WO2002095975A1 - Appareil de traitement d'echos - Google Patents
Appareil de traitement d'echos Download PDFInfo
- Publication number
- WO2002095975A1 WO2002095975A1 PCT/JP2002/004860 JP0204860W WO02095975A1 WO 2002095975 A1 WO2002095975 A1 WO 2002095975A1 JP 0204860 W JP0204860 W JP 0204860W WO 02095975 A1 WO02095975 A1 WO 02095975A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- echo
- signal
- frequency
- low
- digital signal
- Prior art date
Links
- 238000012545 processing Methods 0.000 title claims abstract description 41
- 230000005540 biological transmission Effects 0.000 claims abstract description 19
- 230000001629 suppression Effects 0.000 claims description 37
- 230000005236 sound signal Effects 0.000 claims description 29
- 238000001514 detection method Methods 0.000 claims description 10
- 238000000034 method Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 239000000284 extract Substances 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 description 58
- 238000010586 diagram Methods 0.000 description 12
- 230000000694 effects Effects 0.000 description 10
- 230000002542 deteriorative effect Effects 0.000 description 9
- 230000014509 gene expression Effects 0.000 description 9
- 230000007423 decrease Effects 0.000 description 8
- 230000006866 deterioration Effects 0.000 description 7
- 238000012935 Averaging Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 101100521097 Caenorhabditis elegans pri-1 gene Proteins 0.000 description 1
- 101000822695 Clostridium perfringens (strain 13 / Type A) Small, acid-soluble spore protein C1 Proteins 0.000 description 1
- 101000655262 Clostridium perfringens (strain 13 / Type A) Small, acid-soluble spore protein C2 Proteins 0.000 description 1
- 101000655256 Paraclostridium bifermentans Small, acid-soluble spore protein alpha Proteins 0.000 description 1
- 101000655264 Paraclostridium bifermentans Small, acid-soluble spore protein beta Proteins 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 244000309464 bull Species 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000011410 subtraction method Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R3/00—Circuits for transducers, loudspeakers or microphones
- H04R3/02—Circuits for transducers, loudspeakers or microphones for preventing acoustic reaction, i.e. acoustic oscillatory feedback
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M19/00—Current supply arrangements for telephone systems
- H04M19/08—Current supply arrangements for telephone systems with current supply sources at the substations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M9/00—Arrangements for interconnection not involving centralised switching
- H04M9/08—Two-way loud-speaking telephone systems with means for conditioning the signal, e.g. for suppressing echoes for one or both directions of traffic
- H04M9/082—Two-way loud-speaking telephone systems with means for conditioning the signal, e.g. for suppressing echoes for one or both directions of traffic using echo cancellers
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
- G10L2021/02082—Noise filtering the noise being echo, reverberation of the speech
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/60—Substation equipment, e.g. for use by subscribers including speech amplifiers
- H04M1/6033—Substation equipment, e.g. for use by subscribers including speech amplifiers for providing handsfree use or a loudspeaker mode in telephone sets
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/72—Mobile telephones; Cordless telephones, i.e. devices for establishing wireless links to base stations without route selection
- H04M1/725—Cordless telephones
Definitions
- the present invention relates to an echo processing device that reduces acoustic echo output from sub power and input to a microphone via a reverberation path in voice communication of a vehicle-mounted phone, a mobile phone, or the like.
- FIG. 1 is a block circuit diagram showing a configuration of a conventional echo processing apparatus.
- 1 is a DZA converter for converting a digital received signal into an analog audio signal
- 2 is an amplifier for amplifying an audio signal
- 3 Is a speaker that outputs sound according to the amplified audio signal.
- Reference numeral 4 denotes a microphone that converts the input sound into an audio signal
- 5 denotes an amplifier that amplifies the audio signal
- 6 denotes an A / D converter that converts the amplified audio signal into a digital signal.
- ⁇ is an echo canceller means for generating a pseudo echo signal based on the received signal and subtracting the pseudo echo signal from the output digital signal converted by the AZD converter 6 to remove one echo component.
- the echo canceller means 7, 8 is an adaptive filter means for generating a pseudo echo signal based on the received signal, the output digital signal and the determination signal, and 9 is an addition for adding the output digital signal and the pseudo echo signal. It is a vessel.
- Numeral 10 denotes a double talk detecting means for outputting a silent or double talk determination signal of the received signal to the echo canceller means 7 based on the received signal, the output digital signal and the transmission signal.
- the digital signal sent from the far-end talker is used as the received signal, and the audio signal input from the near-end talker is amplified by the amplifier 5 and the output digital signal converted by the A / D converter 6.
- a signal in which the echo component is suppressed by the echo canceller means 7 is defined as a transmission signal.
- the received signal from the far-end speaker is input to the echo canceller means 7 and the double talk detection means 10, and is also converted into an analog signal, that is, a voice signal by the D / A converter 1, and is amplified by the amplifier 2.
- the sound is output from speaker 3 as sound.
- a part of the output from the speaker 3 is input to the microphone 4 through an echo path, and is mixed with an audio signal to be transmitted as an echo signal.
- the adaptive filter means 8 calculates an adaptive filter coefficient using the received signal and the transmission signal to be fed back, and generates a pseudo echo signal.
- the adder 9 subtracts the pseudo echo signal from the output digital signal, suppresses an echo component included in the output digital signal, and generates a transmission signal.
- the adaptive filter coefficient is sequentially obtained and updated in order to cope with fluctuations in the echo path.
- the double talk detecting means 10 determines whether or not the output digital signal is in a double talk state in which the echo signal and the voice signal of the near-end speaker are simultaneously included, and whether or not the received signal is silent. To the adaptive filter means 8 as a judgment signal. The adaptive filter means 8 stops updating the adaptive filter coefficient during double talk or when the received signal is silent, thereby preventing the calculation accuracy of the filter coefficient from deteriorating due to the double talk.
- the method of double-talk detection in the double-talk detecting means 10 is disclosed, for example, in Japanese Patent Application Laid-Open No. H10-242,911, and is performed as follows. .
- the double talk detecting means 10 is configured to output the average power S of the output digital signal, The average power X of the transmission signal and the average power E of the transmission signal are obtained, and double talk is detected by a combination of the following equations (1) to (3).
- P i to p 3 in each equation are predetermined constants, which are determined according to the use environment.
- equation (1) is not satisfied and equation (2) is satisfied, it is determined that a double torque is obtained.
- Equation (3) it is determined that double talk is caused by a small amount of echo suppression and a large amount of input sound other than echo signals.
- the conventional echo processing apparatus is configured as described above, the conventional echo processing apparatus shown in FIG.
- the conventional echo processing apparatus shown in FIG. In a device, when low-frequency components that cannot be reproduced as sound are input to a speaker, vibration of high-frequency sound input at the same time may be disturbed, resulting in non-linear distortion. In this case, non-linear distortion occurs in the echo signal input from the microphone, the accuracy of calculating the adaptive filter coefficient in the adaptive filter means 8 deteriorates, and the generated pseudo echo signal and the echo signal are reduced. There was a problem that the difference became large and the amount of echo suppression was reduced. ,
- Japanese Patent Application Laid-Open No. Hei 6-22069 discloses a high-pass filter before a spin force.
- a method has been proposed in which a low frequency region component that cannot be reproduced by the beeker is inserted in advance. Applying this method to the conventional echo processor shown in Fig. 1 and introducing a high-pass filter between the DZA converter 1 and the amplifier 2 can be considered.
- the characteristics of the received signal input to 8 and the echo signal output from loudspeaker 3 and input to microphone 4 in the low frequency band greatly differ. This is because the received signal has low frequency region components but the echo signal does not.
- the calculation accuracy of the adaptive filter coefficient in the adaptive filter means 8 deteriorates, the difference between the generated pseudo echo signal and the echo signal increases, and the echo suppression amount decreases.
- the double talk detecting means 10 of the conventional echo processor shown in FIG. 1 determines whether S and X in equation (2) have close values, or that S and E in equation (3) are close. If you do, you may make a mistake in determining double talk. For example, if the value of p 2 in equation (2) is set low so that it is easy to judge a double-talk in order to prevent the adaptive filter coefficient from being erroneously updated, the single talk (only the received signal) (X is large), but when the distance between the speaker 3 and the microphone 4 is short, or when the amplification values of the amplifiers 2 and 5 are large, the power of the echo signal mixed into the output digital signal increases. S is large), it may be erroneously determined as double talk.
- the conventional echo processing device it is difficult to clearly determine the single talk and the double talk, and the determination may be erroneous. As a result, the stop and start times of the update of the adaptive filter coefficient may be incorrectly adapted. There was a problem that the calculation accuracy of the fill coefficient was degraded and the echo suppression amount was reduced.
- the present invention has been made to solve the above problems, and reduces a non-linear distortion of a sound output from a spurious force, prevents a deterioration in calculation accuracy of an adaptive filter coefficient, and generates a pseudo echo signal and an acoustic echo. It is an object of the present invention to obtain an echo processing device that reduces the difference between the two and reduces the amount of echo suppression. Also, the present invention provides an echo processing that accurately determines double talk, correctly stops and starts updating of the adaptive filter coefficient, prevents deterioration in the calculation accuracy of the adaptive filter coefficient, and prevents a decrease in the amount of echo suppression. The purpose is to obtain the device. Disclosure of the invention
- An echo processing device includes: a high-pass filter that suppresses a low-frequency component of a received signal formed of a digital signal; a DZA converter that converts a low-band suppressed received signal that has passed through the high-pass filter into an audio signal.
- a speaker that outputs sound based on the audio signal, a microphone to which an echo of the sound output from the speaker may be input, and a conversion of the audio signal output from the microphone into a digital signal.
- a pseudo echo signal is generated based on the AZD converter and the low-frequency-suppressed received signal obtained through the high-pass filter, and the pseudo echo signal is subtracted from the digital signal output by the AZD converter.
- an echo canceller means for generating a transmission signal.
- the nonlinear distortion of the sound output from the speaker can be reduced, and the acoustic echo mixed into the digital signal output by the A / D converter and the low-frequency suppressed reception signal input to the echo canceller means can be reduced.
- the accuracy of calculating the adaptive filter coefficient by the echo canceller is prevented from deteriorating, and the difference between the pseudo echo signal and the acoustic echo is reduced. This has the effect of preventing a reduction in the amount of echo suppression.
- An echo processing device comprises: a high-pass filter for suppressing a low frequency domain component of a reception signal formed of a digital signal; and a D / A conversion for converting a low-frequency suppression reception signal having passed through the high-pass filter into an audio signal.
- a high-pass filter for suppressing a low frequency domain component of a reception signal formed of a digital signal
- a D / A conversion for converting a low-frequency suppression reception signal having passed through the high-pass filter into an audio signal.
- a speech force for outputting sound based on the audio signal
- a microphone to which an echo of the sound output from the speaker may be input
- an AZD for converting the audio signal output from the microphone into a digital signal
- a pseudo-echo signal based on the low-pass suppressed reception signal obtained by passing through the high-pass filter, and subtracting the pseudo-echo signal from the digital signal output by the A / D converter.
- An echo canceller means for generating a transmission signal by extracting a low frequency domain component of the digital signal output by the AZD converter, performing a double talk determination based on the low frequency domain component, and performing a double talk determination based on the double talk determination.
- a double-talk detecting means for controlling the stop and start of the updating of the filter coefficient of the echo canceller means.
- the nonlinear distortion of the sound output from the loudspeaker can be reduced, and the acoustic echo mixed into the digital signal output from the A / D converter and the low-frequency-suppressed reception signal input to the echo canceller means can be reduced.
- the accuracy of calculating the adaptive filter coefficient in the echo canceller is prevented from deteriorating, the difference between the pseudo echo signal and the acoustic echo is reduced, and the amount of echo suppression is reduced. Can be prevented from decreasing.
- the low-frequency domain component of the digital signal output by the A / D converter is extracted, and double-talk determination is performed based on the low-frequency domain component. Stopping and starting the update of the filter coefficient correctly and preventing the accuracy of the adaptive filter coefficient from deteriorating further reduces the difference between the pseudo echo signal and the acoustic echo, and has the effect of preventing a decrease in the amount of echo suppression. .
- the double talk detecting means when the double talk detecting means has a small low frequency region component of the digital signal output by the AZD converter, The more difficult it is to determine double talk.
- the double talk detecting means makes it easier to determine that the digital signal output from the AZD converter has double talk as the low frequency region component of the digital signal is larger.
- the double talk detecting means calculates an amount of a background noise component included in the digital signal output from the A / D converter, and when the background noise component is small, The greater the low-frequency component of the signal, the easier it is to determine that double talk is occurring.
- the echo processing apparatus includes a noise canceller that suppresses a background noise component of the digital signal output by the A / D converter, and the double talk detecting means includes a background noise output from the noise sublesser.
- the digital signal in which is suppressed is input.
- the double talk detecting means converts a low-frequency component of the digital signal output from the A / D converter by a low-pass filter having a cutoff frequency corresponding to a cutoff frequency of the high-pass filter. It is to extract.
- An echo processing device includes: a storage unit that stores a set frequency; a control CPU that reads a set frequency from the storage unit and outputs the read frequency to a high-pass filter and a double talk detection unit; A plurality of high-pass filters having different cut-off frequencies in advance, and selecting a high-pass filter having a cut-off frequency corresponding to a set frequency output from the control CPU to suppress a low frequency region component of a received signal;
- the talk detecting means is for extracting a low frequency region component equal to or lower than a set frequency output from the control CPU when extracting a low frequency region component of the digital signal output by the A / D converter. is there.
- the double talk detecting means includes a double toe of a low frequency region component of the digital signal output by the A / D converter.
- the degree of contribution to the judgment is changed by the set frequency.
- the double talk detecting means includes a plurality of low-pass filters having different cut-off frequencies in advance, and selects a low-bass filter having a cut-off frequency corresponding to the set frequency output from the control CPU. Then, it suppresses the high frequency region component of the digital signal output by the A / D converter and extracts the low frequency region component.
- FIG. 1 is a block diagram showing a configuration of a conventional echo processor.
- FIG. 2 is a block diagram showing a configuration of an echo processor according to Embodiment 1 of the present invention.
- FIG. 3 is an explanatory diagram showing an operation of a high-pass fill according to the first embodiment of the present invention.
- FIG. 4 is a block diagram showing a configuration of an echo processing apparatus according to Embodiment 3 of the present invention.
- FIG. 5 shows a configuration of an echo processing apparatus according to Embodiment 4 of the present invention.
- FIG. 3 is a block circuit diagram.
- FIG. 6 is an explanatory diagram showing an operation of a high-pass fill according to Embodiment 4 of the present invention.
- FIG. 2 is a block circuit diagram showing a configuration of the echo processing device according to the first embodiment of the present invention.
- 11 is a high-pass filter that suppresses low-frequency components of a received signal composed of a digital signal
- 1 is a high-pass filter that converts a low-band suppressed received signal that has passed through 1 into an analog audio signal.
- An A / A converter 2 is an amplifier for amplifying an audio signal
- 3 is a speaker for outputting sound according to the amplified audio signal.
- Reference numeral 4 denotes a microphone for inputting sound and converting the sound signal into an audio signal
- 5 an amplifier for amplifying the audio signal
- 6 an A / D converter for converting the amplified analog audio signal into a digital signal.
- the echo canceller means 7, 8 is an adaptive filter means for generating a pseudo echo signal based on the low-frequency-suppressed reception signal, the transmission signal and the judgment signal, and 9 is a digital signal from the A / D converter 6. And a pseudo echo signal.
- Reference numeral 12 denotes double talk detecting means for judging silence or double talk of the received signal based on the low-frequency suppressed reception signal, output digital signal and transmission signal, and outputting the judgment signal to the echo canceller means 7. Next, the operation will be described.
- the high-pass filter 11 When the received signal is input to the high-pass filter 11, the high-pass filter 11 suppresses signal components below the cut-off frequency of the filter and outputs a low-band suppressed received signal.
- the low-band suppression received signal is supplied to the adaptive filter means 8 and the D / A converter 1 and also to the double talk detection means 12.
- the high-pass filter 11 is composed of, for example, an 8th-order IR type digital filter.
- the cut-off frequency of the high-pass filter 11 is determined in advance by checking the lower limit frequency of the reproducible low frequency of the speaker 3, and is set in accordance with this value. If the lower limit of the reproducible frequency of the speaker 3 is 400 Hz, the coefficient of the digital filter is designed so that the cutoff frequency is 400 Hz.
- FIG. 3 is an explanatory diagram showing the operation of the filter according to the first embodiment of the present invention.
- (a) to (c) show an operation example of the high-pass filter 11.
- (A) is the frequency characteristic when the received signal is a sound
- (b) is the response characteristic of the high-pass filter 11
- (c) is the frequency characteristic of the low-band suppression received signal output from the high-pass filter 11. It is.
- fc is the cutoff frequency of the high-pass filter 11.
- the low-frequency-suppressed reception signal input to the D / A converter 1 is converted into an analog signal, that is, an audio signal, amplified by the amplifier 2, and output from the speaker 3 as sound.
- an analog signal that is, an audio signal, amplified by the amplifier 2, and output from the speaker 3 as sound.
- low-frequency components that cannot be reproduced by the speaker 3 are suppressed by the high-pass filter 11, so that non-linear distortion does not occur in the sound output from the speaker 3. Therefore, non-linear distortion does not occur in the acoustic echo input from the microphone 4.
- the adaptive filter means 8 calculates an adaptive filter coefficient based on the low-frequency suppressed reception signal and the transmission signal to generate a pseudo echo signal, and the adder 9 subtracts the pseudo echo signal from the output digital signal.
- non-linear distortion does not occur in the acoustic echo, and the acoustic echo mixed into the audio signal and the low-frequency-suppressed received signal input to the adaptive filter means 8 are converted to the low-frequency
- the adaptive filter means 8 can accurately calculate the adaptive filter coefficient and generate a pseudo echo signal close to the actual echo signal.
- the double talk detecting means 12 determines double talk based on the low-frequency suppressed reception signal, the output digital signal, and the transmission signal. Hereinafter, the operation of the double talk detecting means 12 will be described in detail.
- the double talk detecting means 1 2 samples the average power Xh of the low-frequency-suppressed reception signal, the average power S of the output digital signal, and the average power E of the transmission signal, for example, by sampling the level of each signal within the period at regular intervals Then, the sum of squares of each value is calculated by averaging the number of samples.
- the high-frequency components hereinafter referred to as high-frequency components
- the mouth-to-pass filter is composed of, for example, an 8th-order IIR digital filter.
- the cut-off frequency of the low-pass fill is set to the same value as the cut-off frequency of the no-pass fill.
- the output digital signal The low-frequency power and the low-frequency power of the low-frequency suppression received signal
- S l and XI are defined as S l and XI, for example, by averaging the sum of squares of the sample values with the number of samples.
- the double talk detecting means 12 calculates the minimum value of the average power S of the output digital signal obtained in a predetermined period (for example, one second) as a periodicity in the output digital signal in the period in which the minimum value is obtained. Only when there is no background noise Extract as one N s value.
- the presence or absence of the periodicity can be obtained by, for example, determining the maximum value of the autocorrelation coefficient of the output digital signal using a threshold. If the maximum value of the self-correlation number exceeds the threshold value, it is determined that there is periodicity, and if not, there is no value.
- the section of the output digital signal in which the value of the average power S is larger than the preset threshold value from N s and has periodicity is detected as a sound section, and for example, the sound section in the past 1 second is detected.
- the average power of the section is obtained as V s, and the ratio of V s to N s (V s / N s) is obtained as R s.
- double talk is determined using, for example, the following equations (4) to (7).
- Equations (4) and (5) are not satisfied and Equations (6) and (7) are satisfied, the echo suppression is small, and the double talk due to the large amount of input sound other than the echo signal is considered. judge.
- p 4 is the variable value that changes depending on the value of the previous obtained R s, large R s, for example if it exceeds 3 6 2, i.e. when the amount of background noise 1.0, does not exceed That is, if the amount of background noise is large, set to 1.5.
- whether or not the received signal is silent is determined using the average power Xh of the low-band suppressed received signal, but the average power X of the received signal before passing through the high-pass filter 11 is determined. May be applied to equation (4).
- (a) is an example of the frequency characteristic when the received signal is a sound
- (b) is the response characteristic of the high-pass filter 11
- (c) is the low-band suppressed received signal as described above. It is a frequency characteristic.
- (D) is the same low-frequency-suppressed received signal as in (c)
- (e) is the response characteristic of the low-pass filter in the double talk detecting means 12
- the cut-off frequency fc is the same as that of the high-pass filter 11 Is the same.
- (F) is a single-pass filter output obtained from the signal (d) through this low-pass filter.
- the average power of the low-pass filter output (f) is equivalent to XI in equation (7), but the value of XI is small because of the low-pass suppression effect of the high-pass filter 11.
- (g) shows the frequency characteristic when the low-band suppression received signal having the frequency characteristic of (c) is converted from speech 3 into voice, and only the echo signal component obtained by inputting to the microphone 4 through the echo path is shown. It is.
- (H) is the response characteristic of the mouth-to-pass fill in the same double talk detecting means 12 as (e), and (i) is the output of this mouth-to-pass fill.
- the average power of this low-pass fill output (i) is equivalent to S1 in equation (7).
- S 1 is as small as X 1.
- (j) is the frequency characteristic when the output digital signal is composed of the echo signal and the voice signal of the near-end speaker
- (k) is the one-pass filter in the double talk detecting means 12 same as (e).
- (1) is the output of the low-pass filter.
- the average power of the mouth-to-passfill evening output (1) is the formula This corresponds to S1 in (7), but in this case, the output digital signal includes the voice of the near-end speaker, so that the low-frequency component is included.
- the low-pass filter output (i) This is a large value compared to S 1 of FIG.
- the double-talk detecting means 12 is used when the value of S 1 is small and close to X 1, that is, the equation (7) If does not hold, it is determined that the output digital signal is single talk with only echo.
- the double talk detecting means 12 uses the double digital signal composed of the echo signal and the near-end talker's voice signal in the output digital signal. Judge as talk.
- S1 is the power of the low-frequency component of the output digital signal, and S can be regarded as the power of the entire band of the output digital signal.
- the double-talk determination means 12 uses the equations (5), (6) ) Holds, but if equation (7) using S 1 does not hold, it is not determined that the signal is a double-talk. Therefore, the single-talk echo only is doubled compared to the conventional echo processor that does not use equation (7). There is little misjudgment as a talk, and accurate double talk judgment can be performed. However, when the amount of background noise is small, the background noise included in S1 is small, and the reliability of the value of S1 is determined to be high, and p4 is set to a small value. It's getting easier.
- the double talk detecting means 12 determines whether the received signal is silent or in a double talk state based on the above conditions, and transmits the result to the adaptive filter means 8 as a determination signal.
- the adaptive filter means 8 stops updating the adaptive filter coefficient during double talk or when the received signal is silent, thereby preventing the calculation accuracy of the filter coefficient from deteriorating due to double talk. Also, the adaptive filter means 8 starts updating the adaptive filter coefficient when double talk is not performed and the received signal is not silent.
- low-frequency components of a received signal that cannot be reproduced by speaker 3 are suppressed by high-pass filter 11, and the output of high-pass filter 11 is adaptively filtered. Since the signal is also input to the means 8, the nonlinear distortion of the echo signal output from the speaker 3 can be reduced. The difference between echo signals can be reduced to prevent a reduction in the amount of echo suppression.
- the fact that the low frequency is suppressed by the high-pass filter 11 of the echo signal included in the output digital signal is used to determine the double talk, so that the double talk can be determined with high accuracy and the adaptive filter can be determined. It correctly stops and starts updating the evening coefficient, prevents the accuracy of the adaptive fill coefficient from deteriorating, reduces the difference between the pseudo echo signal and the echo signal, and prevents a reduction in the amount of echo suppression.
- the low-frequency components of the output digital signal and the low-frequency suppression received signal are extracted by low-pass filtering, the low-frequency components of the output digital signal and the low-frequency suppression reception signal can be obtained by relatively simple processing. it can. Embodiment 2.
- double-talk detection is performed when Expression (4) is not satisfied and Expressions (5) and (7) are satisfied, or when Expressions (4) and (7) are satisfied. If (5) is not satisfied and Eqs. (6) and (7) are satisfied, double talk is determined. However, again with R s is large, for example 3 when more than 6 2, i.e. smaller if the amount of background noise equation (4) is Equation (7) is established not satisfied. Based on this fact, double talk may be determined regardless of whether Expressions (5) and (6) hold. Therefore, when the amount of background noise is small, even if the value of S or E is small, In this case, it is possible to accurately detect the talk, and to prevent false determination of double talk when the amount of background noise is large.
- the background noise component included in the output digital signal is utilized by utilizing the fact that the low frequency band of the echo signal included in the output digital signal is suppressed by the high-pass filter 11.
- the double talk is determined mainly by using the low frequency region component of the output digital signal, so that the double talk can be determined with high accuracy, and the stop and start of the update of the adaptive filter coefficient can be performed. It is possible to reduce the difference between the pseudo echo signal and the echo signal by preventing the deterioration of the precision of the adaptive filter coefficient by performing the correction correctly, thereby preventing the reduction of the echo suppression amount.
- FIG. 4 is a block circuit diagram showing a configuration of an echo processing apparatus according to Embodiment 3 of the present invention.
- reference numeral 17 denotes a noise suppressor.
- Other configurations are the same as those in FIG. 2, and the operation of the double talk detecting means 12 is the same as that of the second embodiment.
- the noise sub-laser 17 inputs the output digital signal from the AZD converter 6, suppresses the background noise component included in the output digital signal, and outputs the signal to the double talk detecting means 12.
- Bol 1 “S uppression of A cousticnoiseinseechus ingspectrals ub traction”, IEEETrans. ASSP. Vol. ASSP-27, No. 2, A This is performed by the spectral subtraction method disclosed in pri 1 19 79 (Reference 4).
- Double talk detection means Since the amount of the background noise component included in the output digital signal is reduced by the noise sub-lesser 17, the calculated value of R s increases, and the case where R s exceeds 36 2 increases.
- Equation (5) and (6) can be satisfied, if (4) is not satisfied and Equation (7) is satisfied, the number of cases where double talk is determined increases, and the values of S and E increase. Even if is small, the possibility of accurately detecting a double torque increases.
- the noise sub-lesser 17 is arranged at a position bypassing the echo canceller means 7, and the output signal which has been transformed by the noise sub-lesser 17 with its background noise suppressed is input to the double-talk detecting means 12.
- the pseudo echo signal generated by the adaptive filter means 8 does not deteriorate due to the introduction of the noise sub-sensor 17.
- the noise sublesser 17 suppresses the background noise component included in the output digital signal, so that even when the output digital signal contains background noise, Double talk can be determined with high accuracy, the update of the adaptive filter coefficient is correctly stopped and started, the accuracy of the adaptive filter coefficient is prevented from deteriorating, the difference between the pseudo echo signal and the echo signal is reduced, and the echo suppression amount is reduced. Reduction can be prevented.
- FIG. 5 is a block circuit diagram showing a configuration of an echo processing apparatus according to Embodiment 4 of the present invention.
- reference numeral 13 denotes storage means for storing a set frequency
- 14 denotes a set frequency of the storage means 13.
- This is a control CPU for reading and outputting to the high-pass filter 15 and the double talk detecting means 16.
- the high-pass filter 15 is composed of a plurality of filter elements having different cut-off frequencies in advance, and a high-pass filter for selecting and setting a cut-off frequency corresponding to the set frequency output from the control CPU 14.
- the lower limit of the low-frequency range that the beeker 3 can reproduce depends on the performance of the speaker.
- a filter element with a plurality of cutoff frequencies corresponding to the lower limit frequency of reproducible low frequency assumed by the speaker 3 is prepared in the high pass filter 15. For example, if the lower reproducible lower-limit frequency expected from speaker 3 is 300 Hz to 600 Hz, then 300 Hz, 400 Hz, 500 Hz every 100 Hz Prepare filter elements with four types of cutoff frequencies, Hz and 600 Hz.
- the mouth-to-pass filter inside the double talk detecting means 16 also holds filter elements having four kinds of cutoff frequencies at intervals of 100 Hz from 300 Hz to 600 Hz.
- the storage means 13 is composed of, for example, a ROM, and measures a lower reproducible lower frequency limit of the speaker 3 at the time of design, determines a set frequency corresponding to the lower limit frequency, and stores it at the time of manufacture.
- the control CPU 14 reads out the value of the set frequency stored in the storage means 13 and outputs it to the high-pass filter 15 and the double talk detection means 16.
- the high-pass filter 15 selects a cutoff frequency corresponding to this set frequency value, for example, when the set frequency is 400 Hz, selects a filter element having a cutoff frequency of 400 Hz to select a low-frequency component of the received signal. Suppress the component and output the low-band suppression received signal.
- the double talk detecting means 16 selects a filter element having a cutoff frequency corresponding to the set frequency value and suppresses the high frequency components of the output digital signal and the low frequency suppression received signal. Then, the double toe described in FIG. Xh, XI, S, Sl, and E are respectively obtained by the same processing as that of the detection means 12 and the double talk is determined using, for example, Equations (8) to (10).
- Equation (10) it is determined that double talk is caused by a small amount of echo suppression and a large amount of input sound other than echo signals. I do. '
- the value in equation (9) is a parameter representing the contribution of the output digital signal and the low-band suppressed received signal to the low-band power S 1 and X 1 in equation (9).
- FIG. 6 is an explanatory diagram showing an operation of a filter according to Embodiment 4 of the present invention, and shows a frequency characteristic of an output digital signal that has passed through a mouth-to-pass filter of a double talk detecting means 16. It is.
- (a) shows the case where the output digital signal is an echo signal only and the low-pass lower limit frequency is low
- (C) is the case where the output digital signal is composed of the echo signal and the voice signal of the near-end speaker, and the low-pass lower limit frequency of the one-pass filter is low.
- (D) is the case where the lower limit frequency of the low frequency band is high with the same signal configuration.
- the power difference between (b) and (d) is larger than the power difference between (a) and (c), and the reliability of S1 is higher when the lower-limit frequency in the mouth-to-pass filter is higher.
- a plurality of filters having different cutoff frequencies are provided in a high-pass filter 15 composed of a plurality of filter elements, and similarly, a plurality of filters having different cutoff frequencies are prepared.
- a low-pass filter consisting of an element is prepared in the double-talk detecting means 16 and the cut-off frequency of the high-pass filter and the one-pass filter is switched in accordance with the set frequency stored in the storage device 13.
- Embodiment 5 depending on the cutoff frequency of the high-pass fill and low-pass fill, Since the contribution of the low-frequency power of the output digital signal in the Bull Talk condition formula is changed, it is possible to determine the double talk with high accuracy, stop and start updating the adaptive filter coefficient correctly, and By preventing the coefficient from deteriorating in accuracy, it is possible to reduce the difference between the pseudo echo signal and the echo signal, thereby preventing a reduction in echo suppression.
- the low-frequency power is obtained by suppressing the high-frequency component of the output digital signal with a low-pass filter. Instead, the output digital signal is converted to a power spectrum using a fast Fourier transform (FFT), and low-frequency power is obtained by adding components below the cut-off frequency on the power spectrum. good.
- FFT fast Fourier transform
- the low-pass power is obtained on the frequency axis using the FFT, and thus the mouth-to-pass filter characteristic shown in (h) of FIG.
- the slope can be cut out steeply without a slope, and the power can be determined, and the accurate low-frequency power can be determined.
- the echo processing device is suitable for reducing the acoustic echo output from the speaker and input to the microphone via the echo path. .
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Otolaryngology (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Telephone Function (AREA)
- Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)
Description
Claims
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP02771733A EP1300963A4 (en) | 2001-05-22 | 2002-05-20 | APPARATUS FOR TREATING ECHOS |
JP2002592318A JPWO2002095975A1 (ja) | 2001-05-22 | 2002-05-20 | エコー処理装置 |
US10/333,224 US6868158B2 (en) | 2001-05-22 | 2002-05-20 | Echo processing apparatus |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2001152888 | 2001-05-22 | ||
JP2001-152888 | 2001-05-22 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2002095975A1 true WO2002095975A1 (fr) | 2002-11-28 |
Family
ID=18997497
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2002/004860 WO2002095975A1 (fr) | 2001-05-22 | 2002-05-20 | Appareil de traitement d'echos |
Country Status (5)
Country | Link |
---|---|
US (1) | US6868158B2 (ja) |
EP (1) | EP1300963A4 (ja) |
JP (1) | JPWO2002095975A1 (ja) |
CN (1) | CN1320782C (ja) |
WO (1) | WO2002095975A1 (ja) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2004064366A1 (en) * | 2003-01-08 | 2004-07-29 | Philips Intellectual Property & Standards Gmbh | Non-linear acoustic echo canceller |
WO2005043772A1 (ja) * | 2003-11-04 | 2005-05-12 | Oki Electric Industry Co., Ltd. | エコーキャンセラ |
WO2005086525A1 (ja) * | 2004-03-05 | 2005-09-15 | Rohm Co., Ltd | フィルタ回路およびそれを利用した再生装置 |
JP2009246628A (ja) * | 2008-03-31 | 2009-10-22 | Yamaha Corp | 音響エコー除去装置 |
JP2015170867A (ja) * | 2014-03-04 | 2015-09-28 | 沖電気工業株式会社 | トーク状態検出器、トーク状態検出方法、エコーキャンセラ及びエコーサプレッサ |
JP2019020600A (ja) * | 2017-07-18 | 2019-02-07 | 富士通株式会社 | 評価プログラム、評価方法および評価装置 |
Families Citing this family (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7046794B2 (en) * | 2003-12-12 | 2006-05-16 | Motorola, Inc. | Double talk activity detector and method for an echo canceler circuit |
US7099458B2 (en) | 2003-12-12 | 2006-08-29 | Motorola, Inc. | Downlink activity and double talk probability detector and method for an echo canceler circuit |
EP1704748B1 (en) * | 2004-01-07 | 2008-04-30 | Koninklijke Philips Electronics N.V. | Audio system providing for filter coefficient copying |
KR100582728B1 (ko) * | 2004-01-30 | 2006-05-22 | 삼성전자주식회사 | 케이블 모뎀 및 그 케이블 모뎀의 주파수 대역별 필터링방법 |
JP4613705B2 (ja) * | 2005-06-15 | 2011-01-19 | 沖電気工業株式会社 | エコーキャンセラ |
JP4771532B2 (ja) * | 2006-02-01 | 2011-09-14 | オンセミコンダクター・トレーディング・リミテッド | エコー防止回路及びデジタル信号処理回路 |
US20080159551A1 (en) * | 2006-12-28 | 2008-07-03 | Texas Instruments Incorporated | System and Method for Acoustic Echo Removal (AER) |
US8077857B1 (en) * | 2007-03-14 | 2011-12-13 | Clearone Communications, Inc. | Portable speakerphone device with selective mixing |
US7912211B1 (en) | 2007-03-14 | 2011-03-22 | Clearone Communications, Inc. | Portable speakerphone device and subsystem |
US8654955B1 (en) | 2007-03-14 | 2014-02-18 | Clearone Communications, Inc. | Portable conferencing device with videoconferencing option |
US8406415B1 (en) | 2007-03-14 | 2013-03-26 | Clearone Communications, Inc. | Privacy modes in an open-air multi-port conferencing device |
GB0725111D0 (en) * | 2007-12-21 | 2008-01-30 | Wolfson Microelectronics Plc | Lower rate emulation |
PL216396B1 (pl) * | 2008-03-06 | 2014-03-31 | Politechnika Gdańska | Sposób i układ tłumienia echa akustycznego w terminalu VoIP |
US20120195423A1 (en) * | 2011-01-31 | 2012-08-02 | Empire Technology Development Llc | Speech quality enhancement in telecommunication system |
WO2012122698A1 (zh) * | 2011-03-14 | 2012-09-20 | 苏州全波通信技术有限公司 | 用于覆盖补点的同频转发器的回波消除系统及其方法 |
JP5175948B2 (ja) * | 2011-03-22 | 2013-04-03 | 株式会社東芝 | 映像表示装置、ノイズ除去方法および映像表示システム |
CN103428607A (zh) * | 2012-05-25 | 2013-12-04 | 华为技术有限公司 | 一种音频信号播放系统及电子设备 |
JPWO2015128960A1 (ja) * | 2014-02-26 | 2017-03-30 | 三菱電機株式会社 | 車載制御装置および車載制御方法 |
US10373608B2 (en) | 2015-10-22 | 2019-08-06 | Texas Instruments Incorporated | Time-based frequency tuning of analog-to-information feature extraction |
CN111345048A (zh) * | 2017-11-14 | 2020-06-26 | 日本电信电话株式会社 | 语音通信装置、语音通信方法、程序 |
TWI672690B (zh) * | 2018-03-21 | 2019-09-21 | 塞席爾商元鼎音訊股份有限公司 | 人工智慧語音互動之方法、電腦程式產品及其近端電子裝置 |
JP7195344B2 (ja) | 2018-07-27 | 2022-12-23 | ドルビー ラボラトリーズ ライセンシング コーポレイション | パーベイシブ・リステニングのための強制ギャップ挿入 |
US10636435B1 (en) * | 2018-12-22 | 2020-04-28 | Microsemi Semiconductor (U.S.) Inc. | Acoustic echo cancellation using low-frequency double talk detection |
CN113035219B (zh) * | 2019-12-09 | 2023-10-31 | 北京君正集成电路股份有限公司 | 一种消除回音提高音频质量的方法 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH01198155A (ja) * | 1988-02-03 | 1989-08-09 | Hitachi Ltd | 拡声電話機 |
JPH0766756A (ja) * | 1993-08-30 | 1995-03-10 | Kyocera Corp | 音響エコーキャンセラ |
JPH08256089A (ja) * | 1995-03-17 | 1996-10-01 | Toshiba Corp | エコー打消装置 |
JPH11289283A (ja) * | 1998-03-31 | 1999-10-19 | Sanyo Electric Co Ltd | エコー除去方法およびエコーキャンセラ |
JP2000101484A (ja) * | 1998-09-18 | 2000-04-07 | Fujitsu Ltd | エコーキャンセラ |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH06202669A (ja) | 1992-12-28 | 1994-07-22 | Toshiba Corp | 能動消音装置 |
JP2907091B2 (ja) | 1996-01-26 | 1999-06-21 | 日本電気株式会社 | 適応型雑音除去自動車電話装置 |
JP3139405B2 (ja) | 1997-02-28 | 2001-02-26 | 日本電気株式会社 | エコーキャンセラ |
-
2002
- 2002-05-20 CN CNB028017870A patent/CN1320782C/zh not_active Expired - Fee Related
- 2002-05-20 EP EP02771733A patent/EP1300963A4/en not_active Withdrawn
- 2002-05-20 US US10/333,224 patent/US6868158B2/en not_active Expired - Fee Related
- 2002-05-20 WO PCT/JP2002/004860 patent/WO2002095975A1/ja active Application Filing
- 2002-05-20 JP JP2002592318A patent/JPWO2002095975A1/ja not_active Abandoned
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH01198155A (ja) * | 1988-02-03 | 1989-08-09 | Hitachi Ltd | 拡声電話機 |
JPH0766756A (ja) * | 1993-08-30 | 1995-03-10 | Kyocera Corp | 音響エコーキャンセラ |
JPH08256089A (ja) * | 1995-03-17 | 1996-10-01 | Toshiba Corp | エコー打消装置 |
JPH11289283A (ja) * | 1998-03-31 | 1999-10-19 | Sanyo Electric Co Ltd | エコー除去方法およびエコーキャンセラ |
JP2000101484A (ja) * | 1998-09-18 | 2000-04-07 | Fujitsu Ltd | エコーキャンセラ |
Non-Patent Citations (3)
Title |
---|
HIROSHI RIKIMARU: "Chokakukei ni okeru joho shori no kozo", SURI KAGAKU, vol. 36, no. 1, 1 January 1998 (1998-01-01), pages 22 - 33, XP002955036 * |
See also references of EP1300963A4 * |
YURIKO TSUKAHARA, SHIGENOBU MINAMI: "Echo canceller-noise canceller heiyo no okeru kadai to taisaku", 5 September 1994, 1994-NEN THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS SHUKI TAIKAI KOEN RONBUNSHU, B-331, XP002955037 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006513627A (ja) * | 2003-01-08 | 2006-04-20 | コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ | 非線形音響エコーキャンセラ |
WO2004064366A1 (en) * | 2003-01-08 | 2004-07-29 | Philips Intellectual Property & Standards Gmbh | Non-linear acoustic echo canceller |
GB2422280B (en) * | 2003-11-04 | 2008-06-25 | Oki Electric Ind Co Ltd | Echo canceler |
GB2422280A (en) * | 2003-11-04 | 2006-07-19 | Oki Electric Ind Co Ltd | Echo canceler |
WO2005043772A1 (ja) * | 2003-11-04 | 2005-05-12 | Oki Electric Industry Co., Ltd. | エコーキャンセラ |
US7564964B2 (en) | 2003-11-04 | 2009-07-21 | Oki Electric Industry Co., Ltd. | Echo canceller |
WO2005086525A1 (ja) * | 2004-03-05 | 2005-09-15 | Rohm Co., Ltd | フィルタ回路およびそれを利用した再生装置 |
JPWO2005086525A1 (ja) * | 2004-03-05 | 2008-01-24 | ローム株式会社 | フィルタ回路およびそれを利用した再生装置 |
JP4658924B2 (ja) * | 2004-03-05 | 2011-03-23 | ローム株式会社 | フィルタ回路およびそれを利用した再生装置 |
JP2009246628A (ja) * | 2008-03-31 | 2009-10-22 | Yamaha Corp | 音響エコー除去装置 |
JP2015170867A (ja) * | 2014-03-04 | 2015-09-28 | 沖電気工業株式会社 | トーク状態検出器、トーク状態検出方法、エコーキャンセラ及びエコーサプレッサ |
JP2019020600A (ja) * | 2017-07-18 | 2019-02-07 | 富士通株式会社 | 評価プログラム、評価方法および評価装置 |
JP7143574B2 (ja) | 2017-07-18 | 2022-09-29 | 富士通株式会社 | 評価プログラム、評価方法および評価装置 |
Also Published As
Publication number | Publication date |
---|---|
CN1320782C (zh) | 2007-06-06 |
JPWO2002095975A1 (ja) | 2004-09-09 |
US6868158B2 (en) | 2005-03-15 |
CN1463507A (zh) | 2003-12-24 |
US20030156711A1 (en) | 2003-08-21 |
EP1300963A4 (en) | 2009-03-04 |
EP1300963A1 (en) | 2003-04-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2002095975A1 (fr) | Appareil de traitement d'echos | |
US7856097B2 (en) | Echo canceling apparatus, telephone set using the same, and echo canceling method | |
JP5036874B2 (ja) | エコー消去装置 | |
US8644496B2 (en) | Echo suppressor, echo suppressing method, and computer readable storage medium | |
US8472616B1 (en) | Self calibration of envelope-based acoustic echo cancellation | |
US7826799B2 (en) | System for calibrating a hands-free system | |
US20090238373A1 (en) | System and method for envelope-based acoustic echo cancellation | |
JP4678349B2 (ja) | 通話判定装置 | |
JP3406590B2 (ja) | 音声通信装置、およびエコー処理プロセッサ | |
US20090103743A1 (en) | Echo canceller | |
US6377679B1 (en) | Speakerphone | |
JP2004165888A (ja) | ハウリング処理装置 | |
US8208623B2 (en) | Echo processing method and device | |
WO2020203258A1 (ja) | エコー抑圧装置、エコー抑圧方法及びエコー抑圧プログラム | |
JP4887181B2 (ja) | エコー防止装置及びプログラム | |
JP2006279191A (ja) | エコーキャンセラ | |
JP5183506B2 (ja) | ハウリング防止装置 | |
WO2021210473A1 (ja) | エコー抑圧装置、エコー抑圧方法及びエコー抑圧プログラム | |
JP5346350B2 (ja) | 反響消去装置とその方法とプログラム | |
JP3404236B2 (ja) | 拡声通話装置 | |
JP4534529B2 (ja) | ハウリング抑圧方法及び装置 | |
JP2004228939A (ja) | エコー処理装置及び音声処理装置 | |
JP3357284B2 (ja) | ダブルトーク検出制御装置およびダブルトーク検出制御方法 | |
JPS6053490B2 (ja) | 適応型反響打消装置 | |
JP2004040161A (ja) | ハンズフリー携帯電話端末装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): CN JP US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR |
|
ENP | Entry into the national phase |
Ref country code: JP Ref document number: 2002 592318 Kind code of ref document: A Format of ref document f/p: F |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2002771733 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 10333224 Country of ref document: US Ref document number: 028017870 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWP | Wipo information: published in national office |
Ref document number: 2002771733 Country of ref document: EP |