Summary of the invention
In view of this, this application provides a kind of earphone noise-reduction method, device, equipment and readable storage medium storing program for executing.It can solveCertainly existing earphone noise-reduction method eliminates whole external environmental sounds, wears the disadvantage that object brings great inconvenience to earphone.
To achieve the goals above, it is proposed that scheme it is as follows:
A kind of earphone noise-reduction method, comprising:
Obtain ambient sound data；
The content for whether wearing subject interests in the ambient sound data comprising earphone determined；
If so, the content in the ambient sound data in addition to the earphone wears the content of subject interests is disappearedIt removes, and plays the content that the earphone wears subject interests.
Preferably, the content of subject interests, packet whether are worn in the determination ambient sound data comprising earphoneIt includes:
According to preconfigured information of interest corresponding with earphone wearing object, the ambient sound data are determinedIn whether include voice data corresponding with the information of interest.
Preferably, the information of interest is interest word, described to wear object pair with the earphone according to preconfiguredThe information of interest answered determines in the ambient sound data whether include voice data corresponding with the information of interest,Include:
By the ambient sound data conversion at identification text；
It is searched whether in the identification text there are preconfigured interest word corresponding with earphone wearing object,If it exists, it is determined that include voice data corresponding with the information of interest in the ambient sound data, if it does not exist, thenIt determines and does not include voice data corresponding with the information of interest in the ambient sound data.
Preferably, the information of interest is interest vocal print, described to wear object with the earphone according to preconfiguredWhether corresponding information of interest determines in the ambient sound data comprising sound number corresponding with the information of interestAccording to, comprising:
It determines in the ambient sound data, if there are the corresponding voice datas of the interest vocal print；
If it exists, it is determined that it include voice data corresponding with the information of interest in the ambient sound data, ifIt is not present, it is determined that do not include voice data corresponding with the information of interest in the ambient sound data.
Preferably, the content of subject interests, packet whether are worn in the determination ambient sound data comprising earphoneIt includes:
The characteristic information that the earphone wears object is obtained, the characteristic information and the earphone are worn in subject interestsHold related；
Determine whether the ambient sound data match with the characteristic information, if so, determining the ambient soundThe content of subject interests is worn in data comprising earphone, if not, it is determined that do not include earphone in the ambient sound data and wearWear the content of subject interests.
Preferably, if determining the ring according to preconfigured information of interest corresponding with earphone wearing objectDo not include voice data corresponding with the information of interest in the voice data of border, then this method further include:
The characteristic information that the earphone wears object is obtained, the characteristic information and the earphone are worn in subject interestsHold related；
Determine whether the ambient sound data match with the characteristic information, if so, determining the ambient soundThe content of subject interests is worn in data comprising earphone, if not, it is determined that do not include earphone in the ambient sound data and wearWear the content of subject interests.
Preferably, the characteristic information for obtaining the earphone and wearing object, comprising:
The personal information that the pre-set earphone wears object is obtained, as characteristic information；
Whether the determination ambient sound data match with the characteristic information, comprising:
Obtain label corresponding with the personal information；
It determines whether comprising the label in the corresponding identification text of the ambient sound data, if then determining matching,If not, it is determined that mismatch.
Preferably, the characteristic information for obtaining the earphone and wearing object, comprising:
It obtains the earphone and wears the material time point for including in subjects history speech content and event information, as featureInformation；
Whether the determination ambient sound data match with the characteristic information, comprising:
It whether determines in the corresponding identification text of the ambient sound data comprising the material time point and/or the thingPart information, if then determining matching, if not, it is determined that mismatch.
Preferably, the characteristic information for obtaining the earphone and wearing object, comprising:
It obtains and wears the pass corresponding with interest-degree of interest content determined by subjects history speech content according to the earphoneSystem, as characteristic information；
Whether the determination ambient sound data match with the characteristic information, comprising:
It whether determines in the corresponding identification text of the ambient sound data comprising target interest content；
If it is determined that comprising target interest content, then the corresponding relationship according to the interest content and interest-degree, described in inquiryThe corresponding target interest-degree of target interest content；
According to the size relation of the target interest-degree and default interest-degree threshold value, the ambient sound data and institute are determinedState the match condition of characteristic information.
Preferably, the content for playing the earphone and wearing subject interests, comprising:
Identify that earphone described in the ambient sound data wears the corresponding target vocal print of content of subject interests；
Into the dialogue mode with the target vocal print corresponding objects, the dialogue mode includes: by the ambient soundThe corresponding voice data of target vocal print described in data plays.
Preferably, the content for playing the earphone and wearing subject interests, further includes:
Record starts to obtain the initial time of the corresponding voice data of the target vocal print；
Detection is from the setting time length after the initial time, if gets the corresponding sound of the target vocal printData；
If it is not, then exiting the dialogue mode with the target vocal print corresponding objects.
A kind of noise reduction device of earphone, comprising:
Ambient sound data capture unit, for obtaining ambient sound data；
Content of interest judging unit, it is whether emerging comprising earphone wearing object sense in the ambient sound data for determiningThe content of interest；
Noise canceling unit, for including the content of earphone wearing subject interests in determining the ambient sound dataWhen, the content in the ambient sound data in addition to the earphone wears the content of subject interests is eliminated；
Content of interest broadcast unit, for including that earphone wears subject interests in determining the ambient sound dataContent when, play the content that the earphone wears subject interests.
Preferably, the content of interest judging unit includes:
First content of interest judgment sub-unit, for according to preconfigured sense corresponding with earphone wearing objectWhether interest information determines in the ambient sound data comprising voice data corresponding with the information of interest.
Preferably, the information of interest is interest word, and the first content of interest judgment sub-unit includes:
It identifies text acquiring unit, is used for the ambient sound data conversion into identification text；
Interest word searching unit is worn there are preconfigured with the earphone for searching whether in the identification textWear the corresponding interest word of object, and if it exists, then determine in the ambient sound data comprising corresponding with the information of interestVoice data, if it does not exist, it is determined that do not include sound number corresponding with the information of interest in the ambient sound dataAccording to.
Preferably, the information of interest is interest vocal print, and the first content of interest judgment sub-unit includes:
Interest vocal print determination unit determines in the ambient sound data, if there are the corresponding sound of the interest vocal printSound data；If it exists, it is determined that include voice data corresponding with the information of interest in the ambient sound data, if notIn the presence of, it is determined that do not include voice data corresponding with the information of interest in the ambient sound data.
Preferably, above-mentioned content of interest judging unit includes:
Second content of interest judgment sub-unit wears the characteristic information of object, the feature for obtaining the earphoneInformation is related to earphone wearing subject interests content；
Third content of interest judgment sub-unit, for determine the ambient sound data whether with the characteristic information phaseMatching, if so, determining the content for wearing subject interests in the ambient sound data comprising earphone, if not, it is determined that instituteIt states and does not include the content that earphone wears subject interests in ambient sound data.
Preferably, above-mentioned content of interest judging unit further include:
Second content of interest judgment sub-unit, for determining the ring in the first content of interest judgment sub-unitWhen not including voice data corresponding with the information of interest in the voice data of border, the feature that the earphone wears object is obtainedInformation, the characteristic information are related to earphone wearing subject interests content；
Third content of interest judgment sub-unit, determine the ambient sound data whether with the characteristic information phaseMatch, if so, determining the content for wearing subject interests in the ambient sound data comprising earphone, if not, it is determined that describedThe content that earphone wears subject interests is not included in ambient sound data.
Preferably, the second content of interest judgment sub-unit, comprising:
Fisrt feature information acquisition unit is worn the personal information of object for obtaining the pre-set earphone, is madeIt is characterized information；
The third content of interest judgment sub-unit, comprising:
Label acquiring unit, for obtaining label corresponding with the personal information；
Label lookup unit, for whether determining in the corresponding identification text of the ambient sound data comprising the markLabel, if then determining matching, if not, it is determined that mismatch.
Preferably, the second content of interest judgment sub-unit, comprising:
Second feature information acquisition unit wears the key for including in subjects history speech content for obtaining the earphoneTime point and event information, as characteristic information；
The third content of interest judgment sub-unit, comprising:
Text reference unit is identified, for whether determining in the corresponding identification text of the ambient sound data comprising describedMaterial time point and/or the event information, if then determining matching, if not, it is determined that mismatch.
Preferably, the second content of interest judgment sub-unit, comprising:
Third feature information acquisition unit is worn determined by subjects history speech content for obtaining according to the earphoneThe corresponding relationship of interest content and interest-degree, as characteristic information；
The third content of interest judgment sub-unit, comprising:
Interest content judging unit, for whether determining in the corresponding identification text of the ambient sound data comprising targetInterest content；
Target interest-degree query unit, if for above-mentioned interest content judging unit judging result be it is yes, according to described inThe corresponding relationship of interest content and interest-degree inquires the corresponding target interest-degree of the target interest content；
Interest-degree comparing unit is determined for the size relation according to the target interest-degree and default interest-degree threshold valueThe match condition of the ambient sound data and the characteristic information.
Preferably, the content of interest broadcast unit, comprising:
Target Application on Voiceprint Recognition unit, earphone described in the ambient sound data wears the interior of subject interests for identificationHold corresponding target vocal print；
Dialogue mode enters unit, for entering the dialogue mode with the target vocal print corresponding objects, the dialogue mouldFormula includes: to play the corresponding voice data of target vocal print described in the ambient sound data.
Preferably, the content of interest broadcast unit, further includes:
Initial time recording unit, when for recording the starting for starting to obtain the corresponding voice data of the target vocal printBetween；
Duration detection unit is set, for detecting from the setting time length after the initial time, if getThe corresponding voice data of the target vocal print；
Dialogue mode exits unit, for the target vocal print detection unit testing result be it is no when, exit with it is describedThe dialogue mode of target vocal print corresponding objects.
A kind of earphone noise reducing apparatus, including memory and processor；
The memory, for storing program；
The processor realizes each step of earphone noise-reduction method any one of as described above for executing described programSuddenly.
A kind of readable storage medium storing program for executing is stored thereon with computer program, real when the computer program is executed by processorNow each step of earphone noise-reduction method any one of as described above.
It can be seen from the above technical scheme that earphone noise-reduction method provided by the embodiments of the present application, obtains earphone and wearsWhether the external environmental sounds data in object local environment determine in the ambient sound data comprising earphone wearing object senseThe content of interest, here, interested content refer to that earphone wears object and wants the content heard, further, when determining ringAfter the content for wearing subject interests in the voice data of border comprising earphone, worn the earphone is removed in the ambient sound dataContent except the content of subject interests eliminates, and plays the content that the earphone wears subject interests.From the foregoing, it will be observed thatThis method plays earphone and wears the content of subject interests, and the content in addition to earphone wears the content of subject interests is disappearedIt removes, so that earphone is worn the interference that object avoids ambient noise as a result, and be at quiet state, and be avoided that earphoneIt wears object and misses the sound-content for wanting to hear, improve the convenience that earphone wears object.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, completeSite preparation description, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.It is based onEmbodiment in the application, it is obtained by those of ordinary skill in the art without making creative efforts every otherEmbodiment shall fall in the protection scope of this application.
Earphone noise-reduction method provided by the present application, can be applied to noise cancelling headphone, need to wear noise reduction ear suitable for userMachine carries out ambient sound under the scene of noise reduction process, such as when user is in noisy public place, needs to wear noise reductionEarphone is to experience quiet rest environment.At this point, existing earphone noise-reduction method is that whole ambient sounds are carried out noise reduction process,But this method can eliminate together the ambient sound that earphone wears subject interests, for example, in the dialogue of other people talkAppearance or broadcasted content.Therefore the application in ambient sound data noise carry out noise reduction process during, by earphoneThe content for wearing subject interests plays, and while capable of guaranteeing that earphone wearing object is in rest state, can also hear ringInterested content in the sound of border.
Next, the application earphone noise-reduction method is introduced in conjunction with attached drawing 1, as Fig. 1 illustrates a kind of earphone noise reductionMethod flow diagram, this method include: in detail
Step S100, ambient sound data are obtained.
Specifically, the external environmental sounds data in earphone wearing object local environment, a kind of optional realization way are obtainedDiameter is setting recording module, when detecting the sound of external environment, starts recording module, carries out to extraneous environmental sound signalAcquisition.
Step S110, the content for whether wearing subject interests in the ambient sound data comprising earphone determined, if so,Execute step S120.
Specifically, the content that earphone wears subject interests refers to that earphone wearing object is wanted to listen in ambient sound dataThe content arrived, for example, the broadcasted content of waiting hall or other people and the conversation content of oneself etc., this step is obtaining environmentAfter voice data, further determine that whether wearing object comprising earphone in ambient sound data wants the content heard.
Step S120, by the content in the ambient sound data in addition to the earphone wears the content of subject interestsIt eliminates, and plays the content that the earphone wears subject interests.
Specifically, after determining the content comprising earphone wearing subject interests in ambient sound data in step S110,Content in ambient sound data in addition to earphone wears the content of subject interests is eliminated as noise, removing methodPrinciple can be identical as existing noise-reduction method, i.e., noise sound data content is neutralized by generating the sound wave opposite with noiseFall, further, plays the content that earphone wears subject interests, allow users to hear the voice data content for wanting to hear.
It can be seen from the above technical scheme that earphone noise-reduction method provided by the embodiments of the present application, obtains earphone and wearsWhether the external environmental sounds data in object local environment determine in the ambient sound data comprising earphone wearing object senseThe content of interest, here, interested content refer to that earphone wears object and wants the sound-content heard, further, when trueDetermine after wearing the content of subject interests in ambient sound data comprising earphone, the earphone will be removed in the ambient sound dataIt wears the content except the content of subject interests to eliminate, and plays the content that the earphone wears subject interests.By upperIt is found that this method plays the content that earphone wears subject interests, and will be in addition to earphone wears the content of subject interestsContent eliminates, and so that earphone is worn the interference that object avoids ambient noise as a result, and had been at quiet state, and can keep awayExempt from earphone wearing object and miss the sound-content for wanting to hear, improves the convenience that earphone wears object.
Next the one embodiment for proposing the application, in determination in the step S110 ambient sound data whetherThe content for wearing subject interests comprising earphone is introduced.
It is aforementioned it is stated that earphone wear subject interests content refer to earphone wear object want the sound heardData content, there are many embodiments for determining the content for whether subject interests being worn comprising earphone in ambient sound data,The present embodiment introduces one of optional embodiment.
Specifically, the embodiment of the present embodiment introduction is based on information of interest, and here, information of interest is to be pre-configured withWear the corresponding information of interest of object with earphone, it is clear that these information of interest are wearing pair with earphone of being arranged in advanceAs the corresponding information of the sound-content for wanting to hear, can by determine in ambient sound data whether include and information of interestCorresponding voice data determines the content for whether wearing subject interests in ambient sound data comprising earphone.
It is different according to the preconfigured type for wearing the corresponding information of interest of object from earphone, determine ambient sound numberIn whether the method comprising voice data corresponding with information of interest include it is a variety of, next introduce respectively wherein optionalTwo kinds of embodiments.
The first, information of interest be interest word.
Here, interest word is that the preconfigured one or more words of object are worn by earphone, may include time, thingPart, personage, place etc., it is to be understood that interest word is that the voice data content that earphone wearing object is wanted to hear includesWord, for example, name, the name of companion that earphone wears object are preconfigured to interest word, or in waiting room scene,If wanting to hear broadcasted content, the keywords such as " passenger ", " broadcast ", " train number " can be preconfigured to interest word in advance.
Based on this, determine in ambient sound data whether include voice data corresponding with information of interest process, canTo include:
Firstly, by the ambient sound data conversion at identification text.
Specifically, the ambient sound data conversion that will acquire by speech recognition technology is at identification text.Here voice is knownOther technology can choose the prior art, and the present embodiment does not repeat them here.
Further, search whether that there are preconfigured corresponding with earphone wearing object in the identification textInterest word.
Specifically, above-mentioned preconfigured interest word is searched in the identification text of ambient sound data, due to interest wordIt is information of interest, so in the presence of lookup result is, it is determined that comprising corresponding with information of interest in ambient sound dataVoice data, if lookup result be there is no, it is determined that in ambient sound data do not include sound corresponding with information of interestSound data.
Show for example, earphone wearing object is wanted to hear the sound-content comprising oneself name, therefore earphone is worn objectName " Li Ming " is preconfigured to interest word, after earphone begins to use, persistently obtains ambient sound data, and by ambient soundData are converted to identification text, further search for wherein wearing the preconfigured interest word " Li Ming " of object with the presence or absence of earphone,If it exists, it is determined that include voice data corresponding with earphone wearing object name in ambient sound data, that is, determine ambient soundIt include voice data corresponding with information of interest in sound data.
Second, information of interest is interest vocal print.
Here, interest vocal print refers to that earphone wears object and wants the corresponding vocal print of the sound heard, for example, earphone is wornWhen object wants to hear companion the conversation content with oneself, the vocal print of companion's sound can be preconfigured to interest soundLine.
At this point, determine in ambient sound data whether include voice data corresponding with information of interest process, can be withInclude:
It determines in the ambient sound data, if there are the corresponding voice datas of the interest vocal print.
Specifically, ambient sound data are searched, determine whether that there are the corresponding sound of preconfigured interest vocal printData.It is understood that interest vocal print is that earphone wears the preconfigured information of interest of object, so when searching ambient soundSound data determine when wherein there is the corresponding voice data of interest vocal print, it is determined that in ambient sound data comprising with it is interestedThe corresponding voice data of information, if lookup result be there is no, it is determined that do not include in ambient sound data and information of interestCorresponding voice data.
Show for example, earphone wearing object wants to hear sound-content when companion speaks to oneself, so by the sound of companionSound is preconfigured to interest vocal print, after earphone begins to use, persistently obtains ambient sound data, determines whether that there are earsMachine wears the preconfigured interest vocal print of object, and if it exists, then determines that the vocal print in ambient sound data comprising companion is correspondingVoice data, that is, determine in ambient sound data comprising voice data corresponding with information of interest.
It should be noted that whether including sound corresponding with information of interest in above two determining ambient sound dataThe method of data may include online or offline two kinds of realization processes.
Wherein during canbe used on line, background server obtains preconfigured information of interest and stores.Earphone is obtainingTake and be sent to server after ambient sound data, by server be determined in ambient sound data whether comprising with it is interestedThe process of the corresponding voice data of information, and definitive result is sent to earphone.
Off-line procedure is that earphone obtains preconfigured information of interest and locally stored, this is based on, in local progressDetermine in ambient sound data whether include voice data corresponding with information of interest process.
Pass through determination using preconfigured interest word or interest vocal print as information of interest based on above two methodWhether there is interest word or interest vocal print in ambient sound data, determine in ambient sound data whether include and letter interestedCorresponding voice data is ceased, because information of interest is that the preconfigured content for wanting to hear with earphone wearing object is relatedInterest word or interest vocal print, so whether may further determine that in ambient sound data according to above-mentioned definitive result comprising earMachine wears the content of subject interests, i.e., if it is determined that including voice data corresponding with information of interest in ambient sound data,It can then determine the content for wearing subject interests in ambient sound data comprising earphone, however, it is determined that do not wrap in ambient sound dataContaining voice data corresponding with information of interest, then it can determine that not including earphone in ambient sound data wears subject interestsContent.
Next, whether the present embodiment introduction step S110, determine in ambient sound data comprising earphone wearing object senseThe optional embodiment of another kind of the content of interest may include:
A1, the characteristic information that the earphone wears object is obtained.
Specifically, characteristic information is related to earphone wearing subject interests content.Usually like for example, earphone wears objectIt watches movie, likes moving, it is believed that in ambient sound data is earphone wearing pair with film or the related topic of movementAs wanting the content heard.Then can information that the hobby that object worn with earphone is related as characteristic information, such as filmAnd the title or movement and sports events of film.Obviously, characteristic information is to wear subject interests content phase with earphoneIt closes.
A2, determine whether the ambient sound data match with the characteristic information.
Specifically, since characteristic information is related to the earphone wearing content of subject interests, so if it is determined that ambient soundData match with characteristic information, it is determined that wear the content of subject interests in ambient sound data comprising earphone, however, it is determined thatAmbient sound data and characteristic information mismatch, it is determined that do not include earphone in ambient sound data and wear the interior of subject interestsHold.
It should be noted that can be further increased secondary when determining that ambient sound data and characteristic information matchThe process of confirmation: whether setting voice prompting or vibration prompting are worn in object confirmation ambient sound data by earphone comprising senseThe content of interest.Obvious increased secondary-confirmation process can make judging result more accurate.
Whether above-mentioned two embodiment describes in two different determining ambient sound data comprising earphone wearing objectThe embodiment of interested content can also combine above two embodiment in the present embodiment, and detailed process can be withInclude:
B1, the corresponding information of interest of object is worn with the earphone according to preconfigured, determine the ambient soundIt whether include voice data corresponding with the information of interest in data.
Specifically, information of interest corresponding with earphone wearing object may include preconfigured interest word or interest soundLine includes voice data corresponding with information of interest in ambient sound data when determining, it is determined that wrap in ambient sound dataThe content of subject interests is worn containing earphone.
If it is determined that not including voice data corresponding with the information of interest in ambient sound data, then B2 is executed.IfIt determines in ambient sound data comprising voice data corresponding with the information of interest, then can determine in ambient sound dataThe content of subject interests is worn comprising earphone.
B2, the characteristic information that the earphone wears object is obtained.
Wherein, characteristic information is related to earphone wearing subject interests content.This step is referred to the implementation of above-mentioned A1Process, this will not be repeated here.
B3, determine whether the ambient sound data match with the characteristic information.
It is understood that if the definitive result of B3 be it is yes, can determine in the ambient sound data comprising earphoneThe content of subject interests is worn, otherwise, can determine that not including earphone in ambient sound data wears the interior of subject interestsHold.
The implementation process of B3 step is referred to above-mentioned A2 step, and this will not be repeated here for detailed process.
It is understood that characteristic information can be there are many optional acquisition modes, for example, can be according to earphone wearing pairThe personal information of elephant determines characteristic information, or wears subjects history speech content according to earphone and determine characteristic information.According to spyInformation acquiring pattern difference is levied, the embodiment for determining that whether ambient sound data match with characteristic information in the present embodiment canWith include it is a variety of, next optional three kinds of embodiments are introduced.
The first,
The personal information that pre-set earphone wears object is obtained, as characteristic information.
Here, personal information refers to pre-set some personal information, such as occupation, age, home address etc..It can be withUnderstand, ambient sound data relevant to these personal information may be the content that earphone wears subject interests.For example,Pre-set professional " programmer " is used as characteristic information, then ambient sound data related with programmer may be that earphone is wornObject is worn to want to hear content.
Based on this, above-mentioned A2 or B3 determine the process whether the ambient sound data match with the characteristic information,May include:
Firstly, obtaining label corresponding with the personal information.
Specifically, the corresponding label of preset personal information is obtained, these labels are all relevant to personal information contentWord, every personal information can correspond to one or more labels.For example, the occupation that earphone wears object is programmer, thenGet " programmer " relevant label, such as " code agriculture ", " program ape ", " programming ".
Further, it determines whether comprising the label in the corresponding identification text of the ambient sound data, if then reallyFixed matching, if not, it is determined that mismatch.
Specifically, it is searched whether in the corresponding identification text of ambient sound data comprising the corresponding label of personal information,When in the corresponding identification text of ambient sound data comprising one or more labels, ambient sound data and characteristic information are determinedMatch, further determines that the content for wearing subject interests in ambient sound data comprising earphone.For example, earphone wears objectOccupation be programmer, when in the corresponding identification text of ambient sound data include label " program ape ", then it is determined that ambient soundSound data match with characteristic information, further determine that in ambient sound data comprising earphone wear subject interests and oneselfThe relevant content of occupation.
When not including label in the corresponding identification text of ambient sound data, ambient sound data and characteristic information are determinedIt mismatches.
Second,
It obtains the earphone and wears the material time point for including in subjects history speech content and event information, as featureInformation.Here, history speech content may include the identification text that the history speech audio collected is converted into.
Specifically, material time point and event information refer to that earphone whithin a period of time wears the history speech of objectThe time point for including in content and event, it is to be understood that material time point and event information are and earphone wearing object senseThe relevant information of the content of interest, it is possible to which the material time point and temporal information that will acquire are as characteristic information.In general,Material time point or event information can be also possible to separately as characteristic information by material time point and corresponding eventInformation is together as one group of characteristic information, for example, earphone wears object discussed " afternoon next Thursday two o'clock before one week with others" two o'clock in afternoon next Thursday " can be then used as material time point by the related content of meeting ", and by " meeting " as crucial with thisTime point corresponding event information.
Further, according to earphone wear subjects history speech content determine material time point and event information itAfterwards, suggestion voice prompt earphone can be set and wear object, if the material time point is set as a time-setting alarm clock.?After obtaining the confirmation that earphone wears object, the material time point can be set to a time-setting alarm clock, and when systemBetween when reaching the time-setting alarm clock moment, earphone can play corresponding prompt tone reminding earphone and wear the object handles material time point pairThe event answered.
When characteristic information includes material time point and event information, above-mentioned A2 or B3 determine that the ambient sound data areThe no process to match with the characteristic information may include:
It whether determines in the corresponding identification text of ambient sound data comprising material time point and/or event information.
A kind of optional mode, as long as believing in the corresponding identification text of ambient sound data comprising material time point or eventWhen at least one of breath, then it can determine that ambient sound data match with characteristic information.For example, ambient sound data are correspondingIt identifies in text to include material time point " two o'clock in afternoon next Thursday ", then can determine ambient sound data and characteristic information phaseMatch.
Another optional mode, when include simultaneously in the corresponding identification text of ambient sound data material time point and withCorresponding event information when, just determine ambient sound data match with characteristic information.For example ambient sound data are correspondingIdentify text in include material time point " two o'clock in afternoon next Thursday " and, time point corresponding event information " meeting " then may be usedTo determine that ambient sound data and characteristic information match.
Since characteristic information is to wear subjects history speech content according to earphone to determine, so when determining ambient sound numberWhen according to matching with characteristic information, it may further determine that in ambient sound data and wearing the interior of subject interests comprising earphoneHold.
It should be noted that can be material time point since material time point and event information have timelinessTime tag is set with event information, the timeliness of the time tag is determined by the timeliness of material time point and event information.
Then determine ambient sound data whether with the process that characteristic information matches may include:
It determines in the corresponding identification text of ambient sound data, if when being in the key in validity period comprising time tagBetween put and event information, however, it is determined that include, it may be considered that successful match, otherwise it is assumed that matching is unsuccessful.
The third,
It obtains and wears the pass corresponding with interest-degree of interest content determined by subjects history speech content according to the earphoneSystem, as characteristic information.Here, history speech content may include the identification text that the history speech audio collected is converted into.
Specifically, interest content includes the items of interest under category of interest, and wherein category of interest may include film, soundPleasure, books etc., items of interest may include movie name, books title and musical designation etc., can be by wearing to earphoneSubjects history speech content is analyzed, determine it includes interest pattern set and items of interest set.By interest patternAny items of interest forms an interest content in any interest pattern, with items of interest set in set.
It is understood that earphone wears object difference possible for the interest level of each interest content, so intoOne step wears subjects history speech content according to earphone and determines the interest-degree of each interest content, determines that the method for interest-degree can be withThere are many, for example number or frequency that interest content occurs in history speech content can be calculated, the present embodiment introduction canA kind of method of calculating interest-degree of choosing is as follows:
1) support of interest content is calculated.
The support of interest content refers to: wearing the interest class that interest content includes in subjects history speech content in earphoneType and items of interest appear in the frequency in a word simultaneously.It is directed to an interest content, if the support of the interest contentIt is high, then it represents that the interest content frequency of occurrences is high, for example, the support of " programmer " and " Java ", are higher than " programmer " and " teaThe support of leaf " indicates that " programmer " and " Java " appears in the frequency in a word simultaneously and be greater than " programmer " and " tealeaves ".Interest pattern and items of interest are expressed as X and Y, the specific formula for calculation of support is as follows:
Wherein, total word number refers to that earphone wears the total quantity for the word for including in subjects history speech content.For example, obtainHistory speech content includes 100 words altogether, wherein the number that " programmer " and " Java " occurs simultaneously in a word is 8 times,Then determine the support of the interest content including " programmer " and " Java " are as follows: support (programmer, JAVA)=8/100=0.08。
2) confidence level of interest content is calculated.
Specifically, the confidence level of interest content refers to: wearing in subjects history speech content and occurs in the interest in earphoneOn the basis of the interest pattern of appearance, probability that the items of interest of the interest content also occurs.Interest pattern is expressed as X, it will be emergingInteresting project is expressed as Y, and the specific formula for calculation of confidence level is as follows:
Wherein, the total degree for X occur refers to that earphone wears the object total degree that X occurs in history speech content.ExampleSuch as, the history speech content of acquisition includes 100 words altogether, wherein the number of the appearance of " programmer " be 20 times, " programmer " andThe number that " Java " occurs simultaneously in a word is 8 times, it is determined that the interest content including " programmer " and " Java " is setReliability are as follows: confidence (programmer, JAVA)=8/20=0.4.
Based on the above method, the support of each interest content is come out with confidence calculations.As the following table 1 illustrates oneThe correspondence table of kind interest content support and confidence level.
Table 1
3) interest-degree of interest content is calculated
It is above-mentioned that the support and confidence level of every interest content has been determined, by the corresponding support of each interest content and setReliability storage.The process for calculating interest-degree, which can directly select the support of interest content or confidence level, to be indicated, alternatively, this methodThe influence for considering support and confidence level simultaneously determines that interest content is corresponding with interest-degree according to support and confidence level and closesSystem.
Specifically, interest-degree refers to that earphone wears object to the interested degree of interest content, by interest pattern and interestProject is expressed as X and Y, and interest-degree usually uses I (X, Y) to indicate that specific calculation formula can be as follows:
Show for example, the interest content of above-mentioned " programmer " and " Java " support be 0.08, confidence level 0.4, then itsInterest-degree can be calculated by above formula:
The corresponding relationship of each interest content and its interest-degree is obtained using the above method, using the corresponding relationship as featureInformation.It is understood that the corresponding relationship of interest content and interest-degree, which reflects earphone, wears sense of the object to interest contentLevel of interest, and corresponding relationship is constantly updated with the increase of history speech content.
Based on this, above-mentioned A2 or B3 determine the process whether the ambient sound data match with the characteristic information,May include:
C1, it whether determines in the corresponding identification text of the ambient sound data comprising target interest content.
Specifically, it is searched in the corresponding identification text of ambient sound data, if emerging comprising the target by above-mentioned determinationInteresting content.Here, as long as determining emerging comprising target interest pattern set or target in the corresponding identification text of ambient sound dataAny one of interesting project set is multinomial, it is determined that includes target interest content in the corresponding identification text of ambient sound data.For example, determining and being wrapped in the corresponding identification text of ambient sound data when being searched in the corresponding identification text of ambient sound data" avenger alliance " in the set of items of interest containing target, then can determine in the corresponding identification text of ambient sound data and includeTarget interest content.Or when searching in the corresponding identification text of ambient sound data, determine that ambient sound data are correspondingIt identifies in text comprising " Hollywood movie " in target interest pattern set, then can determine the corresponding knowledge of ambient sound dataIt include target interest content in other text.
It is understood that if in the corresponding identification text of ambient sound data only including target interest pattern, it can be withThe target interest content for being included as identification text using the interest content comprising target interest pattern；If ambient sound data pairIt only include target items of interest in the identification text answered, then it can be using the interest content comprising target items of interest as identification textThis target interest content for being included；If in the corresponding identification text of ambient sound data simultaneously including target interest pattern and meshItems of interest is marked, then can be included using the interest content comprising target interest pattern and target items of interest as identification textTarget interest content.
C2, if it is determined that comprising target interest content, then the corresponding relationship according to the interest content and interest-degree, inquires instituteState the corresponding target interest-degree of target interest content.
It can be understood that the target interest content of above-mentioned determination may include a plurality of, such as determine ambient sound data pairComprising " the avenger alliance " in target items of interest set in the identification text answered, can be determined by table 1 comprising target interestTwo are had in the target interest of project " avenger alliance ", and the target interest pattern of two target interest contents is respectively" Hollywood movie " and " unrestrained prestige ".
The corresponding target interest-degree of each target interest content may further be obtained by inquiry, when target interest contentIncluding it is a plurality of when, then the corresponding target interest-degree of target interest content can be obtained according to multiple target interest-degrees, such as can be withBy the addition of multiple target interest-degrees or weighted sum.For example, comprising " multiple in the corresponding identification text of current environmental sound dataChou Zhe alliance ", then inquiry includes the target interest content of " avenger alliance ", and discovery one is under " unrestrained prestige " interest pattern" avenger alliance ", another is " the avenger alliance " under " Hollywood movie " interest pattern, at this point, in inquiry target interestHolding corresponding target interest-degree is to be added the interest-degree of two target interest contents, obtains " avenger alliance " corresponding targetInterest-degree.
C3, according to the size relation of the target interest-degree and default interest-degree threshold value, determine the ambient sound dataWith the match condition of the characteristic information.
Specifically, the use habit that object can be worn according to earphone presets interest-degree threshold value, when target interest content pairWhen the target interest-degree answered is more than or equal to default interest-degree threshold value, it can determine that ambient sound data match with characteristic information,Otherwise, it can determine that ambient sound data and characteristic information mismatch.Since target interest content is corresponding with target interest-degreeRelationship is to wear subjects history speech content according to earphone to determine, it is possible to further determine that in ambient sound data and includesThe content of earphone wearing subject interests.
Above describe three kinds of modes for determining whether ambient sound data match with characteristic information, due to characteristic informationIt is to be determined according to the preset personal information of earphone wearing object or history speech content, so characteristic information and earphone are wornThe content of subject interests be it is relevant, when determining that ambient sound data match with characteristic information by any of the above-described kind of methodWhen, then it can determine the content for wearing subject interests in ambient sound data comprising earphone.
It should be noted that the matching process of above-mentioned three kinds of characteristic informations can have, online or offline two kinds were realizedJourney.
Wherein during canbe used on line, background server calls open application program API (ApplicationProgram Interface) interface obtains preset personal information or history speech content, and it obtains characteristic information and stores, intoWhen row matching, the corresponding identification text of ambient sound data is sent to background server by earphone, is carried out by background serverThe process whether determining ambient sound data match with characteristic information is stated, and matching result is sent to earphone.
Off-line procedure is that the characteristic information that background server stores is downloaded to local in advance by earphone, and is based on this, at thisGround carries out the process whether above-mentioned determining ambient sound data match with characteristic information.
Next the embodiment of the present application is proposed, to the content for playing the earphone wearing subject interests in step S120Process is introduced, which may include:
Earphone described in D1, the identification ambient sound data wears the corresponding target vocal print of content of subject interests.
Specifically, the corresponding target vocal print of content that earphone wears subject interests may include one or more, successivelyIt identifies target vocal print, and makes marks to different target vocal prints.For example the scene in more people talk, A and B simultaneously wear earphoneObject is spoken, and the earphone in ambient sound data wears the content of subject interests respectively from A and B, then passes through Application on Voiceprint RecognitionTechnology identifies the vocal print of A and B respectively, successively determines that target vocal print is target vocal print 1 and target vocal print 2.
Distinguishingly, when information of interest is interest vocal print, however, it is determined that it is corresponding that there are interest vocal prints in ambient sound dataVoice data, directly using interest vocal print as earphone wear subject interests the corresponding target vocal print of content.
D2, into the dialogue mode with the target vocal print corresponding objects.
Specifically, the corresponding voice data of target vocal print includes the content that earphone wears subject interests, therefore dialogue modeIt is to play the corresponding voice data of target vocal print each in ambient sound data.
On the basis of the above, into after the dialogue mode of target vocal print corresponding objects, earphone, which wears object, to be heardThe sound of target vocal print corresponding objects, it is to be understood that may no longer include mesh in ambient sound data as time goes byMark the corresponding voice data of vocal print, such as under session operational scenarios, after earphone wears object and object A talk, object A fromIt opens, does not include the corresponding voice data of object A in ambient sound data at this time, worn so playing the earphone in step S120The process of the content of subject interests can also include:
Firstly, record obtains the initial time of the corresponding voice data of each target vocal print.Further, for each targetVocal print is detected from the setting time length after initial time, if the corresponding voice data of target vocal print is got, if reallyDetermine not getting the corresponding voice data of target vocal print in setting time length, then exits and the target vocal print corresponding objectsDialogue mode, but the dialogue mode of other target vocal prints is not influenced.
For example, the content that the earphone in ambient sound data wears subject interests comes from object A, then pass through Application on Voiceprint RecognitionTechnology identifies that the vocal print of A is target vocal print, and the initial time for obtaining the corresponding voice data of target vocal print is 10:00, it is assumed thatSetting time length is 5 minutes, then the corresponding sound of target vocal print is not got since initial time 10:00, in 5 minutesData then exit the dialogue mode with target vocal print corresponding objects A.
Noise reduction device of earphone provided by the embodiments of the present application is described below, noise reduction device of earphone described below withAbove-described earphone noise-reduction method can correspond to each other reference.
Referring to fig. 2, Fig. 2 is a kind of noise reduction device of earphone structural schematic diagram disclosed in the embodiment of the present application.
As shown in Fig. 2, the apparatus may include:
Ambient sound data capture unit 11, for obtaining ambient sound data；
Content of interest judging unit 12, for whether determining in the ambient sound data comprising earphone wearing object senseThe content of interest；
Noise canceling unit 13, for including that earphone is worn in subject interests in determining the ambient sound dataRong Shi eliminates the content in the ambient sound data in addition to the earphone wears the content of subject interests；
Content of interest broadcast unit 14, it is emerging in determining the ambient sound data including earphone wearing object senseWhen the content of interest, the content that the earphone wears subject interests is played.
Optionally, in the various structures that the embodiment of the present application may include to the content of interest judging unit 12One kind is introduced:
The content of interest judging unit may include:
First content of interest judgment sub-unit, for according to preconfigured sense corresponding with earphone wearing objectWhether interest information determines in the ambient sound data comprising voice data corresponding with the information of interest.
Optionally, when the information of interest is interest word, the first content of interest judgment sub-unit may include:
It identifies text acquiring unit, is used for the ambient sound data conversion into identification text；
Interest word searching unit is worn there are preconfigured with the earphone for searching whether in the identification textWear the corresponding interest word of object, and if it exists, then determine in the ambient sound data comprising corresponding with the information of interestVoice data, if it does not exist, it is determined that do not include sound number corresponding with the information of interest in the ambient sound dataAccording to.
Optionally, when the information of interest is interest vocal print, the first content of interest judgment sub-unit can be withInclude:
Interest vocal print determination unit determines in the ambient sound data, if there are the corresponding sound of the interest vocal printSound data；If it exists, it is determined that include voice data corresponding with the information of interest in the ambient sound data, if notIn the presence of, it is determined that do not include voice data corresponding with the information of interest in the ambient sound data.
Optionally, second in various structures that the embodiment of the present application may include to the content of interest judging unitKind is introduced:
The content of interest judging unit may include:
Second content of interest judgment sub-unit wears the characteristic information of object, the feature for obtaining the earphoneInformation is related to earphone wearing subject interests content；
Third content of interest judgment sub-unit, determine the ambient sound data whether with the characteristic information phaseMatch, if so, determining the content for wearing subject interests in the ambient sound data comprising earphone, if not, it is determined that describedThe content that earphone wears subject interests is not included in ambient sound data.
It optionally, can also will above describe two kinds of different structures of content of interest judging unit, in the present embodimentDescribed two structures combine, and propose the third alternative construction of content of interest judging unit, as follows:
Content of interest judging unit may include:
First content of interest judgment sub-unit, for according to preconfigured sense corresponding with earphone wearing objectInterest information determines in the ambient sound data whether include voice data corresponding with the information of interest, if so,Determine the content for wearing subject interests in the ambient sound data comprising earphone；
Second content of interest judgment sub-unit, for determining the ring in the first content of interest judgment sub-unitWhen not including voice data corresponding with the information of interest in the voice data of border, the feature that the earphone wears object is obtainedInformation, the characteristic information are related to earphone wearing subject interests content；
Third content of interest judgment sub-unit, for determine the ambient sound data whether with the characteristic information phaseMatching, if so, determining the content for wearing subject interests in the ambient sound data comprising earphone, if not, it is determined that instituteIt states and does not include the content that earphone wears subject interests in ambient sound data.
Optionally, the various structures that the embodiment of the present application may include to above-mentioned third content of interest judgment sub-unit intoRow is introduced:
The first,
Second content of interest judgment sub-unit may include:
Fisrt feature information acquisition unit is worn the personal information of object for obtaining the pre-set earphone, is madeIt is characterized information.
Based on this, the third content of interest judgment sub-unit may include:
Label acquiring unit, for obtaining label corresponding with the personal information；
Label lookup unit, for whether determining in the corresponding identification text of the ambient sound data comprising the markLabel, if then determining matching, if not, it is determined that mismatch.
Second,
Second content of interest judgment sub-unit may include:
Second feature information acquisition unit wears the key for including in subjects history speech content for obtaining the earphoneTime point and event information, as characteristic information.
Based on this, the third content of interest judgment sub-unit may include:
Text reference unit is identified, for whether determining in the corresponding identification text of the ambient sound data comprising describedMaterial time point and/or the event information, if then determining matching, if not, it is determined that mismatch.
The third,
Second content of interest judgment sub-unit may include:
Third feature information acquisition unit is worn determined by subjects history speech content for obtaining according to the earphoneThe corresponding relationship of interest content and interest-degree, as characteristic information.
Based on this, the third content of interest judgment sub-unit may include:
Interest content judging unit, for whether determining in the corresponding identification text of the ambient sound data comprising targetInterest content；
Target interest-degree query unit, if for the interest content judging unit judging result be it is yes, according to described inThe corresponding relationship of interest content and interest-degree inquires the corresponding target interest-degree of the target interest content；
Interest-degree comparing unit is determined for the size relation according to the target interest-degree and default interest-degree threshold valueThe match condition of the ambient sound data and the characteristic information.
Optionally, above-mentioned content of interest broadcast unit is introduced in the embodiment of the present application, may include:
Target Application on Voiceprint Recognition unit, earphone described in the ambient sound data wears the interior of subject interests for identificationHold corresponding target vocal print；
Dialogue mode enters unit, for entering the dialogue mode with the target vocal print corresponding objects, the dialogue mouldFormula includes: to play the corresponding voice data of target vocal print described in the ambient sound data.
Optionally, above-mentioned content of interest broadcast unit can also include:
Initial time recording unit, when for recording the starting for starting to obtain the corresponding voice data of the target vocal printBetween；
Setting time detection unit, for detecting from the setting time length after the initial time, if getThe corresponding voice data of the target vocal print；
Dialogue mode exits unit, for the target vocal print detection unit testing result be it is no when, exit with it is describedThe dialogue mode of target vocal print corresponding objects.
Noise reduction device of earphone provided by the embodiments of the present application can be applied to earphone noise reducing apparatus, such as PC terminal, cloud platform, clothesBusiness device and server cluster etc..Optionally, Fig. 3 shows the hardware block diagram of earphone noise reducing apparatus, referring to Fig. 3, earphone dropThe hardware configuration for equipment of making an uproar may include: at least one processor 1, at least one communication interface 2,3 He of at least one processorAt least one communication bus 4；
In the embodiment of the present application, processor 1, communication interface 2, memory 3, communication bus 4 quantity be at least one,And processor 1, communication interface 2, memory 3 complete mutual communication by communication bus 4；
Processor 1 may be a central processor CPU or specific integrated circuit ASIC (ApplicationSpecific Integrated Circuit), or be arranged to implement the integrated electricity of one or more of the embodiment of the present inventionRoad etc.；
Memory 3 may include high speed RAM memory, it is also possible to further include nonvolatile memory (non-volatileMemory) etc., a for example, at least magnetic disk storage；
Wherein, memory is stored with program, the program that processor can call memory to store, and described program is used for:
Obtain ambient sound data；
The content for whether wearing subject interests in the ambient sound data comprising earphone determined；
If so, the content in the ambient sound data in addition to the earphone wears the content of subject interests is disappearedIt removes, and plays the content that the earphone wears subject interests.
Optionally, the refinement function of described program and extension function can refer to above description.
The embodiment of the present application also provides a kind of readable storage medium storing program for executing, which can be stored with and hold suitable for processorCapable program, described program are used for:
Obtain ambient sound data；
The content for whether wearing subject interests in the ambient sound data comprising earphone determined；
If so, the content in the ambient sound data in addition to the earphone wears the content of subject interests is disappearedIt removes, and plays the content that the earphone wears subject interests.
Optionally, the refinement function of described program and extension function can refer to above description.
Finally, it is to be noted that, herein, relational terms such as first and second and the like be used merely to byOne entity or operation are distinguished with another entity or operation, without necessarily requiring or implying these entities or operationBetween there are any actual relationship or orders.Moreover, the terms "include", "comprise" or its any other variant meaningCovering non-exclusive inclusion, so that the process, method, article or equipment for including a series of elements not only includes thatA little elements, but also including other elements that are not explicitly listed, or further include for this process, method, article orThe intrinsic element of equipment.In the absence of more restrictions, the element limited by sentence "including a ...", is not arrangedExcept there is also other identical elements in the process, method, article or apparatus that includes the element.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with otherThe difference of embodiment, the same or similar parts in each embodiment may refer to each other.
The foregoing description of the disclosed embodiments makes professional and technical personnel in the field can be realized or use the application.Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined hereinGeneral Principle can be realized in other embodiments without departing from the spirit or scope of the application.Therefore, the applicationIt is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase oneThe widest scope of cause.