CN103811013A - Noise suppression method, device thereof, electronic equipment and communication processing method - Google Patents
Noise suppression method, device thereof, electronic equipment and communication processing method Download PDFInfo
- Publication number
- CN103811013A CN103811013A CN201210442169.3A CN201210442169A CN103811013A CN 103811013 A CN103811013 A CN 103811013A CN 201210442169 A CN201210442169 A CN 201210442169A CN 103811013 A CN103811013 A CN 103811013A
- Authority
- CN
- China
- Prior art keywords
- terminal
- voice signal
- noise
- module
- suppressing method
- 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.)
- Granted
Links
Images
Abstract
The invention provides a noise suppression method, a device thereof, electronic equipment and a communication processing method. The noise suppression method comprises the steps of obtaining a first speech signal collected by a first terminal in a user speech collection state, obtaining at least one second speech signal collected by a second terminal in a non user speech collection state, and using the second speech signal to carry out noise elimination operation on the first speech signal and obtaining a third speech signal. According to the noise suppression method, the device thereof, the electronic equipment and the communication processing method, the noise suppression of the terminal of a single microphone can be carried out, and the noise suppression of the terminals of multiple microphones can be carried out too.
Description
Technical field
The present invention relates to mobile multimedia field, relate in particular to noise suppressing method, device, electronic equipment and communication processing method.
Background technology
Along with the develop rapidly of computer technology, computer has entered into huge numbers of families, and it is infinite convenient to have brought to the mankind.Meanwhile, people are also more and more higher to the intelligent requirements of computing machine, certainly the most outstanding demand is embodied in man-machine interface, if can realize the direct dialogue between man-machine, allow " machine " to understand people's language, and go executor's intention according to its information, this is undoubtedly optimal human-machine intelligence's interface mode so, therefore speech recognition is arisen at the historic moment as the subject of very attractive, and a lot of experts point out that speech recognition technology will be Future Ten year one of areas of information technology ten large important development in science and technology technology.
The basic goal of speech recognition (Speech Recognition) research is to work out a kind of machine with auditory function, and directly acceptor's mortar is exhaled order, understands people's intention and makes corresponding reaction.Speech recognition technology, take voice signal as research object, relates to the numerous areas such as linguistics, computer science, signal processing, physiology, psychology, is the important branch of pattern-recognition.This technology has boundless application prospect, such as: the computing machine with speech interface can change people's mode of operation to computer at present, causes the revolution of operating system; Dictation machine will bring great change for office automation, also make the computer input of some non-alphabetic writing (as Chinese) be not a kind of technical ability that needs training simultaneously: communication aspects more magnificent plan be the direct communication realizing between bilingual, by " speech recognition one mechanical translation one phonetic synthesis ", a kind of language is directly changed into another kind of language; Speech recognition can make user pass through the direct searching database of voice, not only economy but also rapid; Also have in some special industries, as automobile, battlebus, aviator carries out dialing or issue an order in running at high speed.
At nearly 2 years, speech recognition technology was obtained rapid progress, and the many companies take Apple, NTT Docomo as representative have released the products such as Siri voice assistant, voiced translation, and these products have been subject to extensive concern.
Speech recognition is tested, found to exist under the true environment for use of noise, noise identification rate declines very fast.As shown in Figure 1, the available threshold of voice signal discrimination is at 65%-75%, and in the time that signal to noise ratio (S/N ratio) reaches 12dB, if do not carry out squelch, the discrimination of voice signal declines and is exceedingly fast, and will drop to below 60%.
Patent publication No. is CN101873363A, has proposed a kind of method and terminal that uses couple Mikes to suppress noise in the patent document that open day is on October 27th, 2010, comprising: sense terminals current talking state, judges whether terminal enters earphone talking state; In the time that terminal enters non-earphone talking state, be that main microphone and reference microphone are selected voice-frequency channel by time-multiplexed mode; By main microphone collection call sound and by reference to microphone collection ground unrest; Utilize ground unrest to carry out squelch processing to the noise in call sound.
First this scheme judges whether to adopt headset input, if do not adopt headset input, by 2 Mic, is set to main Mic and from Mic, main Mic gathers voice, and calculates with the sound gathering from Mic, realizes squelch.This scheme requires described terminal must possess two Mic, and the Mic number that must increase terminal carries out the inhibition of noise.
Summary of the invention
The object of this invention is to provide a kind of noise suppressing method, device, electronic equipment and communication processing method, can carry out squelch to the terminal of single Mic, also can carry out squelch to the terminal of multiple Mic.
To achieve these goals, the embodiment of the present invention provides a kind of noise suppressing method, comprising:
Obtain the first voice signal that the first terminal under user speech acquisition state collects;
Obtain the second voice signal that at least one second terminal under non-user speech acquisition state collects;
Utilize described the second voice signal to carry out noise cancellation operation to described the first voice signal, obtain the 3rd voice signal.
Wherein, preferably, described noise suppressing method is for telex network process or speech recognition process, and when described noise suppressing method is used for telex network process, described first terminal and the second terminal are wireless communication terminal.
Wherein, preferably, described noise suppressing method is for first terminal, described in obtain the second voice signal that at least one second terminal under non-user speech acquisition state collects and specifically comprise:
Broadcast comprises the request message of self-position information;
Receive the response message that carries described the second voice signal that described the second terminal is returned; Described the second terminal is to receive in the terminal of described request message, is less than the terminal of predeterminable range thresholding with the distance of described first terminal.
Wherein, preferably, described noise suppressing method is for a network equipment, and described the first voice signal is specially described first terminal and sends under talking state the voice signal of described network equipment;
Described the second voice signal corresponding at least one the second terminal under non-user speech acquisition state that obtain specifically comprises:
Receive from described the second terminal the response message that carries described the second voice signal;
Described the second terminal is to receive in the terminal of the request message that comprises self-position information of described first terminal broadcast, judges according to described self-position information the terminal that is less than predeterminable range thresholding with the distance of described first terminal.
Wherein, preferably, obtain the second voice signal corresponding at least one the second terminal under non-user speech acquisition state and specifically comprise:
Determine the first collection moment of the original sound signal that described the first voice signal is corresponding;
Determine propagation delay time according to the distance between the second terminal and described first terminal;
In the voice signal that the sound that collects from the second terminal generates, be chosen in voice signal that signal that the second acquisition time gathers generates as described the second voice signal under non-talking state; Described the second collection moment is the difference of the first collection moment and described propagation delay time.
Wherein, preferably, described the second voice signal exists when multiple, utilizes described the second voice signal to carry out noise cancellation operation to described the first voice signal, obtains the 3rd voice signal and specifically comprises:
According to all the second voice signals short-time average energy separately, determine current background noise pattern;
Determine each the second voice signal weight separately according to current background noise pattern;
According to described the second voice signal and weight separately, described the first voice signal is carried out to noise cancellation operation, obtain described the 3rd voice signal.
To achieve these goals, the embodiment of the present invention also provides a kind of Noise Suppression Device, comprising:
Sound collection module, the first voice signal collecting for the first terminal obtaining under user speech acquisition state;
Neighbourhood noise acquisition module, the second voice signal collecting for obtaining at least one second terminal under non-user speech acquisition state;
Noise processed module, for utilizing described the second voice signal to carry out noise cancellation operation to described the first voice signal, obtains the 3rd voice signal.
Wherein, preferably, described Noise Suppression Device is for telex network process or speech recognition process, and when described Noise Suppression Device is used for telex network process, described first terminal and the second terminal are wireless communication terminal.
Wherein, preferably, described Noise Suppression Device is for first terminal, and described neighbourhood noise acquisition module specifically comprises:
User interactive module, for broadcasting the request message that comprises self-position information;
Neighbourhood noise receiver module, the response message that carries described the second voice signal returning for receiving described the second terminal; Described the second terminal is to receive in the terminal of described request message, is less than the terminal of predeterminable range thresholding with the distance of described first terminal.
Wherein, preferably, described Noise Suppression Device is for a network equipment, and described the first voice signal is specially described first terminal and sends under talking state the voice signal of described network equipment;
Described neighbourhood noise acquisition module specifically comprises:
Neighbourhood noise receiver module, for receiving from described the second terminal the response message that carries described the second voice signal;
Described the second terminal is to receive in the terminal of the request message that comprises self-position information of described first terminal broadcast, judges according to described self-position information the terminal that is less than predeterminable range thresholding with the distance of described first terminal.
Wherein, preferably, neighbourhood noise acquisition module specifically comprises:
Gather moment determination module, for determining the first collection moment of the original sound signal that described the first voice signal is corresponding;
Transfer delay determination module, for determining propagation delay time according to the distance between the second terminal and described first terminal;
Select module, the voice signal generating for the sound that collects under non-talking state from the second terminal, is chosen in voice signal that signal that the second acquisition time gathers generates as described the second voice signal; Described the second collection moment is the difference of the first collection moment and described propagation delay time.
Wherein, preferably, described the second voice signal exists when multiple, and noise processed module specifically comprises:
Noise pattern determination module, for according to all the second voice signals short-time average energy separately, determines current background noise pattern;
Weighting coefficient determination module, for determining each the second voice signal weight separately according to current background noise pattern;
Voice signal extraction module, for according to described the second voice signal and weight separately, described the first voice signal being carried out to noise cancellation operation, obtains described the 3rd voice signal.
To achieve these goals, the embodiment of the present invention also provides a kind of electronic equipment, comprises the Noise Suppression Device described in above-mentioned any one.
Wherein, preferably, described electronic equipment is a terminal.
Wherein, preferably, described electronic equipment is a network equipment.
To achieve these goals, the embodiment of the present invention also provides the communication processing method of noise suppressing method described in above-mentioned any one, comprising:
Send described the 3rd voice signal to Correspondent Node.
The embodiment of the present invention has following beneficial effect:
The embodiment of the present invention, can carry out squelch to single Mic terminal, solves the problem that needs constantly increase terminal Mic number to obtain better de-noising effect;
The embodiment of the present invention, with existing audio recognition method and system compatible, the acoustic information that different terminals is collected, process in terminal or on network equipment, realize and improve the performance that noise is eliminated, and then improving quality of service lifting user and experience;
The embodiment of the present invention, does not significantly increase user's operation, supports the miscellaneous service of operator, can realize smooth upgrade in existing terminal and network equipment, and system improving cost is low, can greatly promote related service development, increases business business revenue.
Accompanying drawing explanation
Fig. 1 is noise identification rate schematic diagram;
Fig. 2 is the schematic flow sheet of the noise suppressing method of the embodiment of the present invention;
Fig. 3 is the structural representation of the noise suppressing system of the embodiment of the present invention;
Fig. 4 is the schematic flow sheet of the noise suppressing function module implementation procedure of the embodiment of the present invention;
Fig. 5 is the structural representation of the Noise Suppression Device of the embodiment of the present invention;
Fig. 6 is the structural representation of the noise suppressing function entity of the embodiment of the present invention;
Fig. 7 is the structural representation that the embodiment of the present invention is applied to the noise suppressing function entity of terminal;
Fig. 8 is the squelch operation flow schematic diagram of the embodiment of the present invention.
Embodiment
For technical matters, technical scheme and advantage that the embodiment of the present invention will be solved are clearer, be described in detail below in conjunction with the accompanying drawings and the specific embodiments.
The embodiment of the present invention provides a kind of noise suppressing method, as shown in Figure 2, comprising:
Step 21, obtains the first voice signal that the first terminal under user speech acquisition state collects;
Step 22, obtains the second voice signal that at least one second terminal under non-user speech acquisition state collects;
Step 23, utilizes described the second voice signal to carry out noise cancellation operation to described the first voice signal, obtains the 3rd voice signal.
Gather the first voice signal by the first terminal under user speech acquisition state, the second terminal by least one under user speech acquisition state gathers the second voice signal, the first voice signal that first terminal collects is the synthetic speech of efficient voice and noise voice, the second terminal only gathers first terminal noise voice around, utilize the second voice signal to carry out noise elimination to the first voice signal, the 3rd voice signal of the noise that has been eliminated.
As shown in Figure 3, terminal U obtains phonetic entry, i.e. the first voice signal, and terminal A1-An carries out Noise Acquisition, i.e. the second voice signal.Said process can carry out noise elimination to single Mic terminal, solves the problem that needs constantly increase terminal Mic number to obtain better de-noising effect.
Certainly, the method for the embodiment of the present invention is also applicable to comprise the terminal of multiple MIC.
Simultaneously, above-mentioned noise suppressing method is applicable to communication or carries out in the process of speech recognition, in the time communicating, first terminal and the second terminal are common wireless communication terminal, as mobile phone etc., be that described noise suppressing method can be for telex network process or speech recognition process, when described noise suppressing method is used for telex network process, described first terminal and the second terminal are wireless communication terminal.Describe respectively below.
When described noise suppressing method is used for first terminal, step 22 specifically comprises:
Step 221, broadcast comprises the request message of self-position information;
Step 222, receives the response message that carries described the second voice signal that described the second terminal is returned; Described the second terminal is to receive in the terminal of described request message, is less than the terminal of predeterminable range thresholding with the distance of described first terminal.
First terminal is broadcast request message " Ask Environment Noise " towards periphery, sends the position of oneself simultaneously; When the second terminal receives after described request message, judge the distance of self and first terminal, distance is in predeterminable range thresholding time, and the second terminal just can be returned to the response message that carries described the second voice signal.Second terminal here can be the terminal device that other users use, and can be also the terminal device setting in advance in each place, and predeterminable range thresholding is traditionally arranged to be 5m.Exceed the second terminal of predeterminable range thresholding with the distance of first terminal, do not need to return described response message, this is to be the noise signal that gathers surrounding environment for first terminal because gather the second voice signal, from the second terminal of first terminal hypertelorism, the second voice signal collecting can not have been served as the noise signal of first terminal environment of living in.
First terminal is receiving after the response message that carries described the second voice signal that described the second terminal returns, utilize the second voice signal to carry out noise cancellation operation to the first voice signal obtaining, obtain the 3rd voice signal, the first terminal here must meet computation requirement, can carry out sound and " subtract each other ".If when first terminal can not reach computation requirement, the embodiment of the present invention also provides another noise suppressing method, is applied on network equipment.
Described noise suppressing method is when the network equipment, and described the first voice signal is specially described first terminal and sends under talking state the voice signal of described network equipment;
Step 22 specifically comprises:
Step 221, receives from described the second terminal the response message that carries described the second voice signal;
Described the second terminal is to receive in the terminal of the request message that comprises self-position information of described first terminal broadcast, judges according to described self-position information the terminal that is less than predeterminable range thresholding with the distance of described first terminal.
Network equipment receives the first voice signal that first terminal sends under talking state, receive the response message that has carried the second voice signal that the second terminal sends simultaneously, wherein the second terminal is to receive in the terminal of the request message that comprises self-position information of described first terminal broadcast, judges according to described self-position information the terminal that is less than predeterminable range thresholding with the distance of described first terminal.
Can not carry out noise cancellation operation when terminal, can not reach and calculate while needs, the first voice signal, the second voice signal can be sent on network equipment, utilize the powerful computing power of network equipment to complete noise cancellation operation.
The acoustic information that the embodiment of the present invention collects different terminals is processed on terminal or network equipment, to realize and to improve the performance that noise is eliminated, and then improves quality of service, has promoted user's experience.
Described noise suppressing method, while being applied to terminal or network equipment, the definite of the second voice signal is specially:
Determine the first collection moment of the original sound signal that described the first voice signal is corresponding;
Determine propagation delay time according to the distance between the second terminal and described first terminal;
In the voice signal that the sound that collects from the second terminal generates, be chosen in voice signal that signal that the second acquisition time gathers generates as described the second voice signal under non-talking state; Described the second collection moment is the difference of the first collection moment and described propagation delay time.
Owing to there being certain distance between the second terminal and first terminal, utilizing the second voice signal to carry out noise cancellation operation to the first voice signal, while obtaining the 3rd voice signal, can not carry out noise cancellation operation to first voice signal in the first moment with second voice signal in the first moment, the first the first voice signal that gathers the moment should be corresponding with described the second voice signal that the second acquisition time gathers, and described the second collection moment is the difference of the first collection moment and described propagation delay time; If in the first collection moment, the second terminal does not also start to gather the second voice signal, now first gathers first voice signal in moment and should receive described request information with the second terminal and start the second voice signal of the initial time that gathers the second voice signal corresponding.
For example: first terminal starts to gather the first voice signal in the t1 moment, the second terminal starts to gather the second voice signal in the t2 moment, between the second terminal and first terminal, there is the propagation delay time of Δ t, in the time carrying out squelch processing, the corresponding relation of the first voice signal and the second voice signal is as shown in table 1:
| The first voice signal | t1 | t1+Δt | tn |
| The second voice signal | t2 | t2 | tn-Δt |
Table 1
In described noise suppressing method, when described the second voice signal exists when multiple, step 23 specifically comprises:
Step 231, according to all the second voice signals short-time average energy separately, determines current background noise pattern;
Step 232, determines each the second voice signal weight separately according to current background noise pattern;
Step 233, carries out noise cancellation operation according to described the second voice signal and weight separately to described the first voice signal, obtains described the 3rd voice signal.
According to all the second voice signals short-time average energy separately, determine current background noise pattern; Again according to each the second voice signal of definite ground unrest mode decision weight separately; According to described the second voice signal and weight separately, described the first voice signal is carried out to noise cancellation operation, obtain described the 3rd voice signal.The noise pattern is here smooth noise pattern, single-point noise pattern or multiple spot noise pattern.
The implementation procedure of the noise suppressing method that the embodiment of the present invention provides, as shown in Figure 4, the sound that terminal U and terminal An are collected, is designated as Su (x, y, and Sn (x t), y, t), be illustrated respectively in the t time, the sound that is positioned at the terminal record of (x, y) position is S;
The efficient voice that terminal U is gathered, is designated as Vu (x, y, t);
Noise when terminal U is gathered, is designated as Nu (x, y, t);
The object of above-mentioned noise suppressing method is to recover PVu (x from Su (x, y, t) and Sn (x, y, t), y, t), make PVu (x, y, t) and Vu (x, y, t) between error e (x, y, t) minimum.
The method that the present embodiment provides, with existing audio recognition method and system compatible, the acoustic information that different terminals is collected, process in terminal or on network equipment, realize and improve the performance that noise is eliminated, and then improving quality of service lifting user and experience;
Significantly do not increase user's operation, support the miscellaneous service of operator, can realize smooth upgrade in existing terminal and network equipment, system improving cost is low simultaneously, can greatly promote related service development, increases business business revenue.
The embodiment of the present invention also provides a kind of Noise Suppression Device, as shown in Figure 5, comprising:
Sound collection module, the first voice signal collecting for the first terminal obtaining under user speech acquisition state;
Neighbourhood noise acquisition module, the second voice signal collecting for obtaining at least one second terminal under non-user speech acquisition state;
Noise processed module, for utilizing described the second voice signal to carry out noise cancellation operation to described the first voice signal, obtains the 3rd voice signal.
Sound collection module gathers the first voice signal of the first terminal under user speech acquisition state, neighbourhood noise acquisition module gathers the second voice signal of at least one the second terminal under user speech acquisition state, the first voice signal that first terminal collects is the synthetic speech of efficient voice and noise voice, the second terminal only gathers first terminal noise voice around, noise processed module utilizes the second voice signal to carry out noise elimination to the first voice signal, the 3rd voice signal of the noise that has been eliminated.The de-noising of the Noise Suppression Device support that the embodiment of the present invention provides to single Mic terminal, solves the problem that needs constantly increase terminal Mic number to obtain better de-noising effect.
Wherein, preferably, described Noise Suppression Device is for telex network process or speech recognition process, and when described Noise Suppression Device is used for telex network process, described first terminal and the second terminal are wireless communication terminal.
Wherein, preferably, described Noise Suppression Device is for first terminal, and described neighbourhood noise acquisition module specifically comprises:
User interactive module, for broadcasting the request message that comprises self-position information;
Neighbourhood noise receiver module, the response message that carries described the second voice signal returning for receiving described the second terminal; Described the second terminal is to receive in the terminal of described request message, is less than the terminal of predeterminable range thresholding with the distance of described first terminal.
Wherein, preferably, described Noise Suppression Device is for a network equipment, and described the first voice signal is specially described first terminal and sends under talking state the voice signal of described network equipment;
Described neighbourhood noise acquisition module specifically comprises:
Neighbourhood noise receiver module, for receiving from described the second terminal the response message that carries described the second voice signal;
Described the second terminal is to receive in the terminal of the request message that comprises self-position information of described first terminal broadcast, judges according to described self-position information the terminal that is less than predeterminable range thresholding with the distance of described first terminal.
Wherein, preferably, neighbourhood noise acquisition module specifically comprises:
Gather moment determination module, for determining the first collection moment of the original sound signal that described the first voice signal is corresponding;
Transfer delay determination module, for determining propagation delay time according to the distance between the second terminal and described first terminal;
Select module, the voice signal generating for the sound that collects under non-talking state from the second terminal, is chosen in voice signal that signal that the second acquisition time gathers generates as described the second voice signal; Described the second collection moment is the difference of the first collection moment and described propagation delay time.
Wherein, preferably, described the second voice signal exists when multiple, and noise processed module specifically comprises:
Noise pattern determination module, for according to all the second voice signals short-time average energy separately, determines current background noise pattern;
Weighting coefficient determination module, for determining each the second voice signal weight separately according to current background noise pattern;
Voice signal extraction module, for according to described the second voice signal and weight separately, described the first voice signal being carried out to noise cancellation operation, obtains described the 3rd voice signal.
The embodiment of the present invention is according to above-mentioned noise processed Process Design and realize a functional module entity, as shown in Figure 6, comprising:
Distance calculation module: calculate the position to terminal U and terminal An, and take terminal U as initial point, calculate distance and the angle of each terminal An, be designated as ρ n and θ n;
Background noise graph generation module: sound that each terminal An is collected calculates short-time average energy, based on En, and completes background noise graph, wherein the method for expressing of background noise graph has multiplely, wherein a kind ofly represents with forms mode, as shown in table 2:
| Distance | Short-time energy | Time | Additional information |
| {ρ1，，θ1} | E1 | T1 | {m1，ip1，....} |
| {ρ2，，θ2} | E2 | T2 | {m2，ip2，....} |
| {ρ3，，θ3} | E3 | T3 | {m3，ip3，....} |
Table 2
Noise pattern discrimination module: as less in each En difference, be smooth noise pattern, as exist an En value to be obviously greater than other, be single-point noise pattern, as exist several En values to be obviously greater than other, be multiple spot noise pattern;
Weighting coefficient and time calculations of offset module: according to the distance of An and U, and noise pattern calculates weighting coefficient, as be smooth noise pattern, the weighting coefficient Wn that each An is corresponding is identical, as is single-point noise pattern, the Wn that maximal value An is corresponding is calculated by En, for the ratio of En and average En, as be multiple spot noise pattern, the Wn that each extreme value An is corresponding is calculated by En, for the ratio of En and average En, Wn recycling range information is revised;
According to the distance of An and U, computing time, side-play amount, namely estimated that An place noise is sent to the time delay that U is ordered, and is designated as μ n;
Voice signal extraction module: the voice signal gathering from terminal U, (t-μ n) for x, y for noise signal PVu (x, y, t)=Su (x, y, t)-∑ Wn*Sn of " deducting " weighting terminal An.
Described noise suppressing method can be applied to terminal or network equipment, and the embodiment of the present invention provides a kind of functional module entity of realizing in terminal, as shown in Figure 7.
Terminal U comprises:
User interactive module, for carrying out alternately with terminal An, sends the request message that comprises self-position information;
The first sound acquisition module, for gathering the first voice messaging Su (x, y, t);
Neighbourhood noise request module, the neighbourhood noise voice messaging sending for receiving terminal An, i.e. the second voice messaging Sn (x, y, t);
Distance calculation module, for determining the distance of terminal An and terminal U;
Background noise graph generation module, calculates short-time average energy for the sound that terminal An is collected, and based on En, and completes background noise graph;
Noise pattern discrimination module, for determining noise pattern according to En;
Weighting coefficient and time calculations of offset module, for according to the distance of An and U, and noise pattern calculates weighting coefficient and determines weighting system Wn and time migration μ n;
Voice signal extraction module, for " deducting " noise signal of weighting terminal An from the voice signal of terminal U collection.
Terminal An comprises:
Request-reply module, for returning to the response message that carries described the second voice signal;
Location and time logging modle, for recording terminal An and the distance of terminal U and the time of acquisition noise;
The second sound acquisition module, for acquisition noise, and sends to terminal U by the noise collecting, terminal An and the distance of terminal U and the time of acquisition noise.
By said process, terminal U has obtained eliminating the voice signal after noise, can realize the noise of single Mic terminal and eliminate.
Above-mentioned operation flow is carried out to overall description below, as shown in Figure 8.
When terminal U starts to carry out phonetic entry, broadcast request " Ask Environment Noise " towards periphery, the position of transmission simultaneously oneself;
Terminal An around, is receiving after request, the distance of first relatively own and terminal U, if within the scope of threshold value Dist, and response request; The threshold value of Dist is traditionally arranged to be 5m, and An can be A1, A2, A3......;
Subsequently, terminal U and terminal An carry out record, in the time that system is moved, utilize the Mic acquisition noise information of terminal (An) around, then send to terminal U, or send to server, wherein terminal U is recorded as voice and noise, and terminal An is recorded as neighbourhood noise; Meanwhile, record position and the time;
Stopping, after record, being divided into following two kinds of situations:
A. terminal U judges that own performance can meet computation requirement, requires terminal An to send data to oneself, oneself carries out sound and " subtracts each other ";
B. terminal U judges that own performance can not meet computation requirement, requires terminal An to send data to network equipment, carries out sound " subtract each other " by network equipment.
To achieve these goals, the embodiment of the present invention also provides a kind of electronic equipment, comprises the Noise Suppression Device described in above-mentioned arbitrary.
Wherein, preferably, described electronic equipment is a terminal.
Wherein, preferably, described electronic equipment is a network equipment.
To achieve these goals, the embodiment of the present invention also provides the communication processing method that utilizes the noise suppressing method described in above-mentioned any one, also comprises:
Send described the 3rd voice signal to Correspondent Node.
The above is only the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (16)
1. a noise suppressing method, is characterized in that, comprising:
Obtain the first voice signal that the first terminal under user speech acquisition state collects;
Obtain the second voice signal that at least one second terminal under non-user speech acquisition state collects;
Utilize described the second voice signal to carry out noise cancellation operation to described the first voice signal, obtain the 3rd voice signal.
2. noise suppressing method according to claim 1, it is characterized in that, described noise suppressing method is for telex network process or speech recognition process, and when described noise suppressing method is used for telex network process, described first terminal and the second terminal are wireless communication terminal.
3. noise suppressing method according to claim 2, is characterized in that, described noise suppressing method is for first terminal, described in obtain the second voice signal that at least one second terminal under non-user speech acquisition state collects and specifically comprise:
Broadcast comprises the request message of self-position information;
Receive the response message that carries described the second voice signal that described the second terminal is returned; Described the second terminal is to receive in the terminal of described request message, is less than the terminal of predeterminable range thresholding with the distance of described first terminal.
4. noise suppressing method according to claim 2, is characterized in that, described noise suppressing method is for a network equipment, and described the first voice signal is specially described first terminal and sends under talking state the voice signal of described network equipment;
Described the second voice signal corresponding at least one the second terminal under non-user speech acquisition state that obtain specifically comprises:
Receive from described the second terminal the response message that carries described the second voice signal;
Described the second terminal is to receive in the terminal of the request message that comprises self-position information of described first terminal broadcast, judges according to described self-position information the terminal that is less than predeterminable range thresholding with the distance of described first terminal.
5. according to the noise suppressing method described in any one in claim 1-4, it is characterized in that, obtain the second voice signal corresponding at least one the second terminal under non-user speech acquisition state and specifically comprise:
Determine the first collection moment of the original sound signal that described the first voice signal is corresponding;
Determine propagation delay time according to the distance between the second terminal and described first terminal;
In the voice signal that the sound that collects from the second terminal generates, be chosen in voice signal that signal that the second acquisition time gathers generates as described the second voice signal under non-talking state; Described the second collection moment is the difference of the first collection moment and described propagation delay time.
6. noise suppressing method according to claim 5, is characterized in that, described the second voice signal exists when multiple, utilizes described the second voice signal to carry out noise cancellation operation to described the first voice signal, obtains the 3rd voice signal and specifically comprises:
According to all the second voice signals short-time average energy separately, determine current background noise pattern;
Determine each the second voice signal weight separately according to current background noise pattern;
According to described the second voice signal and weight separately, described the first voice signal is carried out to noise cancellation operation, obtain described the 3rd voice signal.
7. a Noise Suppression Device, is characterized in that, comprising:
Sound collection module, the first voice signal collecting for the first terminal obtaining under user speech acquisition state;
Neighbourhood noise acquisition module, the second voice signal collecting for obtaining at least one second terminal under non-user speech acquisition state;
Noise processed module, for utilizing described the second voice signal to carry out noise cancellation operation to described the first voice signal, obtains the 3rd voice signal.
8. Noise Suppression Device according to claim 7, it is characterized in that, described Noise Suppression Device is for telex network process or speech recognition process, and when described Noise Suppression Device is used for telex network process, described first terminal and the second terminal are wireless communication terminal.
9. Noise Suppression Device according to claim 8, is characterized in that, described Noise Suppression Device is for first terminal, and described neighbourhood noise acquisition module specifically comprises:
User interactive module, for broadcasting the request message that comprises self-position information;
Neighbourhood noise receiver module, the response message that carries described the second voice signal returning for receiving described the second terminal; Described the second terminal is to receive in the terminal of described request message, is less than the terminal of predeterminable range thresholding with the distance of described first terminal.
10. Noise Suppression Device according to claim 8, is characterized in that, described Noise Suppression Device is for a network equipment, and described the first voice signal is specially described first terminal and sends under talking state the voice signal of described network equipment;
Described neighbourhood noise acquisition module specifically comprises:
Neighbourhood noise receiver module, for receiving from described the second terminal the response message that carries described the second voice signal;
Described the second terminal is to receive in the terminal of the request message that comprises self-position information of described first terminal broadcast, judges according to described self-position information the terminal that is less than predeterminable range thresholding with the distance of described first terminal.
11. Noise Suppression Devices according to any one of claims of claim 7-10, is characterized in that, neighbourhood noise acquisition module specifically comprises:
Gather moment determination module, for determining the first collection moment of the original sound signal that described the first voice signal is corresponding;
Transfer delay determination module, for determining propagation delay time according to the distance between the second terminal and described first terminal;
Select module, the voice signal generating for the sound that collects under non-talking state from the second terminal, is chosen in voice signal that signal that the second acquisition time gathers generates as described the second voice signal; Described the second collection moment is the difference of the first collection moment and described propagation delay time.
12. Noise Suppression Devices according to claim 11, is characterized in that, described the second voice signal exists when multiple, and noise processed module specifically comprises:
Noise pattern determination module, for according to all the second voice signals short-time average energy separately, determines current background noise pattern;
Weighting coefficient determination module, for determining each the second voice signal weight separately according to current background noise pattern;
Voice signal extraction module, for according to described the second voice signal and weight separately, described the first voice signal being carried out to noise cancellation operation, obtains described the 3rd voice signal.
13. 1 kinds of electronic equipments, is characterized in that, comprise the Noise Suppression Device described in claim 7 or 8.
14. electronic equipments as claimed in claim 13, is characterized in that, described electronic equipment is a terminal.
15. electronic equipments as claimed in claim 13, is characterized in that, described electronic equipment is a network equipment.
16. 1 kinds of communication processing methods, is characterized in that, comprise the noise suppressing method described in any one in claim 1-4, also comprise:
Send described the 3rd voice signal to Correspondent Node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210442169.3A CN103811013B (en) | 2012-11-07 | 2012-11-07 | Noise suppression method, device thereof, electronic equipment and communication processing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210442169.3A CN103811013B (en) | 2012-11-07 | 2012-11-07 | Noise suppression method, device thereof, electronic equipment and communication processing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103811013A true CN103811013A (en) | 2014-05-21 |
| CN103811013B CN103811013B (en) | 2017-05-03 |
Family
ID=50707686
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210442169.3A Active CN103811013B (en) | 2012-11-07 | 2012-11-07 | Noise suppression method, device thereof, electronic equipment and communication processing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103811013B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104092802A (en) * | 2014-05-27 | 2014-10-08 | 中兴通讯股份有限公司 | Method and system for de-noising audio signal |
| WO2015176372A1 (en) * | 2014-05-22 | 2015-11-26 | 中兴通讯股份有限公司 | Method for reducing communication noise of intelligent terminal and intelligent terminal thereof |
| CN105227725A (en) * | 2014-07-02 | 2016-01-06 | 中兴通讯股份有限公司 | Realize method and the terminal of call de-noising |
| CN105635379A (en) * | 2015-12-29 | 2016-06-01 | 腾讯科技（深圳）有限公司 | Noise suppression method and device |
| CN105654960A (en) * | 2015-09-21 | 2016-06-08 | 宇龙计算机通信科技(深圳)有限公司 | Terminal sound denoising processing method and apparatus thereof |
| CN107623783A (en) * | 2017-09-11 | 2018-01-23 | 安徽省未来博学信息技术有限公司 | A kind of call voice optimizes system |
| CN109817209A (en) * | 2019-01-16 | 2019-05-28 | 深圳市友杰智新科技有限公司 | A kind of intelligent speech interactive system based on two-microphone array |
| CN110070881A (en) * | 2014-06-14 | 2019-07-30 | 宝利通公司 | For reducing the acoustics circumference for the noise that communication equipment is transmitted in open environment |
| CN110335589A (en) * | 2018-03-29 | 2019-10-15 | 松下电器产业株式会社 | Speech translation apparatus, voice translation method and recording medium |
| CN111052228A (en) * | 2017-09-28 | 2020-04-21 | 哈曼国际工业有限公司 | Method and apparatus for speech recognition |
| CN112303829A (en) * | 2020-09-18 | 2021-02-02 | 珠海格力电器股份有限公司 | Wind noise intelligent wind speed adjusting method and system, air conditioner, remote controller and storage medium |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040158457A1 (en) * | 2003-02-12 | 2004-08-12 | Peter Veprek | Intermediary for speech processing in network environments |
| US20070067423A1 (en) * | 2003-05-22 | 2007-03-22 | Anthony Dauguet | Context management system for a network including a heterogenous set of terminals |
| CN101193139A (en) * | 2006-11-20 | 2008-06-04 | 鸿富锦精密工业（深圳）有限公司 | A method and its mobile phone for filtering environmental noise |
| CN101193384A (en) * | 2006-11-20 | 2008-06-04 | 鸿富锦精密工业（深圳）有限公司 | Method and mobile phone for filtering environmental noise via pattern recognition |
| CN201417603Y (en) * | 2009-06-23 | 2010-03-03 | 深圳市奥尼电子工业有限公司 | Double microphone speech processor |
| CN101699831A (en) * | 2009-10-23 | 2010-04-28 | 深圳华为通信技术有限公司 | Terminal speech transmitting method, system and equipment |
| CN1753084B (en) * | 2004-09-23 | 2010-06-23 | 纽昂斯通讯公司 | Multi-channel adaptive speech signal processing with noise reduction |
| CN101911724A (en) * | 2008-03-18 | 2010-12-08 | 高通股份有限公司 | Speech enhancement using multiple microphones on multiple devices |
| US20110179180A1 (en) * | 2010-01-20 | 2011-07-21 | Microsoft Corporation | Communication sessions among devices and interfaces with mixed capabilities |
| CN102257566A (en) * | 2008-12-19 | 2011-11-23 | 皇家飞利浦电子股份有限公司 | Method and system for adapting communications |
| CN102318230A (en) * | 2009-02-03 | 2012-01-11 | Itec东京株式会社 | Method for detecting distributed peer to transmit multimedia data in wireless peer-to-peer network |
| CN102355748A (en) * | 2010-06-02 | 2012-02-15 | 索尼公司 | Method for determining a processed audio signal and a handheld device |
| WO2012020394A2 (en) * | 2010-08-11 | 2012-02-16 | Bone Tone Communications Ltd. | Background sound removal for privacy and personalization use |
| US20120078397A1 (en) * | 2010-04-08 | 2012-03-29 | Qualcomm Incorporated | System and method of smart audio logging for mobile devices |
| CN102543096A (en) * | 2011-12-26 | 2012-07-04 | 上海聚力传媒技术有限公司 | Method and device for suppressing scene noise during media file playing |
| CN102779525A (en) * | 2012-07-23 | 2012-11-14 | 华为终端有限公司 | Noise reduction method and terminal |
-
2012
- 2012-11-07 CN CN201210442169.3A patent/CN103811013B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1748249A (en) * | 2003-02-12 | 2006-03-15 | 松下电器产业株式会社 | Intermediary for speech processing in network environments |
| US20040158457A1 (en) * | 2003-02-12 | 2004-08-12 | Peter Veprek | Intermediary for speech processing in network environments |
| US20070067423A1 (en) * | 2003-05-22 | 2007-03-22 | Anthony Dauguet | Context management system for a network including a heterogenous set of terminals |
| CN1753084B (en) * | 2004-09-23 | 2010-06-23 | 纽昂斯通讯公司 | Multi-channel adaptive speech signal processing with noise reduction |
| CN101193139A (en) * | 2006-11-20 | 2008-06-04 | 鸿富锦精密工业（深圳）有限公司 | A method and its mobile phone for filtering environmental noise |
| CN101193384A (en) * | 2006-11-20 | 2008-06-04 | 鸿富锦精密工业（深圳）有限公司 | Method and mobile phone for filtering environmental noise via pattern recognition |
| CN101911724A (en) * | 2008-03-18 | 2010-12-08 | 高通股份有限公司 | Speech enhancement using multiple microphones on multiple devices |
| CN102257566A (en) * | 2008-12-19 | 2011-11-23 | 皇家飞利浦电子股份有限公司 | Method and system for adapting communications |
| CN102318230A (en) * | 2009-02-03 | 2012-01-11 | Itec东京株式会社 | Method for detecting distributed peer to transmit multimedia data in wireless peer-to-peer network |
| CN201417603Y (en) * | 2009-06-23 | 2010-03-03 | 深圳市奥尼电子工业有限公司 | Double microphone speech processor |
| CN101699831A (en) * | 2009-10-23 | 2010-04-28 | 深圳华为通信技术有限公司 | Terminal speech transmitting method, system and equipment |
| US20110179180A1 (en) * | 2010-01-20 | 2011-07-21 | Microsoft Corporation | Communication sessions among devices and interfaces with mixed capabilities |
| US20120078397A1 (en) * | 2010-04-08 | 2012-03-29 | Qualcomm Incorporated | System and method of smart audio logging for mobile devices |
| CN102355748A (en) * | 2010-06-02 | 2012-02-15 | 索尼公司 | Method for determining a processed audio signal and a handheld device |
| WO2012020394A2 (en) * | 2010-08-11 | 2012-02-16 | Bone Tone Communications Ltd. | Background sound removal for privacy and personalization use |
| CN102543096A (en) * | 2011-12-26 | 2012-07-04 | 上海聚力传媒技术有限公司 | Method and device for suppressing scene noise during media file playing |
| CN102779525A (en) * | 2012-07-23 | 2012-11-14 | 华为终端有限公司 | Noise reduction method and terminal |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015176372A1 (en) * | 2014-05-22 | 2015-11-26 | 中兴通讯股份有限公司 | Method for reducing communication noise of intelligent terminal and intelligent terminal thereof |
| CN104092802A (en) * | 2014-05-27 | 2014-10-08 | 中兴通讯股份有限公司 | Method and system for de-noising audio signal |
| CN110070881A (en) * | 2014-06-14 | 2019-07-30 | 宝利通公司 | For reducing the acoustics circumference for the noise that communication equipment is transmitted in open environment |
| CN105227725A (en) * | 2014-07-02 | 2016-01-06 | 中兴通讯股份有限公司 | Realize method and the terminal of call de-noising |
| CN105654960A (en) * | 2015-09-21 | 2016-06-08 | 宇龙计算机通信科技(深圳)有限公司 | Terminal sound denoising processing method and apparatus thereof |
| CN105635379B (en) * | 2015-12-29 | 2020-02-14 | 腾讯科技（深圳）有限公司 | Noise suppression method and device |
| CN105635379A (en) * | 2015-12-29 | 2016-06-01 | 腾讯科技（深圳）有限公司 | Noise suppression method and device |
| CN107623783A (en) * | 2017-09-11 | 2018-01-23 | 安徽省未来博学信息技术有限公司 | A kind of call voice optimizes system |
| CN111052228A (en) * | 2017-09-28 | 2020-04-21 | 哈曼国际工业有限公司 | Method and apparatus for speech recognition |
| CN111052228B (en) * | 2017-09-28 | 2023-09-05 | 哈曼国际工业有限公司 | Method and device for speech recognition |
| CN110335589A (en) * | 2018-03-29 | 2019-10-15 | 松下电器产业株式会社 | Speech translation apparatus, voice translation method and recording medium |
| CN109817209A (en) * | 2019-01-16 | 2019-05-28 | 深圳市友杰智新科技有限公司 | A kind of intelligent speech interactive system based on two-microphone array |
| CN112303829A (en) * | 2020-09-18 | 2021-02-02 | 珠海格力电器股份有限公司 | Wind noise intelligent wind speed adjusting method and system, air conditioner, remote controller and storage medium |
| CN112303829B (en) * | 2020-09-18 | 2021-11-30 | 珠海格力电器股份有限公司 | Wind noise intelligent wind speed adjusting method and system, air conditioner, remote controller and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103811013B (en) | 2017-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103811013A (en) | Noise suppression method, device thereof, electronic equipment and communication processing method | |
| CN106910500B (en) | Method and device for voice control of device with microphone array | |
| US20190244628A1 (en) | Device and method for cancelling echo | |
| EP3923273B1 (en) | Voice recognition method and device, storage medium, and air conditioner | |
| CN108681440A (en) | A kind of smart machine method for controlling volume and system | |
| US20160078880A1 (en) | Systems and Methods for Restoration of Speech Components | |
| CN108962263A (en) | A kind of smart machine control method and system | |
| WO2016112113A1 (en) | Utilizing digital microphones for low power keyword detection and noise suppression | |
| CN110780741B (en) | Model training method, application running method, device, medium and electronic equipment | |
| CN102197422A (en) | Audio source proximity estimation using sensor array for noise reduction | |
| WO2020088153A1 (en) | Speech processing method and apparatus, storage medium and electronic device | |
| CN103137129A (en) | Voice recognition method and electronic device | |
| CN106572418A (en) | Voice assistant expansion device and working method therefor | |
| CN105654767A (en) | Station-arrival reminding processing method, device and terminal | |
| CN103886861A (en) | Method for controlling electronic equipment and electronic equipment | |
| CN109087662B (en) | Echo cancellation method and device | |
| CN107680594A (en) | A kind of distributed intelligence voice collecting identifying system and its collection and recognition method | |
| US20120004913A1 (en) | Method and apparatus for controlling operation of portable terminal using microphone | |
| CN103346897A (en) | Group joining method and relative device and system | |
| CN101867742A (en) | Television system based on sound control | |
| CN103873978A (en) | Information processing method and electronic equipment | |
| CN105719672A (en) | Recording scene mode switching method and recording scene mode switching mode device | |
| CN107358958A (en) | Intercommunication method, apparatus and system | |
| CN106953962A (en) | A kind of call recording method and device | |
| CN101950564A (en) | Remote digital voice acquisition, analysis and identification system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |