CN106571147B - Method for suppressing acoustic echo of network telephone - Google Patents
Method for suppressing acoustic echo of network telephone Download PDFInfo
- Publication number
- CN106571147B CN106571147B CN201610996518.4A CN201610996518A CN106571147B CN 106571147 B CN106571147 B CN 106571147B CN 201610996518 A CN201610996518 A CN 201610996518A CN 106571147 B CN106571147 B CN 106571147B
- Authority
- CN
- China
- Prior art keywords
- echo
- signal
- far
- adaptive filter
- adaptive
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Processing of the speech or voice signal 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
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Processing of the speech or voice signal 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
- G10L21/0216—Noise filtering characterised by the method used for estimating noise
- G10L21/0232—Processing in the frequency domain
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/253—Telephone sets using digital voice transmission
- H04M1/2535—Telephone sets using digital voice transmission adapted for voice communication over an Internet Protocol [IP] network
-
- 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
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Processing of the speech or voice signal 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
Abstract
The invention discloses a method for suppressing acoustic echo of a network telephone, which comprises an adaptive filter, a double-end talkback detector, a nonlinear processor, a comfortable noise generator, a far-end signal x (n), an echo signal d (n), a near-end voice signal s (n), a near-end signal y (n) and a near-end output e (n), wherein an echo estimation signal d '(n) similar to a real echo signal d (n) is generated by utilizing an input far-end signal x (n) and the adaptive filter W (n) according to an adaptive calculation mode, and then the echo estimation signal d' (n) is subtracted from the near-end signal y (n) so as to generate a near-end output e (n); when the adaptive filter w (n) is exactly matched with the echo path h (n), the echo estimation signal d' (n) is an exact estimate of the echo signal d (n); the echo generated by the far-end signal x (n) is completely cancelled by the echo estimation signal d' (n), and has no influence on the near-end speech signal s (n); the technical scheme has the advantages of good inhibition effect, stable and reliable operation, good portability and low computation workload, and is suitable for embedded equipment.
Description
Technical Field
The invention relates to a method for inhibiting echo in a network telephone, in particular to a method for realizing acoustic echo inhibition by the network telephone, belonging to the technical field of embedded computer systems, network communication and media information processing.
Background
In daily life or office, the hand-free function of the network phone is widely used, when the network phone carries out hand-free call, the phone generates echo, which affects the normal call quality of a user, and echo suppression is a necessary function of the network phone.
Compared with the traditional telephone, the real-time voice transmission on the Internet has the fatal weakness that the voice quality is poor, the factors influencing the voice quality of the Internet are various, and one of the most key factors is the influence of echo; therefore, to improve the voice quality of the internet, the echo cancellation processing must be performed during the voice transmission process of the internet, that is, the internet phone gateway, as a voice access device of the internet, must have the echo cancellation function; because the voice transmission of the internet is a brand-new telecommunication service realized by adopting a packet switching technology, the transmitted voice signal needs to be subjected to a series of processing such as encoding, compression, packaging and the like, which not only causes larger delay of an echo path, but also causes larger delay jitter; therefore, the echo problem is particularly prominent in the voice transmission process of the internet, and has the following characteristics.
First, the echo source is complicated, and in the conventional telephone system, there is a so-called "line echo", which is mainly caused by the 2-4 line conversion in the system, and the mixer that completes the 2-4 line conversion causes "leakage" due to impedance matching, thereby causing "line echo", as can be seen from the connection manner of the internet IP telephone gateway, one end of which is connected to the public telephone network, and the other end of which is connected to the internet.
Although the circuit echo is generated in the public telephone network, it is also transmitted to the IP telephone gateway, which is one of the echo sources in the internet voice transmission, and the second echo source in the internet voice transmission is the so-called "acoustic echo"; the acoustic echo refers to the sound played by the loudspeaker and picked up by the microphone and sent back to the far end, so that a far-end talker can hear the sound of the talker, the acoustic echo is divided into direct echo and indirect echo, and the direct echo refers to the sound played by the loudspeaker and directly enters the microphone without any reflection; the echo delay is the shortest, and is related to the voice energy of a far-end speaker, the distance and the angle between the speaker and a microphone, the playing volume of the speaker, the pick-up sensitivity of the microphone and other factors; indirect echo is an echo set generated by a microphone after sound played by a loudspeaker is reflected once or for multiple times through different paths, and the return path of the echo is changed due to the change of surrounding objects, such as the walking of people, and the like, because the echo is characterized by multipath and time-varying, and in addition, background noise is also one of factors for generating the echo.
Secondly, the echo path has a large delay, and in the voice transmission in the internet, there are three delay sources: the method comprises compression delay, packet transmission delay and processing delay, wherein voice compression delay is the main delay for generating echo, for example, in the g.723.1 standard, the maximum delay for compressing a frame for 30ms is 37.5ms, packet transmission delay is also a very important source, tests show that the maximum transmission delay from end to end can reach more than 250ms, and processing delay refers to the encapsulation delay, the buffering delay and the like of a voice packet.
Thirdly, the echo path has large delay jitter, and in the voice transmission process of the internet, because the echo path, the voice compression delay, the packet transmission route and the like have a plurality of uncertain factors, the fluctuation range is large, and is generally between 20 ms and 50 ms.
With the development of echo cancellation technology, the focus of current echo cancellation research has been shifted to "acoustic echo" by the cancellation of "line echo".
Analyzing the mechanism of generation of acoustic echo, it is known that: the simplest control method of acoustic echo is to improve the surrounding environment of the speaker and minimize the reflection of the sound played by the speaker, because such environment can control the reflection and not make the speaker feel uncomfortable, and the improved environment can effectively suppress indirect acoustic echo, but cannot suppress direct acoustic echo.
Another way is to use an echo suppressor, which is an echo control method used earlier, which is a nonlinear echo cancellation that compares the received sound played by the loudspeaker with the level of the sound picked up by the current microphone by a simple comparator; if the former is above a certain threshold, then passage to the speaker is allowed and the microphone is turned off to prevent it from picking up the sound played by the speaker and causing a far-end echo; if the sound level picked up by the microphone is higher than a certain full threshold value, the loudspeaker is forbidden to achieve the purpose of eliminating echo; since echo suppression is a nonlinear echo control method, it can cause discontinuity of loudspeaker playing and affect echo cancellation effect, and with the appearance of high-performance echo cancellers, the echo suppressors have been used by few people.
Therefore, we can know that several current echo cancellation methods almost have some insurmountable problems, for line echoes and acoustic echoes, the line echo delay is small, within 10ms, the amplitude is small, and the processing is easy, while the acoustic echoes reach more than 100ms due to complex paths and large delay, and the echo amplitude is very large, and even exceeds the amplitude of the original signal, so that the echoes are difficult to be completely cancelled, the traditional echo cancellation method has high calculation complexity, the echo cancellation is not thorough, and the effect is not ideal.
Disclosure of Invention
In order to solve the above-mentioned defects in the prior art, the technical solution is to adopt a new adaptive calculation method, to find a corresponding block according to the echo delay estimation, to perform a block normalization least mean square adaptive calculation method, which can reduce the computation amount by one order of magnitude, and to perform block division, and to adopt frequency domain calculation to process the echo in the frequency domain, so as to eliminate the echo more thoroughly, and better suppress and eliminate the echo.
The purpose of the invention is realized by the following technical scheme:
a method for network phone acoustic echo suppression, characterized by: the method comprises the following steps of generating a far-end signal x (n), an echo signal d (n), a near-end voice signal s (n), a near-end signal y (n) and a near-end output e (n), generating an echo estimation signal d '(n) approximate to a real echo signal d (n) by utilizing an input far-end signal x (n) and an adaptive filter W (n) according to an adaptive calculation mode, and then subtracting the echo estimation signal d' (n) from the near-end signal y (n) to further generate a near-end output e (n); when the adaptive filter w (n) is exactly matched with the echo path h (n), the echo estimation signal d' (n) is an exact estimate of the echo signal d (n); the echo generated by the far-end signal x (n) is completely cancelled by the echo estimation signal d' (n), and has no influence on the near-end speech signal s (n).
Further, the far-end signal x (n): is a signal transmitted from a remote end to the remote end through a network.
Further, the echo signal d (n): is the echo generated at the far end by the far end signal.
Further, the near-end speech signal s (n): is a voice or noise signal input from a microphone at the near end.
Further, the near-end signal y (n): the signal of the near-end network afferent system, y (n) ═ s (n) + d (n).
Further, the near-end output e (n): the echo signal d (n) in the near-end signal y (n) is the cancellation output, i.e. y (n) -d' (n).
Furthermore, the adaptive filter is used for generating an echo estimation signal d '(n), and the echo intensity is greatly reduced after the true echo d (n) is subtracted from the echo estimation signal d' (n); in the adaptive echo cancellation process, the adaptive filter adopts a transverse finite impulse response filter.
Furthermore, the adaptive calculation mode is to automatically adjust the coefficient of the adaptive filter according to a fixed criterion so as to achieve the purpose of optimal filtering.
Furthermore, the double-end talkback detector controls the adjustment mode of the adaptive filter according to the double-end talkback state on the basis of double-end talkback detection by taking a near-end speech signal s (n) as interference of adaptive calculation under the condition of double-end talkback.
Furthermore, the non-linear processor, because the adaptive filter cannot completely eliminate the echo, the residual echo is subjected to non-linear processing by using an echo suppression technology, so that the power of the residual echo is further reduced.
Furthermore, when the comfort noise generator performs nonlinear processing on the residual echo by using an echo suppression technology, the speech path is disconnected, so that complete silence is caused to the local user or the local user is mistaken for the line interruption and quits the call, and therefore, the occurrence of the situation is avoided, and noise similar to the far-end background noise needs to be provided for the local user, and the noise is provided by the comfort noise generator.
The invention has the following advantages and beneficial effects:
the method for suppressing the acoustic echo of the network phone, which is provided by the technical scheme, is a new algorithm, a corresponding block is found according to the echo time delay estimation, a block normalization least mean square self-adaptive algorithm is carried out, the operation amount can be reduced by one order of magnitude, and simultaneously, after the blocks are adopted, the echo is processed in a frequency domain by adopting a frequency domain algorithm, and the echo is eliminated; the acoustic echo suppression method has the advantages of good effect, stable and reliable operation, good portability and low computation amount, and is suitable for being realized by embedded equipment.
Drawings
FIG. 1 is a schematic diagram of a method for acoustic echo suppression for a network phone;
fig. 2 is a schematic diagram of a near-end output filter calculation formula according to the present embodiment;
fig. 3 is a schematic diagram of an iterative calculation formula of the adaptive filter according to the present embodiment;
Detailed Description
The technical scheme of the invention is detailed in the following with the accompanying drawings of the specification:
as shown in fig. 1, a method for acoustic echo suppression of a network phone is characterized by: the method comprises the following steps of generating a far-end signal x (n), an echo signal d (n), a near-end voice signal s (n), a near-end signal y (n) and a near-end output e (n), generating an echo estimation signal d '(n) approximate to a real echo signal d (n) by utilizing an input far-end signal x (n) and an adaptive filter W (n) according to an adaptive calculation mode, and then subtracting the echo estimation signal d' (n) from the near-end signal y (n) to further generate a near-end output e (n); when the adaptive filter w (n) is exactly matched with the echo path h (n), the echo estimation signal d' (n) is an exact estimate of the echo signal d (n); the echo generated by the far-end signal x (n) is completely cancelled by the echo estimation signal d' (n), and has no influence on the near-end speech signal s (n).
Further, in the embodiment, setting 1 indicates spoken speech, 0 indicates silent speech, i.e. silence or very weak sound, far-end signal x (n) and near-end signal y (n).
The power spectrum of the remote signal obtained by windowing the input remote signal x is represented by Xw (P, Q) at a set time interval P, i.e., P is 1 or 2 … P, and the frequency bands Q and Q are 1 or 2 … Q, and a threshold value Xw (P, Q) _ th is set for the power spectrum in each frequency band.
When Xw (p, q) > is Xw (p, q) _ th, Xw (p, q) > is 1;
when Xw (p, q) < Xw (p, q) _ th, Xw (p, q) ═ 0;
similarly, for the near-end signal y (n), the windowed signal power spectrum Yw (p, q) and the threshold Yw (p, q) _ th,
when Yw (p, q) > is Yw (p, q) _ th, Yw (p, q) is 1;
when Yw (p, q) < Yw (p, q) _ th, Yw (p, q) ═ 0;
further, the fourier-transformed frequency domain power spectrum is divided into 32 sub-bands, and the value of each fixed sub-band Xw (p, q) can be represented by 1 bit, and 32 bits in total, and is represented by one 32-bit data type.
1) The segmented far-end signal is set to be one block every 8ms, so 128ms is divided into 16 blocks, and the signal power spectrum Xw (p, q) of each block can be represented by 32-bit type data, stored by 16 32-bit arrays and set as records [16 ].
2) When the near-end signal power spectrum Yw (p, q) is processed each time, the difference between the near-end signal power spectrum Yw and the near-end signal power spectrum Yw (p, q) and the 16 far-end signal power spectrums is respectively compared, and the most reliable reference signal is represented by the smallest difference.
3) When the spectrum of the near-end signal power spectrum Yw (p, q) is closest to the record [ k ], the time delay of echo is represented as 8ms k milliseconds, wherein k is an integer from 0 to 15.
After the accurate echo delay is obtained through calculation, historical data of the delay can be obtained, and a self-adaptive calculation mode is carried out, wherein the historical data is one of 16 blocks, so that the calculation amount is reduced to 1/16, the resource consumption is reduced, and the hardware cost is reduced.
As shown in fig. 2, the present technical solution uses a least mean square algorithm normalization calculation mode, a filter calculation formula.
As shown in fig. 3, the least mean square algorithm normalizes the calculation mode to iterate the calculation formula, and the range is selected to be 0< μ <2, and 0< γ < 2.
In summary, the method for suppressing acoustic echo of a network phone provided by the present invention is only a specific application example of the present invention, and during the actual application process, alternative device devices and actual input parameters may be selected as appropriate according to specific situations, but the scope of the present invention is not limited at all.
Claims (9)
1. A method for network phone acoustic echo suppression, characterized by: the system comprises an adaptive filter, a double-end talkback detector, a nonlinear processor, a comfort noise generator, a far-end signal x (n), an echo signal d (n), a near-end voice signal s (n), a near-end signal y (n) and a near-end output e (n), wherein a frequency domain power spectrum subjected to Fourier transform is divided into 32 sub-bands according to an adaptive calculation mode, the value of each fixed sub-band is represented by one bit, the total number of the fixed sub-bands is 32 bits, and the fixed sub-bands are represented by one 32-bit data type; setting the segmented far-end signal as one block every 8ms, so that 128ms is divided into 16 blocks, wherein a signal power spectrum Xw (p, q) of each block is represented by 32-bit type data, is stored by 16 32-bit arrays and is set as a record [16], and p is a time interval; q is a frequency band; when the near-end signal power spectrum Yw (p, q) is processed each time, the difference is respectively compared with the 16 far-end signal frequency spectrums, and the most reliable reference signal is represented by the minimum difference; when the frequency spectrum of the near-end signal power spectrum Yw (p, q) is closest to the record [ k ], the time delay of echo is represented as 8ms k milliseconds, wherein k is an integer from 0 to 15; after obtaining the accurate echo delay through calculation, historical data of the delay can be obtained, and a self-adaptive calculation mode is carried out, namely one of 16 blocks, so that the calculation amount is reduced to 1/16, an echo estimation signal d '(n) similar to a real echo signal d (n) is generated by utilizing an input far-end signal x (n) and a self-adaptive filter W (n), and then the echo estimation signal d' (n) is subtracted from a near-end signal y (n), so that a near-end output e (n) is generated; when the adaptive filter w (n) is exactly matched with the echo path h (n), the echo estimation signal d' (n) is an exact estimate of the echo signal d (n); the echo generated by the far-end signal x (n) is completely cancelled by the echo estimation signal d' (n), and has no influence on the near-end speech signal s (n); when the comfortable noise generator carries out nonlinear processing on the residual echo by using an echo suppression technology, the speech path is disconnected, and noise similar to the far-end background noise is provided for the local user, and the noise is provided by the comfortable noise generator.
2. The method for network phone acoustic echo suppression according to claim 1, wherein: the far-end signal x (n): is a signal transmitted from a remote end to the remote end through a network.
3. The method for network phone acoustic echo suppression according to claim 1, wherein: the echo signal d (n): is the echo generated at the far end by the far end signal.
4. The method for network phone acoustic echo suppression according to claim 1, wherein: the near-end speech signal s (n): is a voice or noise signal input from a microphone at the near end.
5. The method for network phone acoustic echo suppression according to claim 1, wherein: the near-end signal y (n): the signal of the near-end network afferent system, y (n) ═ s (n) + d (n).
6. The method for network phone acoustic echo suppression according to claim 1, wherein: the near-end output e (n): the echo signal d (n) in the near-end signal y (n) is the cancellation output, i.e. y (n) -d' (n).
7. The method for network phone acoustic echo suppression according to claim 1, wherein: the adaptive filter is used for generating an echo estimation signal d '(n), and the echo intensity is greatly reduced after the real echo d (n) is subtracted from the echo estimation signal d' (n); in the adaptive echo cancellation process, the adaptive filter adopts a transverse finite impulse response filter.
8. The method for network phone acoustic echo suppression according to claim 1, wherein: the double-end talkback detector takes a near-end voice signal s (n) as interference of self-adaptive calculation under the condition of double-end talkback, and controls an adjusting mode of a self-adaptive filter according to a double-end talkback state on the basis of double-end talkback detection.
9. The method for network phone acoustic echo suppression according to claim 1, wherein: because the adaptive filter can not completely eliminate the echo, the residual echo is subjected to nonlinear processing by adopting an echo suppression technology, and the power of the residual echo is further reduced.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610996518.4A CN106571147B (en) | 2016-11-13 | 2016-11-13 | Method for suppressing acoustic echo of network telephone |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610996518.4A CN106571147B (en) | 2016-11-13 | 2016-11-13 | Method for suppressing acoustic echo of network telephone |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106571147A CN106571147A (en) | 2017-04-19 |
CN106571147B true CN106571147B (en) | 2021-05-28 |
Family
ID=58542553
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610996518.4A Active CN106571147B (en) | 2016-11-13 | 2016-11-13 | Method for suppressing acoustic echo of network telephone |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106571147B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108831496B (en) * | 2018-05-22 | 2019-12-10 | 出门问问信息科技有限公司 | echo suppression method and device, storage medium and electronic equipment |
CN109658950B (en) * | 2018-11-13 | 2022-11-11 | 南京南大电子智慧型服务机器人研究院有限公司 | Mixed frequency domain self-adaptive algorithm |
CN109448748B (en) * | 2018-12-17 | 2021-08-03 | 嘉楠明芯(北京)科技有限公司 | Filtering method, device, equipment and medium for echo cancellation |
CN109547655A (en) * | 2018-12-30 | 2019-03-29 | 广东大仓机器人科技有限公司 | A kind of method of the echo cancellation process of voice-over-net call |
CN110310653A (en) * | 2019-07-09 | 2019-10-08 | 杭州国芯科技股份有限公司 | A kind of echo cancel method |
CN112502784B (en) * | 2020-12-11 | 2022-11-15 | 国网智能科技股份有限公司 | Tunnel bidirectional visual voice intercom method, robot and system |
CN114141224B (en) * | 2021-11-30 | 2023-06-09 | 北京百度网讯科技有限公司 | Signal processing method and device, electronic equipment and computer readable medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1499809A (en) * | 2002-11-04 | 2004-05-26 | 华为技术有限公司 | Echo eliminator and method for evaluating echo |
CN1668058A (en) * | 2005-02-21 | 2005-09-14 | 南望信息产业集团有限公司 | Recursive least square difference based subband echo canceller |
CN102625006A (en) * | 2011-01-31 | 2012-08-01 | 深圳三石科技有限公司 | Method and system for synchronization and alignment of echo cancellation data and audio communication equipment |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101185820B1 (en) * | 2004-10-13 | 2012-10-02 | 코닌클리케 필립스 일렉트로닉스 엔.브이. | Echo cancellation |
CN1842110B (en) * | 2005-03-28 | 2010-04-28 | 华为技术有限公司 | Echo eliminating device and method |
WO2007123048A1 (en) * | 2006-04-20 | 2007-11-01 | Nec Corporation | Adaptive array control device, method, and program, and its applied adaptive array processing device, method, and program |
CN101106405A (en) * | 2006-07-12 | 2008-01-16 | 北京大学深圳研究生院 | Method for eliminating echo in echo eliminator and its dual end communication detection system |
CN101179294B (en) * | 2006-11-09 | 2012-07-04 | 黄山好视达通信技术有限公司 | Self-adaptive echo eliminator and echo eliminating method thereof |
JP5555987B2 (en) * | 2008-07-11 | 2014-07-23 | 富士通株式会社 | Noise suppression device, mobile phone, noise suppression method, and computer program |
GB2473267A (en) * | 2009-09-07 | 2011-03-09 | Nokia Corp | Processing audio signals to reduce noise |
CN101888455B (en) * | 2010-04-09 | 2013-07-03 | 熔点网讯(北京)科技有限公司 | Self-adaptive echo counteracting method for frequency domain |
CN105791611B (en) * | 2016-02-22 | 2020-07-07 | 腾讯科技(深圳)有限公司 | Echo cancellation method, device, terminal and storage medium |
-
2016
- 2016-11-13 CN CN201610996518.4A patent/CN106571147B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1499809A (en) * | 2002-11-04 | 2004-05-26 | 华为技术有限公司 | Echo eliminator and method for evaluating echo |
CN1668058A (en) * | 2005-02-21 | 2005-09-14 | 南望信息产业集团有限公司 | Recursive least square difference based subband echo canceller |
CN102625006A (en) * | 2011-01-31 | 2012-08-01 | 深圳三石科技有限公司 | Method and system for synchronization and alignment of echo cancellation data and audio communication equipment |
Also Published As
Publication number | Publication date |
---|---|
CN106571147A (en) | 2017-04-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106571147B (en) | Method for suppressing acoustic echo of network telephone | |
US10074380B2 (en) | System and method for performing speech enhancement using a deep neural network-based signal | |
US6526139B1 (en) | Consolidated noise injection in a voice processing system | |
KR100989266B1 (en) | Double talk detection method based on spectral acoustic properties | |
US8290141B2 (en) | Techniques for comfort noise generation in a communication system | |
US5598468A (en) | Method and apparatus for echo removal in a communication system | |
US9443528B2 (en) | Method and device for eliminating echoes | |
US10192567B1 (en) | Echo cancellation and suppression in electronic device | |
US9800734B2 (en) | Echo cancellation | |
CN105825864B (en) | Both-end based on zero-crossing rate index is spoken detection and echo cancel method | |
US8861713B2 (en) | Clipping based on cepstral distance for acoustic echo canceller | |
US9699554B1 (en) | Adaptive signal equalization | |
KR20130108063A (en) | Multi-microphone robust noise suppression | |
WO2012142270A1 (en) | Systems, methods, apparatus, and computer readable media for equalization | |
CN109273019B (en) | Method for double-talk detection for echo suppression and echo suppression | |
CN110995951B (en) | Echo cancellation method, device and system based on double-end sounding detection | |
US20220301577A1 (en) | Echo cancellation method and apparatus | |
US20070092074A1 (en) | Echo canceller | |
CN108076239B (en) | Method for improving IP telephone echo | |
EP3469591B1 (en) | Echo estimation and management with adaptation of sparse prediction filter set | |
US20080152156A1 (en) | Robust Method of Echo Suppressor | |
Chandra et al. | Performance Evaluation of Adaptive Algorithms for Monophonic Acoustic Echo Cancellation: A Technical | |
CN102739287A (en) | Echo cancelling method based on self-adaptive time delay estimation | |
Helwani et al. | A single-channel MVDR filter for acoustic echo suppression | |
Fukui et al. | Acoustic echo canceller software for VoIP hands-free application on smartphone and tablet devices |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |