JPH07154306A - Acoustic echo canceller - Google Patents
Acoustic echo cancellerInfo
- Publication number
- JPH07154306A JPH07154306A JP29937493A JP29937493A JPH07154306A JP H07154306 A JPH07154306 A JP H07154306A JP 29937493 A JP29937493 A JP 29937493A JP 29937493 A JP29937493 A JP 29937493A JP H07154306 A JPH07154306 A JP H07154306A
- Authority
- JP
- Japan
- Prior art keywords
- circuit
- reception signal
- signal input
- input terminal
- pseudo
- 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.)
- Pending
Links
Landscapes
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
Abstract
(57)【要約】
【目的】従属変数に外乱として混入する近端雑音の時間
的変化を相対的に検出し、その状態に合わせた係数修正
ループゲインの閾値を逐次的に変更する事で、安定性と
真値への収束特性をを向上させた音響反響除去装置を提
供する事を目的とする。
【構成】最大周期系列符号の様な既知の低電力レベルの
擬似白色雑音信号を重畳させた参照信号を音場へ入力
し、音場からの出力である送話信号の電力と擬似白色雑
音信号電力の差を第一の評価値とし、誤差信号電力を第
二の入力として係数修正ループゲインの内挿閾値を回線
状態に対応させて、動作安定性と真値への収束速度を向
上させることができる。
(57) [Summary] [Purpose] By relatively detecting the temporal change of near-end noise mixed as a disturbance in the dependent variable and sequentially changing the coefficient correction loop gain threshold value according to the state, An object of the present invention is to provide an acoustic echo canceller with improved stability and convergence to a true value. [Structure] A reference signal on which a pseudo white noise signal of a known low power level such as a maximum period sequence code is superimposed is input to the sound field, and the power of the transmission signal as the output from the sound field and the pseudo white noise signal Using the power difference as the first evaluation value, the error signal power as the second input, and making the interpolation threshold of the coefficient correction loop gain correspond to the line state to improve the operation stability and the convergence speed to the true value. You can
Description
【0001】[0001]
【産業上の利用分野】本発明は、通信回線、室内音場制
御装置そして高品質な音声通信会議装置に使用され、受
話径路の信号が送話径路を介して現れる音響残響成分を
除去する音響反響除去装置に関する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used in a communication line, a room sound field control device, and a high-quality voice communication conference device, and removes acoustic reverberation components in which a signal on a receiving path appears through a transmitting path. The present invention relates to an echo canceller.
【0002】[0002]
【従来の技術】一般に音響反響除去装置は通信衛生およ
び海底ケーブルを利用した長距離電話回線において、2
線4線変換器のインピーダンス不整合により生ずる反射
を除去するものと、テレビ会議システムなどの拡声電話
において、話者音声の音響結合による反響を除去するも
のとに大別でき、修正量演算回路、擬似エコーを発生す
る可変係数フィルタおよび減算回路から構成されてい
る。以下に音響反響除去装置の基本動作を述べる。2. Description of the Related Art Generally, an acoustic echo canceller is used in communication hygiene and long-distance telephone lines using a submarine cable.
It can be roughly divided into one that removes reflection caused by impedance mismatch of the line-to-four-line converter and one that removes reverberation due to acoustic coupling of speaker's voice in a loudspeaker telephone such as a video conference system. It is composed of a variable coefficient filter that generates a pseudo echo and a subtraction circuit. The basic operation of the acoustic echo canceller will be described below.
【0003】図3は音響反響除去装置の基本構成を示す
図である。受話信号入力端1は受話信号出力2に接続さ
れ、その受話信号入力端子1の受話信号は可変係数フィ
ルタ3に分岐供給され、擬似エコーを生成させる。送話
信号入力端子4からの送話信号と可変係数フィルタ3の
出力である擬似エコーは減算回路5へ入力され、送話信
号中のエコー成分が除去され、その減算回路5の出力は
送話信号出力端子6へ出力される。送話信号出力端子6
の出力と受話信号入力端子1の信号が修正量演算回路7
に入力され、修正量演算回路7の出力により可変係数フ
ィルタ3のフィルタ係数が修正される。可変係数フィル
タ3内で受話信号は受話信号入力レジスタ8に入力さ
れ、その受話信号入力レジスタ8の受話信号と擬似イン
パルス応答レジスタ9の擬似インパルス応答との積和が
積和回路10でとられ、積和回路10の出力が擬似エコ
ーとして出力される。受話信号出力端子2および送話信
号入力端子4は長距離電話回線の場合、2線4線変換器
に、拡声電話システムの場合、スピーカとマイクロホン
へと接続されている。FIG. 3 is a diagram showing the basic structure of an acoustic echo canceller. The reception signal input terminal 1 is connected to the reception signal output 2, and the reception signal of the reception signal input terminal 1 is branched and supplied to the variable coefficient filter 3 to generate a pseudo echo. The transmission signal from the transmission signal input terminal 4 and the pseudo echo that is the output of the variable coefficient filter 3 are input to the subtraction circuit 5, the echo component in the transmission signal is removed, and the output of the subtraction circuit 5 is the transmission signal. The signal is output to the signal output terminal 6. Transmission signal output terminal 6
Output and the signal of the reception signal input terminal 1 are the correction amount calculation circuit 7
And the filter coefficient of the variable coefficient filter 3 is corrected by the output of the correction amount calculation circuit 7. The reception signal is input to the reception signal input register 8 in the variable coefficient filter 3, and the sum of products of the reception signal of the reception signal input register 8 and the pseudo impulse response of the pseudo impulse response register 9 is obtained by the sum of products circuit 10. The output of the product-sum circuit 10 is output as a pseudo echo. The reception signal output terminal 2 and the transmission signal input terminal 4 are connected to a two-wire to four-wire converter in the case of a long-distance telephone line, and to a speaker and a microphone in the case of a public telephone system.
【0004】エコー経路の信号伝搬特性を線形で、且つ
FIR形ディジタルフィルタで表されると仮定し、その
インパルス応答h(t)と入力受話信号x(t)とを用
いれば、サンプル時間間隔をTとし、時刻kTにおける
エコーyk は、 yk = h’xk (1) で表される。但し、 h=[h1 ,h2 ,・・・,hn ]’ (2) x=[xk-1 ,・・・,xk-n ]’ ’:べクトルの転置 である。Assuming that the signal propagation characteristic of the echo path is linear and represented by a FIR type digital filter, if the impulse response h (t) and the input received signal x (t) are used, the sampling time interval is The echo y k at time kT is represented by y k = h'x k (1). However, h = [h 1 , h 2 , ..., H n ] '(2) x = [x k-1 , ..., x kn ]'': transposition of the vector.
【0005】一方、 時刻kTにおけるhの推定値をh
sk とすれば、yk の推定値yskは、 ysk = hsk ’xk (3) で与えられる。 音響反響除去装置では、受話信号入力
端子1に音声信号があり、送話信号入力端子4に音声信
号がなくエコーのみが存在している時、適応動作状態と
して反響除去動作を行う。この適応動作アルゴリズムに
は、一般に学習同定法が採用される。学習同定法による
hsk の逐次修正は hsk+1 =hsk +α(xk ek )/xk ’xk (4) によって行われる。但し、 ek =yk −ysk 、 0<α≦1 (5) であり、ek を残留エコーと呼ぶ。この様な演算動作が
修正量演算回路7において処理実行されている。擬似イ
ンパルス応答レジスタ9の内容には上記可変係数hsk
が格納されている。αは推定の敏感さを決める為の修正
ループゲインで1.0に近いほど大きな修正量を与える
事が出来るが、近端雑音や回線状態によって変えてやる
必要がある。On the other hand, the estimated value of h at time kT is h
If s k , the estimated value ys k of y k is given by ys k = hs k ′ x k (3). In the acoustic echo canceller, when there is a voice signal at the reception signal input terminal 1 and there is no voice signal at the transmission signal input terminal 4 but only an echo exists, the echo elimination operation is performed as an adaptive operation state. A learning identification method is generally adopted for this adaptive operation algorithm. Successive correction of hsk by the learning identification method is performed by hs k + 1 = hs k + α (x k e k) / x k 'x k (4). However, a e k = y k -ys k, 0 <α ≦ 1 (5), the e k is referred to as residual echo. Such an arithmetic operation is processed in the correction amount arithmetic circuit 7. The contents of the pseudo impulse response register 9 include the variable coefficient hs k
Is stored. α is a correction loop gain for determining the sensitivity of estimation, and a larger correction amount can be given as it approaches 1.0, but it must be changed depending on near-end noise and line conditions.
【0006】[0006]
【発明が解決しようとする課題】上記送話信号入力端子
に送話信号に対して比較的大きなレベルの近端雑音が入
力された場合、上記誤差信号の一定区間平均電力のレベ
ルのみに対応させた修正ループゲイン設定では近端雑音
などの外乱が存在する場合にその雑音レベルが、修正ル
ープゲインの切り替え閾値を桁上げしてしまい、必要以
上の大きな修正ループゲインを設定してしまう結果とな
り、耳障りな残響音声が発生してしまう。又、現状以上
に同定し得ないにも関わらず、大きな修正量が加わる為
に推定系インパルス応答係数系列を発振させ精度の高い
音響反響除去が行えないと言う様な問題点があった。When a near-end noise of a relatively large level is inputted to the transmission signal input terminal, it is made to correspond only to the level of the average power of the error signal in a certain section. In the modified loop gain setting, when there is a disturbance such as near-end noise, the noise level carries the modified loop gain switching threshold to a carry value, resulting in setting a larger modified loop gain than necessary. Annoying reverberant sound is generated. In addition, there is a problem in that the estimated system impulse response coefficient sequence is oscillated and acoustic echo removal cannot be performed with high accuracy because a large amount of correction is added, although it cannot be identified more than at present.
【0007】本発明は上述の点に鑑みてなされたもの
で、上記問題点を除去し、説明変数以外の近端雑音など
の外乱による影響を検知し、制御対象となる音場のイン
パルス応答の推定の安定度と真値への収束速度を向上さ
せた音響反響除去装置を提供する事を目的とする。The present invention has been made in view of the above points, and eliminates the above problems, detects the influence of a disturbance such as near-end noise other than the explanatory variables, and detects the impulse response of the sound field to be controlled. It is an object of the present invention to provide an acoustic echo canceller with improved stability of estimation and speed of convergence to a true value.
【0008】[0008]
【課題を解決するための手段】本発明はこれらの課題を
解決するためのものであり、受話信号入力端子と、受話
信号出力端子と、送話信号入力端子と、送話信号出力端
子と、前記受話信号入力端子から入力された受話信号を
入力とする可変係数ディジタルフィルタと、前記可変係
数ディジタルフィルタの係数系列を格納した擬似インパ
ルス応答レジスタと、前記擬似インパルス応答レジスタ
の内容と前記受話信号入力端子からの受話信号を格納し
た前記受話信号入力レジスタの内容との畳み込み積分演
算を行う積和演算回路と、前記積和演算回路により生成
された擬似音響反響と前記送話信号入力端子より入力さ
れる音響反響とを差分して残差エコーを求める減算回路
と、前記可変係数ディジタルフィルタに該反響の推定近
似値が供給される前記擬似インパルス応答レジスタの係
数系列を自動的に更新する制御を行う係数修正量演算回
路とを有する音響反響除去装置において、前記受話信号
入力端子の受話信号に関係なく平均値“零”となる擬似
的な白色雑音を生成し発生する擬似白色雑音発生回路
と、前記擬似白色雑音の一定区間平均電力を算出する第
1の積算回路と、前記受話信号入力端子の受話信号と前
記疑似白色信号とが加算され前記受話信号出力端子の受
話信号を生成する加算回路と、前記送話信号入力端子の
送話信号の一定区間平均電力を算出する第2の積算回路
と、前記送話信号の有声音区間と無声音区間を検出する
有声無声判定回路と、前記送話信号の無声音区間の一定
区間平均電力と前記擬似白色雑音の平均電力を比較する
レベル比較回路と、前記誤差信号の一定区間平均電力を
算出する第3の積算回路と、前記レベル比較回路で求め
られるレベル差と前記誤差信号の一定区間平均電力との
比よりループゲインを決定するループゲイン設定回路と
から構成され、前記擬似白色雑音を前記受話信号入力端
子と前記送話信号出力端子の間の基準レベルとして、前
記擬似インパルス応答の修正量を決定する修正ループゲ
インを基準レベルとして任意に変更することによって音
響反響信号の収束特性を向上させる音響反響除去装置を
提供する。さらに、前記ループゲイン設定回路の内部評
価値によって前記送話信号入力端子から入力される近端
雑音の電力レベルを検出し、その検出された電力レベル
が双方向通信検出用の検出閾値として使用される請求項
1記載の音響反響除去装置を提供する。SUMMARY OF THE INVENTION The present invention is to solve these problems and comprises a reception signal input terminal, a reception signal output terminal, a transmission signal input terminal, and a transmission signal output terminal. A variable coefficient digital filter that receives the received signal input from the received signal input terminal, a pseudo impulse response register that stores the coefficient series of the variable coefficient digital filter, the contents of the pseudo impulse response register, and the received signal input. A product-sum operation circuit that performs a convolution integral operation with the content of the reception-signal input register that stores the reception signal from the terminal, a pseudo-acoustic echo generated by the product-sum operation circuit, and the input from the transmission-signal input terminal. The subtraction circuit for obtaining the residual echo by subtracting the acoustic echo from the acoustic echo and the variable coefficient digital filter are supplied with the estimated approximate value of the echo. In an acoustic echo canceller having a coefficient correction amount calculation circuit for automatically updating a coefficient sequence of a pseudo impulse response register, a pseudo echo having an average value "zero" regardless of a reception signal at the reception signal input terminal. White noise generating circuit that generates and generates a typical white noise, a first integrating circuit that calculates the constant section average power of the pseudo white noise, a reception signal at the reception signal input terminal, and the pseudo white signal. An adder circuit that is added to generate a reception signal at the reception signal output terminal, a second integrating circuit that calculates a constant section average power of the transmission signal at the transmission signal input terminal, and a voiced section of the transmission signal. And a voiced unvoiced determination circuit for detecting an unvoiced sound section, a level comparison circuit for comparing a constant section average power of the unvoiced sound section of the transmission signal and an average power of the pseudo white noise, and the error signal And a loop gain setting circuit that determines a loop gain based on a ratio of a level difference obtained by the level comparison circuit and a constant section average power of the error signal. Pseudo white noise is used as a reference level between the reception signal input terminal and the transmission signal output terminal, and a correction loop gain that determines the correction amount of the pseudo impulse response is arbitrarily changed as a reference level to obtain the acoustic echo signal. Provided is an acoustic echo canceller that improves convergence characteristics. Furthermore, the power level of the near-end noise input from the transmission signal input terminal is detected by the internal evaluation value of the loop gain setting circuit, and the detected power level is used as a detection threshold value for two-way communication detection. An acoustic echo canceller according to claim 1 is provided.
【0009】[0009]
【作用】本発明では、上記手段により説明変数として無
意味な近端雑音を未知のまま従属変数として取り扱うの
ではなく、近端雑音の状況を逐次的に観測しながら修正
ループゲインを適応的に変更して残響制御を行うので、
常時安定で高速な残響除去が可能となる。また本方式で
は通信回線の定常レベル基準を自分で設定でき、この基
準を対象として残響制御を行える。そして、この基準を
用いる事で高精度な双方向通信検出も容易に行う事が出
来、高品質な音声通信会議空間を提供する事が出来る。According to the present invention, the above-mentioned means does not treat meaningless near-end noise as an explanatory variable as a dependent variable while remaining unknown, but adaptively adjusts the modified loop gain while sequentially observing the situation of near-end noise. Since it is changed and reverberation control is performed,
It is always stable and enables high-speed dereverberation. Also, in this method, the steady level standard of the communication line can be set by oneself, and reverberation control can be performed with this standard as the target. By using this standard, it is possible to easily perform high-precision bidirectional communication detection and provide a high-quality voice communication conference space.
【0010】[0010]
【実施例】以下本発明の一実施例を図面に基づいて詳細
に説明する。図1は本発明の音響反響除去装置の構成を
示すブロック図である。図に示すように、本発明は従来
の受話信号入力端子1、受話信号出力端子2、可変係数
ディジタルフィルタ3、送話信号入力端子4、減算回路
5、送話信号出力端子6、係数修正量演算回路7、受話
信号入力レジスタ8、擬似インパルス応答レジスタ9、
積和演算回路10、からなる適応アルゴリズムとして学
習同定法を採用した音響反響除去装置と同一構成の装置
に、擬似白色雑音発生回路11、加算回路12、積算回
路13、積算回路14、有声無声判定回路15、レベル
比較回路16、積算回路17、ループゲイン設定回路1
8、排他的論理加算回路19、遅延回路20を追加した
構成になっている。DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described in detail below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of the acoustic echo canceller of the present invention. As shown in the figure, the present invention uses the conventional reception signal input terminal 1, reception signal output terminal 2, variable coefficient digital filter 3, transmission signal input terminal 4, subtraction circuit 5, transmission signal output terminal 6, coefficient correction amount. Arithmetic circuit 7, reception signal input register 8, pseudo impulse response register 9,
A pseudo white noise generating circuit 11, an adding circuit 12, an integrating circuit 13, an integrating circuit 14, a voiced / unvoiced determination is provided in a device having the same configuration as the acoustic echo canceling device adopting a learning identification method as an adaptive algorithm including a product-sum operation circuit 10. Circuit 15, level comparison circuit 16, integration circuit 17, loop gain setting circuit 1
8, an exclusive logical addition circuit 19, and a delay circuit 20 are added.
【0011】上記受話信号入力端子1と、上記受話信号
出力端子2と、上記送話信号入力端子4と、上記送話信
号出力端子6と、上記受話信号入力端子1から入力され
た受話信号を入力とする上記可変係数ディジタルフィル
タ3と、上記可変係数ディジタルフィルタ3の係数系列
を格納した上記擬似インパルス応答レジスタ9と、上記
擬似インパルス応答レジスタ9の内容と上記受話信号入
力端子1からの受話信号を格納した上記受話信号入力レ
ジスタ8の内容との畳み込み積分演算を行う上記積和演
算回路10と、上記積和演算回路10により生成された
擬似音響反響と上記送話信号入力端子4より入力される
音響反響との差分値をとる上記減算回路5と、上記可変
係数ディジタルフィルタ3が該反響の近似値を供給する
様に上記擬似インパルス応答レジスタの係数系列を自動
的に更新される様に制御を行う係数修正量演算回路7と
で構成される音響反響除去装置において、上記擬似白色
発生回路11で生起された擬似白色雑音は、上記加算回
路12により受話信号に印加され上記受話信号出力端子
から通信回線外へ出力される。この受話信号が反響路を
経由し、上記送話信号入力端子4より送話信号として通
信回線へ入力される。この送話信号の一定区間平均電力
を上記積算回路14で算出し、上記有声無声判定回路1
5によって無声音部分の検出を行う。次にこの無声音部
分の電力レベルと上記擬似白色雑音発生回路11で生起
された擬似白色雑音の一定区間平均電力を求める上記積
算回路13の出力電力レベルの比較を上記レベル比較回
路16において行う。The reception signal input terminal 1, the reception signal output terminal 2, the transmission signal input terminal 4, the transmission signal output terminal 6, and the reception signal input from the reception signal input terminal 1 The variable coefficient digital filter 3 to be input, the pseudo impulse response register 9 storing the coefficient series of the variable coefficient digital filter 3, the contents of the pseudo impulse response register 9 and the reception signal from the reception signal input terminal 1. The sum of products calculation circuit 10 for performing a convolution integral calculation with the contents of the reception signal input register 8 storing the above, the pseudo acoustic echo generated by the sum of products calculation circuit 10 and the transmission signal input terminal 4 are inputted. The subtraction circuit 5 which takes the difference value from the acoustic echo and the variable coefficient digital filter 3 supply the pseudo-in so that the approximate value of the echo is supplied. In the acoustic echo canceller configured with the coefficient correction amount calculation circuit 7 for controlling so that the coefficient series of the loose response register is automatically updated, the pseudo white noise generated by the pseudo white generation circuit 11 is The adding circuit 12 applies the received signal to the received signal and outputs the signal from the received signal output terminal to the outside of the communication line. This reception signal passes through the echo path and is input to the communication line as the transmission signal from the transmission signal input terminal 4. The constant section average power of the transmission signal is calculated by the integration circuit 14, and the voiced / unvoiced determination circuit 1
5, the unvoiced sound portion is detected. Next, the level comparing circuit 16 compares the power level of the unvoiced sound portion with the output power level of the integrating circuit 13 for obtaining the constant section average power of the pseudo white noise generated by the pseudo white noise generating circuit 11.
【0012】一方、上記送話信号入力端子4の送話信号
と上記受話信号入力端子1より入力した受話信号から発
生させた擬似残響の減算が上記減算回路5で行われ上記
誤差信号が出力され積算回路14において上記誤差信号
の一定期間平均電力が算出される。この平均電力と上記
レベル比較回路16で求めた上記擬似白色雑音の一定区
間平均電力と上記無声音部分の電力レベル差とを二つの
入力として、上記ループゲイン設定回路18において回
線状態に対応した修正ループゲインを設定して適応処理
を動作させる。つまり、上記レベル比較回路16の出力
は未知系の減衰特性と近端雑音の影響を評価する基準と
なり、又、上記ループゲイン設定回路18での比率は近
端雑音のレベルを評価する基準となっている。この基準
値は制御動作中において一貫しているので、更に通信中
上記受話信号出力端子2の受話信号と上記送話信号入力
端子4に発言者の音声信号が同時に存在するような双方
向通信の際に、この上記ループゲイン設定回路18の内
部基準である電力比率を双方向通信の検出基準として使
用する。On the other hand, the subtraction circuit 5 subtracts the pseudo reverberation generated from the transmission signal from the transmission signal input terminal 4 and the reception signal input from the reception signal input terminal 1, and outputs the error signal. The integrating circuit 14 calculates the average power of the error signal for a certain period. This average power, the constant section average power of the pseudo white noise obtained by the level comparison circuit 16 and the power level difference of the unvoiced part are used as two inputs, and the loop gain setting circuit 18 makes a correction loop corresponding to the line state. Set the gain to operate the adaptive processing. That is, the output of the level comparison circuit 16 serves as a standard for evaluating the influence of the attenuation characteristic of the unknown system and the near-end noise, and the ratio in the loop gain setting circuit 18 serves as a standard for evaluating the level of the near-end noise. ing. Since this reference value is consistent during the control operation, the two-way communication in which the voice signal of the speaker is simultaneously present in the voice signal of the voice signal output terminal 2 and the voice signal input terminal 4 during communication is performed. At this time, the power ratio, which is an internal reference of the loop gain setting circuit 18, is used as a detection reference for bidirectional communication.
【0013】図2に示したのは、上記擬似白色雑音発生
回路11の一構成例である。これは10段の最大周期系
列符号で、上記排他的論理加算回路19と上記遅延回路
20の2要素から成る簡単な構成で、周期が1023サ
ンプルの擬似白色雑音を生起し発生させる事の出来る回
路である。FIG. 2 shows an example of the configuration of the pseudo white noise generating circuit 11 described above. This is a 10-stage maximum period sequence code, and a circuit which can generate and generate pseudo white noise having a period of 1023 samples with a simple configuration including two elements of the exclusive logical adder circuit 19 and the delay circuit 20. Is.
【0014】[0014]
【発明の効果】以上、詳細に説明したように本発明によ
れば、下記のような優れた効果が期待される。As described in detail above, according to the present invention, the following excellent effects are expected.
【0015】(1)近端雑音が相対的に観測できるの
で、説明変数と従属変数の関係が明確化されて、近端雑
音の影響を考慮した安定な残響除去制御が行える。(1) Since the near-end noise can be relatively observed, the relationship between the explanatory variable and the dependent variable is clarified, and stable dereverberation control considering the influence of the near-end noise can be performed.
【0016】(2)制御基準を自分自身で設定でき、到
達残響除去量が既知となるので無理に大きな修正量を与
えなくてもよくなり、発振などの発生がなくなる。(2) Since the control reference can be set by oneself and the arrival dereverberation amount is known, it is not necessary to forcibly give a large correction amount, and oscillation does not occur.
【0017】(3) 近端雑音が大きな場合でも小さな
場合にでも制御動作中は一貫した基準が得られているの
で、効率の良い修正ループゲイン設定が行え、真値への
収束速度を高速化出来る。(3) Since a consistent reference is obtained during the control operation regardless of whether the near-end noise is large or small, efficient correction loop gain setting can be performed, and the speed of convergence to the true value can be increased. I can.
【0018】(4)一定レベルの擬似白色雑音が背景雑
音の様に重畳されているので、残留誤差信号の様な耳障
りな不規則音はかき消されて、良好な通信会議システム
としての音場を提供できる。(4) Since a constant level of pseudo white noise is superimposed like background noise, annoying irregular sounds such as residual error signals are erased, and a sound field as a good communication conference system is provided. Can be provided.
【0019】(5)近端雑音の電力レベルを逐次的に既
知情報として得られるので、非線形変換処理などの閾値
を常時明確化でき、確実な残響制御を行う事が出来る。(5) Since the power level of the near-end noise is sequentially obtained as known information, the threshold value for nonlinear conversion processing can be always clarified, and reliable reverberation control can be performed.
【0020】(6)回線状態に対して常時安定した一定
基準を自立して持っている為に高速で且つ精度の高い双
方向通信検出が出来、高品質な作業空間を使用している
人に対して提供する事が出来る。(6) Since it has a stable and constant standard independently for the line status, it is possible to detect bi-directional communication at high speed and with high accuracy, and to those who use a high-quality work space. It can be provided to.
【図1】本説明の音響反響除去装置の一構成例を示すブ
ロック図である。FIG. 1 is a block diagram showing a configuration example of an acoustic echo canceller of the present description.
【図2】本発明で用いた擬似白色雑音発生回路の一構成
例を示すブロック図である。FIG. 2 is a block diagram showing a configuration example of a pseudo white noise generation circuit used in the present invention.
【図3】従来の一般的な学習同定法を用いた音響反響除
去装置の基本構成を示したブロック図である。FIG. 3 is a block diagram showing a basic configuration of an acoustic echo canceller using a conventional general learning identification method.
1 受話信号入力端子 2 受話信号出力端子 3 可変係数ディジタルフィルタ 4 送話信号入力端子 5 減算回路 6 送話信号出力端子 7 係数修正量演算回路 8 受話信号入力レジスタ 9 擬似インパルス応答レジスタ 10 積和演算回路 11 擬似白色発生回路 12 加算回路 13 積算回路 14 積算回路 15 有声無声判定回路 16 レベル比較回路 17 積算回路 18 ループゲイン設定回路 19 排他的論理加算回路 20 遅延回路 1 reception signal input terminal 2 reception signal output terminal 3 variable coefficient digital filter 4 transmission signal input terminal 5 subtraction circuit 6 transmission signal output terminal 7 coefficient correction amount calculation circuit 8 reception signal input register 9 pseudo impulse response register 10 sum of products calculation Circuit 11 Pseudo white generation circuit 12 Addition circuit 13 Integration circuit 14 Integration circuit 15 Voiced / unvoiced determination circuit 16 Level comparison circuit 17 Integration circuit 18 Loop gain setting circuit 19 Exclusive logic addition circuit 20 Delay circuit
Claims (2)
と、送話信号入力端子と、送話信号出力端子と、前記受
話信号入力端子から入力された受話信号を入力とする可
変係数ディジタルフィルタと、前記可変係数ディジタル
フィルタの係数系列を格納した擬似インパルス応答レジ
スタと、前記擬似インパルス応答レジスタの内容と前記
受話信号入力端子からの受話信号を格納した前記受話信
号入力レジスタの内容との畳み込み積分演算を行う積和
演算回路と、前記積和演算回路により生成された擬似音
響反響と前記送話信号入力端子より入力される音響反響
とを差分して残差エコーを求める減算回路と、前記可変
係数ディジタルフィルタに該反響の推定近似値が供給さ
れる前記擬似インパルス応答レジスタの係数系列を自動
的に更新する制御を行う係数修正量演算回路とを有する
音響反響除去装置において、前記受話信号入力端子の受
話信号に関係なく平均値“零”となる擬似的な白色雑音
を生成し発生する擬似白色雑音発生回路と、前記擬似白
色雑音の一定区間平均電力を算出する第1の積算回路
と、前記受話信号入力端子の受話信号と前記疑似白色信
号とが加算され前記受話信号出力端子の受話信号を生成
する加算回路と、前記送話信号入力端子の送話信号の一
定区間平均電力を算出する第2の積算回路と、前記送話
信号の有声音区間と無声音区間を検出する有声無声判定
回路と、前記送話信号の無声音区間の一定区間平均電力
と前記擬似白色雑音の平均電力を比較するレベル比較回
路と、前記誤差信号の一定区間平均電力を算出する第3
の積算回路と、前記レベル比較回路で求められるレベル
差と前記誤差信号の一定区間平均電力との比よりループ
ゲインを決定するループゲイン設定回路とから構成さ
れ、前記擬似白色雑音を前記受話信号入力端子と前記送
話信号出力端子の間の基準レベルとして、前記擬似イン
パルス応答の修正量を決定する修正ループゲインを基準
レベルとして任意に変更することによって音響反響信号
の収束特性を向上させたことを特徴とする音響反響除去
装置。1. A variable coefficient digital filter which receives a reception signal input terminal, a reception signal output terminal, a transmission signal input terminal, a transmission signal output terminal, and a reception signal input from the reception signal input terminal. And a pseudo impulse response register that stores the coefficient sequence of the variable coefficient digital filter, and a convolution integral of the contents of the pseudo impulse response register and the contents of the reception signal input register that stores the reception signal from the reception signal input terminal. A sum-of-products calculation circuit for performing a calculation, a subtraction circuit for obtaining a residual echo by making a difference between the pseudo-acoustic echo generated by the sum-of-products calculation circuit and the acoustic echo input from the transmission signal input terminal, and the variable A control for automatically updating the coefficient sequence of the pseudo impulse response register to which the estimated approximation value of the echo is supplied to the coefficient digital filter is provided. In the acoustic echo canceller having a coefficient correction amount calculation circuit, a pseudo white noise generating circuit for generating and generating pseudo white noise having an average value "zero" regardless of the reception signal at the reception signal input terminal, A first integrating circuit for calculating the constant section average power of the pseudo white noise; and an adder circuit for adding the reception signal at the reception signal input terminal and the pseudo white signal to generate a reception signal at the reception signal output terminal. A second integrating circuit for calculating a constant section average power of the transmission signal at the transmission signal input terminal, a voiced unvoiced determination circuit for detecting a voiced sound section and an unvoiced sound section of the transmission signal, and the transmission signal Level comparing circuit for comparing the average power of the unvoiced sound in the constant section and the average power of the pseudo white noise, and the third level for calculating the average power of the constant section of the error signal.
And a loop gain setting circuit that determines a loop gain from the ratio of the level difference obtained by the level comparison circuit and the constant section average power of the error signal, and inputs the pseudo white noise to the reception signal input. As a reference level between the terminal and the transmission signal output terminal, it is possible to improve the convergence characteristic of the acoustic echo signal by arbitrarily changing the correction loop gain that determines the correction amount of the pseudo impulse response as the reference level. Characteristic acoustic echo canceller.
よって前記送話信号入力端子から入力される近端雑音の
電力レベルを検出し、その検出された電力レベルが双方
向通信検出用の検出閾値として使用されることを特徴と
する請求項1記載の音響反響除去装置。2. A power level of near-end noise input from the transmission signal input terminal is detected by an internal evaluation value of the loop gain setting circuit, and the detected power level is a detection threshold value for detecting bidirectional communication. The acoustic echo canceller according to claim 1, wherein the acoustic echo canceller is used as.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP29937493A JPH07154306A (en) | 1993-11-30 | 1993-11-30 | Acoustic echo canceller |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP29937493A JPH07154306A (en) | 1993-11-30 | 1993-11-30 | Acoustic echo canceller |
Publications (1)
Publication Number | Publication Date |
---|---|
JPH07154306A true JPH07154306A (en) | 1995-06-16 |
Family
ID=17871743
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP29937493A Pending JPH07154306A (en) | 1993-11-30 | 1993-11-30 | Acoustic echo canceller |
Country Status (1)
Country | Link |
---|---|
JP (1) | JPH07154306A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2437260A2 (en) | 2010-09-30 | 2012-04-04 | Roland Corporation | Sound signal processing device |
CN114171049A (en) * | 2021-12-24 | 2022-03-11 | 上海领世通信技术发展有限公司 | Echo cancellation method and device, electronic device and storage medium |
-
1993
- 1993-11-30 JP JP29937493A patent/JPH07154306A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2437260A2 (en) | 2010-09-30 | 2012-04-04 | Roland Corporation | Sound signal processing device |
US8908881B2 (en) | 2010-09-30 | 2014-12-09 | Roland Corporation | Sound signal processing device |
CN114171049A (en) * | 2021-12-24 | 2022-03-11 | 上海领世通信技术发展有限公司 | Echo cancellation method and device, electronic device and storage medium |
CN114171049B (en) * | 2021-12-24 | 2024-09-17 | 上海领世通信技术发展有限公司 | Echo cancellation method and device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JPH03218150A (en) | Double talk detection circuit | |
JPH09139696A (en) | Method and device for both adaptive identification and related adaptive echo canceler thereto | |
JPH10242891A (en) | Echo canceller | |
CA2200672C (en) | Two-wire conference apparatus with echo canceler without training | |
JP3211884B2 (en) | Acoustic echo canceller | |
JPH07154306A (en) | Acoustic echo canceller | |
JP3121983B2 (en) | Acoustic echo canceller | |
JP3152815B2 (en) | Acoustic echo canceller | |
JP3217614B2 (en) | Acoustic echo canceller | |
JPH0722983A (en) | Acoustic dereverberation device | |
JP3217618B2 (en) | Acoustic echo canceller | |
JPH0758674A (en) | Howling detector | |
JP3121969B2 (en) | Acoustic echo canceller | |
JPH0746160A (en) | Acoustic echo canceller | |
JP3187715B2 (en) | Echo canceller | |
JP3152825B2 (en) | Acoustic echo canceller | |
JP2551869B2 (en) | Echo canceller | |
JPH0946276A (en) | Public information communication system | |
JPS6053490B2 (en) | Adaptive echo canceller | |
JPH0746161A (en) | Acoustic echo canceller | |
JP3152822B2 (en) | Acoustic echo canceller | |
JP3121988B2 (en) | Acoustic echo canceller | |
JPH09162787A (en) | Echo canceller | |
JPH09307483A (en) | Echo canceller | |
JPH10145487A (en) | High quality voice information communication system |