CN206489876U - Self-alignment far field interactive voice equipment - Google Patents
Self-alignment far field interactive voice equipment Download PDFInfo
- Publication number
- CN206489876U CN206489876U CN201621197888.3U CN201621197888U CN206489876U CN 206489876 U CN206489876 U CN 206489876U CN 201621197888 U CN201621197888 U CN 201621197888U CN 206489876 U CN206489876 U CN 206489876U
- Authority
- CN
- China
- Prior art keywords
- far field
- voice
- microphone
- signal
- calibration
- 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
Landscapes
- Circuit For Audible Band Transducer (AREA)
Abstract
The utility model provides a kind of self-alignment far field interactive voice equipment, and microphone array collection far field voice signal, self calibration and voice recognition terminal are calibrated to the error as caused by microphone channel gain and signal propagation attenuation, obtain strengthening voice output.The utility model utilizes the mode that channel gain is calibrated, it is ensured that each microphone channel relative gain is consistent, improves rear end GSC performance;Using voice segments as calibration sound source, subsequent treatment ensures that speech damage is smaller, is conducive to the elimination of echo signal and the reservation of noise interferences;It can ensure that under microphone damaged condition, equipment remains able to normal work with automatic decision microphone works state, improve the robustness of equipment.
Description
Technical field
The utility model relates generally to voice processing technology field, more particularly to self-alignment far field interactive voice is set
It is standby.
Background technology
Microphone array technology receives the extensive concern of researcher, particularly recent years in recent decades,
Driven by artificial intelligence prosperity and development, far field speech recognition technology obtains bigger concern, to being used as far field
The microphone array technology of speech recognition front-ends proposes bigger requirement.Currently for the main method bag of far field speech recognition
Include MVDR (Minimum Variance DistortionlessResponse, minimum variance distortionless response) and multichannel dimension
Nanofiltration wave method, no matter any method is all to have carried out far field hypothesis, it is generally recognized that passage reaches the amplitude of each microphone
Unanimously, simply sound wave arrival time is different, but is typically different passage microphone gain difference in practice, simultaneously because range attenuation
Amplitude difference is caused etc. factor, it is therefore desirable to microphone gain and signal amplitude decay are compensated, algorithm is just can guarantee that
Energy.Above-mentioned algorithm is required for the accurate location of clear and definite microphone simultaneously, and a large amount of methods are also based on the accurate position of microphone with research
Install meter, and all microphones can normal work, under extreme case, once microphone can not normal recording work, will
Algorithm performance is caused drastically to decline, in actual product application, microphone position is commonly known, but in use very
It is difficult to ensure that problem occurs for card some microphone therein, it is impossible to rejecting problem microphone is needed under normal recording, this situation, now
Original microphone position information needs to update, to ensure that array algorithm remains able to normal work, now needs calibration to determine
New microphone array and its relevant information that normal microphone array is constituted.
Calibration of the prior art to microphone signal, is generally completed using large-scale calibrator (-ter) unit in special laboratory,
Waste time and energy very much, be not particularly suited for consumer electronics.For example, prior art one (CN200810213962, a kind of microphone array
And the method and module of microphone array calibration) using the ambient noise of quiet period as calibration sound source, calculate different passages
Gain, adjusts different passages to identical gain.The technical scheme is used as school using the relatively low ambient noise of the coherence of quiet period
Quasi- sound source, the signal to noise ratio that microphone receives signal is relatively low, and calibration error is larger, while only considering the gain of microphone itself, does not have
There is the range error for considering that propagation attenuation etc. is caused, especially under the situation of far field, between the voice signal of each microphone still
In the presence of error by a relatively large margin, it is not suitable for microphone array signals processing.
" the Robust speech recognition using beamforming with of prior art two
Adaptivemicrophone gains and multichannel noise reduction " use voice segment signal conduct
Sound source is calibrated, the gain of different passages is calculated.On the one hand the technical scheme does not pick out the influence of noise in gain calculating, makes an uproar
Acoustic gain calculate it is inaccurate, on the other hand for extreme case, i.e., microphone damage can not normal work in the case of, for wheat
Gram wind array does not propose solution.
Utility model content
(1) technical problem to be solved
The utility model provides a kind of self-alignment far field interactive voice equipment.
(2) technical scheme
The utility model provides a kind of self-alignment far field interactive voice equipment, including:Microphone array and self calibration
And voice recognition terminal, the microphone array and the self calibration and voice recognition terminal electric connection;The microphone array
Row collection far field voice signal, and by the far field transmitting voice signal to the self calibration and voice recognition terminal, wherein, institute
State far field voice signal and include the error as caused by microphone channel gain and signal propagation attenuation;The self calibration and voice are known
Error described in other terminal-pair is calibrated, and obtains strengthening voice output.
Preferably, interactive voice equipment in far field also includes control end, and the control end is mobile phone.
Preferably, the self calibration and voice recognition terminal are computer or server.
Preferably, the self calibration and voice recognition terminal include:Self calibration module and general matching law, the wheat
Gram wind array connection self calibration module, the self calibration module connects the general matching law, wherein, the self-correcting quasi-mode
Block is calibrated to the error, the far field transmitting voice signal after calibration to the general matching law;By the broad sense
Valve canceller is handled the far field voice signal after calibration, obtains strengthening voice output.
Preferably, the self calibration module includes:Voice Activity Detection module, relative gain computing module and gain calibration
Module;Wherein, the Voice Activity Detection module chooses a microphone of microphone array as reference channel, using described
Reference channel extracts the voice segments in the voice signal of far field;The relative gain computing module connects the voice activity detection mould
Block, relative gain of other microphones relative to the reference channel in microphone array is calculated using institute's speech segment;Institute
State gain calibration module and connect the relative gain computing module, according to the relative gain, by the voice segments of other microphones
The gain level of reference channel is adjusted to, the far field voice signal after being calibrated.
Preferably, the general matching law includes:Fixed beam forms module, blocking matrix module, adaptively made an uproar
Sound cancellation module;Wherein, the far field voice signal after calibration respectively enters fixed beam formation module and blocking matrix module;Gu
Determine Wave beam forming module to handle the far field voice signal after calibration, generate speech reference signal;Blocking matrix module pair
Far field voice signal after calibration is handled, and generates noise reference signal, speech reference signal by adaptive noise with supporting
The noise reference signal for the module that disappears asks poor, obtains strengthening voice output.
Preferably, the general matching law also includes:Divider and voice segments determination module, the divider are obtained
The speech reference signal and the ratio of the noise reference signal, and the ratio is fed back into institute's speech segment judgement mould
Block;Institute's speech segment determination module extracts the voice segments in the voice signal of far field according to the ratio.
Preferably, the self calibration module also includes:Fixed beam formation parametric calibration module;The fixed beam is formed
Parametric calibration module connects the relative gain computing module, according to the relative gain detect can not normal work Mike
Wind, the fixed beam formation parameter after being calibrated;General matching law utilizes the fixed beam formation parameter pair after calibration
Far field voice signal after calibration is handled, and obtains strengthening voice output.
Preferably, interactive voice equipment in far field also includes:Control end, the control end controls the fixed beam formation ginseng
The unlatching of number calibration module.
Preferably, the microphone of the microphone array is provided with button, passes through the fixed beam as described in key control
Form the unlatching of parametric calibration module.
(3) beneficial effect
It can be seen from the above technical proposal that self-alignment far field interactive voice equipment of the present utility model has and following had
Beneficial effect:
(1) mode calibrated using channel gain, it is ensured that each microphone channel relative gain is consistent so that enter GSC
Each passage expectation target signal amplitude it is consistent, improve rear end GSC performance;
(2) not only contribute to fixed beam as calibration sound source using voice segments and formed, simultaneously for blocking matrix mould
Block, when voice segments amplitude is consistent, the voice segments remained after blocking matrix resume module are less, and subsequent treatment ensures language
Sound damage is smaller, is conducive to echo signal to eliminate, and retains noise or interference signal;
(3) it can be ensure that with automatic decision microphone works state under microphone damaged condition, equipment still can
Enough normal works, improve the robustness of equipment.
Brief description of the drawings
In order to be more fully understood from the utility model and its advantage, referring now to the following description with reference to accompanying drawing, wherein:
Fig. 1 is the overall structure figure of the self-alignment far field interactive voice equipment of the utility model embodiment;
Fig. 2 is the structural representation of the self-alignment far field interactive voice equipment of the utility model embodiment;
Fig. 3 is the structural representation of the self-alignment far field interactive voice equipment of another embodiment of the utility model;
Fig. 4 is the structural representation of the self-alignment far field interactive voice equipment of the another embodiment of the utility model;
Fig. 5 is the overall structure figure of the self-alignment far field interactive voice equipment of the utility model another embodiment;
Fig. 6 is the flow chart of far field voice method for self-calibrating.
【Symbol description】
10- microphone arrays;
20- self calibrations and voice recognition terminal;
21- Voice Activity Detection modules;22- relative gain computing modules;23- gain calibration modules;24- fixed beam shapes
Into parametric calibration module;25- signal-to-noise ratio computation modules;
31- fixed beams formation module;32- blocking matrix modules;33- adaptive noise cancellation modules;34- strengthens voice
Output;
35- voice segments determination modules;
40- control ends.
Embodiment
According to reference to accompanying drawing to the described in detail below of the utility model exemplary embodiment, other sides of the present utility model
Face, advantage and prominent features will become obvious for those skilled in the art.
In the utility model, term " comprising " and " containing " and its derivative mean including and it is unrestricted;Term "or"
Inclusive, mean and/or.
In this manual, following various embodiments for being used to describe the utility model principle are explanation, should not be with
Any mode is construed to limit the scope of utility model.The comprehensive understanding described below that is used to help referring to the drawings is by claim
And its exemplary embodiment of the present utility model that equivalent is limited.It is described below to help to understand including a variety of details,
But these details are considered as what is be merely exemplary.Therefore, it will be appreciated by those of ordinary skill in the art that without departing substantially from this practicality
In the case of new scope and spirit, embodiment described herein can be made various changes and modifications.In addition, in order to
Understand and for purpose of brevity, eliminate the description of known function and structure.In addition, through accompanying drawing, same reference numbers are used for similar
Function and operation.
The calibration of voice signal, target includes microphone own gain difference and the amplitude fading caused by propagation attenuation etc.
The error of composition, these errors will cause far field speech enhan-cement hydraulic performance decline.Under extreme case, the damage of microphone will cause far
Field speech enhan-cement is entirely ineffective.Therefore, error of the utility model not only to microphone is calibrated, while also to microphone array
Row are calibrated so that far field speech enhan-cement remains able to normal work in the case where some microphone damages situation.
A kind of far field interactive voice equipment of the utility model embodiment, referring to Fig. 1, it carries self-calibration function, and this is remote
Field interactive voice equipment includes:Microphone array 10 and self calibration and voice recognition terminal 20, microphone array 10 and self calibration
It is electrically connected with voice recognition terminal 20.
Self calibration and voice recognition terminal 20 include:Self calibration module and general matching law (GSC, General
Sidelobe Cancellation), the connection self calibration module of microphone array 10, self calibration module connection GSC, wherein,
Microphone array 10 gathers far field voice signal, and by far field transmitting voice signal to self calibration module, self calibration
Module is calibrated to far field voice signal error as caused by microphone channel gain and signal propagation attenuation, and is detected
The working condition of microphone array, calibration fixed beam formation (FBF, Fixed Beamforming) parameter, the far field after calibration
Transmitting voice signal is handled the far field voice signal after calibration using the FBF parameters after calibration, increased to GSC, GSC
Strong voice output 34.
The closed array that microphone array 10 is made up of multiple microphones, in Fig. 1, microphone array include 4
Microphone (micl, mic2, mic3, mic4), but the quantity of microphone can be arranged as required to, and be greater than being equal to 5, wheat
The shape of gram wind array can be rectangle, circle, ellipse etc..Each microphone of microphone array is used to gather original remote
Voice signal, due to the channel gain difference of each microphone so that far field voice signal passes through microphone array 10
Afterwards, there is amplitude difference between the far field voice signal of each microphone output;It is by source of sound additionally, due to far field voice signal
Microphone array 10 is traveled to, because far field voice signal has decay in communication process, the remote of each microphone is reached
Can also there is the amplitude difference as caused by propagation attenuation between the voice signal of field.Meanwhile, microphone array 10 also likely to be present damage
Bad microphone, microphone can not normal work situation, if can not in time detect and update FBF parameters, will also influence
The effect of speech enhan-cement.The above-mentioned amplitude difference as caused by microphone channel gain and signal propagation attenuation and FBF parameters will
Calibrated by self calibration module.
Self calibration and voice recognition terminal 20 can have the equipment of data-handling capacity using computer, server etc..
Referring to Fig. 2, self calibration module includes:Voice activity detection (VAD) module 21, relative gain computing module 22, gain calibration mould
Block 23 and fixed beam formation parametric calibration module 24.
Wherein, VAD module 21 chooses a microphone channel as reference channel from microphone array 10, utilizes the ginseng
Examine passage to detect far field voice signal, extract the voice segments in the voice signal of far field, calibration sound is used as using voice segments
Source carries out follow-up calibration.
Relative gain computing module 22 using voice segments calculate microphone array 10 in other microphone channels relative to
The relative gain of reference channel.
The voice segments of other microphone channels are adjusted to the increasing of reference channel by gain calibration module 23 according to relative gain
Beneficial level, so that amplitude difference caused by microphone channel gain and signal propagation attenuation is eliminated, the far field language after being calibrated
Message number.
FBF parametric calibrations module 24 according to relative gain, detection can not normal work microphone, FBF ginsengs are updated accordingly
Number, the FBF parameters after being calibrated.
Wherein, VAD module 21 can choose any one microphone in microphone array 10 as reference channel, below with
Exemplified by 4 microphone channels, illustrate how to calculate relative gain and calibrate far field voice signal.
The voice segments current frame signal of 4 microphone channels is respectively s1(t), s2(t), s3(t), s4(t), its frequency domain is believed
Number be respectively S1(j ω), S2(j ω), S3(j ω), S4(j ω), the 1st microphone channel of selection is used as reference channel.
The relative gain of other 3 passages is calculated by following wave filters, wave filter is as follows:
Wherein, μ is a limit coefficient, and which limit the minimum value of relative gain, it is ensured that relative gain is not excessive;α is one
Rank smoothing factor, n (n=2,3,4 ...) is signal frame number, Pxx、PxyThe respectively auto-power spectrum and crosspower spectrum of frame signal;Represent second order norm, PnnFor noise power spectrum, it may be referred to existing a variety of methods and calculate acquisition, wherein Pxx
(j ω, 1), PxyThe power spectrum and cross-power that the initial value of (j ω, 1) can be set as the first frame voice are general.
Therefore the far field voice signal of each microphone channel after calibration is:
Although above by taking 4 microphone channels as an example, illustrating how to calculate relative gain and calibrating far field voice signal,
Obvious aforesaid way is equally applicable to the microphone array of 2,3 or more than or equal to 5 microphone compositions.
FBF parametric calibrations module 24 according to the relative gain of each passage judge whether can not normal work Mike
Wind, is specifically included:
If the relative gain of each microphone channel is approached, and is all higher than a relative gain threshold value, then judge with reference to logical
The corresponding microphone in road for can not normal work microphone;
If the relative gain of some microphone channel differs larger with the relative gain of other microphone channels, and the wheat
The relative gain of gram wind passage is less than a relative gain threshold value, then judges that the corresponding microphone of the microphone channel can not normal work
Make;
If the relative gain of each microphone channel is approached, and a respectively less than relative gain threshold value, then all wheats are judged
Gram equal normal work of wind.
FBF parametric calibrations module 24 updates FBF parameters, including microphone position information, Mike according to above-mentioned judged result
Wind array weight and other information related to array are (if it is determined that the equal normal work of all microphones, then need not update FBF
Parameter), and the FBF parameters after calibration are sent to GSC.
In the utility model, the working condition of microphone can be not only detected using relative gain, while can use
Other judgment modes, such as energy method, or a variety of methods joint judges, so as to improve the accuracy of detection.
GSC use standard general matching law, including fixed beam formation (FBF) module 31, blocking matrix (BM,
Block Matrix) module 32, adaptive noise cancellation module 33.
Far field voice signal after calibration respectively enters FBF modules 31 and BM modules 32, after FBF modules 31 are using calibration
FBF parameters are handled the far field voice signal after calibration, generate speech reference signal, the far field after the 32 pairs of calibrations of BM modules
Voice signal is handled, and generates noise reference signal, speech reference signal and the noise Jing Guo adaptive noise cancellation module
Reference signal asks poor, obtains strengthening voice output 34.
Wherein, GSC of the present utility model also includes a divider and voice segments determination module 35, and language is obtained using divider
The ratio of sound reference signal and noise reference signal, and ratio is fed back into voice segments determination module 35, voice segments determination module
35 extract the voice segments in the voice signal of far field according to the ratio, specifically, and the section is thought when the ratio is less than a threshold value
Signal is noise or interference, and it is voice segments that the segment signal is thought during more than the threshold value.
As can be seen here, interactive voice equipment in far field of the present utility model, the mode calibrated using channel gain, it is ensured that each
Microphone channel relative gain is consistent so that each passage expectation target signal amplitude for entering GSC is consistent, improves rear end
GSC performance.Meanwhile, using voice segments as calibration sound source, not only contribute to fixed beam and formed, simultaneously for blocking matrix
Module, when voice segments amplitude is consistent, the voice segments remained after blocking matrix resume module are less, and subsequent treatment ensures language
Sound damage is smaller, is conducive to echo signal to eliminate, and retains noise or interference signal.And can be with automatic decision microphone
Working condition, ensure that under microphone damaged condition, equipment remains able to normal work, improve the robustness of equipment.
The far field interactive voice equipment of another embodiment of the utility model, referring to Fig. 3, its far field with above-described embodiment
The 26S Proteasome Structure and Function of interactive voice equipment is essentially identical, and difference is, the far field voice signal after calibration is only transmitted to BM
Module 32 and be not transmitted to FBF modules 31, microphone array 10 gather far field voice signal be transferred directly to FBF modules 31,
The far field voice signal that FBF modules 31 are gathered using the FBF parameters after calibration to microphone array 10 is handled, and generates voice
Reference signal, and ask poor with the noise reference signal Jing Guo adaptive noise cancellation module, obtain strengthening voice output 34.
The far field interactive voice equipment of another embodiment of the utility model, referring to Fig. 4, its far field with above-described embodiment
The 26S Proteasome Structure and Function of interactive voice equipment is essentially identical, and difference is, self calibration module also includes a signal-to-noise ratio computation mould
Block 25, it calculates the signal to noise ratio of the far field voice signal of each microphone channel, and the far field language that microphone array 10 is gathered
Message number is sent to FBF modules 31, and FBF parametric calibrations module 24 updates FBF parameters according to signal to noise ratio, and by the FBF after calibration
Parameter is sent to FBF modules 31, the far field voice that FBF modules 31 are gathered using the FBF parameters after calibration to microphone array 10
Signal is handled, and generates speech reference signal, and asks poor with the noise reference signal Jing Guo adaptive noise cancellation module, is obtained
To enhancing voice output 34.
The far field interactive voice equipment of the utility model above-mentioned two embodiment can equally improve rear end GSC performance,
Ensure that the speech damage of subsequent treatment is smaller, improve the robustness of equipment.
Due to microphone occur can not the failure of normal work belong to more extreme situation, therefore, FBF parametric calibration moulds
Block 24 only can start and work in far field interactive voice device power-up, complete microphone array detection and FBF parameters more
It can be closed after new, i.e., the detection of microphone array is only carried out in far field interactive voice device power-up and FBF parameters are updated,
Without being carried out in real time in equipment running process, to save calculation resources, equipment power dissipation is reduced.
In addition, the utility model can also carry out the detection of microphone array by remote control, referring to Fig. 5, far field
Interactive voice equipment can also include a control end 40, and self calibration and voice recognition terminal 20 can also include a control module.
Control end 40 can be mobile phone, itself and control module wireless connection, and control end and control module can be for example, by
The wireless protocol communications such as Wi-Fi or bluetooth.When needing to carry out microphone array detection, user can be by the spy of control end 40
Determine the control module that application program sends control signal, self calibration and voice recognition terminal to self calibration and voice recognition terminal 20
Receive after control signal, control FBF parametric calibrations module 24 starts and worked, carry out microphone array detection and FBF ginsengs
Number updates.Or, user can set the detection cycle of microphone array by the application-specific of control end 40, for example often
It carries out weekly a microphone array detection, and detection cycle is sent to control module, when reaching detection time,
Control module control FBF parametric calibrations module 24 starts and worked, and carries out microphone array detection and FBF parameters update.
In addition to this it is possible to set physical button or membrane keyboard on microphone, when needing to carry out microphone array
During detection, user can be with the button on Manual press microphone to produce a trigger signal, and control module receives trigger signal
Afterwards, control FBF parametric calibrations module 24 starts and worked, and carries out microphone array detection and FBF parameters update.
As can be seen here, the utility model can start microphone array detection in several ways and FBF parameters update, side
Just flexibly, it is convenient for the user to operate.
Referring to Fig. 6, it is possible to use above-mentioned far field interactive voice equipment carries out self calibration, including:
Far field voice signal is gathered using microphone array 10, and by far field transmitting voice signal to self calibration module;
Using self calibration module to far field voice signal as caused by microphone channel gain and signal propagation attenuation by mistake
Difference is calibrated, and detects the working condition of microphone array 10, calibrates FBF parameters;
Far field transmitting voice signal after calibration is to GSC, and GSC is using the FBF parameters after calibration to the far field language after calibration
Message is handled, and obtains strengthening voice output 34.
Wherein, GSC can also be gathered at the voice signal of far field using the FBF parameters after calibration to microphone array 10
Reason, obtains strengthening voice output 34.
Wherein, microphone array 10 is detected according to the signal to noise ratio of the relative gain of each passage or far field voice signal
Working condition, calibrates FBF parameters.
So far, the present embodiment is described in detail combined accompanying drawing.According to above description, those skilled in the art
There should be clear understanding to self-alignment far field interactive voice equipment of the present utility model.
It should be noted that in accompanying drawing or specification text, the implementation for not illustrating or describing is affiliated technology
Form known to a person of ordinary skill in the art, is not described in detail in field.In addition, above-mentioned definition to each element and not only limiting
Various concrete structures, shape or the mode mentioned in embodiment, those of ordinary skill in the art can be carried out simply more to it
Change or replace, for example:
(1) VAD and microphone works state-detection can also use other modes;
(2) direction term mentioned in embodiment, is only ginseng such as " on ", " under ", "front", "rear", "left", "right"
The direction of accompanying drawing is examined, not for limiting protection domain of the present utility model;
(3) consideration that above-described embodiment can be based on design and reliability, the collocation that is mixed with each other is used or and other embodiment
Mix and match is used, i.e., technical characteristic not in be the same as Example can freely form more embodiments.
Particular embodiments described above, has carried out entering one to the purpose of this utility model, technical scheme and beneficial effect
Step is described in detail, be should be understood that and be the foregoing is only specific embodiment of the utility model, is not limited to this
Utility model, all within spirit of the present utility model and principle, any modification, equivalent substitution and improvements done etc. all should be wrapped
It is contained within protection domain of the present utility model.
Claims (5)
1. a kind of self-alignment far field interactive voice equipment, it is characterised in that including:Microphone array is known with self calibration and voice
Other terminal, the microphone array is electrically connected with the self calibration and voice recognition terminal;
The microphone array gathers far field voice signal, and by the far field transmitting voice signal to the self calibration and voice
Identification terminal, wherein, the far field voice signal includes the error as caused by microphone channel gain and signal propagation attenuation;
The self calibration and voice recognition terminal include:Self calibration module and general matching law, the microphone array connect
Calibration module is connected to, the self calibration module connects the general matching law.
2. interactive voice equipment in far field as claimed in claim 1, it is characterised in that also include:Control end, the control end is
Mobile phone.
3. interactive voice equipment in far field as claimed in claim 1, it is characterised in that the self calibration and voice recognition terminal are
Computer or server.
4. far field as claimed in claim 1 interactive voice equipment, it is characterised in that also include:Control end.
5. interactive voice equipment in far field as claimed in claim 1, it is characterised in that the microphone of the microphone array is set
There is button.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201621197888.3U CN206489876U (en) | 2016-11-04 | 2016-11-04 | Self-alignment far field interactive voice equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201621197888.3U CN206489876U (en) | 2016-11-04 | 2016-11-04 | Self-alignment far field interactive voice equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN206489876U true CN206489876U (en) | 2017-09-12 |
Family
ID=59768644
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201621197888.3U Active CN206489876U (en) | 2016-11-04 | 2016-11-04 | Self-alignment far field interactive voice equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN206489876U (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110751946A (en) * | 2019-11-01 | 2020-02-04 | 达闼科技成都有限公司 | Robot and voice recognition device and method thereof |
CN113113037A (en) * | 2021-04-15 | 2021-07-13 | 南京邮电大学 | Microphone array voice beam forming system based on OMAP-L137 |
WO2023201886A1 (en) * | 2022-04-22 | 2023-10-26 | 歌尔股份有限公司 | Sound signal processing method and apparatus, device, and storage medium |
-
2016
- 2016-11-04 CN CN201621197888.3U patent/CN206489876U/en active Active
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110751946A (en) * | 2019-11-01 | 2020-02-04 | 达闼科技成都有限公司 | Robot and voice recognition device and method thereof |
CN113113037A (en) * | 2021-04-15 | 2021-07-13 | 南京邮电大学 | Microphone array voice beam forming system based on OMAP-L137 |
WO2023201886A1 (en) * | 2022-04-22 | 2023-10-26 | 歌尔股份有限公司 | Sound signal processing method and apparatus, device, and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106340305B (en) | Self-calibration far-field voice interaction device and far-field voice self-calibration method | |
US11798531B2 (en) | Speech recognition method and apparatus, and method and apparatus for training speech recognition model | |
CN100535992C (en) | Small scale microphone array speech enhancement system and method | |
CN206489876U (en) | Self-alignment far field interactive voice equipment | |
US11094334B2 (en) | Sound processing method and apparatus | |
US10412518B2 (en) | Blocked microphone detection | |
CN107221336A (en) | It is a kind of to strengthen the devices and methods therefor of target voice | |
CN106251857A (en) | Sounnd source direction judgment means, method and mike directivity regulation system, method | |
CN104158990A (en) | Method for processing an audio signal and audio receiving circuit | |
CN107154256B (en) | Sound masking system based on sound source positioning and self-adaptive adjusting method | |
CN103426440A (en) | Voice endpoint detection device and voice endpoint detection method utilizing energy spectrum entropy spatial information | |
CN102710839B (en) | A kind of method and communication terminal for lifting voice call effect | |
CN106210962A (en) | A kind of ear stricture of vagina acoustic characteristic ONLINE RECOGNITION method and system being applicable to noise cancelling headphone design | |
CN110234044A (en) | A kind of voice awakening method, voice Rouser and earphone | |
CN108122559A (en) | Binaural sound sources localization method based on deep learning in a kind of digital deaf-aid | |
CN111640431A (en) | Equipment response processing method and device | |
CN107580155A (en) | Networking telephone quality determination method, device, computer equipment and storage medium | |
Lin et al. | Bionic optimization of MFCC features based on speaker fast recognition | |
CN106618499B (en) | Fall detection device, fall detection method and device | |
CN108680902A (en) | A kind of sonic location system based on multi-microphone array | |
CN107592600A (en) | A kind of pickup screening technique and sound pick up equipment based on distributed microphone | |
CN106409278A (en) | Drone active noise control device | |
EP4021012B1 (en) | Method and apparatus for recognizing wind noise of earphone | |
CN104424954B (en) | noise estimation method and device | |
CN100559817C (en) | Stereo echo removing method and system thereof based on the voice signal disjunctive model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
GR01 | Patent grant | ||
GR01 | Patent grant |