Embodiment
The embodiment to the present invention is described in detail below in conjunction with the accompanying drawings.It should be appreciated that this place is retouchedThe embodiment stated is merely to illustrate and explain the present invention, and is not intended to limit the invention.
Fig. 1 is a kind of schematic flow sheet of voice interactive method of the embodiment of the present invention；As shown in figure 1, methods described includes：
Step 101, when receiving the first control instruction, voice enabled acquisition function exports the first voice message, and startVoice collecting progress icon, makes the voice collecting progress icon be moved along direction initialization；
The embodiment of the present invention is mainly used in voice interaction device, and described device can specifically be provided with voice APPElectronic equipment, the function that the voice interactive method is realized can by the processor caller code in electronic equipment comeRealize, certain program code can be stored in computer-readable storage medium, it is seen then that the electronic equipment at least includes processor and depositedStorage media.
The electronic equipment includes：Mobile terminal, Wearable terminal, fixed terminal, car-mounted terminal, bank transaction are wholeThe delivery terminal at end, supermarket's transaction terminal and express delivery mailbag.Wherein mobile terminal can at least include mobile phone, it is tablet personal computer, individualPersonal digital assistant (PDA, Personal Digital Assistant), navigator, game machine, intelligent toy etc., Wearable are wholeEnd can at least include intelligent watch, intelligent glasses, intelligent running shoes etc., and fixed terminal can at least include desktop computer, tableIntelligence in face computer, integral computer, television set, projecting apparatus, sound equipment etc., above intelligent toy, intelligent watch refers to equipmentInclude processor and storage medium, so as to automatically or according to the setting of operator such as user perform some sequencingInstruction.
In the embodiment of the present invention, first control instruction that the electronic equipment is received is opening for speech recognition modeWhen opening instruction, the voice collecting progress icon is shown, and start counting up；Or, first control instruction received isWhen wake-up in the local instruction database is instructed, the voice collecting progress icon is shown, and is started counting up, is adopted with voice enabledCollect function, export the first voice message.First voice message is used to inform that custom system immediately enters speech recognition state,User is reminded to start phonetic entry.And start voice collecting progress icon, make the voice collecting progress icon along direction initializationMotion.Here, the voice collecting progress icon can be voice progress bar, progress circle or progress percentage.In addition, describedOne voice message can be the prompt tone defined by user oneself, for example：The voice message such as " please say " or " please indicate ".The electricitySub- equipment immediately enters speech recognition state when first voice message output is finished.
Step 102, when the voice collecting progress icon is moved along direction initialization, to the voice signal in current environmentIt is acquired；
In the embodiment of the present invention, the electronic equipment is after speech recognition state is entered, the voice collecting progress iconMoved with uniform velocity along direction initialization to position is placed restrictions on, and timing of starting from scratch, now, the electronic equipment is in current environmentVoice signal be acquired.And during voice signal is gathered, voice collecting progress chart target described in real-time updateProgress, i.e., described voice collecting progress chart target progress gradually increases.For example, the voice collecting progress icon is voice progressDuring bar, the voice collecting progress icon at least includes being provided with uniform motion in tempo instructions frame, the tempo instructions frameProgress indicator strip；The progress indicator strip is referred to from one end of the tempo instructions frame to another end motion, and the arrival progressStop motion when showing the other end of frame.Here, the time that the voice collecting progress icon moves to the other end from one end is 1-15 seconds.In order to prevent the voice collecting progress chart target movement velocity too fast, the electronic equipment does not collect user's inputVoice signal, or, the voice collecting progress chart target movement velocity is too slow, influences the acquisition time of the electronic equipmentAnd recognition accuracy, in the embodiment of the present invention, it may be preferable that be set to voice collecting progress chart target motion duration 5 seconds.Here, it is within described 5 seconds the upper limit of the electronic equipment single acquisition Speech time, if the electronic equipment was completed in 3 secondsVoice collecting, the voice collecting progress icon also can stop motion immediately, then the time of this interactive voice is exactly 3 seconds.
User is hearing that first voice message or display interface in the voice APP see that the voice collecting entersWhen degree icon starts timing, phonetic entry is proceeded by, and position is placed restrictions on ensuring that the voice collecting progress icon is reachedBefore, complete the phonetic entry.Here, it is the voice collecting progress icon that the voice collecting progress chart target, which places restrictions on position,Maximum progress threshold value, that is to say, that the electronic equipment single allows the maximum time value of phonetic entry, i.e. time-out time.Such asThis, the time of the electronic equipment single acquisition voice signal is limited according to voice collecting progress icon, can shorten the electricitySub- equipment is to the recognition time of voice signal, while improving the signal identification efficiency of the electronic equipment.
Step 103, voice letter is collected before the voice collecting progress icon moves to along direction initialization and places restrictions on positionNumber when, parse the voice signal, obtain speech data；
In the embodiment of the present invention, the electronic equipment determines to move to along direction initialization in the voice collecting progress iconPlace restrictions on before position, when collecting voice signal, show this phonetic entry success, then the voice signal is divided into length certainSpeech frame, is then asked for each frame speech data the average pitch cycle, obtains voice number corresponding with the voice signalAccording to.
If on the contrary, the electronic equipment is moved in the voice collecting progress icon along direction initialization places restrictions on positionBefore, when not collecting the voice signal, show that phonetic entry fails, then terminate this interactive voice, the first language is exported againSound is pointed out.
Step 104, the speech data is matched with the director data in local instruction database；
In the embodiment of the present invention, the electronic equipment is then searched after the speech data is obtained in preset instructions storehouseDirector data corresponding with the speech data, and obtain lookup result.Here, the preset instructions storehouse is according to certainly by userThe instruction database of own requirement definition.Specifically, user by the electronic equipment to APP pairs of the voice installed on the electronic equipmentThe voice server answered sends instruction database request to create, and the voice server is responded after the request to create, controls the electricityThe establishment interface in the display interface idsplay order storehouse of sub- equipment, user is at the establishment interface of the instruction database according to the demand of oneselfCreate the instruction database.For example, user can carry out phonetic entry by the speech voice input function in the establishment interface, with completeIt into the establishment of the instruction database, can also directly be inputted at the establishment interface by word, complete the establishment of the instruction database.Wherein, the instruction database of establishment includes father's instruction and sub-instructions.For example, father's instruction is：Music, video, intelligent familyOccupy, then the sub-instructions of music can be：Next bent, upper one bent, Chinese song, English song；The sub-instructions of video can be：Tengxun regardsFrequently, QQ videos, youku.com's video；The sub-instructions of smart home can be：Curtain, bedroom air-conditioning, electric light etc..In this way, electronic equipmentThe corresponding speech data of the voice signal of collection is identified by user-defined instruction database, can effectively ensure voiceThe accuracy rate of identification.
Further, since the default instruction database is limited instruction set, alone word identification technology is used,Therefore, the embodiment of the present invention, without carrying out semantic analysis by voice cloud service, can be provided the user in speech recognition processOffline service.
In embodiments of the present invention, user can also be configured in preset instructions storehouse to signal acquisition periods.SpecificallyGround, user is asked by the voice APP settings for sending signal acquisition periods to the voice server, the voice serviceDevice is received after the setting request of the signal acquisition periods, controls the display interface of the voice APP to show that the signal is adoptedThe setting interface in collection cycle, user is according to oneself demand at the setting interface of the signal acquisition periods to signal acquisition weekPhase is configured, and after the setup, is sent to the voice server and set successfully request, and the voice server existsReceive after the successful request of the setting, preserve the setting of the signal acquisition periods, and in interactive voice next time,Electronic equipment is acquired according to the signal acquisition periods of preservation to the voice signal of user.
In embodiments of the present invention, user can also be configured to the voice collecting progress chart target type, specificallyGround, user sends voice collecting progress chart target type to the voice server by the voice APP and sets request, describedVoice server receives the voice collecting progress chart target and set after request, controls the display interface of the voice APP to showShow that voice collecting progress chart target sets interface, user sets interface to select oneself needs according to oneself demand in progress chart targetVoice collecting progress icon, and after the setup, sent to the voice server and successfully request, the voice be setServer preserves the voice collecting progress chart target and set after the successful request of the setting is received, and next timeInteractive voice in, show preserve voice collecting progress icon.
Step 105, the director data for determining in the speech data and local instruction database is when the match is successful, output with it is describedCorresponding second voice message of director data.
In the embodiment of the present invention, work instruction data, dormancy instruction data and wake-up are included in the preset instructions storehouseDirector data, the work instruction data that the electronic equipment determines in the speech data and local instruction database is when the match is successful,Show that the speech data is recognized successfully, then the electronic equipment exports the second voice corresponding with the work instruction data and carriedShow, for example, the work instruction data is：" music ", then second voice message is " music ", for reporting to user's inputSpeech data recognize successfully；Or, when the electronic equipment determines that the speech data refers to the work in preset instructions storehouseData are made to mismatch, and during with the dormancy instruction Data Matching, then the electronic equipment is exported and the dormancy instruction dataCorresponding dormancy prompt tone.For example, the dormancy instruction data are：" rest ", then the dormancy prompt tone be：" I rests, it is busy to be me ".Afterwards, the speech recognition mode is changed to park mode；Conversely, when the electronic equipment determines institute's predicateWhen sound data are mismatched with all director datas in preset instructions storehouse, show this speech data recognition failures, terminate thisSecondary interactive voice, and first voice message is exported again, interactive voice is realized in the way of continuously circulating.Here, it is describedWork instruction data refers to that the voice signal that the electronic equipment is inputted according to active user performs the data of command adapted thereto group.ExampleSuch as, the instruction group included in the electronic equipment has " bedroom ", " amusement ", if the voice signal corresponding instruction group being currently receivedWhen " bedroom ", then the electronic equipment performs the instruction group " bedroom ", and shows that the son in the instruction group " bedroom " refers toOrder, for example, sub-instructions are：Lamp, bedroom air-conditioning, curtain；If during the voice signal corresponding instruction group " amusement " being currently received,The electronic equipment performs the instruction group " amusement ", and shows the sub-instructions in the instruction group " amusement ", for example, describedSub-instructions are：Music, game.
Step 106, the corresponding phonetic order of the speech data is performed.
In the embodiment of the present invention, the electronic equipment the second voice message output finish after, immediately hop to it is describedThe corresponding sub-instructions storehouse of speech data, and in the sub-instructions storehouse, continue to gather the voice signal that user sends.For example, instituteStating speech data is：Parlor, then the electronic equipment is when successfully identifying " parlor ", playing alert tones " parlor ", and redirectsTo parlor sub-instructions storehouse corresponding with parlor.For example, the parlor sub-instructions storehouse includes：Curtain, lamp, bedroom air-conditioning, then it is describedElectronic equipment continues to gather the voice signal that user sends in the parlor sub-instructions storehouse, for example, collecting user's transmissionThe corresponding instruction of voice signal is " lamp ", then the electronic equipment performs the control operation to " lamp ".
By the voice collecting progress icon in the embodiment of the present invention, user can be helped to understand when oneself should sendPhonetic order, also, according to voice collecting progress chart target state change and prompt tone, user can be made to understand oneself inputWhether phonetic order is successfully identified, so that user has at fingertips whole speech control process.
Fig. 2 is the implementation process schematic diagram of interactive voice of the embodiment of the present invention；As shown in Figure 2：Including：
Step 201, instruction database is created；
Here, the instruction database is the instruction database defined by user oneself by way of phonetic entry or word input.For example, user-defined instruction database includes：Voice message data, work instruction data, dormancy instruction data, wake-up director dataWith voice collecting progress icon.Wherein, the work instruction data includes at least one work sub-instructions data.For example, describedWork instruction data is amusement, then also includes in the work instruction data：The sons such as game, TV, film and camera refer toMake data.In this way, the voice signal that the local instruction database that system is created according to user oneself is sent to user is identified, not onlySpeech recognition accuracy can be improved, and without carrying out semantic analysis by cloud service, offline service can be provided the user.
Step 202, the first voice message is exported；
Here, first voice message can be the prompt tone of system default, for example, " please say " or by withThe prompt tone that family is defined, for example, the voice message sound such as " owner please tell ", first voice message is mainly used in reminding userIt is ready for phonetic entry.And when first voice message is finished, system immediately enters speech recognition state.
In the embodiment of the present invention, while the electronic equipment is playing first voice message, start voice and adoptCollection progress icon, makes the voice collecting progress icon be moved along direction initialization, voice collecting progress icon edge setting sidePlaced restrictions on to moving to before position, when not collecting voice signal, the voice collecting progress icon makees even from one end to the other sideSpeed motion, and stop motion when reaching the other end.For example, the voice collecting progress icon include tempo instructions frame, it is described enterDegree indicates to be provided with the progress indicator strip of uniform motion in frame；The progress indicator strip is from one end of the tempo instructions frame to anotherOne end motion, and stop motion when reaching the other end of the tempo instructions frame.
The electronic equipment does not collect voice before voice collecting progress icon moves to along direction initialization and places restrictions on positionDuring input, show that phonetic entry fails, first voice message is exported again, remind user to re-start phonetic entry.
In the embodiment of the present invention, the electronic equipment does not inquire what is matched with the phonetic order in preset instructions storehouseDuring sub-instructions, show that this interactive voice fails, terminate this interactive voice, resend first voice message, now,The voice collecting progress chart target progress zero, and after first voice message is finished, the voice collectingProgress chart indicated weight is newly started from scratch timing, and is moved along direction initialization.
In the embodiment of the present invention, the electronic equipment exports first voice message when receiving wake-up instruction, thisWhen, the voice collecting progress icon zero.User is reminded to start phonetic entry.And played in first voice messageBi Hou, the voice collecting progress chart indicated weight is newly started from scratch timing.
Step 203, when the voice collecting progress icon is moved along direction initialization, to the voice signal in current environmentIt is acquired；
Here, user is after first voice message is finished, or is seeing the voice collecting progress iconWhen being moved along direction initialization, phonetic entry is carried out, the electronic equipment is transported in the voice collecting progress icon along direction initializationMove to placing restrictions on before position, gather voice signal.
Step 204, before the voice collecting progress icon moves to along direction initialization and places restrictions on position, voice letter is collectedNumber when, perform step 205；When not collecting voice signal, return and perform step 202；
Here, the electronic equipment is adopted before the voice collecting progress icon moves to along direction initialization and places restrictions on positionWhen collecting voice signal, show the phonetic entry success of user；, whereas if in voice collecting progress icon edge setting sidePlaced restrictions on to moving to before position, voice signal is not collected, then show that this phonetic entry fails.
Step 205, speech data corresponding with the voice signal is searched in preset instructions storehouse, the voice number is determinedAccording to whether being matched with work instruction data, when being matched with work instruction data, step 206 is performed, with work instruction data notTiming, performs step 208；Here, the work instruction data includes father's director data and sub-instructions data, for example, father instructsData are：" amusement ", sub-instructions data are：" game ".The work instruction data can be described in detail below.
Step 206, prompt tone corresponding with the phonetic order data is played；
For example, the phonetic order data are " amusements ", then the corresponding prompt tone of the phonetic order data is " amusement ",To remind the phonetic entry of user to be identified successfully.
Step 207, the corresponding instruction of the phonetic order data is performed；
Here, the electronic equipment finds work instruction data and the speech data collected in preset instructions storehouseTiming, exports work prompt tone corresponding with the work instruction data, and the phonetic entry to report to user is recognized successfully, andAfter the work prompt tone is finished, sub-instructions storehouse corresponding with the phonetic order data is immediately hopped to, this is representedInteractive voice is completed, and re-executes step 203.Or, the electronic equipment is after the work prompt tone is finished, immediatelyCorresponding function is performed, without jump instruction storehouse.For example, the work prompt tone is to play music, then the electronic equipment existsAfter the work prompt tone is finished, music playback function is immediately performed.
Step 208, speech data corresponding with the voice signal is searched in preset instructions storehouse, the voice number is determinedAccording to whether with dormancy instruction Data Matching, during with dormancy instruction Data Matching, perform step 209, with dormancy instruction data notTiming, performs step 202；
Here, dormancy instruction data refer to allow the voice APP in the electronic equipment to enter the instruction of resting state.ExampleSuch as, dormancy instruction data are " heronsbill rests ".
Step 209, dormancy prompt tone corresponding with dormancy instruction data is sent；
Here, the electronic equipment determines the speech data collected and the dormancy instruction data in preset instructions storehouseDuring matching, dormancy prompt tone corresponding with dormancy instruction data is played, and after the dormancy prompt tone is finished, institute's predicateSound APP enters resting state, while control voice collection progress icon enters park mode, and performs step 210.
Step 210, wait and wake up instruction；
Here, the voice APP of the electronic equipment in a dormant state when, only receive wake up instruction, it is other instruction without exceptionDo not receive.
Step 211, if receive wake-up instruction, when having been received by wake-up instruction, performs step 212, exits dormancy shapeState, re-executes step 202, when not receiving wake-up instruction, re-executes step 210.
Step 212, resting state is exited.
Fig. 3 is the view of voice APP installations on mobile terminals；As shown in figure 3, the mobile terminal is handMachine, and the entitled heronsbill voice assistant of the voice APP on mobile phone, including heronsbill voice assistant work shapeState schematic diagram 301a and heronsbill voice assistant resting state schematic diagram 301b.Wherein, Figure 30 1a in the operating condition, progress chartMode of operation is designated as, corresponding instruction group is shown.For example, parlor, bedroom, amusement, navigation, return.And it can receive anyPhonetic order；And Figure 30 1b are in the dormant state, progress chart is designated as resting state, and system then only receives and wakes up instruction, do not receiveOther any phonetic orders.
In the embodiment of the present invention, each node in multiway tree can regard the corresponding instruction of a work instruction data asGroup.As shown in Fig. 3-301a, " happy cabin ", " bedroom ", " parlor ", " navigation ", " amusement " are considered as an instruction group,And each instruction group includes multiple sub-instructions again.For example, instruction group " happy cabin " include " bedroom ", " parlor ", " navigation "," amusement " four sub-instructions, instruction group " bedroom " includes：Curtain, lamp, three sub-instructions of bedroom air-conditioning；In instruction group " parlor "Including：Parlor monitoring, parlor air-conditioning, robot, television set, five sub-instructions of video recorder；Instruction group " amusement " includes：ElectricityShadow, music, three sub-instructions of camera；Then without sub-instructions in instruction group " navigation ".
Instruction in the embodiment of the present invention includes two types, is respectively：Jump instruction and execute instruction.Wherein, it is describedJump instruction refers to the instruction that turn function is performed between each instruction.For example, the instruction group for being currently at working condition is " fastHappy cabin ", then when the phonetic order received is " bedroom ", be currently at instruction group " happy cabin " switching of working conditionTo instruction group " bedroom ".
The execute instruction refers to the instruction for performing specific function.For example, the instruction group for being currently at working condition is " soundIt is happy ", then when it is " music " to receive phonetic order, then music playback function is performed, not cutting between execute instruction groupChange operation.
In the embodiment of the present invention, only allow an instruction group in running order in the same time, such as, currently, instructionWhen group " amusement " is in running order, system currently only supports " film ", " music " and " camera " corresponding phonetic order.
Shown in Fig. 3-301, in running order instruction group is " happy cabin ", and the rightmost side of current display interface is arrangedList all instructions of present instruction group " happy cabin " in table, including " parlor ", " bedroom ", " amusement ", " navigation ", " returnReturn " five instructions.Wherein, in five instructions, " parlor ", " bedroom ", " amusement ", " navigation " are jump instructions, are respectively used toExecute instruction turn function, for example, " parlor " instruction performs and jumps to instruction group " parlor " from present instruction group " happy cabin ".And " return " instruction is then used to jump to upper level instruction from present instruction group.For example, being currently at the instruction group of working conditionIt is " music ", then when performing " return " instruction, then jumps to father's instruction group " amusement " of " music " instruction.
Fig. 3 also realizes schematic diagram 302a including phonetic order (one)；As shown in Figure 30 2a：
User inputs phonetic order " amusement ", electronic equipment identification when instruction group " happy cabin " is in running orderWhen to go out the phonetic order be " amusement ", output with after the corresponding voice message of " amusement " instruction, it is " happy small from instruction group immediatelyRoom " jumps to instruction group " amusement ", and the control voice APP sub-instructions information that includes of display screen idsplay order group " amusement ".For example, the command information that instruction group " amusement " is included is：Music, film, camera, return.
Fig. 3 also realizes schematic diagram 302b including phonetic order (two)；As shown in Figure 30 2b：
User's input speech signal " music ", electronic equipment identifies that the corresponding phonetic order of the voice signal is " soundIt is happy " when, export with after the corresponding voice message of " music " instruction, jumping to instruction group " sound from present instruction group " amusement " immediatelyIt is happy ", and the control voice APP command information that includes of display screen idsplay order group " music ".For example, the instruction group " music "Including command information be：" next ", " pause ", " broadcasting ", " upper one is first ", " end ", " song of Little Bear ", " English songSong ", " national language song ", " Music on Demand " and " return ".Wherein, " next ", " pause ", " broadcasting ", " upper one is first ", " knotBeam ", " song of Little Bear ", " English songs ", " national language song " are execute instructions." Music on Demand " and " return " is jump instruction.
Fig. 3 also realizes schematic diagram 302c including phonetic order (three), as shown in Figure 30 2c：
User's input phonetic order " broadcastings ", electronic equipment identifies the phonetic order when being " broadcasting ", and output is with " broadcastingPut " instruct after corresponding voice message, the music in current music storehouse is played immediately, and here, the music of broadcasting can be upper oneA song or the song of system shuffle for the last broadcasting of subsystem record, can also be and set according to userThe song for the music sequential selection put.
Fig. 3 also realizes schematic diagram 302d including phonetic order (four).As shown in Figure 30 2d：
User's input phonetic order " Aladdin rest ", electronic equipment identifies that the phonetic order is that " Aladdin is stoppedDuring breath ", after output voice message " I rests, busy to be me " corresponding with " Aladdin rest " instruction, systemImmediately enter resting state.
Fig. 4 is the view that voice APP is arranged on wearable device；As shown in figure 4, the wearable device is handTable, including working state schematic representation 401a and resting state schematic diagram 401b, when voice APP is in running order, such as schemeShown in 401a, progress icon is in mode of operation, and display multiple instruction group, and can receive any phonetic order；Work as languageSound APP in a dormant state when, as shown in Figure 40 1b, progress icon be in park mode, except wake up instruction in addition to do not receive appointWhat phonetic order.Voice APP recognizes that the method for phonetic order is consistent with Fig. 3 in described Fig. 4, its voice APP's distinguishedCarrier is installed different, in the embodiment of the present invention, method reference picture 1, Fig. 2 of the voice APP identifications and execution phonetic orderWith described by Fig. 3, it will not be repeated here.
Fig. 5 is Fig. 3 and Fig. 4 workflow schematic diagram；As shown in figure 5, including：
Step 501, voice APP loads default instruction database；
Here, when the voice APP in a dormant state when, user can send to electronic equipment and wake up instruction to startThe voice APP, and after the electronic equipment opens the voice APP, load user-defined local instruction database, exampleSuch as, the local instruction database includes：Voice message sound data " please say ", work instruction data " bedroom, parlor, navigation, amusement,Return ", dormancy instruction data " Aladdin rest ", dormancy prompt tone data " I rests, busy to be me ", wake-up instructionData " calling Latin " and voice collecting progress icon " progress circle " (referring to shown in Figure 30 1a).
Step 502, voice message sound " please say " is played；
Here, user starts phonetic entry after the voice message sound is heard.
Step 503, voice collecting progress icon is moved along direction initialization, and timing of starting from scratch, and voice APP opens languageSound identification function；
Step 504, user's input phonetic order " amusement " (referring to shown in Figure 30 2a)；
Step 505, when electronic equipment collects voice signal, the voice signal is parsed into speech data and to describedSpeech data is identified, if recognizing successfully, performs step 506, if recognition failures, re-executes step 502；
Step 506, whether the speech data matches with work instruction data.When being matched with work instruction data, performStep 507, when being mismatched with work instruction data, step 509 is performed；
Step 507, instruction group " amusement " matching corresponding with work instruction data, plays voice message sound " amusement "；
Here, the voice message sound is used to point out user to identify phonetic order " amusement ".
Step 508, phonetic order " amusement " is performed；
Here, the electronic equipment jumps to instruction group " amusement " node from current instruction group node, and describedInstruction group " amusement " node includes sub-instructions " film, music, camera, return ", afterwards, re-executes step 503.
In the embodiment of the present invention, the voice APP circulations perform step 502 to step 505, it is determined that collecting user's transmissionDuring voice signal, the voice signal collected is identified electronic equipment, for example, identifying that the voice signal correspondingly refers toWhen making group " music ", instruction group " music " is jumped to from present instruction group.It can refer in the instruction group " music " including sonMake " next, pause, play, upper one, end, the song of Little Bear, English songs, national language song, Music on Demand, return " (joinAs shown in Figure 30 2b), afterwards, re-execute step 503.
In the embodiment of the present invention, the voice APP circulations perform step 502 to step 505, collect user and send voiceDuring signal, the voice signal is identified.For example, (referring to figure when determining voice signal corresponding instruction " broadcasting "Shown in 302c), music playback function is performed, afterwards, step 503 is re-executed.
Step 509, the speech data whether with dormancy instruction Data Matching, determine the speech data and dormancy instructionDuring Data Matching, step 510 is performed, when determining that the speech data is mismatched with dormancy instruction data, step is re-executed502；
Here, the voice APP that the electronic equipment is installed identifies the voice letter that user sends in default instruction databaseIt is number corresponding when being dormancy instruction " Aladdin rest ", perform step 510, it is unidentified go out dormancy instruction " Aladdin rest "When, re-execute step 502.
Step 510, dormancy prompt tone " I rests, busy to be me " is sent；
Here, voice APP is sent after the dormancy prompt tone, and the voice APP enters resting state (referring to Figure 30 2d institutesShow), perform step 511；
Step 511, wait and wake up instruction " calling Latin ", and perform step 512；
Step 512, if receive wake-up instruction " calling Latin ", when having been received by wake-up instruction " calling Latin ", holdsRow step 513, when not receiving wake-up instruction " calling Latin ", re-executes step 511；
Step 513, resting state is exited, step 502 is re-executed.
Fig. 6 is a kind of composition schematic diagram of voice interaction device of the embodiment of the present invention：As shown in fig. 6, described device includes：Output unit 601, collecting unit 602, resolution unit 603, judging unit 604 and execution unit 605；
Wherein, the output unit 601, for receiving during the first control instruction, voice enabled acquisition function, output theOne voice message, and start voice collecting progress icon, the voice collecting progress icon is moved along direction initialization；It is additionally operable toThe director data for determining in speech data and local instruction database exports the second language corresponding with the director data when the match is successfulSound is pointed out；
The collecting unit 602, for when the voice collecting progress icon is moved along direction initialization, to current environmentIn voice signal be acquired；
The resolution unit 603, for before the voice collecting progress icon moves to along direction initialization and places restrictions on positionWhen collecting voice signal, the voice signal is parsed, speech data is obtained；
The judging unit 604, for the speech data to be matched with the director data in local instruction database；
The execution unit 605, for determining the speech data, the match is successful with the director data in local instruction databaseWhen, perform the corresponding phonetic order of the speech data.
In the embodiment of the present invention, described device can be specifically the electronic equipment for being provided with voice APP.The electronic equipmentIncluding：Mobile terminal, Wearable terminal, fixed terminal, car-mounted terminal, bank transaction terminal, supermarket's transaction terminal and express deliveryThe delivery terminal of mailbag.Wherein mobile terminal can at least include mobile phone, tablet personal computer, PDA, navigator, game machine, intelligence object for appreciationTool etc., Wearable terminal can at least include intelligent watch, intelligent glasses, intelligent running shoes etc., and fixed terminal can at least be wrappedInclude in desktop computer, desktop computer, integral computer, television set, projecting apparatus, sound equipment etc., above intelligent toy, intelligent watchIntelligence refers to that equipment includes processor and storage medium, so as to automatically or according to the setting of operator such as user holdThe instruction of some sequencing of row.
In the embodiment of the present invention, described device also includes display unit 606, for referring in first control receivedWhen order is the open command of speech recognition mode, the voice collecting progress icon is shown；Or, first control receivedWhen system instruction is the wake-up instruction in the local instruction database, the voice collecting progress icon is shown, to enable voice collectingFunction, and the first voice message is exported from the output unit 601 to user, first voice message is used to inform that user isSystem immediately enters speech recognition state, reminds user to start phonetic entry.And start voice collecting progress icon, make the voiceCollection progress icon is moved along direction initialization.Here, the voice collecting progress icon can be voice progress bar, progress circle orProgress percentage.In addition, first voice message can be the prompt tone defined by user oneself, such as：" please say " or " pleaseThe voice messages such as instruction ".The electronic equipment immediately enters speech recognition shape when first voice message output is finishedState.
In the embodiment of the present invention, voice APP in said device is installed after speech recognition state is entered, the voiceCollection progress icon moves with uniform velocity along direction initialization to position is placed restrictions on, and timing of starting from scratch.Trigger the collecting unitVoice signal in 602 pairs of current environments is acquired.The collecting unit 602 is additionally operable to the gatherer process in voice signalIn, voice collecting progress chart target progress described in real-time update, i.e., described voice collecting progress chart target progress gradually increases.ExampleSuch as, when the voice collecting progress icon is voice progress bar, the voice collecting progress icon at least includes tempo instructions frame,The progress indicator strip of uniform motion is provided with the tempo instructions frame；The progress indicator strip is by the one of the tempo instructions frameStop motion when holding to another end motion, and reaching the other end of the tempo instructions frame.Here, the voice collecting progress chartIt is 1-15 seconds to mark the time for moving to the other end from one end.In order to prevent the voice collecting progress chart target movement velocity tooIt hurry up, described device does not collect the voice signal of user's input, or, the voice collecting progress chart target movement velocity is tooSlowly, influence in the acquisition time and recognition accuracy of described device, the embodiment of the present invention, it is preferable that enter the voice collectingThe motion duration of degree icon is set to 5 seconds.Specifically, it is within described 5 seconds the upper of the single acquisition Speech time of collecting unit 602Limit, if the collecting unit 602 completed voice collecting in 3 seconds, the voice collecting progress icon can also stop immediatelyMotion, then the time of this interactive voice is exactly 3 seconds.
In the embodiment of the present invention, user is hearing that first voice message or display interface in the voice APP seeTo the voice collecting progress chart timestamp, phonetic entry is proceeded by, and ensuring the voice collecting progress icon arrivalPlace restrictions on before position, complete the phonetic entry.Here, it is the voice collecting that the voice collecting progress chart target, which places restrictions on position,Progress chart target maximum progress threshold value, that is to say, that described device single allows the maximum time value of phonetic entry, i.e., when overtimeBetween.In this way, limiting the time of described device single acquisition voice signal according to voice collecting progress icon, the dress can be shortenedThe recognition time to voice signal is put, while improving the signal identification efficiency of described device.
In the embodiment of the present invention, the collecting unit 602 is moved in the voice collecting progress icon along direction initializationPlace restrictions on before position, it is determined that when collecting voice signal, showing this phonetic entry success, the resolution unit 603 being triggered, by instituteState resolution unit 603 and the voice signal is divided into the certain speech frame of length, then each frame speech data is asked for averagePitch period, obtains speech data corresponding with the voice signal.
If on the contrary, voice collecting progress icon described in the collecting unit 602 moves to along direction initialization and places restrictions on positionBefore, when not collecting the voice signal, show that phonetic entry fails, then terminate this interactive voice, trigger the output singleMember 601 exports the first voice message again.
In the embodiment of the present invention, the resolution unit 603 triggers the judging unit after the speech data is obtained604, director data corresponding with the speech data is searched in preset instructions storehouse by the judging unit 604, and looked intoLook for result.Here, the preset instructions storehouse is the instruction database according to oneself requirement definition by user.Specifically, user passes through electricitySub- equipment sends instruction database request to create, institute's predicate to the corresponding voice servers of the voice APP installed on the electronic equipmentSound server is responded after the request to create, controls the establishment interface in the display interface idsplay order storehouse of the voice APP, userThe instruction database is created according to the demand of oneself at the interface that creates of the instruction database.For example, user can be created by describedSpeech voice input function in interface carries out phonetic entry, to complete the establishment of the instruction database, directly can also be created describedInterface is inputted by word, completes the establishment of the instruction database.Wherein, father's instruction can be included in the instruction database of establishment and son refers toOrder, wherein, father's instruction can be：Music, video, smart home, the sub-instructions of music can be：A next bent, upper song,Chinese song, English song；The sub-instructions of video can be：Tengxun's video, QQ videos, youku.com's video；The sub-instructions of smart home canTo be：Curtain, bedroom air-conditioning, electric light etc..In this way, electronic equipment by user-defined instruction database to the voice signal that identifiesCorresponding phonetic order is identified, and can effectively ensure the accuracy rate of speech recognition.
Further, since the default instruction database is limited instruction set, alone word identification technology is used,Therefore, the embodiment of the present invention, without carrying out semantic analysis by voice cloud service, can be provided the user in speech recognition processOffline service.
In embodiments of the present invention, user can be to be configured in preset instructions storehouse to signal acquisition periods.Specifically,User sends signal acquisition periods by electronic equipment to the corresponding voice servers of the voice APP installed on the electronic equipmentSetting request, the voice server received after the setting request of the signal acquisition periods, controls the voice APP'sDisplay interface shows the setting interface of the signal acquisition periods, user's setting in the signal acquisition periods according to oneself demandPut interface to be configured the signal acquisition periods, and after the setup, send and set successfully to the voice serverRequest, the voice server receiving the setting successfully after request, preserving the setting of the signal acquisition periods,And in interactive voice next time, the collecting unit 602 is according to the voice signal of the signal acquisition periods of preservation to userIt is acquired.
In embodiments of the present invention, user can also be configured to voice collecting progress icon, and specifically, user passes throughElectronic equipment sends voice collecting progress chart target to voice server and sets request, and the voice server receives institute's predicateSound collection progress chart target is set after request, controls the display interface of the voice APP to show that voice collecting progress chart target is setInterface is put, user sets the voice collecting progress that interface selects oneself to need according to oneself demand in voice collecting progress chart targetIcon, and after the setup, sent to the voice server and successfully request is set, the voice server is being receivedThe setting successfully after request, preserves the voice collecting progress chart target and sets, and the display unit 606 is next timeIn interactive voice, the voice collecting progress icon preserved is shown.
In the embodiment of the present invention, include in the preset instructions storehouse work instruction data, dormancy instruction data, wake up refer toData are made, the output unit 601 determines the work in the speech data and preset instructions storehouse that the collecting unit 602 is collectedWhen director data is matched, show that the speech data is recognized successfully, then export the second language corresponding with the work instruction dataSound is pointed out, for example, the work instruction data is：" music ", then second voice message is " music ", for reporting to userThe phonetic order of input is recognized successfully；Or, when the output unit 601 is determined in the speech data and preset instructions storehouseWork instruction data is mismatched, and during with the dormancy instruction Data Matching, is then exported corresponding with the dormancy instruction dataDormancy prompt tone.For example, the dormancy instruction data are：" heronsbill rest ", then the dormancy prompt tone is：" I rests, it is busy to be me ".Afterwards, the speech recognition mode is changed to park mode；Conversely, described in being determined when the output unit 601When speech data is mismatched with all director datas in preset instructions storehouse, show this phonetic order recognition failures, terminateThis interactive voice, and first voice message is exported again.Interactive voice is realized in the way of continuously circulating.
In the embodiment of the present invention, the output unit 601 is after the second voice message output is finished, and triggering is described to perform listMember 605, immediately hops to sub-instructions storehouse corresponding with the phonetic order data, and trigger described by the execution unit 605Collecting unit 602 continues to gather the voice signal that user sends in the sub-instructions storehouse.For example, the phonetic order data are：Parlor, then described device is when successfully identifying " parlor ", after the playing alert tones of output unit 601 " parlor " finish, instituteState execution unit 605 and immediately hop to parlor sub-instructions storehouse corresponding with parlor.For example, the parlor sub-instructions storehouse includes：WindowCurtain, lamp, bedroom air-conditioning.Fig. 2 descriptions in specific interactive voice implementation process reference method embodiment, will not be repeated here.
It should be understood by those skilled in the art that, embodiments of the invention can be provided as method, system or computer programProduct.Therefore, the shape of the embodiment in terms of the present invention can use hardware embodiment, software implementation or combine software and hardwareFormula.Moreover, the present invention can be used can use storage in one or more computers for wherein including computer usable program codeThe form for the computer program product that medium is implemented on (including but is not limited to magnetic disk storage and optical memory etc.).
The present invention is the flow with reference to method according to embodiments of the present invention, equipment (system) and computer program productFigure and/or block diagram are described.It should be understood that can be by every first-class in computer program instructions implementation process figure and/or block diagramJourney and/or the flow in square frame and flow chart and/or block diagram and/or the combination of square frame.These computer programs can be providedThe processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produceA raw machine so that produced by the instruction of computer or the computing device of other programmable data processing devices for realThe device for the function of being specified in present one flow of flow chart or one square frame of multiple flows and/or block diagram or multiple square frames.
These computer program instructions, which may be alternatively stored in, can guide computer or other programmable data processing devices with spyDetermine in the computer-readable memory that mode works so that the instruction being stored in the computer-readable memory, which is produced, to be included referring toThe manufacture set by dress is made, the command device is realized in one flow of flow chart or multiple flows and/or one side of block diagramThe function of being specified in frame or multiple square frames.
These computer program instructions can be also loaded into computer or other programmable data processing devices so that in meterSeries of operation steps is performed on calculation machine or other programmable devices to produce computer implemented processing, thus in computer orThe instruction performed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram oneThe step of function of being specified in individual square frame or multiple square frames.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the scope of the present invention.