CN108039174A - Speech recognition system, method and apparatus - Google Patents
Speech recognition system, method and apparatus Download PDFInfo
- Publication number
- CN108039174A CN108039174A CN201810015848.XA CN201810015848A CN108039174A CN 108039174 A CN108039174 A CN 108039174A CN 201810015848 A CN201810015848 A CN 201810015848A CN 108039174 A CN108039174 A CN 108039174A
- Authority
- CN
- China
- Prior art keywords
- voice
- speech recognition
- microphone array
- plate
- recognition result
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/20—Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/223—Execution procedure of a spoken command
Abstract
The invention discloses a kind of speech recognition system, method and apparatus.Wherein, which includes：Microphone array and speech recognition apparatus, wherein, microphone array, for gathering voice；Speech recognition apparatus includes：Voice plate and loudspeaker, wherein, voice plate, communicates with microphone array, for receiving the voice of microphone array collection, and docks received voice and is identified, obtains recognition result；Loudspeaker, with voice board communications, the recognition result for being identified to voice plate is reported；Wherein, microphone array is placed outside speech recognition apparatus.The present invention is solved since acoustic echo caused by microphone array and the limitation of loudspeaker installation site eliminates the technical problem of AEC debugging and place electrical structure difficult design.
Description
Technical field
The present invention relates to air-conditioning voice control field, in particular to a kind of speech recognition system, method and apparatus.
Background technology
In the related art, when being controlled to air-conditioning, the mode of control is varied, for example, can be directly using sky
Adjust controller to be controlled, can also be controlled according to gesture, can also be controlled according to voice.In air-conditioning voice control
In, speech recognition system is mainly made of microphone array, voice module, loudspeaker, in correlation technique when being designed on product,
Often microphone array design in product structure, while loudspeaker is also built in product structure, by both installation sites
Limitation, microphone array and loudspeaker distance are closer, and the acoustic echo be easy to causeing in speech recognition system eliminates AEC
(Acoustic Echo Chancellor) debugging is relatively difficult, while structure design and appearance design demand also more difficult knot
Close.
For it is above-mentioned the problem of, not yet propose effective solution at present.
The content of the invention
An embodiment of the present invention provides a kind of speech recognition system, method and apparatus, at least to solve due to microphone array
Acoustic echo caused by row and the limitation of loudspeaker installation site eliminates AEC debugging and the technology of place electrical structure difficult design is asked
Topic.
One side according to embodiments of the present invention, there is provided a kind of speech recognition system, including：Microphone array and language
Sound identification equipment, wherein, microphone array, for gathering voice；The speech recognition apparatus includes：Voice plate and loudspeaker,
Wherein, the voice plate, communicates with the microphone array, for receiving the voice of the microphone array collection, and docks
The received voice is identified, and obtains recognition result；The loudspeaker, and voice board communications, for institute's predicate
The recognition result that soundboard identifies is reported；Wherein, the microphone array is placed outside the speech recognition apparatus.
Optionally, the voice plate is multiple that the multiple voice plate is located in a different geographical location respectively.
Another aspect according to embodiments of the present invention, additionally provides a kind of audio recognition method, including：Receive microphone array
Arrange the voice of collection；The voice of the microphone array collection is identified by the voice plate in speech recognition apparatus, is identified
As a result；The recognition result identified by the loudspeaker report voice plate in the speech recognition apparatus is reported, its
In, the microphone array is placed outside the speech recognition apparatus.
Optionally, the voice of the microphone array collection is identified by the voice plate in speech recognition apparatus, obtains institute
Stating recognition result includes：Determine the speech recognition modeling for speech recognition, wherein, the speech recognition modeling is using multigroup
Data show that every group of data in the multi-group data include by machine learning training：Voice and corresponding with the voice
Recognition result；By the voice plate in speech recognition apparatus using the definite speech recognition modeling, the wheat is identified
The corresponding recognition result of voice of gram wind array acquisition.
Optionally, determine that the speech recognition modeling for speech recognition includes：Sample different age group, different tone colors
User voice, and recognition result corresponding with the voice of sampling；Voice to sampling and corresponding with the voice of sampling
Recognition result be trained, obtain the speech recognition modeling.
Optionally, the voice plate in by the speech recognition apparatus identifies the language of the microphone array collection
Sound, before obtaining the recognition result, further includes：The voice plate in the speech recognition apparatus is multiple situation
Under, receive wake-up word；The voice plate of speech recognition will be carried out by being waken up according to the wake-up word.
Optionally, the voice plate in by the speech recognition apparatus identifies the language of the microphone array collection
Sound, after obtaining the recognition result, further includes：Parsed from the recognition result for controlling the control of predetermined electric appliance to refer to
Order；The control instruction is sent to the electric appliance master control for controlling the predetermined electric appliance.
Another aspect according to embodiments of the present invention, additionally provides a kind of speech recognition equipment, including：First receives mould
Block, for receiving the voice of microphone array collection；Module is obtained, for identifying institute by the voice plate in speech recognition apparatus
The voice of microphone array collection is stated, obtains recognition result；Broadcasting module, for passing through raising one's voice in the speech recognition apparatus
Device is reported the recognition result that the voice plate identifies and is reported, wherein, the microphone array, which is placed outside the voice, to be known
Other equipment.
Optionally, the module that obtains includes：Determination unit, for determining the speech recognition modeling for speech recognition,
Wherein, the speech recognition modeling is drawn using multi-group data by machine learning training, every in the multi-group data
Group data include：Voice and recognition result corresponding with the voice；Recognition unit, for passing through the language in speech recognition apparatus
Soundboard identifies the corresponding recognition result of voice that the microphone array gathers using the definite speech recognition modeling.
Optionally, the speech recognition equipment further includes：Second receiving module, for passing through the speech recognition apparatus
In the voice plate identify the voice of microphone array collection, before obtaining the recognition result, know in the voice
In the case that the voice plate in other equipment is multiple, wake-up word is received；Wake-up module, for being called out according to the wake-up word
The voice plate of speech recognition will be carried out by waking up.
Optionally, the speech recognition equipment further includes：Parsing module, in by the speech recognition apparatus
The voice plate identifies the voice of the microphone array collection, after obtaining the recognition result, from the recognition result
Parse the control instruction for controlling predetermined electric appliance；Sending module, it is described pre- for the control instruction to be sent to control
Determine the electric appliance master control of electric appliance.
In embodiments of the present invention, by the way of external microphone wind array, there is provided one kind include microphone array and
The speech recognition system of speech recognition apparatus, wherein, microphone array, for gathering voice；Speech recognition apparatus includes：Voice
Plate and loudspeaker, wherein, voice plate, communicates with microphone array, for receiving the voice of microphone array collection, and to receiving
To voice be identified, obtain recognition result；Loudspeaker, communicates with microphone array, for the knowledge identified to voice plate
Other result is reported.By the speech recognition system of the embodiment of the present invention, reach and realized microphone array with loudspeaker certainly
By the purpose combined, it is achieved thereby that improve speech recognition anti-acoustic capability and meet the technique effect of appearance consistency requirement, into
And solve since acoustic echo caused by microphone array and the limitation of loudspeaker installation site eliminates AEC debugging and place electric appliance
The technical problem of structure design difficulty.
Brief description of the drawings
Attached drawing described herein is used for providing a further understanding of the present invention, forms the part of the application, this hair
Bright schematic description and description is used to explain the present invention, does not form inappropriate limitation of the present invention.In the accompanying drawings：
Fig. 1 is a kind of structure diagram of speech recognition system 10 according to embodiments of the present invention；
Fig. 2 is the flow chart of audio recognition method according to embodiments of the present invention；
Fig. 3 is the structure diagram of another speech recognition system 30 of preferred embodiment according to the present invention；
Fig. 4 is the single-link voice control schematic diagram of the air-conditioning of preferred embodiment according to the present invention；
Fig. 5 is the multilink voice control schematic diagram of the air-conditioning of preferred embodiment according to the present invention；
Fig. 6 is the structure diagram of speech recognition equipment according to embodiments of the present invention；
Fig. 7 is that speech recognition equipment according to embodiments of the present invention obtains the structure diagram of module 64；
Fig. 8 is the preferred structure block diagram one of speech recognition equipment according to embodiments of the present invention；
Fig. 9 is the preferred structure block diagram two of speech recognition equipment according to embodiments of the present invention.
Embodiment
In order to make those skilled in the art more fully understand the present invention program, below in conjunction with the embodiment of the present invention
Attached drawing, is clearly and completely described the technical solution in the embodiment of the present invention, it is clear that described embodiment is only
The embodiment of a part of the invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill people
Member's all other embodiments obtained without making creative work, should all belong to the model that the present invention protects
Enclose.
It should be noted that term " first " in description and claims of this specification and above-mentioned attached drawing, "
Two " etc. be for distinguishing similar object, without for describing specific order or precedence.It should be appreciated that so use
Data can exchange in the appropriate case, so as to the embodiment of the present invention described herein can with except illustrating herein or
Order beyond those of description is implemented.In addition, term " comprising " and " having " and their any deformation, it is intended that cover
Cover it is non-exclusive include, be not necessarily limited to for example, containing the process of series of steps or unit, method, system, product or equipment
Those steps or unit clearly listed, but may include not list clearly or for these processes, method, product
Or the intrinsic other steps of equipment or unit.
In embodiments of the present invention, there is provided a kind of speech recognition system, Fig. 1 are a kind of languages according to embodiments of the present invention
The structure diagram of sound identifying system 10, as shown in Figure 1, the system includes：Microphone array 12 and speech recognition apparatus 14, below
The speech recognition system 10 is illustrated.
Microphone array 12, for gathering voice；
Speech recognition apparatus 14, including：Voice plate 142 and loudspeaker 144, wherein, voice plate 142, with microphone array
12 communications, for receiving the voice of the collection of microphone array 12, and dock received voice and are identified, obtain recognition result；
Loudspeaker 144, communicates with voice plate 142, and the recognition result for being identified to voice plate 142 is reported；
Wherein, microphone array 12 is placed outside speech recognition apparatus 14.
Meanwhile in order to improve the voice control convenience in more spaces, it is preferred that voice plate 142 can be it is multiple, its
In, multiple voice plates can be located in a different geographical location respectively, so that the voice that microphone array 12 receives can be at the same time
Identification is handled by the voice plate of multiple positions, and then voice control is carried out to the electric appliance residing for each voice plate.
In embodiments of the present invention, by the way of external microphone wind array 12, there is provided one kind includes microphone array
12 and the speech recognition system 10 of speech recognition apparatus 14.By the speech recognition system of the embodiment of the present invention, realization is reached
The purpose of microphone array and speech recognition apparatus independent assortment, it is achieved thereby that improving speech recognition anti-acoustic capability and meeting institute
In the technique effect of electric appliance appearance design coherence request.
According to embodiments of the present invention, a kind of embodiment of the method for speech recognition is additionally provided, it is necessary to illustrate, in attached drawing
Flow the step of illustrating can be performed in the computer system of such as a group of computer-executable instructions, although also,
Show logical order in flow charts, but in some cases, can with different from order herein perform it is shown or
The step of description.
Fig. 2 is the flow chart of audio recognition method according to embodiments of the present invention, as shown in Fig. 2, this method is including as follows
Step：
Step S202, receives the voice of microphone array collection；
Step S204, the voice of microphone array collection is identified by the voice plate in speech recognition apparatus, is identified
As a result；
Step S206, the recognition result identified by the loudspeaker report voice plate in speech recognition apparatus are broadcast
Report, wherein, microphone array is placed outside speech recognition apparatus.
Pass through above-mentioned steps, it is possible to achieve in embodiments of the present invention, speech recognition is placed outside by microphone array and is set
Standby mode, achievees the purpose that microphone array and loudspeaker being freely combined, it is achieved thereby that it is noise reduction to improve speech recognition
The technique effect of electric appliance appearance consistency requirement where and meeting, and then solve due to microphone array and loudspeaker installation
Acoustic echo caused by position limits eliminates the technical problem of AEC debugging and place electrical structure difficult design.
Preferably, the voice of microphone array collection is identified by the voice plate in speech recognition apparatus, obtains identification knot
Fruit can include：Determine the speech recognition modeling for speech recognition, wherein, speech recognition modeling is to be passed through using multi-group data
Machine learning training show that every group of data in multi-group data include：Voice and recognition result corresponding with the voice；It is logical
The voice plate in speech recognition apparatus is crossed using definite speech recognition modeling, identifies that the voice of microphone array collection corresponds to
Recognition result.The voice of collection is identified by way of above-mentioned speech recognition modeling, i.e., using the side of artificial intelligence
The voice of collection is identified in formula, and not only intelligence is quick but also accurate, can effectively improve user's body to a certain extent
Test.
It should be noted that above-mentioned every group of training data is obtained by experiment or widely applies this
Constantly collection accumulates what is reported to the electric appliance of speech recognition apparatus in use, passes through and the electric appliance sold away is carried out
Tracking, may be incorporated for training so as to obtain substantial amounts of data.Optionally, also may be used in the electric appliance of the application speech recognition apparatus
To pre-set communication module, during multiple electric appliances can upload onto the server the data collected in real time, so that machine is trained
Use.Wherein, communication module can include but is not limited to：Wireless network card, bluetooth etc..
For the problem that user type scope present in voice control is wide, since maloperation easily occurs for accent, age, really
Surely being used for the speech recognition modeling of speech recognition can include：Sampling different age group, the voice of the user of different tone colors, and
Recognition result corresponding with the voice of sampling；Voice and recognition result corresponding with the voice of sampling to sampling are instructed
Practice, obtain speech recognition modeling.By the above method, sampling instruction can be carried out to different age group, the voice of different tone colors
Practice so that the speech recognition modeling trained more fully, so as to effectively improve the knowledge identified using the speech recognition modeling
Not as a result, making it more accurate.
Alternatively, to realize different control to different crowds, or realize some cannot be allowed to perform what is controlled
Personnel limit control electric appliance, and the voice of voice plate identification microphone array collection that can be in by speech recognition apparatus, obtains
To before recognition result, the user identity of the corresponding user of voice of collection is determined；It is guardian's in the user identity of user
In the case of, pass through the voice of the voice plate identification microphone array collection in speech recognition apparatus.Pass through the voice to collection
The identity of user, determines whether the user possesses the authority by voice control electric appliance.For example, if the voice of collection corresponds to
In the case of children, the control instruction for not performing the voice of collection children is set, on the one hand it is possible to prevente effectively from the uneasiness of operation
Quan Xing, and the intentional or unintentional maloperation of children is effectively avoided to a certain extent.
It should be noted that it is determined here that the user identity of the corresponding user of voice of collection can use various ways,
For example, it according to the tone color of the predicate sound of collection, can determine the user identity of the corresponding user of voice of collection；Can also basis
The volume of the predicate sound of collection, determines the user identity of the corresponding user of voice of collection；Can also be according to the predicate sound of collection
Tone, determine collection the corresponding user of voice user identity.
In addition, in order to improve the voice control convenience in more spaces, the voice plate in speech recognition apparatus can be more
It is a, wherein, multiple voice plates can be located in a different geographical location respectively, so that the voice that microphone array receives can be same
When by multiple positions voice plate handle identification, and then to residing for each voice plate electric appliance carry out voice control.For realization pair
Multiple voice plate identification voices carry out flexibly accurate control, it is preferred that the voice plate identification wheat in by speech recognition apparatus
The voice of gram wind array acquisition, before obtaining recognition result, can also include：Voice plate in speech recognition apparatus is multiple
In the case of, receive wake-up word；The voice plate of speech recognition will be carried out according to word wake-up is waken up.Wherein, waking up word can be with
It is fixedly installed when being manufactured for the speech recognition apparatus, or voluntarily set during user's use.By to difference
Voice plate identification voice different wake-up word is set, on the one hand not only can effectively realize the voice plate to multiple and different positions
Control, and the accuracy of control can be effectively improved, effectively improve the intelligentized experience of user.
Preferably, the voice of the voice plate identification microphone array collection in by speech recognition apparatus, is identified
As a result after, can also include：The control instruction for controlling predetermined electric appliance is parsed from recognition result；Control instruction is sent out
Give the electric appliance master control for controlling predetermined electric appliance.By parsing control instruction from recognition result, and it is sent to predetermined electric appliance
Electric appliance master control, realize the complete control to electric appliance.It should be noted that the species of electric appliance herein can be a variety of, for example,
It can be air-conditioning, can be refrigerator, can be humidifier etc..
It is above-mentioned that different wake-up words is set to different voice plate, to realize voice pair that same microphone array receives
It can identify that multiple voice plates of voice realize Dock With Precision Position at the same time, and then avoid that the voice control to electric appliance where voice plate occurs
Mistake processed.
In addition, when the control instruction identified by way of above-mentioned artificial intelligence is controlled electric appliance, can also
Compatibility is controlled electric appliance by appliance controller, for example, being controlled when being identified by way of artificial intelligence to electric appliance
While the control instruction of system, also receive appliance controller and the controller of electric appliance is instructed, by setting control instruction and control
The mode of the priority of device processed instruction performs different control.For example, work as the control instruction for setting artificial intelligence to identify
In the case of priority of the priority higher than controller instruction, performed according to the control instruction that artificial intelligence identifies to electric appliance
Control；In the case of the control instruction for setting the priority that controller instructs to be identified higher than artificial intelligence, according to controller
Instruction performs the control to electric appliance.
In embodiments of the present invention, another speech recognition system is additionally provided, Fig. 3 is the side of being preferable to carry out according to the present invention
The structure diagram of another speech recognition system 30 of case, as shown in figure 3, the system includes：Microphone array 32, voice plate
34, loudspeaker 36, below illustrates the speech recognition system.
Above-mentioned speech recognition system 30 includes：Microphone array 32 (with above-mentioned microphone array 12), voice plate 34 are (ibid
Predicate soundboard 142), loudspeaker 36 (with above-mentioned loudspeaker 144).Wherein, above-mentioned microphone array 32 can be placed outside voice plate
34 and loudspeaker 36.
Microphone, the signal processing DSP (Digital of pickup can be integrated in the external microphone array 32
Signal Process) chip and communication module, wherein, above-mentioned microphone can be electret or silicon wheat, and can include
Two, the microphone of four or more.Above-mentioned communication module is used to carry out wireless connection between voice plate 34, can be blue
Tooth, wireless telecommunications Zigbee and WIFI (Wireless Fidelity) etc..Meanwhile also it is built-in with power supply in microphone array 32
Management module, can be powered by external power supply or built-in rechargeable battery is powered.
Built-in communication module and phonation unit body in loudspeaker 36, while power management module is also built-in with, can also be outer
Connect power supply power supply or built-in rechargeable battery power supply.
Above-mentioned microphone array 32 is wirelessly attached with voice plate 34, and loudspeaker 36 is carried out with voice plate 34
Wireless connection, wherein, in electric appliance where voice plate 34 can be designed in the speech recognition system.
Optionally, above-mentioned microphone array 32 can be individually integrally formed, and be placed on room Anywhere, Ke Yishi
At the top of tea table or ceiling；Loudspeaker 36 can also be placed on room Anywhere, can be beside air-conditioning, or smallpox
At the top of plate.
Preferably, voice plate 34 is connected with the electric appliance master control of the speech recognition system 30, when user carries out speech recognition,
Microphone array 32 receives right instructions, and the parsing of phonetic order is carried out by voice plate 34, sends and refers to after resolve command word
Make and give electric appliance master control, master control receives after control command and then goes to control corresponding electric appliance load to work.
By the external speech recognition system of above-mentioned microphone array, pulled open microphone array 32 and loudspeaker 36 away from
From, help to lift the optimization of acoustic echo elimination AEC, and then solve because microphone array is near with loudspeaker, cause noise reduction
The problem of poor.
In embodiments of the present invention, a kind of air-conditioning for including above-mentioned speech recognition system 30 is additionally provided, Fig. 4 is according to this
The single-link voice control schematic diagram of the air-conditioning of invention preferred embodiment, Fig. 5 are the skies of preferred embodiment according to the present invention
The multilink voice control schematic diagram of tune, as shown in figure 4, above-mentioned speech recognition system 30 can include 1 voice plate 34, it is optional
, as shown in figure 5, the speech recognition system 30 can also include multiple voice plates 34, wherein, multiple voice plates 34 can be distinguished
Lie in the air-conditioning of diverse geographic location, a microphone array can be with multiple 34 companies of communicating wirelessly of voice plate
Connect.For example, a microphone array is set in parlor, while bedroom room and parlor room are respectively provided with the sky with voice plate
Adjust, user is set by application APP (Application), and two voice plates are matched somebody with somebody with this microphone array at the same time
To connection, and it is bedroom air-conditioning that can set No. 1 in APP sets interface, while the self-defined wake-up word on APP, such as
" bedroom bedroom ", after being provided with and preserves；Same setting 2 is parlor air-conditioning, and the self-defined wake-up word on APP,
Such as " parlor parlor ", after being provided with and preserve.
Loudspeaker 36 can share a configuration at the same time, as shown in figure 5, can also be independently connected with each voice plate 34,
This connection mode can be completed to set on APP.When user is in parlor, to the voice air conditioner using parlor, it need to only say and call out
Wake up word " parlor parlor ", carry out the identification and control of corresponding airconditioning control order again after waking up speech recognition system, obtain correct
After identification, loudspeaker 36 can carry out report feedback, and equally, will go back bedroom from parlor, " crouch in bedroom as long as being said in parlor and waking up word
Room ", after waking up speech recognition system, bedroom airconditioning control is carried out with corresponding airconditioning control order, and loudspeaker 36 can be with
Result feedback is carried out with the loudspeaker in parlor to report.
It should be noted that such a voice control mode can cover the household appliances of all voice controls, and the voice is known
Other function can include identified off-line, online recognition, the identification of offline and on-line mixing.
Meanwhile the design method of above-mentioned speech recognition system 30, it is possible to achieve free group of microphone array and loudspeaker
Close, and then effective effect for solving lifting acoustic echo and eliminating AEC, discrimination is improved, and realize the place of speech recognition system 30
The product versatility design of electric appliance, not because microphone array perforate causes appearance impacted, meets appearance consistency requirement.
In embodiments of the present invention, a kind of speech recognition equipment is additionally provided, Fig. 6 is voice according to embodiments of the present invention
The structure diagram of identification device, as shown in fig. 6, the device includes：First receiving module 62, obtains module 64, broadcasting module 66.
The speech recognition equipment is illustrated below.
First receiving module 62, for receiving the voice of microphone array collection；
Module 64 is obtained, is connected to above-mentioned first receiving module 62, for being known by the voice plate in speech recognition apparatus
The voice of other microphone array collection, obtains recognition result；
Broadcasting module 66, be connected to it is above-mentioned obtain module 64, for reporting language by loudspeaker in speech recognition apparatus
The recognition result that soundboard identifies is reported, wherein, microphone array is placed outside speech recognition apparatus.
Fig. 7 is that speech recognition equipment according to embodiments of the present invention obtains the structure diagram of module 64, as shown in fig. 7,
This, which obtains module 64, includes：Determination unit 72, recognition unit 74.Module 64 is obtained to this below to illustrate.
Determination unit 72, for determining the speech recognition modeling for speech recognition, wherein, speech recognition modeling is use
Multi-group data show that every group of data in multi-group data include by machine learning training：Voice and corresponding with the voice
Recognition result；
Recognition unit 74, is connected to above-mentioned determination unit 72, true for being used by the voice plate in speech recognition apparatus
Fixed speech recognition modeling, identifies the corresponding recognition result of voice of microphone array collection.
Fig. 8 is the preferred structure block diagram one of speech recognition equipment according to embodiments of the present invention, as shown in figure 8, the voice
Identification device in addition to all structures, is further included in containing Fig. 6：Second receiving module 82, wake-up module 84.Below to the speech recognition
Device illustrates.
Second receiving module 82, is connected to above-mentioned first receiving module 62, for the language in by speech recognition apparatus
The voice of soundboard identification microphone array collection, before obtaining recognition result, the voice plate in speech recognition apparatus is multiple
In the case of, receive wake-up word；
Wake-up module 84, is connected to above-mentioned second receiving module 82 and obtains module 64, will for being waken up according to wake-up word
Carry out the voice plate of speech recognition.
Fig. 9 is the preferred structure block diagram two of speech recognition equipment according to embodiments of the present invention, as shown in figure 9, the voice
Identification device in addition to all structures, is further included in containing Fig. 6：Parsing module 92, sending module 94.Below to the speech recognition equipment
Illustrate.
Parsing module 92, be connected to it is above-mentioned obtain module 64, in by speech recognition apparatus voice plate identification
The voice of microphone array collection, after obtaining recognition result, parses the control for controlling predetermined electric appliance from recognition result
System instruction；
Sending module 94, is connected to above-mentioned parsing module 92, and the electricity of predetermined electric appliance is controlled for control instruction to be sent to
Device master control.
Another aspect according to embodiments of the present invention, additionally provides a kind of storage medium, which includes storage
Program, wherein, equipment performs the audio recognition method of above-mentioned any one where controlling storage medium when program is run.
Another aspect according to embodiments of the present invention, additionally provides a kind of processor, which is used for operation program, its
In, program performs the audio recognition method of above-mentioned any one when running.
The embodiments of the present invention are for illustration only, do not represent the quality of embodiment.
In the above embodiment of the present invention, the description to each embodiment all emphasizes particularly on different fields, and does not have in some embodiment
The part of detailed description, may refer to the associated description of other embodiment.
In several embodiments provided herein, it should be understood that disclosed technology contents, can pass through others
Mode is realized.Wherein, device embodiment described above is only schematical, such as the division of the unit, Ke Yiwei
A kind of division of logic function, can there is an other dividing mode when actually realizing, for example, multiple units or component can combine or
Person is desirably integrated into another system, or some features can be ignored, or does not perform.Another, shown or discussed is mutual
Between coupling, direct-coupling or communication connection can be INDIRECT COUPLING or communication link by some interfaces, unit or module
Connect, can be electrical or other forms.
The unit illustrated as separating component may or may not be physically separate, be shown as unit
The component shown may or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On unit.Some or all of unit therein can be selected to realize the purpose of this embodiment scheme according to the actual needs.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, can also
That unit is individually physically present, can also two or more units integrate in a unit.Above-mentioned integrated list
Member can both be realized in the form of hardware, can also be realized in the form of SFU software functional unit.
If the integrated unit is realized in the form of SFU software functional unit and is used as independent production marketing or use
When, it can be stored in a computer read/write memory medium.Based on such understanding, technical scheme is substantially
The part to contribute in other words to the prior art or all or part of the technical solution can be in the form of software products
Embody, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or network equipment etc.) perform each embodiment the method for the present invention whole or
Part steps.And foregoing storage medium includes：USB flash disk, read-only storage (ROM, Read-Only Memory), arbitrary access are deposited
Reservoir (RAM, Random Access Memory), mobile hard disk, magnetic disc or CD etc. are various can be with store program codes
Medium.
The above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (11)
- A kind of 1. speech recognition system, it is characterised in that including：Microphone array and speech recognition apparatus, wherein,Microphone array, for gathering voice；The speech recognition apparatus includes：Voice plate and loudspeaker, wherein,The voice plate, communicates with the microphone array, for receiving the voice of the microphone array collection, and to receiving To the voice be identified, obtain recognition result；The loudspeaker, with the voice board communications, the recognition result for being identified to the voice plate is reported；Wherein, the microphone array is placed outside the speech recognition apparatus.
- 2. speech recognition system according to claim 1, it is characterised in that the voice plate is multiple, multiple voice plates It is located in a different geographical location respectively.
- A kind of 3. audio recognition method, it is characterised in that including：Receive the voice of microphone array collection；The voice of the microphone array collection is identified by the voice plate in speech recognition apparatus, obtains recognition result；The recognition result identified by the loudspeaker report voice plate in the speech recognition apparatus is reported, its In, the microphone array is placed outside the speech recognition apparatus.
- 4. audio recognition method according to claim 3, it is characterised in that known by the voice plate in speech recognition apparatus The voice of not described microphone array collection, obtaining the recognition result includes：Determine the speech recognition modeling for speech recognition, wherein, the speech recognition modeling is to pass through machine using multi-group data Device learning training show that every group of data in the multi-group data include：Voice and recognition result corresponding with the voice；By the voice plate in speech recognition apparatus using the definite speech recognition modeling, the microphone array is identified The corresponding recognition result of voice of collection.
- 5. audio recognition method according to claim 4, it is characterised in that determine the voice knowledge for speech recognition Other model includes：Sample different age group, the voice of the user of different tone colors, and recognition result corresponding with the voice of sampling；Voice and recognition result corresponding with the voice of sampling to sampling are trained, and obtain the speech recognition modeling.
- 6. audio recognition method according to claim 3, it is characterised in that the institute in by the speech recognition apparatus Predicate soundboard identifies the voice of the microphone array collection, before obtaining the recognition result, further includes：In the case that the voice plate in the speech recognition apparatus is multiple, wake-up word is received；The voice plate of speech recognition will be carried out by being waken up according to the wake-up word.
- 7. the audio recognition method according to any one of claim 3 to 6, it is characterised in that know by the voice The voice plate in other equipment identifies the voice of the microphone array collection, after obtaining the recognition result, further includes：The control instruction for controlling predetermined electric appliance is parsed from the recognition result；The control instruction is sent to the electric appliance master control for controlling the predetermined electric appliance.
- A kind of 8. speech recognition equipment, it is characterised in that including：First receiving module, for receiving the voice of microphone array collection；Module is obtained, for identifying the voice of the microphone array collection by the voice plate in speech recognition apparatus, is obtained Recognition result；Broadcasting module, for the recognition result identified by the loudspeaker report voice plate in the speech recognition apparatus Reported, wherein, the microphone array is placed outside the speech recognition apparatus.
- 9. speech recognition equipment according to claim 8, it is characterised in that the module that obtains includes：Determination unit, for determining the speech recognition modeling for speech recognition, wherein, the speech recognition modeling is using more Group data show that every group of data in the multi-group data include by machine learning training：Voice and recognition result corresponding with the voice；Recognition unit, for using the definite speech recognition modeling by the voice plate in speech recognition apparatus,Identify the corresponding recognition result of voice of the microphone array collection.
- 10. speech recognition equipment according to claim 8, it is characterised in that further include：Second receiving module, identifies that the microphone array is adopted for the voice plate in by the speech recognition apparatus The voice of collection, before obtaining the recognition result, in the case that the voice plate in the speech recognition apparatus is multiple, Receive wake-up word；Wake-up module, the voice plate of speech recognition will be carried out for being waken up according to the wake-up word.
- 11. the speech recognition equipment according to any one of claim 8 to 10, it is characterised in that further include：Parsing module, the microphone array collection is identified for the voice plate in by the speech recognition apparatus Voice, after obtaining the recognition result, parses the control instruction for controlling predetermined electric appliance from the recognition result；Sending module, the electric appliance master control of the predetermined electric appliance is controlled for the control instruction to be sent to.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810015848.XA CN108039174A (en) | 2018-01-08 | 2018-01-08 | Speech recognition system, method and apparatus |
| PCT/CN2018/118962 WO2019134473A1 (en) | 2018-01-08 | 2018-12-03 | Speech recognition system, method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810015848.XA CN108039174A (en) | 2018-01-08 | 2018-01-08 | Speech recognition system, method and apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108039174A true CN108039174A (en) | 2018-05-15 |
Family
ID=62099339
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810015848.XA Pending CN108039174A (en) | 2018-01-08 | 2018-01-08 | Speech recognition system, method and apparatus |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN108039174A (en) |
| WO (1) | WO2019134473A1 (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108712704A (en) * | 2018-06-13 | 2018-10-26 | 腾讯科技（深圳）有限公司 | Speaker, the playback method of audio data, device, storage medium and electronic device |
| CN109119071A (en) * | 2018-09-26 | 2019-01-01 | 珠海格力电器股份有限公司 | A kind of training method and device of speech recognition modeling |
| WO2019134473A1 (en) * | 2018-01-08 | 2019-07-11 | 珠海格力电器股份有限公司 | Speech recognition system, method and apparatus |
| CN110837234A (en) * | 2018-08-17 | 2020-02-25 | 阿里巴巴集团控股有限公司 | Intelligent voice control panel and panel switch socket |
| CN110868648A (en) * | 2018-08-27 | 2020-03-06 | 杭州海康威视数字技术股份有限公司 | Intelligent voice realization method of indoor intercom device and indoor intercom device |
| CN110986293A (en) * | 2019-12-12 | 2020-04-10 | 珠海格力电器股份有限公司 | Voice board assembly and air conditioner |
| CN111128194A (en) * | 2019-12-31 | 2020-05-08 | 云知声智能科技股份有限公司 | System and method for improving online voice recognition effect |
| CN111182412A (en) * | 2019-12-31 | 2020-05-19 | 联想(北京)有限公司 | Electronic equipment, data processing method for electronic equipment and conference system equipment |
| CN112731831A (en) * | 2020-12-18 | 2021-04-30 | 宁波向往智能科技有限公司 | Intelligent switch panel |
| CN113819585A (en) * | 2021-09-16 | 2021-12-21 | 青岛海尔空调器有限总公司 | Microphone device, method and device for matching voice air conditioner microphone and air conditioner |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103186232A (en) * | 2011-12-30 | 2013-07-03 | 上海博泰悦臻电子设备制造有限公司 | Voice keyboard device |
| CN104538030A (en) * | 2014-12-11 | 2015-04-22 | 科大讯飞股份有限公司 | Control system and method for controlling household appliances through voice |
| CN105931633A (en) * | 2016-05-30 | 2016-09-07 | 深圳市鼎盛智能科技有限公司 | Speech recognition method and system |
| CN106679326A (en) * | 2017-01-25 | 2017-05-17 | 北京通远科技有限公司 | Intelligent refrigerator controlled on basis of voice recognition |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107123421A (en) * | 2017-04-11 | 2017-09-01 | 广东美的制冷设备有限公司 | Sound control method, device and home appliance |
| CN108039174A (en) * | 2018-01-08 | 2018-05-15 | 珠海格力电器股份有限公司 | Speech recognition system, method and apparatus |
-
2018
- 2018-01-08 CN CN201810015848.XA patent/CN108039174A/en active Pending
- 2018-12-03 WO PCT/CN2018/118962 patent/WO2019134473A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103186232A (en) * | 2011-12-30 | 2013-07-03 | 上海博泰悦臻电子设备制造有限公司 | Voice keyboard device |
| CN104538030A (en) * | 2014-12-11 | 2015-04-22 | 科大讯飞股份有限公司 | Control system and method for controlling household appliances through voice |
| CN105931633A (en) * | 2016-05-30 | 2016-09-07 | 深圳市鼎盛智能科技有限公司 | Speech recognition method and system |
| CN106679326A (en) * | 2017-01-25 | 2017-05-17 | 北京通远科技有限公司 | Intelligent refrigerator controlled on basis of voice recognition |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019134473A1 (en) * | 2018-01-08 | 2019-07-11 | 珠海格力电器股份有限公司 | Speech recognition system, method and apparatus |
| CN108712704A (en) * | 2018-06-13 | 2018-10-26 | 腾讯科技（深圳）有限公司 | Speaker, the playback method of audio data, device, storage medium and electronic device |
| CN108712704B (en) * | 2018-06-13 | 2021-08-06 | 腾讯科技（深圳）有限公司 | Sound box, audio data playing method and device, storage medium and electronic device |
| CN110837234A (en) * | 2018-08-17 | 2020-02-25 | 阿里巴巴集团控股有限公司 | Intelligent voice control panel and panel switch socket |
| CN110868648A (en) * | 2018-08-27 | 2020-03-06 | 杭州海康威视数字技术股份有限公司 | Intelligent voice realization method of indoor intercom device and indoor intercom device |
| CN109119071A (en) * | 2018-09-26 | 2019-01-01 | 珠海格力电器股份有限公司 | A kind of training method and device of speech recognition modeling |
| CN110986293A (en) * | 2019-12-12 | 2020-04-10 | 珠海格力电器股份有限公司 | Voice board assembly and air conditioner |
| CN111128194A (en) * | 2019-12-31 | 2020-05-08 | 云知声智能科技股份有限公司 | System and method for improving online voice recognition effect |
| CN111182412A (en) * | 2019-12-31 | 2020-05-19 | 联想(北京)有限公司 | Electronic equipment, data processing method for electronic equipment and conference system equipment |
| CN111182412B (en) * | 2019-12-31 | 2021-04-13 | 联想(北京)有限公司 | Electronic equipment, data processing method for electronic equipment and conference system equipment |
| CN112731831A (en) * | 2020-12-18 | 2021-04-30 | 宁波向往智能科技有限公司 | Intelligent switch panel |
| CN113819585A (en) * | 2021-09-16 | 2021-12-21 | 青岛海尔空调器有限总公司 | Microphone device, method and device for matching voice air conditioner microphone and air conditioner |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019134473A1 (en) | 2019-07-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108039174A (en) | Speech recognition system, method and apparatus | |
| CN108320753A (en) | Control method, the device and system of electrical equipment | |
| CN107135443A (en) | A kind of signal processing method and electronic equipment | |
| CN104216351B (en) | Household electrical appliance sound control method and system | |
| CN107682240A (en) | A kind of distributed sound interactive system for intelligent domestic | |
| CN110336723A (en) | Control method and device, the intelligent appliance equipment of intelligent appliance | |
| CN107388487B (en) | The method and apparatus for controlling air-conditioning | |
| CN106921911B (en) | Voice acquisition method and device | |
| CN109493849A (en) | Voice awakening method, device and electronic equipment | |
| CN206696909U (en) | A kind of classroom based on Application on Voiceprint Recognition takes attendance in class system | |
| CN105276751B (en) | Speech playing method and system | |
| CN109741747B (en) | Voice scene recognition method and device, voice control method and device and air conditioner | |
| CN108592349A (en) | A kind of air-conditioner control system | |
| CN106572228A (en) | Volume adjusting method, volume adjusting device and mobile terminal | |
| CN111798852A (en) | Voice wake-up recognition performance test method, device and system and terminal equipment | |
| CN109737521A (en) | Air purifier with voice control function | |
| CN110767225B (en) | Voice interaction method, device and system | |
| CN109524013A (en) | A kind of method of speech processing, device, medium and smart machine | |
| CN109360564A (en) | The selection method and device of language identification mode, household electrical appliance | |
| CN105182763A (en) | Intelligent remote controller based on voice recognition and realization method thereof | |
| CN108922528A (en) | Method and apparatus for handling voice | |
| CN113470634A (en) | Control method of voice interaction equipment, server and voice interaction equipment | |
| CN109582275A (en) | Voice regulation method, device, storage medium and electronic device | |
| CN108882103A (en) | Intelligent sound box, sound collection equipment and intelligent sound box system | |
| CN103645690A (en) | Method for controlling digital home smart box by using voices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180515 |