JP5105336B2 - Sound source separation apparatus, program and method - Google Patents

Description

  The present invention relates to a sound source separation device, a program, and a method, and can be applied to noise removal in speech capture such as a telephone device and a speech recognition device.

  In a telephone device or a voice recognition device, a user's voice is captured by a microphone, but the accuracy of voice recognition may be extremely deteriorated due to ambient noise, or the recorded voice may be difficult to hear due to noise.

  For this reason, in the past, attempts have been made to selectively capture only a desired target sound by controlling the directional characteristics using a microphone array. It was difficult to separate and extract from background noise.

  Conventional microphone array technologies include, for example, a delay-and-sum array (DSA), a technology related to directivity control called BF (Beam-Forming), or a DCMP (Directly Constrained Minimization of Power) adaptive array. There are technologies related to characteristic control.

  On the other hand, as a technology for separating speech by remote utterance, a technology (referred to as SAFIA) that performs narrowband spectrum analysis on the output signals of a plurality of fixed microphones and assigns the sound in that frequency band to the microphone that gives the largest amplitude for each frequency band. Is described in Patent Document 1. In the sound separation technology by band selection (BS) described in Patent Document 1, in order to obtain a desired sound, a microphone closest to a sound source that emits the desired sound is selected and assigned to the microphone. Synthesizes speech using sound in the frequency band.

  As a further technique, Patent Document 2 discloses a technique obtained by improving the band selection method.

  In the technique described in Patent Document 2, the target sound to be captured is suppressed by using the signals input to the two microphones arranged side by side in a direction perpendicular or substantially perpendicular to the direction of arrival of the target sound. And a target sound inferior signal in which the target sound is suppressed and the disturbance sound is emphasized, and the two kinds of signals are used for separation of the target sound and the interference sound.

  In Patent Document 2, the generation of the target sound superior signal and the target sound inferior signal is realized by using a filter called “spatial filter”.

  FIG. 3 is an explanatory diagram showing the characteristics of the spatial filter.

  Hereinafter, a vertical plane with respect to a line connecting the two microphones M1 and M2 is referred to as a 0 degree direction, and the clockwise direction is a positive angle and the counterclockwise direction is a negative angle. That is, the above-described direction is expressed in a range of −180 degrees to 180 degrees (−180 degrees and 180 degrees are the same direction).

FIG. 3 illustrates a case where there is a sound source that is input from the direction of the angle θ with respect to the two microphones M1 and M2 arranged at the interval d. In this case, a distance difference of d × sin θ occurs in the distance from the sound source input from the direction of the angle θ to the two microphones M1 and M2, and as a result, the sound arrival time is between the microphones M1 and M2. A time difference τ expressed by the following equation (1) is generated.

  Therefore, when the output obtained by delaying the output of the microphone M1 by the time difference τ is subtracted from the output of the microphone M2, they cancel each other and the sound in the θ direction is suppressed. Hereinafter, an angle (θ in the above example) in a direction in which sound is suppressed in the spatial filter is referred to as a “suppression angle”.

  FIG. 4 is an explanatory diagram showing directivity characteristics in the spatial filter.

  In FIG. 4, a curve L represents the directivity when the suppression angle of the spatial filter is set to θ, and the gain increases as the distance from the midpoint of the line connecting the microphones M1 and M2 increases. It indicates that the gain is small (the suppression intensity is strong) as it is large (the suppression intensity is weak) and the distance is short.

  In FIG. 4, since the suppression angle of the spatial filter is set in the direction of θ, the state in which the intensity of suppression in that direction is set to the maximum is shown.

Japanese Patent Laid-Open No. 10-313497 JP 2006-197552 A

  However, in the technique described in Patent Document 1, the two sounds can be separated well in a situation where two sounds overlap. However, if there are three or more sound sources, the separation is theoretically possible. Performance is extremely degraded. Therefore, it is difficult to accurately separate the target sound from the plurality of noises in a situation where there are a plurality of noise sources.

  In the technique described in Patent Document 2, the target sound is separated using the spatial filter. However, when the direction of arrival of the target sound is shifted during the separation process, the spatial filter is used. May affect the quality of the target sound after separation. Hereinafter, in the method described in Patent Document 2, the influence that the characteristics of the spatial filter may have on the target sound after separation will be described.

  FIG. 5 is an explanatory diagram showing the change characteristic of the gain in the direction close to the suppression angle in the spatial filter.

  In FIG. 5, the suppression angle of the spatial filter is θ, and the target sound arrives from a direction slightly shifted counterclockwise from G1 when the target sound arrives from the direction of 0 degrees (front). The case gain is described as G2.

  In the spatial filter, when the rate of change in gain according to the change in angle is large near the suppression angle, as shown in FIG. 5, the deviation of the angle between the direction in which the gain is G1 and the direction in which it is G2 Even if it is slight, the difference between G1 and G2 may become large.

  In the target sound inferior signal generating means described in Patent Document 2 described above, the target sound component is suppressed by directing the suppression angle of the spatial filter in the direction in which the target sound is expected to arrive, and the interference sound component is reduced. As described above, if the direction of arrival of the target sound is shifted during the process of separating the target sound and the interference sound as described above, even if there is a slight shift, the output sound will be greatly shaken. May result.

  Therefore, in the process of separating the target sound and the interfering sound coming from an arbitrary direction other than the direction of arrival of the target sound, the quality of the sound after the separation process can be maintained even when the direction of arrival of the target sound is deviated. A sound source separation device, a program, and a method that can be used are desired.

The sound source separation apparatus according to the first aspect of the present invention is (1) an assumed arrival that a target sound is assumed to arrive in the spectrum of the received signal of two microphones among a plurality of microphones arranged at intervals. The target sound component is suppressed from the spectrum of the received signal using a plurality of target sound suppression units that process the directivity of component suppression in different directions within a predetermined range including the direction. A target sound suppression spectrum generating means for generating a target sound suppression spectrum; and (2) generating a target sound dominant spectrum in which interference sound coming from an arbitrary direction other than the predetermined range is suppressed for the spectrum of the received signal. Using the target sound dominant spectrum generating means, (3) the target sound suppression spectrum, and the target sound dominant spectrum, Have a separating means for separating the components of the target sound, (4) the target sound suppression spectrum generating unit for each component of the target sound suppressed spectrum, the most absolute value of the processing result of the target sound suppressing unit It is characterized by applying a small value .

The sound source separation program according to the second aspect of the present invention provides a computer mounted on a sound source separation device, for (1) a spectrum of received sound signals of two microphones among a plurality of microphones arranged at intervals. A spectrum of the received signal using a plurality of target sound suppression units that process the directivity of component suppression in different directions within a predetermined range including an assumed direction of arrival where the target sound is expected to arrive. And (2) interference sound coming from an arbitrary direction other than the predetermined range for the spectrum of the received signal. A target sound dominance spectrum generating means for generating a target sound dominance spectrum with suppressed sound, and (3) a target sound suppression spectrum and a target sound dominance spectrum. For the received sound signal, to function as a separating means for separating the components of the component and the target sound of the interference sound, (4) the target sound suppression spectrum generating unit for each component of the target sound suppression spectrum, the A value having the smallest absolute value among the processing results of the target sound suppressing unit is applied .

According to a third aspect of the present invention, in the sound source separation method performed by the sound source separation device, (1) a target sound suppression spectrum generation unit, a target sound dominant spectrum generation unit, and a separation unit are provided, and (2) the target sound suppression spectrum generation is performed. The means is different from each other within a predetermined range including an assumed arrival direction in which the target sound is expected to be received, with respect to the spectrum of the received signal of two microphones among a plurality of microphones arranged at intervals. A target sound suppression spectrum in which the target sound component is suppressed is generated from the spectrum of the received signal using a plurality of target sound suppression units that process the directivity of component suppression in the direction, (3) The target sound dominance spectrum generating means suppresses the interference sound coming from an arbitrary direction other than the predetermined range in the spectrum of the received signal. Generates a spectrum, (4) the separating means uses a target sound suppressed spectrum, the target sound predominant spectrum for the received sound signal, to separate the components of the component and the target sound of the interference sound, (5) The target sound suppression spectrum generating means applies a value having the smallest absolute value among the processing results of the target sound suppression unit to each component of the target sound suppression spectrum .

  According to the present invention, in the process of separating the target sound and the disturbing sound coming from an arbitrary direction other than the direction of arrival of the target sound, the quality of the sound after the separation process even when the direction of arrival of the target sound is deviated. Can keep.

It is the block diagram shown about the functional structure of the sound source separation apparatus which concerns on 1st Embodiment. It is the block diagram shown about the functional structure of the sound source separation apparatus which concerns on 2nd Embodiment. It is explanatory drawing shown about the characteristic of the conventional spatial filter. It is explanatory drawing shown about the directional characteristic in the conventional spatial filter. It is explanatory drawing shown about the change characteristic of the gain of the direction close | similar to the suppression angle in the conventional spatial filter.

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

(A-1) Configuration and Operation of First Embodiment FIG. 1 is a block diagram showing a functional configuration of a sound source separation device 10 of the first embodiment.

  The sound source separation device 10 separates the target sound and the disturbing sound coming from an arbitrary direction other than the arrival direction of the target sound. The use of the sound source separation device 10 is not limited. For example, the sound source separation device 10 may be mounted on a voice recognition device or a telephone device such as a mobile phone and used for voice capture. Specifically, for example, the sound source separation device 10 is installed in a teleconference device, and a voice of an arbitrary speaker is separated as a target sound from a mixed voice of a plurality of speakers performing remote speech, or remote speech is performed. It may be used to separate the speaker's voice as the target sound from the mixed sound of the speaker's voice and other sounds. Further, for example, it may be used for separation of a user's voice, which is a target sound, in voice recognition for in-vehicle devices such as a robot that performs a voice dialogue, a car navigation system, and a meeting minutes.

  The sound source separation device 10 mainly includes an input unit 20, an analysis unit 30, a separation unit 40, a removal unit 50, and a generation unit 60.

  The sound source separation device 10 may be realized by installing the sound source separation program of the embodiment in a device having a processor (CPU or the like) regarding components other than hardware such as a microphone, or a part thereof. Alternatively, all the components may be realized using dedicated hardware (for example, a semiconductor chip).

  The input means 20 digitally converts two microphones 21 and 22 arranged at intervals, and the received signals of these two microphones 21 and 22 using an analog / digital signal converter (not shown). The signal is converted into a signal, and the digital signal is given to the analyzing means 30.

  In the following, as in FIGS. 3 to 5 described above, the vertical plane with respect to the line connecting the two microphones 21 and 22 is referred to as 0 degree direction, the clockwise direction is a positive angle, and the counterclockwise direction is negative. The direction is expressed as an angle. That is, the above-described direction is expressed in a range of −180 degrees to 180 degrees (−180 degrees and 180 degrees are the same direction).

  Further, hereinafter, as an example, the sound source separation device 10 will be described as a configuration assuming that the target sound comes from a direction of approximately 0 degrees.

  In the following description, the digital audio signal output from the microphone 21 is assumed to be x1 (n). Similarly, let the digital audio signal output from the microphone 22 be x2 (n). However, n represents the nth data (sample).

  The digital audio signals x1 (n) and x2 (n) are obtained, for example, by analog / digital conversion of an analog audio signal input from an audio input device such as a microphone and sampling at every sampling period T. Is. It is desirable that the sampling period T is, for example, about 31.25 microseconds to 125 microseconds.

  With the N consecutive x1 (n) and x2 (n) as one analysis unit (frame) in the same time interval, the analysis means 30, separation means 40, removal means 50, and generation means 60 described later are performed. Shall be.

  In the following description, the sound source separation apparatus 10 assumes N = 1024 as an example. When the sound source separation apparatus 10 completes a series of the sound source separation processes for the processing target analysis unit, 3N / 4 of the latter half of x1 (n) and x2 (n) are shifted to the first half, and a new It is assumed that N / 4 continuous data input to the terminal is connected to the latter half. Thus, the sound source separation apparatus 10 generates new N consecutive x1 (n) and x2 (n), and performs a new process as one analysis unit. It is assumed that the sound source separation apparatus 10 repeats such processing for the processing target analysis unit.

  Note that the digital audio signal input to the analyzing means 30 is not limited to the one obtained by the microphone and analog / digital converted. For example, it may be read from a recording medium or the like, or may be given by communication from another device. That is, in the sound source separation device 10, the input unit 20 may be omitted as long as x1 (n) and x2 (n) can be held.

When the digital voice signals x1 (n) and x2 (n) mixed with noise are given from the input means 20, the analyzing means 30 is supplied with x1 (n) by the frequency analyzing section 31 and x2 (n) by the frequency analyzing section. At 32, FFT (fast Fourier transform) processing or the like is performed, and the result is given to the separating means 40. In the analysis unit 30, in the FFT processing, a window function is applied to N consecutive x1 (n) and x2 (n). Various window functions can be applied as the window function w (n). For example, a Hanning window as shown in the following equation (2) may be applied.

  The window processing described above by the analysis unit 30 is a process performed in consideration of the analysis unit connection processing in the generation unit 60 described later. However, although it is preferable to apply the above window function, it is not essential.

  Below, the output of the frequency analysis parts 31 and 32 shall be represented as D1 (m) and D2 (m), respectively. D1 (m) and D2 (m) are complex numbers.

  The analysis method in the analysis means 30 is not limited to FFT, and other frequency analysis methods such as DFT (Discrete Fourier Transform) may be applied.

  Further, depending on the device on which the sound source separation device 10 is mounted, the configuration relating to the analysis in the other purpose processing device may be used as the configuration of the sound source separation device 10. For example, when the device on which the sound source separation device 10 is mounted is an IP telephone device, such diversion is possible. In the case of an IP telephone device, an encoded FFT output is inserted into the payload of an IP packet, and the FFT output can be used as the output of the analyzing means 30 described above.

  Further, in the process of the separating means 40 described later, the property D (m) = D * (N−m) of the spectrum D (m) (where 1 ≦ m ≦ N / 2-1, D * (N−m)). Represents a conjugate complex number of D (N−m)) to 0 ≦ m ≦ N / 2.

  The separating means 40 includes a disturbance sound suppressing unit 41 and a target sound suppressing unit 42.

  The interfering sound suppressing unit 41 uses D1 (m) and D2 (m) to suppress the interfering sound component and generate a spectrum in which the target sound component is emphasized. Then, the target sound suppression unit 42 uses D1 (m) and D2 (m) to suppress the target sound component and generate a spectrum in which the disturbing sound component is emphasized.

  Next, the configuration of the interference sound suppression unit 41 will be described.

  The interference sound suppression unit 41 includes two spatial filters 411 and 412 and a minimum selection unit 413.

  It is assumed that the suppression angles of the spatial filters 411 and 412 are set to 90 degrees and −90 degrees, respectively.

  As described above, since the target sound is assumed to arrive from a direction of approximately 0 degrees in the sound source separation device 10, the interference sound suppression unit 41 has a space in a direction different from the direction in which the target sound arrives. Although the suppression angle of the filter is directed, the number of spatial filters and the combination of suppression angles may be changed according to the direction in which the target sound is expected to arrive.

  As a specific process of the spatial filter 411, E1 (m) is obtained using the following equation (3). The spatial filter 412 calculates E2 (m) using the following equation (4). In the following formulas (3) and (4), f is a sampling frequency, and for example, 1600 Hz may be applied.

Then, as shown in the following equation (5), the minimum selection unit 413 determines the absolute value M (m) of the absolute value of the output E1 (m) of the spatial filter 411 and the output E2 (m) of the spatial filter 412 as follows: Calculate M (m). The output M (m) is given from the minimum selection unit 413 to the removing unit 50 as an extracted target sound component.

  Next, the configuration of the target sound suppression unit 42 will be described.

  The target sound suppression unit 42 includes three spatial filters 421, 422, and 423 and a minimum selection unit 424.

  It is assumed that the suppression angles of the spatial filters 421, 422, and 423 are set in directions of 0 degrees, 5 degrees, and -5 degrees, respectively.

  As described above, in the sound source separation device 10, since the target sound is assumed to come from a direction of approximately 0 degrees, the target sound suppression unit 42 sets the suppression angle of the spatial filter 421 to 0 degrees, The suppression angles of the spatial filter 422 and the spatial filter 423 are set in a direction slightly shifted (about ± 5 degrees) from the 0 degree direction. In the sound source separation device 10, it is desirable to set the suppression angle of the spatial filter so as to form a symmetric pair around the direction in which the target sound is expected to arrive, as in the above example.

  The target sound suppression unit 42 uses three spatial filters, but within a predetermined range including the direction in which the target sound is expected to arrive (0 degree in the sound source separation apparatus 10) (-5 degrees in the sound source separation apparatus 10). As long as different suppression angles are directed by a plurality of spatial filters, the number of spatial filters and combinations of the suppression angles are not limited.

  As specific processing of the spatial filter 421, F0 (m) is obtained using the following equation (6).

The spatial filter 422 calculates F1 (m) using the following equation (7). In equation (7), τ ₅ is a delay corresponding to the suppression angle = + 5 degrees.

The spatial filter 423 calculates F2 (m) using the following equation (8). In the equation (8), τ ₋₅ is a delay corresponding to the suppression angle = −5 degrees.

And the minimum selection part 424 calculates the minimum value N (m) of the absolute value of F0 (m), F1 (m), and F2 (m), as shown to the following (9) Formula. This output N (m) is provided from the minimum selection unit 424 to the removing means 50 as an extracted component of the interference sound.

  Next, the configuration of the removing unit 50 will be described.

  The removing means 50 uses M (m) and N (m) given from the separating means 40 to obtain a disturbing sound removal spectrum H (m) for removing the disturbing sound in D1 (m). This is given to the generating means 60.

  Hereinafter, an example of the interference noise removal spectrum H (m) required by the removal unit 50 will be described.

The removing unit 50 obtains S (m) from the output M (m) of the minimum selection unit 413 and the output N (m) of the minimum selection unit 424 using the following equation (10). Further, the removal means 50 uses the following equation (11) for S (m) obtained in the range of 0 ≦ m ≦ N / 2, and the interference noise removal spectrum H that is the output of the removal means 50: Find (m). Note that D1 may be replaced with D2 in the expressions (10) and (11).

H (m) = S (m) D1 (m) (11)
Further, the removing means 50 uses the property of H (m) = H * (N−m) (where N / 2 + 1 ≦ m ≦ N−1) and obstructs in the range of 0 ≦ m ≦ N−1. A sound removal spectrum H (m) is obtained and given to the generation means 60.

  The generation unit 60 performs N-point inverse FFT processing on the interference noise removal spectrum H (m) to obtain a sound source separation signal h (n). Then, as shown in the following equation (12), the generation unit 60 generates 3N / 4 signals in the latter half of the current sound source separation signal h (n) and the sound source separation signal h ′ (n) for the immediately preceding analysis unit. Are added to obtain an output y (n).

y (n) = h (n) + h ′ (n + N / 4) (12)
In the sound source separation apparatus 10, the example in which the above-described processing is performed while shifting N / 4 data so that data (samples) are overlapped in successive analysis units has been described. Since it is for smooth execution, it is not always necessary, and N pieces may be processed. In the case of processing while shifting N / 4 data, the time required for the above-described series of processing from the analysis unit 30 to the generation unit 60 is limited to NT / 4 for one analysis unit. It is desirable.

(A-2) Effects of First Embodiment According to the first embodiment, the following effects can be achieved.

  In the sound source separation device 10, the directivity of 0 degree, 5 degrees, and -5 degrees is given to the three spatial filters of the target sound suppression unit 42, respectively, and the minimum selection unit 424 outputs the three spatial filters. An output value having the smallest absolute value among the values is applied to N (m). That is, in the target sound suppression unit 42, when the target sound comes from the vicinity of the 0 degree direction, the absolute value of the output value of the spatial filter 421 is the smallest for the component in the vicinity of the 0 degree direction. Reflected in (m). On the other hand, when the target sound arrives from the vicinity in the 5 degree direction, the output value of the spatial filter 422 is reflected in N (m) for the component in the vicinity in the 5 degree direction. In this way, the target sound suppression unit 42 provides a spatial filter group that is selected and applied according to the direction of arrival of the target sound, so that N (m ) Is prevented from being mixed with the target sound component, and deterioration of the sound quality output by the sound source separation device 10 is prevented.

  Therefore, as described above, by configuring the target sound suppression unit 42 using the spatial filter group that is selected and applied according to the direction of arrival of the target sound, even when the direction of arrival of the target sound is deviated, separation is performed. It is possible to improve the quality of the later target sound and make it easier to hear.

(B) Second Embodiment Hereinafter, a second embodiment of a sound source separation device, program, and method according to the present invention will be described in detail with reference to the drawings.

(B-1) Configuration and Operation of the Second Embodiment FIG. 2 is a block diagram showing the overall configuration of the sound source separation device 10A of the second embodiment.

  In the sound source separation apparatus 10 of the first embodiment, the input means 20, the analysis means 30, and the separation means 40 are each provided, but in the sound source separation apparatus 10A of the second embodiment, the input means 20 The difference is that a plurality of sets of analysis means 30 and separation means 40 are provided. Further, the sound source separation device 10A of the second embodiment is different from the first embodiment in that the removing unit 50 is replaced with the removing unit 50A.

  As shown in FIG. 2, the sound source separation device 10 </ b> A has two sets of input means 20, analysis means 30, and separation means 40. That is, it has two input means 20 (20-1, 20-2), two analysis means 30 (30-1, 30-2), and two separation means 40 (40-1, 40-2). ing. The input unit 20-1 has two microphones 21-1 and 22-1, and the input unit 20-2 also has two microphones 21-2 and 22-2.

  Regarding the processes of the input means 20-1, 20-2, the analysis means 30-1, 30-2, and the separation means 40-1, 40-2, the input means 20, the analysis means 30, and the first embodiment, Since it is the same as that of the separation means 40, detailed description is abbreviate | omitted.

  In the following, the output of the interference sound suppressing unit in the separating unit 40-1 is expressed as MA (m), and the output of the target sound suppressing unit is expressed as NA (m). Further, in the separating unit 40-2, the output of the interference sound suppressing unit is represented as MB (m), and the output of the target sound suppressing unit is represented as NB (m). In addition, a signal obtained by processing the signal from the microphone 21-1 by the analyzing unit 30-1 is represented as D1 (m).

  Next, the configuration of the removing unit 50A will be described.

  The removing means 50A uses MA (m) and NA (m) given from the separating means 40-1 and MB (m) and NB (m) given from 40-2, and uses D (m). Then, the interference sound removal spectrum H (m) for removing the interference sound is obtained and given to the generating means 60.

  Hereinafter, an example of the interference noise removal spectrum H (m) required by the removing unit 50A will be described.

The removing unit 50A converts MA (m) and NA (m) given from the separating unit 40-1 and MB (m) and NB (m) given from 40-2 into the following equation (13). Apply and determine S (m). Further, the removing unit 50A uses the following equation (14) for S (m) obtained in the range of 0 ≦ m ≦ N / 2, and uses the following equation (14) to remove the interference sound removal spectrum H that is the output of the removing unit 50A. Find (m). Note that in the equations (13) and (14), D1 may be replaced with a spectrum based on signals from other microphones.

H (m) = S (m) D1 (m) (14)
Further, by utilizing the property of H (m) = H * (N−m) (where N / 2 + 1 ≦ m ≦ N−1), the interference noise elimination spectrum H (range of 0 ≦ m ≦ N−1) ( m) is obtained and given to the generating means 60.

  Since the processing of the generation unit 60 is the same as that of the first embodiment, description thereof is omitted.

(B-2) Effects of Second Embodiment In the sound source separation device 10A of the second embodiment, even when more than two microphones are used in the input means, the same as in the first embodiment There is an effect.

(C) Other Embodiments The present invention is not limited to the above-described embodiments, and may include modified embodiments as exemplified below.

(C-1) In the first embodiment, depending on the application of the sound source separation device 10, the generation unit 60 can be omitted, or a generation unit included in another device can be used. For example, if the sound source separation device is used for a speech recognition device, the generation means 60 can be omitted by using the separated spectrum H (m) as the recognition feature amount. Further, for example, if the sound source separation device is used for an IP phone, the IP phone has means corresponding to the generation unit, and the generation unit may be used.

(C-2) In the second embodiment, the example using the four microphones 21-1, 22-1, 21-2, and 22-2 has been described. However, the input unit 20-1, the input unit 20-2, In this case, one microphone may be used in common, and three microphones may be used. In this case, the amount of calculation can be reduced because the processing of the signals received by the commonly used microphones can be made common. Further, even when the number of microphones to be used is further increased, a common microphone may be used between the input means.

  DESCRIPTION OF SYMBOLS 10 ... Sound source separation apparatus, 20 ... Input means, 21, 22 ... Microphone, 30 ... Analysis means, 31, 32 ... Frequency analysis part, 40 ... Separation means, 41 ... Interference sound suppression part, 411, 412 ... Spatial filter, 413 ... minimum selection unit, 42 ... target sound suppression unit, 421, 422, 423 spatial filter, 424 ... minimum selection unit, 50 ... removal means, 60 ... generation means.

Claims (4)

Among the plurality of microphones arranged at intervals, the spectrum of the received sound signal of two microphones is in a different direction within a predetermined range including an assumed arrival direction where the target sound is expected to arrive. A target sound suppression spectrum generating means for generating a target sound suppression spectrum in which a component of the target sound is suppressed from a spectrum of the received signal using a plurality of target sound suppression units that process the directivity of component suppression; ,
About the spectrum of the received sound signal, target sound dominant spectrum generating means for generating a target sound dominant spectrum suppressing interfering sound coming from any direction other than the predetermined range;
Separating means for separating the disturbing sound component and the target sound component from the received signal using the target sound suppression spectrum and the target sound dominant spectrum ;
The sound source separation device, wherein the target sound suppression spectrum generating means applies a value having the smallest absolute value among the processing results of the target sound suppression unit to each component of the target sound suppression spectrum . A plurality of spectrum generation processing units having the target sound suppression spectrum generation means and the target sound dominant spectrum generation means,
The separation means separates the interference sound component and the target sound component from the received signal using the target sound suppression spectrum and the target sound dominant spectrum generated by each of the spectrum generation processing units. The sound source separation device according to claim 1, wherein: The computer installed in the sound source separation device
Among the plurality of microphones arranged at intervals, the spectrum of the received sound signal of two microphones is in a different direction within a predetermined range including an assumed arrival direction where the target sound is expected to arrive. A target sound suppression spectrum generating means for generating a target sound suppression spectrum in which a component of the target sound is suppressed from a spectrum of the received signal using a plurality of target sound suppression units that process the directivity of component suppression; ,
About the spectrum of the received sound signal, target sound dominant spectrum generating means for generating a target sound dominant spectrum suppressing interfering sound coming from any direction other than the predetermined range;
Using a target sound suppressed spectrum, the target sound predominant spectrum for the received sound signal, to function as a separating means for separating the components of the component and the target sound of the interference sound,
The sound source separation program, wherein the target sound suppression spectrum generating means applies a value having the smallest absolute value among the processing results of the target sound suppression unit to each component of the target sound suppression spectrum . In the sound source separation method performed by the sound source separation device,
A target sound suppression spectrum generating means, a target sound dominant spectrum generating means, and a separating means;
The target sound suppression spectrum generating means includes a predetermined number of directions including an assumed arrival direction in which the target sound is expected to arrive in the spectrum of the received signal of two microphones among a plurality of microphones arranged at intervals. A target sound suppression spectrum in which the target sound component is suppressed from the spectrum of the received signal using a plurality of target sound suppression units that process the component suppression directivity in different directions within the range. Generate
The target sound dominant spectrum generating means generates a target sound dominant spectrum in which a disturbing sound coming from an arbitrary direction other than the predetermined range is suppressed for the spectrum of the received signal, and the separating means suppresses the target sound suppression Using the spectrum and the target sound dominant spectrum, for the received signal, separating the disturbing sound component and the target sound component ,
The sound source separation method, wherein the target sound suppression spectrum generating means applies a value having the smallest absolute value among the processing results of the target sound suppression unit to each component of the target sound suppression spectrum .

Images