Embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is
A part of embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art
The every other embodiment obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.
The robot awakening method and device that the present invention is provided are applied to application scenarios as shown in Figure 2, including multiple machines
It can be communicated between device people 20, multiple robots 20 by network 30, wherein, network 30 can be such as family or company
LAN, or particular network etc..Robot 20 has at least one network interface, sets up and communicates to connect with network 30,
Data or instruction are obtained from network 30.User 10 can be configured or issue order to multiple robots 20.
There is corresponding wake-up word to be used to itself is waken up or responded use from resting state for each robot
The call (wake up word and be generally one, or be multiple) at family.Wherein, it is described to wake up the name that word be robot, know
Other code or other any vocabulary, the wake-up word can be configured by user, can also be carried when dispatching from the factory.
Each robot, which has, is used to place the wake-up dictionary for waking up word, and the wake-up dictionary can be when dispatching from the factory certainly
Band, can also be set by user.Identical can be shared between each robot of same user and wakes up dictionary, with shown in Fig. 2
Embodiment exemplified by, it is assumed that user 10 possesses three robots, and the wake-up word of three robots is respectively Mike, Tom and Jerry
(this, which is sentenced exemplified by each robot correspondence one wakes up word, illustrates, but is not limited to this, and each robot can also associate two
Wake-up word above), then the wake-up dictionary of three robots can be arranged to include Mike, Tom and Jerry.Work as user
When issuing voice command " Mike, Tom, Jerry help me that room is tidied up " to three robots simultaneously, due to each
The wake-up dictionary of robot includes Mike, Tom and Jerry, and according to the wake-up dictionary, robot can parse wake-up word
Mike, Tom and Jerry, the command context " helping me that room is tidied up " after word is waken up so as to correctly parse.
In actual use, acoustic model can be used to the parsing that word is waken up in voice command, wakes up calling out in dictionary
Word of waking up uses aligned phoneme sequence corresponding with waking up word, and the voice sent according to user and default acoustic model decode phoneme sequence
Row, then the aligned phoneme sequence is matched with wake-up word aligned phoneme sequence, so as to parse wake-up word.Parse after wake-up word, language
The content behind word is waken up in sound order and is used as command context.The semantic parsing of command context needs to build in advance in voice command
Order word grammar file, command context that user sends needs to be present in the order word grammar file built in advance, according to
Voice and order word grammar file that family is sent parse the semanteme of command context.
Waking up the renewal of dictionary can be configured by user, can also voluntarily complete to update between robot.For example
Current wake-up dictionary includes tri- wake-up words of Mike, Tom and Jerry, when again user buys a robot (assuming that waking up
Word is John) and by between network and other robot set up communication connection after, the robot newly added can by network to
Other robot broadcasts the wake-up word John of oneself, and other robot is received after the wake-up word, can be added to the wake-up word
Wake up in dictionary, and the wake-up dictionary after renewal is sent to the robot newly added.
It should be noted that, although only show 1 user 10,3 robots 20 in fig. 2.But people in the art
Member is it is understood that in actual application, the application scenarios can also include more users 10 and robot 20.This
The robot awakening method provided and device are provided and are also applied to the occasion that user is controlled to one or two robot.
The embodiments of the invention provide a kind of robot awakening method, any machine that the awakening method can be as shown in Figure 2
Device people performs, as shown in figure 3, the flow chart of one embodiment for the awakening method, the awakening method includes:
Step 101:Obtain voice command;
In actual applications, microphone can be set to be used for real-time reception voice signal with robot.The voice
Order can be the voice signal of real-time reception.But although user sends voice sometimes, but it is not necessarily for robot
Voice command, it is therefore desirable to determined whether to voice messaging, if the voice messaging is for waking up appointing in dictionary
One wakes up the order of the corresponding robot of word, then regard the voice messaging as voice command.
Step 102:According to institute's speech commands and default wake-up dictionary, the wake-up word in institute's speech commands is parsed
And command context, the wake-up dictionary includes several wake-up words, wake-up words correspondence at least two in the wake-up dictionary
Robot;
Wherein, the command context can be the corresponding language of the command context intercepted from the voice command that user sends
Sound or the result after semantic parsing is carried out to the voice.If the former, robot is corresponding in execution command context
Also need to carry out semantic parsing during operation.Described several include one or more, i.e., described wake-up dictionary can include one
Individual or multiple wake-up words.
Step 103:Wake-up word and command context according to parsing perform operation.
Optionally, robot is parsed after wake-up word and command context, if the wake-up word in parsing includes itself correspondence
Wake-up word, then perform the corresponding operation of command context, if not including itself corresponding wake-up word, any behaviour do not performed
Make.Or, whether the wake-up word no matter parsed includes itself corresponding wake-up word, and command context all is sent into wake-up word pair
The other robot answered, the task of user's explanation is completed by the corresponding robot of wake-up word jointly.Can also by robot according to
The Task-decomposing that command context hands over user notifies to wake up other corresponding machines of word respectively into several subtasks, by subtask
Device people, by corresponding robot cooperated the completing user's explanation of the task of wake-up word.
It should be noted that above-mentioned steps 101,102 and 103 are not necessarily carried out to Ren Yiji by each robot
Device people is possible to execution Overall Steps and is also possible to only carry out one or two steps therein.Using the application scenarios shown in Fig. 2 as
Example, when user sends voice " Mike, Tom, Jerry help me that room is tidied up ".Robot Mike, Tom, Jerry can
To obtain the voice command that user sends respectively, parse after wake-up word and command context, if it find that oneself being the life of user
Object is made, then operation is performed according to the command context parsed.In this case, three robots be carried out step 101,
102 and 103.May also only have robot Mike to hear the voice command that user sends, Mike is obtained after voice command, solve
Separate out and wake up word and command context, command context is sent to Tom and Jerry, three robots perform task jointly.This
In the case of only Mike perform step 101,102 and 103, and Tom and Jerry have only carried out step 103.Step 102 may
Only performed by a robot or several robots, the wake-up word and command context that parse can be passed through net by the robot of execution
Network shares to other robot.
Awakening method provided in an embodiment of the present invention, includes several wake-up words for waking up word by being preset in robot
Storehouse, such as user possess multiple robots, and the corresponding word that wakes up of the plurality of robot can be pre-set in wake-up dictionary,
When user issues voice command to multiple robots simultaneously, robot can correctly be parsed according to default wake-up dictionary
The wake-up word contained in voice command, so as to correctly parse the command context in voice command, and then completes user's explanation
Task.
Wherein, specifically, in certain embodiments of the present invention, as shown in fig. 4 a, the acquisition voice command include with
Lower step;
Step 1011:Monitor voice messaging;
Each robot can monitoring users are sent in real time voice messaging;
Step 1012:Whether confirm the voice messaging is order for waking up the corresponding robot of dictionary;
Step 1013:If the voice messaging is the order for waking up the corresponding robot of dictionary, record is described
Voice messaging is used as voice command.
Whether confirm above-mentioned voice messaging is that user sends for waking up the corresponding robot of any wake-up word in dictionary
Order, if it is, the voice command that the voice messaging, which is user, to be sent.Specifically, whether confirming the voice messaging
For the order for waking up the corresponding robot of dictionary, first confirm that whether the voice messaging includes default wake up in dictionary
Any wake-up word, if including waking up any wake-up word in dictionary, whether the appearance for continuing to judge the wake-up word is user
Order or call to robot, rather than user mention robot in the call with other people.Judge what user sent
Whether voice messaging is order or call to robot, can be stopped by judging to wake up between word and voice content below
Whether the time interval paused is more than a preset time, if it exceeds a preset time, then the voice messaging is the life to robot
Order is called.Or can be by judging first wakes up before word whether have other voice contents, if without other languages
Sound content, then the voice messaging is the order or call to robot.
Optionally, in the other embodiment of the awakening method, as shown in Figure 4 b, the acquisition voice command includes:
Step 1014:Monitor voice messaging;
Each robot can monitoring users are sent in real time voice messaging;
Step 1015:Whether confirm the voice messaging is order for waking up the corresponding robot of dictionary;
Step 1016:If the voice messaging is the order for waking up the corresponding robot of dictionary, record is described
Voice messaging and the initial time for listening to voice messaging, and add default interim order record group;
Whether if robot listens to the voice messaging that user sends, it is user that can determine whether the voice messaging
For waking up the order that sends of the corresponding robot of any wake-up word in dictionary, if it is, record the voice messaging with
And robot is at the time of hear voice messaging, and add default interim order record group.
In actual applications, robot can be by the voice messaging and deposit robot at the time of hearing voice messaging
In caching.The purpose for setting up interim order group is to obtain relatively complete and clearly user command, with three machines shown in Fig. 2
Exemplified by people, it is assumed that the voice command of user's issue is " Mike, Tom, Jerry help me that room is tidied up ", then hears
Its timestamp when hearing above-mentioned wake-up word of each robot records of " Mike " beginning, " Tom " beginning or " Jerry " beginning
T (initial time), each might not can hear complete order by robot, presumable robot is just a long way off to user
It is mobile, " Mike " most started is not heard, what is heard is " Tom, Jerry help me that room is tidied up ", is recorded herein
Timestamp is exactly to prevent this half sentence to be taken as the situation of complete order to occur.
Specifically, confirming whether the voice messaging is order for waking up the corresponding robot of dictionary, first confirm that
Whether the voice messaging includes the default any wake-up word waken up in dictionary, if including waking up any wake-up in dictionary
Word, then whether the appearance for continuing to judge the wake-up word is order or call of the user to robot, rather than user with its
Robot is mentioned in other people call.Whether be order or call to robot, can if judging voice messaging that user sends
With by whether judging to wake up the time interval paused between word and voice content below more than a preset time, if it exceeds
One preset time, then the voice messaging is the order or call to robot.Or can be by judging first wake-up word
Above whether have other voice contents, if without other voice contents, the voice messaging be order to robot or
Call.
Step 1017:Confirm the earliest moment in the initial time of the robot records in interim order record group, according to
The earliest moment determines initial time section;
Compare the initial time of the robot records in interim order record group, it is determined that at the time of earliest, because earliest
The moment voice messaging of record is relatively most complete.Such as earliest moment is t1, using earliest moment t1 as starting point, and according to a warp
Test given threshold t0 and determine initial time Duan Weicong t1 to t1+t0.
T0 setting can be set according to the performance and experience of robot, such as 0.1s.On the one hand, different machines people
React faster or slower, it is therefore desirable to set a time difference, on the other hand, one is had between two wake-up words that user sends
Relatively-stationary interval, thus setting time difference should be less than this interval, to prevent miss one wake-up word.
Step 1018:Obtain definition highest voice in the voice messaging that initial time is located in initial time section
Information is used as voice command.
After voicemail logging terminates, the robot in interim order record group can be carried out to the voice messaging of own cache
Definition judges, obtains fractional value x, takes the definition highest voice messaging in initial time section (t1 to t1+t0) as language
Sound order.
It should be noted that above-mentioned steps 1014,1015,1016,1017 and 1018 are not necessarily by each robot
Go to, step 1014,1015 and 1016 can all be performed by typically hearing the robot of voice messaging, but step 1017 and 1018
Only it may be performed by a robot or several robots, each robot can broadcast the work shape of oneself to other robot
State, is performed by most idle robot, the robot then performed can by implementing result by network share to other machines
People.
The embodiment of the present invention listens to the initial time of voice messaging by each robot records, and adds interim order note
Record group, by the initial time for comparing the robot records in interim order record group, it may be determined that hear voice messaging most
At the early moment, relatively complete voice command can be obtained.By determining that definition highest voice messaging is used as in initial time section
Voice command, can obtain relatively complete and clearly voice command, enhance the reliability for obtaining user voice command.Voice is ordered
The clear of order is easy to the correct parsing to institute's speech commands, voice command it is complete, it is ensured that can complete solution as far as possible
The wake-up word that user sends is separated out, is easy to the cooperative cooperating between the corresponding robot of each wake-up word.
Specifically, in some embodiments of the awakening method, as shown in figure 5, it is described according to institute's speech commands and
Default wake-up dictionary, parses the wake-up word and command context in institute's speech commands, including:
Step 1021:The wake-up word in voice command is parsed according to voice command and wake-up dictionary;
Because user may order one or while order multiple robots, language of the user to the voice command of robot
Method form is typically as follows:
<name>, [with | and | also have | that | uh | ...]<name>, [with | and | also have | that | uh | ...]<name
>,……,<command>。
Wherein,<name>For the wake-up word of some robot,<name>Occurrence number can be one or more;
[with | and | also have | that | uh | ...] for two the conjunction that is likely to occur between words is waken up, such as " Zhang San also has Li Si ",
What this conjunction not necessarily occurred, therefore represented with [];
<command>To wake up the command context after word.
For example, user voice command is " Zhang San, also Li Si, king five get out lunch before helping me at ten two points ", this word
Applied mechanically according to above-mentioned syntax format, then can determine that " Zhang San, Li Si, king five " is the wake-up word of three robots, and " also having " is company
Word or pet phrase, " help me ten two points before get out lunch " is command context.
By the way that the wake-up word in voice command and wake-up dictionary is matched one by one, it is possible to parse in voice command
Wake-up word.
Step 1022:The command context in voice command is parsed according to voice command and wake-up word.
Parse after the wake-up word in voice command, it is the actual life of user that the content behind word is waken up in voice command
Content is made, is parsed by the semanteme to the command context, the true purpose with regard to user can be known.
It should be noted that above-mentioned steps 1021 and 1022 are not necessarily carried out to a possible quilt by each robot
One robot or several robots are performed, and each robot can broadcast the working condition of oneself to other robot, by most
Idle robot is performed.
Specifically, in some embodiments of the awakening method, as shown in Figure 6 a, the wake-up word that the basis is parsed
Operation is performed with command context, including:
Step 1031:Command context is notified into robot corresponding with waking up word, so that robot corresponding with waking up word
Perform the corresponding operation of command context.
Can parse that to wake up the robot of word and command context can also wake up word and order by Network Capture
Command context is notified robot corresponding with waking up word by the other robot of content.For example, the voice command of user's issue is
" Mike, Tom, Jerry help me that room is tidied up ", it is assumed that the robot of parsing voice command is Mike, then Mike meetings
Command context " helping me that room is tidied up " is sent to Tom and Jerry, then Mike, Tom and Jeery tidy up room together
Between.Wherein, it can carry out result or language after semantic parsing that Mike, which is sent to Tom and Jeery command context,
Sound is, it is necessary to which Tom and Jerry voluntarily carry out semantic parsing to voice.
Optionally, in the other embodiment of the awakening method, as shown in Figure 6 b, the wake-up word that the basis is parsed
Operation is performed with command context, including:
Step 1032:According to command context task resolution, the task after decomposition is notified into machine corresponding with waking up word respectively
Device people, so that the corresponding operation of robot cooperated execution command context corresponding with waking up word.
In actual applications, robot is parsed after wake-up word and command context, can parse wake-up word and order
The robot of content can also be the other robot that word and command context are waken up by Network Capture, can notify to wake up word pair
The other robot answered waits task, each robot is set up between task groups, task group membership and starts synchronous (such as sharing position
Information, unique identifier, self-ability etc.), then wait task.The robot of voice command is parsed according to command context knot
Position and the self-ability of each robot are closed, the task that user hands over is disassembled, is divided into several subtasks, then will
Subtask after disassembling is sent to times of the robot cooperated completion user explanation in other members in task groups, task groups
Business.For example, by taking Fig. 2 as an example explanation, it is assumed that doing one's room for task is divided into by the machine artificial Mike of parsing voice command, Mike
Three subtasks:Parlor and bedroom are tidied up, toilet is tidied up and tidies up kitchen, Mike can will tidy up toilet and tidy up kitchen
The subtask in room is sent respectively to Tom and Jerry, and oneself tidies up parlor and bedroom.
The embodiment of the present invention is relative to prior art, and the main body for performing task is no longer single robot, but will be many
Individual robot combined common execution task, execution efficiency is higher, and Consumer's Experience is good.
As shown in Figure 7a, be the awakening method one embodiment schematic flow sheet, it is in this embodiment, described to call out
Awake method, including:
Step 201:Monitor voice messaging;
Each robot can monitoring users are sent in real time voice messaging;
Step 202:Whether confirm the voice messaging is order for waking up the corresponding robot of dictionary;
Step 203:If the voice messaging is the order for waking up the corresponding robot of dictionary, record institute predicate
Message breath is used as voice command.
Each robot for listening to voice messaging can confirm whether above-mentioned voice messaging is that user is directed to wake-up dictionary
In the order that sends of the corresponding robot of any wake-up word, if it is, regarding the voice messaging as voice command.
Step 204:According to institute's speech commands and default wake-up dictionary, the wake-up word in institute's speech commands is parsed
And command context.
In this embodiment, each listening to the robot of voice command can parse to voice command, parse and call out
Awake word and command context.
Step 205:Command context is notified into robot corresponding with waking up word, so that robot corresponding with waking up word is held
The corresponding operation of line command content.
The robot of voice command is parsed, if the wake-up word parsed does not include oneself, command context is sent to
Other wake up the corresponding robot of word;If the wake-up word parsed includes oneself, start to perform the corresponding behaviour of command context
Make, and command context is sent to other related robots.If robot itself parses command context, according to itself
The command context parsed performs operation, if itself does not parse command context, the life sent according to other robot
Content is made to perform operation.
Awakening method provided in an embodiment of the present invention, includes several wake-up words for waking up word by being preset in robot
Storehouse, robot can correctly parse the wake-up word contained in voice command according to default wake-up dictionary, so that correctly
Parse the command context in voice command, and then correct the completing user's explanation of the task.User can be to multiple robots
Instruction is simultaneously emitted by, multiple robots can join together to perform the task of user's issue jointly, and execution efficiency is higher, user's body
Test.
As shown in Figure 7b, be the awakening method one embodiment schematic flow sheet, it is in this embodiment, described to call out
Awake method, including:
Step 301:Monitor voice messaging;
Each robot can monitoring users are sent in real time voice messaging;
Step 302:Whether confirm the voice messaging is order for waking up the corresponding robot of dictionary;
Step 303:If the voice messaging is the order for waking up the corresponding robot of dictionary, record institute predicate
Message ceases and listened to the initial time of voice messaging, and adds default interim order record group;
Each robot for listening to voice messaging can confirm whether above-mentioned voice messaging is that user is directed to wake-up dictionary
In the order that sends of the corresponding robot of any wake-up word, if it is, recording the voice messaging and listening to voice
The initial time of information, and add default interim order record group.Then, the robot for adding interim order record group can be to
Other robot, which is broadcasted, oneself adds interim order record group, so that the group member in interim order record group can know other groups
The presence of member.
Step 304:The earliest moment in the initial time of the robot records in interim order record group is confirmed, according to institute
State the earliest moment determine initial time section;
Robot in interim order record group can be into group other robot broadcast oneself record initial time, certain
Individual robot can determine whether the earliest moment in each initial time, and determine initial time section according to the earliest moment.
Step 305:Obtain definition highest voice in the voice messaging that initial time is located in initial time section
Information is used as voice command
After voicemail logging terminates, the robot in interim order record group carries out clear to the voice messaging of self record
Clear degree judges, obtains articulation score value, then the other robot into interim order record group broadcasts the articulation score
Value, some robot can find out the definition highest voice messaging in initial time section.
Step 306:According to institute's speech commands and default wake-up dictionary, the wake-up word in institute's speech commands is parsed
And command context;
Step 307:According to command context task resolution, the task after decomposition is notified into machine corresponding with waking up word respectively
People, so that the corresponding operation of robot cooperated execution command context corresponding with waking up word.
It should be noted that above-mentioned steps 304,305,306 and 307 are not necessarily performed by each robot, Mei Geji
Device people can broadcast the working condition of oneself to other robot, be performed by most idle robot.
Awakening method provided in an embodiment of the present invention, includes several wake-up words for waking up word by being preset in robot
Storehouse, robot can correctly parse the wake-up word contained in voice command according to default wake-up dictionary, so that correctly
Parse the command context in voice command, and then correct the completing user's explanation of the task.Monitored by each robot records
To the initial time of voice messaging, and interim order record group is added, remembered by comparing the robot in interim order record group
The initial time of record, it may be determined that hear the earliest moment of voice messaging, can obtain relatively complete voice command.By determining
Definition highest voice messaging is as voice command in initial time section, can obtain relatively completely with clearly voice command,
Enhance the reliability for obtaining user voice command.User can be simultaneously emitted by instruction to multiple robots, and multiple robots can
To complete the task of user's issue cooperatively, execution efficiency is higher, and Consumer's Experience is good.
Optionally, in the other embodiment of the awakening method, methods described also includes:
Update and wake up dictionary.
When user buys a robot again, and by being set up between network and other robot after communication connection, it is new plus
The robot entered can broadcast the wake-up word of oneself by network to other robot, receive the robot of wake-up word and can wake up this
Word is synchronized to the wake-up dictionary of itself, and the complete or collected works after synchronization are waken up into the robot that dictionary sends back to the new addition.
Optionally, in the other embodiment of the awakening method, methods described also includes:
Dismiss interim order record group.
After task is assigned, each robot starts execution task, can now dismiss interim order record group, release
Internal memory, improves resource utilization.
Accordingly, the embodiment of the present invention additionally provides a kind of robot Rouser, and the Rouser is arranged at Fig. 2 institutes
In any robot shown, as shown in figure 8, the Rouser 400 includes:
Voice command acquisition module 401, for obtaining voice command;
Voice command parsing module 402, for according to institute's speech commands and default wake-up dictionary, parsing institute's predicate
Wake-up word and command context in sound order, the wake-up dictionary include calling out in several wake-up words, the wake-up dictionary
At least two robots of word correspondence of waking up;
Performing module 403, for performing operation according to the wake-up word and command context that parse.
Rouser provided in an embodiment of the present invention, includes several wake-up words for waking up word by being preset in robot
Storehouse, robot can correctly parse the wake-up word contained in voice command according to default wake-up dictionary, so that correctly
Parse the command context in voice command, and then correct the completing user's explanation of the task.
Specifically, in some embodiments of the Rouser, institute's speech commands acquisition module 401 includes:
Voice messaging monitors submodule, for monitoring voice messaging;
Voice command confirms submodule, for confirming whether the voice messaging is for waking up the corresponding robot of dictionary
Order;
First voice command acquisition submodule, if being for waking up the corresponding robot of dictionary for the voice messaging
Order, then record the voice messaging as voice command.
Optionally, in the other embodiment of the Rouser, as shown in figure 9, institute's speech commands acquisition module 401
Including:
Voice messaging monitors submodule 4011, for monitoring voice messaging;
Voice command confirms submodule 4012, for confirming whether the voice messaging is for waking up the corresponding machine of dictionary
The order of device people;
Voice command record sub module 4013, if being for waking up the corresponding robot of dictionary for the voice messaging
Order, then record the voice messaging and listen to the initial time of voice messaging, and add default interim order note
Record group;
When initial time segment acknowledgement submodule 4014, starting for confirming the robot records in interim order record group
At the earliest moment in quarter, initial time section is determined according to the earliest moment;
Second voice command acquisition submodule 4015, for obtaining voice of the initial time in initial time section
Definition highest voice messaging is used as voice command in information.
Specifically, in some embodiments of the Rouser, institute's speech commands confirm that submodule includes:
Voice command confirms subelement, if including the default any wake-up word waken up in dictionary for the voice messaging
And the appearance of the wake-up word is calls, then the voice messaging is the order for waking up the corresponding robot of dictionary.
Optionally, as shown in Figure 10, in the other embodiment of the Rouser, the Rouser 500 is except bag
Include outside voice command acquisition module 501, voice command parsing module 502 and performing module 503, in addition to:
Word library updating module 504 is waken up, dictionary is waken up for updating.
Wherein, in certain embodiments, the wake-up Word library updating module includes:
First wakes up Word library updating submodule, wakes up word for setting instruction to set according to wake-up word, and called out described in broadcast
Awake word.
In further embodiments, the wake-up Word library updating module also includes:
Second wakes up Word library updating submodule, the wake-up word for receiving broadcast, and the wake-up word is added into default call out
Wake up in dictionary, and the wake-up dictionary after renewal is sent to the robot of broadcast wakeup word.
Specifically, in some embodiments of the Rouser, institute's speech commands parsing module includes:
Word analyzing sub-module is waken up, for parsing the wake-up word in voice command according to voice command and wake-up dictionary;
Command context analyzing sub-module, for being parsed according to voice command and wake-up word in the order in voice command
Hold.
Specifically, in some embodiments of the Rouser, the performing module includes:
First implementation sub-module, for command context to be notified into robot corresponding with waking up word, so that with waking up word pair
The robot answered performs the corresponding operation of command context.
In the other embodiment of the Rouser, the performing module includes:
Second implementation sub-module, for according to command context task resolution, the task after decomposition being notified and being waken up respectively
The corresponding robot of word, so that the corresponding operation of robot cooperated execution command context corresponding with waking up word.
As shown in figure 11, be the Rouser one embodiment structural representation, it is in this embodiment, described to call out
Device 600 of waking up includes:
Voice command acquisition module 601, for obtaining voice command;Wherein, institute's speech commands acquisition module 601 is wrapped
Include:
Voice messaging monitors submodule 6011, for monitoring voice messaging;
Voice command confirms submodule 6012, for confirming whether the voice messaging is for waking up the corresponding machine of dictionary
The order of device people;
Voice command record sub module 6013, if being for waking up the corresponding robot of dictionary for the voice messaging
Order, then record the voice messaging and listen to the initial time of voice messaging, and add default interim order note
Record group;
When initial time segment acknowledgement submodule 6014, starting for confirming the robot records in interim order record group
At the earliest moment in quarter, initial time section is determined according to the earliest moment;
Second voice command acquisition submodule 6015, for obtaining voice of the initial time in initial time section
Definition highest voice messaging is used as voice command in information.
Voice command parsing module 602, for according to institute's speech commands and default wake-up dictionary, parsing institute's predicate
Wake-up word and command context in sound order, the wake-up word are used to wake up robot;Wherein, institute's speech commands parsing mould
Block 602 includes:
Word analyzing sub-module 6021 is waken up, for parsing the wake-up in voice command according to voice command and wake-up dictionary
Word;
Command context analyzing sub-module 6022, for parsing the order in voice command according to voice command and wake-up word
Content.
Performing module 603, for performing operation according to the wake-up word and command context that parse, wherein, the execution mould
Block 603 includes:
Second implementation sub-module 6031, for according to command context task resolution, by the task after decomposition notify respectively with
The corresponding robot of word is waken up, so that the corresponding operation of robot cooperated execution command context corresponding with waking up word.
Word library updating module 604 is waken up, dictionary is waken up for updating.
Voice messaging monitors the voice that the real-time monitoring users of submodule 6011 are sent, and voice command confirms that submodule 6012 is true
Whether recognize voice messaging to monitor the voice messaging that listens to of submodule 6011 is user for waking up the corresponding robot of dictionary
Order, if the voice messaging is the order for waking up the corresponding robot of dictionary, voice command record sub module 6013
The voice messaging and the initial time for listening to voice messaging are recorded, and add default interim order record
Group.Robot in interim order record group can broadcast the initial time of oneself record to other robot, and initial time section is really
Recognize submodule 6014 to be confirmed to listen to the earliest moment of user voice command according to each initial time, and according to the earliest moment
Determine initial time section.After voicemail logging terminates, the robot in interim order record group is believed the voice of self record
Cease and judge into line definition, obtain articulation score value, then to broadcast this clear for the other robot into interim order record group
Clear degree fractional value, the second voice command acquisition submodule 6015 determines the voice letter that initial time is located in initial time section
Definition highest voice messaging is used as voice command in breath.Word analyzing sub-module 6021 is waken up to be parsed according to the voice command
Word is waken up, command context analyzing sub-module 6022 parses command context according to voice command and above-mentioned wake-up word.Second performs
Submodule 6031, according to mentioned order content resolution task, promotes robot cooperated execution command context corresponding with waking up word
Corresponding operation.
Rouser provided in an embodiment of the present invention, includes several wake-up words for waking up word by being preset in robot
Storehouse, robot can correctly parse the wake-up word contained in voice command according to default wake-up dictionary, so that correctly
Parse the command context in voice command, and then correct the completing user's explanation of the task.Monitored by each robot records
To the initial time of voice messaging, and interim order record group is added, remembered by comparing the robot in interim order record group
The initial time of record, it may be determined that hear the earliest moment of voice messaging, can obtain relatively complete voice command.By determining
Definition highest voice messaging is as voice command in initial time section, can obtain relatively completely with clearly voice command,
Enhance the reliability for obtaining user voice command.User can be simultaneously emitted by instruction to multiple robots, and multiple robots can
To complete the task of user's issue cooperatively, execution efficiency is higher, and Consumer's Experience is good.
It should be noted that above-mentioned Rouser can perform the awakening method that the embodiment of the present invention is provided, possesses execution
The corresponding functional module of method and beneficial effect.The ins and outs of detailed description not in Rouser embodiment, reference can be made to this
The awakening method that inventive embodiments are provided.
Figure 12 is the hardware architecture diagram of the robot 700 of robot awakening method provided in an embodiment of the present invention, such as
Shown in Figure 12, the robot 700 includes:
In one or more processors 701 and memory 702, Figure 12 by taking a processor 701 as an example.
Processor 701 and memory 702 can be connected by bus or other modes, to be connected by bus in Figure 12
Exemplified by.
Memory 702 is as a kind of non-volatile computer readable storage medium storing program for executing, available for storage non-volatile software journey
The corresponding programmed instruction of awakening method in sequence, non-volatile computer executable program and module, such as embodiment of the present invention/
Module (for example, voice command acquisition module 401, voice command parsing module 402, performing module 403 shown in accompanying drawing 8).Place
Reason device 701 is stored in non-volatile software program, instruction and module in memory 702 by operation, so as to perform service
The various function application of device and data processing, that is, realize the awakening method of above method embodiment.
Memory 702 can include storing program area and storage data field, wherein, storing program area can store operation system
Application program required for system, at least one function;Storage data field can be stored uses created number according to Rouser
According to etc..In addition, memory 702 can include high-speed random access memory, nonvolatile memory can also be included, for example extremely
A few disk memory, flush memory device or other non-volatile solid state memory parts.In certain embodiments, memory
702 is optional including the memory remotely located relative to processor 701, and these remote memories can be by network connection to calling out
Awake device.The example of above-mentioned network includes but is not limited to internet, intranet, LAN, mobile radio communication and combinations thereof.
One or more of modules are stored in the memory 702, when by one or more of processors
During 701 execution, the awakening method in above-mentioned any means embodiment is performed, for example, performing the method step in Fig. 3 described above
Rapid 101 to step 103, the method and step 1011 in Fig. 4 a to step 1013, the method and step 1014 in Fig. 4 b to step 1018,
Method and step 1021 in Fig. 5 is to step 1022, the method and step 1031 in Fig. 6 a, the method and step 1032 in Fig. 6 b, Fig. 7 a
In method and step 201-205, Fig. 7 b in method and step 301-307;Realize submodule in module 401-403, Fig. 9 in Fig. 8
Module 601-604, submodule 6011-6015, submodule 6021- in module 501-504, Figure 11 in block 4011 and 4015, Figure 10
6022nd, the function of submodule 6031.
The said goods can perform the method that the embodiment of the present invention is provided, and possesses the corresponding functional module of execution method and has
Beneficial effect.Not ins and outs of detailed description in the present embodiment, reference can be made to the method that the embodiment of the present invention is provided.
The embodiments of the invention provide a kind of non-volatile computer readable storage medium storing program for executing, the computer-readable storage medium
Matter is stored with computer executable instructions, and the computer executable instructions are executed by one or more processors, such as in Figure 12
A processor 701, may be such that said one or multiple processors can perform the wake-up side in above-mentioned any means embodiment
Method, for example, method and step 101 in Fig. 3 described above is performed to step 103, the method and step 1011 in Fig. 4 a to step
Method and step 1014 in 1013, Fig. 4 b is to step 1018, method and step 1021 in Fig. 5 to step 1022, the side in Fig. 6 a
The method and step 301- in method and step 201-205, Fig. 7 b in method and step 1032 in method step 1031, Fig. 6 b, Fig. 7 a
307;Realize in module 401-403, Fig. 9 Neutron module 4011 and 4015, Figure 10 in Fig. 8 module in module 501-504, Figure 11
601-604, submodule 6011-6015, submodule 6021-6022, the function of submodule 6031.
Device embodiment described above is only schematical, wherein the unit illustrated as separating component can
To be or may not be physically separate, the part shown as unit can be or may not be physics list
Member, you can with positioned at a place, or can also be distributed on multiple NEs.It can be selected according to the actual needs
In some or all of module realize the purpose of this embodiment scheme.
Through the above description of the embodiments, those of ordinary skill in the art can be understood that each embodiment
The mode of general hardware platform can be added to realize by software, naturally it is also possible to pass through hardware.Those of ordinary skill in the art can
Realize that all or part of flow in above-described embodiment method can be by computer program to instruct the hard of correlation with understanding
Part is completed, and described program can be stored in a computer read/write memory medium, the program is upon execution, it may include as above
State the flow of the embodiment of each method.Wherein, described storage medium can be magnetic disc, CD, read-only memory (Read-
Only Memory, ROM) or random access memory (Random Access Memory, RAM) etc..
Finally it should be noted that:The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;At this
Under the thinking of invention, it can also be combined between the technical characteristic in above example or non-be the same as Example, step can be with
Realized with random order, and there are many other changes of the different aspect of the present invention as described above, for simplicity, they do not have
Have and provided in details;Although the present invention is described in detail with reference to the foregoing embodiments, the ordinary skill people of this area
Member should be understood:It can still modify to the technical scheme described in foregoing embodiments, or to which part skill
Art feature carries out equivalent substitution;And these modifications or replacement, the essence of appropriate technical solution is departed from each reality of the invention
Apply the scope of a technical scheme.