JPH02238730A - Two-way call detection system and echo canceller and loudspeaking telephone set using the system - Google Patents

Abstract

PURPOSE:To evade mis-detection and malfunction of an echo canceller by detecting the presence of 2-way call not by level comparison of signals but by difference from input and output signal patterns of an echo path. CONSTITUTION:A signal pattern extraction circuit 7 extracts a characteristic time series signal pattern of input and output signals of an echo path 1, a pattern matching circuit 8 calculates the distance of patterns, that is, the difference and outputs the result to a judging circuit 9. The judging circuit 9 compares a threshold level with the output signal and discriminates it to be the state of mixture of a signal other than an echo signal to an output of the echo path 1, that is, 2-way talking when the difference is over the threshold level, and outputs a 2-way talking detection signal to a terminal 3-4 of the echo canceller 3. Thus, the production of mis-detection depending on the echo level is not caused and malfunction of the echo canceller is evaded.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an echo canceler.
), and particularly relates to a two-way conversation detection method for suitably controlling the echo path characteristic estimation function of a device.

[Prior Art] An echo cancellation device adaptively estimates the characteristics of the echo path using the input signal to the echo path and the echo signal (echo), and generates a pseudo echo signal (echo signal) based on the obtained echo characteristics. The echo signal is eliminated by creating a replica of the echo signal and subtracting the replica of the echo signal from the actual echo signal. It is used to eliminate echoes caused by unbalanced 2-wire to 4-wire conversion circuits in communication lines, or echoes caused by acoustic coupling between the speaker and microphone of public address telephones. On the other hand, the conventional so-called echo suppression device ti! (EchoS
Uppressor) detects the speaker from the transmitting and receiving signal levels, assumes that the one with the higher level is the current speaker, and disconnects the signal line with the lower level with a switch or inserts a large loss into the signal line. This prevents repercussions.

Although echo suppression devices can be constructed with simple circuits, they have the problem of causing various communication disturbances, such as cutting off of the beginning and end of words due to signal line switching.

In this respect, it is well known that echo canceling devices do not have the above-mentioned problems and are superior to echo blocking devices.

The biggest problem with echo cancellers is the two-way communication (do
When the adaptive estimation of the echo path is disturbed and the echo path characteristics are different from the actual echo path characteristics, the echo cancellation ability is reduced and howling occurs, or the device itself is damaged. becomes unstable and malfunctions. This is because, during two-way communication, a high-level signal other than the echo signal is superimposed on the echo signal, which is the basis for calculating the characteristics of the echo path.

Conventional equipment monitors the signal levels at the input and output sides of the echo path, detects bidirectional communication based on these levels, and stops adaptive estimation of the echo path when bidirectional communication occurs. There is. However, in an actual circuit, the echo path (in this case, 2I!-
Attenuation amount (εcho Retur
nLoss) may be small, and even during non-two-way communication, the signal level at the anti-1jl output terminal becomes quite large due to the echo signal itself. Furthermore, in a loudspeaker telephone, the amount of attenuation in the echo path (in this case, the acoustic path from the speaker to the microphone) varies greatly depending on the positional relationship between the input and output ends of the echo path, that is, the positional relationship between the speaker and the microphone. Depending on this positional relationship or gain distribution of amplifiers, etc., the amount of attenuation may be zero or negative, that is, the echo signal may not be attenuated but on the contrary may be amplified.

Therefore, if a simple level comparison is made, the echo signal itself is large, so even if it is not a two-way conversation, it will be determined that a two-way conversation is occurring, and estimation of the echo path may be stopped. be. However, it is precisely in cases like this that echoes must be eliminated. In addition, even during two-way communication, the echo signal level is high, making it impossible to detect the original transmission signal (near-end speaker signal) and estimating the echo path, which may differ from the actual echo path characteristics. It has i-like echo path characteristics, and its echo cancellation ability decreases. Therefore, such a method of detecting two-way communication simply by comparing levels and controlling the echo path estimation operation of the echo canceller is not very effective.

Therefore, as described in Japanese Patent Application Laid-Open No. 62-29910, the level of the residual echo (remaining signal that has not been completely eliminated from the echo signal) is monitored at the same time, and the comparison standard value ( This publication discloses a double talk detection circuit for this purpose. In other words, this double talk detection circuit includes a power comparison circuit that detects a change in the amount of echo cancellation, which is the difference between the received input power and the transmitted output power of the echo canceller, and a power comparison circuit that detects a change in the amount of echo cancellation, which is the difference between the received input power and the transmitted output power of the echo canceller. and a threshold circuit for changing the monitoring reference threshold.

[Problem to be solved by the invention] However, in the above-mentioned prior art,
Even if residual echo is used as described in Publication No. 0, two-way communication is basically detected by comparing signal levels, so as explained earlier, if the level difference is too large, detection errors may occur. There was a problem. SUMMARY OF THE INVENTION An object of the present invention is to provide a two-way conversation detection method that does not rely on signal level comparison, and to provide an echo canceling device and a loudspeaker telephone device that avoid malfunctions and performance deterioration.

[Means for Solving the Problems] The above object is to perform two-way communication from a signal pattern extracting means, a pattern matching means, and a determining means provided at the input end and the output end of the echo path, respectively, in an echo canceling device. This is accomplished by configuring a detection circuit and controlling the adaptive estimation of the echo path of the echo canceller with the output of the two-way conversation detection circuit. The signal pattern extraction means extracts a time-series pattern of t/{ parameters indicating the characteristics of the signal. The time-series pattern of this parameter is one that remains unchanged even when the signal passes through the echo path. [Operation] Next, the principle of the present invention will be described. Now, consider only the echo signal as the echo path output.

In this case, the input signal to the echo path 1 is affected by the transfer characteristics of the echo path 1 and is output from the echo path 1. As is well known, the output signal waveform is y (t), the spectrum is Y (ω),
The input signal waveform is x (t), the spectrum is X (ω), the impulse response of the echo path is h (t), and the transfer function is H (ω).
), the output signal waveform y(t) and the spectrum Y(
ω) is expressed by the following formula. y(t)=h(t)
* x (t)Y(ω)=H(ω)・X(u) In other words, the output signal is expressed as the convolution integral of the input signal and the impulse response, and the output spectrum is expressed as the product of the input spectrum and the transfer function. Therefore, even if you directly compare the input and output, waveforms, or spectra, they will not match. However, for example, if you pass a single sine wave through the echo path,
The amplitude and phase change, but the frequency value does not change.
In the case of speech, the formant structure that represents phoneme does not change. Specifically, the phoneme "a" does not change to the phoneme "a" even if it passes through the echo path. This is because the positional relationship of each formant frequency that characterizes the phoneme ``a'' and its temporal change pattern are preserved. It is also clear that although the energy value of the input signal itself changes as it passes through the echo path, the temporal change pattern of the energy does not change. The present invention detects two-way calls based on the above facts. The signal pattern extraction circuit outputs time-series signals representing the characteristics of the input and output signals of the echo path, respectively. This characteristic time-series signal is called a pattern. This pattern uses feature parameters that basically do not change depending on the transfer function of the echo path. The pattern matching circuit calculates the distance between the feature time-series signals, that is, the amount of difference, and outputs it to the determination circuit. The judgment circuit compares and judges a predetermined value, that is, a threshold value, with the output signal. If the amount of difference is equal to or greater than the threshold, it is determined that the output of the echo path is mixed with a signal other than the echo signal, that is, the signal input to the echo path and the signal that is reflected and output, that is, a two-way conversation, A signal is output to the echo cancellation device to inform that the echo path estimation operation of the echo cancellation device should be stopped. As a result, the echo canceling device stops estimating the echo path, and therefore does not malfunction.

(The following is a blank space.) [Example] An example of the present invention will be explained below with reference to Fig. 1. In Figure 1, 1 is an echo path, 2 is a transmission path, 3 is an echo canceller, 4 is a two-way communication detection circuit, 5 is an input terminal of the echo path, 6 is an output terminal of the echo path, and 7 is a signal pattern extraction 8 is a pattern matching circuit, and 9 is a judgment circuit. Figure 2 is a diagram showing a concrete example of the echo path 1. Echo path 1 is an equivalent representation of the left side of the broken line in Figure 2. The echo path 1 is formed by a wrap-around signal (echo signal) due to unbalance of a 2-wire/4-wire conversion circuit in a communication line, or a echo wrap-around signal due to a wall or the like in an acoustic path between a speaker and a microphone.

The transmission path 2 is a signal transmission path.

The echo cancellation circuit 3 adaptively estimates the characteristics of the echo path using the input signal to the echo path and the echo signal, creates a pseudo echo signal based on the obtained echo path characteristics, and generates a pseudo echo signal from the actual echo signal. The echo signal is eliminated by subtracting the pseudo echo signal. FIG. 3 shows an example of a concrete echo canceling device.

The echo canceling device 3 includes a pseudo echo path circuit 3-5, a pseudo echo path correction circuit 3-6, and a subtraction circuit 3-7. The pseudo echo path circuit 3-5 creates a pseudo echo signal from the signal input to the input terminal 3-3 and output to the echo path 1.

The subtraction circuit 3-7 subtracts the pseudo echo signal created by the pseudo echo circuit 3-5 from the echo signal input to the input terminal 3-1. This result is output as an error signal at output terminal 3-
2 is output.

The pseudo echo path correction circuit 3-6 receives this error signal and the terminal 3-
The pseudo-echo characteristics in the pseudo-echo circuit 3-5 are corrected based on the correlation value with the output signal from the echo path 1 from the pseudo-echo circuit 3-5. This modification is performed in such a way that the error signal is minimized. Also, terminal 3
-4 stops the operation of the pseudo echo path correction circuit 3-6 when two-way communication is detected at the two-way communication input terminal.

Note that of the signals input to the terminal 3-1, the signals other than the echo signal are output as they are to the terminal 3-2.

The two-way conversation detection circuit 4 includes a signal pattern extraction circuit 7, a pattern matching circuit 8, and a determination circuit 9.

The signal pattern extraction circuit 7 extracts time-series signals representing the characteristics of the input and output signals of the echo path 1. This characteristic time-series signal is called a pattern.

The pattern matching circuit 8 calculates the distance between the characteristic time series signals, that is, the amount of difference, and outputs it to the judgment circuit 9.

The judgment circuit 9 compares and judges a predetermined value, that is, a threshold value, with the output signal. If the amount of difference is equal to or greater than the threshold value, the output of echo path 1 is mixed with a signal other than the echo signal (i.e., the signal that is reflected from the input signal to echo path 1), i.e., two-way communication is established. It is determined that this is the case, and a two-way conversation detection signal is output to the terminal 3-4 of the echo canceling device 3.

As a result, the operation of the pseudo echo path correction circuit 3-6 is stopped, and the estimated pseudo echo path is prevented from being disturbed by signals other than the echo signal. Note that the two-way communication state, referring to FIG. 2, refers to a state in which a transmission signal and a reception signal (near-end and far-end talker signals) exist simultaneously.

The present invention detects two-way communication based on the principle described above. The signal pattern extraction circuit 7 extracts a time-series pattern of characteristic parameters that is basically not changed by the transfer function of the echo path.

The extracted time-series patterns are used in a pattern matching circuit to calculate the differences between the patterns. When only the echo signal is input to the echo path output terminal 6, each signal pattern extraction circuit 7 connected to the echo path input terminal 5 and the output terminal 6 outputs almost the same time series pattern, so the pattern matching circuit 6 The output of is almost zero. When other signals coexist with the echo signal, the output of the signal pattern extraction circuit 7 connected to the output terminal 6 is different from the output of the signal pattern extraction circuit 7 connected to the input terminal 5, and the output of the pattern matching circuit 6 is different from the output of the signal pattern extraction circuit 7 connected to the input terminal 5. The output deviates significantly from zero.

The judgment circuit 9 compares and judges this deviation with a predetermined threshold value, outputs a two-way communication detection signal to the echo canceling device 3, and stops the correction operation of the pseudo echo path.

As described above, according to the present invention, two-way communication is detected not only by level comparison but also by the difference between patterns, so there is no possibility of false detection due to the level of the echo signal, and therefore the echo canceller does not malfunction. .

FIG. 4 shows another embodiment of the present invention. This embodiment is an example in which pattern matching is performed by correcting the amount of attenuation in the echo path. The other configurations are the same as those shown in FIG. 1 above. In FIG. 4, the same reference numerals as in FIG. 1 indicate the same parts.

10 is an amplifier circuit. When the time-series pattern extracted by the signal pattern extraction circuit 7 has the same shape but different absolute values, such as a time-varying pattern of signal energy, the so-called attenuation amount in the echo path 1 (Echo Re
It is necessary to perform pattern matching after correcting (turn loss). The role of the amplifier circuit 10 is to do this. In advance, in a state where only the echo signal is applied to the echo path output terminal 6, the gain of the amplifier circuit 10 is adjusted so that the output signal of the pattern matching circuit 8 becomes zero.

In this embodiment, the operations other than the amplifier circuit 10 are the same as those in FIG. 1, so the explanation will be omitted.

FIG. 5 shows another embodiment of the present invention. This embodiment is an example of correcting a time shift between time series patterns. The other configurations are the same as the embodiment shown in FIG. 1 above. In FIG. 5, the same symbols as in FIG. 1 indicate the same parts.

11 is a delay circuit. This delay circuit 11 corrects the time difference between time-series patterns.

Generally, there is a time delay in the echo path 1. This is a so-called flat delay, and specifically, between the speaker power and the acoustic path of the microphone, there is a time delay equal to the distance between the speaker and the microphone divided by the speed of sound. There is no problem if the speaker and microphone are close to each other, but if the distance is several tens of meters, there will be a delay of several hundred ls. Therefore, in this case, a difference of several hundred l15 occurs between the time series patterns extracted by the signal pattern extraction circuit 7, and the output of the pattern matching circuit 8 does not become zero.

The input signal to the echo path 1 is given a time delay by the delay circuit 11 and is input to the signal pattern extraction circuit 7. The input signal that has passed through the echo path 1 is input as a echo signal to the signal pattern extraction circuit 7 connected to the output terminal 6 of the echo path 1. If the delay times of the delay circuit 7 and the echo path 1 are the same, the output of the pattern matching circuit 8 can be made zero. In advance, in a state where only the echo signal is applied to the echo path output terminal 6, the delay amount of the delay circuit 11 is adjusted so that the output of the pattern matching circuit 8 becomes zero. The other operations are the same as those in the embodiment shown in FIG. 1, so the explanation will be omitted. FIG. 6 shows another embodiment of the present invention. In this embodiment, the amplifier circuit 10 of the embodiment shown in FIG. 4 is adapted to automatically adjust the gain. The other configurations are the same as the embodiment shown in FIG. 4 above. The same symbols as in Figure 1 indicate the same items. In FIG. 6, 12 is a variable gain amplifier circuit, 13 is a gain control circuit that controls the gain of the variable gain amplifier circuit 12, and 14 is a gain control circuit that controls the gain of the variable gain amplifier circuit 12.
1 is a power detection circuit that detects the power of the signal line, and 15 is an echo cancellation amount calculation circuit.

Input terminal 3-1 and output terminal 3-2 of echo canceller 3
A power detection circuit 14 connected to detects the power of each signal line. Based on the respective power values, the echo cancellation amount calculation circuit 15 calculates the echo signal cancellation amount of the echo cancellation device 3, that is, the echo signal inputted to the input terminal 3-1, to the output terminal 3.
Calculate the amount that indicates how much the output is attenuated by −2°.

The amount of echo cancellation is proportional to the ratio of the echo signal to other signals. This can be seen from the fact that signals other than the echo signal are not erased and appear as they are from the input terminal 3-1 to the output terminal 3-2. Therefore, when the amount of echo cancellation is sufficiently large, the signal applied to the echo path output terminal 6 can be considered to be almost an echo signal.

At this time, the output of the pattern matching circuit 8 should be zero,
The gain control circuit 13 controls the gain of the variable gain amplifier circuit 12 based on the echo cancellation amount output from the echo cancellation amount calculation circuit 15 and the output from the pattern matching circuit 8. When the amount of echo cancellation exceeds a certain value, the gain of the variable gain amplifier circuit 12 is automatically adjusted so that the output of the pattern matching circuit 8 becomes zero!
! ! I! ! be done.

This embodiment is effective in a public address system where the positional relationship between the microphone and the speaker changes, and the amount of attenuation in the echo path changes.

The operation of this embodiment is the same as that shown in FIG. 4 except for what is uniquely added to this embodiment, so a description thereof will be omitted. FIG. 7 shows another embodiment of the present invention. In addition to the embodiment shown in FIG. 6, this embodiment also automatically adjusts the delay amount of the delay circuit. The other configurations are the same as those shown in Figures 5 and 6 above. In Figure 7, the same symbols as in Figure 6 indicate the same items. In FIG. 7, 16 is a variable delay circuit, and 17 is a delay amount control circuit.

When the amount of echo cancellation exceeds a certain value, the delay amount control circuit 1
7 adjusts the delay amount of the variable delay circuit 16 so that the output of the pattern matching circuit 8 becomes zero.

It is desirable that the gain of the variable gain amplifier circuit 12 and the amount of delay of the variable delay circuit 16 be shifted in time. For example, first, the gain is automatically adjusted to minimize the output of the pattern matching circuit 8, and then the delay amount is automatically adjusted to make the output of the pattern matching circuit 8 almost zero.

This embodiment is effective in a public address system where the positional relationship between a microphone and a speaker changes, and the amount of attenuation and delay in the echo path changes accordingly.

The other operations of this embodiment are the same as those in FIGS. 5 and 6, so their explanation will be omitted. FIG. 8 shows another embodiment of the present invention. In this embodiment, the time series pattern extracted in the embodiment shown in FIG. 4 is a signal energy pattern, and the echo path 1, signal pattern extraction circuit 7, pattern matching circuit 8, and judgment circuit 9 are made concrete. . In FIG. 8, the same symbols as in FIG. 4 indicate the same parts.

In FIG. 8, 18 is a speaker, 19 is a microphone, and 20
is a rectifier circuit, 21 is a smoothing circuit, 22 is a differential amplifier circuit, 2
3 is a comparison circuit, 24 is a threshold circuit, and 35 is a volume adjustment circuit.

The echo path 1 shown in FIG. 4 is an acoustic path between the speaker 18 and the microphone 19 in this embodiment. The rectifier circuit 20 and the smoothing circuit 21 correspond to the signal pattern extraction circuit 7 in FIG. 4, and extract the average amplitude pattern of the signal emitted from the speaker 5 and the signal input to the microphone 19, respectively. It is clear that this corresponds to the energy (power) pattern of the signal. Now assume that the microphone and speaker are placed close to each other, the delay can be ignored, and the loss can be compensated for by the amplifier circuit 10. In this case, if the signal input to the microphone 19 is only a signal output from the speaker 18 and echoed (echo signal), the patterns output by each smoothing circuit 21 will be the same. When a sound other than the echo signal is input to the microphone 19, the microphone 19
The output pattern of the rectifier circuit 19 and smoothing circuit 20 connected to the speaker 18 is an energy pattern of a signal in which the echo signal and other signals are superimposed, and the output pattern of the rectifier circuit 20 and the smoothing circuit 21 connected to the speaker 18 The energy pattern will be different. The differential amplifier circuit 22 corresponds to the pattern matching circuit shown in FIG.
time series signal) is input. When the signal input to the microphone 19 is only a echo signal. Since the patterns are the same,
The output of the differential amplifier circuit 22 is approximately zero. During two-way communication in which the echo signal and other signals are input to the microphone 19, the output of the differential amplifier circuit 22 deviates from zero.

Comparison circuit 23 and threshold circuit 24 correspond to determination circuit 9 in FIG. 4. The comparison circuit 23 compares the output of the differential amplification circuit 22 with a predetermined threshold of the threshold circuit 24, and when the output of the differential amplification circuit 22 exceeds this threshold, it is determined that two-way communication is established. This is output to the echo canceller 3.

The volume adjustment circuit 35 is for adjusting the volume of the speaker. The reason for inserting the volume adjustment circuit 35 at this position is to prevent the gain between the rectifier circuits 20 and 20 from changing depending on the volume of the speaker. If the volume adjustment circuit 35 is inserted between the rectifier circuit 20 connected to the echo path input terminal 5 and the speaker 18, the gain of the amplifier circuit 10 should be adjusted to increase or decrease the volume so as not to change the previous gain. It becomes necessary to increase or decrease proportionately. The other operations of this embodiment are similar to those of the embodiment shown in FIG. 4 and FIG. 1, and therefore their explanation will be omitted.

FIG. 9 shows another embodiment of the present invention. This embodiment detects two-way communication using the properties of zero-crossing waves. In FIG. 9, the same symbols as in FIG. 8 indicate the same parts.

25 is a zero crossing wave circuit that converts the signal into a zero crossing wave, 26 is a counting circuit, 27 is a count difference calculation circuit, and 28 is 10 to 30 m.
This is a frame periodic circuit that generates a timing signal at regular intervals of s.

The zero-crossing wave circuit 25 converts the amplitude of the signal f (t) into f (
t)≦0, the constant value A. When f(t)<O -A
, so that it becomes . As is well known, it has been confirmed both mathematically and aurally that this zero-crossing wave reflects the formant structure of speech. Even if it passes through the echo path, it does not change much in terms of audibility. Signal to speaker 18 (i.e. input signal to echo path)
The signals from the microphone 19 and the microphone 19 (ie, the output signal from the echo path) are converted into zero-crossing waves by a zero-crossing wave circuit 25, respectively. Then, the counting circuit 26 counts the number of times each crosses zero. Then, the difference between the respective count values is calculated by the count difference calculation circuit 27, and a signal proportional to the difference is outputted to the comparison circuit 23. These counts and calculation of the calculation difference are performed at fixed intervals (10 to 3
This is done synchronously every 0ms).

When the signal input to the microphone 19 is only the echo signal, these count values are approximately equal, and the output of the count difference calculation circuit 27 is approximately zero. When the echo signal and other signals are applied to the microphone 19 at the same time (during two-way communication), the count values differ greatly, and the output of the count difference calculation circuit 27 deviates significantly from zero. This deviation is compared with the threshold value of the threshold circuit 24 in the comparator circuit 23 and detected. This two-way communication detection signal is then transmitted to the echo canceling device 3.

The zero crossing wave circuit 25 and the counting circuit 26 correspond to the signal pattern extraction circuit 7 of the embodiment shown in FIG. 1, and the count difference calculation circuit 27 corresponds to the pattern matching circuit 8 shown in FIG.

This embodiment is characterized in that by converting to a zero-crossing wave, a pattern that does not depend on the amplitude level, that is, a pattern that does not depend on changes in the amount of attenuation between the speaker and the microphone, can be obtained. That is, in this embodiment, the amplifier circuit 1 as shown in FIG.
Does not require 0. Further, since counting is also performed at regular intervals, if the delay between the speaker and the microphone is smaller than this interval time, it will not have a large effect on the counted value itself.

The other operations of this embodiment are the same as those of the embodiment shown in FIG. 8, so the explanation will be omitted. FIG. 10 shows another embodiment of the present invention.
Similar to the embodiment shown in FIG. 9, two-way communication is detected using the properties of zero-crossing waves. In FIG. 10, the same reference numerals as in FIG. 9 indicate the same parts.

29 is a zero crossing wave analysis circuit, and 30 is an inter-pattern distance calculation circuit. The other configurations are the same as shown in Figure 9. The zero-crossing wave analysis circuit 29 measures the interval between zero-crossing points of the zero-crossing wave signal and outputs the classified output.

The spacing between zero crossings is related to the formant frequency. Therefore, measuring and classifying this corresponds to analyzing the formant structure without depending on the amplitude level, and the operation of the zero-crossing wave analysis circuit 29 will be specifically explained.

First, a plurality of predetermined interval times, for example, 10
~5IIls, 5~IIIs, 1ms~500μs,
An interval of 500 μs to 250 μs, . . . is determined. Then, the time between zero-crossing points of the zero-crossing signal within a certain frame time is measured and classified into the above-mentioned intervals. For example, 10~
The zero crossing interval of 5+ms is performed ae times, the zero crossing interval of 5 to lms is performed a2 times, and so on.

The interval of 10 to 5 as is 100 to 200 in terms of frequency.
Corresponds to Hz. The presence of this interval time means that the above frequency component was present.

In this way, the zero-crossing wave analysis circuit 29 outputs (all aay...) vectors as a time-series pattern for each frame period. The inter-pattern distance calculation circuit 30 calculates the difference, that is, the distance, between these time-series vector patterns, and outputs a signal proportional to the difference to the comparison circuit 23.

When the signal input to the microphone 19 is only the echo signal, these time-series vector patterns are approximately equal, and the output of the inter-pattern distance calculation circuit 30 is approximately zero. When a reverberant signal and other signals are simultaneously applied to the microphone 19, the time-series vector patterns differ greatly and are detected by an inter-pattern distance calculating circuit 23 by comparison with a threshold value of a threshold circuit 24. This two-way conversation detection signal is then transmitted to the echo canceller 3. This embodiment allows more accurate detection than that shown in FIG. 9.

The zero-crossing wave circuit 25 and the zero-crossing wave analysis circuit 29
This corresponds to the signal pattern extraction circuit 7 shown in the figure, and the inter-pattern distance calculation circuit 30 corresponds to the pattern matching circuit shown in FIG.

The other operations of this embodiment are the same as those shown in FIG. 9, so their explanation will be omitted.

FIG. 11 shows another embodiment of the present invention. In this example,
Two-way communication is detected using time-series patterns of parameters that express the resonance characteristics of the vocal tract. In FIG. 11, the same reference numerals as in FIG. 10 indicate the same parts.

31 is a voice parameter extraction circuit. The other configurations are the same as those shown in FIG. 10 above.

The voice parameter extraction circuit 31 extracts parameters expressing the resonance characteristics of the vocal tract, and extracts parameters expressing the resonance characteristics of the vocal tract, and includes linear prediction coefficients,
PARCOR coefficients, LSP coefficients, etc. are known as parameters. It is also known that these parameters equivalently express the shape of the human throat, and do not change even when passing through the echo path. In this embodiment, first, the audio parameter extraction circuit 31 extracts preset parameters for the signal to the speaker 18 and the signal from the microphone 19. The extracted parameters are used in an inter-pattern distance detection circuit 30 to calculate the difference between these parameters. This calculation result is sent to the comparison circuit 23, where two-way communication is detected in the same manner as in the embodiment shown in FIG. The configuration and operation of the audio parameter extraction circuit 31 of this embodiment will be specifically explained. FIG. 14 shows an embodiment of the audio parameter extraction circuit 31 that extracts linear prediction coefficients or PARCOR coefficients as parameters. In Fig. 14, 14-1 is a buffer memory that temporarily stores input data, and the value converted to a digital value is stored. 14-2 is an autocorrelation coefficient calculation circuit that extracts the autocorrelation coefficient. ．． 14-3 is a linear prediction calculation circuit that extracts linear prediction coefficients or PARCOR coefficients from autocorrelation coefficients. The input data for one frame cycle time is stored in buffer memory 1.
4-1. Next, the autocorrelation coefficient calculation circuit 14-2 calculates the autocorrelation coefficient ρ. is calculated using the formula below.

Here, N is the number of data samples in one frame time, and X, indicates one data sample.

The linear prediction calculation circuit 14-3 uses the above autocorrelation coefficient ρ to calculate the recurrence formula shown in the following equation to obtain the linear prediction coefficient α. ，
PARCOR coefficient K. Calculate.

This recurrence formula is based on the known literature (Itakura et al., rPARCOR
"Formal speech analysis/synthesis system", i! ! Electric Corporation Tsuken Research Practical Report, Vol. 27-6.1978). w,=ρ,,U,=ρa l (X00 = 1 1
αl:'+r = O ・= (2) Ka + r
=Wa/U*; n=0~p-1
-(3)U. +, =U, -K. ．． 1-
W. =U, (1-K2..I)-(4)ατ =α”I
K@+1'+! 'a+1-,;i=1~n+1
・(s) Here, set the initial value using formula (2), and (3) ~
Calculate α7 and K8 sequentially using the recurrence formula (6). usually.
n=10. In other words, α,,α2,...,α1. , or K, K2,..., K1. About 10 parameters are used.

This parameter is extracted for each frame period and output to the inter-pattern distance calculation circuit 30.

The other operations are the same as those in FIG. 10, so their explanation will be omitted.

FIG. 12 shows another embodiment of the present invention. In this example,
This is an example in which the detection signal of the two-way communication detection circuit 4 provided in the acoustic path is supplied to both the acoustic echo canceling device and the line echo canceling device. In FIG. 12, the same symbols as in FIG. 8 indicate the same parts.

32 is an acoustic echo canceler, 33 is a line echo canceler, 3
4 is a 2-wire 4-wire conversion circuit. This embodiment is an example applied to a public address telephone. Note that the other configurations are the same as those shown in FIG.

The acoustic echo canceler 32 is the same as the echo canceler 3 and cancels the echo signal in the acoustic path between the speaker 18 and the microphone 19. The line echo canceling device 33 is the echo canceling device 3
This is the same as and eliminates the anti-\I signal due to the unbalance of the 2-wire 4-wire conversion circuit 34. The 2-wire/4-wire conversion circuit 34 is a type of bridge circuit that converts a 4-wire signal of a public address telephone into a bidirectional 2-wire signal of a telephone line.

The two-way communication detection circuit 4 is constructed in the same manner as that shown in FIG. 8, and operates in the same manner, so a description of its operation will be omitted.

In a loudspeaker telephone, only a transmitted signal and a received signal exist, and a two-way call occurs when the transmitted signal and the received signal overlap in time. Therefore, the two-way conversation for the acoustic echo canceller 32 and the two-way conversation for the line echo canceler 33 are the same. In other words, it is possible to provide the two-way communication detection circuit 4 on either the echo path between the speaker and the microphone or the echo path in the 2-wire 4-wire conversion circuit, and supply this detection signal to both echo cancelers. Become. 1st
Figure 3 shows another embodiment of the present invention. In Figure 13, the same symbols as in Figure 12 indicate the same items. In this example, 2
The detection signal of the bidirectional detection circuit 4 provided in the anti-vIM in the line 4Is circuit is transmitted to the acoustic echo canceler 32 and the line echo canceler 3.
This is the case when supplying to both of 3. Each of the embodiments described above can be applied to a telephone device, particularly a loudspeaker telephone device. Examples of the loudspeaker telephone device include a hands-free car telephone device.

[Effects of the Invention] As described above, according to the present invention, two-way communication is detected based on the difference in the pattern (shape) of the input and output signals of the echo path, not based on signal level comparison. However, there will be no detection error. As a result, malfunctions and performance deterioration of the echo canceling device can be avoided.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing an embodiment of the present invention, Fig. 2 is an explanatory diagram showing a specific example of an echo path, Fig. 3 is a block diagram showing an example of an echo canceling device, and Figs. 4 to 13 are respectively FIG. 14 is a block diagram showing another embodiment of the present invention. FIG. 14 is a block diagram showing an example of an audio parameter extraction circuit. DESCRIPTION OF SYMBOLS 1... Echo path, 3... Echo cancellation device, 7... Signal pattern extraction circuit, 8... Pattern matching circuit, 9...
・Judgment circuit, 10... Amplification circuit, 11... Delay circuit, 12... Variable gain amplifier circuit, 13... Gain control circuit, 14. Power detection circuit, 15... Echo cancellation amount calculation circuit , 16... variable delay circuit, 17... delay amount control circuit,
18... Speaker, 19... Microphone, 20... Rectifier circuit, 21... Smoothing circuit, 22... Differential amplifier circuit,
24...Threshold circuit, 25...Zero crossing wave circuit, 2
6... Counting circuit, 27... Counting difference calculation circuit, 28...
・Frame period circuit, 29...Zero crossing analysis circuit, 30
... Inter-pattern distance calculating circuit, 31... Audio parameter extraction circuit, 32... Acoustic echo canceller, 33...
Line echo canceller, 34...2-wire 4-wire conversion circuit, 35
...Volume adjustment circuit. Figure 1 Figure 2 Figure 4 Figure 8 Figure 9 Figure 11

Claims (1)

[Claims] 1. First and second signal pattern extraction means for extracting time-series patterns of parameters indicating characteristics of signals;
, pattern matching means for calculating a distance indicating a difference between the time series patterns output by the second signal pattern extraction means;
determining means for determining a difference between patterns based on the output of the pattern matching means based on a predetermined threshold; the first signal pattern extracting means is connected to an input terminal of the echo path; A two-way call detection method, characterized in that a signal pattern extraction means is connected to an output terminal of an echo path, and when a difference between the patterns exceeds a threshold value, a two-way call is determined. 2. The two-way conversation detection method according to claim 1, wherein the time-series pattern of the parameters is one that remains unchanged even when the signal passes through an echo path. 3. Two-way call detection according to claim 2, characterized in that a delay means for delaying the signal is provided, and the input terminal of the echo path and the first signal pattern extraction means are connected via the delay means. method. 4. Two-way call detection according to claim 2, characterized in that an amplification means for amplifying the signal is provided, and the output terminal of the echo path and the second signal pattern extraction means are connected via the amplification means. method. 5. The two-way call detection method according to claim 2 or 4, wherein the time-series pattern of the parameters is a time-series pattern of signal power. 6. The two-way call detection method according to claim 2 or 3, wherein the time series pattern of the parameters is a time series pattern of the number of zero crossings at fixed time intervals. 7. The two-way call detection method according to claim 2 or 3, wherein the time-series pattern of the parameters is a time-series pattern of time classification parameters between zero crossing points at regular time intervals. 8. The two-way conversation detection method according to claim 2 or 4, wherein the time-series pattern of the parameters is a time-series pattern of parameters expressing resonance characteristics of the vocal tract. 9. An echo cancellation amount calculation means for calculating the amount of echo signal cancellation of the echo cancellation device is provided, and the delay means is configured such that the output of the pattern matching means becomes zero when the amount of echo cancellation exceeds a predetermined value. 4. The two-way call detection method according to claim 3, further comprising controlling the amount of delay. 10. The device is characterized by providing the echo cancellation amount calculation means, and controlling the gain of the amplification means so that the output of the pattern matching device becomes zero when the echo cancellation amount exceeds a predetermined value. The two-way call detection method according to claim 4. 11. An echo canceling device comprising the two-way conversation detection method according to any one of claims 2 to 10. 12. An acoustic echo canceling device comprising the two-way conversation detection method according to any one of claims 2 to 10. 13. A line echo canceling device characterized by having the two-way conversation detection method according to any one of claims 2 to 10. 14. A loudspeaker telephone comprising the echo canceling device according to claim 11. 15. an acoustic echo canceler for canceling echo signals in the acoustic path caused by acoustic coupling between the speaker and the microphone;
In a loudspeaker telephone equipped with a line echo cancellation device that cancels echo signals in a line echo path caused by unbalance in line four-wire conversion, a claim is made for any one of the acoustic echo path or the line echo path. A loudspeaker telephone, characterized in that it applies the two-way conversation detection method according to any one of Items 2 to 10, and simultaneously controls both the acoustic echo canceling device and the line echo canceling device based on this detection signal. . 16. A car telephone device comprising the loudspeaker telephone according to claim 16.