CN109688036A - Control method and device of intelligent household appliance, intelligent household appliance and storage medium - Google Patents
Control method and device of intelligent household appliance, intelligent household appliance and storage medium Download PDFInfo
- Publication number
- CN109688036A CN109688036A CN201910126388.2A CN201910126388A CN109688036A CN 109688036 A CN109688036 A CN 109688036A CN 201910126388 A CN201910126388 A CN 201910126388A CN 109688036 A CN109688036 A CN 109688036A
- Authority
- CN
- China
- Prior art keywords
- intelligent appliance
- speech recognition
- intelligent
- voice
- cpu
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 80
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 69
- 230000008569 process Effects 0.000 claims abstract description 38
- 238000013138 pruning Methods 0.000 claims description 37
- 230000015654 memory Effects 0.000 claims description 23
- 238000012549 training Methods 0.000 claims description 23
- 238000004088 simulation Methods 0.000 claims description 14
- 230000005611 electricity Effects 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 24
- 230000004044 response Effects 0.000 abstract description 7
- 230000006870 function Effects 0.000 description 13
- 238000010586 diagram Methods 0.000 description 8
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 8
- 238000012545 processing Methods 0.000 description 5
- 238000013507 mapping Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 239000000779 smoke Substances 0.000 description 3
- 238000004378 air conditioning Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010411 cooking Methods 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 208000019300 CLIPPERS Diseases 0.000 description 1
- 241001672694 Citrus reticulata Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000010009 beating Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 208000021930 chronic lymphocytic inflammation with pontine perivascular enhancement responsive to steroids Diseases 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010297 mechanical methods and process Methods 0.000 description 1
- 230000005226 mechanical processes and functions Effects 0.000 description 1
- 238000011022 operating instruction Methods 0.000 description 1
- 238000005057 refrigeration Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
- H04L12/2803—Home automation networks
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; 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 TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
- H04L12/2803—Home automation networks
- H04L12/2816—Controlling appliance services of a home automation network by calling their functionalities
- H04L12/282—Controlling appliance services of a home automation network by calling their functionalities based on user interaction within the home
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Automation & Control Theory (AREA)
- Artificial Intelligence (AREA)
- Selective Calling Equipment (AREA)
Abstract
The invention discloses a control method and device of an intelligent household appliance, the intelligent household appliance and a storage medium. The method comprises the following steps: receiving voice information; acquiring voice command words, wherein the voice command words are obtained by inputting voice information and a voice recognition algorithm into a voice recognition model, and the voice recognition model is determined according to the working scene of the intelligent household appliance; and controlling the intelligent household appliance according to the voice command words. When the intelligent household appliance is in a working scene with small calculation amount, the voice recognition model with small calculation amount is used, so that the hardware cost, the CPU calculation resource and the CPU power consumption of the intelligent household appliance are reduced; when the intelligent household appliance is in a working scene with a large calculation amount, the voice recognition model with the large calculation amount is used in a self-adaptive mode, so that CPU calculation resources are reasonably utilized, and the system response speed and the overall power consumption of the working process of the intelligent household appliance are guaranteed to be good.
Description
Technical field
The present embodiments relate to intelligent appliance technology more particularly to a kind of control methods of intelligent appliance, device, intelligence
Household electrical appliances and storage medium.
Background technique
With the fast development of science and technology, more and more intelligent appliance increases have speech identifying function.In order to guarantee voice
Discrimination generally requires the speech recognition algorithm high using discrimination, correspondingly, discrimination it is high speech recognition algorithm it is corresponding
The complexity of speech recognition modeling is also relatively high, to consume biggish CPU operation money when running speech recognition algorithm
Source, and then the system of intelligent appliance is caused the high problem of low-response, power consumption occur.
Summary of the invention
The present invention provides control method, device, intelligent appliance and the storage medium of a kind of intelligent appliance, to solve intelligent family
System low-response, power consumption high problem when electricity voice controls.
In a first aspect, the embodiment of the invention provides a kind of control methods of intelligent appliance, comprising:
Receive voice messaging;
Voice command words are obtained, the voice command words are that the voice messaging and speech recognition modeling input speech recognition
Algorithm obtains, and the speech recognition modeling is determined according to the operative scenario of intelligent appliance;
The intelligent appliance is controlled according to the voice command words.
Further, the search precision of the speech recognition algorithm is determined according to the working performance of the intelligent appliance.
Further, the working performance of the intelligent appliance includes at least the CPU usage and CPU function of the intelligent appliance
One of consumption.
Further, the search precision of the speech recognition algorithm is determined according to the working performance of the intelligent appliance and is wrapped
It includes:
Obtain the current process runing time of the intelligent appliance;
The beta pruning coefficient of the speech recognition algorithm is calculated, the beta pruning coefficient is used according to the CPU of the intelligent appliance
Rate determines that the CPU usage is the ratio that current process runing time accounts for CPU total run time, and search precision is cut with described
Branch coefficient is inversely proportional.
Further, before the acquisition voice command words, further includes:
Training speech recognition modeling, the training input parameter of the speech recognition modeling include the simulation of the intelligent appliance
Operative scenario information and analog voice order word.
Further, the speech recognition modeling includes acoustic model and language model.
Further, the operative scenario of the intelligent appliance is transported according to the load operating mode of the intelligent appliance, backstage
At least one line program and current time determine.
Second aspect, the embodiment of the invention also provides a kind of control devices of intelligent appliance, comprising:
Receiving module, for receiving voice messaging;
Module is obtained, for obtaining voice command words, the voice command words are the voice messaging and speech recognition mould
Type input speech recognition algorithm obtains, and the speech recognition modeling is determined according to the operative scenario of intelligent appliance;
Control module, for controlling the intelligent appliance according to the voice command words.
Further, the control device of the intelligent appliance, further includes:
Determining module, the search precision for the speech recognition algorithm are true according to the working performance of the intelligent appliance
It is fixed.
Further, the working performance of the intelligent appliance includes at least the CPU usage and CPU function of the intelligent appliance
One of consumption.
Further, the determining module includes:
Acquiring unit, for obtaining the current process runing time of the intelligent appliance;
Computing unit, for calculating the beta pruning coefficient of the speech recognition algorithm, the beta pruning coefficient is according to the intelligence
The CPU usage of household electrical appliances determines that the CPU usage is the ratio that current process runing time accounts for CPU total run time, search
Precision is inversely proportional with the beta pruning coefficient.
Further, the control device of the intelligent appliance further include:
Training module, for training speech recognition modeling, the speech recognition mould before the acquisition voice command words
The training input parameter of type includes the simulation operative scenario information and analog voice order word of the intelligent appliance.
Further, the speech recognition modeling includes acoustic model and language model.
Further, the operative scenario of the intelligent appliance is transported according to the load operating mode of the intelligent appliance, backstage
At least one line program and current time determine.
The third aspect, the embodiment of the invention also provides a kind of intelligent appliances, comprising: display screen, memory and one
Or multiple processors;
The display screen, for showing the status information of intelligent appliance;
The memory, for storing one or more programs;
When one or more of programs are executed by one or more of processors, so that one or more of processing
Device realizes the control method of intelligent appliance as described in relation to the first aspect.
Fourth aspect, the embodiment of the invention also provides a kind of computer readable storage mediums, are stored thereon with computer
Program realizes the control method of intelligent appliance as described in relation to the first aspect when the program is executed by processor.
The present invention determines suitable language according to the operative scenario of intelligent appliance by receiving the voice messaging of user's input
Sound identification model, then by voice messaging and speech recognition modeling input speech recognition algorithm to obtain voice command words, and root
Intelligent appliance is controlled according to voice command words.It is small using operand when intelligent appliance is in the lesser operative scenario of operand
Speech recognition modeling reduces hardware cost, CPU calculation resources and the CPU power consumption of intelligent appliance;Operand is in intelligent appliance
When biggish operative scenario, the speech recognition modeling for adaptively using operand big, to be carried out rationally to CPU calculation resources
It utilizes, guarantees that the system response time of the course of work of intelligent appliance is good and overall power is low.
Detailed description of the invention
Fig. 1 is a kind of flow chart of the control method for intelligent appliance that the embodiment of the present invention one provides;
Fig. 2 is a kind of flow chart of the control method of intelligent appliance provided by Embodiment 2 of the present invention;
Fig. 3 is a kind of flow chart of the control method for intelligent appliance that the embodiment of the present invention three provides;
Fig. 4 is a kind of structural schematic diagram for multiple intelligent appliances that the embodiment of the present invention three provides;
Fig. 5 is a kind of display schematic diagram for TV programme pause that the embodiment of the present invention three provides;
Fig. 6 is a kind of structural schematic diagram of the control device for intelligent appliance that the embodiment of the present invention four provides;
Fig. 7 is a kind of structural schematic diagram for intelligent appliance that the embodiment of the present invention five provides.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining the present invention rather than limiting the invention.It also should be noted that in order to just
Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
Embodiment one
Fig. 1 is a kind of flow chart of the control method for intelligent appliance that the embodiment of the present invention one provides, and is provided in embodiment
The control method of intelligent appliance can be executed by intelligent appliance, which includes two or more physical entities,
It or is a physical entity.For example, intelligent appliance includes but is not limited to intelligent range hood, intelligent air condition, smart television and intelligence
Energy water heater etc..
As shown in Figure 1, the control method of the intelligent appliance specifically comprises the following steps:
S110, voice messaging is received.
Wherein, voice messaging refers to the voice operating instruction that user is issued to intelligent appliance.The languages of voice messaging without
It needs to limit, for example, mandarin, English or various types of dialects;The sentence sequence of voice messaging, statement type etc. without
It limits, for example, voice messaging can be " tuning up air quantity " etc. for " closing blower " or voice messaging.
In embodiment, user issues voice messaging to intelligent appliance, to wake up the intelligent appliance.Such as: user is cooking
When preparing food, the current air volume of intelligent range hood is unable to satisfy air quantity requirement, then user can say to intelligent range hood and such as " tune up wind
The voice messaging of amount " or the like, the air quantity of intelligent range hood is tuned up by the voice messaging.
S120, voice command words are obtained.
Wherein, voice command words are that voice messaging and speech recognition modeling input speech recognition algorithm and obtain, which knows
Other model is determined according to the operative scenario of intelligent appliance.
The voice messaging of user's input is generally colloquial command information, voice messaging need to be passed through speech recognition modeling
It identifies acoustic feature, to identify the voice messaging of user's input, and voice messaging and speech recognition modeling is input in advance
Determining speech recognition algorithm, to obtain the corresponding standard commands word of voice messaging, i.e. voice command words.Illustratively, it is assumed that
The voice messaging of user's input is " turning the air quantity of kitchen ventilator to one grade down ", the then voice command words that intelligent range hood receives
For " air quantity is turned down to one grade ".
Specifically, after the voice messaging that intelligent appliance receives user's input, intelligent appliance can extract voice letter
Acoustic feature in breath, and acoustic feature is input in predetermined speech recognition modeling, known by speech recognition modeling
Other acoustic feature to identify the voice messaging of user's input, and voice messaging and speech recognition modeling is input in advance really
In fixed speech recognition algorithm, to obtain corresponding voice command words.
Wherein, speech recognition modeling includes acoustic model and language model.Wherein, acoustic model be to acoustics, linguistics,
The representation of knowledge of the difference of variable, speaker's gender, the accent of environment etc., and language model is known one group of word Sequence composition
Knowing indicates.In general, acoustic model lstm+ctc training, obtains the mapping of acoustic feature to phoneme;Language model is used
The training that SRILM tool is LM obtains 3-gram and4-gram, is the mapping of word Yu word, word and sentence.For speech recognition
System, the output valve of acoustic model is exactly total each calculated acoustic feature of frame, to each phoneme in voice messaging
Mapping;And speech model is the mapping to each word, sentence in voice messaging.
Wherein, the operative scenario of intelligent appliance, it can be understood as the currently used scene of intelligent appliance.Specifically, when
Intelligent appliance be different terminal devices when comprising operative scenario be also distinguishing.For example, when intelligent appliance is intelligence
When kitchen ventilator, operative scenario may include standby scene, culinary art scene and audio and video playing scene etc.;When intelligent appliance is intelligence
When air-conditioning, operative scenario may include standby scene, timing working scene, refrigeration scene, heating scene etc.;Work as intelligent appliance
When for smart television, operative scenario may include standby scene, audio and video playing scene etc., not to each intelligence in embodiment
The above-mentioned operative scenario of energy household electrical appliances is defined.Wherein, audio and video playing scene can be individual music scene, can also
The scene used simultaneously for audio and video.
In embodiment, load operating mode of the operative scenario of intelligent appliance according to intelligent appliance, running background program
It is determining at least one current time.Wherein, load operating mode can be understood as locating for each load in intelligent appliance
Operating mode, for example, load operating mode is active operation mode and standby mode;Running background program, it can be understood as intelligence
The relative program for the communication being currently running inside energy household electrical appliances;Current time is current Beijing time.Under normal circumstances, current
Time is between 5. -7 point of -8 point of 6:00 AM or 11. -12 point of noon or afternoon, it is believed that is that intelligent range hood is in
Cook scene.Illustratively, it is assumed that intelligent appliance is intelligent range hood, then the load of intelligent range hood may include blower, illumination
Lamp, display screen etc., then blower is in active operation mode, and current time is 7 points of morning, can determine that the work of intelligent range hood
Make scene as culinary art scene.And the display screen for working as intelligent range hood is in active operation mode, and running background program is in logical
Interrogate the interaction stage, while blower and headlamp are in close state, and can only loudspeaker in kitchen ventilator it is in the open state, then
The operative scenario for determining intelligent range hood is audio and video playing scene.Certainly, it when intelligent appliance is different terminal device, needs
Operative scenario is determined using different foundation conditions, in embodiment only by taking intelligent range hood as an example to operative scenario
Determination be illustrated, but be defined not to this.
It should be noted that being provided with initialized data base in intelligent appliance, it is stored in initialized data base pre-
Different phonetic identification model obtained from first training.In embodiment, it when intelligent appliance is in different operating scene, corresponds to
Speech recognition modeling be also different.Wherein, the corresponding specified operative scenario of each speech recognition modeling.It is exemplary
Ground, it is assumed that intelligent appliance is intelligent range hood, then the operative scenario of intelligent range hood may include standby scene, culinary art scene and sound
Video playing scene etc..Certainly, under different operating scene, high probability uses in corresponding speech recognition modeling order word
The order word list that list and low probability use is entirely different.For example, when the operative scenario of intelligent range hood is culinary art field
Jing Shi, then the corresponding speech recognition modeling of scene is cooked in selection from initialized data base, under the culinary art scene, due to needing to beat
Machine blow in handle oil smoke, alternatively, opening headlamp to illuminate environment locating for intelligent range hood, then high probability uses
Order word list in may include blower, the relevant order word of Lighting control, and can be wrapped in the order word list that low probability uses
It includes and switches to the relevant order words of control such as next video, pause broadcasting.
S130, intelligent appliance is controlled according to voice command words.
Specifically, after intelligent appliance gets voice command words, corresponding order is executed according to voice command words
Operation, so that the operative scenario to intelligent appliance is adaptively adjusted, to meet the use demand of user.For example, working as intelligence
Energy kitchen ventilator is in culinary art scene, i.e. blower is in the open state, at this point, the voice of user's capable of emitting " organ of keeping watch closes " etc
Information is then calculated by the way that voice messaging speech recognition modeling corresponding with culinary art scene is input to predetermined speech recognition
Method receives the voice command of " blower closing " in intelligent range hood to get the voice command words of " blower closing " etc
After word, the operating mode of blower is switched into closing by opening according to the voice command words, to realize to intelligent range hood
Voice control.It is described herein to be, the process that " blower closing " is a speech-to-text is converted by " organ of keeping watch closes ", i.e.,
The voice messaging that user inputs is converted to the process of corresponding voice command words.
The technical solution of the present embodiment, by receiving the voice messaging of user's input, and according to the yard of intelligent appliance
Scape determines suitable speech recognition modeling, then by voice messaging and speech recognition modeling input speech recognition algorithm to obtain language
Sound order word, and intelligent appliance is controlled according to voice command words.When intelligent appliance is in the lesser operative scenario of operand, make
With the small speech recognition modeling of operand, hardware cost, CPU calculation resources and the CPU power consumption of intelligent appliance are reduced;In intelligent family
When electric operative scenario biggish in operand, the speech recognition modeling for adaptively using operand big, thus to CPU operation
Resource is rationally utilized, and guarantees that the system response time of the course of work of intelligent appliance is good and overall power is low.
Embodiment two
Fig. 2 is a kind of flow chart of the control method of intelligent appliance provided by Embodiment 2 of the present invention.The present embodiment be
On the basis of above-described embodiment, the control method of intelligent appliance is embodied.As shown in Fig. 2, the intelligent family of the present embodiment
The control method of electricity specifically comprises the following steps:
S210, training speech recognition modeling.
It should be noted that user in input voice information, may be subjected to the influence of ambient enviroment and make
With certain accent when having noise or user's input voice information in the voice messaging that intelligent appliance receives, and or
There is the case where pronouncing indistinctly Chu etc. in the voice messaging that person user is issued, these factors may will affect speech recognition modeling
Recognition result, and then influence acquisition of the intelligent appliance to voice command words.Therefore, in order to improve the identification of speech recognition modeling
Rate can be trained speech recognition modeling before receiving voice messaging.
Wherein, the training input parameter of speech recognition modeling includes the simulation operative scenario information and simulation language of intelligent appliance
Sound order word.In embodiment, simulation operative scenario information is to be in the letter that different operating scene is simulated to intelligent appliance
Breath.Specifically, when intelligent appliance is in different operating scene, locating for ambient enviroment be also different, for example, when to intelligence
Can kitchen ventilator be in culinary art scene when being simulated, since blower is in the open state, then surrounding's ring locating for intelligent range hood
There are noises in border, when in order to make intelligent range hood get the voice messaging of user's input, more can accurately identify that voice is believed
Breath, when being trained to acoustic model, need to consider the noise of fan operation, analog voice order word is carried out plus made an uproar is trained
Processing.
In embodiment, analog voice order word is to be in the voice that score is different under different operating scene to intelligent appliance
Order word, it can be understood as the voice command words under simulation operative scenario.In order to improve the phonetic recognization rate of intelligent appliance,
It, can be according to different simulation operative scenarios to simulation when being trained to simulation operative scenario information and analog voice order word
Voice command words are set.Specifically, the different operating scene according to locating for intelligent appliance, it is right under the operative scenario to determine
The voice command word list that the voice command word list and low probability that the high probability answered uses use, then according to selected voice
Order word list is trained simulation operative scenario information and analog voice order word.Wherein, each voice command word list
A corresponding simulation operative scenario, while including several analog voice order words in voice command word list.
It is in different operative scenarios in intelligent appliance, it, can be for not when being trained to acoustic model and language model
With the voice that uses of analog voice order word the voice command word list and low probability that are trained, while using in high probability
Voice command words appeared in order word list are also different.
By taking the culinary art scene of intelligent range hood as an example, training speech recognition modeling is specifically described.For culinary art field
Scape is then used in high probability since the score that blower, headlamp etc. stress the relevant voice command words of air-blower control is relatively high
It may include closing blower, opening the sides such as blower, air quantity tunes up, air quantity is turned down, one/bis-/third gear air quantity in voice command word list
The relevant analog voice order word of weight air-blower control, and may include volume tune in the voice command word list that low probability uses
Greatly, next song, a upper first song, pause, play etc. and to stress the relevant analog voice order word of music control.Meanwhile to sound
When learning model progress simulated training, the yard in simulation operative scenario information and actual mechanical process in order to make simulated training
Scape is more nearly, and need to consider the noise of fan operation, and need to carry out while inputting analog voice order word plus make an uproar trained place
Reason.Certainly, when being trained to language model, it need to be promoted in language model and stress the relevant analog voice life of air-blower control
The scoring for enabling word reduces the score that other analog voice order words correspond to language model, and then reduces searching for speech recognition algorithm
Rope time and CPU calculation resources.
Specifically, it is assumed that the range of analog voice order word score is at 0-100 points, wherein 0 indicates the voice obtained life
Word and the analog voice order word of the intelligent appliance is enabled to mismatch completely, 100 points indicate the voice command words obtained and the intelligence
The analog voice order words of household electrical appliances exactly matches, wherein can be by the analog voice in voice command word list that high probability uses
The preset threshold of order word is set as 80 points, is denoted as the first preset threshold, and the mould in the voice command word list that low probability uses
The preset threshold of quasi- voice command words is set as 60 points, is denoted as the second preset threshold.Illustratively, when intelligent range hood is in culinary art
When scene, then by " closing blower ", " opening blower ", " air quantity tunes up ", " air quantity is turned down " and " one/bis-/third gear air quantity " etc.
The scoring for stressing the relevant analog voice order word of air-blower control is arranged at 80 points or more, and incite somebody to action " volume tunes up, next is sung,
A upper first song, plays at pause " etc. stress music and control the scoring of relevant analog voice order word to be arranged at 60 points hereinafter, i.e.
It reduces and stresses music and control relevant analog voice order word to correspond to the score of language model, to reduce speech recognition modeling
Search time, thereby reduce the waste of CPU calculation resources.
It should be noted that can pass through when carrying out simulated training to acoustic model for different operative scenarios
The score of acoustic model is configured, and also may be embodied in the setting of searching route.Illustratively, it is cooked when intelligent range hood is in
When scene of preparing food, specially treated can be carried out to the analog voice order word under culinary art scene.For example, by " component tunes up ", " air quantity
Fall big " etc. be similar to the analog voice order word of " air quantity tunes up ", also correspond to " air quantity tunes up " this voice command words, thus
Increase the phonetic recognization rate of the voice command words.Meanwhile it being not required to stress " upper one first song " " switching to next video " etc.
Music controls relevant analog voice order word and carries out specially treated, to be effectively utilized CPU calculation resources.
Certainly, when being trained to speech recognition modeling, in order to guarantee to simulate the standard of phonetic recognization rate in working environment
True property, only one intelligent appliance under the same environment.But in practical application scene, not only one intelligent appliance.
For example, being generally fitted with intelligent range hood, intelligent air condition, smart television and intelligent water heater in the family of some user
Equal terminal devices, at this point, just will appear the case where same voice command words are likely to be suited for multiple terminal devices, to can make
At the false triggering rate of intelligent appliance.
In order to guarantee the phonetic recognization rate of intelligent appliance and increase diversity, multiple analog voice order word difference can be chosen
The multiple speech recognition modelings of training, and be independent from each other between each speech recognition modeling.Meanwhile in order to reduce intelligent appliance
False triggering rate can carry out corresponding setting to analog voice order word when being trained to speech recognition modeling.For example,
Intelligent range hood and smart television are under audio and video playing scene, and all there are sounds in the order word list that high probability uses
Amount, program, next/music such as a upper head control relevant order word, then audio and video playing field is being in intelligent range hood
When speech recognition modeling under scape is trained, analog voice order word can be set in " volume of kitchen ventilator is turned down " " oil smoke
Machine-cut shifts to next " or the like word;Correspondingly, speech recognition modeling of the smart television under audio and video playing scene into
When row training, analog voice order word can be set to the classes such as " volume of TV is turned down ", " TV switches to next channel "
As word, i.e., the title and operational order of intelligent appliance are provided in analog voice order word.
Certainly, title and the behaviour of intelligent appliance are only illustratively provided in analog voice order word in embodiment
It orders, a kind of intelligent appliance can also be represented using specified voice command words, using different intelligent appliances as differentiation,
For example, intelligent water heater is named as " small love ", thus when intelligent water heater is trained, it need to be in analog voice order word
It is provided with small love and corresponding operation instruction.
S220, voice messaging is received.
S230, voice command words are obtained.
Wherein, voice command words are that voice messaging and speech recognition modeling input speech recognition algorithm and obtain, speech recognition
Model is determined according to the operative scenario of intelligent appliance.
Wherein, the search precision of speech recognition algorithm is determined according to the working performance of intelligent appliance.In embodiment, intelligence
The working performance of household electrical appliances includes at least one of CPU usage and CPU power consumption of intelligent appliance.
Wherein, CPU usage is the CPU calculation resources that current operation process occupies, and indicates the intelligent appliance at some
Between put operation process the case where.In general, CPU usage is the feelings that current all operation processes occupy CPU calculation resources
The case where condition, some not single operation process occupies CPU calculation resources.Specifically, CPU usage is higher, illustrate the intelligence
The process that energy household electrical appliances are run on this time point is more, otherwise fewer.Certainly, the height of CPU usage with currently run into
It is not stringent proportional relationship between the quantity of journey, for example, process is in operating status there are two current, but each process
CPU consumption be all 30%;It is of course also possible to currently have 10 processes in operating status, but the CPU consumption of each process
Only 1%.Specifically, when CPU usage increases, then it represents that speech recognition algorithm consumes biggish CPU calculation resources, this
When answer the search precision of appropriate adjustment speech recognition algorithm, to reduce the utilization rate of CPU, to make the Caton of intelligent appliance, hair
Phenomena such as hot serious, is improved;Conversely, when CPU usage reduces, then it represents that speech recognition algorithm consumes lesser CPU fortune
Resource is calculated, the search precision of speech recognition algorithm can be properly increased at this time.
Correspondingly, CPU power consumption is the current value for flowing through the processor core of intelligent appliance and the core electricity on the processor
The product of pressure value, and the Rule of judgment to intelligent appliance cooling system.In general, if CPU power consumption is low, illustrate that CPU is sent out
Heat is small, illustrates that CPU usage is also relatively low indirectly;Conversely, illustrating that CPU calorific value is big, indirectly if CPU power consumption is high
Illustrate that CPU usage is also relatively high.Under normal conditions, CPU power consumption rises, and is because the work of CPU itself is frequently in intelligent appliance
Rate is improved, for example, being promoted the working frequency of CPU in intelligent appliance itself from 600MHz to 1GHz, then CPU power consumption
Also it rises with it.Meanwhile working frequency is directly proportional to CPU usage, it can be understood as, since CPU usage is higher, lead to intelligence
The current working frequency of energy household electrical appliances has been unable to satisfy demand, then needs the working frequency for improving intelligent appliance.
Specifically, the search precision of speech recognition algorithm according to the working performance of intelligent appliance determine, specifically include as
Lower step S10-S20:
S10, the current process runing time for obtaining intelligent appliance.
Wherein, current process runing time is the time of CPU occupied by current process in intelligent appliance.In general, exist
Time-interval control is all that time-slotting is used to CPU, it can be understood as when A process is occupying CPU, B into
Journey and vacant CPU.In embodiment, when intelligent appliance is in the open state, each load, backstage in intelligent appliance are transported
Line program and front stage operation program have corresponding process, can be according to the calculating side of current process runing time in the prior art
Formula calculates the runing time of current process in intelligent appliance and is obtained.
S20, the beta pruning coefficient for calculating speech recognition algorithm.
Wherein, beta pruning coefficient is determined according to the CPU usage of intelligent appliance, and CPU usage is current process runing time
The ratio of CPU total run time is accounted for, search precision is inversely proportional with beta pruning coefficient.
In embodiment, CPU usage is the ratio that current process runing time accounts for CPU total run time.Illustratively,
Assuming that operation has A process and B process in intelligent appliance, wherein A process occupies 10ms, and B process occupies 30ms, idle 60ms;So
It is that A process occupies 10ms again afterwards, B process occupies 30ms, idle 60ms, if being not always the case in a period of time, in this period
CPU usage be 40%.Wherein, CPU usage is CPU usage.
Wherein, beta pruning coefficient is the ginseng for carrying out beta pruning to decision hedge clipper branch algorithm in speech recognition algorithm and using
Number.Specifically, speech recognition algorithm search precision on the one hand is determined by the working performance of intelligent appliance, another aspect also with
The complexity of speech recognition algorithm is related.When beta pruning coefficient is higher, show that the complexity of speech recognition algorithm is lower, conversely,
When beta pruning coefficient is lower, show that the complexity of speech recognition algorithm is higher.Wherein, beta pruning coefficient is according to the CPU of intelligent appliance
Utilization rate determines that specifically, the relationship between beta pruning coefficient and CPU usage is known with the currently employed voice of intelligent appliance
Other algorithm is related with speech recognition modeling, and depending on the increase with beta pruning coefficient, the operation of speech recognition algorithm is by which kind of pass
System is declined.In embodiment, beta pruning coefficient is higher, then when carrying out optional path searching in speech recognition algorithm, more
It is easy to crop the not high enough individual path of score, to make CPU operand reduce, but search precision also declines therewith.Example
Property, the relational expression between beta pruning coefficient and CPU usage is b=A*C+B, wherein b is beta pruning coefficient, and C is that intelligent appliance is worked as
Preceding CPU usage, A and B are empirical coefficients.Wherein, A and B is to be tested to obtain by many experiments in development phase
's.As can be seen that beta pruning coefficient and CPU usage be it is in a linear relationship, when beta pruning coefficient b increase when, to empirical coefficient A and B
It is adjusted, so that CPU usage C declines therewith.
It is described herein to be, when being scanned in speech recognition algorithm to some voice command words, has several and search
It rope path need to will be general in speech recognition algorithm in order to reduce CPU calculation resources of the intelligent appliance when running speech recognition algorithm
The not high enough individual path of rate crops.Specifically, there are several layers in the decision tree of speech recognition algorithm, have in each layer
Several leaf nodes need to be according to preset beta pruning coefficient to branch path then during cutting to individual path
Diameter is cut, and the individual path by probability lower than beta pruning coefficient crops, to reduce the use of CPU calculation resources.It is exemplary
Ground, it is assumed that beta pruning coefficient is 0.9, is respectively A, B and C in speech recognition algorithm three individual paths therein, wherein A, B and C
The probability in path is respectively 90%, 85% and 80%, then when cutting to individual path, can cut B path and the path C
Fall, to identify the best searching route in speech recognition algorithm, and then reduces search time.
It is worth noting that, the operation relation of beta pruning coefficient and CPU usage is only illustratively listed in embodiment,
And beta pruning coefficient and CPU usage are in a linear relationship, but and be not limited thereof, in specific practical operation, beta pruning system
Several and CPU usage can also be in non-linear relation.
S240, intelligent appliance is controlled according to voice command words.
The technical solution of the present embodiment, on the basis of the above embodiments, the current process by obtaining intelligent appliance are transported
The row time to obtain CPU usage, and determines according to CPU usage the beta pruning coefficient of speech recognition algorithm, to crop voice
The not high enough individual path of probability in recognizer, thus the case where realizing according to current CPU calculation resources, dynamic adjusts voice
The search precision of recognizer, so that CPU calculation resources are reasonably utilized, and then the CPU for improving intelligent appliance is utilized
Rate.
Embodiment three
Fig. 3 is a kind of flow chart of the control method for intelligent appliance that the embodiment of the present invention three provides.The present embodiment be
On the basis of above-described embodiment, as a preferred embodiment, the control method of intelligent appliance is specifically described.The control
Method specifically comprises the following steps:
S310, the usage scenario for judging current device.
Wherein, equipment is a kind of terminal device, for example, the intelligent appliances such as intelligent range hood, intelligent air condition, smart television.
In embodiment, usage scenario is the operative scenario of current intelligent appliance.Specifically, in the operative scenario to current intelligent appliance
When being determined, the various information of intelligent appliance need to be first acquired, for example, the foreground and backstage that can acquire current intelligent appliance are
The application program of operation, the information such as operating condition and current time of each load in current intelligent appliance, and according to above-mentioned
Information determines operative scenario.It illustratively, can be to blower, the headlamp in intelligent appliance when intelligent appliance is intelligent range hood
The operating condition of equal loads is acquired, when blower is in the operating mode in being currently running, it is possible to determine that the intelligence oil smoke
Machine is in culinary art scene.Or when current time is 6 pm half, it also can determine that the intelligent range hood is in culinary art field
Scape.
Certainly, in order to guarantee determine operative scenario accuracy, multiple information of intelligent appliance can be acquired, for example, according to
The operating condition of load, current time and front stage operation program, are determined the operative scenario of intelligent appliance.It is exemplary
The headlamp on ground, intelligent range hood is in the open state, and current time is 5 thirty of afternoon, but blower is in close state, and
Video playing interface is shown on display screen, while loudspeaker also issues audio-frequency information, then can determine that intelligent range hood is in sound
Video playing scene.
S320, inquiry database, select suitable algorithm model.
Wherein, it is stored with speech recognition modeling trained in advance in the database.In embodiment, algorithm model is to transport
Used speech recognition modeling when row speech recognition algorithm specifically, can root after determining the operative scenario of intelligent appliance
According to the relevant information of current intelligent appliance operative scenario, preset database is inquired, selects suitable acoustic model and voice mould
Type.Wherein, preset database is confirmed in equipment development phase, first according to the type of equipment to operative scenario into
Row divides, for example, operative scenario can be divided into standby scene, culinary art scene and audio and video playing when equipment is intelligent range hood
Scene etc..According to each different operative scenario, order word list that the high probability under the operative scenario uses and low general is determined
The order word list that rate uses.The training of corresponding speech recognition modeling is carried out for selected order word list.For example, being directed to
Scene, the relevant order word of acoustic model emphasis training air-blower control are cooked, while considering the noise of fan operation, carries out plus makes an uproar
Trained processing, promotes the scoring of air-blower control related command word in language model, while reducing air-blower control related voice
The related rejection path of order word, to reduce CPU operand.Wherein, rejection path, it can be understood as one kind is in language model
Prevent nearly sound word from leading to the path of the misrecognition to voice command words and increased " refusal identification ".Illustratively, for " beating
Blow in machine " this voice command words, then the settable one rejection path for being similar to " open fan ", to prevent user from issuing
When the voice messaging of " opening fan ", intelligent appliance is identified as " opening blower " and leads to the maloperation to intelligent appliance.
Especially in the case where cooking scene, since user need to carry out control operation to blower, then under the scene, " opening blower " this language
Sound order word uses probability relatively high, if reducing the corresponding related rejection path of air-blower control related voice order word,
So that the misoperation probability of intelligent appliance declines, so that reducing speech recognition algorithm is carrying out the corresponding voice life of voice messaging
Searching route when word being enabled to search for, and then search time is shortened, save CPU calculation resources.
Illustratively, configured with the intelligent range hood of speech identifying function, there is range hood fan control and music control
The preset analog voice order word of two classes is made, for example, the analog voice order word of range hood fan control includes: " to close wind
Machine ", " air quantity tunes up ", " third gear air quantity " etc.;The analog voice order word of music control includes: " volume tunes up ", " next
Song " etc..When monitoring that user opens intelligent range hood, speech recognition algorithm selection is done using for air-blower control order
Training, while the trained acoustic model that carried out plus made an uproar, and selection stress the language model of air-blower control related command word;And
When monitoring that user opens music player interface, then speech recognition algorithm selection done for music control command stress it is excellent
The acoustic model and language model of change.
S330, current CPU usage is calculated.
Specifically, the CPU usage current to intelligent appliance calculates, in order to calculate cutting for speech recognition algorithm
Branch coefficient.Wherein, it to there are many calculation methods of CPU usage, is always run for example, carrying out runing time according to the free time and accounting for CPU
The ratio of time calculates CPU usage, the ratio of CPU total run time can also be accounted for according to current process runing time to calculate
CPU usage.
S340, suitable beta pruning coefficient is selected according to current CPU usage.
In embodiment, it according to the CPU usage of current intelligent appliance, determines and searches for best point in speech recognition algorithm
Beta pruning coefficient when branch path.Wherein, beta pruning coefficient is higher, easier to crop when carrying out best branch route searching
The not high enough individual path of probability, to make CPU operand reduce, correspondingly, the search precision of speech recognition algorithm also declines.
Specifically, show that the CPU usage of current intelligent appliance is higher when the CPU for monitoring current intelligent appliance consumes relatively high,
The beta pruning coefficient for suitably increasing speech recognition algorithm is then needed, to reduce search precision, and then reduces CPU consumption;Conversely, working as
When CPU consumption is relatively low, shows that the CPU usage of current intelligent appliance is lower, then need suitably to reduce speech recognition algorithm
Beta pruning coefficient, to improve the search precision of speech recognition algorithm.
S350, operation speech recognition algorithm, identify user voice command.
Specifically, determining cutting for speech recognition modeling and speech recognition algorithm in the operative scenario according to intelligent appliance
After branch coefficient, the voice messaging that speech recognition modeling and user issue is input to speech recognition algorithm, to run voice knowledge
Other algorithm, the voice messaging issued to user identifies, obtains voice command words, to carry out voice control to intelligent appliance.
The technical solution of the present embodiment selects suitable speech recognition modeling according to the usage scenario of intelligent appliance, is protecting
Under the premise of demonstrate,proving phonetic recognization rate, the consumption of CPU calculation resources is reduced as far as possible;Meanwhile it being used according to the CPU of current intelligent appliance
Rate, dynamic adjust the search precision of speech recognition algorithm, improve the utilization rate of CPU.
Fig. 4 is a kind of structural schematic diagram for multiple intelligent appliances that the embodiment of the present invention three provides.As shown in figure 4, with certain
In the family of a user configured with intelligent range hood 301, intelligent air condition 302, smart television 303 and intelligent water heater 304 this four
The control method of intelligent appliance is specifically described in a terminal device.
Assuming that user 300 issues the voice messaging of " suspending TV programme ", to carry out voice control to smart television 303
System, if but the working environment of intelligent range hood 301 is also in audio and video playing scene at this time, and height is general in speech recognition modeling
Also include " pause " similar analog voice order word in the order word list that rate uses, is then needed at this time according to voice command words
In device name select corresponding intelligent appliance, to carry out voice control to it.Specifically, the language is issued in user 300
After message breath, this four terminal devices under current environment can receive the voice messaging, due to intelligent range hood
301 and smart television 303 be in audio and video playing scene, then smart television 303 is when receiving the voice messaging, by the language
Message breath and the speech recognition modeling determined previously according to audio and video playing scene are input in speech recognition algorithm, to obtain pair
The voice command words answered, then according to the voice command words by the program break of smart television.Fig. 5 is that the embodiment of the present invention three mentions
The display schematic diagram of a kind of TV programme pause supplied, as shown in figure 5, being shown on the display screen 3031 of smart television 303
The interface of pause button 3032, to prompt user to suspend TV programme, so as to see user intuitively on a display screen
Observe the effect after the execution voice command words of smart television 303;And intelligent range hood 301 is when receiving the voice messaging,
Due to not including " kitchen ventilator " this device name in the voice messaging, then intelligent range hood 301 and not according to the voice
Information obtains corresponding voice command words, i.e. intelligent range hood 301 and without any response.
Certainly, in embodiment only illustratively with " suspending TV programme " and intelligent range hood 301, intelligent air condition
302, for smart television 303 and intelligent water heater 304, the control method of intelligent appliance is illustrated.In practical operation
In the process, user can set according to the actual situation, be not limited thereto.
Example IV
Fig. 6 is a kind of structural schematic diagram of the control device for intelligent appliance that the embodiment of the present invention four provides.The present embodiment
The control device of the intelligent appliance of offer can integrate in intelligent appliance, and it is real which can be two or more physics
Body is constituted, and is also possible to a physical entity and is constituted, which can be intelligent appliance, for example, intelligent range hood, intelligence
Air-conditioning, smart television, intelligent water heater etc..With reference to Fig. 6, the control device of intelligent appliance provided in this embodiment is specifically included:
Receiving module 410 obtains module 420 and control module 430.
Wherein, receiving module 410, for receiving voice messaging;
Module 420 is obtained, for obtaining voice command words, which is that voice messaging and speech recognition modeling are defeated
Enter speech recognition algorithm to obtain, which determines according to the operative scenario of intelligent appliance;
Control module 430, for controlling intelligent appliance according to voice command words.
The technical solution of the present embodiment, by receiving the voice messaging of user's input, and according to the yard of intelligent appliance
Scape determines suitable speech recognition modeling, then by voice messaging and speech recognition modeling input speech recognition algorithm to obtain language
Sound order word, and intelligent appliance is controlled according to voice command words.When intelligent appliance is in the lesser operative scenario of operand, make
With the small speech recognition modeling of operand, hardware cost, CPU calculation resources and the CPU power consumption of intelligent appliance are reduced;In intelligent family
When electric operative scenario biggish in operand, the speech recognition modeling for adaptively using operand big, thus to CPU operation
Resource is rationally utilized, and guarantees that the system response time of the course of work of intelligent appliance is good and overall power is low.
On the basis of the above embodiments, the control device of the intelligent appliance, further includes:
Determining module, the search precision for speech recognition algorithm are determined according to the working performance of intelligent appliance.
On the basis of the above embodiments, the CPU that the working performance of the intelligent appliance includes at least intelligent appliance is used
One of rate and CPU power consumption.
On the basis of the above embodiments, the determining module, further includes:
Acquiring unit, for obtaining the current process runing time of intelligent appliance;
Computing unit, for calculating the beta pruning coefficient of speech recognition algorithm, which makes according to the CPU of intelligent appliance
It is determined with rate, which is the ratio that current process runing time accounts for CPU total run time, search precision and beta pruning system
Number is inversely proportional.
On the basis of the above embodiments, the control device of the intelligent appliance, further includes:
Training module, for training speech recognition modeling, the instruction of the speech recognition modeling before obtaining voice command words
Practice the simulation operative scenario information and analog voice order word that input parameter includes intelligent appliance.
On the basis of the above embodiments, speech recognition modeling includes acoustic model and language model.
On the basis of the above embodiments, the operative scenario of intelligent appliance according to the load operating mode of intelligent appliance, after
Platform runs at least one program and current time and determines.
The controlling party of intelligent appliance provided by any embodiment of the invention can be performed in the control device of above-mentioned intelligent appliance
Method has the corresponding functional module of execution method and beneficial effect.
Embodiment five
Fig. 7 is a kind of structural schematic diagram for intelligent appliance that the embodiment of the present invention five provides.The intelligent appliance includes: processing
Device 510, memory 520, input unit 530, output device 540 and display screen 550.Processor 510 in the intelligent appliance
Quantity can be one or more, in Fig. 7 by taking a processor 510 as an example.The quantity of memory 520 can in the intelligent appliance
To be one or more, in Fig. 7 by taking a memory 520 as an example.It is the processor 510 of the intelligent appliance, memory 520, defeated
Entering device 530, output device 540 and display screen 550 can be connected by bus or other modes, by total in Fig. 7
For line connection.In embodiment, intelligent appliance can be intelligent range hood, intelligent air condition, smart television, intelligent water heater etc..
Memory 520 is used as a kind of computer readable storage medium, can be used for storing software program, journey can be performed in computer
Sequence and module, the corresponding program instruction/module of intelligent appliance as described in any embodiment of that present invention is (for example, intelligent appliance
Control device in receiving module 510, obtain module 520 and control module 530).Memory 520 can mainly include storage
Program area and storage data area, wherein storing program area can application program needed for storage program area, at least one function;
Storage data area, which can be stored, uses created data etc. according to equipment.In addition, memory 520 may include that high speed is deposited at random
Access to memory, can also include nonvolatile memory, a for example, at least disk memory, flush memory device or other are non-
Volatile solid-state part.In some instances, memory 520 can further comprise remotely located relative to processor 510
Memory, these remote memories can pass through network connection to equipment.The example of above-mentioned network includes but is not limited to interconnect
Net, intranet, local area network, mobile radio communication and combinations thereof.
Input unit 530 can be used for receiving the number or character information of input, and generate the user with intelligent appliance
Setting and the related key signals input of function control, can also be the camera for obtaining image and obtain audio data
Pick up facility.Output device 540 may include the audio frequency apparatuses such as loudspeaker.It should be noted that input unit 530 and output
The concrete composition of device 540 may be set according to actual conditions.
Display screen 550 can be the display screen 550 with touch function, can be capacitance plate, electromagnetic screen or infrared screen.
In general, display screen 550 is used to show data, such as the status information of display intelligent appliance according to the instruction of processor 510,
For example, be also used to receive the touch operation for acting on display screen 550 for showing the information such as information about power, locating operating mode,
For example, can be to tune up the options such as air quantity, switching working mode, and corresponding signal is sent when intelligent appliance is kitchen ventilator
To processor 510 or other devices.
Software program, instruction and the module that processor 510 is stored in memory 520 by operation, thereby executing setting
Standby various function application and data processing, that is, realize the control method of above-mentioned intelligent appliance.
The intelligent appliance of above-mentioned offer can be used for executing the control method for the intelligent appliance that above-mentioned any embodiment provides, tool
Standby corresponding function and beneficial effect.
Embodiment six
The embodiment of the present invention six also provides a kind of storage medium comprising computer executable instructions, and the computer can be held
Row instruction by computer processor when being executed for executing a kind of control method of intelligent appliance, comprising:
Receive voice messaging;
Voice command words are obtained, which is that voice messaging and speech recognition modeling input speech recognition algorithm and obtain
It arrives, which determines according to the operative scenario of intelligent appliance;
Intelligent appliance is controlled according to voice command words.
Certainly, a kind of storage medium comprising computer executable instructions, computer provided by the embodiment of the present invention
The control method for the intelligent appliance that executable instruction is not limited to the described above operates, and any embodiment of that present invention institute can also be performed
Relevant operation in the control method of the intelligent appliance of offer, and have corresponding function and beneficial effect.
By the description above with respect to embodiment, it is apparent to those skilled in the art that, the present invention
It can be realized by software and required common hardware, naturally it is also possible to which by hardware realization, but in many cases, the former is more
Good embodiment.Based on this understanding, technical solution of the present invention substantially in other words contributes to the prior art
Part can be embodied in the form of software products, which can store in computer readable storage medium
In, floppy disk, read-only memory (Read-Only Memory, ROM), random access memory (Random such as computer
Access Memory, RAM), flash memory (FLASH), hard disk or CD etc., including some instructions are with so that a computer is set
Standby (can be robot, personal computer, server or the network equipment etc.) executes described in any embodiment of that present invention
The test method of VBO display interface.
It is worth noting that, in the control device of above-mentioned intelligent appliance, included each unit and module only according to
What function logic was divided, but be not limited to the above division, as long as corresponding functions can be realized;In addition, each
The specific name of functional unit is also only for convenience of distinguishing each other, the protection scope being not intended to restrict the invention.
It should be appreciated that each section of the invention can be realized with hardware, software, firmware or their combination.Above-mentioned
In embodiment, software that multiple steps or method can be executed in memory and by suitable instruction execution system with storage
Or firmware is realized.It, and in another embodiment, can be under well known in the art for example, if realized with hardware
Any one of column technology or their combination are realized: having a logic gates for realizing logic function to data-signal
Discrete logic, with suitable combinational logic gate circuit specific integrated circuit, programmable gate array (PGA), scene
Programmable gate array (FPGA) etc..
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
Centainly refer to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be any
One or more embodiment or examples in can be combined in any suitable manner.
Note that the above is only a better embodiment of the present invention and the applied technical principle.It will be appreciated by those skilled in the art that
The invention is not limited to the specific embodiments described herein, be able to carry out for a person skilled in the art it is various it is apparent variation,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above embodiments to the present invention
It is described in further detail, but the present invention is not limited to the above embodiments only, without departing from the inventive concept, also
It may include more other equivalent embodiments, and the scope of the invention is determined by the scope of the appended claims.
Claims (10)
1. a kind of control method of intelligent appliance characterized by comprising
Receive voice messaging;
Voice command words are obtained, the voice command words are that the voice messaging and speech recognition modeling input speech recognition algorithm
It obtains, the speech recognition modeling is determined according to the operative scenario of intelligent appliance;
The intelligent appliance is controlled according to the voice command words.
2. the method according to claim 1, wherein the search precision of the speech recognition algorithm is according to the intelligence
The working performance of energy household electrical appliances determines.
3. according to the method described in claim 2, it is characterized in that, the working performance of the intelligent appliance includes at least the intelligence
One of CPU usage and CPU power consumption of energy household electrical appliances.
4. according to the method described in claim 3, it is characterized in that, the search precision of the speech recognition algorithm is according to the intelligence
Can the working performance determinations of household electrical appliances include:
Obtain the current process runing time of the intelligent appliance;
The beta pruning coefficient of the speech recognition algorithm is calculated, the beta pruning coefficient is true according to the CPU usage of the intelligent appliance
Fixed, the CPU usage is the ratio that current process runing time accounts for CPU total run time, search precision and the beta pruning system
Number is inversely proportional.
5. the method according to claim 1, wherein before the acquisition voice command words, further includes:
Training speech recognition modeling, the training input parameter of the speech recognition modeling include the simulation work of the intelligent appliance
Scene information and analog voice order word.
6. the method according to claim 1, wherein the speech recognition modeling includes acoustic model and language mould
Type.
7. the method according to claim 1, wherein the operative scenario of the intelligent appliance is according to the intelligent family
At least one load operating mode, running background program and current time of electricity determine.
8. a kind of control device of intelligent appliance characterized by comprising
Receiving module, for receiving voice messaging;
Module is obtained, for obtaining voice command words, the voice command words are that the voice messaging and speech recognition modeling are defeated
Enter speech recognition algorithm to obtain, the speech recognition modeling is determined according to the operative scenario of intelligent appliance;
Control module, for controlling the intelligent appliance according to the voice command words.
9. a kind of intelligent appliance characterized by comprising display screen, memory and one or more processors;
The display screen, for showing the status information of intelligent appliance;
The memory, for storing one or more programs;
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
The now control method of the intelligent appliance as described in any in claim 1-7.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor
The control method of the intelligent appliance as described in any in claim 1-7 is realized when execution.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910126388.2A CN109688036B (en) | 2019-02-20 | 2019-02-20 | Control method and device of intelligent household appliance, intelligent household appliance and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910126388.2A CN109688036B (en) | 2019-02-20 | 2019-02-20 | Control method and device of intelligent household appliance, intelligent household appliance and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109688036A true CN109688036A (en) | 2019-04-26 |
CN109688036B CN109688036B (en) | 2021-12-03 |
Family
ID=66196593
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910126388.2A Active CN109688036B (en) | 2019-02-20 | 2019-02-20 | Control method and device of intelligent household appliance, intelligent household appliance and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109688036B (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110136717A (en) * | 2019-05-22 | 2019-08-16 | 北京汽车股份有限公司 | Voice recognition processing method, apparatus, system and vehicle |
CN110597139A (en) * | 2019-09-25 | 2019-12-20 | 珠海格力电器股份有限公司 | Self-learning method and system of cooking appliance |
CN110706706A (en) * | 2019-11-01 | 2020-01-17 | 北京声智科技有限公司 | Voice recognition method, device, server and storage medium |
CN111369992A (en) * | 2020-02-27 | 2020-07-03 | Oppo(重庆)智能科技有限公司 | Instruction execution method and device, storage medium and electronic equipment |
CN111669304A (en) * | 2020-05-19 | 2020-09-15 | 广东科徕尼智能科技有限公司 | Intelligent household scene control method and equipment based on edge gateway and storage medium |
CN112579249A (en) * | 2019-09-30 | 2021-03-30 | 奇安信安全技术(珠海)有限公司 | Multi-CPU virtual machine operation method and device, storage medium and computer equipment |
CN112781248A (en) * | 2020-10-28 | 2021-05-11 | 青岛经济技术开发区海尔热水器有限公司 | Voice control method and device for intelligent water heater, electronic equipment and storage medium |
CN112890572A (en) * | 2021-02-07 | 2021-06-04 | 广州一盒科技有限公司 | Intelligent control system and method for cooking food materials |
CN113284489A (en) * | 2021-04-16 | 2021-08-20 | 珠海格力电器股份有限公司 | Voice equipment control method and device, storage medium and voice equipment |
CN113518023A (en) * | 2021-09-13 | 2021-10-19 | 深圳小小小科技有限公司 | Control method and device for household appliance |
CN115031266A (en) * | 2022-05-10 | 2022-09-09 | 海信(山东)厨卫有限公司 | Range hood and voice recognition method thereof |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101740028A (en) * | 2009-11-20 | 2010-06-16 | 四川长虹电器股份有限公司 | Voice control system of household appliance |
CN101751924A (en) * | 2009-12-10 | 2010-06-23 | 清华大学 | Embedded platform recognition method for large vocabulary quantity of voice command words |
CN105301982A (en) * | 2015-11-23 | 2016-02-03 | 广东美的暖通设备有限公司 | Control method and system for intelligent household electrical appliance |
CN106997762A (en) * | 2017-03-08 | 2017-08-01 | 广东美的制冷设备有限公司 | The sound control method and device of household electrical appliance |
CN107527615A (en) * | 2017-09-13 | 2017-12-29 | 联想(北京)有限公司 | Information processing method, device, equipment, system and server |
CN207074554U (en) * | 2017-05-17 | 2018-03-06 | 詹翀 | A kind of more scene command word speech recognition devices |
EP3382697A1 (en) * | 2017-03-30 | 2018-10-03 | LG Electronics Inc. | Voice server, voice recognition server system, and method of operating the same |
-
2019
- 2019-02-20 CN CN201910126388.2A patent/CN109688036B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101740028A (en) * | 2009-11-20 | 2010-06-16 | 四川长虹电器股份有限公司 | Voice control system of household appliance |
CN101751924A (en) * | 2009-12-10 | 2010-06-23 | 清华大学 | Embedded platform recognition method for large vocabulary quantity of voice command words |
CN105301982A (en) * | 2015-11-23 | 2016-02-03 | 广东美的暖通设备有限公司 | Control method and system for intelligent household electrical appliance |
CN106997762A (en) * | 2017-03-08 | 2017-08-01 | 广东美的制冷设备有限公司 | The sound control method and device of household electrical appliance |
EP3382697A1 (en) * | 2017-03-30 | 2018-10-03 | LG Electronics Inc. | Voice server, voice recognition server system, and method of operating the same |
CN207074554U (en) * | 2017-05-17 | 2018-03-06 | 詹翀 | A kind of more scene command word speech recognition devices |
CN107527615A (en) * | 2017-09-13 | 2017-12-29 | 联想(北京)有限公司 | Information processing method, device, equipment, system and server |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110136717A (en) * | 2019-05-22 | 2019-08-16 | 北京汽车股份有限公司 | Voice recognition processing method, apparatus, system and vehicle |
CN110597139A (en) * | 2019-09-25 | 2019-12-20 | 珠海格力电器股份有限公司 | Self-learning method and system of cooking appliance |
CN112579249A (en) * | 2019-09-30 | 2021-03-30 | 奇安信安全技术(珠海)有限公司 | Multi-CPU virtual machine operation method and device, storage medium and computer equipment |
CN110706706A (en) * | 2019-11-01 | 2020-01-17 | 北京声智科技有限公司 | Voice recognition method, device, server and storage medium |
CN111369992A (en) * | 2020-02-27 | 2020-07-03 | Oppo(重庆)智能科技有限公司 | Instruction execution method and device, storage medium and electronic equipment |
CN111669304A (en) * | 2020-05-19 | 2020-09-15 | 广东科徕尼智能科技有限公司 | Intelligent household scene control method and equipment based on edge gateway and storage medium |
CN111669304B (en) * | 2020-05-19 | 2022-03-15 | 广东好太太智能家居有限公司 | Intelligent household scene control method and equipment based on edge gateway and storage medium |
CN112781248A (en) * | 2020-10-28 | 2021-05-11 | 青岛经济技术开发区海尔热水器有限公司 | Voice control method and device for intelligent water heater, electronic equipment and storage medium |
CN112890572A (en) * | 2021-02-07 | 2021-06-04 | 广州一盒科技有限公司 | Intelligent control system and method for cooking food materials |
CN113284489A (en) * | 2021-04-16 | 2021-08-20 | 珠海格力电器股份有限公司 | Voice equipment control method and device, storage medium and voice equipment |
CN113518023A (en) * | 2021-09-13 | 2021-10-19 | 深圳小小小科技有限公司 | Control method and device for household appliance |
CN115031266A (en) * | 2022-05-10 | 2022-09-09 | 海信(山东)厨卫有限公司 | Range hood and voice recognition method thereof |
Also Published As
Publication number | Publication date |
---|---|
CN109688036B (en) | 2021-12-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109688036A (en) | Control method and device of intelligent household appliance, intelligent household appliance and storage medium | |
KR102373905B1 (en) | Shortened voice user interface for assistant applications | |
US11860913B2 (en) | Streaming real-time dialog management | |
KR101359718B1 (en) | Conversation Managemnt System and Method Thereof | |
EP3158427B1 (en) | System and method for speech-enabled personalized operation of devices and services in multiple operating environments | |
CN105575386B (en) | Audio recognition method and device | |
CN110286601A (en) | Method and device for controlling intelligent household equipment, control equipment and storage medium | |
CN108737933A (en) | A kind of dialogue method, device and electronic equipment based on intelligent sound box | |
US20200365150A1 (en) | Home appliance and control method therefor | |
CN109147782A (en) | Control method, device and the air-conditioning of air-conditioning | |
CN110299136A (en) | A kind of processing method and its system for speech recognition | |
CN106452987B (en) | A kind of sound control method and device, equipment | |
CN109597313A (en) | Method for changing scenes and device | |
CN112634897B (en) | Equipment awakening method and device, storage medium and electronic device | |
CN110176230A (en) | A kind of audio recognition method, device, equipment and storage medium | |
CN110503954A (en) | Voice technical ability starts method, apparatus, equipment and storage medium | |
CN111933135A (en) | Terminal control method and device, intelligent terminal and computer readable storage medium | |
CN112463108B (en) | Voice interaction processing method and device, electronic equipment and storage medium | |
CN111756603B (en) | Control method and device of intelligent home system, electronic equipment and readable medium | |
CN109215645A (en) | A kind of voice messaging exchange method and intelligent electric appliance | |
CN117253479A (en) | Voice control method and system applied to wax-melting aromatherapy lamp | |
CN112787899B (en) | Equipment voice interaction method, computer readable storage medium and refrigerator | |
CN110517680A (en) | A kind of data detection method and device, storage medium of artificial intelligence | |
CN109346081A (en) | Voice control method, device, equipment and storage medium | |
CN114694635A (en) | Sleep scene setting method and device |
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 | ||
TA01 | Transfer of patent application right | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20200407 Address after: 510530, No. four, 6, Po Po Road, Yun Po Industrial Zone, Whampoa District, Guangzhou, Guangdong Applicant after: Guangzhou Shiyuan Electronic Technology Company Limited Address before: 510530, No. four, 6, Po Po Road, Yun Po Industrial Zone, Whampoa District, Guangzhou, Guangdong Applicant before: Guangzhou Shiyuan Electronic Technology Company Limited Applicant before: GUANGZHOU RUIXIN ELECTRONICS Co.,Ltd. |
|
GR01 | Patent grant | ||
GR01 | Patent grant |