JP4504892B2 - Echo canceling method, echo canceling apparatus, program, recording medium - Google Patents

Echo canceling method, echo canceling apparatus, program, recording medium Download PDF

Info

Publication number
JP4504892B2
JP4504892B2 JP2005251648A JP2005251648A JP4504892B2 JP 4504892 B2 JP4504892 B2 JP 4504892B2 JP 2005251648 A JP2005251648 A JP 2005251648A JP 2005251648 A JP2005251648 A JP 2005251648A JP 4504892 B2 JP4504892 B2 JP 4504892B2
Authority
JP
Japan
Prior art keywords
signal
power
estimated value
value
echo
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
JP2005251648A
Other languages
Japanese (ja)
Other versions
JP2007067855A (en
Inventor
末廣 島内
陽一 羽田
章俊 片岡
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nippon Telegraph and Telephone Corp
Original Assignee
Nippon Telegraph and Telephone Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nippon Telegraph and Telephone Corp filed Critical Nippon Telegraph and Telephone Corp
Priority to JP2005251648A priority Critical patent/JP4504892B2/en
Publication of JP2007067855A publication Critical patent/JP2007067855A/en
Application granted granted Critical
Publication of JP4504892B2 publication Critical patent/JP4504892B2/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Circuit For Audible Band Transducer (AREA)
  • Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)

Abstract

<P>PROBLEM TO BE SOLVED: To accurately cancel an echo signal from a sound collecting signal obtained by mixing acoustic signals to be reproduced from two speakers and to be collected into a single microphone. <P>SOLUTION: Values are output, as power deviation values, which is different depending on relation between the power of a first reproduction signal and a second reproduction signal, and a fifth estimation value obtained by differentiating the weighted average of a first estimation value and a third estimation value is obtained depending on this power deviation. Further, a sixth estimation value obtained by differentiating the proportion of weighted average of a second estimation error value and a fourth estimation error value is obtained depending on the power deviation value, a first adaptive simulated echo path characteristic is updated based on the fifth estimation value, and a second adaptive simulated echo path characteristic is updated based on a sixth estimation value. <P>COPYRIGHT: (C)2007,JPO&amp;INPIT

Description

本発明は、ステレオ音声などを再生可能な２つのスピーカから一つ以上のマイクロホンへと回り込む反響を消去するための反響消去方法、反響消去装置、反響消去プログラム、記録媒体に関する。   The present invention relates to an echo canceling method, an echo canceling device, an echo canceling program, and a recording medium for canceling echo that circulates from two speakers capable of reproducing stereo sound and the like to one or more microphones.

ステレオ音声などを再生可能な２つのスピーカから少なくとも一つ以上のマイクロホンへ回り込む反響を消去する反響消去装置は、図２のように接続される。図２では、マイクロホンは一つの場合を挙げているが、複数のマイクロホンが存在する場合も各マイクロホン毎に同様な構成がとられる。従来、反響消去装置１００内では、第一スピーカ１とマイクロホン３間の第一反響路４のインパルス応答に相当する長さLのベクトルH1を模擬する第一模擬特性Hm1(k)を保持する第一適応型模擬反響路１０１と、第二スピーカ２とマイクロホン３間の第二反響路５のインパルス応答に相当する長さLのベクトルH2を模擬する第二模擬特性Hm2(k)を保持する第二適応型模擬反響路１０２を有する。ここで、kは、所定間隔の離散時間を指すステップ数である。第一の再生信号x1(k)および第二の再生信号x2(k)は、それぞれ、第一反響路４および第二反響路５の各ベクトルH1とH2との畳込演算により第一反響信号d1(k)および第二反響信号d2(k)として、マイクロホン３において混合され収音される。   An echo canceling device that cancels the echo that circulates from at least one microphone to two speakers capable of reproducing stereo sound or the like is connected as shown in FIG. Although FIG. 2 shows a case where there is one microphone, the same configuration is adopted for each microphone even when there are a plurality of microphones. Conventionally, in the echo canceling apparatus 100, a first simulation characteristic Hm1 (k) that simulates a vector H1 of length L corresponding to the impulse response of the first echo path 4 between the first speaker 1 and the microphone 3 is held. A second simulation characteristic Hm2 (k) for simulating a vector H2 of length L corresponding to an impulse response of the second echo path 101 between the one adaptive type simulation echo path 101 and the second speaker 2 and the microphone 3 is held. There are two adaptive simulated echo paths 102. Here, k is the number of steps indicating discrete time at a predetermined interval. The first reproduction signal x1 (k) and the second reproduction signal x2 (k) are obtained by performing a convolution operation with the vectors H1 and H2 of the first echo path 4 and the second echo path 5, respectively. As d1 (k) and the second echo signal d2 (k), they are mixed and picked up by the microphone 3.

このとき、第一適応型模擬反響路１０１と第二適応型模擬反響路１０２とは、それぞれ、第一の再生信号x1(k)および第二の再生信号x2(k)を入力として、第一模擬特性Hm1(k)および第二模擬特性Hm2(k)との畳込演算により第一模擬反響信号dm1(k)および第二模擬反響信号dm2(k)を生成し、加算器１０３において加算した後、減算器１０４において、第一反響信号d1(k)および第二反響信号d2(k)を含むマイクロホン３の収音信号y(k)から減算することで、反響消去装置の出力信号でもある誤差信号e(k)を計算する。なお、加算器１０３と減算器１０４の機能を統合した誤差算出手段３０４で同様な処理を行ってもよい。   At this time, the first adaptive simulated echo path 101 and the second adaptive simulated echo path 102 receive the first reproduction signal x1 (k) and the second reproduction signal x2 (k) as inputs, respectively. A first simulated echo signal dm1 (k) and a second simulated echo signal dm2 (k) are generated by convolution with the simulated characteristics Hm1 (k) and the second simulated characteristics Hm2 (k) and added in the adder 103. After that, the subtracter 104 subtracts from the collected sound signal y (k) of the microphone 3 including the first echo signal d1 (k) and the second echo signal d2 (k), so that it is also an output signal of the echo canceller. Calculate the error signal e (k). Note that the same processing may be performed by the error calculation unit 304 that integrates the functions of the adder 103 and the subtracter 104.

さて、第一適応型模擬反響路１０１および第二適応型模擬反響路１０２において、第一模擬特性Hm1(k)および第二模擬特性Hm2(k)は、それぞれ、

Figure 0004504892
と更新される。ここでX1(k)=[x1(k), x1(k−1), …, x1(k−L＋1)]T , X2(k)=[x2(k), x2(k−1), …, x2(k−L＋1)]Tであり、μは更新量を調整する正の数、‖X‖はベクトルXのノルムを表す。また、e1(k)=d1(k)−dm1(k),e2(k)=d2(k)−dm2(k)として表せるが、例え、マイクロホン３の収音信号y(k)が反響信号以外の信号を含まなかったとしても、実際に観測できる信号は、
y(k)＝d1(k)＋d2(k)であり、d1(k)とd2(k)を個別に観測することは不可能である。 In the first adaptive simulated echo path 101 and the second adaptive simulated echo path 102, the first simulated characteristic Hm1 (k) and the second simulated characteristic Hm2 (k) are respectively
Figure 0004504892
 And updated. Where X1 (k) = [x1 (k), x1 (k−1),…, x1 (k−L + 1)] T , X2 (k) = [x2 (k), x2 (k−1),… , x2 (k−L + 1)] T , μ is a positive number for adjusting the update amount, and ‖X‖ represents the norm of the vector X. Also, e1 (k) = d1 (k) −dm1 (k), e2 (k) = d2 (k) −dm2 (k) can be expressed, for example, the collected sound signal y (k) of the microphone 3 is an echo signal. Even if the signal other than is not included, the signal that can actually be observed is
y (k) = d1 (k) + d2 (k), and d1 (k) and d2 (k) cannot be observed individually.

実際には、図２の例では、誤差信号e(k)＝y(k)−[dm1(k)＋dm2(k)]に基づき推定値を与えている。すなわち、信号電力加重誤差配分手段１０５において、e1(k)の推定値として、

Figure 0004504892
を求め、第一の再生信号x1(k)および第二の再生信号x2(k)の電力に相当する量により重み付けを行い、e(k)を配分し、（式１）および（式２）を適用する形で、第一模擬特性Hm1(k)および第二模擬特性Hm2(k)の特性更新を行っている。ここで、分母の‖X1(k)‖2+‖X2(k)‖2に、零除算防止を目的とした小さい正の実数を加える場合もあるが、ここでの議論の本質には影響しない。なお、このような特性更新は、非特許文献１記載の「学習同定法を用いた線形結合型多チャンネル適応フィルタ」の２チャンネルの場合の特性更新と等価である。 Actually, in the example of FIG. 2, the estimated value is given based on the error signal e (k) = y (k) − [dm1 (k) + dm2 (k)]. That is, in the signal power weighted error distribution means 105, as an estimated value of e1 (k),
Figure 0004504892
 Is weighted by an amount corresponding to the power of the first reproduction signal x1 (k) and the second reproduction signal x2 (k), and e (k) is allocated, and (Equation 1) and (Equation 2) The characteristic update of the first simulation characteristic Hm1 (k) and the second simulation characteristic Hm2 (k) is performed by applying the above. Here, a small positive real number may be added to the denominator ‖X1 (k) ‖ 2 + ‖X2 (k) 目的2 to prevent division by zero, but this does not affect the nature of the discussion here. . Such a characteristic update is equivalent to a characteristic update in the case of two channels of “a linear combination type multi-channel adaptive filter using a learning identification method” described in Non-Patent Document 1.

また、図３の例では、均等誤差配分手段２０５により、e1(k)の推定値として、
em1(k)＝（1／2）e(k) （式５）
e2(k)の推定値として、
em2(k)＝（1／2）e(k) （式６）
とe(k)を均等に配分し、（式１）および（式２）に適用する形で、第一模擬特性Hm1(k)および第二模擬特性Hm2(k)の特性更新を行っている。このような特性更新は、非特許文献１記載の「正規化法を用いた線形結合型多チャンネル適応フィルタ」の２チャンネルの場合の特性更新に相当する。 Further, in the example of FIG. 3, the equal error distribution means 205 uses the estimated value of e1 (k) as
em1 (k) = (1/2) e (k) (Formula 5)
As an estimate of e2 (k),
em2 (k) = (1/2) e (k) (Formula 6)
And e (k) are evenly distributed, and the characteristics of the first simulated characteristic Hm1 (k) and the second simulated characteristic Hm2 (k) are updated in a form that is applied to (Expression 1) and (Expression 2). . Such a characteristic update corresponds to a characteristic update in the case of two channels of “a linearly coupled multi-channel adaptive filter using a normalization method” described in Non-Patent Document 1.

ここで、図２および図３に示した従来技術において、（式１）および（式２）で本来用いられるべきe1(k)、e2(k)に対して実際に適用される第一推定値および第二推定値em1(k)、em2(k)の精度について考える。まず、
e1(k)＝ec(k)＋ea(k) （式７）
e2(k)＝ec(k)−ea(k) （式８）
とおく。ただし、

Figure 0004504892
と分解できる。このとき上記の関係より、
Figure 0004504892
 は、観測不可能であるため、eu(k)を含むea(k)の値は正確に知ることができない。そこで、ea(k)の推定値ema(k)を用いて、（式７）、（式８）に従い、推定値em1(k)、em2(k)を
em1(k)＝ec(k)＋ema(k) （式９）
em2(k)＝ec(k)−ema(k) （式１０）
と表すことを考える。 Here, in the prior art shown in FIGS. 2 and 3, the first estimated value actually applied to e1 (k) and e2 (k) that should be originally used in (Expression 1) and (Expression 2). Consider the accuracy of the second estimated values em1 (k) and em2 (k). First,
e1 (k) = ec (k) + ea (k) (Formula 7)
e2 (k) = ec (k) −ea (k) (Formula 8)
far. However,
Figure 0004504892
 Can be disassembled. At this time, from the above relationship,
Figure 0004504892
 Cannot be observed, so the value of ea (k) including eu (k) cannot be known accurately. Therefore, using the estimated value ema (k) of ea (k), the estimated values em1 (k) and em2 (k) are calculated according to (Expression 7) and (Expression 8).
em1 (k) = ec (k) + ema (k) (Formula 9)
em2 (k) = ec (k) −ema (k) (Formula 10)
Think of the expression.

このとき、（式３）、（式４）に基づく図２の構成では、（式９）、（式１０）において、

Figure 0004504892
なる関係から、eu(k)の大きさは、αに依存するため、反響路の模擬精度もαの値に依存してしまう。
一方、（式５）、（式６）に基づく図３の構成では（式９）、（式１０）において、
ema(k)＝0 （式１２）
と与えたことに相当する。（式１２）では、ema(k)の平均値を零と仮定し、eu(k)の大きさ、すなわちαの値によらない反響路の模擬を実現している。
藤井 哲郎、島田 正治：「多チャンネル適応ディジタルフィルタ」電子通信学会 論文誌（Ａ），J69-A No.10, pp1226-1233, 1986. At this time, in the configuration of FIG. 2 based on (Expression 3) and (Expression 4), in (Expression 9) and (Expression 10),
Figure 0004504892
 Therefore, since the magnitude of eu (k) depends on α, the simulation accuracy of the echo path also depends on the value of α.
On the other hand, in the configuration of FIG. 3 based on (Equation 5) and (Equation 6), in (Equation 9) and (Equation 10),
ema (k) = 0 (Formula 12)
It is equivalent to giving. In (Expression 12), the average value of ema (k) is assumed to be zero, and the echo path is simulated regardless of the size of eu (k), that is, the value of α.
Tetsuro Fujii, Masaharu Shimada: “Multi-channel adaptive digital filter” IEICE Transactions (A), J69-A No.10, pp1226-1233, 1986.

背景技術でのe1(k)、e2(k)に対する推定値em1(k)、em2(k)に関する考察より、図２の構成においては、αの大きさに対する、推定精度の依存性が強く、図３の構成においては、αの大きさによらず、固定的な誤差配分により、推定を実現している。図２の構成の推定精度は、αの値に依存して、図３の構成より優れる場合もあるが、劣る場合もある。
本発明の目的は、図２の構成の推定値em1(k)、em2(k)について、αの値に依存した精度劣化を抑え、反響路の模擬精度を高めることにより、反響消去性能の優れた反響消去方法及び装置を提案することにある。 From the consideration of the estimated values em1 (k) and em2 (k) for e1 (k) and e2 (k) in the background art, in the configuration of FIG. 2, the dependence of the estimation accuracy on the magnitude of α is strong. In the configuration of FIG. 3, estimation is realized by fixed error distribution regardless of the magnitude of α. The estimation accuracy of the configuration of FIG. 2 may be better than the configuration of FIG. 3 depending on the value of α, but may be inferior.
The object of the present invention is to improve the echo canceling performance by suppressing the accuracy deterioration depending on the value of α for the estimated values em1 (k) and em2 (k) of the configuration of FIG. 2 and increasing the simulation accuracy of the echo path. Another object is to propose an echo canceling method and apparatus.

本発明による反響消去装置は同一空間内に存在する第一の音響再生手段と音響収音手段との間の音響的な伝達経路である第一の反響路を介して第一の音響再生手段から再生する第一の再生信号が音響収音手段により収音される第一の反響信号と、第一の音響再生手段の同一の空間内に存在する第二の音響再生手段と音響収音手段との間の音響的な伝達経路である第二の反響路を介して、第二の音響再生手段から再生する第二の再生信号が音響収音手段により第一の反響信号と混合して収音される第二の反響信号とを、音響収音手段で収音する全ての収音信号の中から消去または低減する反響消去装置において、第一の反響路の模擬特性を有し、第一の再生信号を入力し、第一の反響信号を模擬する第一の反響模擬信号を出力する第一の適応型模擬反響路と、第二の反響路の模擬特性を有し、第二の再生信号を入力し、第二の反響信号を模擬する第二の反響模擬信号を出力する第二の適応型模擬反響路と、第一の反響模擬信号および第二の反響模擬信号を収音信号から差引き誤差信号を取得し、この誤差信号を出力する誤差算出手段と、及び誤差信号を均等割合で配分して得られる第一の推定値と第二の推定値を出力する均等誤差配分手段か、誤差信号を第一の再生信号の電力と第二の再生信号の電力との比に応じて配分して得られる第三の推定値と第四の推定値を出力する信号電力加重誤差配分手段との何れかを備えた反響消去装置は上記したように従来より存在するが、
本発明の特徴とする構成はこれら均等誤差配分手段と、信号電力加重誤差配分手段の双方を備える点と、均等誤差配分手段は誤差信号を均等割合で配分して得られる第一の推定値と第二の推定値とを出力する点と、信号電力加重誤差配分手段は第一の再生信号の電力と第二の再生信号の電力との比に応じて配分して得られる第三の推定値と第四の推定値を出力する点を特徴とし、更に第一の再生信号の電力と第二の再生信号の電力の大小関係に応じて異なる値を電力偏差として出力する電力偏差計算手段と、
電力偏差の値が、第一の再生信号の電力が第二の再生信号の電力より大きいことを示すほど、第三の推定値を第一の推定値よりも大きな割合で加重平均し、電力偏差の値が、第二の再生信号の電力が第一の再生信号の電力より大きいことを示すほど、第一の推定値を第三の推定値よりも大きな割合で加重平均し、第五の推定値として出力し、第一の適応型模擬反響路に与え、第一の適応型模擬反響路が有する模擬特性を更新させる第一の電力偏差加重平均手段と、電力偏差の値が、第一の再生信号の電力が第二の再生信号の電力より大きいことを示すほど、第二の推定値を第四の推定値よりも大きな割合で加重平均し、電力偏差の値が、第二の再生信号の電力が第一の再生信号の電力より大きいことを示すほど、第四の推定値を第二の推定値よりも大きな割合で加重平均し、第六の推定値として出力し、第二の適応型模擬反響路に与え、第二の適応型模擬反響路が有する模擬特性を更新させる第二の電力偏差加重平均手段とを備える構成とした点を特徴とする。 The echo canceling apparatus according to the present invention is connected to the first sound reproducing means via the first echo path which is an acoustic transmission path between the first sound reproducing means and the sound collecting means existing in the same space. A first reverberation signal in which a first reproduction signal to be reproduced is picked up by the sound pickup means; a second sound reproduction means and a sound pickup means that exist in the same space of the first sound reproduction means; The second reproduction signal reproduced from the second sound reproducing means is mixed with the first echo signal by the sound collecting means via the second echo path which is an acoustic transmission path between the two. In the echo canceling device for canceling or reducing the second echo signal to be collected from all the collected sound signals picked up by the sound pickup means, the first echo path has a simulation characteristic, A first adaptive type that inputs a playback signal and outputs a first echo simulation signal that simulates the first echo signal A second adaptive simulated echo that has simulated characteristics of a pseudo echo path and a second echo path, inputs a second reproduction signal, and outputs a second echo simulation signal that simulates the second echo signal. A path, a first echo simulation signal, and a second echo simulation signal are obtained by subtracting an error signal from the collected signal, and an error calculation means for outputting the error signal, and the error signal is distributed at an equal ratio. An equal error distribution means for outputting the first estimated value and the second estimated value obtained, or by distributing the error signal according to the ratio of the power of the first reproduction signal and the power of the second reproduction signal. As described above, there is a conventional echo canceller including any one of the third estimated value and the signal power weighted error distribution means for outputting the fourth estimated value.
The characteristic feature of the present invention is that both of the equal error distribution means and the signal power weighted error distribution means are provided, and the equal error distribution means has a first estimated value obtained by distributing the error signal at an equal ratio. The point of outputting the second estimated value and the third estimated value obtained by distributing the signal power weighted error distribution means according to the ratio of the power of the first reproduced signal and the power of the second reproduced signal And a fourth estimated value, and further, a power deviation calculating means for outputting a value different as a power deviation according to the magnitude relationship between the power of the first reproduced signal and the power of the second reproduced signal;
As the power deviation value indicates that the power of the first reproduced signal is greater than the power of the second reproduced signal, the third estimated value is weighted and averaged at a rate greater than the first estimated value. As the value of indicates that the power of the second reproduced signal is greater than the power of the first reproduced signal, the first estimated value is weighted average at a rate greater than the third estimated value, and the fifth estimated Output as a value, given to the first adaptive simulated echo path, the first power deviation weighted average means for updating the simulation characteristics of the first adaptive simulated echo path, the value of the power deviation is the first The second estimated value is weighted and averaged at a rate greater than the fourth estimated value, indicating that the power of the reproduced signal is greater than the power of the second reproduced signal, and the power deviation value is the second reproduced signal. The fourth estimate is the second estimate as the power of is greater than the power of the first playback signal. The second power deviation weight is weighted average with a larger ratio, output as a sixth estimated value, given to the second adaptive simulated echo path, and updating the simulated characteristics of the second adaptive simulated echo path It is characterized by having a configuration comprising an averaging means.

本発明による反響消去装置は更に、電力偏差計算手段において、電力偏差として、第一の再生信号の電力と第二の再生信号の電力の差を、第一の再生信号の電力と第二の再生信号の和で除した値を用いることを特徴とする。
本発明による反響消去装置は更に、第一の電力偏差加重平均手段において、第五の推定値を求める第一の推定値と第三の推定値の加重平均の値を、１から電力偏差を引いた値の二分の一と第一の推定値との積と、１に電力偏差を加えた値の二分の一と第三の推定値との積と、の和として得ることを特徴とする。 The echo canceling apparatus according to the present invention further includes, in the power deviation calculation means, the difference between the power of the first reproduction signal and the power of the second reproduction signal as the power deviation, and the power of the first reproduction signal and the second reproduction signal. A value obtained by dividing the sum of signals is used.
The echo canceling apparatus according to the present invention further subtracts the power deviation from 1 for the weighted average value of the first estimated value and the third estimated value for obtaining the fifth estimated value in the first power deviation weighted average means. The product of a half of the obtained value and the first estimated value and the product of the half of the value obtained by adding 1 to the power deviation and the third estimated value are obtained.

本発明による反響消去装置は更に、第二の電力偏差加重平均手段において、第六の推定値を求める第二の推定値と第四の推定値の加重平均の値を、１に電力偏差を加えた値の二分の一と第二の推定値との積と、１から電力偏差を引いた値の二分の一と第四の推定値との積と、の和として得ることを特徴とする。
更に具体的な説明をするならば、本発明では、背景技術で説明した図２の誤差配分と図３の誤差配分を、αの値に応じて、組み合わせることにより、前記課題を解決する。図２の構成において、ema(k)の推定精度が高くなるのは、eu(k)が十分小さい、すなわち、αが、１、または、−１に近いときである。さらに、（式３）、（式４）を、αを用いて表すと、

Figure 0004504892
であるから、αが１に近いときは、推定値em2(k)に比べ、推定値em1(k)のほうが、eu(k)の占める割合が小さく、推定精度が高くなり、逆に、αが−１に近いときは、推定値em1(k)に比べ、推定値em2(k)のほうが、eu(k)に占める割合が小さく、推定精度が高くなる。この性質を踏まえ、図２の構成で用いる推定値em1(k)、em2(k)と、図３の構成で用いる、αの値に依存しない、推定値em1(k)、em2(k)とを組合せる。αの値が１に近づくことは、第一の再生信号ベクトルX1(k)の大きさが第二の再生信号ベクトルX2(k)の大きさよりも大きくなることを意味し、αの値が−１に近づくことは、第一の再生信号ベクトルX1(k)の大きさが第二の再生信号ベクトルX2(k)の大きさよりも小さくなることを意味する。 The echo canceling apparatus according to the present invention further adds a power deviation to a weighted average value of the second estimated value and the fourth estimated value for obtaining the sixth estimated value in the second power deviation weighted average means. It is characterized in that it is obtained as the sum of the product of one half of the obtained value and the second estimated value and the product of one half of the value obtained by subtracting the power deviation from 1 and the fourth estimated value.
More specifically, in the present invention, the above problem is solved by combining the error distribution of FIG. 2 described in the background art and the error distribution of FIG. 3 according to the value of α. In the configuration of FIG. 2, the estimation accuracy of ema (k) is high when eu (k) is sufficiently small, that is, when α is close to 1 or −1. Furthermore, when (Equation 3) and (Equation 4) are expressed using α,
Figure 0004504892
 Therefore, when α is close to 1, the estimated value em1 (k) has a smaller proportion of eu (k) and the estimated accuracy is higher than the estimated value em2 (k). Is close to −1, the estimated value em2 (k) has a smaller proportion of eu (k) than the estimated value em1 (k), and the estimation accuracy is high. Based on this property, the estimated values em1 (k) and em2 (k) used in the configuration of FIG. 2 and the estimated values em1 (k) and em2 (k) used in the configuration of FIG. Combine. When the value of α approaches 1, it means that the first reproduction signal vector X1 (k) is larger than the second reproduction signal vector X2 (k), and the value of α is − To approach 1 means that the magnitude of the first reproduction signal vector X1 (k) is smaller than the magnitude of the second reproduction signal vector X2 (k).

そこで、０以上１以下の値を取る変数γ１に対して、第一の再生信号ベクトルX1(k)の大きさが第二の再生信号ベクトルX2(k)の大きさより大きくなるほど、小さい値を与え、０以上１以下の値を取る変数γ２に対して、第一の再生信号ベクトルX1(k)の大きさが第二の再生信号ベクトルX2(k)の大きさよりも小さくなるほど、小さい値を与えるものとして、以下のように組合せて得られる第五推定値em5(k)、及び第六の推定値em6(k)を、本発明では用いる。
なお、本発明では、第一の再生信号の大きさと第二の再生信号の大きさの大小関係を表す量を電力偏差と呼び、例えば、前述のαを電力偏差として用いる。

Figure 0004504892
なる関係を用いた。 Therefore, a smaller value is given to the variable γ1 that takes a value between 0 and 1 as the first reproduction signal vector X1 (k) is larger than the second reproduction signal vector X2 (k). For a variable γ2 that takes a value between 0 and 1, a smaller value is given as the magnitude of the first reproduction signal vector X1 (k) becomes smaller than the magnitude of the second reproduction signal vector X2 (k). As an example, the fifth estimated value em5 (k) and the sixth estimated value em6 (k) obtained in combination as described below are used in the present invention.
In the present invention, an amount representing the magnitude relationship between the magnitude of the first reproduction signal and the magnitude of the second reproduction signal is called a power deviation, and for example, the aforementioned α is used as the power deviation.
Figure 0004504892
 The following relationship was used.

本発明による反響消去装置は、（式１）、（式２）の中でe1(k)、e2(k)の代わりに適用する第五の推定値em5(k)と第六の推定値em6(k)の算出において、従来の（式３）、（式４）による推定値、および、従来の（式５）、（式６）による推定値それぞれの長所を活かすため、それらの貢献度を、eu(k)のパワー期待値と、e1(k)からeu(k)を除いた成分とパワー期待値、および、e2(k)から−eu(k)を除いた成分とのパワー期待値それぞれとの比に応じ、上記２種類の従来の推定値の貢献度を制御する手段を設けたため、従来の推定値よりも精度を高めることができる。このため、第一模擬特性Hm1(k)および第二模擬特性Hm2(k)の模擬精度を従来の二つの技術より高めることができる。   The echo canceling apparatus according to the present invention includes a fifth estimated value em5 (k) and a sixth estimated value em6 applied in place of e1 (k) and e2 (k) in (Expression 1) and (Expression 2). In the calculation of (k), in order to make use of the advantages of the conventional estimated values of (Equation 3) and (Equation 4) and the estimated values of the conventional (Equation 5) and (Equation 6), , The expected power value of eu (k), the expected power value of the component excluding eu (k) from e1 (k), and the expected power value of the component excluding −eu (k) from e2 (k) Since the means for controlling the contribution of the two types of conventional estimated values is provided in accordance with the ratio with each of them, the accuracy can be improved more than the conventional estimated values. For this reason, the simulation accuracy of the first simulation characteristic Hm1 (k) and the second simulation characteristic Hm2 (k) can be improved compared to the two conventional techniques.

本発明による反響消去装置はハードウェアによって構成することが可能であるが、最も簡素に実現するにはコンピュータに本発明で提案する反響消去プログラムをインストールし、コンピュータに本発明による反響消去装置として機能させる実施形態が最良の形態である。
コンピュータを本発明による反響消去装置として機能させる場合、コンピュータには第一の反響路の模擬特性を有し、第一の再生信号を入力し、第一の反響信号を模擬する第一の反響模擬信号を生成する第一の適応型模擬反響路と、
第二の反響路の模擬特性を有し、第二の再生信号を入力し、第二の反響信号を模擬する第二の反響模擬信号を生成する第二の適応型模擬反響路と、
第一の反響模擬信号および第二の模擬信号を収音信号から差引き誤差信号を得る誤差産出手段と
誤差信号を均等割合で配分して得られる第一の推定値と第二の推定値を出力する均等誤差配分手段と、
誤差信号を第一の再生信号の電力と第二の再生信号の電力との比に応じて配分して得られる第三の推定値と第四の推定値を出力する信号電力加重誤差配分手段と、
第一の再生信号の電力と第二の再生信号の電力の大小関係に応じて異なる値を電力偏差として出力する電力偏差計算手段と、
電力偏差の値が、第一の再生信号の電力が第二の再生信号の電力より大きいことを示すほど、第三の推定値を第一の推定値よりも大きな割合で加重平均し、電力偏差の値が、第二の再生信号の電力が第一の再生信号の電力より大きいことを示すほど、第一の推定値を第三の推定値よりも大きな割合で加重平均し、第五の推定値として出力し、第一の適応型模擬反響路に与え、第一の適応型模擬反響路が有する模擬特性を更新させる第一の電力偏差加重平均手段と、
電力偏差の値が、第一の再生信号の電力が第二の再生信号の電力より大きいことを示すほど、第二の推定値を第四の推定値よりも大きな割合で加重平均し、電力偏差の値が、第二の再生信号の電力が第一の再生信号の電力より大きいことを示すほど、第四の推定値を第二の推定値よりも大きな割合で加重平均し、第六の推定値として出力し、第二の適応型模擬反響路に与え、第二の適応型模擬反響路が有する模擬特性を更新させる第二の電力偏差加重平均手段とを構築し、反響消去装置として機能させる。 The echo canceling apparatus according to the present invention can be configured by hardware, but in order to realize the simplest, the echo canceling program proposed in the present invention is installed in a computer, and the computer functions as the echo canceling apparatus according to the present invention. The embodiment to be made is the best mode.
When the computer functions as the echo canceling apparatus according to the present invention, the computer has a first echo path simulation characteristic, the first reproduction signal is input, and the first echo simulation that simulates the first echo signal A first adaptive simulated echo path for generating a signal;
A second adaptive simulated reverberation path that has a second reverberation path simulation characteristic, inputs a second reproduction signal, and generates a second reverberation simulation signal that simulates the second reverberation signal;
An error generating means for subtracting the first echo simulation signal and the second simulation signal from the collected sound signal and the first estimate value and the second estimate value obtained by allocating the error signal in equal proportions Uniform error distribution means for outputting;
Signal power weighted error distribution means for outputting a third estimated value and a fourth estimated value obtained by allocating the error signal according to the ratio of the power of the first reproduced signal and the power of the second reproduced signal; ,
Power deviation calculation means for outputting a value different as a power deviation according to the magnitude relationship between the power of the first reproduction signal and the power of the second reproduction signal;
As the power deviation value indicates that the power of the first reproduced signal is greater than the power of the second reproduced signal, the third estimated value is weighted and averaged at a rate greater than the first estimated value. As the value of indicates that the power of the second reproduced signal is greater than the power of the first reproduced signal, the first estimated value is weighted average at a rate greater than the third estimated value, and the fifth estimated A first power deviation weighted averaging means for outputting to the first adaptive simulated echo path as a value and updating a simulation characteristic of the first adaptive simulated echo path;
As the value of the power deviation indicates that the power of the first reproduction signal is larger than the power of the second reproduction signal, the second estimated value is weighted and averaged at a rate larger than the fourth estimated value, and the power deviation As the value of indicates that the power of the second reproduced signal is greater than the power of the first reproduced signal, the fourth estimated value is weighted and averaged at a rate greater than the second estimated value, and the sixth estimated A second power deviation weighted averaging means that outputs the value as a value, gives it to the second adaptive simulated echo path, and updates the simulation characteristics of the second adaptive simulated echo path, and functions as an echo canceller .

図１に本発明による反響消去装置３００の実施例を示す。この実施例において、第一スピーカとマイクロホン３間の第一反響路４のインパルス応答に相当する長さLのベクトルH1の第一模擬特性Hm1(k)を保持する第一適応型模擬反響路１０１と、第二スピーカ２とマイクロホン３間の第二反響路５のインパルス応答に相当する長さLのベクトルH2の第二模擬特性Hm2(k)を保持する第二適応型模擬反響路１０２と、第一適応型模擬反響路１０１と第二適応型模擬反響路１０２とにおいて、それぞれ、第一の再生信号x1(k)および第二の再生信号x2(k)を入力として第一模擬特性Hm1(k)および第二模擬特性Hm2(k)との畳込演算により生成される第一模擬反響信号dm1(k)および第二模擬反響信号dm2(k)を、第一反響路４および第二反響路５を経て生成される第一反響信号d1(k)および第二反響信号d2(k)が混合してマイクロホン３において得られる収音信号y(k)から差引き、誤差信号e(k)を出力する誤差算出手段３０４とから成る構成に加えて（式３）、（式４）に基づき、第一の再生信号x1(k)の電力と、第一の再生信号x1(k)の電力と第二の再生信号x2(k)の電力の和との比を誤差信号e(k)に乗ずることにより第三の推定値を出力するとともに、第二の再生信号x2(k)の電力と、第一の再生信号x1(k)の電力と第二の再生信号x2(k)の電力の和との比を誤差信号e(k)に乗ずることにより第四の推定値を出力する信号電力加重誤差配分手段１０５とを備えた図２の構成か又は（式５）、（式６）に基づき、誤差信号e(k)を二分の一に均等に分割し、第一の推定値と第二の推定値を出力する均等誤差配分手段２０５を備えた図３の構成は従来から知られている。   FIG. 1 shows an embodiment of an echo canceling apparatus 300 according to the present invention. In this embodiment, a first adaptive simulated echo path 101 that holds a first simulated characteristic Hm1 (k) of a vector H1 of length L corresponding to the impulse response of the first echo path 4 between the first speaker and the microphone 3 is used. A second adaptive simulated echo path 102 that holds a second simulated characteristic Hm2 (k) of a vector H2 of length L corresponding to the impulse response of the second echo path 5 between the second speaker 2 and the microphone 3; In the first adaptive simulated reverberation path 101 and the second adaptive simulated reverberation path 102, the first simulated characteristics Hm1 (1) are inputted with the first reproduction signal x1 (k) and the second reproduction signal x2 (k), respectively. k) and the second simulated characteristic signal Hm2 (k), the first simulated echo signal dm1 (k) and the second simulated echo signal dm2 (k) generated by the convolution calculation are used as the first echo path 4 and the second echo characteristic. The first echo signal d1 (k) and the second echo signal d2 (k) generated via the path 5 are mixed and sent to the microphone 3. In addition to the structure comprising the error calculation means 304 for subtracting the sound pickup signal y (k) obtained in this way and outputting the error signal e (k), the first reproduction is performed based on (Expression 3) and (Expression 4). The error signal e (k) is multiplied by the ratio of the power of the signal x1 (k) and the sum of the power of the first reproduction signal x1 (k) and the power of the second reproduction signal x2 (k) And the ratio of the power of the second reproduction signal x2 (k) and the sum of the power of the first reproduction signal x1 (k) and the power of the second reproduction signal x2 (k). The error signal e based on the configuration of FIG. 2 provided with the signal power weighted error distribution means 105 that outputs the fourth estimated value by multiplying the error signal e (k) or (Equation 5) or (Equation 6). The structure of FIG. 3 provided with the equal error distribution means 205 which divides | segments (k) equally by a half and outputs a 1st estimated value and a 2nd estimated value is conventionally known.

本発明の特徴とする構成は図２で説明した信号電力加重誤差配分手段１０５と図３で説明した均等誤差配分手段２０５とを共に有する点と、電力偏差計算手段３０１と、第一電力偏差加重平均手段３０２と、第二電力偏差加重平均手段３０３とを設けた構成を特徴とするものである。
以下本発明の特徴とする部分の動作を説明する。
均等誤差配分手段２０５は図３で説明したのと同様に、（式５）、（式６）に基づき、誤差信号e(k)を二分の一に均等分割し、第一の推定値em1(k)＝(1／2)・e(k)と第二推定値em2(k)＝(1／2)・e(k)を出力する。 The feature of the present invention is that it has both the signal power weighted error distribution means 105 described in FIG. 2 and the equal error distribution means 205 described in FIG. 3, a power deviation calculation means 301, and a first power deviation weighting. The present invention is characterized in that an averaging means 302 and a second power deviation weighted averaging means 303 are provided.
The operation of the characteristic features of the present invention will be described below.
The equal error distribution unit 205 equally divides the error signal e (k) by a half based on (Equation 5) and (Equation 6) as described in FIG. k) = (1/2) · e (k) and the second estimated value em2 (k) = (1/2) · e (k) are output.

信号電力加重誤差配分手段１０５は（式３）、（式４）に基づき、第一の再生信号x1(k)の電力‖x1(k)‖2と、第一の再生信号x1(k)の電力‖x1(k)‖2と第二の再生信号x2(k)の電力‖x2(k)‖2の和との比‖x1(k)‖2／（‖x1(k)‖2＋‖x2(k)‖2）を誤差信号e(k)に乗ずることにより第三推定値em3(k)を出力すると共に、第二の再生信号x2(k)との電力‖x2(k)‖2と、第一の再生信号x1(k)の電力‖x1(k)‖2との和との比‖x2(k)‖2／（‖x1(k)‖2＋‖x2(k)‖2）を誤差信号e(k)に乗ずることにより第四推定値em4(k)を出力する。
つまり、第三推定値em3(k)と第四推定値em4(k)は

Figure 0004504892
で求められる。
電力偏差計算手段３０１は第一の再生信号x1(k)の電力‖x1(k)‖2と第二の再生信号x2(k)の電力‖x2(k)‖2の差を、第一の再生信号x1(k)の電力と第二の再生信号x2(k)の電力の和で除した値を電力偏差αとして出力する。 Signal power weighted error distribution means 105 (equation 3), the basis of the equation (4), a power ‖X1 (k) ‖ 2 of the first reproduction signal x1 (k), the first reproduction signal x1 (k) power ‖x1 (k) ‖ 2 and the second ratio of the sum of the power ‖x2 (k) ‖ 2 of the reproduced signal x2 (k) ‖x1 (k) ‖ 2 / (‖x1 (k) ‖ 2 + ‖ x2 (k) ‖ 2 ) is multiplied by the error signal e (k) to output the third estimated value em3 (k) and the power of the second reproduction signal x2 (k) 再生 x2 (k) ‖ 2 When the ratio of the sum of the power ‖x1 (k) ‖ 2 of the first reproduction signal x1 (k) ‖x2 (k) ‖ 2 / (‖x1 (k) ‖ 2 + ‖x2 (k) ‖ 2 ) Is multiplied by the error signal e (k) to output a fourth estimated value em4 (k).
That is, the third estimated value em3 (k) and the fourth estimated value em4 (k) are
Figure 0004504892
 Is required.
The difference in power ‖x2 (k) ‖ 2 power ‖x1 (k) ‖ 2 and the second reproduction signal x2 (k) of the first reproduction signal x1 is power deviation calculation means 301 (k), the first A value obtained by dividing the power of the reproduction signal x1 (k) by the sum of the power of the second reproduction signal x2 (k) is output as a power deviation α.

Figure 0004504892
第一電力偏差加重平均手段３０２は、１から電力偏差αを引いた値の二分の一と第一推定値em1(k)との積(1／2)・(1−α)・em1(k)と、１に電力偏差αを加えた値の二分の一と第三の推定値em3(k)との積(1／2)・(1＋α)・em3(k)と、の和を第五推定値em5(k)
em5＝(1／2)・(1−α)・em1(k)＋(1／2)・(1＋α)・em3(k)
として出力し、この第五推定値em5(k)を第一適応型模擬反響路１０１に与え、第一模擬特性Hm1(k)を特性更新させる。
Figure 0004504892
 The first power deviation weighted averaging means 302 is a product (1/2) · (1−α) · em1 (k) of a half of the value obtained by subtracting the power deviation α from 1 and the first estimated value em1 (k). ) And 1/2 of the value obtained by adding the power deviation α to 1 and the third estimated value em3 (k) (1/2) · (1 + α) · em3 (k) Estimated value em5 (k)
em5 = (1/2) ・ (1−α) ・ em1 (k) + (1/2) ・ (1 + α) ・ em3 (k)
The fifth estimated value em5 (k) is given to the first adaptive simulated echo path 101, and the first simulated characteristic Hm1 (k) is updated.

第二電力偏差加重平均手段３０３は（式１４）に基づき、１に電力偏差αを加えた値の二分の一と第二推定値との積(1／2)・(1＋α)・em2(k)と、１から電力偏差αを引いた値の二分の一と第四推定値em4(k)との積(1／2)・(1−α)・em4(k)と、の和を第６推定値em6(k)
em6＝(1／2)・(1＋α)・em2(k)＋(1／2)・(1−α)・em4(k)
として出力し、この第六推定値em6(k)を第二適応型模擬反響路１０２に与え、第二模擬特性Hm2(k)を特性更新させる。 Based on (Equation 14), the second power deviation weighted average means 303 is a product (1/2) · (1 + α) · em2 (k) of a half of the value obtained by adding the power deviation α to 1 and the second estimated value. ) And the product of 1/2 of the value obtained by subtracting the power deviation α from 1 and the fourth estimated value em4 (k) (1/2) · (1−α) · em4 (k) 6 Estimated value em6 (k)
em6 = (1/2) ・ (1 + α) ・ em2 (k) + (1/2) ・ (1−α) ・ em4 (k)
And the sixth estimated value em6 (k) is given to the second adaptive simulated echo path 102 to update the second simulated characteristic Hm2 (k).

このように本発明によれば第三推定値em3(k)と第四推定値em4(k)の算出において、従来の（式３）、（式４）による推定値、および従来の（式５）、（式６）による推定値のそれぞれの長所を活かすため、従来の推定値よりも推定精度を高めることができる。この結果として第一模擬特性Hm1(k)および第二模擬特性の模擬精度を従来の二つの技術より高めることができ、複数チャネルの反響を高い精度で除去乃至は消去することができる。   Thus, according to the present invention, in the calculation of the third estimated value em3 (k) and the fourth estimated value em4 (k), the estimated values based on the conventional (Expression 3) and (Expression 4), and the conventional (Expression 5) ) And (6), the estimation accuracy can be improved as compared with the conventional estimation value. As a result, the simulation accuracy of the first simulation characteristic Hm1 (k) and the second simulation characteristic can be improved as compared with the conventional two techniques, and the echoes of a plurality of channels can be removed or eliminated with high accuracy.

図１の第一電力偏差加重平均手段３０２、および、第二電力偏差加重平均手段３０３において、（式１３）および（式１４）を適用するにあたり、（式１６）、（式１８）のSNR1、SNR2は、

Figure 0004504892
として、振幅期待値の比から計算してもよい。 In applying the (Equation 13) and (Equation 14) in the first power deviation weighted average means 302 and the second power deviation weighted average means 303 in FIG. 1, SNR1 in (Equation 16) and (Equation 18), SNR2 is
Figure 0004504892
 As an alternative, it may be calculated from the ratio of expected amplitude values.

図１の第一電力偏差加重平均手段３０２、および、第二電力偏差加重平均手段３０３において、（式１３）および（式１４）の代わりに

Figure 0004504892
を与えてもよい。
以上説明した本発明による反響消去装置はハードウェアによって構成することも可能であるが現実的ではない。本発明による反響消去装置を簡素に実現するにはコンピュータに本発明で提案する反響消去プログラムをインストールし、コンピュータに上述した反響消去装置として機能させる実施形態が最良の実施形態である。本発明による反響消去プログラムはコンピュータが解読可能なプログラム言語によって記述され、磁気ディスク或いはCD-ROM、半導体メモリのような記録媒体に記録される。コンピュータへのインストールはこれらの記録媒体から或いは通信回線を通じてインストールされる。コンピュータへインストールされた状態ではコンピュータに備えられたCPUまたはDSPで解読され、コンピュータを反響消去装置として機能させる。 In the first power deviation weighted average means 302 and the second power deviation weighted average means 303 in FIG. 1, instead of (Equation 13) and (Equation 14)
Figure 0004504892
 May be given.
The echo canceling apparatus according to the present invention described above can be configured by hardware, but it is not realistic. In order to simply realize the echo canceling apparatus according to the present invention, an embodiment in which the echo canceling program proposed in the present invention is installed in a computer and the computer functions as the echo canceling apparatus described above is the best embodiment. The echo cancellation program according to the present invention is described in a computer-readable program language, and is recorded on a recording medium such as a magnetic disk, a CD-ROM, or a semiconductor memory. The computer is installed from these recording media or through a communication line. When installed in the computer, it is decoded by a CPU or DSP provided in the computer, and the computer is made to function as an echo canceller.

本発明による反響消去装置はハンズフリー通話、ハンズフリー音声認識などへの適用が可能である。   The echo canceling apparatus according to the present invention can be applied to hands-free calling and hands-free voice recognition.

本発明による反響消去装置の一実施例を説明するためのブロック図。The block diagram for demonstrating one Example of the echo cancellation apparatus by this invention. 従来の反響消去装置を説明するためのブロック図。The block diagram for demonstrating the conventional echo cancellation apparatus. 従来の反響消去装置の他の例を説明するためのブロック図。The block diagram for demonstrating the other example of the conventional echo cancellation apparatus.

符号の説明Explanation of symbols

１ 第一スピーカ １０５ 信号電力加重誤差配分手段
２ 第二スピーカ ２０５ 均等誤差配分手段
３ マイクロホン ３０１ 電力偏差計算手段
４ 第一反響路 ３０２ 第一電力偏差加重平均手段
５ 第二反響路 ３０３ 第二電力偏差加重平均手段
１０１ 第一適応型模擬反響路
１０２ 第二適応型模擬反響路 DESCRIPTION OF SYMBOLS 1 1st speaker 105 Signal power weighting error distribution means 2 2nd speaker 205 Equal error distribution means 3 Microphone 301 Power deviation calculation means 4 1st echo path 302 1st power deviation weighted average means 5 2nd echo path 303 2nd power deviation Weighted average means 101 First adaptive simulated echo path 102 Second adaptive simulated echo path

Claims (10)

同一空間内で実行される第一の音響再生ステップと音響収音ステップとの間の音響的な伝達経路である第一の反響路を介して前記第一の音響再生ステップで再生する第一の再生信号が前記音響収音ステップで収音される第一の反響信号と、前記第一の音響再生ステップの同一の空間内で実行される第二の音響再生ステップと前記音響収音ステップとの間の音響的な伝達経路である第二の反響路を介して、前記第二の音響再生ステップで再生する第二の再生信号が前記音響収音ステップにより前記第一の反響信号と混合して収音される第二の反響信号とを、前記音響収音ステップで収音する全ての収音信号の中から消去または低減する反響消去方法において、
前記第一の反響路の模擬特性を有し、前記第一の再生信号を入力し、前記第一の反響信号を模擬する第一の反響模擬信号を生成する第一の適応型反響模擬ステップと、
前記第二の反響路の模擬特性を有し、前記第二の再生信号を入力し、前記第二の反響信号を模擬する第二の反響模擬信号を生成する第二の適応型反響模擬ステップと、
前記第一の反響模擬信号および前記第二の反響模擬信号を前記収音信号から差引き誤差信号を得る誤差算出ステップと、
前記誤差信号を均等割合で配分して得られる第一の推定値と第二の推定値を出力する均等誤差配分ステップと、
前記誤差信号を前記第一の再生信号の電力と前記第二の再生信号の電力との比に応じて配分して得られる第三の推定値と第四の推定値を出力する信号電力加重誤差配分ステップと、
前記第一の再生信号の電力と前記第二の再生信号の電力の大小関係に応じて異なる値を電力偏差として出力する電力偏差計算ステップと、
前記電力偏差の値が、前記第一の再生信号の電力が前記第二の再生信号の電力より大きいことを示すほど、前記第三の推定値を前記第一の推定値よりも大きな割合で加重平均し、前記電力偏差の値が、前記第二の再生信号の電力が前記第一の再生信号の電力より大きいことを示すほど、前記第一の推定値を前記第三の推定値よりも大きな割合で加重平均し、第五の推定値として出力し、前記第一の適応型反響模擬ステップに与え、前記第一の適応型反響模擬ステップが有する模擬特性を更新させる第一の電力偏差加重平均ステップと、
前記電力偏差の値が、前記第一の再生信号の電力が前記第二の再生信号の電力より大きいことを示すほど、前記第二の推定値を前記第四の推定値よりも大きな割合で加重平均し、前記電力偏差の値が、前記第二の再生信号の電力が前記第一の再生信号の電力よりも大きいことを示すほど、前記第四の推定値を前記第二の推定値よりも大きな割合で加重平均し、第六の推定値として出力し、前記第二の適応型反響模擬ステップに与え、前記第二の適応型反響模擬ステップが有する模擬特性を更新させる第二の電力偏差加重平均ステップと、
を含むことを特徴とする反響消去方法。 A first sound reproduced in the first sound reproduction step via a first echo path that is an acoustic transmission path between the first sound reproduction step and the sound pickup step executed in the same space. A first reverberation signal in which a reproduction signal is collected in the sound collection step, a second sound reproduction step executed in the same space of the first sound reproduction step, and the sound collection step. The second reproduction signal reproduced in the second sound reproduction step is mixed with the first echo signal in the sound collection step via the second echo path which is an acoustic transmission path between In the echo canceling method, the second echo signal to be collected is erased or reduced from all the collected sound signals collected in the acoustic sound collecting step.
A first adaptive reverberation simulation step having simulation characteristics of the first reverberation path, inputting the first reproduction signal, and generating a first reverberation simulation signal for simulating the first reverberation signal; ,
A second adaptive reverberation simulation step having simulation characteristics of the second reverberation path, inputting the second reproduction signal, and generating a second reverberation simulation signal for simulating the second reverberation signal; ,
An error calculating step of obtaining a subtraction error signal from the collected sound signal of the first echo simulation signal and the second echo simulation signal;
An equal error distribution step of outputting a first estimated value and a second estimated value obtained by distributing the error signal at an equal ratio;
A signal power weighting error that outputs a third estimated value and a fourth estimated value obtained by allocating the error signal according to the ratio of the power of the first reproduced signal and the power of the second reproduced signal. An allocation step;
A power deviation calculation step of outputting a value different as a power deviation according to the magnitude relationship between the power of the first reproduction signal and the power of the second reproduction signal;
The third estimated value is weighted at a rate greater than the first estimated value, as the power deviation value indicates that the power of the first reproduced signal is greater than the power of the second reproduced signal. On average, the first estimated value is larger than the third estimated value as the power deviation value indicates that the power of the second reproduced signal is greater than the power of the first reproduced signal. A first power deviation weighted average that is weighted averaged by a ratio, output as a fifth estimated value, given to the first adaptive echo simulation step, and updates a simulation characteristic of the first adaptive echo simulation step Steps,
The value of the power deviation indicates that the power of the first reproduction signal is greater than the power of the second reproduction signal, and the second estimation value is weighted at a rate larger than the fourth estimation value. On average, the value of the power deviation indicates that the power of the second reproduction signal is greater than the power of the first reproduction signal, so that the fourth estimated value is more than the second estimated value. Weighted average with a large proportion, output as a sixth estimated value, given to the second adaptive echo simulation step, the second power deviation weight to update the simulation characteristics of the second adaptive echo simulation step Average step and
The echo cancellation method characterized by including. 前記電力偏差計算ステップにおいて、電力偏差として、前記第一の再生信号の電力と前記第二の再生信号の電力の差を、前記第一の再生信号の電力と前記第二の再生信号の和で除した値を用いることを特徴とする請求項１記載の反響消去方法。   In the power deviation calculating step, as a power deviation, a difference between the power of the first reproduction signal and the power of the second reproduction signal is calculated by adding the power of the first reproduction signal and the second reproduction signal. 2. The echo canceling method according to claim 1, wherein the divided value is used. 前記第一の電力偏差加重平均ステップにおいて、前記第五の推定値を求める前記第一の推定値と前記第三の推定値の加重平均の値を、１から請求項２に基づく電力偏差を引いた値の二分の一と前記第一の推定値との積と、１に請求項２に基づく前記電力偏差を加えた値の二分の一と前記第三の推定値との積と、の和として得ることを特徴とする請求項１又は２の何れかに記載の反響消去方法。   In the first power deviation weighted average step, the weighted average value of the first estimated value and the third estimated value for obtaining the fifth estimated value is subtracted from 1 to the power deviation based on claim 2. The product of one half of the measured value and the first estimated value, and the product of one half of the value obtained by adding the power deviation according to claim 2 to the third estimated value. The echo canceling method according to claim 1, wherein the echo canceling method is obtained as follows. 前記第二の電力偏差加重平均ステップにおいて、前記第六の推定値を求める前記第二の推定値と前記第四の推定値の加重平均の値を、
１に請求項２に基づく電力偏差を加えた値の二分の一と前記第二の推定値との積と、１から請求項２に基づく電力偏差を引いた値の二分の一と前記第四の推定値との積と、の和として得ることを特徴とする請求項１乃至３の何れかに記載の反響消去方法。 In the second power deviation weighted average step, a value of a weighted average of the second estimated value and the fourth estimated value for obtaining the sixth estimated value,
The product of one half of the value obtained by adding the power deviation based on claim 2 to the second estimated value and one half of the value obtained by subtracting the power deviation based on claim 2 from 1 and the fourth The echo canceling method according to claim 1, wherein the echo canceling method is obtained as a sum of a product of the estimated value and the estimated value. 同一空間内に存在する第一の音響再生手段と音響収音手段との間の音響的な伝達経路である第一の反響路を介して前記第一の音響再生手段から再生する第一の再生信号が前記音響収音手段により収音される第一の反響信号と、前記第一の音響再生手段の同一の空間内に存在する第二の音響再生手段と前記音響収音手段との間の音響的な伝達経路である第二の反響路を介して、前記第二の音響再生手段から再生する第二の再生信号が前記音響収音手段により前記第一の反響信号と混合して収音される第二の反響信号とを、前記音響収音手段が収音する全ての収音信号の中から消去または低減する反響消去装置において、
前記第一の反響路の模擬特性を有し、前記第一の再生信号を入力し、前記第一の反響信号を模擬する第一の反響模擬信号を生成する第一の適応型模擬反響路と、
前記第二の反響路の模擬特性を有し、前記第二の再生信号を入力し、前記第二の反響信号を模擬する第二の反響模擬信号を生成する第二の適応型模擬反響路と、
前記第一の反響模擬信号および前記第二の反響模擬信号を前記収音信号から差引き誤差信号を得て、この誤差信号を出力する誤差算出手段と、
前記誤差信号を均等割合で配分して得られる第一の推定値と第二の推定値を出力する均等誤差配分手段と、
前記誤差値を前記第一の再生信号の電力と前記第二の再生信号の電力との比に応じて配分して得られる第三の推定値と第四の推定値を出力する信号電力加重誤差配分手段と、
前記第一の再生信号の電力と前記第二の再生信号の電力の大小関係に応じて異なる値を電力偏差として出力する電力偏差計算手段と、
前記電力偏差の値が、前記第一の再生信号の電力が前記第二の再生信号の電力より大きいことを示すほど、前記第三の推定値を前記第一の推定値よりも大きな割合で加重平均し、前記電力偏差の値が、前記第二の再生信号の電力が前記第一の再生信号の電力より大きいことを示すほど、前記第一の推定値を前記第三の推定値よりも大きな割合で加重平均し、第五の推定値として出力し、前記第一の適応型模擬反響路に与え、前記第一の適応型模擬反響路が有する模擬特性を更新させる第一の電力偏差加重平均手段と、
前記電力偏差の値が、前記第一の再生信号の電力が前記第二の再生信号の電力より大きいことを示すほど、前記第二の推定値を前記第四の推定値よりも大きな割合で加重平均し、前記電力偏差の値が、前記第二の再生信号の電力が前記第一の再生信号の電力より大きいことを示すほど、前記第四の推定値を前記第二の推定値よりも大きな割合で加重平均し、第六の推定値として出力し、前記第二の適応型模擬反響路に与え、前記第二の適応型模擬反響路が有する模擬特性を更新させる第二の電力偏差加重平均手段と、
を有することを特徴とする反響消去装置。 1st reproduction | regeneration reproduced | regenerated from said 1st sound reproduction | regeneration means via the 1st echo path which is an acoustic transmission path | route between the 1st sound reproduction | regeneration means and sound collection means which exist in the same space Between the first reverberation signal in which the signal is picked up by the sound pickup means, and between the second sound reproduction means and the sound pickup means existing in the same space of the first sound reproduction means. A second reproduction signal reproduced from the second sound reproducing means is mixed with the first echo signal by the acoustic sound collecting means via a second echo path that is an acoustic transmission path, and collected. In the echo canceling apparatus for canceling or reducing the second echo signal to be collected from all the collected sound signals collected by the acoustic sound collecting means,
A first adaptive simulated reverberation path having simulation characteristics of the first reverberation path, receiving the first reproduction signal and generating a first reverberation simulation signal for simulating the first reverberation signal; ,
A second adaptive simulated reverberation path having simulation characteristics of the second reverberation path, generating the second reverberation simulation signal for inputting the second reproduction signal and simulating the second reverberation signal; ,
An error calculating means for obtaining a subtraction error signal from the collected sound signal and outputting the error signal of the first echo simulation signal and the second echo simulation signal;
Equal error distribution means for outputting a first estimated value and a second estimated value obtained by distributing the error signal at an equal ratio;
A signal power weighting error that outputs a third estimated value and a fourth estimated value obtained by allocating the error value according to the ratio of the power of the first reproduced signal and the power of the second reproduced signal. Distribution means,
A power deviation calculating means for outputting a value different as a power deviation according to the magnitude relationship between the power of the first reproduction signal and the power of the second reproduction signal;
The third estimated value is weighted at a rate greater than the first estimated value, as the power deviation value indicates that the power of the first reproduced signal is greater than the power of the second reproduced signal. On average, the first estimated value is larger than the third estimated value as the power deviation value indicates that the power of the second reproduced signal is greater than the power of the first reproduced signal. A first power deviation weighted average that is weighted averaged as a ratio, output as a fifth estimated value, given to the first adaptive simulated echo path, and updates the simulated characteristics of the first adaptive simulated echo path Means,
The value of the power deviation indicates that the power of the first reproduction signal is greater than the power of the second reproduction signal, and the second estimation value is weighted at a rate larger than the fourth estimation value. On average, the fourth estimated value is larger than the second estimated value as the power deviation value indicates that the power of the second reproduced signal is larger than the power of the first reproduced signal. Weighted average with a ratio, output as a sixth estimated value, given to the second adaptive simulated echo path, and a second power deviation weighted average that updates the simulation characteristics of the second adaptive simulated echo path Means,
An echo canceling apparatus characterized by comprising: 前記電力偏差計算手段において、電力偏差として、前記第一の再生信号の電力と前記第二の再生信号の電力の差を、前記第一の再生信号の電力と前記第二の再生信号の和で除した値を用いることを特徴とする請求項５記載の反響消去装置。   In the power deviation calculation means, as a power deviation, a difference between the power of the first reproduction signal and the power of the second reproduction signal is obtained by adding the power of the first reproduction signal and the second reproduction signal. 6. The echo canceling apparatus according to claim 5, wherein the divided value is used. 前記第一の電力偏差加重平均手段において、前記第五の推定値を求める前記第一の推定値と前記第三の推定値の加重平均の値を、１から請求項６に基づく電力偏差を引いた値の二分の一と前記第一の推定値との積と、１に請求項６に基づく前記電力偏差を加えた値の二分の一と前記第三の推定値との積と、の和として得ることを特徴とする請求項５又は６の何れかに記載の反響消去装置。   In the first power deviation weighted average means, the weighted average value of the first estimated value and the third estimated value for obtaining the fifth estimated value is subtracted from 1 to the power deviation based on claim 6. The product of one half of the measured value and the first estimated value, and the product of one half of the value obtained by adding the power deviation according to claim 6 to the third estimated value. The echo canceling device according to claim 5, wherein the echo canceling device is obtained as follows. 前記第二の電力偏差加重平均手段において、前記第六の推定値を求める前記第二の推定値と前記第四の推定値の加重平均の値を、
１に請求項６に基づく電力偏差を加えた値の二分の一と前記第二の推定値との積と、１から請求項６に基づく電力偏差を引いた値の二分の一と前記第四の推定値との積と、の和として得ることを特徴とする請求項５乃至７の何れかに記載の反響消去装置。 In the second power deviation weighted average means, a weighted average value of the second estimated value and the fourth estimated value for obtaining the sixth estimated value,
The product of one half of the value obtained by adding the power deviation based on claim 6 to the second estimated value and one half of the value obtained by subtracting the power deviation based on 1 from 6 and the fourth The echo canceling apparatus according to claim 5, wherein the echo canceling apparatus is obtained as a sum of a product with an estimated value of コンピユータが解読可能なプログラム言語によって記述され、コンピユータに請求項５乃至８の何れかに記載の反響消去装置として機能させる反響消去プログラム。   9. An echo canceling program that is described in a program language that can be read by a computer, and that causes the computer to function as the echo canceling device according to claim 5. コンピュータが読み取り可能な記録媒体によって構成され、この記録媒体に請求項９記載の反響消去プログラムを記録した記録媒体。   A recording medium comprising a computer-readable recording medium, wherein the echo canceling program according to claim 9 is recorded on the recording medium.
JP2005251648A 2005-08-31 2005-08-31 Echo canceling method, echo canceling apparatus, program, recording medium Expired - Fee Related JP4504892B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2005251648A JP4504892B2 (en) 2005-08-31 2005-08-31 Echo canceling method, echo canceling apparatus, program, recording medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2005251648A JP4504892B2 (en) 2005-08-31 2005-08-31 Echo canceling method, echo canceling apparatus, program, recording medium

Publications (2)

Publication Number Publication Date
JP2007067855A JP2007067855A (en) 2007-03-15
JP4504892B2 true JP4504892B2 (en) 2010-07-14

Family

ID=37929516

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2005251648A Expired - Fee Related JP4504892B2 (en) 2005-08-31 2005-08-31 Echo canceling method, echo canceling apparatus, program, recording medium

Country Status (1)

Country Link
JP (1) JP4504892B2 (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5438629B2 (en) * 2010-08-23 2014-03-12 日本電信電話株式会社 Stereo echo canceling method, stereo echo canceling device, stereo echo canceling program

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002261659A (en) * 2001-02-27 2002-09-13 Nippon Telegr & Teleph Corp <Ntt> Multi-channel echo cancellation method, its apparatus, its program, and its storage medium
JP2002261660A (en) * 2001-02-27 2002-09-13 Nippon Telegr & Teleph Corp <Ntt> Multi-channel echo cancellation method, its apparatus, its program, and its storage medium

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002261659A (en) * 2001-02-27 2002-09-13 Nippon Telegr & Teleph Corp <Ntt> Multi-channel echo cancellation method, its apparatus, its program, and its storage medium
JP2002261660A (en) * 2001-02-27 2002-09-13 Nippon Telegr & Teleph Corp <Ntt> Multi-channel echo cancellation method, its apparatus, its program, and its storage medium

Also Published As

Publication number Publication date
JP2007067855A (en) 2007-03-15

Similar Documents

Publication Publication Date Title
EP3791565B1 (en) Method and apparatus utilizing residual echo estimate information to derive secondary echo reduction parameters
JP4954334B2 (en) Apparatus and method for calculating filter coefficients for echo suppression
JP5102371B2 (en) Apparatus and method for calculating control information of echo suppression filter, and apparatus and method for calculating delay value
CN103137136B (en) Sound processing device
JP2008517317A (en) Audio data processing system, method, program element, and computer readable medium
JP5762479B2 (en) Voice switch device, voice switch method, and program thereof
JP4504892B2 (en) Echo canceling method, echo canceling apparatus, program, recording medium
JP4504891B2 (en) Echo canceling method, echo canceling apparatus, program, recording medium
JP2003309493A (en) Method, device and program for reducing echo
JP5937451B2 (en) Echo canceling apparatus, echo canceling method and program
JP2003250193A (en) Echo elimination method, device for executing the method, program and recording medium therefor
JP5438629B2 (en) Stereo echo canceling method, stereo echo canceling device, stereo echo canceling program
JP5583181B2 (en) Cascade connection type transmission system parameter estimation method, cascade connection type transmission system parameter estimation device, program
JP5774062B2 (en) Echo canceling apparatus, echo canceling method, and program thereof
JP5826712B2 (en) Multi-channel echo canceling apparatus, multi-channel echo canceling method, and program
JP2017191992A (en) Echo suppressor, method therefor, program, and recording medium
JP2017191987A (en) Echo suppressor, method thereof, program, and record medium
JP6790659B2 (en) Sound processing equipment and sound processing method
JP4559989B2 (en) Echo suppression method, echo suppression program, recording medium
JP6075783B2 (en) Echo canceling apparatus, echo canceling method and program
JP5925149B2 (en) Acoustic coupling amount estimating apparatus, echo canceling apparatus, method and program thereof
Cifani et al. A PEM-AFROW based algorithm for acoustic feedback control in automotive speech reinforcement systems
CN113841198B (en) Signal component estimation using coherence
JP2002261659A (en) Multi-channel echo cancellation method, its apparatus, its program, and its storage medium
JP5325134B2 (en) Echo canceling method, echo canceling apparatus, program thereof, and recording medium

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20070810

TRDD Decision of grant or rejection written
A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

Effective date: 20100413

A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20100423

R150 Certificate of patent or registration of utility model

Ref document number: 4504892

Country of ref document: JP

Free format text: JAPANESE INTERMEDIATE CODE: R150

Free format text: JAPANESE INTERMEDIATE CODE: R150

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20130430

Year of fee payment: 3

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20140430

Year of fee payment: 4

S531 Written request for registration of change of domicile

Free format text: JAPANESE INTERMEDIATE CODE: R313531

R350 Written notification of registration of transfer

Free format text: JAPANESE INTERMEDIATE CODE: R350

LAPS Cancellation because of no payment of annual fees