A kind of method of voice controlTechnical field
The present invention relates to a kind of methods of voice control.
Background technique
It is various with voice to have become user and mobile phone, electric appliance etc. for means of the voice command as intelligence interactionThe common exchange way of identification function equipment.The accuracy of the speech recognition relationship directly proportional to power consumption general state, large size are setStandby not mind power consumption, speech recognition can identify (or even can be implanted into dialect databases) comprehensively substantially, but for bottom power consumptionEquipment, power consumption is that various functions are necessarily required to consider the problems of when integrated.There is this language mostly on current mobile phoneThe module of sound identification can be associated with the program progress specific program of mobile phone such as the SIRI of apple after content is arrived in speech recognitionIt executes.The functions of modules of the speech recognition integrated in current mobile phone is powerful, and the power consumption needed is also larger.
Above- mentioned information are only used for reinforcing the understanding to the background of the disclosure, therefore it disclosed in the background technology partIt may include the information not constituted to the prior art known to persons of ordinary skill in the art.
Summary of the invention
The present invention for the above technical issues, provides a kind of method of speech recognition, in this way, be integrated with compared withIn the case where for accurate control command, by the unique logic calculation of the present invention of finite number of time, can accurately it knowNot Chu user instruction, biggest advantage is to realize very low land power consumption.
A kind of method of voice control, including,
Acquire voice messaging；
Voice recognition unit identifies voice messaging, obtains first content information；
First content information and pre-stored voice control enabled instruction are compared, if the first content information and the voiceControl enabled instruction matches, then starts voice control；
It acquires voice messaging and obtains the second content information through voice recognition unit；
Second content information and pre-stored control instruction are compared, if second content information and the control instructionMatch, then executes the corresponding movement of the control instruction.
Further, the learning method of pre-stored the voice control enabled instruction and control instruction includes: that user opensDynamic study module and selection instruction content；
Acquire voice data；
Voice recognition unit identifies voice data, obtains the first learning Content；
After second time threshold, the study module prompt inputs voice data again；
Acquire voice data；
Voice recognition unit identifies voice data, obtains the second learning Content；
After second time threshold, the study module prompt inputs voice data again；
Acquire voice data；
Voice recognition unit identifies voice data, obtains third learning Content；
If first learning Content, the second learning Content are consistent or almost the same with the feature of third learning Content, instituteStudy module is stated using any learning Content in the first learning Content, the second learning Content and third learning Content as the fingerEnable the corresponding instruction of content.
Further, the pre-stored voice control enabled instruction is the first speech database；Described is pre-storedControl instruction be second speech data library, first and second speech database is that system is pre-stored or user is pre-The voice data first stored.
Further, the voice recognition unit judges the voice number of each word in the first/second content informationAccording to the presence or absence of in the first/second speech database.
Further, the processor judges continuously to be identified and be stored in the first/second content information describedWhether two words in first/second speech database are adjacent or interval context, if it is not, then processor judgementWith failure, if so, persistently being matched.
Further, the processor judge in the first/second content information identified number of words and with it is identifiedWhether the ratio of the total number of word of the corresponding wrist-watch control instruction of number of words is greater than proportion threshold value, if being less than, the processor judgementIt fails to match, if more than then the processor judges successful match.
Further, after starting voice control, if not collecting any voice letter within the scope of first time thresholdBreath, then terminated speech controls.
Through the above technical solution, user can simply control equipment by voice control, and can be to controlSystem instruction is learnt, and control method is simple, and power consumption is very low.
Detailed description of the invention
Its example embodiment is described in detail by referring to accompanying drawing, above and other target of the invention, feature and advantage willIt becomes more fully apparent.
Fig. 1 is flow chart of the invention.
Fig. 2 is the flow chart of matching judgment of the present invention.
Specific embodiment
Example embodiment is described more fully with reference to the drawings.However, example embodiment can be with a variety of shapesFormula is implemented, and is not understood as limited to example set forth herein；On the contrary, thesing embodiments are provided so that the present invention will moreFully and completely, and by the design of example embodiment comprehensively it is communicated to those skilled in the art.Attached drawing is only the present inventionSchematic illustrations, be not necessarily drawn to scale.Identical appended drawing reference indicates same or similar part in figure, thusRepetition thereof will be omitted.
In addition, described feature, structure or characteristic can be incorporated in one or more implementations in any suitable mannerIn mode.In the following description, many details are provided to provide and fully understand to embodiments of the present invention.SoAnd it will be appreciated by persons skilled in the art that technical solution of the present invention can be practiced and omit one in the specific detailOr more, or can be using other methods, constituent element, device, step etc..In other cases, it is not shown in detail or describesKnown features, method, apparatus, realization or operation are to avoid a presumptuous guest usurps the role of the host and each aspect of the present invention is made to thicken.
Embodiment one.
As shown in Fig. 1, the method for voice control acquires voice messaging first.
Equipment includes voice recognition unit, and voice recognition unit capable of speech information obtains first content information.ForThe specific research of identification voice messaging has been compared skillfully, and the present invention is based on the modules of current speech recognition can recognize that voiceOn the basis of and carry out algorithm.
Voice recognition unit identifies voice messaging, obtains first content information.
Voice control enabled instruction is stored in the memory module of equipment, which is stored in the first languageIn sound database.
It is " voice control " that (default setting) is defaulted in voice control enabled instruction, and user can be according to oneself speech habitsMultiple voice control enabled instructions are set, it might even be possible to are arranged in the form of dialect.Such as " starting voice ", " opening voice "Deng.
When the first content information and the voice control enabled instruction match, equipment starts voice control.
Equipment continues to acquire voice messaging, and the voice recognition unit identifies the voice messaging, obtains in secondHold information.
One or more control instruction is stored in the memory module of equipment, one or more control instruction storesIn second speech data library.
Control instruction can be some control instructions of default (default setting).User can also to phonetic control command intoRow is customized.
When second content information and the control instruction match, the equipment executes the control instruction pairThe movement answered.
After starting voice control, if not collecting any voice messaging within the scope of first time threshold, then terminating languageSound control system.
Embodiment two.
The voice control enabled instruction and control instruction are referred to as instructing, the equipment includes study module, describedStudy module can learn described instruction.The method of study is as follows:
User starts study module and selection instruction content；
Acquire voice data；
Voice recognition unit identifies voice data, obtains the first learning Content；
After second time threshold, the study module prompt inputs voice data again；
Acquire voice data；
Voice recognition unit identifies voice data, obtains the second learning Content；
After second time threshold, the study module prompt inputs voice data again；
Acquire voice data；
Voice recognition unit identifies voice data, obtains third learning Content；
If first learning Content, the second learning Content are consistent or almost the same with the feature of third learning Content, instituteStudy module is stated using any learning Content in the first learning Content, the second learning Content and third learning Content as the fingerEnable the corresponding instruction of content.
The command content that can choose in starting module includes voice control enabled instruction and control instruction, wherein control refers toIt again include that user wants some instructions controlled equipment, such as takes pictures, makes a phone call in order.
Such as, when user selects the content of the voice control enabled instruction in study module, user can input voice" 1 ", after second time threshold, such as after 3 seconds, user says " 1 " again, and after second time threshold, such as after 3 seconds, user is againSay " 1 ".When input three times it is consistent or almost the same when (method that the present invention mainly describes voice control, concrete sound phaseJudgement like degree has been the prior art, and the state modulator that consistent or almost the same error range gives judgement can be justThe range of negative 5-10%), it is determined as the phonetic order that the phonetic order of the input meets the requirements, as voice control enabled instruction.
Certainly, user also can choose any control instruction in control instruction and learn.
Embodiment three.
As shown in Fig. 2, on the basis of example 1, to it is therein match make following algorithm and limit.
The pre-stored voice control enabled instruction is the first speech database；The pre-stored control instructionFor second speech data library, first and second speech database is that system is pre-stored or the pre-stored language of userSound data.The voice recognition unit judges that the voice data of each word in the first/second content information whether there isIn the first/second speech database.The processor judge in the first/second content information to be continuously identified andWhether two words being stored in the first/second speech database are adjacent or interval context, if it is not, then locatingManaging device judgement, it fails to match, if so, persistently being matched.The processor judges quilt in the first/second content informationWhether the ratio of the total number of word of the number of words of identification and wrist-watch control instruction corresponding with identified number of words is greater than proportion threshold value, if smallIn then it fails to match for the processor judgement, if more than then the processor judges successful match.
Such as:
Voice recognition unit judges that the voice data of each word in the content information whether there is in the speech databaseIn.
If some word is present in the speech database, then it is assumed that have identified this from the voice signal receivedWord is denoted as the first word.
The processor maps out that be stored in include the in the speech database according to identified first wordThe control instruction of one word.
Voice recognition unit continues to judge the subsequent voice word of the first word in content information, when the voice recognition unit is knownIt is clipped to another word, and the word is present in the speech database, is then denoted as the second word；If it is not, then continuing to identify nextWord.
The processor is sought in all control instructions for including the first word according to identified second wordIt out include the control instruction of second word.
If the first word and the second word are the positional relationship that front and back is close in the control instruction having been found, and the control refers toOrder only includes two words, then the processor judges that the content information matches with the control instruction.
If the first word and the second word in the control instruction having been found are the positional relationship of one word of time interval, andThe control instruction only includes three words, then the processor still judges that the content information matches with the control instruction.
If the first word and the second word are that front and back is close to or one word of time interval in the control instruction having been foundPositional relationship, and the control instruction includes being more than or equal to four words, then the voice recognition unit continues to identify second wordContent information later.
If the distance between the first word and the second word are more than or equal to two words, this time in the control instruction having been foundVoice match failure.
The voice recognition unit to one word of multi-compartment identifies remaining word, when interval is more than or equal to two words, instituteProcessor is stated to judge the voice recognition unit it fails to match.
Example IV.
As shown in Fig. 2, the voice recognition unit at least identifies [the total word of 66%* of wrist-watch control instruction total number of wordNumber] it just can determine that and want to match, otherwise assert that it fails to match.[] indicates the meaning being rounded, such as instruction of two words, it is necessary to all knowNot；Triliteral instruction, it is necessary to identify at least two；The instruction of four words, it is necessary to identify at least three words.
As when identifying " delay is taken pictures ", when the first word is identified as " prolonging ", the second word must identify " when " or" bat " could be regarded as matching by processor.
Such as when to identify " cloud data are synchronous ", if only identification " cloud ", " end " and " number " three words, or" cloud ", " number ", " same " three words, since the total number of word of identification only has 50%, processor identification matching is unsuccessful.
Through the above technical solution, user can simply control equipment by voice control, and can be to controlSystem instruction is learnt, and control method is simple, and power consumption is very low.
Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extentPipe present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: its according toSo be possible to modify the technical solutions described in the foregoing embodiments, or to some or all of the technical features intoRow equivalent replacement；And these are modified or replaceed, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solutionThe range of scheme.