CN106936991A - The method and terminal of a kind of automatic regulating volume - Google Patents
The method and terminal of a kind of automatic regulating volume Download PDFInfo
- Publication number
- CN106936991A CN106936991A CN201710124718.5A CN201710124718A CN106936991A CN 106936991 A CN106936991 A CN 106936991A CN 201710124718 A CN201710124718 A CN 201710124718A CN 106936991 A CN106936991 A CN 106936991A
- Authority
- CN
- China
- Prior art keywords
- parameter
- feature parameter
- noise
- environment
- speech model
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 230000001105 regulatory effect Effects 0.000 title claims abstract description 20
- 238000001514 detection method Methods 0.000 claims abstract description 49
- 230000002159 abnormal effect Effects 0.000 claims description 22
- 238000004364 calculation method Methods 0.000 claims description 7
- 241000208340 Araliaceae Species 0.000 claims description 5
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims description 5
- 235000003140 Panax quinquefolius Nutrition 0.000 claims description 5
- 235000008434 ginseng Nutrition 0.000 claims description 5
- 238000007689 inspection Methods 0.000 claims 1
- 230000006870 function Effects 0.000 description 60
- 238000010586 diagram Methods 0.000 description 17
- 238000003860 storage Methods 0.000 description 7
- 230000006399 behavior Effects 0.000 description 5
- 230000007613 environmental effect Effects 0.000 description 4
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000013499 data model Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/72—Mobile telephones; Cordless telephones, i.e. devices for establishing wireless links to base stations without route selection
- H04M1/724—User interfaces specially adapted for cordless or mobile telephones
- H04M1/72448—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions
- H04M1/72454—User interfaces specially adapted for cordless or mobile telephones with means for adapting the functionality of the device according to specific conditions according to context-related or environment-related conditions
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0272—Voice signal separating
- G10L21/028—Voice signal separating using properties of sound source
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M1/00—Substation equipment, e.g. for use by subscribers
- H04M1/60—Substation equipment, e.g. for use by subscribers including speech amplifiers
- H04M1/6033—Substation equipment, e.g. for use by subscribers including speech amplifiers for providing handsfree use or a loudspeaker mode in telephone sets
- H04M1/6041—Portable telephones adapted for handsfree use
- H04M1/605—Portable telephones adapted for handsfree use involving control of the receiver volume to provide a dual operational mode at close or far distance from the user
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
- G10L2015/0635—Training updating or merging of old and new templates; Mean values; Weighting
- G10L2015/0636—Threshold criteria for the updating
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L2015/088—Word spotting
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
- G10L2021/02087—Noise filtering the noise being separate speech, e.g. cocktail party
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Environmental & Geological Engineering (AREA)
- Computer Networks & Wireless Communication (AREA)
- Telephone Function (AREA)
Abstract
The embodiment of the invention discloses the method and terminal of a kind of automatic regulating volume, methods described includes：The current environment voice signal of detection；The characteristic parameter of environment voice signal is obtained, whether characteristic parameter is non-noise signal for environment-identification voice signal；Whether it is non-noise signal according to characteristic parameter and default speech model environment-identification voice signal, speech model includes the scope of characteristic parameter in non-noise region；If non-noise signal, current volume is reduced to preset value.The method of automatic regulating volume provided in an embodiment of the present invention can effectively prevent user from causing missing or omitting for information because video or audio broadcast sound volume are too high with automatic regulating volume.
Description
Technical field
The present invention relates to electronic technology field, more particularly to a kind of automatic regulating volume method and terminal.
Background technology
With the development of intelligent electronic device, intelligent electronic device brings great convenience to the life of user, especially
It is that smart mobile phone has become one of daily necessities of user.It should be appreciated that user often listens music or viewing to regard using mobile phone
Volume is larger and cause user usually not hear or do not hear other people dialogues with itself when frequently, due to playing music or video
Content and cause missing or omitting for information, and the volume adjusting mode of existing mobile phone is often the control of user's manually operated volume
Panel or operating physical button realize regulation, therefore cannot in time avoid above-mentioned situation by automatic regulating volume in time
Occur.
The content of the invention
The embodiment of the present invention provides a kind of method and terminal of automatic regulating volume, can be reduced and used with automatic regulating volume
Family because the broadcast sound volume of video or audio is too high cause to be linked up face to face with other people when there is the situation missed or omit of information.
In a first aspect, the embodiment of the invention provides a kind of method of automatic regulating volume, the above method includes：
The current environment voice signal of detection；The characteristic parameter of environment voice signal is obtained, characteristic parameter is used for identification ring
Whether border voice signal is non-noise signal；According to characteristic parameter and default speech model environment-identification voice signal whether
It is non-noise signal, speech model includes the scope of characteristic parameter in non-noise region；If non-noise signal, by current sound
Amount is reduced to preset value.
On the other hand, a kind of terminal is the embodiment of the invention provides, the terminal includes：Detection unit, collecting unit, identification
Unit and control unit,
Detection unit is used to detect current environment voice signal；Collecting unit is used to obtain the feature of environment voice signal
Whether parameter, characteristic parameter is non-noise signal for environment-identification voice signal；Recognition unit be used for according to characteristic parameter with
And whether default speech model environment-identification voice signal is non-noise signal, speech model includes feature in non-noise region
The scope of parameter；If control unit is used for non-noise signal, current volume is reduced to preset value.
The method and terminal of a kind of automatic regulating volume disclosed in the embodiment of the present invention are by gathering environment voice signal
Characteristic parameter come whether environment-identification voice is non-noise signal, if non-noise signal, then current volume is reduced to
Preset value, can effectively prevent user from causing missing or omitting for information because the broadcast sound volume of video or audio is too high.
Brief description of the drawings
Technical scheme in order to illustrate more clearly the embodiments of the present invention, embodiment will be described below needed for be used
Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are some embodiments of the present invention, general for this area
For logical technical staff, on the premise of not paying creative work, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is a kind of indicative flowchart of the method for automatic regulating volume provided in an embodiment of the present invention；
Fig. 2 is the schematic diagram of the sub-process of Fig. 1 provided in an embodiment of the present invention；
Fig. 3 a are a kind of function curve schematic diagrames of characteristic function formula provided in an embodiment of the present invention；
Fig. 3 b are the function curve schematic diagrames of another characteristic function formula provided in an embodiment of the present invention；
Fig. 4 is a kind of schematic block diagram of terminal provided in an embodiment of the present invention；
Fig. 5 is a kind of schematic block diagram of the recognition unit in Fig. 4 provided in an embodiment of the present invention；
Fig. 6 is the schematic block diagram of another terminal provided in an embodiment of the present invention.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is a part of embodiment of the invention, rather than whole embodiments.Based on this hair
Embodiment in bright, the every other implementation that those of ordinary skill in the art are obtained under the premise of creative work is not made
Example, belongs to the scope of protection of the invention.
It should be appreciated that when using in this specification and in the appended claims, term " including " and "comprising" instruction
The presence of described feature, entirety, step, operation, element and/or component, but be not precluded from one or several further features,
The presence or addition of entirety, step, operation, element, component and/or its set.It is also understood that in this description of the invention
The term for being used is not intended to limit the present invention merely for the sake of the purpose of description specific embodiment.Such as in explanation of the invention
As used in book and appended claims, unless context clearly indicates other situations, otherwise singulative
" one ", " one " and " being somebody's turn to do " are intended to include plural form.
A kind of method for automatically adjusting provided in an embodiment of the present invention, the above method runs on terminal, and terminal is included still
It is not restricted to any electronic product that man-machine interaction can be carried out with user, for example smart mobile phone (such as Android phone, ios
Mobile phone, Windows Phone mobile phones etc.), flat board electric energy, palm electric energy, notebook computer, mobile internet device etc..More than
Electronic equipment is only citing, and non exhaustive, and but terminal provided in an embodiment of the present invention includes is not restricted to above-mentioned electronic equipment.
It should be noted that terminal provided in an embodiment of the present invention should have the element of collection environment voice signal, such as microphone.
1 is please see Figure, is a kind of indicative flowchart of the method for automatic regulating volume provided in an embodiment of the present invention, should
Understand, the automatic regulating volume described in the embodiment of the present invention be based on current on the premise of audio or video resource etc. is played
Carry out.As illustrated, a kind of method of automatic regulating volume provided in an embodiment of the present invention includes following S101~S105：
S101, detects current environment voice signal.
Specifically, the embodiment of the present invention detects environment voice signal preferably by microphone, in other feasible embodiments
Can be used to detect environment voice signal using other electronic components, the embodiment of the present invention does not limit this specifically.
S102, obtains the characteristic parameter of environment voice signal, and whether characteristic parameter is non-noise for environment-identification voice
Signal.
Specifically, the classification of environment voice signal includes noise signal and non-noise signal, noise signal in the present embodiment
Including environmental noise signal and voice noise signal, environmental noise includes building site noise, vehicle noise, office's keyboard noise
Include under dialogue noise and quiet environment under noisy environment Deng, voice noise other people talk with noise, such as subway traffic work
Colleague's talk sound transmitted at a distance in dialogue noise, office noisy in tool etc..It should be appreciated that non-noise signal is on non-
The voice signal of noise signal is stated, being depending on non-noise signal possible be with user-related dialogic voice signal.
Preferably, the characteristic parameter of the environment voice signal for getting includes fisrt feature parameter and second feature parameter.
In other feasible embodiments, the characteristic parameter of the environment voice signal for getting can be a characteristic parameter or three or
The characteristic parameter that person is more than three.The embodiment of the present invention will include that fisrt feature parameter and second feature parameter are with characteristic parameter
Example is illustrated.
Preferably, fisrt feature parameter is the short-time zero-crossing rate of environment voice signal；Second feature parameter is environment voice
The short-time average energy of signal, i.e. fisrt feature parameter are the short-time zero-crossing rates of the whole environment voice signal for obtaining, and second is special
Levy the short-time average energy that parameter is the whole environment voice signal for obtaining.Short-time zero-crossing rate represents voice signal in a frame voice
Number of times of the waveform through zero level；Short-time average energy represents the energy of n moment voice signals.Wherein short-time zero-crossing rate Zn and short
When average energy computing formula it is as follows：
Wherein, N is window long, short-time energy average value be a frame sample value weighted average and；Sgn [] is sign function, x
During (n)≤0, sgn [x (n)]=1；x(n)<When 0, sgn [x (n)]=- 1.
It should be appreciated that in the present embodiment, the short-time zero-crossing rate and short-time average energy of noise signal and non-noise signal
It is different and then can to efficiently differentiate environment voice signal be noise signal or non-noise signal.For example, under quiet environment
The short-time average energy of the voice noise that the colleague transmitted at a distance under voice noise such as office environment talks, its use with collection
The short-time average energy of the related dialogic voice signal in family is different, this be based on it is general with user-related dialogue betide with
Family is separated by relatively closely, and the voice signal that the colleague of distant place talks has been lost energy, therefore its short-time average energy by propagating
Amount is different, and then can distinguish voice noise signal and non-noise signal.And voice noise and environment under noisy environment
Noise is significantly different with the short-time zero-crossing rate and short-time average energy of non-noise, can more efficiently distinguish.
Whether S103, be non-noise signal according to characteristic parameter and default speech model environment-identification voice signal,
Speech model includes the scope of characteristic parameter in non-noise region, if non-noise signal, performs S104, if not non-noise is believed
Number, perform S105.
If specifically, obtain characteristic parameter in the range of characteristic parameter in non-noise region, be non-noise signal；
It is noise signal if the characteristic parameter for obtaining is in the range of characteristic parameter in noise region.Due to noise signal and non-make an uproar
The scope of the corresponding characteristic parameter of message number is different, and included in speech model characteristic parameter in noise region scope and
The scope of characteristic parameter in non-noise region, therefore can identify that environment voice signal is noise according to default speech model
Signal is also non-noise signal.It should be noted that speech model is trained according to some data samples.
Then be included in when characteristic parameter is only a parameter, in speech model in noise region and non-noise region this one
The scope of individual characteristic parameter；If characteristic parameter is two or more, speech model is then included in noise region and non-makes an uproar
The scope of two or more the characteristic parameter in sound region, such as in a kind of mode, first is special in non-noise region
The scope of parameter is levied for (a1, a2), the scope (b1, b2) of second feature parameter, in another way, first in non-noise region
The scope of characteristic parameter is (a1, a2), and when fisrt feature parameter is a1, the scope of second feature parameter is (b1, b2), now
Each specific fisrt feature parameter corresponds to a scope for second feature parameter in non-noise region.
S104, preset value is reduced to by current volume, and flow terminates.
S105, holding current volume is constant, and flow terminates.
If specifically, non-noise signal, then it represents that environment voice be particularly likely that with user-related voice, then should
Reducing current volume can in time hear voice to ensure user；Then it is noise signal if not non-noise signal, then can keeps
Current volume is constant.
It should be appreciated that preset value can be a fixed value in certain embodiments, no matter that is, current volume is how many
Current volume is reduced to a fixed value；In further embodiments, preset value can be the dynamic related to current volume
Value, such as the 1/3 of current volume, the present invention are not limited specifically this.
In sum, the embodiment of the present invention by characteristic parameter come recognition detection to environment voice signal whether be non-making an uproar
Message number determines whether to reduce current volume automatically, due to being depending on non-noise signal and user-related dialogic voice letter
Number, therefore detect environment voice signal and can effectively prevent user because of current sound to reduce current volume after non-noise signal
The broadcast sound volume of frequency or video is too high and miss voice messaging.On the other hand, embodiment of the present invention preferred feature parameter includes
Fisrt feature parameter and second feature parameter, and fisrt feature parameter and second feature parameter are combined for recognizing that noise is believed
Number and non-noise signal, improve identification accuracy.
Preferably, before current volume is reduced to preset value by S015, the above method also includes：
Whether there is default keyword in environment-identification voice signal, default keyword is related to user, for knowing
Whether other environment voice is related to user；And if there is default keyword, then perform current volume is reduced to it is default
Value；If not existing default keyword, keep current volume constant.
Specifically, default keyword is related to user, the name of such as user, the pet name and academic title etc., Ke Yiyong
In the identity for representing user.Whether there is the default keyword can further environment-identification voice in environment-identification voice signal
Whether be with it is user-related, if related to user, then reduce current volume；If not it is related to user, then keep current
Volume is constant.For example can efficiently differentiate be sent in voice noise using which and be separated by close range still with user
The dialogic voice unrelated with user and with user-related dialogic voice.Recognize whether that default keyword can be logical
Keyword identification (KWR) system is crossed, that is, make use of the phonetic features such as energy, cepstrum coefficient in voice.
Preferably, default speech model includes the speech model of several different scenes, and different scenes can be expressed as
In the scene of different location, such as office, family, on subway, the different time scene in same place is also denoted as, such as handled official business
Different time sections in room：When go to work in the morning, during nooning, when go to work in the afternoon.
And then preferably, S103 according to the characteristic parameter and default speech model environment-identification voice signal whether be
Before non-noise signal, the above method also includes：
Obtain current time；
Scene is currently located according to current time and the historical behavior information identifying user for prestoring；
The speech model that matches is searched from several speech models according to scene is currently located.
Specifically, the historical behavior information analysis user according to user is currently located scene, such as according to workaday difference
Time period identifies user in office or family or way on and off duty.It should be appreciated that distinguishing the language of multiple different scenes
Sound model is that, in order to more accurately identify noise signal and non-noise signal, this is based on the noise under the different scenes having
There is larger difference in signal, similarly have the non-noise signal under a different scenes to there is also larger difference.
It is also understood that speech model is the data model obtained according to some data samples of collection.Data sample is
The characteristic parameter gathered under noisy environment or under non-noisy environment.
Preferably, the characteristic parameter for being obtained in S102 includes fisrt feature parameter and second feature parameter.Now, please see Figure
Whether according to characteristic parameter and default speech model environment-identification voice signal it is non-noise signal in 2, S103, including：
S201, the second spy of standard is calculated according to characteristic function formula in fisrt feature parameter and default speech model
Parameter is levied, characteristic function formula is the relational expression of fisrt feature parameter and second feature parameter.
Specifically, preferred fisrt feature parameter is the short-time zero-crossing rate of environment voice signal in the present embodiment；Second feature
Parameter is the short-time average energy of environment voice signal.Now characteristic function formula is the pass of short-time zero-crossing rate and short-time average energy
It is formula, preferably polynomial function, because short-time zero-crossing rate is bent with polynomial function with the relational expression of short-time average energy
Line is similar to.
In certain embodiments, characteristic function formula is by the fisrt feature parameter and second feature in the case of some noises
What the data sample of fisrt feature parameter and second feature parameter in the case of the data sample of parameter and non-noise was obtained makes an uproar
The boundary function of message number and non-noise signal, boundary function is specifically the first spy of noise signal and non-noise signal boundary
Levy the relational expression of function and second feature function.For example, a kind of function curve diagram of characteristic function formula as shown in Figure 3 a is illustrated
Figure, curve L1 is the coordinate curve schematic diagram of characteristic function.The region S1 of curve L1 L1 included below is expressed as non-noise region
The scope of middle fisrt feature parameter and second feature parameter；Curve L1 area above S2 be noise region in fisrt feature parameter with
The scope of second feature parameter.It should be appreciated that dividing non-noise region and noise region according to characteristic function formula in speech model
Division rule be according to data sample derive.The second feature parameter of standard is now calculated according to fisrt feature parameter
Afterwards, can be according to the model of the second feature parameter in the second feature parameter acquiring of standard to noise region and non-noise region
Enclose.
In further embodiments, characteristic function formula is special by the fisrt feature parameter in the case of some noises and second
Levy the relational expression or some non-noise situations of fisrt feature parameter that the data sample of parameter obtains and second feature parameter
Under fisrt feature parameter and second feature parameter the fisrt feature parameter that obtains of data sample and second feature parameter pass
It is formula.Now characteristic function formula be embodied as in noise signal the relational expression of fisrt feature parameter and second feature parameter or
The relational expression of fisrt feature parameter and second feature parameter in non-noise signal.Divide non-according to characteristic function formula in speech model
The division rule of the scope of corresponding fisrt feature parameter and second feature parameter is according to number in noise region and noise region
According to sample derive, preferably division rule for keep fisrt feature functional value it is constant, second feature functional value increase or
Preset value is reduced to generate the scope of second feature parameter in non-noise region and noise region.For example, as shown in Figure 3 b another
A kind of function curve diagram schematic diagram of characteristic function formula, curve L2 is that the coordinate curve of characteristic function in non-noise scene is illustrated
Figure.When fisrt feature parameter is Z1, corresponding second feature parameter is E0 in characteristic function formula, now second in non-noise region
Characteristic parameter scope is (E0- △ E2, E0+ △ E1)；Now, the scope of second feature parameter is less than E0- △ in noise region
E2 and more than E0+ △ E1, wherein △ E2 and △ E1 are obtained according to sample data, be can be the same or different.
S202, second feature parameter in the non-noise region of second feature parameter and preset rules acquisition according to standard
Scope.
Specifically, preset rules divide voice in representing speech model according to the second feature parameter of standard in the present embodiment
The division rule in noise region and non-noise region in model.
As shown in Figure 3 a, if characteristic function formula is boundary function, the of standard is calculated according to fisrt feature parameter z1
After two characteristic parameter E0, second feature parameter E0 and division rule according to standard have got the second spy in non-noise region
The scope of parameter is levied for S1 is less than or equal to E0.
S203, detect obtain second feature parameter whether in non-noise region second feature parameter scope, if
In non-noise region in the range of second feature parameter, perform S204, if not in non-noise region second feature parameter model
In enclosing, S205 is performed；
S204, environment voice signal is non-noise signal.
S205, environment voice signal is not non-noise signal.
Similarly, if characteristic parameter includes three or more than three, it is also applied for the above method and carrys out environment-identification voice letter
Whether number it is non-noise signal, such as when including three characteristic parameters, two of which characteristic parameter is calculated according to preset rules
Go out fourth feature parameter, retraining goes out fourth feature parameter with a remaining characteristic function formula for characteristic parameter.
When further, due to the method for performing automatic regulating volume, speech model has built up, in order to improve
The accuracy of identification, to prevent speech model no longer where applicable from persistently occurring using the situation of the speech model, the above method is also
Including：
If receiving the abnormal information of detection, whether the abnormal frequency of occurrences of recognition detection has exceeded particular value；
If having exceeded particular value, estimating for current detection error according to calculation of characteristic parameters when detection is abnormal every time
Value；And the scope of characteristic parameter in the non-noise region of default speech model is adjusted according to discreet value；
If not less than particular value, not processed.
Specifically, the abnormal information of detection is to be operated according to user and generated, for example, adjust volume display reminding information
Volume will be adjusted, if user does not receive to reduce the request of volume, is considered as one-time detection exception or is reduced user after volume
Elevated volume is also considered as detection exception immediately.It should be understood that detecting whether that more than particular value be in order to the event for preventing detection abnormal is
Once in a while speech model is have adjusted in the case of event.
It should be noted that the discreet value of detection error is the gap of the result and actual result drawn according to characteristic parameter
Obtain.Feature ginseng in scope and noise region further according to characteristic parameter in non-noise region in error amount regulation speech model
Several scopes, makes the speech model after adjustment be applied to current scene, improves detection accuracy.
For example, when characteristic parameter includes fisrt feature parameter and second feature parameter, the preferably discreet value of detection error
It is the second feature parameter and the quadratic sum of the deviation of the second feature parameter of standard of acquisition when detection is abnormal every time, such as obtains
Second feature parameter and standard second feature parameter difference quadratic sum.Now, it is preferably default according to discreet value regulation
Speech model non-noise region in the scope of characteristic parameter be specifically：Fisrt feature parameter and the is adjusted according to discreet value
The characteristic function formula of two characteristic parameters.
It should be appreciated that detection accuracy can be improved by monitor and detection exception and dynamic regulation speech model.
4 are please see Figure, is a kind of schematic block diagram of terminal provided in an embodiment of the present invention, the side of above-mentioned automatic regulating volume
Method runs on terminal, as illustrated, the terminal 400 includes：Detection unit 401, collecting unit 402, recognition unit 403 and control
Unit processed 404.
Wherein, detection unit 401, for detecting current environment voice signal.
Collecting unit 402, the characteristic parameter for obtaining environment voice signal, characteristic parameter is for environment-identification voice
No is non-noise signal.
Specifically, the classification of environment voice signal includes noise signal and non-noise signal, noise signal in the present embodiment
Including environmental noise signal and voice noise signal, environmental noise includes building site noise, vehicle noise, office's keyboard noise
Include under dialogue noise and quiet environment under noisy environment Deng, voice noise other people talk with noise.It should be appreciated that non-noise
Signal is the voice signal of non-above-mentioned noise signal, and being depending on non-noise signal possible be with user-related dialogic voice signal.
Preferably, the characteristic parameter of the environment voice signal for getting includes fisrt feature parameter and second feature parameter.
In other feasible embodiments, the characteristic parameter of the environment voice signal for getting can be a characteristic parameter or three or
The characteristic parameter that person is more than three.The embodiment of the present invention will include that fisrt feature parameter and second feature parameter are with characteristic parameter
Example is illustrated.
Preferably, fisrt feature parameter is the short-time zero-crossing rate of environment voice signal；Second feature parameter is environment voice
The short-time average energy of signal, i.e. fisrt feature parameter are the short-time zero-crossing rates of the whole environment voice signal for obtaining, and second is special
Levy the short-time average energy that parameter is the whole environment voice signal for obtaining.Short-time zero-crossing rate represents voice signal in a frame voice
Number of times of the waveform through zero level；Short-time average energy represents the energy of n moment voice signals.Wherein short-time zero-crossing rate Zn and short
When average energy computing formula it is as follows：
Wherein, N is window long, short-time energy average value be a frame sample value weighted average and；Sgn [] is sign function, x
During (n)≤0, sgn [x (n)]=1；x(n)<When 0, sgn [x (n)]=- 1.
It should be appreciated that in the present embodiment, the short-time zero-crossing rate and short-time average energy of noise signal and non-noise signal
It is different and then can to efficiently differentiate environment voice signal be noise signal or non-noise signal.
Recognition unit 403, for according to characteristic parameter and default speech model environment-identification voice signal whether be
Non- noise signal, speech model includes the scope of characteristic parameter in non-noise region.
Control unit 404, if for non-noise signal, current volume is reduced into preset value；And be additionally operable to, if
It is not non-noise signal, control keeps current volume constant.
If specifically, obtain characteristic parameter in the range of characteristic parameter in non-noise region, be non-noise signal；
It is noise signal if the characteristic parameter for obtaining is in the range of characteristic parameter in noise region.Due to noise signal and non-make an uproar
The scope of the corresponding characteristic parameter of message number is different, and included in speech model characteristic parameter in noise region scope and
The scope of characteristic parameter in non-noise region, therefore can identify that environment voice signal is noise according to default speech model
Signal is also non-noise signal.It should be noted that speech model is trained according to some data samples.
Then be included in when characteristic parameter is only a parameter, in speech model in noise region and non-noise region this one
The scope of individual characteristic parameter；If characteristic parameter is two or more, speech model is then included in noise region and non-makes an uproar
The scope of two or more the characteristic parameter in sound region, such as in a kind of mode, first is special in non-noise region
The scope of parameter is levied for (a1, a2), the scope (b1, b2) of second feature parameter, in another way, first in non-noise region
The scope of characteristic parameter is (a1, a2), and when fisrt feature parameter is a1, the scope of second feature parameter is (b1, b2), now
Each specific fisrt feature parameter corresponds to a scope for second feature parameter in non-noise region.
It should be appreciated that preset value can be a fixed value in certain embodiments, no matter that is, current volume is how many
Current volume is reduced to a fixed value；In further embodiments, preset value can be the dynamic related to current volume
Value, such as the 1/3 of current volume, the present invention are not limited specifically this.
Preferably, above-mentioned terminal 400 also includes discriminating unit 405.
Discriminating unit 405, in environment-identification voice signal whether there is default keyword, default keyword with
Whether user is related, related to user for environment-identification voice.
Control unit 404, if being additionally operable to have default keyword, then is reduced to preset value by current volume；If no
There is default keyword, control keeps current volume constant.
Specifically, default keyword is related to user, the name of such as user, the pet name and academic title etc., Ke Yiyong
In the identity for representing user.Whether there is the default keyword can further environment-identification voice in environment-identification voice signal
Whether be with it is user-related, if related to user, then reduce current volume；If not it is related to user, then keep current
Volume is constant.
Preferably, default speech model includes the speech model of several different scenes, and different scenes can be expressed as
In the scene of different location, such as office, family, on subway, the different time scene in same place is also denoted as, such as handled official business
Different time sections in room：When go to work in the morning, during nooning, when go to work in the afternoon.And then preferably above-mentioned terminal 400 also includes：Obtain
Take unit 406, matching unit 407 and searching unit 408.
Acquiring unit 406, for obtaining current time.
Matching unit 407, for being currently located field according to current time and the historical behavior information identifying user for prestoring
Scape.
Searching unit 408, is currently located scene and the voice mould for matching is searched from several speech models for basis
Type.
Specifically, the historical behavior information analysis user according to user is currently located scene, such as according to workaday difference
Time period identifies user in office or family or way on and off duty.It should be appreciated that distinguishing the language of multiple different scenes
Sound model is that, in order to more accurately identify noise signal and non-noise signal, this is based on the noise under the different scenes having
There is larger difference in signal, similarly have the non-noise signal under a different scenes to there is also larger difference.
It is also understood that speech model is the data model obtained according to some data samples of collection.Data sample is
The characteristic parameter gathered under noisy environment or under non-noisy environment.
Preferably, the characteristic parameter that collecting unit 402 is obtained includes fisrt feature parameter and second feature parameter.Now,
5 are please see Figure, recognition unit 403 includes：Computing unit 501, results unit 502 and judging unit 503.
Computing unit 501, for being calculated according to characteristic function formula in fisrt feature parameter and default speech model
The second feature parameter of standard, characteristic function formula is the relational expression of fisrt feature parameter and second feature parameter.
Specifically, preferred fisrt feature parameter is the short-time zero-crossing rate of environment voice signal in the present embodiment；Second feature
Parameter is the short-time average energy of environment voice signal.Now characteristic function formula is the pass of short-time zero-crossing rate and short-time average energy
It is formula, preferably polynomial function, because short-time zero-crossing rate is bent with polynomial function with the relational expression of short-time average energy
Line is similar to.
In certain embodiments, characteristic function formula is by the fisrt feature parameter and second feature in the case of some noises
What the data sample of fisrt feature parameter and second feature parameter in the case of the data sample of parameter and non-noise was obtained makes an uproar
The boundary function of message number and non-noise signal, boundary function is specifically the first spy of noise signal and non-noise signal boundary
Levy the relational expression of function and second feature function.For example, a kind of curve map schematic diagram of characteristic function formula as shown in Figure 3 a, bent
Line L1 is the coordinate curve schematic diagram of characteristic function.The region S1 of curve L1 L1 included below is expressed as first in non-noise region
The scope of characteristic parameter and second feature parameter；Curve L1 area above S2 is fisrt feature parameter and the second spy in noise region
Levy the scope of parameter.It should be appreciated that the division in non-noise region and noise region is divided in speech model according to characteristic function formula
Rule is derived according to data sample.After the second feature parameter of standard is now calculated according to fisrt feature parameter, can be with
The scope of second feature parameter acquiring according to standard to the second feature parameter in noise region and non-noise region.
In further embodiments, characteristic function formula is special by the fisrt feature parameter in the case of some noises and second
Levy the relational expression or some non-noise situations of fisrt feature parameter that the data sample of parameter obtains and second feature parameter
Under fisrt feature parameter and second feature parameter the fisrt feature parameter that obtains of data sample and second feature parameter pass
It is formula.Now characteristic function formula be embodied as in noise signal the relational expression of fisrt feature parameter and second feature parameter or
The relational expression of fisrt feature parameter and second feature parameter in non-noise signal.Divide non-according to characteristic function formula in speech model
The division rule of the scope of corresponding fisrt feature parameter and second feature parameter is according to number in noise region and noise region
According to sample derive, preferably division rule for keep fisrt feature functional value it is constant, second feature functional value increase or
Preset value is reduced to generate the scope of second feature parameter in non-noise region and noise region.For example, as shown in Figure 3 b another
A kind of curve map schematic diagram of characteristic function formula, curve L2 is the coordinate curve schematic diagram of characteristic function in non-noise scene.The
When one characteristic parameter is Z1, corresponding second feature parameter is E0 in characteristic function formula, now second feature in non-noise region
Parameter area is (E0- △ E2, E0+ △ E1)；Now, in noise region the scope of second feature parameter be less than E0- △ E2 and
More than E0+ △ E1, wherein △ E2 and △ E1 are obtained according to sample data, be can be the same or different.
Unit 502 is harvested, for obtaining in non-noise region the according to the second feature parameter and preset rules of standard
The scope of two characteristic parameters.
Specifically, preset rules divide voice in representing speech model according to the second feature parameter of standard in the present embodiment
The division rule in noise region and non-noise region in model.
As shown in Figure 3 a, if characteristic function formula is boundary function, the of standard is calculated according to fisrt feature parameter z1
After two characteristic parameter E0, second feature parameter E0 and division rule according to standard have got the second spy in non-noise region
The scope of parameter is levied for S1 is less than or equal to E0.
For example, the curve map schematic diagram of characteristic function formula as shown in Figure 3.If shown curve L is showing for characteristic function formula
It is intended to, the second feature parameter E2 of standard is calculated according to fisrt feature parameter Z1, the second feature parameter E2 of standard adds respectively
Subtract default value △ E1 and △ E2 and generate the scope of the second feature parameter in non-noise region, other scopes are noise region
The scope of second feature parameter, △ E1 and △ E2 can be with identical also different.
Judging unit 503, for detect obtain second feature parameter whether in non-noise region second feature parameter
Scope, if in the range of second feature parameter in non-noise region, environment voice signal be non-noise signal, if not non-
In noise region in the range of second feature parameter, environment voice signal is not non-noise signal.
Similarly, if characteristic parameter includes three or more than three, it is also applied for the above method and carrys out environment-identification voice letter
Whether number it is non-noise signal, such as when including three characteristic parameters, two of which characteristic parameter is calculated according to preset rules
Go out fourth feature parameter, retraining goes out fourth feature parameter with a remaining characteristic function formula for characteristic parameter.
When further, due to the method for performing automatic regulating volume, speech model has built up, in order to improve
The accuracy of identification, to prevent speech model no longer where applicable from persistently occurring using the situation of the speech model, above-mentioned terminal 400
Also include identification unit 409, arithmetic element 410 and adjustment unit 411.
Wherein, unit 409 is identified, if the information abnormal for receiving detection, the abnormal frequency of occurrences of recognition detection is
It is no to have exceeded particular value, wherein, if not less than particular value, not processed.
Arithmetic element 410, if for having exceeded particular value, being gone out currently according to calculation of characteristic parameters when detection is abnormal every time
The discreet value of detection error.
Adjustment unit 411, for characteristic parameter in the non-noise region that default speech model is adjusted according to discreet value
Scope.
Specifically, the abnormal information of detection is to be operated according to user and generated, for example saving volume display reminding information will
Regulation volume, if user does not receive to reduce the request of volume, be considered as one-time detection it is abnormal or reduce after volume user with
I.e. elevated volume is also considered as detection exception.It should be understood that detecting whether that more than particular value be in order to the event for preventing detection abnormal is even
Speech model is have adjusted in the case of your event.
It should be noted that the discreet value of detection error is the gap of the result and actual result drawn according to characteristic parameter
Obtain.Feature ginseng in scope and noise region further according to characteristic parameter in non-noise region in error amount regulation speech model
Several scopes, makes the speech model after adjustment be applied to current scene, improves detection accuracy.
For example, when characteristic parameter includes fisrt feature parameter and second feature parameter, the preferably discreet value of detection error
It is the second feature parameter and the quadratic sum of the deviation of the second feature parameter of standard of acquisition when detection is abnormal every time, such as obtains
Second feature parameter and standard second feature parameter difference quadratic sum.Now, it is preferably default according to discreet value regulation
Speech model non-noise region in the scope of characteristic parameter be specifically：Fisrt feature parameter and the is adjusted according to discreet value
The characteristic function formula of two characteristic parameters.
It should be appreciated that detection accuracy can be improved by monitor and detection exception and dynamic regulation speech model.
It is the schematic block diagram of another terminal that embodiment of the present invention is provided referring to Fig. 6.The present embodiment as depicted
In terminal 600 can include one or several processors 601, or several input units 602, or
Several output devices 603 and memory 604.Above-mentioned processor 601, input unit 602, output device 603 and storage
Device 604 is connected by bus.
Input unit 602 is used to receive the information of input.In implementing, the input unit 602 of the embodiment of the present invention can
Including keyboard, mouse, light device of electrical input, acoustic input dephonoprojectoscope, touch input unit, scanner, microphone etc..
Output device 603 is used for external output information to user.In implementing, the output device of the embodiment of the present invention
603 may include display, loudspeaker, printer etc..
Memory 604 is used to store the routine data with various functions, in implementing, the storage of the embodiment of the present invention
Device 604 can be system storage, such as, and volatile (such as RAM), and non-volatile (such as ROM, flash memory etc.), or
Both combinations.In implementing, the memory 604 of the embodiment of the present invention can also be the external memory storage outside system, than
Such as, disk, CD, tape etc..
Processor 601 is used for the instruction for calling the routine data stored in memory 604 to perform the storage of memory 604,
And perform following operation：
The current environment voice signal of detection；The characteristic parameter of environment voice signal is obtained, characteristic parameter is used for identification ring
Whether border voice is non-noise signal；Whether it is non-according to characteristic parameter and default speech model environment-identification voice signal
Noise signal, speech model includes the scope of characteristic parameter in non-noise region, if non-noise signal, by current volume decrease
As little as preset value, if not non-noise signal, keeps current volume constant.
Preferably, before current volume is reduced to preset value by processor 601, and following operation is specifically performed：
Whether there is default keyword in environment-identification voice signal, default keyword is related to user, for knowing
Whether other environment voice is related to user；And if there is default keyword, then perform current volume is reduced to it is default
Value；If not existing default keyword, keep current volume constant.
Preferably, processor 601 is according to the characteristic parameter and default speech model environment-identification voice signal
It is no for non-noise signal before, and specifically perform following operation：
Obtain current time；Scene is currently located according to current time and the historical behavior information identifying user for prestoring；
And the speech model that matches is searched from several speech models according to scene is currently located.
Preferably, the characteristic parameter for being obtained in S102 includes fisrt feature parameter and second feature parameter, processor 601
When according to characteristic parameter and default speech model environment-identification voice signal whether being non-noise signal, and specifically perform as follows
Operation：
The second feature ginseng of standard is calculated according to characteristic function formula in fisrt feature parameter and default speech model
Number, characteristic function formula is the relational expression of fisrt feature parameter and second feature parameter；Second feature parameter according to standard and
Preset rules obtain the scope of second feature parameter in non-noise region；Detect the second feature parameter for obtaining whether in non-noise
The scope of second feature parameter in region, if in the range of second feature parameter in non-noise region, environment voice signal is
Non- noise signal, if not in the range of second feature parameter in non-noise region, environment voice signal is not non-noise signal.
When further, due to the method for performing automatic regulating volume, speech model has built up, in order to improve
The accuracy of identification, to prevent speech model no longer where applicable from persistently occurring using the situation of the speech model, processor 601 is also
Perform following operation：
If receiving the abnormal information of detection, whether the abnormal frequency of occurrences of recognition detection has exceeded particular value；
If having exceeded particular value, estimating for current detection error according to calculation of characteristic parameters when detection is abnormal every time
Value；And the scope of characteristic parameter in the non-noise region of default speech model is adjusted according to discreet value；
If not less than particular value, not processed.
It should be appreciated that in embodiments of the present invention, alleged processor 601 can be central first processing units (Central
Processing Unit, CPU), the processor 601 can also be other general processors, digital signal processor (Digital
Signal Processor, DSP), application specific integrated circuit (Application Specific Integrated Circuit,
ASIC), ready-made programmable gate array (Field-Programmable Gate Array, FPGA) or other FPGAs
Device, discrete gate or transistor logic, discrete hardware components etc..General processor can be microprocessor or this at
Reason device can also be any conventional processor etc..
It should be noted that the step in present invention method can according to actual needs carry out order adjustment, close
And and delete.
Unit in embodiment of the present invention terminal can according to actual needs be merged, divides and deleted.
It is apparent to those skilled in the art that, for convenience of description and succinctly, the end of foregoing description
End and the specific work process of unit, may be referred to the corresponding process in preceding method embodiment, will not be repeated here.In this Shen
In the several embodiments for please being provided, it should be understood that disclosed terminal and method, can realize by another way.
For example, device embodiment described above is only schematical, for example, the division of the unit, only a kind of logic
Function is divided, and can have other dividing mode when actually realizing, such as several units or component can combine or can be with
Another system is integrated into, or some features can be ignored, or do not perform.In addition, shown or discussed coupling each other
Conjunction or direct-coupling or communication connection can be INDIRECT COUPLING or the communication connection by some interfaces, device or unit, also may be used
To be electric, machinery or other forms are connected.
In addition, during each functional unit in each embodiment of the invention can be integrated in a first processing units,
Can be that unit is individually physically present, or two or more units are integrated in a unit.Above-mentioned collection
Into unit can both be realized in the form of hardware, it would however also be possible to employ the form of SFU software functional unit is realized.It is described integrated
If unit is to realize in the form of SFU software functional unit and as independent production marketing or when using, can store in a meter
In calculation machine read/write memory medium.Based on such understanding, technical scheme is substantially done to prior art in other words
Go out the part of contribution, or all or part of the technical scheme can be embodied in the form of software product, the computer
Software product is stored in a storage medium, including some instructions are used to so that a computer equipment (can be personal meter
Calculation machine, server, or network equipment etc.) perform all or part of step of each embodiment methods described of the invention.And it is preceding
The storage medium stated includes：USB flash disk, mobile hard disk, read-only storage (ROM, Read-Only Memory), random access memory
(RAM, Random Access Memory), magnetic disc or CD etc. are various can be with the medium of store program codes.
The above, is specific embodiment of the invention, but protection scope of the present invention is not limited thereto, any ripe
Know those skilled in the art the invention discloses technical scope in, various equivalent modifications can be readily occurred in or replaced
Change, these modifications or replacement should all be included within the scope of the present invention.Therefore, protection scope of the present invention should be with right
It is required that protection domain be defined.
Claims (10)
1. a kind of method of automatic regulating volume, it is characterised in that including：
The current environment voice signal of detection；
The characteristic parameter of the environment voice signal is obtained, whether the characteristic parameter is non-making an uproar for environment-identification voice signal
Message number；
Whether it is non-noise signal, institute's predicate according to the characteristic parameter and default speech model environment-identification voice signal
Sound model includes the scope of characteristic parameter in non-noise region；
If non-noise signal, current volume is reduced to preset value.
2. method according to claim 1, it is characterised in that the characteristic parameter includes fisrt feature parameter and second feature
Whether parameter, be non-noise signal according to the characteristic parameter and default speech model environment-identification voice signal, including：
The second feature ginseng of standard is calculated according to characteristic function formula in the fisrt feature parameter and default speech model
Number, the characteristic function formula is the relational expression of fisrt feature parameter and second feature parameter, the second feature parameter of the standard
Scope for obtaining second feature parameter in noise region and non-noise region in the speech model；
Second feature parameter and preset rules according to standard obtain the scope of second feature parameter in non-noise region；
Detect obtain the second feature parameter whether in non-noise region second feature parameter scope；
If in the range of second feature parameter in non-noise region, the environment voice signal is non-noise signal.
3. method according to claim 1, it is characterised in that methods described also includes：
If receiving the abnormal information of detection, whether the abnormal frequency of occurrences of recognition detection has exceeded particular value；
If having exceeded particular value, estimating for current detection error according to calculation of characteristic parameters when detection is abnormal every time
Value；
The scope of characteristic parameter in the non-noise region of the default speech model is adjusted according to the discreet value.
4. method according to claim 1, it is characterised in that the default speech model includes several different scenes
Speech model, whether described be non-noise according to the characteristic parameter and default speech model environment-identification voice signal
Before signal, including：
Obtain current time；
Scene is currently located according to the current time and the historical behavior information identifying user for prestoring；
The speech model that matches is searched from several speech models according to the scene that is currently located.
5. method according to claim 1, it is characterised in that before current volume is reduced into preset value, the side
Method also includes：
Whether there is default keyword in environment-identification voice signal, the default keyword is related to user, for knowing
Whether other environment voice is related to user；
If there is default keyword, then perform and described current volume is reduced to preset value.
6. a kind of terminal, it is characterised in that the terminal includes：
Detection unit, for detecting current environment voice signal；
Collecting unit, the characteristic parameter for obtaining the environment voice signal, the characteristic parameter is used for environment-identification voice
Whether signal is non-noise signal；
Recognition unit, for whether being non-making an uproar according to the characteristic parameter and default speech model environment-identification voice signal
Message number, the speech model includes the scope of characteristic parameter in non-noise region；
Control unit, if for non-noise signal, current volume is reduced into preset value.
7. terminal according to claim 6, it is characterised in that the characteristic parameter includes fisrt feature parameter and second special
Levy parameter, the recognition unit, including：
Computing unit, for calculating standard according to characteristic function formula in the fisrt feature parameter and default speech model
Second feature parameter, the characteristic function formula is the relational expression of fisrt feature parameter and second feature parameter, the standard
Second feature parameter is used to obtain the scope of second feature parameter in noise region and non-noise region in the speech model；
Unit is harvested, for according to second feature ginseng in the non-noise region of the second feature parameter and preset rules of standard acquisition
Several scopes；
Judging unit, for detect obtain the second feature parameter whether in non-noise region second feature parameter model
Enclose, if in the range of second feature parameter in non-noise region, the environment voice signal is non-noise signal.
8. terminal according to claim 6, it is characterised in that the terminal also includes：
Identification unit, if the information abnormal for receiving detection, it is specific whether the abnormal frequency of occurrences of recognition detection has exceeded
Value；
Arithmetic element, if for having exceeded particular value, current inspection is gone out according to calculation of characteristic parameters when detection is abnormal every time
Survey the discreet value of error；
Adjustment unit, for characteristic parameter in the non-noise region that the default speech model is adjusted according to the discreet value
Scope.
9. terminal according to claim 6, it is characterised in that the default speech model includes several different scenes
Speech model, the terminal also includes：
Acquiring unit, for obtaining current time；
Matching unit, for being currently located scene according to the current time and the historical behavior information identifying user for prestoring；
Searching unit, the speech model for matching is searched for being currently located scene according to from several speech models.
10. terminal according to claim 6, it is characterised in that the terminal also includes discriminating unit,
The discriminating unit, for whether there is default keyword, the default keyword in environment-identification voice signal
It is related to user, it is whether related to user for environment-identification voice；If there is default keyword, described control unit ought
Preceding volume is reduced to preset value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710124718.5A CN106936991A (en) | 2017-03-03 | 2017-03-03 | The method and terminal of a kind of automatic regulating volume |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710124718.5A CN106936991A (en) | 2017-03-03 | 2017-03-03 | The method and terminal of a kind of automatic regulating volume |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106936991A true CN106936991A (en) | 2017-07-07 |
Family
ID=59424072
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710124718.5A Withdrawn CN106936991A (en) | 2017-03-03 | 2017-03-03 | The method and terminal of a kind of automatic regulating volume |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106936991A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107484000A (en) * | 2017-09-29 | 2017-12-15 | 北京奇艺世纪科技有限公司 | A kind of volume adjusting method of terminal, device and voice remote controller |
| CN107728990A (en) * | 2017-09-30 | 2018-02-23 | 努比亚技术有限公司 | A kind of audio frequency playing method, mobile terminal and computer-readable recording medium |
| CN108064007A (en) * | 2017-11-07 | 2018-05-22 | 苏宁云商集团股份有限公司 | Know method for distinguishing and microcontroller and intelligent sound box for the enhancing voice of intelligent sound box |
| CN108494975A (en) * | 2018-06-27 | 2018-09-04 | 上海连尚网络科技有限公司 | Incoming call response method and equipment |
| CN109408025A (en) * | 2018-10-25 | 2019-03-01 | 北京小米移动软件有限公司 | Audio frequency playing method, device and storage medium |
| CN110428835A (en) * | 2019-08-22 | 2019-11-08 | 深圳市优必选科技股份有限公司 | Voice equipment adjusting method and device, storage medium and voice equipment |
| WO2020224126A1 (en) * | 2019-05-06 | 2020-11-12 | 平安科技（深圳）有限公司 | Facial recognition-based adaptive adjustment method, system and readable storage medium |
| CN112887857A (en) * | 2021-01-25 | 2021-06-01 | 湖南普奇水环境研究院有限公司 | Hearing protection method and system for eliminating reception noise |
| CN112992153A (en) * | 2021-04-27 | 2021-06-18 | 太平金融科技服务（上海）有限公司 | Audio processing method, voiceprint recognition device and computer equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101227759A (en) * | 2007-01-15 | 2008-07-23 | 上海杰得微电子有限公司 | Earphone device with automatic volume control function |
| CN101415045A (en) * | 2007-10-17 | 2009-04-22 | 北京三星通信技术研究有限公司 | Method and apparatus for implementing intelligent automatic level control in communication network |
| CN102915753A (en) * | 2012-10-23 | 2013-02-06 | 华为终端有限公司 | Method for intelligently controlling volume of electronic device and implementation device of method |
-
2017
- 2017-03-03 CN CN201710124718.5A patent/CN106936991A/en not_active Withdrawn
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101227759A (en) * | 2007-01-15 | 2008-07-23 | 上海杰得微电子有限公司 | Earphone device with automatic volume control function |
| CN101415045A (en) * | 2007-10-17 | 2009-04-22 | 北京三星通信技术研究有限公司 | Method and apparatus for implementing intelligent automatic level control in communication network |
| CN102915753A (en) * | 2012-10-23 | 2013-02-06 | 华为终端有限公司 | Method for intelligently controlling volume of electronic device and implementation device of method |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107484000A (en) * | 2017-09-29 | 2017-12-15 | 北京奇艺世纪科技有限公司 | A kind of volume adjusting method of terminal, device and voice remote controller |
| CN107728990A (en) * | 2017-09-30 | 2018-02-23 | 努比亚技术有限公司 | A kind of audio frequency playing method, mobile terminal and computer-readable recording medium |
| CN107728990B (en) * | 2017-09-30 | 2020-09-01 | 苏州红森软件有限公司 | Audio playing method, mobile terminal and computer readable storage medium |
| CN108064007A (en) * | 2017-11-07 | 2018-05-22 | 苏宁云商集团股份有限公司 | Know method for distinguishing and microcontroller and intelligent sound box for the enhancing voice of intelligent sound box |
| CN108494975A (en) * | 2018-06-27 | 2018-09-04 | 上海连尚网络科技有限公司 | Incoming call response method and equipment |
| CN109408025A (en) * | 2018-10-25 | 2019-03-01 | 北京小米移动软件有限公司 | Audio frequency playing method, device and storage medium |
| CN109408025B (en) * | 2018-10-25 | 2021-10-22 | 北京小米移动软件有限公司 | Audio playing method, device and storage medium |
| WO2020224126A1 (en) * | 2019-05-06 | 2020-11-12 | 平安科技（深圳）有限公司 | Facial recognition-based adaptive adjustment method, system and readable storage medium |
| CN110428835A (en) * | 2019-08-22 | 2019-11-08 | 深圳市优必选科技股份有限公司 | Voice equipment adjusting method and device, storage medium and voice equipment |
| CN112887857A (en) * | 2021-01-25 | 2021-06-01 | 湖南普奇水环境研究院有限公司 | Hearing protection method and system for eliminating reception noise |
| CN112992153A (en) * | 2021-04-27 | 2021-06-18 | 太平金融科技服务（上海）有限公司 | Audio processing method, voiceprint recognition device and computer equipment |
| CN112992153B (en) * | 2021-04-27 | 2021-08-17 | 太平金融科技服务（上海）有限公司 | Audio processing method, voiceprint recognition device and computer equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106936991A (en) | The method and terminal of a kind of automatic regulating volume | |
| CN107731223B (en) | Voice activity detection method, related device and equipment | |
| WO2021208287A1 (en) | Voice activity detection method and apparatus for emotion recognition, electronic device, and storage medium | |
| CN109087670B (en) | Emotion analysis method, system, server and storage medium | |
| CN107644642B (en) | Semantic recognition method and device, storage medium and electronic equipment | |
| WO2018232233A1 (en) | Data update method, client, and electronic device | |
| CN107610698A (en) | A kind of method for realizing Voice command, robot and computer-readable recording medium | |
| CN103971681A (en) | Voice recognition method and system | |
| CN107193527A (en) | A kind of audio control method, terminal and computer-readable recording medium | |
| CN113707173B (en) | Voice separation method, device, equipment and storage medium based on audio segmentation | |
| CN112652306A (en) | Voice wake-up method and device, computer equipment and storage medium | |
| CN110246502A (en) | Voice noise reduction method and device and terminal equipment | |
| CN108924343A (en) | Control method of electronic device, device, storage medium and electronic equipment | |
| CN112951243A (en) | Voice awakening method, device, chip, electronic equipment and storage medium | |
| CN113593597A (en) | Voice noise filtering method and device, electronic equipment and medium | |
| CN113223554A (en) | Wind noise detection method, device, equipment and storage medium | |
| CN112669822A (en) | Audio processing method and device, electronic equipment and storage medium | |
| CN112669837B (en) | Awakening method and device of intelligent terminal and electronic equipment | |
| CN112581937A (en) | Method and device for acquiring voice instruction | |
| CN113658586A (en) | Training method of voice recognition model, voice interaction method and device | |
| CN111755029B (en) | Voice processing method, device, storage medium and electronic equipment | |
| CN112289311A (en) | Voice wake-up method and device, electronic equipment and storage medium | |
| CN105788590A (en) | Speech recognition method, device, mobile terminal | |
| CN114121022A (en) | Voice wake-up method and device, electronic equipment and storage medium | |
| CN114399992A (en) | Voice instruction response method, device and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20170707 |
|
| WW01 | Invention patent application withdrawn after publication |