JP2004320204A - Method and apparatus for echo cancellation and echo cancellation program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accelerate a speed for approaching a filter coefficient used in a filter processing means to true characteristics for echo cancellation. <P>SOLUTION: An apparatus for the echo cancellation includes a disturbance ratio estimating means for sequentially estimating a disturbance ratio representing the ratio of the amplitude of the disturbance signal component, except an echo signal included in a sound pickup signal to the amplitude of simulated error outputted from a simulated error output means at each frequency range; and a second updating quantity regulating means for multiplying the characteristics error of the filter coefficient or the characteristic error of the filter coefficient multiplied by a first regulating coefficient by a second regulating coefficient taking a predetermined value from "0" at each frequency range, in such a manner the larger the disturbance ratio is, the larger the second regulating coefficient in response to the amplitude of the disturbance ratio in the front or rear stage of a first updating quantity regulating means. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a reverberation canceling device for canceling reverberation circulating from a sound reproducing unit such as a speaker to a sound pickup unit such as a microphone.
[0002]
[Prior art]
A reverberation canceling device for canceling reverberation circulating from the sound reproducing means to the sound collecting means is connected as shown in FIG. Reference numeral 10 shown in FIG. 6 indicates a conventional echo canceller. In the conventional echo canceller 10, the impulse response h of the echo path between the sound reproduction means 1 and the sound pickup means 2 is estimated, and the estimated impulse response h 'and the reproduction signal x input to the reproduction signal input terminal 3 are obtained. generating a convolution h'* ^x, is subtracted from the actual echo signal y, to obtain a echo cancellation signal e to the echo cancellation signal output terminal 4.
However, the estimated impulse response and the reproduction signal convolution operation require a large amount of operation, which is a mounting problem.
[0003]
In recent years, in order to solve this problem, the reproduction signal and the reverberation signal are once frequency-domain transformed, parameters corresponding to the frequency domain transformation of the impulse response of the reverberation path are estimated, and multiplication processing is used instead of convolution (non-convolution). There has been proposed a method of reducing the amount of calculation by, for example, dividing the image into smaller convolution operations (Patent Document 1) or a smaller convolution operation (Patent Document 1). Examples of frequency domain transforms include (fast) discrete Fourier transform, (fast) discrete cosine transform, and (fast) Hartley transform.
Although a speaker is shown as the sound reproducing means 1 in FIG. 6, the sound reproducing means 1 also includes an amplifier and a buffer in a stage before reproduction. Similarly, the sound pickup means 2 also includes an amplifier and a buffer at the subsequent stage of the microphone.
[0004]
FIG. 7 schematically shows a configuration described in (Non-Patent Document 1) and the like. The configuration of FIG. 7 includes the following means. That is, a reproduction signal input means 101 for inputting a reproduction signal input from the reproduction signal input terminal 3 and output to the sound reproduction means 1 and accumulating the reproduction signal for a certain period of time to obtain a reproduction signal sequence; A sound pickup signal input means 102 for inputting a sound pickup signal from the sound pickup means 2 and accumulating it for a certain period of time to obtain a sound pickup signal sequence, and a reproduction signal conversion means 103 for frequency domain transforming the reproduction signal sequence to obtain a reproduction signal conversion sequence. Then, a simulated reverberation signal conversion sequence that simulates a frequency domain conversion sequence of a reverberation signal circulating from the sound reproducing unit 1 to the sound pickup unit 2 is generated by inputting the reproduction signal conversion sequence for each frequency domain and performing a filtering process. Filter processing means 104, simulation error output means 105 for inputting the simulated reverberation signal conversion sequence and the picked-up signal sequence, and outputting a simulated error of the simulated reverberation signal conversion sequence; A filter error calculating means for inputting and calculating a characteristic error of the filter coefficient of the filter processing means for each frequency domain, and a first adjusting coefficient for multiplying the characteristic error calculated by the filter error calculating means for each frequency domain. It includes an update amount adjustment unit 107 and a filter update unit 108 that updates the filter coefficient of the filter processing unit 104 by adding a characteristic error multiplied by the first adjustment coefficient.
[0005]
[Patent Document 1]
JP-A-9-116472 (FIG. 6)
[Non-patent document 1]
Simon Haykin, "Adaptive Filter Theory," Science and Technology Publishing, January 10, 2001, pp. 139-214. 500-541
[0006]
[Problems to be solved by the invention]
In the configuration of FIG. 7, if the filter error calculation means 106 can correctly calculate and output the characteristic error between the true characteristic for echo cancellation and the filter coefficient used in the filter processing means 104, the filter error calculation means 106 outputs The first adjustment coefficient by which the first update amount adjustment unit 107 multiplies the characteristic error may be one. That is, the characteristic error can be corrected most quickly.
However, it is difficult for the filter error calculation means 106 to correctly calculate and output the characteristic error. One of the causes is the influence of disturbance components such as ambient noise and circuit noise other than the reverberation signal included in the sound pickup signal picked up by the sound pickup means 2. This disturbance component has an adverse effect on the calculation of the characteristic error output from the filter error calculation means 106. Therefore, in order to avoid this influence, the conventional technique introduces the first update amount adjusting means 107 and sets the first adjustment coefficient small.
[0007]
Conventionally, since the first coefficient is set to be small, the speed at which the filter coefficient used in the filter processing means 104 approaches the true characteristic for echo cancellation is low, and it takes a long time to achieve complete echo cancellation. Was.
As described above, the object of the present invention is to minimize the influence of the disturbance component included in the sound pickup signal picked up by the sound pickup means 2 and to make the first update adjustment means 107 take the largest possible first adjustment coefficient. The purpose is to increase the speed at which the filter coefficient used in the filter processing means 104 approaches the true characteristic for echo cancellation.
[0008]
[Means for Solving the Problems]
According to the first aspect of the present invention, a reproduction signal input process for inputting a reproduction signal output to the sound reproduction means and accumulating the reproduction signal for a certain period of time to obtain a reproduction signal sequence, and collecting the reproduction signal from the sound pickup means existing in the same space as the sound reproduction means. A sound collection signal input process for inputting a sound signal and accumulating it for a certain period of time to obtain a sound collection signal sequence, a reproduction signal conversion process for frequency domain transforming a reproduction signal sequence to obtain a reproduction signal conversion sequence, and a reproduction signal conversion sequence for each frequency domain Filter processing to generate a simulated reverberation signal conversion sequence that simulates a frequency domain conversion sequence of a reverberation signal that circulates from the sound reproduction means to the sound pickup means by inputting and filtering the simulated reverberation signal conversion sequence and the picked-up signal And a simulation error output process for outputting a simulation error of a simulated reverberation signal conversion sequence and a simulation error output process for inputting a reproduction signal conversion sequence and a simulation error, and measuring a characteristic error of a filter coefficient of the filter process for each frequency domain. A filter error calculation process, a first update amount adjustment process of multiplying the characteristic error calculated by the filter error calculation process by a first adjustment coefficient for each frequency domain, and a filter by adding the characteristic error multiplied by the first adjustment coefficient. In the reverberation elimination method of performing the filter update process of updating the filter coefficient of the process, the magnitude of the disturbance signal component other than the reverberation signal included in the sound pickup signal with respect to the magnitude of the simulation error output in the simulation error output process. A disturbance ratio estimation process for sequentially estimating a disturbance ratio representing a ratio for each frequency domain, and a characteristic error of a filter coefficient or a characteristic error of a filter coefficient multiplied by the first adjustment coefficient in a stage before or after the first update amount adjustment process. The second adjustment coefficient is multiplied for each frequency domain by a second adjustment coefficient that takes a smaller value as the disturbance ratio increases according to the magnitude of the disturbance ratio between 0 and a predetermined value. Suggest echo cancellation how to run the new amount adjustment process.
[0009]
According to a second aspect of the present invention, a reproduction signal input means for inputting a reproduction signal output to the sound reproduction means and accumulating the reproduction signal for a certain period of time to obtain a reproduction signal sequence, and a sound pickup means existing in the same space as the sound reproduction means. A sound signal input means for inputting a sound signal and accumulating the sound signal for a certain period of time to obtain a sound signal sequence; a reproduction signal conversion means for frequency domain converting the reproduction signal sequence to obtain a reproduction signal conversion sequence; and a reproduction signal conversion sequence for each frequency region. Filter processing means for generating a simulated reverberation signal conversion sequence that simulates a frequency domain conversion sequence of a reverberation signal circulating from the sound reproducing means to the sound pickup means by inputting and filtering the simulated reverberation signal conversion sequence and sound pickup A simulation error output means for inputting a signal sequence and outputting a simulation error of a simulation echo signal conversion sequence, and a characteristic error for each frequency domain of a filter coefficient of a filter processing means for inputting a reproduction signal conversion sequence and the simulation error A filter error calculating means for calculating, a first update amount adjusting means for multiplying the characteristic error calculated by the filter error calculating means by a first adjusting coefficient for each frequency domain, and adding a characteristic error multiplied by the first adjusting coefficient. An echo canceller having a filter updating means for updating a filter coefficient of the filter processing means, wherein the magnitude of a disturbance signal component other than the reverberation signal included in the picked-up signal with respect to the magnitude of the simulation error output by the simulation error output means is determined. A disturbance ratio estimating means for sequentially estimating a disturbance ratio representing a ratio for each frequency domain, and a preceding or subsequent stage of the first updating amount adjusting means, the characteristic error of the filter coefficient or the characteristic error of the filter coefficient multiplied by the first adjusting coefficient. , A second adjustment coefficient multiplied for each frequency domain by a second adjustment coefficient having a smaller value as the disturbance ratio is larger according to the magnitude of the disturbance ratio between 0 and a predetermined value. Suggest echo canceller and a new quantity adjusting means.
[0010]
According to a third aspect of the present invention, there is provided an echo canceling program which is described by a code string which can be decoded by a computer, and which causes the computer to execute the echo canceling method according to the first aspect.
Action <br/> disturbance component according to the configuration sound pickup unit 2 included in the collected sound signal picked up in this invention, in accordance with the ratio of the filter error calculation means had the characteristic error of the filter coefficients to be output In addition, it is possible to reduce the uncertainty of the calculated characteristic error of the filter coefficient. For this reason, the first updating amount adjusting means can take the largest first adjusting coefficient as much as possible, and can increase the speed at which the filter coefficient used in the filter processing means approaches the true value for echo cancellation.
[0011]
The disturbance ratio estimating means estimates the disturbance ratio representing the ratio of the magnitude of the disturbance signal component other than the reverberant signal included in the sound pickup signal from the sound pickup means to the magnitude of the simulation error output by the simulation error output means. Can be realized as follows. Here, the magnitude of the simulation error output by the simulation error output means in the frequency domain ω (n = 1... N) divided into N is | E (ω _n ) |, and the magnitude of the disturbance in the frequency domain ω _n is estimated. The value is | N (ω _n ) |. As the estimated disturbance value | N (ω _n ) |, a value measured in advance in a non-voice section or the like is used. At this time, the calculation of the disturbance ratio DR (ω _n )
DR (ω _n ) = | N (ω _n ) | / | E (ω _n ) | (1)
Can be calculated as Here, the value of the disturbance ratio DR (ω _n ) is assumed to be, for example, T = 1 from 0 to a predetermined value T. If the value deviates from this range in an actual use environment, a range of 0 to a predetermined value is set. Censor the numerical value so that The value of DR (ω _n ) may be based on a formula other than equation (1) as long as the ratio of | N (ω _n ) | to | E (ω _n ) |
[0012]
The second adjustment coefficient β (ω _n ) of the second update amount adjustment means takes a value from 0 to a predetermined value, for example, 1, and uses the above-described disturbance ratio DR (ω _n )
β (ω _n ) = (T-DR (ω _n ) ^p ) ^{1 / p} (2)
Can be determined as follows. Here, p is a positive real number. In addition, in addition to the equation (2), as shown in FIG. 4A, it is possible to smoothly correspond to the value of DR (ω _n ), or to correspond to discontinuity as shown in FIG. Even if the value is such that the value of β (ω _n ) decreases as the value of DR (ω _n ) increases, it is included in the embodiment in which the effects of the present invention can be obtained. Even when DR (ω _n ) is obtained in a form other than equation (1), an appropriate one can be selected from the correspondence between DR (ω _n ) and β (ω _n ) shown here.
[0013]
BEST MODE FOR CARRYING OUT THE INVENTION
FIG. 1 shows an embodiment of the echo canceling apparatus according to the present invention. The echo canceller shown in FIG. 1 will be described together with the echo canceller method.
Reference numeral 100 shown in FIG. 1 indicates an echo canceller according to the present invention. The echo canceling apparatus 100 according to the present invention is obtained by adding the following means to the conventional echo canceling apparatus 10 shown in FIG. That is, a disturbance ratio DR (ω _n ) representing a ratio of the magnitude of the disturbance signal component other than the reverberation signal included in the sound pickup signal from the sound pickup means 2 to the magnitude of the simulation error output from the simulation error output means 105. Ratio estimation means 111 for successively estimating the frequency ratio for each frequency domain, and the characteristic error of the filter coefficient actually output by the filter error calculation means 106, and the disturbance ratio according to the magnitude of the disturbance ratio estimated by the disturbance ratio estimation means 111. Is a predetermined value when 0% is 0%, for example, 1 when the disturbance ratio is 100%, and a second adjustment coefficient β (ω _n ) which takes a value between 0 and a predetermined value when the disturbance ratio is in between. And a second update amount adjusting unit 110 for multiplying each region.
[0014]
According to this configuration, the disturbance coefficient included in the sound pickup signal picked up by the sound pickup means 2 affects the characteristic error of the filter coefficient output from the filter error calculation means 106 in accordance with the ratio of the calculated filter coefficient. The value of the part where the characteristic error is uncertain can be reduced. For this reason, the first update amount adjusting means 107 can take the largest possible first adjustment coefficient, and can increase the speed at which the filter coefficient used in the filter processing means 104 approaches the true characteristic for echo cancellation.
[0015]
Modified embodiment of the invention In the embodiment shown in FIG. 1, an example in which the second update amount adjusting unit 110 is provided before the first update amount adjusting unit 107 has been described. However, as shown in FIG. Even if the second update amount adjusting unit 110 is arranged at the subsequent stage of the first update amount adjusting unit 107, the same operation and effect as the embodiment shown in FIG. 1 can be obtained.
[0016]
As shown in FIG. 3, a third update amount adjusting unit 112 is arranged between the filter error calculating unit 106 and the filter updating unit 108, and the third update amount adjusting unit 112 For example, the same operation as in the embodiment shown in FIG. 1 may be performed by obtaining a product with the coefficient, using the value of the product as a third adjustment coefficient, and multiplying the third adjustment coefficient by the characteristic error of the filter coefficient for each frequency domain. The effect can be obtained.
The above-described reverberation canceling apparatus and reverberation canceling method according to the present invention are realized by causing a computer to execute a sound canceling program described by a computer-readable code string. FIG. 5 shows a state in which the sound erasing program according to the present invention is installed in the computer 20.
[0017]
As is well known, the computer 20 includes a central processing unit (CPU) 21, a read-only memory (ROM) 22 storing a basic program for controlling start-up and shut-down of the computer, and the like; In addition to the temporary recording, the read / write / read / write memory (RAM) 23 for storing a program for executing the echo canceling method of the present invention, an input port 24, an output port 25, and the like can be used.
The reproduction signal input terminal 3 and the sound pickup means 2 are connected to the input port 24, and the reproduction signal X and the pickup signal are input. The output port 25 is connected to the sound reproducing means 1 and the echo canceling signal output terminal 4, and outputs the sound corresponding to the reproduced signal and the echo canceling signal e = yy- '.
[0018]
In the readable / writable memory 23, in addition to the data storage area 23A, a reproduction signal input processing program 23B operating as the reproduction signal input means 101, and a sound collection signal input processing program 23C operating as the sound collection signal input means 102 A reproduction signal conversion processing program 23D operating as the reproduction signal conversion means 103, a filter processing program 23E operating as the filter processing means 104, a simulation error output processing program 23F operating as the simulation error output means 105, and a filter error calculation means A filter error calculation processing program 23G operating as 106, a first update amount adjustment processing program 23H operating as the first update amount adjustment means 107, a filter update processing program 23I operating as the filter update means 108, and a second update amount Coordinator A second update amount adjustment processing program 23J operating as 110 and a disturbance ratio estimation processing program 23K operating as the disturbance ratio estimating means 111 are installed, and these programs are decoded by the central processing unit 21 to perform the echo cancellation operation. Execute Each of the programs 23B to 23K installed in the readable and writable memory 23 is recorded in advance on a computer-readable magnetic disk or CD-ROM, and is installed from these recording media or installed through a communication circuit. You.
[0019]
【The invention's effect】
The echo canceller according to the present invention suppresses a decrease in the speed at which the filter coefficient used in the filter processing unit approaches the true characteristic for echo cancellation, and suppresses a disturbance component included in the sound pickup signal picked up by the sound pickup unit. The influence can be reduced, and the echo canceling performance can be improved.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating an embodiment of an echo canceller according to the present invention.
FIG. 2 is a block diagram for explaining another embodiment of the echo canceller according to the present invention.
FIG. 3 is a block diagram for explaining still another embodiment of the echo canceller according to the present invention.
FIG. 4 is a characteristic curve diagram for explaining the operation of the main part of the present invention.
FIG. 5 is a conceptual diagram illustrating a configuration when the echo canceller according to the present invention is implemented by a computer.
FIG. 6 is a block diagram for explaining the outline of the echo canceller.
FIG. 7 is a block diagram for explaining a conventional echo canceller.
[Explanation of symbols]
REFERENCE SIGNS LIST 1 sound reproduction means 105 simulated error output means 2 sound collection means 106 filter error calculation means 3 reproduction signal input terminal 107 first update amount adjustment means 4 echo cancellation signal output terminal 108 filter update means 100 echo cancellation device 110 second update amount Adjusting means 101 Reproduction signal input means 111 Disturbance ratio estimation means 102 Sound collection signal input means 103 Reproduction signal conversion means 104 Filter processing means

Claims (3)

A reproduction signal input process for inputting a reproduction signal to be output to the sound reproduction unit and accumulating the reproduction signal for a certain period of time to obtain a reproduction signal sequence; and inputting a predetermined sound collection signal from the sound collection unit existing in the same space as the sound reproduction unit. A sound collection signal input process for accumulating time to obtain a sound collection signal sequence, a reproduction signal conversion process for frequency domain transforming the reproduction signal sequence to obtain a reproduction signal conversion sequence, and a filter for inputting the reproduction signal conversion sequence for each frequency domain A filtering process for generating a simulated reverberation signal conversion sequence that simulates a frequency domain conversion sequence of a reverberation signal that wraps around from the sound reproduction unit to the sound collection unit by processing; And a simulation error output process for inputting the sequence and outputting a simulation error of the simulation echo signal conversion sequence, and for each frequency domain of the filter coefficient of the filter process after inputting the reproduction signal conversion sequence and the simulation error. A filter error calculation process for calculating a characteristic error, a first update amount adjustment process for multiplying the characteristic error calculated by the filter error calculation process by a first adjustment coefficient for each frequency domain, and a characteristic multiplied by the first adjustment coefficient. And a filter updating process for updating a filter coefficient of the filtering process by adding an error.
A disturbance ratio that sequentially estimates a disturbance ratio representing a ratio of a magnitude of a disturbance signal component other than the reverberation signal included in the sound pickup signal to a magnitude of the simulation error output in the simulation error output process for each frequency domain. Estimation processing and before or after the first update amount adjustment processing, the characteristic error of the filter coefficient or the characteristic error of the filter coefficient multiplied by the first adjustment coefficient, between 0 and a predetermined value, the disturbance A second update amount adjustment process of multiplying a second adjustment coefficient having a smaller value as the disturbance ratio increases in accordance with the ratio, for each frequency domain. A reproduction signal input means for inputting a reproduction signal to be output to the sound reproduction means and accumulating the reproduction signal for a certain period of time to obtain a reproduction signal sequence; and inputting a fixed sound signal from a sound collection means existing in the same space as the sound reproduction means. Sound collection signal input means for accumulating time to obtain a sound collection signal sequence, reproduction signal conversion means for frequency domain transforming the reproduction signal sequence to obtain a reproduction signal conversion sequence, and a filter for inputting the reproduction signal conversion sequence for each frequency domain A filter processing unit for generating a simulated reverberation signal conversion sequence that simulates a frequency domain conversion sequence of a reverberation signal circulating from the sound reproduction unit to the sound collection unit by processing; the simulated reverberation signal conversion sequence and the sound collection A simulation error output means for inputting a signal sequence and outputting a simulation error of a simulation echo signal conversion sequence, and a frequency of a filter coefficient of the filter processing means for inputting the reproduction signal conversion sequence and the simulation error Filter error calculating means for calculating a characteristic error for each area, first updating adjusting means for multiplying the characteristic error calculated by the filter error calculating means by a first adjusting coefficient for each frequency area, and multiplying by the first adjusting coefficient. And a filter updating means for updating the filter coefficient of the filtering means by adding the characteristic error.
Disturbance ratio estimation for sequentially estimating a disturbance ratio representing a ratio of a magnitude of a disturbance signal component other than the reverberation signal included in the picked-up signal to a magnitude of the simulation error output by the simulation error output means for each frequency domain. Means, at a stage before or after the first update amount adjusting unit, the characteristic error of the filter coefficient or the characteristic error of the filter coefficient multiplied by the first adjustment coefficient, the disturbance ratio between 0 and a predetermined value A reverberation canceling apparatus, comprising: a second update amount adjusting unit that multiplies, for each frequency region, a second adjustment coefficient that takes a smaller value as the disturbance ratio increases according to the magnitude. An echo cancellation program which is described by a computer-readable code string and causes the computer to execute the echo cancellation method according to claim 1.

Images