete time index.

Step S4
The frame generation unit 220 receives the digital signal x ^→ (t) = [x ₁ (t),..., X _M (t)] ^T output from the AD conversion unit 210 and stores N samples in a buffer for each channel. Thus, the digital signal x ^→ (k) = [x ^→ ₁ (k),..., X ^→ _M (k)] ^T is output in frame units. k is an index of the frame number. x ^→ _m (k) = [x _m ((k−1) N + 1),..., x _m (kN)] (1 ≦ m ≦ M). N depends on the sampling frequency, but in the case of 16kHz sampling, around 512 points is reasonable.

Step S5
The frequency domain transform unit 230 converts the digital signal x ^→ (k) of each frame into the frequency domain signal X ^→ (ω, k) = [X ₁ (ω, k),..., X _M (ω, k)] ^T Convert to and output. ω is an index of discrete frequency. One method for converting a time domain signal to a frequency domain signal is a fast discrete Fourier transform, but the present invention is not limited to this, and other methods for converting to a frequency domain signal may be used. The frequency domain signal X ^→ (ω, k) is output for each frequency ω and for each frame k.

Step S6
The filter application unit 240 changes the frequency domain signal X ^→ (ω, k) = [X ₁ (ω, k),..., X _M (ω, k)] ^T for each frequency ω∈Ω for each frame k. Then, an output signal Y (ω, k, θ _s ) is output by applying the filter W ^→ (ω, θ _s ) corresponding to the target direction θ _s to be emphasized (see Expression (27)). Since the index s of the target direction θ _s is s∈ {1,..., I}, and the filter W ^→ (ω, θ _s ) is stored in the storage unit 290, for example, every time the process of step S6 is performed, The filter application unit 240 may acquire the filter W ^→ (ω, θ _s ) corresponding to the target direction θ _s to be emphasized from the storage unit 290. When the index s of the target direction θ _s does not belong to the set {1,..., I}, that is, when the filter W ^→ (ω, θ _s ) corresponding to the target direction θ _s is not calculated in the process of step S1. The filter design unit 260 may calculate the filter W ^→ (ω, θ _s ) corresponding to the target direction θ _s temporarily, or the filter W ^→ ( ^→ corresponding to the direction θ _{s ′} close to the target direction θ _s (ω, θ _{s ′} ) may be used.

Step S7
The time domain transform unit 250 transforms the output signal Y (ω, k, θ _s ) of each frequency ω∈Ω of the kth frame into the time domain to obtain a frame unit time domain signal y (k) of the kth frame. Further, the obtained frame unit time domain signal y (k) is connected in the order of the frame number index to output the time domain signal y (t) in which the voice in the target direction θ _s is emphasized. The method for converting the frequency domain signal into the time domain signal is an inverse transformation corresponding to the transformation method used in the process of step S5, for example, a fast discrete inverse Fourier transform.

Here, the embodiment in which the filter W ^→ (ω, θ _i ) is calculated in advance in the process of step S1 has been described, but the target direction θ _s is determined according to the calculation processing capability of the narrow directivity speech enhancement apparatus 1 and the like. An embodiment in which the filter design unit 260 calculates the filter W ^→ (ω, θ _s ) for each frequency after it is determined can also be adopted.

<< Applicable form 2 >>
The functional configuration and processing flow of application mode 2 are shown in FIGS. The speech processing apparatus 2 according to Application 2 includes an AD conversion unit 210, a frame generation unit 220, a frequency domain conversion unit 230, a filter application unit 240, a time domain conversion unit 250, a filter design unit 260, and a storage unit 290.

  First, the position of the reflector is determined by the above-described embodiment of the present invention. Subsequently, the following processing continues (see also FIGS. 10 and 11). The formula quoted in the application form 2 is the formula in the <Introduction of distance> column.

Step S1
The filter design unit 260 calculates the filter W ^→ (ω, θ _i , D _g ) for each frequency in advance for each discrete position (θ _i , D _g ) that can be the target of speech enhancement. The total number of discrete directions that can be the target of speech enhancement is I (I is a predetermined integer of 1 or more and satisfies I ≦ P), and the total number of discrete distances is G (G is 1 or more in advance). W ^→ (ω, θ ₁ , D ₁ ), ..., W ^→ (ω, θ _i , D ₁ ), ..., W ^→ (ω, θ _I , D ₁ ), W ^→ (ω, θ ₁ , D ₂ ),…, W ^→ (ω, θ _i , D ₂ ),…, W ^→ (ω, θ _I , D ₂ ),…, W ^→ (ω, θ ₁ , D _g ), ..., W ^→ (ω, θ _i , D _g ), ..., W ^→ (ω, θ _I , D _g ), ..., W ^→ (ω, θ ₁ , D _G ), ..., W ^→ (ω, θ _i , D _G ), ..., W ^→ (ω, θ _I , D _G ) (1 ≦ i ≦ I, 1 ≦ g ≦ G, ω∈Ω; i is an integer, (Ω is a set of frequencies ω) is calculated in advance. For this purpose, the transfer characteristic a ^→ (ω, θ _i , D _g ) = [a ₁ (ω, θ _i , D _g ), ..., a _M (ω, θ _i , D _g )] ^T (1 ≦ i ≤ I, 1 ≤ g ≤ G, ω ∈ Ω), which is determined by the arrangement of the microphones in the microphone array and the positional relationship of the reflectors such as the reflector to the microphone array (this has already been determined) It can be concretely calculated by formula (9a) based on environmental information such as the arrival time difference between the direct sound and the ξ-th (1 ≦ ξ ≦ Ξ) reflected sound and the reflectance of the sound of the reflector (exact) In the equation (9a), θ _i is θ _i and D is D _g ). The number of the reflected sounds is set to an integer satisfying 1 ≦ Ξ, but according to the above-described embodiment, Ξ = 1, and since one reflector 300 is installed in the vicinity of the microphone array, the transfer characteristic a ^→ ( (ω, θ _i ) can be specifically calculated by equation (9b) (more precisely, θ in equation (9b) is θ _i ). For the calculation of the steering vector, for example, Expression (10a), Expression (10b), Expression (11a), Expression (11b), Expression (11c), and Expression (11d) can be used. In addition, you may use the transfer characteristic obtained by actual measurement in a real environment, for example, without relying on Formula (9a). Then, using the transfer characteristic a ^→ (ω, θ _i , D _g ), for example, W ^→ (ω, θ _i according to any one of Expression (7), Expression (18), Expression (19), and Expression (24). , D _g ) (1 ≦ i ≦ I, 1 ≦ g ≦ G). In addition, when using Expression (7), Expression (19), or Expression (24), the spatial correlation matrix Q (ω, D _g ) (or R _xx (ω, D _g )) can be calculated by Expression (8b). (To be exact, D in the formula (8b) is D _g ). When using equation (18), the spatial correlation matrix R _nn (ω, D _g ) can be calculated by equation (16) (exactly, D in equation (16) is D _g ). The I × G × | Ω | filters W ^→ (ω, θ _i , D _g ) (1 ≦ i ≦ I, 1 ≦ g ≦ G, ω∈Ω) are stored in the storage unit 290. | Ω | represents the number of elements of the set Ω.

Step S2
Sound is picked up using M microphones 200-1,..., 200-M constituting the microphone array. M is an integer of 2 or more.

There is no limit to the way the M microphones are arranged. However, by arranging M microphones two-dimensionally or three-dimensionally, there is an advantage that there is no uncertainty in the direction of voice enhancement. In other words, when M microphones are arranged in a straight line in the horizontal direction, for example, it is impossible to distinguish between voices coming from the front direction and voices coming from directly above. Can be prevented. In addition, in order to take a wide range of directions that can be set as the sound collection direction, the directivity of each microphone is a directivity capable of collecting sound with a certain sound pressure in a direction that can be the target direction θ _s that is the sound collection direction. It is better to have Therefore, a microphone having a relatively gentle directivity such as an omnidirectional microphone or a unidirectional microphone is preferable.

Step S3
The AD converter 210 converts an analog signal (sound collected signal) collected by the M microphones 200-1,..., 200-M into a digital signal x ^→ (t) = [x ₁ (t),. _M (t)] Convert to ^T. t represents a discrete time index.

Step S4
The frame generation unit 220 receives the digital signal x ^→ (t) = [x ₁ (t),..., X _M (t)] ^T output from the AD conversion unit 210 and stores N samples in a buffer for each channel. Thus, the digital signal x ^→ (k) = [x ^→ ₁ (k),..., X ^→ _M (k)] ^T is output in frame units. k is an index of the frame number. x ^→ _m (k) = [x _m ((k−1) N + 1),..., x _m (kN)] (1 ≦ m ≦ M). N depends on the sampling frequency, but in the case of 16kHz sampling, around 512 points is reasonable.

Step S5
The frequency domain transform unit 230 converts the digital signal x ^→ (k) of each frame into the frequency domain signal X ^→ (ω, k) = [X ₁ (ω, k),..., X _M (ω, k)] ^T Convert to and output. ω is an index of discrete frequency. One method for converting a time domain signal to a frequency domain signal is a fast discrete Fourier transform, but the present invention is not limited to this, and other methods for converting to a frequency domain signal may be used. The frequency domain signal X ^→ (ω, k) is output for each frequency ω and for each frame k.

Step S6
The filter application unit 240 changes the frequency domain signal X ^→ (ω, k) = [X ₁ (ω, k),..., X _M (ω, k)] ^T for each frequency ω∈Ω for each frame k. Apply the filter W ^→ (ω, θ _s , D _h ) corresponding to the position (θ _s , D _h ) to be emphasized, and output the output signal Y (ω, k, θ _s , D _h ) (formula (Refer to (28)). The indices s, h at the position (θ _s , D _h ) are s∈ {1,…, I}, h∈ {1,…, G}, and the filter W ^→ (ω, θ _s , D _h ) is Since it is stored in the storage unit 290, for example, each time the process of step S6, the filter application unit 240 filters W ^→ (ω, θ _s , D _h ) corresponding to the position (θ _s , D _h ) to be emphasized. May be acquired from the storage unit 290. If the index s in the direction θ _s does not belong to the set {1, ..., I} or the index h in the distance D _h does not belong to the set {1, ..., G}, that is, the position (θ _s , D _h ) If the filter W ^→ (ω, θ _s , D _h ) corresponding to い な い is not calculated in the process of step S1, the filter W ^→ (ω, θ _s , D) corresponding to the temporary position (θ _s , D _h ) it the _h) may be calculated in the filter design unit 260, or the filter corresponds to the _'distance D _h near and distance D _h' direction theta _s close to the direction ^{_{θ s W → (ω, θ}} s', D h ), W ^→ (ω, θ _s , D _{h ′} ) or W ^→ (ω, θ _{s ′} , D _{h ′} ).

Step S7
The time domain conversion unit 250 converts the output signal Y (ω, k, θ _s , D _h ) of each frequency ω∈Ω of the kth frame into the time domain, and converts the frame unit time domain signal y (k) of the kth frame. ), And the obtained frame unit time domain signal y (k) is concatenated in the order of the index of the frame number, and the time domain signal y () in which the speech from the position (θ _s , D _h ) is enhanced. t) is output. The method for converting the frequency domain signal into the time domain signal is an inverse transformation corresponding to the transformation method used in the process of step S5, for example, a fast discrete inverse Fourier transform.

Here, the embodiment in which the filter W ^→ (ω, θ _i , D _g ) is calculated in advance in the process of step S1 has been described, but the position (θ _s ) is determined according to the calculation processing capability of the speech processing device 2 and the like. , D _h ) can be adopted, and the filter design unit 260 can calculate the filter W ^→ (ω, θ _s , D _h ) for each frequency.

<< Applicable form 3 >>
A functional configuration and a processing flow of the application form 3 are shown in FIGS. The speech processing apparatus 3 according to the application form 3 includes an AD conversion unit 210, a frame generation unit 220, a frequency domain conversion unit 230, a filter application unit 240, a time domain conversion unit 250, a filter design unit 260, and a storage unit 290.

  First, the position of the reflector is determined by the above-described embodiment of the present invention. Subsequently, the following processing continues (see also FIGS. 12 and 13).

Step S1
In advance, the filter design unit 260 calculates a filter W ^→ (ω, θ _i ) for each frequency for each discrete direction that can be a target of audio reproduction. The total number of I discrete directions that may be subject to audio playback (I is one or more predetermined integer, satisfying the I ≦ P) _{^{When, W → (ω, θ 1}} ), ..., W → (ω, θ _i ),..., W ^→ (ω, θ _I ) (1 ≦ i ≦ I, ω∈Ω; i is an integer, Ω is a set of frequencies ω) is calculated in advance. For this purpose, transfer characteristic a ^→ (ω, θ _i ) = [a ₁ (ω, θ _i ), ..., a _M (ω, θ _i )] ^T (1 ≦ i ≦ I, ω∈Ω) Although it is necessary to obtain this, the speaker arrangement in the speaker array, the positional relationship of the reflector, for example, the position of the reflector with respect to the speaker array (which has already been determined), the direct sound and the ξth (1 ≦ ξ ≦ Ξ ) Can be concretely calculated by the equation (9a) based on the environmental information such as the time difference from the dual sound and the reflectance of the sound of the reflector (more precisely, θ in the equation (9a) is θ _i. is there). The number of dual sounds 音 is set to an integer satisfying 1 ≦ Ξ. However, according to the above-described embodiment, Ξ = 1, and since one reflector 300 is installed in the vicinity of the microphone array, the transfer characteristic a ^→ ( (ω, θ _i ) can be specifically calculated by equation (9b) (more precisely, θ in equation (9b) is θ _i ). For the calculation of the steering vector, for example, Expression (10a), Expression (10b), Expression (11a), Expression (11b), Expression (11c), and Expression (11d) can be used. In addition, you may use the transfer characteristic obtained by actual measurement in a real environment, for example, without depending on Formula (10a) and Formula (10b). Then, using the transfer characteristic a ^→ (ω, θ _i ), for example, W ^→ (ω, θ _i ) (1) according to any one of the equations (7), (18), (19), and (24). ≦ i ≦ I) is obtained. In addition, when using Formula (7), Formula (19), or Formula (24), the spatial correlation matrix Q (ω) (or R _xx (ω)) can be calculated by Formula (8b). When using equation (18), the spatial correlation matrix R _nn (ω) can be calculated by equation (16). I × | Ω | filters W ^→ (ω, θ _i ) (1 ≦ i ≦ I, ω∈Ω) are stored in the storage unit 290. | Ω | represents the number of elements of the set Ω.

Step S2
The sound source 200 outputs a sound source signal ss (t). In this embodiment, it is assumed that the sound source signal ss (t) from the sound source 200 is an analog signal. However, a digital signal can also be used as the sound source signal.

Step S3
The AD converter 210 AD converts the sound source signal ss (t) into a digital signal s (t). Here, t represents an index of discrete time. When the digital signal is a sound source signal, it is not necessary to perform the process of step S3, and the sound source signal can be regarded as s (t) that is an output signal of the AD conversion unit 210.

Step S4
The frame generation unit 220 receives the digital signal s (t) output from the AD conversion unit 210, stores N samples in a buffer, and outputs a digital signal s (k) in units of frames. k is an index of the frame number. s (k) = [s ((k−1) N + 1),..., s (kN)]. N depends on the sampling frequency, but in the case of 16kHz sampling, around 512 points is reasonable.

Step S5
The frequency domain converter 230 converts the digital signal s (k) of each frame into a frequency domain signal S (ω, k) and outputs it. ω is an index of discrete frequency. One method for converting a time domain signal to a frequency domain signal is a fast discrete Fourier transform, but the present invention is not limited to this, and other methods for converting to a frequency domain signal may be used. The frequency domain signal S (ω, k) is output for each frequency ω and for each frame k.

Step S6
The filter application unit 240 applies the filter W ^→ (ω, θ _s ) corresponding to the desired direction θ _s to be reproduced to the frequency domain signal S (ω, k) for each frequency ω∈Ω for each frame k. Thus, the reproduction signal X ^→ (ω, k) = [X ₁ (ω, k),..., X _M (ω, k)] ^T is output (see Expression (29)). Since the index s of the target direction θ _s is s∈ {1,..., I}, and the filter W ^→ (ω, θ _s ) is stored in the storage unit 290, for example, every time the process of step S6 is performed, The filter application unit 240 may acquire the filter W ^→ (ω, θ _s ) corresponding to the target direction θ _s to be reproduced from the storage unit 290. When the index s of the target direction θ _s does not belong to the set {1,..., I}, that is, when the filter W ^→ (ω, θ _s ) corresponding to the target direction θ _s is not calculated in the process of step S1. The filter design unit 260 may calculate the filter W ^→ (ω, θ _s ) corresponding to the target direction θ _s temporarily, or the filter W ^→ ( ^→ corresponding to the direction θ _{s ′} close to the target direction θ _s (ω, θ _{s ′} ) may be used.

Step S7
The time domain transform unit 250 converts the reproduction signal X of each frequency ω∈Ω of the k-th frame ^→ (ω, k) = [X ₁ (ω, k),..., X _M (ω, k)] ^T into the time domain. To obtain a frame unit time domain signal x ^→ (k) = [x ₁ (k),..., X _M (k)] ^T of the k-th frame, and further obtain the frame unit time domain signal x ^→ (k) = [x ₁ (k),..., X _M (k)] A time domain in which speech is emphasized toward the target direction θ _s that is the playback direction by concatenating ^T in the order of the frame number index. Signal x ^→ (t) = [x ₁ (t),..., X _M (t)] ^T is output. The method for converting the frequency domain signal into the time domain signal is an inverse transformation corresponding to the transformation method used in the process of step S5, for example, a fast discrete inverse Fourier transform.

Step S8
M channel time domain signals x ₁ (t),..., X _M (t) are reproduced by the speakers corresponding to the channel among the M speakers 280-1,. Is done. That is, the time domain signal x _m (t) of the m-th (1 ≦ m ≦ M) channel is reproduced by the m-th speaker 280-m.

There is no limit to the way the M speakers are arranged. An array configuration in which speakers are linearly arranged like a linear speaker array may be used, or an array configuration in which M speakers are two-dimensionally or three-dimensionally arranged. In addition, in order to make the direction that can be set as the reproduction direction wide, the directivity of each speaker is such that the sound can be reproduced with a certain sound pressure in the direction that can be the target direction θ _s that is the reproduction direction. You should have it. Therefore, a speaker having relatively gentle directivity such as a non-directional speaker or a unidirectional speaker is preferable.

Here, the embodiment in which the filter W ^→ (ω, θ _i ) is calculated in advance in the process of step S1 has been described. However, the target direction θ that is the reproduction direction depends on the calculation processing capability of the audio processing device 3 and the like. _An embodiment in which the filter design unit 260 calculates the filter W ^→ (ω, θ _s ) for each frequency after _s is determined may be employed.

<< Applicable form 4 >>
The functional configuration and processing flow of application form 4 are shown in FIGS. The speech processing apparatus 4 according to Application Mode 4 includes an AD conversion unit 210, a frame generation unit 220, a frequency domain conversion unit 230, a filter application unit 240, a time domain conversion unit 250, a filter design unit 260, and a storage unit 290.

  First, the position of the reflector is determined by the above-described embodiment of the present invention. Subsequently, the following processing continues (see also FIGS. 12 and 13). The formula quoted in the application form 4 is the formula in the <Introduction of distance> column.

Step S1
The filter design unit 260 calculates the filter W ^→ (ω, θ _i , D _g ) for each frequency in advance for each discrete position (θ _i , D _g ) that can be an audio spot reproduction target. The total number of discrete directions that can be the target of audio spot reproduction is I (I is a predetermined integer of 1 or more and satisfies I ≦ P), and the total number of discrete distances is G (G is 1 or more). a predetermined integer, when satisfy ^{G ≦ Z), W → (} ω, θ 1, D 1), ..., W → (ω, θ i, D 1), ..., W → (ω, θ _I , D ₁ ), W ^→ (ω, θ ₁ , D ₂ ), ..., W ^→ (ω, θ _i , D ₂ ), ..., W ^→ (ω, θ _I , D ₂ ), ..., W ^→ (ω, θ ₁ , D _g ), ..., W ^→ (ω, θ _i , D _g ), ..., W ^→ (ω, θ _I , D _g ), ..., W ^→ (ω, θ ₁ , D _G ), ..., W ^→ (ω, θ _i , D _G ), ..., W ^→ (ω, θ _I , D _G ) (1 ≦ i ≦ I, 1 ≦ g ≦ G, ω∈Ω; i is an integer , Ω is a set of frequencies ω) in advance. For this purpose, the transfer characteristic a ^→ (ω, θ _i , D _g ) = [a ₁ (ω, θ _i , D _g ), ..., a _M (ω, θ _i , D _g )] ^T (1 ≦ i ≤ I, 1 ≤ g ≤ G, ω ∈ Ω), which is determined by the arrangement of the speakers in the speaker array and the positional relationship of the reflector, for example, the reflector plate (this is already determined) It can be concretely calculated by equation (9a) based on environmental information such as the time difference between the direct sound and the ξth (1 ≦ ξ ≦ Ξ) dual sound, and the reflectance of the sound of the reflector (exactly In the formula (9a) is θ _i and D is D _g ). The number of dual sounds 音 is set to an integer satisfying 1 ≦ Ξ. However, according to the above-described embodiment, Ξ = 1, and since one reflector 300 is installed in the vicinity of the microphone array, the transfer characteristic a ^→ ( (ω, θ _i ) can be specifically calculated by equation (9b) (more precisely, θ in equation (9b) is θ _i ). For the calculation of the steering vector, for example, Expression (10a), Expression (10b), Expression (11a), Expression (11b), Expression (11c), and Expression (11d) can be used. In addition, you may use the transfer characteristic obtained by actual measurement in a real environment, for example, without depending on Formula (10a) and Formula (10b). Then, using the transfer characteristic a ^→ (ω, θ _i , D _g ), for example, W ^→ (ω, θ _i according to any one of Expression (7), Expression (18), Expression (19), and Expression (24). , D _g ) (1 ≦ i ≦ I, 1 ≦ g ≦ G). In addition, when using Expression (7), Expression (19), or Expression (24), the spatial correlation matrix Q (ω, D _g ) (or R _xx (ω, D _g )) can be calculated by Expression (8b). (To be exact, D in the formula (8b) is D _g ). When using equation (18), the spatial correlation matrix R _nn (ω, D _g ) can be calculated by equation (16) (exactly, D in equation (16) is D _g ). The I × G × | Ω | filters W ^→ (ω, θ _i , D _g ) (1 ≦ i ≦ I, 1 ≦ g ≦ G, ω∈Ω) are stored in the storage unit 290. | Ω | represents the number of elements of the set Ω.

Step S2
The sound source 200 outputs a sound source signal ss (t). In this embodiment, it is assumed that the sound source signal ss (t) from the sound source 200 is an analog signal. However, a digital signal can also be used as the sound source signal.

Step S3
The AD converter 210 AD converts the sound source signal ss (t) into a digital signal s (t). Here, t represents an index of discrete time. When the digital signal is a sound source signal, it is not necessary to perform the process of step S3, and the sound source signal can be regarded as s (t) that is an output signal of the AD conversion unit 210.

Step S4
The frame generation unit 220 receives the digital signal s (t) output from the AD conversion unit 210, stores N samples in a buffer, and outputs a digital signal s (k) in units of frames. k is an index of the frame number. s (k) = [s ((k−1) N + 1),..., s (kN)]. N depends on the sampling frequency, but in the case of 16kHz sampling, around 512 points is reasonable.

Step S5
The frequency domain converter 230 converts the digital signal s (k) of each frame into a frequency domain signal S (ω, k) and outputs it. ω is an index of discrete frequency. One method for converting a time domain signal to a frequency domain signal is a fast discrete Fourier transform, but the present invention is not limited to this, and other methods for converting to a frequency domain signal may be used. The frequency domain signal S (ω, k) is output for each frequency ω and for each frame k.

Step S6
For each frequency ω∈Ω, the filter application unit 240 filters the frequency W of the frequency domain signal S (ω, k) corresponding to the position (θ _s , D _h ) to be spot-reproduced for each frequency ω∈Ω ^→ (ω, θ _{s 1} , D _h ), and the reproduction signal X ^→ (ω, k) = [X ₁ (ω, k),..., X _M (ω, k)] ^T is output (see Expression (30)). . The indices s, h at the position (θ _s , D _h ) are s∈ {1,…, I}, h∈ {1,…, G}, and the filter W ^→ (ω, θ _s , D _h ) is Since it is stored in the storage unit 290, for example, each time the process of step S6, the filter application unit 240 filters W ^→ (ω, θ _s , D _h ) corresponding to the position (θ _s , D _h ) at which spot reproduction is desired. ) From the storage unit 290. If the index s in the direction θ _s does not belong to the set {1, ..., I} or the index h in the distance D _h does not belong to the set {1, ..., G}, that is, the position (θ _s , D _h ) If the filter W ^→ (ω, θ _s , D _h ) corresponding to い な い is not calculated in the process of step S1, the filter W ^→ (ω, θ _s , D) corresponding to the temporary position (θ _s , D _h ) it the _h) may be calculated in the filter design unit 260, or the filter corresponds to the _'distance D _h near and distance D _h' direction theta _s close to the direction ^{_{θ s W → (ω, θ}} s', D h ), W ^→ (ω, θ _s , D _{h ′} ) or W ^→ (ω, θ _{s ′} , D _{h ′} ).

Step S7
The time domain transform unit 250 converts the reproduction signal X of each frequency ω∈Ω of the k-th frame ^→ (ω, k) = [X ₁ (ω, k),..., X _M (ω, k)] ^T into the time domain. To obtain a frame unit time domain signal x ^→ (k) = [x ₁ (k),..., X _M (k)] ^T of the k-th frame, and further obtain the frame unit time domain signal x ^→ (k) = [x ₁ (k), ..., x _M (k)] The speech is emphasized toward the position (θ _s , D _h ) where spot playback is desired by concatenating ^T in the order of the frame number index. Time domain signal x ^→ (t) = [x ₁ (t),..., X _M (t)] ^T is output. The method for converting the frequency domain signal into the time domain signal is an inverse transformation corresponding to the transformation method used in the process of step S5, for example, a fast discrete inverse Fourier transform.
Step S8
M channel time domain signals x ₁ (t),..., X _M (t) are reproduced by the speakers corresponding to the channel among the M speakers 280-1,. Is done. That is, the time domain signal x _m (t) of the m-th (1 ≦ m ≦ M) channel is reproduced by the m-th speaker 280-m.

There is no limit to the way the M speakers are arranged. An array configuration in which speakers are linearly arranged like a linear speaker array may be used, or an array configuration in which M speakers are two-dimensionally or three-dimensionally arranged. In addition, in order to make the direction that can be set as the sound collection direction wide, the directivity of each speaker is the directivity that can reproduce sound with a certain sound pressure in the direction that can be the target direction θ _s that is the reproduction direction. It is better to have Therefore, a speaker having relatively gentle directivity such as a non-directional speaker or a unidirectional speaker is preferable.

Here, the embodiment in which the filter W ^→ (ω, θ _i , D _g ) is calculated in advance in the process of step S1 has been described, but the position (θ _s ) is determined according to the calculation processing capability of the speech processing device 4 and the like. , D _h ) can be adopted, and the filter design unit 260 can calculate the filter W ^→ (ω, θ _s , D _h ) for each frequency.

<Hardware configuration example of reflector arrangement determination device>
The reflector arrangement determining apparatus according to the above-described embodiment may include an input unit to which a keyboard or the like can be connected, an output unit to which a liquid crystal display or the like can be connected, a CPU (Central Processing Unit) [cache memory, or the like. ] RAM (Random Access Memory) or ROM (Read Only Memory) and external storage device as a hard disk, and data exchange between these input unit, output unit, CPU, RAM, ROM, and external storage device It has a bus that can be connected. Further, if necessary, the reflector arrangement determining device may be provided with a device (drive) that can read and write a storage medium such as a CD-ROM. A physical entity having such hardware resources includes a general-purpose computer.

  The external storage device of the reflector arrangement determination device stores a program for determining the arrangement of the reflector and data necessary for processing of the program [not limited to the external storage device, for example, reading the program. It may be stored in a ROM which is a dedicated storage device. ]. Data obtained by the processing of these programs is appropriately stored in a RAM or an external storage device. A storage device that stores data, addresses of storage areas, and the like is simply referred to as a “storage unit”.

  The storage unit of the reflector arrangement determination device stores a program for determining the arrangement of the reflector based on the candidates (JK) for the reflector arrangement, Equation (12), Equation (13), and the like. .

  In the reflector arrangement determining apparatus, each program stored in the storage unit and data necessary for processing each program are read into the RAM as necessary, and are interpreted and executed by the CPU. As a result, the CPU determines the arrangement of the reflecting object by realizing a predetermined function (arrangement determining unit).

  Also, the sound processing device can have the same hardware configuration, and the storage unit of the sound processing device has a program for obtaining a filter for each frequency using a spatial correlation matrix, and AD for analog signals. A program for performing conversion, a program for performing frame generation processing, a program for converting a digital signal for each frame into a frequency domain signal in the frequency domain, and a filter corresponding to a desired direction (and desired distance) for each frequency A program for obtaining an output signal by applying to a frequency domain signal and a program for converting the output signal into a time domain signal are stored.

  In the audio processing device, each program stored in the storage unit and data necessary for processing each program are read into the RAM as necessary, and are interpreted and executed by the CPU. As a result, the above-described sound processing is realized by the CPU realizing predetermined functions (filter design unit, AD conversion unit, frame generation unit, frequency domain conversion unit, filter application unit, time domain conversion unit).

<Supplementary note>
The present invention is not limited to the above-described embodiment, and can be appropriately changed without departing from the spirit of the present invention. In addition, the processing described in the above embodiment may be executed not only in time series according to the order of description but also in parallel or individually as required by the processing capability of the apparatus that executes the processing. .

  Further, when the processing functions in the hardware entity (reflecting object arrangement determination device / audio processing device) described in the above embodiment are realized by a computer, the processing contents of the functions that the hardware entity should have are described by a program. Then, by executing this program on a computer, the processing functions in the hardware entity are 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), a CD-ROM (Compact Disc Read Only). Memory), CD-R (Recordable) / RW (ReWritable), etc., magneto-optical recording medium, MO (Magneto-Optical disc), etc., semiconductor memory, EEP-ROM (Electronically Erasable and Programmable-Read Only Memory), etc. Can be used.

  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. Furthermore, the program may be distributed by storing the program in a storage device of the server computer and transferring the program from the server computer to another computer via a network.

  A computer that executes such a program first stores, for example, a program recorded on a portable recording medium or a program transferred from a server computer in its own storage device. When executing the process, the computer reads a program stored in its own recording medium and executes a process according to the read program. As another execution form of the program, the computer may directly read the program from a portable recording medium and execute processing according to the program, and the program is transferred from the server computer to the computer. Each time, the processing according to the received program may be executed sequentially. Also, the program is not transferred from the server computer to the computer, and the above-described processing is executed by a so-called ASP (Application Service Provider) type service that realizes the processing function only by the execution instruction and result acquisition. It is good. Note that the program in this embodiment includes information that is used for processing by an electronic computer and that conforms to the program (data that is not a direct command to the computer but has a property that defines the processing of the computer).

  In this embodiment, a hardware entity is configured by executing a predetermined program on a computer. However, at least a part of these processing contents may be realized by hardware.

Claims (6)

A filter applied to speech-based information is designed based on a predetermined evaluation function using a spatial correlation matrix represented by transfer characteristics in multiple directions in space,
Each of the above transfer characteristics is represented by the sum of the transfer characteristic of the direct sound and each transfer characteristic of one or more reflected sounds reflected by the reflector,
The direct sound transfer characteristic and the reflected sound transfer characteristic are determined based on information (hereinafter referred to as arrangement information) indicating an arrangement relationship of the reflector with respect to a microphone array or a speaker array.
The evaluation function is a function that takes a value that is so small that the voice in at least one target direction selected from the plurality of directions is emphasized,
The storage unit, the arrangement information is stored,
For each candidate of the reflector based on the arrangement information, the arrangement determining unit obtains the value of the evaluation function using the spatial correlation matrix represented by the transfer characteristic specified based on the candidate, and the value A method for determining the arrangement of reflectors, comprising: an arrangement determining step for determining a candidate corresponding to the smallest of the above as the arrangement of the reflectors. The arrangement determination method according to claim 1,
The method for determining the arrangement of reflectors is characterized in that the evaluation function is an evaluation function based on a minimum dispersion no distortion response method. The arrangement determination method according to claim 1,
The method for determining the arrangement of reflectors, wherein the evaluation function is an evaluation function based on an S / N ratio maximization criterion. The arrangement determination method according to claim 1,
The reflection function arrangement determination method, wherein the evaluation function is an evaluation function based on power inversion. A filter applied to speech-based information is designed based on a predetermined evaluation function using a spatial correlation matrix represented by transfer characteristics in multiple directions in space,
Each of the above transfer characteristics is represented by the sum of the transfer characteristic of the direct sound and each transfer characteristic of one or more reflected sounds reflected by the reflector,
The direct sound transfer characteristic and the reflected sound transfer characteristic are determined based on information (hereinafter referred to as arrangement information) indicating an arrangement relationship of the reflector with respect to a microphone array or a speaker array.
The evaluation function is a function that takes a value that is so small that the voice in at least one target direction selected from the plurality of directions is emphasized,
A storage unit for storing the configuration information,
For each candidate for the reflector based on the arrangement information, the value of the evaluation function is obtained using the spatial correlation matrix represented by the transfer characteristic specified based on the candidate, and the smallest value among the values is obtained. And an arrangement determining unit for determining a candidate corresponding to the above as an arrangement of the reflector.       The program for making a computer perform the process of the arrangement | positioning determination method of the reflector in any one of Claims 1-4.

Images