CN109634404A - A kind of system and method for the controllable interactive interface based on air blowing - Google Patents
A kind of system and method for the controllable interactive interface based on air blowing Download PDFInfo
- Publication number
- CN109634404A CN109634404A CN201811296072.XA CN201811296072A CN109634404A CN 109634404 A CN109634404 A CN 109634404A CN 201811296072 A CN201811296072 A CN 201811296072A CN 109634404 A CN109634404 A CN 109634404A
- Authority
- CN
- China
- Prior art keywords
- air blowing
- data
- model
- training
- client
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/011—Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
The system and method for present disclose provides a kind of controllable interactive interface based on air blowing, including server end and client, server end include collecting training data module, are configured as collecting three kinds of air blowing data as training data in the training stage；Data preprocessing module, is configured as the initial data of acquisition is normalized and discretization operations；Characteristic extracting module is configured as extracting feature to the data pre-processed；Model training module is configured as being trained sorter model the characteristic extracted；Identification module is configured as carrying out type identification to the air blowing of user using trained disaggregated model, and sends client for the type identified；The client is configured as detection air blowing interactive signal, is sent to server end and is identified, based on the air blowing type from received server-side, match corresponding interactive operation.
Description
Technical field
This disclosure relates to a kind of system and method for the controllable interactive interface based on air blowing.
Background technique
Only there is provided background technical informations relevant to the disclosure for the statement of this part, it is not necessary to so constitute first skill
Art.
From traditional interactive mode such as keyboard, mouse to multi-point touch, voice, gesture, posture, eye tracking and brain-machine
Natural interactive styles, the human-computer interaction technique fields such as interaction achieve huge progress.However, these interactive modes are not necessarily fitted
For all scenes.For example, when can not carry out mouse when both hands are occupied or touch interaction, in language noisy or inconvenient to use
The people that interactive voice or physical handicaps can not be carried out in the environment of sound can not carry out voice or eye tracking interaction etc..
Breathing is the instinct of the mankind, since breathing can be controlled consciously, is interacted using breathing and is considered as
Influence another controlling mechanism of real physical world and virtual world.The convenience and controllability for breathing interaction are to a certain degree
On the shortcomings that can making up above-mentioned common interactive mode.Therefore, breathing be counted as it is a kind of be only second to touch and voice it is direct
Controllable natural interactive style.
So far, some researchs by breathing or are blown as a kind of interactive controlling mode directly inputted.According to obtaining
The difference of the related sensor wearing position of breath signal is taken, breathing detection method can be divided into two classes: first is that by detection chest
The movement of portion or abdomen obtains breath signal；Another kind be by be placed on the special device of mouth directly detect sucking and exhale
Air-flow obtains breath signal out.Existing research is generally dependent upon special equipment for customizing, as respiration transducer, breathing zone or
Other special sensors, these equipment are typically more complicated, and expensive and inconvenient wearing lacks general in daily life
Portability, there may come a time when to be difficult to control.
Summary of the invention
The disclosure to solve the above-mentioned problems, proposes a kind of system and method for controllable interactive interface based on air blowing,
The disclosure does not need expensive breath signal and obtains equipment, and user can choose simple practice or directly carry out using can
It easily to carry, whenever and wherever possible easily uses, is not limited by complex device and particular service requirement.
To achieve the goals above, the disclosure adopts the following technical scheme that
A kind of system of the controllable interactive interface based on air blowing, including server end and client, the server end packet
It includes:
Collecting training data module is configured as collecting three kinds of air blowing data as training data in the training stage；
Data preprocessing module, is configured as the initial data of acquisition is normalized and discretization operations；
Characteristic extracting module is configured as extracting feature to the data pre-processed；
Model training module is configured as being trained sorter model the characteristic extracted；
Identification module is configured as carrying out type identification to the air blowing of user using trained disaggregated model, and will know
Good type is not sent to client；
The client is configured as detection air blowing interactive signal, is sent to server end and is identified, is based on from service
The received air blowing type in device end, matches corresponding interactive operation.
Controllable exchange method based on air blowing, comprising the following steps:
The a certain number of air blowing data of random acquisition, form the training dataset for training sorter model；
Collected initial data is pre-processed, by data normalization to critical field [- 1,1]；Use sliding window
Mouth carries out sliding-model control to original continuous sound wave data, forms discrete sample；
Extract the data of the ratio of mean value, variance, first-order difference and standardized value greater than 0.4；
It selects support vector machines as sorter model, trains sorter model based on four characteristic signals of extraction,
And the accuracy of disaggregated model is verified using random cross validation；
Using trained model, first determine whether user selects to practice；If so, instruction is added in practice data
Practice data set, sample is handled and extracted again；Otherwise formal to use.
It is limited as further, once detecting that user has carried out interaction of blowing, instruction will be sent to server end；Clothes
Business device end obtains air blowing data, carries out sliding-model control and the extraction of data, carries out identification classification using trained model；Clothes
Classification results are sent client by business device；
According to the received identification types of client, corresponding interactive operation is executed.
It is limited as further, sample frequency 8000Hz is once blown at regular intervals.
It is limited as further, in data acquisition, according to the duration of air blowing and intensity, defines three kinds
Inflatable form is for interacting: it is big to blow, greater than some strength air blowing and at least continue N seconds；Exhale by force: greater than some strength is lasting
Less than N seconds；Featheriness: less than setting intensity and continuously less than N seconds.
It is limited as further, the specific steps of training pattern algorithm include:
By pre-processing blowing action a with the training data d of feature extraction and corresponding thereto；
The training that disaggregated model SVM is carried out using training data d and air blowing type a, obtains new model m；
The model in aol server is updated using new model m timing；
Model m is returned to, the type blown for server identification.
It is limited as further, when selecting support vector machines as sorter model, calls mesh parameter optimizing letter
Number SVMcgForClass carrys out the ginseng in the penalty parameter c in Optimum Classification device SVM and Gaussian radial basis function RBF kernel function
Number g verifies the accuracy of disaggregated model using random cross validation method to improve nicety of grading: by whole numbers of acquisition
According to a training set and a test set is randomly divided into, its ratio be 9:1；It repeatedly carries out crossing over many times testing and obtain averagely accurately
Rate verifies the accuracy and validity of model.
It is limited as further, the specific steps of air blowing type identification algorithm include:
The threshold θ being set in advance, for judging whether user has carried out blowing action；
The volume value v of the sound generated after blowing by microphone is obtained in real time；
Judge the size of volume value v and threshold θ, if v > θ, client sends voice data d, service to server end
Device end obtains voice data d, then carries out pretreatment and feature extraction operation to d, obtains d '；Identify that current blow is moved according to d '
Make, obtains air blowing type a, be sent to client；Client executes corresponding interactive operation according to a received；Otherwise, then
The secondary real-time volume value for obtaining the sound generated after air blowing by microphone.
It is limited as further, corresponding interactive operation includes:
The broadcasting of the end PC video；
The browsing of mobile phone mobile terminal webpage；
The operation of reality-virtualizing game.
The working principle of the disclosure are as follows: Web server is as acquisition and handles air blowing data, training and operation disaggregated model
Server end.When user uses the air blowing interactive interface, the classification prediction model run on the server can be according to the use
Recognition result is sent back client (for example, the application of the end PC, mobile phone movement to identify its blowing action by the air blowing data at family
End application or HTC VIVE virtual system).Once client receives the blowing action type of identification, the application program of operation is just
Corresponding operation can be executed.
Compared with prior art, the disclosure has the beneficial effect that
The disclosure is easy to use, easy to operate, cost is relatively low；
The disclosure can use in the noisy scene of 65db or so in environmental noise, work well, such as ambient music is
64db and source of sound are no more than 0.5 meter apart from user, and have 3-4 people in normal communication in the same space, and interaction accuracy rate is reachable
94%；
The disclosure is very convenient to use in certain special populations, as not being available the deaf and dumb crowd of voice, not being available touching
Control the physical handicaps crowd etc. of interaction；
The disclosure has a wide range of application, in addition to being used in common application, such as the system that the disclosure is mentioned, most probably in similar
It blows when receiving calls, navigating when driving and the amplification of control map and reduces, cooks in noisy kitchen study and by this friendship
Mutual mode, which combines to be generalized in deaf and dumb crowd with number dictionary, a variety of special circumstances such as uses.
Detailed description of the invention
The accompanying drawings constituting a part of this application is used to provide further understanding of the present application, and the application's shows
Meaning property embodiment and its explanation are not constituted an undue limitation on the present application for explaining the application.
Fig. 1 is the system architecture diagram of the disclosure；
Fig. 2 is the use schematic illustration of the disclosure；
Fig. 3 is disclosure algorithmic technique flow chart；
Fig. 4 (a)-Fig. 4 (c) is the example schematic that the disclosure is applied to the end PC；
Fig. 5 (a)-Fig. 5 (b) is the example schematic that the disclosure is applied to mobile phone mobile terminal；
Fig. 6 (a)-Fig. 6 (c) is the example schematic that the disclosure is applied to virtual game.
Specific embodiment:
The disclosure is described further with embodiment with reference to the accompanying drawing.
It is noted that following detailed description is all illustrative, it is intended to provide further instruction to the application.Unless another
It indicates, all technical and scientific terms used herein has usual with the application person of an ordinary skill in the technical field
The identical meanings of understanding.
It should be noted that term used herein above is merely to describe specific embodiment, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singular
Also it is intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet
Include " when, indicate existing characteristics, step, operation, device, component and/or their combination.
In the disclosure, term for example "upper", "lower", "left", "right", "front", "rear", "vertical", "horizontal", " side ",
The orientation or positional relationship of the instructions such as "bottom" is to be based on the orientation or positional relationship shown in the drawings, only to facilitate describing this public affairs
The relative for opening each component or component structure relationship and determination, not refers in particular to either component or element in the disclosure, cannot understand
For the limitation to the disclosure.
In the disclosure, term such as " affixed ", " connected ", " connection " be shall be understood in a broad sense, and indicate may be a fixed connection,
It is also possible to be integrally connected or is detachably connected；It can be directly connected, it can also be indirectly connected through an intermediary.For
The related scientific research of this field or technical staff can determine the concrete meaning of above-mentioned term in the disclosure as the case may be,
It should not be understood as the limitation to the disclosure.
Simple, convenient and directly controllable natural interaction interface system based on air blowing, including collecting training data mould
Block, data preprocessing module, characteristic extracting module, model training module, identification module and system client application module.Its
In,
Collecting training data module, in the training stage, for collecting three kinds of air blowing data as training data；
Data preprocessing module, for the initial data of acquisition being normalized and the pretreatment operations such as discretization；
Characteristic extracting module, for extracting feature to the data pre-processed；
Model training module is trained sorter model using the characteristic extracted；
Identification module carries out type identification, and the class that will have been identified to the air blowing of user using trained disaggregated model
Type is sent to applications client；
Client application module is sent to server end and is identified for detecting air blowing interactive signal, be then based on from
The air blowing type of received server-side, carries out corresponding interactive operation in related application.The related application such as end PC, mobile phone are mobile
End and virtual game system.
Working method based on above system, comprising the following steps:
(1) in the training stage, by randomly choosing two personnelAcquisition is a large amount ofAir blowing data are formed for training classification
The training dataset of device model, wherein sample frequency is 8000Hz；
(2) collected initial data is pre-processed, by data normalization to critical field [- 1,1]；Then it uses
The sliding window that one length is 24000Hz, Duplication is 67% carries out sliding-model control, shape to original continuous sound wave data
At discrete sample；
(3) extract following characteristics to the data pre-processed: mean value, variance, first-order difference and standardized value are greater than 0.4
Ratio；
(4) it selects support vector machines (SVM) as sorter model, is trained based on four characteristic signals of said extracted
Sorter model, and verify using random cross validation the accuracy of disaggregated model；
(5) when user uses trained model in the client, first determine whether user selects to practice；If
It is that training dataset is added in practice data, is executed step (2) (3) (4)；Otherwise, (6) are executed, user formally uses；
(6) client once detects that user has carried out interaction of blowing, and instruction will be sent to server end；Server end obtains
Air blowing data are taken, are executed step (2) (3), step (4) is then executed using trained model and carries out identification classification；Server
Classification results are sent into client；
(7) according to the received identification types of client, related application system implements corresponding interactive operation；
(8) step (6) are executed；
In step (1), instruction sufficiently large, for training sorter model is formed using the air blowing data of two participants
Practice data set, sample frequency 8000hHz.In order to keep training set more standardized, participant carries out one in every 3 seconds consciously
Secondary blowing operation.
In step (1), in data acquisition, according to the duration of air blowing and intensity, three kinds of air blowing shapes are defined
Formula is used to interact: gale (big blow): very strong air blowing and for 3 seconds；Gust (is exhaled by force): strong but of short duration air blowing, is continued
About 1 second；Breeze (featheriness): it is slowly mild and of short duration, the duration about 1 second.
In step (3), based on when feature extraction specific formula is as follows:
Wherein, XiIt is i-th of element in discrete sample, n is the length of each sample.Using these features from air blowing
Intensity and the air blowing that three types are distinguished on the duration.Wherein, average value (formula (1)), variance (formula (2)) and single order side
Poor (formula (3)) are all to commonly use classical statistical nature, and feature p (formula (4)) can better discriminate between the intensity of air blowing and hold
The continuous time.
In step (4), the specific steps of training pattern algorithm include:
Input: by pre-processing blowing action a with the training data d of feature extraction and corresponding thereto；
Output: disaggregated model m；
(4-1) carries out the training of disaggregated model SVM using training data d and air blowing type a, obtains new model m；
(4-2) updates the model in aol server using new model m timing；
(4-3) returns to model m, the type blown for server identification.
When selecting support vector machines (SVM) as sorter model, mesh parameter optimizing function is called
SVMcgForClass carrys out the parameter in the penalty parameter c in Optimum Classification device SVM and Gaussian radial basis function RBF kernel function
G, to improve nicety of grading.The accuracy of disaggregated model is verified using random cross validation method: by the total data of acquisition
It is randomly divided into a training set and a test set, its ratio be 9:1；10 intersections of repeatedly progress, which are tested, obtains Average Accuracy
To verify the accuracy and validity of model.This is independently of client application.
In step (5), before formally using the air blowing interactive interface, user, which can choose, carries out certain air blowing practice
It is familiar with available interactive form, obtains the air blowing data with user tiny individual difference at the same time, increased to
The training data of server end is concentrated, and model is allowed preferably " to remember " blowing method of the user, improves recognition accuracy.Into
When row practice, user executes correct air blowing according to the air blowing type of prompt, and every kind of air blowing type repeats practice 10 times, user
Personal air blowing data be all uploaded in the training set of server.The practice process is optional, so using without practice
The air blowing interactive interface also can be used normally in family.
In the step (6), the specific steps of air blowing type identification algorithm include:
Input: the threshold θ being set in advance, for judging whether user has carried out blowing action；
(6-1) is used when the interactive interface in the disclosure, and circulation executes following step；
(6-2) client obtains the volume value v of the sound generated after air blowing by microphone in real time；
(6-3) judges the size of v value and θ value, if v > θ, client sends voice data d to server end, under execution
State step (6-4)；Otherwise, (6-2) is executed；
(6-4) server end obtains voice data d, then carries out pretreatment and feature extraction operation to d, obtains d '；Root
Current blowing action is identified according to d ', is obtained air blowing type a, is sent to client；
(6-5) client executes corresponding interactive operation according to a received, in related application system.
In the step (7), there are three applications involved by the disclosure, specific as follows:
The broadcasting of the end (7-1) PC video.Client, as video display platform, obtains user in real time and blows by Unity3D
Whether gas sound wave, real time monitoring user start blowing operation: if the wave volume monitored in real time is more than previously given threshold
Value is then judged as user using interaction of blowing.In this case, Unity client can send one to server immediately
Request, server end can obtain voice data and carry out feature extraction to carry out identification classification to it, recognition result is sent to
Unity client.Then, unity client executes corresponding interactive operation according to recognition result: gale control video broadcasting and
Pause, gust control the switching of next video, and breeze is used to be switched to a video.
The browsing of (7-2) mobile phone mobile terminal webpage.User is carried out using the prototype of a simple interactive Web page
Air blowing interactive operation: gust enters the web page interlinkage of current selected, similar to the entrance function of the "enter" key" when the end PC browses webpage
Energy；Breeze control web page interlinkage entry between continuously jumping, similar to the end PC browse webpage when tab key can focus on
Next web page interlinkage.The course of work of server end is similar with the broadcasting of the end PC video.
The operation of (7-3) reality-virtualizing game.Blowing action is incorporated into a " seabed based on Unity3D and HTC VIVE
Seek treasured " VR game in, game background is that player falls into a piece of mysterious sea area and is under a curse into a King Crab, is needed at 6 minutes
Find 8 precious deposits casees inside to restore the identity of oneself, player interacts by blowing action and VIVE handle with virtual environment.Every kind
Blowing action maps different game effects: gale flies marine organisms bullet for shower, the effect of water column sustained firing
It is corresponding with the operational motion breathed out greatly in game；Gust controls the unlatching of precious case for emitting bubble；Breeze is imitated
Crab spits bubble under natural conditions, and with moving about under normal condition, the phenomenon that constantly emitting bubble frequently is corresponding.
As shown in Figure 1, Web server is as the server for obtaining and handling air blowing data, training and operation disaggregated model
End.When user uses the air blowing interactive interface, the classification prediction model run on the server can be according to the air blowing of the user
Data identify its blowing action, and by recognition result send back client (for example, the application of the end PC, mobile phone mobile terminal application or
HTC VIVE virtual system).If user selects exercise mode in client, practice data will be updated the training to period of service end
In data set, and then update training pattern.
As shown in Fig. 2, the part a is user's practice in figure: user's selection first blow by practice, enables and obtains in formal use
To better interaction effect；The part b is the air blowing type identified that client receives, and the application program of operation carries out
Corresponding operation, such as the broadcasting of the end PC video, the operation of the browsing, reality-virtualizing game of mobile phone mobile terminal webpage.
As shown in figure 3, the specific flow chart of the disclosure:
(1) two personnel are randomly choosed, for acquiring the sonic data of 3 kinds of air blowing types；
(2) training set for training sorter model is formed；If there is practicing link, then practice data are added,
Update training set；
(3) initial data is normalized and the pretreatment operations such as discretization；
(4) extract four features of pretreated data: numerical value is greater than after mean value, variance, first-order difference, normalization
0.4 ratio；
(5) it selects support vector machines as disaggregated model, support vector machines is trained using the feature of extraction；
(6) enter the client application stage, threshold θ needed for being set in advance, for judging whether user is blown
Movement；
(7) judge whether user selects to practice, if so, jump procedure (2), updates training dataset；Otherwise, it jumps
Go to step (8)；
(8) user formally uses the air blowing interactive interface, and client monitors air blowing volume v in real time；
(9) whether real-time judge v is greater than θ, if so, jump procedure (10)；Otherwise, jump procedure (8)；
(10) identification classification is carried out using model on trained line in server end；
(11) recognition result is sent to corresponding applications client by server；
(12) according to recognition result, the corresponding interactive operation of client executing；
(13) whether terminate using if it is, terminating；Otherwise, jump procedure (8).
As shown in figure 4, by operation suspension video of breathing out greatly, broadcast button is changed into pause button in Fig. 4 (a)；Scheming
In 4 (b), by exhaling operation to switch next video by force, indicate that the button of next video becomes blue；In Fig. 4 (c), pass through
Featheriness operation is switched to a video, indicates that the button of a upper video becomes blue.
As shown in figure 5, using the web page interlinkage for exhaling movement to enter current selected by force in Fig. 5 (a): successively showing original state
(be shown as white indicate current focus herein), it is ongoing exhale by force operation, into current link (it is shown in red indicate into
Enter link) result；Fig. 5 (b) controls the process continuously jumped between web page interlinkage entry using featheriness: successively showing initial shape
State, ongoing featheriness operation, the result for jumping to next link.
As shown in fig. 6, respectively set different abilities for 3 kinds of air blowing interactions: user is by breathing out greatly in the game system
Gas carrys out shower, such as Fig. 6 (a)；When searching out precious case, user is by exhaling by force discharge bubble to open precious case, such as Fig. 6 (b)；Gently
It blows, is the bubbling operation for triggering leading role crab, such as Fig. 6 (c).
The foregoing is merely preferred embodiment of the present application, are not intended to limit this application, for the skill of this field
For art personnel, various changes and changes are possible in this application.Within the spirit and principles of this application, made any to repair
Change, equivalent replacement, improvement etc., should be included within the scope of protection of this application.
Although above-mentioned be described in conjunction with specific embodiment of the attached drawing to the disclosure, model not is protected to the disclosure
The limitation enclosed, those skilled in the art should understand that, on the basis of the technical solution of the disclosure, those skilled in the art are not
Need to make the creative labor the various modifications or changes that can be made still within the protection scope of the disclosure.
Claims (10)
1. a kind of system of the controllable interactive interface based on air blowing, it is characterized in that: including server end and client, the service
Device end includes:
Collecting training data module is configured as collecting three kinds of air blowing data as training data in the training stage；
Data preprocessing module, is configured as the initial data of acquisition is normalized and discretization operations；
Characteristic extracting module is configured as extracting feature to the data pre-processed；
Model training module is configured as being trained sorter model the characteristic extracted；
Identification module is configured as carrying out type identification to the air blowing of user using trained disaggregated model, and will identify
Type be sent to client；
The client is configured as detection air blowing interactive signal, is sent to server end and is identified, is based on from server end
Received air blowing type, matches corresponding interactive operation.
2. the controllable exchange method based on air blowing, it is characterized in that: the following steps are included:
The a certain number of air blowing data of random acquisition, form the training dataset for training sorter model；
Collected initial data is pre-processed, by data normalization to critical field [- 1,1]；Use sliding window pair
Original continuous sound wave data carry out sliding-model control, form discrete sample；
Extract the data for the ratio that mean value, variance, first-order difference and standardized value are greater than the set value；
It selects support vector machines as sorter model, trains sorter model based on four characteristic signals of extraction, and make
The accuracy of disaggregated model is verified with random cross validation；
Using trained model, first determine whether user selects to practice；If so, training number is added in practice data
According to collection, sample is handled and extracted again；Otherwise formal to use.
3. the controllable exchange method based on air blowing as claimed in claim 2, it is characterized in that: once detecting that user is blown
Gas interaction, will send instruction to server end；Server end obtains air blowing data, carries out sliding-model control and the extraction of data,
Identification classification is carried out using trained model；Classification results are sent client by server；
According to the received identification types of client, corresponding interactive operation is executed.
4. the controllable exchange method based on air blowing as claimed in claim 2, it is characterized in that: sample frequency is 8000Hz, every
A period of time is once blown.
5. the controllable exchange method based on air blowing as claimed in claim 2, it is characterized in that: in data acquisition, according to
The duration of air blowing and intensity define three kinds of inflatable forms for interacting: it is big to blow, greater than the air blowing and extremely of some strength
Continue N seconds less；Exhale by force: greater than some strength is continuously less than N seconds；Featheriness: less than setting intensity and continuously less than N seconds.
6. the controllable exchange method based on air blowing as claimed in claim 2, it is characterized in that: the specific steps of training pattern algorithm
Include:
By pre-processing blowing action a with the training data d of feature extraction and corresponding thereto；
The training that disaggregated model SVM is carried out using training data d and air blowing type a, obtains new model m；
The model in aol server is updated using new model m timing；
Model m is returned to, the type blown for server identification.
7. the controllable exchange method based on air blowing as claimed in claim 2, it is characterized in that: select support vector machines as point
When class device model, mesh parameter optimizing function SVMcgForClass is called to come penalty parameter c and height in Optimum Classification device SVM
Parameter g in this radial basis function RBF kernel function is verified point with improving nicety of grading using random cross validation method
The accuracy of class model: the total data of acquisition is randomly divided into a training set and a test set, its ratio be 9:1；Weight
It carries out crossing over many times again testing and obtains Average Accuracy to verify the accuracy of model and validity.
8. the controllable exchange method based on air blowing as claimed in claim 2, it is characterized in that: air blowing type identification algorithm is specific
Step includes:
The threshold θ being set in advance, for judging whether user has carried out blowing action；
The volume value v of the sound generated after blowing by microphone is obtained in real time；
Judge the size of volume value v and threshold θ, if v > θ, client sends voice data d to server end, and server end obtains
Voice data d is taken, pretreatment and feature extraction operation then are carried out to d, obtain d '；Current blowing action is identified according to d ', is obtained
To air blowing type a, it is sent to client；Client executes corresponding interactive operation according to a received；Otherwise, again in real time
Obtain the volume value of the sound generated after blowing by microphone.
9. the controllable exchange method based on air blowing as claimed in claim 2, it is characterized in that: the setting value is 0.4.
10. the controllable exchange method based on air blowing as claimed in claim 2, it is characterized in that: corresponding interactive operation includes:
The broadcasting of the end PC video；
The browsing of mobile phone mobile terminal webpage；
The operation of reality-virtualizing game.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811296072.XA CN109634404A (en) | 2018-11-01 | 2018-11-01 | A kind of system and method for the controllable interactive interface based on air blowing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811296072.XA CN109634404A (en) | 2018-11-01 | 2018-11-01 | A kind of system and method for the controllable interactive interface based on air blowing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109634404A true CN109634404A (en) | 2019-04-16 |
Family
ID=66067143
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811296072.XA Pending CN109634404A (en) | 2018-11-01 | 2018-11-01 | A kind of system and method for the controllable interactive interface based on air blowing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109634404A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111897416A (en) * | 2020-06-29 | 2020-11-06 | 山东大学 | Self-adaptive blowing interaction method and system based on twin network |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TW308666B (en) * | 1996-11-04 | 1997-06-21 | Lin-Shan Lii | Intelligent Chinese voice learning system and method thereof |
| US20100079370A1 (en) * | 2008-09-30 | 2010-04-01 | Samsung Electronics Co., Ltd. | Apparatus and method for providing interactive user interface that varies according to strength of blowing |
| CN105206258A (en) * | 2015-10-19 | 2015-12-30 | 百度在线网络技术（北京）有限公司 | Generation method and device of acoustic model as well as voice synthetic method and device |
| CN105204750A (en) * | 2015-10-28 | 2015-12-30 | 百度在线网络技术（北京）有限公司 | Method and device of configuring keyboard skin |
| CN105446487A (en) * | 2015-11-30 | 2016-03-30 | 北京奇虎科技有限公司 | E-book page turning control method and apparatus |
| CN106328124A (en) * | 2016-08-24 | 2017-01-11 | 安徽咪鼠科技有限公司 | Voice recognition method based on user behavior characteristics |
| CN107564522A (en) * | 2017-09-18 | 2018-01-09 | 郑州云海信息技术有限公司 | A kind of intelligent control method and device |
| CN107928673A (en) * | 2017-11-06 | 2018-04-20 | 腾讯科技（深圳）有限公司 | Acoustic signal processing method, device, storage medium and computer equipment |
-
2018
- 2018-11-01 CN CN201811296072.XA patent/CN109634404A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TW308666B (en) * | 1996-11-04 | 1997-06-21 | Lin-Shan Lii | Intelligent Chinese voice learning system and method thereof |
| US20100079370A1 (en) * | 2008-09-30 | 2010-04-01 | Samsung Electronics Co., Ltd. | Apparatus and method for providing interactive user interface that varies according to strength of blowing |
| CN105206258A (en) * | 2015-10-19 | 2015-12-30 | 百度在线网络技术（北京）有限公司 | Generation method and device of acoustic model as well as voice synthetic method and device |
| CN105204750A (en) * | 2015-10-28 | 2015-12-30 | 百度在线网络技术（北京）有限公司 | Method and device of configuring keyboard skin |
| CN105446487A (en) * | 2015-11-30 | 2016-03-30 | 北京奇虎科技有限公司 | E-book page turning control method and apparatus |
| CN106328124A (en) * | 2016-08-24 | 2017-01-11 | 安徽咪鼠科技有限公司 | Voice recognition method based on user behavior characteristics |
| CN107564522A (en) * | 2017-09-18 | 2018-01-09 | 郑州云海信息技术有限公司 | A kind of intelligent control method and device |
| CN107928673A (en) * | 2017-11-06 | 2018-04-20 | 腾讯科技（深圳）有限公司 | Acoustic signal processing method, device, storage medium and computer equipment |
Non-Patent Citations (1)
| Title |
|---|
| 段永健: "基于时间序列与支持向量机的信号识别与预测", <<中国优秀硕士学位论文全文数据库>> * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111897416A (en) * | 2020-06-29 | 2020-11-06 | 山东大学 | Self-adaptive blowing interaction method and system based on twin network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104436596B (en) | Device and motion support method are supported in motion | |
| CN108805009A (en) | Classroom learning state monitoring method based on multimodal information fusion and system | |
| CN110705390A (en) | Body posture recognition method and device based on LSTM and storage medium | |
| KR102089002B1 (en) | Method and wearable device for providing feedback on action | |
| CN108683724A (en) | A kind of intelligence children's safety and gait health monitoring system | |
| EP3493032A1 (en) | Robot control method and companion robot | |
| CN105739688A (en) | Man-machine interaction method and device based on emotion system, and man-machine interaction system | |
| CN106462725A (en) | Systems and methods of monitoring activities at a gaming venue | |
| CN109621331A (en) | Fitness-assisting method, apparatus and storage medium, server | |
| CN110490109A (en) | A kind of online human body recovery action identification method based on monocular vision | |
| CN109902564A (en) | A kind of accident detection method based on the sparse autoencoder network of structural similarity | |
| CN110478883A (en) | A kind of body-building movement teaching and correction system and method | |
| CN108717548B (en) | Behavior recognition model updating method and system for dynamic increase of sensors | |
| CN110298309A (en) | Motion characteristic processing method, device, terminal and storage medium based on image | |
| CN113663312B (en) | Micro-inertia-based non-apparatus body-building action quality evaluation method | |
| CN106073793B (en) | Attitude Tracking and recognition methods based on micro-inertia sensor | |
| WO2022121769A1 (en) | Training partner robot control method, system and device, and storage medium | |
| Jin et al. | Virtual personal trainer via the kinect sensor | |
| CN110471529A (en) | Act methods of marking and device | |
| CN107943276A (en) | Based on the human body behavioral value of big data platform and early warning | |
| CN109784178A (en) | Parameter adjusting method, device and body-building equipment based on gesture identification | |
| Kang et al. | The gesture recognition technology based on IMU sensor for personal active spinning | |
| CN111931568A (en) | Human body falling detection method and system based on enhanced learning | |
| Ning et al. | Fall detection algorithm based on gradient boosting decision tree | |
| CN109634404A (en) | A kind of system and method for the controllable interactive interface based on air blowing |
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 |
Application publication date: 20190416 |
|
| RJ01 | Rejection of invention patent application after publication |