CN109858958A - Aim client orientation method, apparatus, equipment and storage medium based on micro- expression - Google Patents
Aim client orientation method, apparatus, equipment and storage medium based on micro- expression Download PDFInfo
- Publication number
- CN109858958A CN109858958A CN201910043485.5A CN201910043485A CN109858958A CN 109858958 A CN109858958 A CN 109858958A CN 201910043485 A CN201910043485 A CN 201910043485A CN 109858958 A CN109858958 A CN 109858958A
- Authority
- CN
- China
- Prior art keywords
- micro
- expression
- target
- images
- recognized
- 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
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000003860 storage Methods 0.000 title claims abstract description 19
- 230000001815 facial effect Effects 0.000 claims description 108
- 230000036651 mood Effects 0.000 claims description 61
- 238000004590 computer program Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 8
- 230000010365 information processing Effects 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000004807 localization Effects 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 description 150
- 238000004422 calculation algorithm Methods 0.000 description 13
- 230000006870 function Effects 0.000 description 9
- 238000004458 analytical method Methods 0.000 description 8
- 210000004709 eyebrow Anatomy 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 238000012549 training Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000010195 expression analysis Methods 0.000 description 5
- 238000013527 convolutional neural network Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 230000001276 controlling effect Effects 0.000 description 3
- 230000008092 positive effect Effects 0.000 description 3
- 230000001360 synchronised effect Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 230000002996 emotional effect Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 239000012491 analyte Substances 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 210000000744 eyelid Anatomy 0.000 description 1
- 210000004209 hair Anatomy 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000007935 neutral effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The present invention discloses a kind of aim client orientation method, apparatus, equipment and storage medium based on micro- expression.This method comprises: obtaining product recommendations request, product recommendations request includes target product ID；Product information database is inquired based on target product ID, obtains target product voice data and target product text data；Displaying target product text data simultaneously plays target product voice data, while starting the video data of camera shooting potential customers, and video data includes an at least frame images to be recognized；Each images to be recognized is identified using micro- Expression Recognition model, obtains corresponding micro- expression type and related coefficient；Micro- expression type and related coefficient are calculated using investment willingness degree formula, obtain target investment willingness degree；If target investment willingness degree is greater than default wish degree threshold value, potential customers are determined as target customer.This method, which can be realized, is accurately positioned target customer, helps to improve the promotion success rate of target product.
Description
Technical field
The present invention relates to micro- Expression Recognition technical field more particularly to a kind of aim client orientation sides based on micro- expression
Method, device, equipment and storage medium.
Background technique
In the business hall or shops of financial institution or other mechanisms, the production of shops's offer is mainly utilized by business personnel
The corresponding terminating machine of product recommendation system (such as self aid integrated machine) or IPAD mobile terminal, to the production of present customers institute promotion
Product, and manually inquire client's intention, client is judged to the degree of recognition of product, to determine visitor according to the dialogue of business personnel and client
The purchase intention at family, by the stronger target customer of purchase intention.The mistake of the purchase intention of current business personnel subjective judgement client
Journey, because can not objective determining client to the degree of recognition of product, make it that can not accurately determine the purchase intention of client, so that target is objective
Family position inaccurate is not high so as to cause the success rate of product recommendations.
Summary of the invention
The embodiment of the present invention provides a kind of aim client orientation method, apparatus, equipment and storage medium based on micro- expression,
To solve currently to position target customer's inaccuracy according to the purchase intention of artificial judgment client, so as to cause product recommendations success rate
Lower problem.
A kind of aim client orientation method based on micro- expression, comprising:
The product recommendations request that promotion terminal is sent is obtained, the product recommendations request includes target product ID；
Product information database is inquired based on the target product ID, obtains target corresponding with the target product ID
Product voice data and target product text data；
The promotion terminal is controlled to show the target product text data and play the target product voice data
Meanwhile the video data of the potential customers of the camera captured in real-time of the promotion terminal is obtained, the video data includes extremely
A few frame images to be recognized；
Each images to be recognized is identified using micro- Expression Recognition model, it is corresponding to obtain the images to be recognized
Micro- expression type and related coefficient；
It is calculated, is obtained using micro- expression type and related coefficient of the investment willingness degree formula to the images to be recognized
Target investment willingness degree；
If the target investment willingness degree is greater than default wish degree threshold value, the potential customers are determined as target visitor
Target image is chosen from the corresponding images to be recognized of the target customer in family；
Target customer's information is obtained based on the target investment willingness degree and the target image, and by the target customer
Information is sent to user terminal.
A kind of aim client orientation device based on micro- expression, comprising:
Product recommendations request module, for obtaining the product recommendations request of promotion terminal transmission, the product recommendations
Request includes target product ID；
Target product data acquisition module, for based on the target product ID inquire product information database, obtain with
The corresponding target product voice data of the target product ID and target product text data；
Video data obtains module, shows the target product text data for controlling the promotion terminal and plays institute
While stating target product voice data, the video counts of the potential customers of the camera captured in real-time of the promotion terminal are obtained
According to the video data includes an at least frame images to be recognized；
Micro- Expression Recognition module is obtained for being identified using micro- Expression Recognition model to each images to be recognized
Take the corresponding micro- expression type of the images to be recognized and related coefficient；
Investment willingness degree obtains module, for micro- expression type using investment willingness degree formula to the images to be recognized
It is calculated with related coefficient, obtains target investment willingness degree；
Target customer's determining module will be described if being greater than default wish degree threshold value for the target investment willingness degree
Potential customers are determined as target customer, choose target image from the corresponding images to be recognized of the target customer；
Customer information processing module, for obtaining target customer based on the target investment willingness degree and the target image
Information, and target customer's information is sent to user terminal.
A kind of computer equipment, including memory, processor and storage are in the memory and can be in the processing
The computer program run on device, the processor realize the above-mentioned target visitor based on micro- expression when executing the computer program
The step of family localization method.
A kind of computer readable storage medium, the computer-readable recording medium storage have computer program, the meter
Calculation machine program realizes the step of above-mentioned aim client orientation method based on micro- expression when being executed by processor.
Aim client orientation method, apparatus, equipment and storage medium based on micro- expression, are requested according to product recommendations
Target product ID, the controllable autonomous displaying target product text data of promotion terminal simultaneously plays target product voice data, with reality
Medium intelligent introduction now is carried out to target product, to save manpower explanation cost.Also, in displaying target product text data and broadcasting
While target product voice data, acquire the video data of potential customers, by the images to be recognized in video data into
The micro- Expression Recognition of row obtains corresponding micro- expression type and related coefficient, using investment willingness degree formula to micro- expression type and
Related coefficient is calculated, can quick obtaining target investment willingness degree, and acquired target investment willingness degree objectivity compared with
By force, it can intuitively reflect the investment willingness of potential customers.When target investment willingness degree is greater than default wish degree threshold value, target is determined
Client and corresponding target image form target customer's information based on target investment willingness degree and target image, so that business people
Member carries out one-to-one service to target customer according to target customer's information, improves the promotion success rate of target product.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below by institute in the description to the embodiment of the present invention
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention
Example, for those of ordinary skill in the art, without any creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is an application environment schematic diagram of the aim client orientation method in one embodiment of the invention based on micro- expression；
Fig. 2 is a flow chart of the aim client orientation method in one embodiment of the invention based on micro- expression；
Fig. 3 is another flow chart of the aim client orientation method in one embodiment of the invention based on micro- expression；
Fig. 4 is another flow chart of the aim client orientation method in one embodiment of the invention based on micro- expression；
Fig. 5 is another flow chart of the aim client orientation method in one embodiment of the invention based on micro- expression；
Fig. 6 is another flow chart of the aim client orientation method in one embodiment of the invention based on micro- expression；
Fig. 7 is another flow chart of the aim client orientation method in one embodiment of the invention based on micro- expression；
Fig. 8 is a schematic diagram of the aim client orientation device in one embodiment of the invention based on micro- expression；
Fig. 9 is a schematic diagram of computer equipment in one embodiment of the invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are some of the embodiments of the present invention, instead of all the embodiments.Based on this hair
Embodiment in bright, every other implementation obtained by those of ordinary skill in the art without making creative efforts
Example, shall fall within the protection scope of the present invention.
Aim client orientation method provided in an embodiment of the present invention based on micro- expression, should the target customer based on micro- expression
Localization method can be using in application environment as shown in Figure 1.It specifically, should the aim client orientation method application based on micro- expression
In product recommendations system, which includes server as shown in Figure 1, promotion terminal and user terminal, promotion
Terminal and user terminal are communicated with server by network, for realizing the video counts for acquiring client by promotion terminal
According to the micro- Expression Recognition of video data progress, to determine target investment meaning according to the micro- expression type and related coefficient that identify
Hope degree determines target customer according to target investment willingness degree, to realize the accurate positioning to target customer, and then improves target and produces
The promotion success rate of product.Wherein, promotion terminal and user terminal are client, refer to corresponding with server, are mentioned for client
For the program of local service.Client it is mountable but be not limited to various personal computers, laptop, smart phone, flat
On plate computer and portable wearable device.Server can use the service of the either multiple server compositions of independent server
Device cluster is realized.In the present embodiment, promotion terminal be arranged in business hall or shops for consulting product for client
The terminal of information, user terminal are the terminals used for the business personnel of business hall or shops, can be business personnel's use
Mobile phone or the mobile terminals such as IPAD.
In one embodiment, as shown in Fig. 2, providing a kind of aim client orientation method based on micro- expression, in this way
It applies and is illustrated for the server in Fig. 1, include the following steps:
S201: the product recommendations request that promotion terminal is sent is obtained, product recommendations request includes target product ID.
Wherein, promotion terminal is that the terminal for carrying out self-service query for client being arranged in business hall or shops is set
It is standby, for example, the self-service query equipment being arranged in business hall or shops.The display screen of the promotion terminal is touching display screen, can
All product informations of this business hall or shops are shown on display interface by the promotion terminal, so that client understands this business
The relevant information of product provided by the Room or shops.Target product ID is the mark for unique identification target product, the target
When product is that client passes through the touching display screen in operation promotion terminal, identified product when a certain product is checked in click.Tool
Body, select any product shown by the touching display screen of promotion terminal to determine target product ID when client clicks, so that
Promotion terminal to server sends product recommendations request.
S202: inquiring product information database based on target product ID, obtains target corresponding with target product ID and produces
Product voice data and target product text data.
Wherein, product information database is the data for storing the product information of all products in this business hall or shops
Library.Each product information in product information database includes product IDs, product voice data and product text data.The product
ID is the mark for a certain product of unique identification.Product voice data is prerecorded for carrying out voice introduction to product
Voice data.Product text data is pre-entered for carrying out the text data of text presentation to product.It is understood that
Ground, the product voice data is associated with product text data, and product is introduced from two levels of voice and text respectively,
So that client understands certain product in more detail.
Specifically, server inquires product information database based on the target product ID in product recommendations request, from product
Target product information corresponding with target product ID is inquired and obtained in information database, which includes target
Product voice data and target product text data.Wherein, it can be used during server inquiry target product information simple
Query sentence of database, can quick obtaining target product information, so as to based on target product voice data and target product text
Notebook data carries out product introduction, when the response that server requests client by the product recommendations that promotion terminal triggers can be improved
Between, help to improve customer satisfaction.
S203: it while controlling promotion terminal displaying target product text data and play target product voice data, obtains
The video data of the potential customers of the camera captured in real-time of promotion terminal is taken, video data includes at least frame figure to be identified
Picture.
Specifically, the promotion terminal being arranged in business hall or shops includes display screen, player and camera.The display
Screen is touching display screen, so that client can carry out human-computer interaction by the touch screen display screen and product recommendations system.Player is
Equipment for playing voice data, camera are the equipment for acquiring video data.Potential customers refer to positioned at the promotion
Client in the coverage of the camera of terminal.In general, the client in the coverage of the push of promotion terminal
It can be 1, or multiple, then its potential customers is at least one.It is to be appreciated that if a client is shown by touch-control
Screen triggers the corresponding product recommendations request of a certain target product, then the client in the coverage of the camera of promotion terminal
For the corresponding potential customers of the target product.
Specifically, server, can be at promotion end after obtaining target product voice data and target product text data
The display screen display at end target product text data, so that clear intuitive understanding of the potential customers to target product.?
While displaying target product text data, which is played by the player voice of promotion terminal, with
Combining target product text data carries out phonetic explaining to the target product, so that potential customers become apparent from directly target product
The understanding of sight.Also, playing target product voice data simultaneously, the camera of triggering starting promotion terminal, using camera
The video data of the potential customers in the coverage for the camera for being currently at promotion terminal is shot, will pass through to the video
Data are analyzed, to determine potential customers to the target investment willingness degree of target product.
S204: identifying each images to be recognized using micro- Expression Recognition model, and it is corresponding to obtain images to be recognized
Micro- expression type and related coefficient.
Wherein, micro- Expression Recognition model is the model of the micro- expression of face in images to be recognized for identification.In the present embodiment,
Micro- Expression Recognition model be by capture images to be recognized in user face local feature, and according to local feature determine to
The each target face motor unit for identifying face in image, determines its feelings further according to the target face motor unit identified
The model of thread.Micro- expression type is the micro- expression of face determined after being identified using micro- Expression Recognition model to images to be recognized
Type, images to be recognized is identified specifically by micro- Expression Recognition model, is acted according to the target face that identifies
Micro- expression that unit determines.
In the present embodiment, micro- Expression Recognition model can be the neural network recognization model based on deep learning, can also be with
It is the local identification model based on classification, can also be based on local binary patterns (Local Binary Pattern, LBP)
Local Emotion identification model.In the present embodiment, micro- Expression Recognition model is the local identification model based on classification, micro- Expression Recognition
It include each face in training image data by collecting a large amount of training image data in advance when model is trained in advance
The positive sample of motor unit and the negative sample of Facial action unit are trained training image data by sorting algorithm, obtain
Take micro- Expression Recognition model.In the present embodiment, it can be and a large amount of training image data are instructed by svm classifier algorithm
Practice, to get SVM classifier corresponding with N number of Facial action unit.For example, it may be 39 Facial action units are corresponding
39 SVM classifiers are also possible to corresponding 54 SVM classifiers of 54 Facial action units, the training image being trained
The positive sample and negative sample for the different Facial action units for including in data are more, then the SVM classifier quantity got is more.
It is to be appreciated that the SVM classifier got is more, then shape by N number of SVM classifier to form micro- Expression Recognition model
At micro- expression for identifying of micro- Expression Recognition model it is more accurate.With the corresponding SVM classifier institute of 54 Facial action units
For the micro- Expression Recognition model formed, 54 kinds of micro- expression types may recognize that using this micro- Expression Recognition model, such as can
Identify to include to like, is interested, pleasantly surprised, expecting ... 54 kinds of micro- expression types such as aggressive, conflict, insult, suspection and fear.
Related coefficient refers to that the corresponding micro- expression type of the images to be recognized and potential customers anticipate to the investment of target product
The degree of correlation of hope.The related coefficient specifically uses Pearson correlation coefficient (Pearson Correlation
Coefficient) it is used to measure two datasets conjunction whether on one wire face, it is used to measure the linear pass between spacing variable
System.In the present embodiment, the value of related coefficient is in [- 1,1], and the absolute value of related coefficient is bigger, and degree of correlation is stronger；It is i.e. related
For coefficient closer to 1 or -1, degree of correlation is stronger；For related coefficient closer to 0, degree of correlation is weaker.Specifically, related coefficient
Value be negatively correlated when being negative, be positive correlation when value is positive number.In general, using the value range of related coefficient
Absolute value is come the correlation intensity that determines: 1) 0.8-1.0, extremely strong correlation；2) 0.6-0.8, strong correlation；3) 0.4-0.6, it is moderate
It is related；4) 0.2-0.4, weak correlation；5) 0.0-0.2 is extremely weak related or without correlation.
Specifically, the corresponding phase relation of all micro- expression types that server in advance identifies micro- Expression Recognition model
Number is determined as 1,0.8,0.5,0.3,0, -0.3, -0.5, -0.8 or -1 equivalence, the corresponding phase relation of each micro- expression type
Number.In the present embodiment, 54 kinds of micro- expression types that a certain micro- Expression Recognition model is identified, such as like, is interested, pleasantly surprised,
Expect ... aggressive, conflict, insult, suspection and fear etc., related coefficient is divided according to the fancy grade of user.Example
Such as, set 1 for love and the related coefficients of the two micro- expression types interested, by it is pleasantly surprised, expect and trust these three micro- tables
The related coefficient of feelings type is set as 0.8, by happy, optimistic, admiration, phase that is grateful, sincere and receiving this expression type slightly
Relationship number is set as 0.5, sets 0.3 for the related coefficient of micro- expression type such as vigor, bravery and worry, by it is sad, shy,
The related coefficient of micro- expression type such as arrogant and quiet is set as 0, will detest and the related coefficient of micro- expression type such as disagreeable is set
It is set to -1, will oppose and the related coefficient of micro- expression type such as discontented is set as -0.8, will ignore, indignation and micro- expression such as despise
The related coefficient of type is set as -0.5, sets -0.3 etc. for the related coefficient of micro- expression type such as suspection, insult and fear.
Specifically, after being identified using micro- Expression Recognition model to images to be recognized, if it is corresponding micro- to recognize images to be recognized
Expression type is love or interested, it is determined that its corresponding related coefficient is 1.
S205: it is calculated, is obtained using micro- expression type and related coefficient of the investment willingness degree formula to images to be recognized
Take target investment willingness degree.
Wherein, investment willingness degree formula is the formula for calculating investment willingness degree.Target investment willingness degree is using throwing
Money wish degree formula micro- expression type corresponding to images to be recognized and related coefficient carry out calculating acquired investment willingness degree.
The target investment willingness degree can objectively respond potential customers to the investment willingness degree of target product.It is to be appreciated that investment willingness
Degree is higher, indicates that potential customers are more possible to investment and buy the target product in other words, therefore, can be based on target investment willingness degree
Determine target customer.
In one embodiment, investment willingness degree formula isWherein, R is target investment willingness degree, piFor
I-th of related coefficient, qiFor the corresponding expression number of attributes of i-th of related coefficient, S is expression total quantity.In step S502, adopt
It is calculated with micro- expression type and related coefficient of the investment willingness degree formula to images to be recognized, obtains target investment willingness
Degree, specifically comprises the following steps:
(1) quantity for counting the corresponding all micro- expression types of same related coefficient, determines that each related coefficient is corresponding
Expression number of attributes calculates expression total quantity according to the expression number of attributes of each related coefficient.
Wherein, expression number of attributes is the quantity of the corresponding all micro- expression types of same related coefficient.Expression total quantity
Refer to the corresponding expression number of attributes of all related coefficients and value, specifically it can be usedIt is calculated, wherein S is
Expression total quantity, qiFor the corresponding expression number of attributes of i-th of related coefficient.Since images to be recognized each in video data is known
It is clipped to a kind of corresponding micro- expression type, the corresponding related coefficient of micro- expression type, and all micro- expression classes in the present embodiment
The related coefficient of type is specially 1,0.8,0.5,0.3,0, -0.3, -0.5, -0.8 or -1 equivalence, and each related coefficient is corresponding extremely
A few micro- expression type.It, can be by counting the corresponding all micro- expression types of same related coefficient in order to simplify calculating process
Quantity, determine the corresponding expression number of attributes of related coefficient.In above-described embodiment, identified using micro- Expression Recognition model
" love ", " pleasantly surprised ", " expectation ", " bravery " quantity be respectively 25,20,40 and 15, then count same related coefficient
When the quantity of corresponding all micro- expression types, since micro- expression type that related coefficient is 1 includes " love ", then related coefficient 1
Corresponding expression number of attributes is 25；It is since micro- expression type that related coefficient is 0.8 includes " pleasantly surprised " and " expectation ", then related
The corresponding expression number of attributes of coefficient 0.8 is 20+40=60, since micro- expression type that related coefficient is 0.3 includes " brave
Dare ", then the corresponding expression number of attributes of related coefficient 0.3 is 15.It can be usedIt is calculated, acquired expression
Total quantity is S=25+20+40+15=100.
(2) related coefficient and the corresponding expression number of attributes of related coefficient are calculated using investment willingness degree formula,
Obtain target investment willingness degree.
Wherein, investment willingness degree formula isWherein, R is target investment willingness degree, piFor i-th of correlation
Coefficient, qiFor the corresponding expression number of attributes of i-th of related coefficient, S is expression total quantity.piCan for 1,0.8,0.5,0.3,
0, i-th of related coefficient in -0.3, -0.5, -0.8 and -1, and qiIt is to belong to piThe corresponding all micro- expression classes of this related coefficient
The expression number of attributes of type.
In the present embodiment, investment willingness degree formula can be converted into R=(1*q1+0.8*q2+0.5*q3+0.3*q4-0.3*
Q6-0.5*q7-0.6*q8-1*q9)/S, wherein 1,0.8 ... -1 is related coefficient, q1, q2 ... q9 are respectively corresponding phase
The expression number of attributes of the corresponding micro- expression type of relationship number, S q1, q2 ... q9's and value.The expression number of attributes can be with
It is interpreted as the quantity of the corresponding all micro- expression types of a certain related coefficient.In the present embodiment, acquired target investment willingness
For the value range of degree in [- 1,1], numerical value is bigger, indicates that potential customers are bigger to the degree of recognition of target product, investment willingness is got over
Greatly, so as to according to the target investment willingness degree, degree of recognition or investment willingness of the objective understanding potential customers to target product.
In the present embodiment, its corresponding expression attribute first is determined based on the corresponding all micro- expression types of same related coefficient
Quantity, and expression total quantity is determined according to expression number of attributes, it is subsequent using acquired in the calculating of investment willingness degree formula to guarantee
Target investment willingness degree accuracy, avoid influencing because not including facial image or other disturbing factors in images to be recognized
Expression total quantity.Also, related coefficient, expression number of attributes and expression total quantity are counted using investment willingness degree formula
Calculate, can quick obtaining target investment willingness degree, and acquired target investment willingness degree objectivity is stronger, can intuitively reflect
The investment willingness of potential customers.
S206: if target investment willingness degree is greater than default wish degree threshold value, being determined as target customer for potential customers, from
Target image is chosen in the corresponding images to be recognized of target customer.
Wherein, presetting wish degree threshold value, to be that server is pre-set regard as target customer for assessing whether to reach
Wish degree threshold value.In the present embodiment, if the target investment willingness degree that server recognizes a certain potential customers is greater than default wish
Threshold value is spent, then illustrates that the potential customers are higher to the degree of recognition of target product, investment willingness is higher, it is most likely that invests the target
Product is accurately positioned target customer at this point, the potential customers are determined as target customer to realize, more to target customer's promotion
The target product for meeting its investment willingness improves the promotion success rate of target product.Since the determination of target customer is by right
The video data of promotion terminal acquisition carries out analysis determination, in order to make the business personnel of business hall or shops recognize target visitor
Family, server can also choose a target image from the images to be recognized for identify target customer, so as to by the target image
It is sent to user terminal, so that business personnel understands the feature of target customer by user terminal, is provided more for target customer
Detailed or professional service improves the promotion success rate of target product.
S207: target customer's information is obtained based on target investment willingness degree and target image, and target customer's information is sent out
Give user terminal.
User terminal refers to the mobile terminal that the business personnel in this business hall or shops uses.Specifically, server exists
After determining target customer, one is chosen from least one corresponding images to be recognized of target customer and is used as target image, base
Target customer's information is formed in the target investment willingness and target image, and target customer's information is sent to user terminal,
So that the business personnel in this business hall or shops is accurately positioned to target customer, so that one-to-one service is provided, to throw its institute
It is good, improve the promotion success rate of target product.
Aim client orientation method based on micro- expression provided by the present embodiment, according to product recommendations request in target
Product IDs, the controllable autonomous displaying target product text data of promotion terminal simultaneously plays target product voice data, with realization pair
Target product carries out medium intelligent introduction, to save manpower explanation cost.Also, in displaying target product text data and play target
While product voice data, the video data of potential customers is acquired, it is micro- by being carried out to the images to be recognized in video data
Expression Recognition obtains corresponding micro- expression type and related coefficient, using investment willingness degree formula to micro- expression type and correlation
Coefficient is calculated, can quick obtaining target investment willingness degree, and acquired target investment willingness degree objectivity is stronger, can
The investment willingness of intuitive reflection potential customers.When target investment willingness degree is greater than default wish degree threshold value, target customer is determined
With corresponding target image, target customer's information is formed based on target investment willingness degree and target image, so that business personnel's root
One-to-one service is carried out to target customer according to target customer's information, improves the promotion success rate of target product.
In one embodiment, it as shown in figure 3, being identified using micro- Expression Recognition model to each images to be recognized, obtains
Take the corresponding micro- expression type of images to be recognized and related coefficient, comprising:
S301: identifying each images to be recognized using micro- Expression Recognition model, obtains opposite with images to be recognized
The micro- expression type answered.
Specifically, server first identifies each images to be recognized using micro- Expression Recognition model, and obtaining should be wait know
The micro- expression of face belongs to the corresponding instantaneous confidence level of each micro- expression type in other image, that is, belongs to each micro- expression type
Probability.Then, the maximum micro- expression type of instantaneous confidence level is chosen, the corresponding micro- expression type of images to be recognized is determined as,
The instantaneous maximum micro- expression type of confidence level is obtained identify to images to be recognized using micro- Expression Recognition model
Final result.For example, recognizing the images to be recognized when being identified to a frame images to be recognized and belonging to " love " this micro- expression
The instantaneous confidence level of type is 0.9, and the instantaneous confidence level for belonging to " suspection " and " quiet " the two micro- expression types is respectively
0.05, then instantaneous confidence level is determined as to micro- expression type of the images to be recognized for 0.9 corresponding micro- expression type.It can manage
Xie Di, due to the corresponding specific micro- expressive features of each micro- expression type, in identification process, if not deposited in images to be recognized
In specific micro- expressive features, then identify that the probability (i.e. instantaneous confidence level) for belonging to certain micro- expression type is 0, at subsequent point
Without considering during analysis.
S302: being based on micro- expression type queries related coefficient conversion table, obtains phase relation corresponding with micro- expression type
Number.
Wherein, the tables of data of the corresponding related coefficient of each micro- expression type is stored in related coefficient conversion table.?
In related coefficient conversion table, all micro- expression types are divided into positive strong correlation to the correlativity of investment willingness degree according to it
Classification, passive strong correlation classification and the weak related category between positive strong correlation classification and passive strong correlation classification.Its
In, positive strong correlation classification, which refers to, to have a positive effect to investment willingness degree and strong associations, such as like, happily, it is optimistic, trust,
Micro- expression type such as acceptable and pleasantly surprised can reflect the micro- expression having a positive effect to investment willingness degree, can anticipate according to it to investment
Its related coefficient is arranged in [0.5,1] in the degree that hope degree has a positive effect, that is, may be configured as 0.5,0.8 and 1.Passive strong correlation class
Do not refer to and negative consequence and strong associations are risen to investment willingness degree, such as detests, dislikes, opposes, is discontented with, ignores and despises micro- table
Feelings type can reflect micro- expression that negative consequence is played to investment willingness degree, can play the journey of negative consequence to investment willingness degree according to it
Its related coefficient is arranged in [- 1, -0.5] in degree, that is, may be configured as -1, -0.8 and -0.5.Weak related category refers to investment willingness
The neutral associations for having spent general action are all micro- expression classes in addition to positive strong correlation classification and passive strong correlation classification
The set that type is formed, related coefficient specifically may be configured as -0.3,0 and 0.3 in (- 0.5,0.5).
Aim client orientation method based on micro- expression provided by the present embodiment is first treated using micro- Expression Recognition model
Identify that image is identified, its corresponding micro- expression type with quick obtaining, then tabled look-up based on micro- expression type, it can quickly obtain
Its corresponding related coefficient is taken, to ensure micro- expression type of images to be recognized and the acquisition efficiency of related coefficient.
In one embodiment, it as shown in figure 4, being identified using micro- Expression Recognition model to each images to be recognized, obtains
Take micro- expression type corresponding with images to be recognized, comprising:
S401: identifying each images to be recognized using micro- Expression Recognition model, obtains at least one identification expression
The corresponding instantaneous confidence level of type.
Wherein, it when identification expression type refers to that the micro- Expression Recognition model of use identifies images to be recognized, recognizes
It belongs to the micro- expression type of preconfigured a certain kind.
It specifically, include N number of SVM classifier, each svm classifier in the preparatory trained micro- Expression Recognition model of server
A kind of device Facial action unit for identification.In the present embodiment, includes 54 SVM classifiers in micro- Expression Recognition model, establish
Facial action unit number mapping table, each Facial action unit are indicated with a prespecified number.For example, AU1 is interior
Eyebrow raises up, AU2 be outer eyebrow raise up, AU5 be upper eyelid raise up with AU26 be lower jaw open etc..Each Facial action unit has training
Good corresponding SVM classifier.For example, being raised up the local feature that corresponding SVM classifier may recognize that interior eyebrow raises up by interior eyebrow
Belong to the probability value that interior eyebrow raises up, is raised up the local feature category that corresponding SVM classifier may recognize that outer eyebrow raises up by outer eyebrow
In the probability value etc. that outer eyebrow raises up.
In the present embodiment, server identifies images to be recognized using preparatory trained micro- Expression Recognition model
When, face critical point detection and feature extraction etc. first can be carried out to each images to be recognized, to obtain the part of images to be recognized
Feature.Wherein, face key point algorithm can be but not limited to Ensemble of Regression Tress (abbreviation ERT) calculation
Method, SIFT (scale-invariant feature transform) algorithm, SURF (Speeded Up Robust
Features) algorithm, LBP (Local Binary Patterns) algorithm and HOG (Histogram of Oriented
Gridients) algorithm.Feature extraction algorithm can be calculated with CNN (Convolutional Neural Network, convolutional Neural net)
Method.The local feature is input in N number of SVM classifier again, passes through all local features of the input of N number of SVM classifier pair
It is identified, obtains the probability value corresponding with the Facial action unit of N number of SVM classifier output, probability value is greater than default
The corresponding Facial action unit of the SVM classifier of threshold value is determined as target face motor unit.Wherein, target face motor unit
Refer to and images to be recognized is identified according to micro- Expression Recognition model, the Facial action unit that gets (Action Unit,
AU).Probability value specifically can be the value between 0-1, if the probability value of output is 0.6, preset threshold 0.5, then probability value
0.6 is greater than preset threshold 0.5, then by 0.6 corresponding Facial action unit, the target face as images to be recognized acts single
Member.Finally, accessed all target face motor units are carried out comprehensive assessment, obtains it and belong to micro- Expression Recognition model
The corresponding probability of preconfigured micro- expression type belongs to instantaneous confidence level of each identification expression type.It will be acquired
To all target face motor units carry out comprehensive assessment specifically refer to obtain based on the combination of all target face motor units
This combination is taken to belong to the probability of preconfigured micro- expression type, to determine that it identifies the instantaneous confidence level of expression type.
S402: the maximum identification expression type of instantaneous confidence level is determined as the corresponding micro- expression type of images to be recognized.
Specifically, recognize each images to be recognized belong to it is at least one identification expression type instantaneous confidence level it
Afterwards, the maximum identification expression type of instantaneous confidence level need to be determined as to the corresponding micro- expression type of images to be recognized.For example, knowing
Being clipped to its images to be recognized to belong to the instantaneous confidence level of " love " this identification expression type is 0.9, and is belonged to " suspection " and " peaceful
It is quiet " the two identification expression types instantaneous confidence level be respectively 0.05, then by instantaneous confidence level be 0.9 corresponding identification expression
Type is determined as micro- expression type of the images to be recognized.
Aim client orientation method based on micro- expression provided by the present embodiment, first using micro- Expression Recognition model to every
One images to be recognized is identified, to determine that its corresponding at least one identifies the corresponding instantaneous confidence level of expression type, and is selected
The maximum identification expression type of instantaneous confidence level is taken to be determined as its corresponding micro- expression type, to guarantee the micro- expression identified
The accuracy of type.
In one embodiment, since the potential customers of the coverage of the camera in promotion terminal may have one,
May also have it is multiple, it is each potential to judge respectively therefore, it is necessary to which each potential customers are carried out with micro- Expression Recognition processing respectively
Whether client is target customer.As shown in figure 5, carrying out micro- Expression analysis to multiple potential customers to realize, micro- expression is being used
Identification model identifies each images to be recognized, obtains the step of images to be recognized corresponding micro- expression type and related coefficient
Before rapid, the aim client orientation method based on micro- expression further include:
S501: images to be recognized is identified using human face recognition model, obtains original facial image.
Specifically, server is provided with human face recognition model in advance, using the human face recognition model to images to be recognized into
Row identification, filters out original facial image, from images to be recognized to exclude the interference of inhuman face image.The original facial image
It is the image for recognizing the human face region comprising potential customers.
S502: judge original facial image with the presence or absence of corresponding benchmark face image.
Wherein, benchmark face image is created based on each potential customers for the face figure as reference standard
Picture.It is to be appreciated that the corresponding benchmark face image of each potential customers, it is corresponding that server can will recognize each potential customers
First more visible original facial image as benchmark face image, and be based on the corresponding people of the benchmark face image creation
Face image database, to store the corresponding all original facial images of same potential customers.
Specifically, server mentions original facial image and benchmark face image progress feature using feature extraction algorithm
It takes, obtains original face characteristic and benchmark face feature respectively；Original face characteristic and benchmark are compared using similarity algorithm again
The characteristic similarity of face characteristic；If characteristic similarity is greater than similarity threshold, illustrate original face characteristic and benchmark face
Feature corresponds to the face characteristic of same potential customers, then there are corresponding benchmark face images for original facial image；Conversely, if special
It levies similarity and is not more than similarity threshold, then illustrate that original face characteristic and benchmark face feature do not correspond to same potential customers'
Face characteristic, then corresponding benchmark face image is not present in original facial image.Wherein, similarity threshold is pre-set use
Whether reach the threshold value for being determined as the face characteristic of same potential customers in assessment similarity.In the present embodiment, feature extraction is calculated
Method includes but is not limited to CNN (Convolutional Neural Network, convolutional Neural net) algorithm.Similarity algorithm packet
It includes but is not limited to cosine similarity algorithm.
S503: original facial image is then stored in corresponding with benchmark face image by benchmark face image if it exists
In face image database.
Specifically, if server determines that original facial image there are corresponding benchmark face image, illustrates to have been based on this
The corresponding face image database of benchmark face image creation of potential customers, at this point, can directly store original facial image
It is corresponding to same potential customers all original to realize in face image database corresponding with benchmark face image
Facial image is managed collectively.
S504: benchmark face image if it does not exist, then using original facial image as benchmark face image, creation and benchmark
The corresponding face image database of facial image.
Specifically, if server determines that corresponding benchmark face image is not present in original facial image, illustrate not create
Face image database corresponding with potential customers, at this point, can using the original facial image as a benchmark face image, with
Face image database corresponding with the benchmark face image is created, it is corresponding potential with the benchmark face image for storing
All original facial images of client.It is to be appreciated that will record when creating face image database and create the facial image
System time when creating the face image database is determined as its corresponding creation time by the creation time of database.
S505: real-time statistics are in the preset time period from the creation time of face image database, face image data
The amount of images of original facial image in library.
Wherein, preset time period is the pre-set period, which can be passed through user terminal by business personnel certainly
Main setting, such as may be configured as 2 minutes or other times section.Specifically, from server is when creating face image database,
Meeting real-time statistics are in the preset time period from the creation time of the face image database, the original in the face image database
The amount of images of beginning facial image, to be determined the need for based on the image data to original in the face image database
Facial image carries out micro- Expression analysis.
S506: it if amount of images is greater than preset quantity threshold value, executes using micro- Expression Recognition model to each to be identified
The step of image is identified, obtains the corresponding micro- expression type of images to be recognized and related coefficient.
Preset quantity threshold value is the pre-set amount threshold for assessing whether to need to be analyzed and processed.Specifically
Ground illustrates the people if the amount of images of the original facial image in any face image database is greater than preset quantity threshold value
It is not accidentally to pass through in the coverage for the camera that face image database corresponding potential customers' long period is in promotion terminal
It crosses, each images to be recognized is identified using micro- Expression Recognition model at this point, server is executed, obtain images to be recognized pair
The step of micro- expression type and related coefficient for answering, i.e. execution step S204, it is possible to understand that ground, images to be recognized herein are specific
For original facial image, to realize the potential customers of the coverage of camera that the long period is in promotion terminal as micro-
The analysis object of Expression analysis saves analysis cost and improves analysis efficiency.
S507: if amount of images is not more than preset quantity threshold value, face image database is deleted.
Specifically, if the amount of images of the original facial image in any face image database is not more than preset quantity threshold
Value, then illustrate that the face image database corresponding potential customers' long period is in the coverage of the camera of promotion terminal
It is interior, it is most likely that collected corresponding original facial image when being the coverage accidentally Jing Guo the promotion terminal, at this point, deleting
Except face image database, no longer the original face image database is monitored, to save memory space, without right to its
The potential customers answered carry out micro- Expression analysis, to save analysis cost.
Aim client orientation method based on micro- expression provided by the present embodiment first treats knowledge using human face recognition model
Other image is identified, original facial image is obtained, to exclude the interference of inhuman face image.Original facial image is judged again whether
There are corresponding benchmark face image, to determine the need for creating benchmark people to the corresponding potential customers of the original facial image
Face image and corresponding face image database, will pass through the face image database primitive man all to a certain potential customers
Face image carries out integrated management.The amount of images of original facial image is greater than preset quantity threshold in a certain face image database
When value, step S204 is executed；When no more than preset quantity threshold value, delete face image database, with realize only to for a long time at
Micro- Expression analysis is carried out in the potential customers in the coverage of the camera of promotion terminal, to realize saving analysis cost and mention
High analyte efficiency, and can effectively save memory space.
In one embodiment, as shown in fig. 6, after the step of potential customers are determined as target customer, it is based on micro- table
The aim client orientation method of feelings further include:
S601: from the corresponding face image database of target customer, the corresponding single frames of each original facial image is obtained
Mood value.
In embodiment as above, each potential customers corresponding one are used to store the face image database of original facial image,
And target customer is determined based on potential customers, then target customer also corresponds to a face image database, the facial image number
According to the storage original facial image of multiframe in library.Specifically, it is corresponding to target customer that micro- Expression Recognition model can be used in server
All original facial images carry out micro- Expression Recognition in face image database, corresponding micro- to obtain each original facial image
Expression type, identification process can refer to step S401-S402, to avoid repeating, not be described in detail one by one herein.Then, based on every
Micro- expression type queries mood value conversion table of one original facial image, to obtain mood corresponding with each micro- expression type
Value, is determined as the corresponding single frames mood value of original facial image for the mood value.Wherein, mood value conversion table is every for storing
A kind of tables of data of the corresponding mood value of micro- expression type.In the present embodiment, each mood value in mood value conversion table is taken
It is worth the feelings in [- 1,1], mood value is bigger, and reflection target customer is interested or more approves.
S602: according to the corresponding shooting time of original facial image, using third party's image processing tool to each original
The single frames mood value of facial image is handled, and original micro- expression trend graph is obtained.
The corresponding shooting time of original facial image refer to the camera of promotion terminal shoot the original facial image when
Between.In the present embodiment, used third party's image procossing library is hightcharts image procossing library, and built-in there are many figures
Table transfer function.Highcharts is a graphical control, generates line chart, column diagram, sector, geographical point according to some data
Butut and radar map etc., are mainly used in Web site, current wired batten, area, areaspline, column diagram, bar chart,
Pie chart and scatterplot graph type.
Specifically, shooting time of the server according to original facial image, using third party's image processing tool to each
The single frames mood value of original facial image is handled, to obtain original micro- expression trend graph, in original micro- expression trend graph,
Using the shooting time of original facial image as horizontal axis coordinate, using the single frames mood value of original facial image as ordinate of orthogonal axes, with reality
Now micro- expression shape change feelings of the reflection target customer when watching target product text data or listening to target product voice data
Condition facilitates the focus variation that business personnel understands target customer intuitively to understand the emotional change of target customer.
S603: according to the play time of target product voice data, carrying out topic point mark to original micro- expression trend graph,
Obtain the micro- expression trend graph of target.
Wherein, the play time of target product voice data refers to the target product voice that promotion terminal plays are prerecorded
The system time of data.In the recording process of target product voice data, topic point in different time periods can be predefined.Tool
Body, according to the play time of target product voice data, topic point mark is carried out to original micro- expression trend graph, is referred to mesh
The play time shooting time association corresponding with original facial image of product voice data is marked, with the original micro- expression tendency of determination
Corresponding topic point, i.e., be labeled in original micro- expression trend graph by the topic point of different time in figure, to form the micro- table of target
Feelings trend graph.In the present embodiment, by the play time of target product voice data shooting time corresponding with original facial image
Association, refers to the shooting time and target for making each original facial image (i.e. benchmark face image) in face image database
The play time of product voice data is synchronous.For example, if while playing target product voice data photographic subjects client couple
The original facial image answered, then its time synchronization；If the N second ability photographic subjects client after playing target product voice data
Corresponding original facial image, then by the play time of target product voice data after each original facial image and N seconds
It is synchronous.
For example, 0-30 seconds are to product base in the 2 minutes target product voice data recorded for an insurance products
The description of this topic point of this information, 30-60 seconds are the descriptions for being directed to this topic point of income, and 60-90 seconds are for reason
The description of this topic point of wealth attribute, and 90-120 is the description for this topic point of safety guarantee.Then in original micro- table
In feelings trend graph, the topic point of 0-30 seconds corresponding micro- expression types is determined as product essential information, 30-60 seconds correspondences
The topic point of micro- expression type be determined as income, the topic point of 60-90 seconds corresponding micro- expression types is determined as financing and belongs to
Property, the topic point of 90-120 seconds corresponding micro- expression types is determined as safety guarantee, so as in the micro- expression trend graph of target
Which topic point intuitive reflection target customer is more concerned about, and is conducive to business personnel and carries out for associated topic point to target product
It is further described, to improve the promotion success rate of target product.
Aim client orientation method based on micro- expression provided by the present embodiment, first based on each original facial image
Shooting time and single frames mood value determine original micro- expression trend graph, so that original micro- expression trend graph can be intuitive and objective anti-
Reflect emotional change situation of target customer during understanding the relevant information of target product.Broadcasting based on target speech data again
The time is put, topic point mark is carried out to original micro- expression trend graph, it is anti-with intuitive and client to obtain the micro- expression trend graph of target
Target customer is reflected to the concern situation of different topic points.
In one embodiment, as shown in fig. 7, in step S603, topic point mark is carried out to original micro- expression trend graph, is obtained
Take the micro- expression trend graph of target, comprising:
S701: topic point mark is carried out to original micro- expression trend graph, obtains 1 original topic points.
Specifically, server carries out topic point mark to original micro- expression trend graph of each target customer, original
1 original topic points are determined in micro- expression trend graph, and determine the corresponding time interval of each original topic point.As above
In embodiment, the time interval of financing this original topic point of attribute of a certain target product is 60-90 seconds.
S702: total number of image frames of the corresponding original facial image of each original topic point is counted.
Specifically, total number of image frames of the corresponding original facial image of each original topic point of server statistics, refers to system
Count the amount of images of all original facial images in the corresponding time interval of each original topic point.In embodiment as above, clothes
It is engaged in device statistics face image database, to all original face figures of the shooting time in every 60-90 seconds this time interval
The quantity of picture is counted, to obtain total number of image frames.
S703: counting in the corresponding original facial image of each original topic point, and single frames mood value is greater than default mood threshold
The target image frame number of the original facial image of value.
Wherein, the default mood threshold value be it is pre-set for assess mood value whether reach determine that it is it is interested
Threshold value.Due to the corresponding single frames mood value of each original facial image, server is determining the corresponding original of different original topic points
After beginning facial image, the single frames mood value of each original facial image can be compared with default mood threshold value, if single frames
Mood value is greater than default mood threshold value, then illustrates target customer in the shooting time of original facial image, to the target heard
Product voice data is interested or relatively approves.Specifically, the corresponding original face of each original topic point of server statistics
In image, single frames mood value is greater than the quantity of all original facial images of default mood threshold value, is determined as original facial image
Corresponding target image frame number.
S704: being based on target image frame number and total number of image frames, obtains and approves mood ratio.
Specifically, server is using approval mood ratio formula to the target image frame number and total figure of each original topic point
As frame number is calculated, the approval mood ratio of each original topic point is obtained.Wherein, approve that mood ratio formula is Q=M/
N, wherein Q is to approve mood ratio, and M is the target image frame number of original topic point, and N is total number of image frames of original topic point.
S705: if approving, mood ratio is greater than default mood ratio, and original topic point is determined as target topic point,
Target topic point is highlighted in the micro- expression trend graph of target.
Wherein, presetting mood ratio is pre-set for assessing whether mood ratio reaches determining target topic point
Value.Specifically, the approval mood ratio of each original topic point is compared by server with default mood ratio, if the approval
Mood ratio is greater than default mood ratio, then illustrates that target customer compares approval or interested to original topic point, by the original
Beginning topic point is determined as target topic point, so as to the hobby of basic target topic point analysis target customer, helps to improve mesh
Client is marked to the promotion success rate of target product.Further, after determining target topic point, can also topic point marked out
The micro- expression trend graph of target in target topic point is highlighted, for example, by using prominent color label target topic point,
With reminding business, personnel pay attention to.
Aim client orientation method based on micro- expression provided by the present embodiment, by counting each original topic point pair
The total number of image frames and target image frame number answered, and target customer is calculated to each original words using approval mood ratio formula
The approval mood ratio for inscribing point, reflects the target customer to the degree of recognition of original topic point.It will recognize that mood ratio is greater than again
The original topic point of default mood ratio is determined as target topic point, and highlights the target in the micro- expression trend graph of target
Topic point is facilitated business personnel and is talked about based on the target to achieve the purpose that highlight the topic point that target customer most pays close attention to
Topic point carries out further genralrlization introduction with target customer, improves the promotion success rate of target product.
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
In one embodiment, a kind of aim client orientation device based on micro- expression is provided, it should the target based on micro- expression
Aim client orientation method in items of customer located equipment and above-described embodiment based on micro- expression corresponds.As shown in figure 8, the base
In the aim client orientation device of micro- expression include product recommendations request module 801, target product data acquisition module
802, video data acquisition module 803, micro- Expression Recognition module 804, investment willingness degree obtain module 805, target customer determines
Module 806 and customer information processing module 807.Detailed description are as follows for each functional module:
Product recommendations request module 801, for obtaining the product recommendations request of promotion terminal transmission, product recommendations are asked
It asks including target product ID.
Target product data acquisition module 802, for inquiring product information database, acquisition and mesh based on target product ID
Mark the corresponding target product voice data of product IDs and target product text data.
Video data obtains module 803, produces for controlling promotion terminal displaying target product text data and playing target
While product voice data, the video data of the potential customers of the camera captured in real-time of promotion terminal, video data packet are obtained
Include an at least frame images to be recognized.
Micro- Expression Recognition module 804 is obtained for being identified using micro- Expression Recognition model to each images to be recognized
The corresponding micro- expression type of images to be recognized and related coefficient.
Investment willingness degree obtains module 805, for micro- expression type using investment willingness degree formula to images to be recognized
It is calculated with related coefficient, obtains target investment willingness degree.
Target customer's determining module 806, if being greater than default wish degree threshold value for target investment willingness degree, by potential visitor
Family is determined as target customer, chooses target image from the corresponding images to be recognized of target customer.
Customer information processing module 807, for obtaining target customer's information based on target investment willingness degree and target image,
And target customer's information is sent to user terminal.
Preferably, micro- Expression Recognition module 804 includes micro- expression type acquisition submodule and related coefficient acquisition submodule.
Micro- expression type acquisition submodule, for being identified using micro- Expression Recognition model to each images to be recognized,
Obtain micro- expression type corresponding with images to be recognized.
Related coefficient acquisition submodule obtains and micro- expression for being based on micro- expression type queries related coefficient conversion table
The corresponding related coefficient of type.
Preferably, micro- expression type acquisition submodule includes that instantaneous confidence level acquiring unit and micro- expression type determine list
Member.
Instantaneous confidence level acquiring unit is obtained for being identified using micro- Expression Recognition model to each images to be recognized
Take the corresponding instantaneous confidence level of at least one identification expression type.
Micro- expression type determining units, for the maximum identification expression type of instantaneous confidence level to be determined as images to be recognized
Corresponding micro- expression type.
Preferably, before micro- Expression Recognition module 804, the aim client orientation device based on micro- expression further includes original
Facial image obtains module, facial image judgment module, facial image memory module, image data base creation module, picture number
Measure statistical module, first processing module and Second processing module.
Original facial image obtains module, for being identified using human face recognition model to images to be recognized, obtains former
Beginning facial image.
Facial image judgment module, for judging original facial image with the presence or absence of corresponding benchmark face image.
Original facial image is then stored in and benchmark by facial image memory module for benchmark face image if it exists
In the corresponding face image database of facial image.
Image data base creation module, for benchmark face image if it does not exist, then using original facial image as benchmark
Facial image creates face image database corresponding with benchmark face image.
Amount of images statistical module, for preset time period of the real-time statistics from the creation time of face image database
It is interior, the amount of images of original facial image in face image database.
First processing module executes if being greater than preset quantity threshold value for amount of images using micro- Expression Recognition model
The step of each images to be recognized is identified, the corresponding micro- expression type of images to be recognized and related coefficient are obtained.
Second processing module deletes face image database if being not more than preset quantity threshold value for amount of images.
Preferably, after target customer's determining module 806, the aim client orientation device based on micro- expression further includes list
Frame mood value obtains module, original trend graph obtains module and target trend graph obtains module.
Single frames mood value obtains module, for obtaining each original from the corresponding face image database of target customer
The corresponding single frames mood value of facial image.
Original trend graph obtains module, is used for according to the corresponding shooting time of original facial image, using third party's image
Handling implement handles the single frames mood value of each original facial image, obtains original micro- expression trend graph.
Target trend graph obtains module, for the play time according to target product voice data, walks to original micro- expression
Gesture figure carries out topic point mark, obtains the micro- expression trend graph of target.
Preferably, it includes original topic point acquiring unit, total number of image frames acquiring unit, mesh that target trend graph, which obtains module,
Logo image frame number acquiring unit approves mood ratio acquiring unit and target topic point determination unit.
Original topic point acquiring unit obtains at least two for carrying out topic point mark to original micro- expression trend graph
Original topic point.
Total number of image frames acquiring unit, for counting total picture frame of the corresponding original facial image of each original topic point
Number.
Target image frame number acquiring unit, for counting in the corresponding original facial image of each original topic point, single frames
Mood value is greater than the target image frame number of the original facial image of default mood threshold value.
Approve mood ratio acquiring unit, for being based on target image frame number and total number of image frames, obtains and approve mood ratio
Value.
Target topic point determination unit, if for approving that mood ratio is greater than default mood ratio, by original topic point
It is determined as target topic point, target topic point is highlighted in the micro- expression trend graph of target.
Preferably, it includes mood quantity statistics unit and investment willingness degree computing unit that investment willingness degree, which obtains module 805,.
Mood quantity statistics unit is determined for counting the quantity of the corresponding all micro- expression types of same related coefficient
The corresponding expression number of attributes of each related coefficient calculates expression total quantity according to the expression number of attributes of each related coefficient.
Investment willingness degree computing unit, for using investment willingness degree formula to related coefficient, the corresponding table of related coefficient
Feelings number of attributes and expression total quantity are calculated, and target investment willingness degree is obtained, and investment willingness degree formula isWherein, R is target investment willingness degree, piFor i-th of related coefficient, qiIt is corresponding for i-th of related coefficient
Expression number of attributes, S are expression total quantity.
Specific restriction about the aim client orientation device based on micro- expression may refer to above for based on micro- table
The restriction of the aim client orientation method of feelings, details are not described herein.In the above-mentioned aim client orientation device based on micro- expression
Modules can be realized fully or partially through software, hardware and combinations thereof.Above-mentioned each module can be embedded in the form of hardware
Or independently of in the processor in computer equipment, can also be stored in a software form in the memory in computer equipment,
The corresponding operation of the above modules is executed in order to which processor calls.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 9.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment is used to store the number for acquiring or being formed during the aim client orientation method based on micro- expression of execution
According to such as target product voice data and target product text data.The network interface of the computer equipment is used for and external end
End passes through network connection communication.To realize that a kind of target customer based on micro- expression is fixed when the computer program is executed by processor
Position method.
In one embodiment, a kind of computer equipment is provided, including memory, processor and storage are on a memory
And the computer program that can be run on a processor, processor are realized in above-described embodiment when executing computer program based on micro- table
The step of aim client orientation method of feelings, such as step S201-S207 or Fig. 3 shown in Fig. 2 is to step shown in fig. 7
Suddenly, to avoid repeating, which is not described herein again.Alternatively, processor realizes the target visitor based on micro- expression when executing computer program
The function of each module/unit in this embodiment of family positioning device, such as product recommendations request module shown in Fig. 8
801, target product data acquisition module 802, video data obtain module 803, micro- Expression Recognition module 804, investment willingness degree
Module 805, the function of target customer's determining module 806 and customer information processing module 807 are obtained, to avoid repeating, here not
It repeats again.
In one embodiment, a computer readable storage medium is provided, meter is stored on the computer readable storage medium
Calculation machine program, the computer program realize the aim client orientation side based on micro- expression in above-described embodiment when being executed by processor
The step of method, such as step S201-S207 or Fig. 3 shown in Fig. 2 is to step shown in fig. 7, to avoid repeating, here
It repeats no more.Alternatively, the computer program realizes the above-mentioned aim client orientation device based on micro- expression when being executed by processor
The function of each module/unit in this embodiment, such as product recommendations request module 801 shown in Fig. 8, target product
Data acquisition module 802, video data obtain module 803, micro- Expression Recognition module 804, investment willingness degree obtain module 805,
The function of target customer's determining module 806 and customer information processing module 807, to avoid repeating, which is not described herein again.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
It is apparent to those skilled in the art that for convenience of description and succinctly, only with above-mentioned each function
Can unit, module division progress for example, in practical application, can according to need and by above-mentioned function distribution by different
Functional unit, module are completed, i.e., the internal structure of described device is divided into different functional unit or module, more than completing
The all or part of function of description.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations；Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features；And these are modified
Or replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution should all
It is included within protection scope of the present invention.
Claims (10)
1. a kind of aim client orientation method based on micro- expression characterized by comprising
The product recommendations request that promotion terminal is sent is obtained, the product recommendations request includes target product ID；
Product information database is inquired based on the target product ID, obtains target product corresponding with the target product ID
Voice data and target product text data；
While controlling the promotion terminal and show the target product text data and play the target product voice data,
The video data of the potential customers of the camera captured in real-time of the promotion terminal is obtained, the video data includes an at least frame
Images to be recognized；
Each images to be recognized is identified using micro- Expression Recognition model, it is corresponding micro- to obtain the images to be recognized
Expression type and related coefficient；
It is calculated using micro- expression type and related coefficient of the investment willingness degree formula to the images to be recognized, obtains target
Investment willingness degree；
If the target investment willingness degree is greater than default wish degree threshold value, the potential customers are determined as target customer, from
Target image is chosen in the corresponding images to be recognized of the target customer；
Target customer's information is obtained based on the target investment willingness degree and the target image, and by target customer's information
It is sent to user terminal.
2. as described in claim 1 based on the aim client orientation method of micro- expression, which is characterized in that described to use micro- expression
Identification model identifies each images to be recognized, obtains the corresponding micro- expression type of the images to be recognized and correlation
Coefficient, comprising:
Each images to be recognized is identified using micro- Expression Recognition model, is obtained corresponding with the images to be recognized
Micro- expression type；
Based on micro- expression type queries related coefficient conversion table, phase relation corresponding with the micro- expression type is obtained
Number.
3. as claimed in claim 2 based on the aim client orientation method of micro- expression, which is characterized in that described to use micro- expression
Identification model identifies each images to be recognized, obtains micro- expression type corresponding with the images to be recognized,
Include:
Each images to be recognized is identified using micro- Expression Recognition model, obtains at least one identification expression type pair
The instantaneous confidence level answered；
The instantaneous maximum identification expression type of confidence level is determined as the corresponding micro- expression type of the images to be recognized.
4. as described in claim 1 based on the aim client orientation method of micro- expression, which is characterized in that use micro- table described
Feelings identification model identifies each images to be recognized, obtains the corresponding micro- expression type of the images to be recognized and phase
Before the step of relationship number, the aim client orientation method based on micro- expression further include:
The images to be recognized is identified using human face recognition model, obtains original facial image；
Judge the original facial image with the presence or absence of corresponding benchmark face image；
The original facial image is then stored in corresponding with the benchmark face image by the benchmark face image if it exists
Face image database in；
The benchmark face image if it does not exist, then using the original facial image as benchmark face image, creation with it is described
The corresponding face image database of benchmark face image；
Real-time statistics are in the preset time period from the creation time of the face image database, the face image database
In original facial image amount of images；
If described image quantity is greater than preset quantity threshold value, the micro- Expression Recognition model of the use is executed to each described wait know
The step of other image is identified, obtains the corresponding micro- expression type of the images to be recognized and related coefficient；
If described image quantity is not more than preset quantity threshold value, the face image database is deleted.
5. as claimed in claim 4 based on the aim client orientation method of micro- expression, which is characterized in that described will dive described
After the step of client is determined as target customer, the aim client orientation method based on micro- expression further include:
From the corresponding face image database of the target customer, the corresponding single frames feelings of each original facial image are obtained
Thread value；
According to the corresponding shooting time of the original facial image, using third party's image processing tool to each primitive man
The single frames mood value of face image is handled, and original micro- expression trend graph is obtained；
According to the play time of the target product voice data, topic point mark is carried out to original micro- expression trend graph,
Obtain the micro- expression trend graph of target.
6. as claimed in claim 5 based on the aim client orientation method of micro- expression, which is characterized in that described to described original
Micro- expression trend graph carries out topic point mark, obtains the micro- expression trend graph of target, comprising:
Topic point mark is carried out to original micro- expression trend graph, obtains 1 original topic points；
Count total number of image frames of the corresponding original facial image of each original topic point；
It counts in the corresponding original facial image of each original topic point, single frames mood value is greater than the original of default mood threshold value
The target image frame number of beginning facial image；
Based on the target image frame number and total number of image frames, obtains and approve mood ratio；
If the approval mood ratio is greater than default mood ratio, the original topic point is determined as target topic point,
The target topic point is highlighted in the micro- expression trend graph of target.
7. as described in claim 1 based on the aim client orientation method of micro- expression, which is characterized in that described using investment meaning
Hope degree formula calculates the micro- expression type and related coefficient of the images to be recognized, obtains target investment willingness degree, packet
It includes:
The quantity for counting the corresponding all micro- expression types of the same related coefficient, determines each related coefficient pair
The expression number of attributes answered calculates expression total quantity according to the expression number of attributes of each related coefficient；
Using the investment willingness degree formula to the related coefficient, the corresponding expression number of attributes of the related coefficient and expression
Total quantity is calculated, and target investment willingness degree is obtained, and the investment willingness degree formula isWherein, R is mesh
Mark investment willingness degree, piFor i-th of related coefficient, qiFor the corresponding expression number of attributes of i-th of related coefficient, S is expression sum
Amount.
8. a kind of aim client orientation device based on micro- expression characterized by comprising
Product recommendations request module, for obtaining the product recommendations request of promotion terminal transmission, the product recommendations request
Including target product ID；
Target product data acquisition module, for based on the target product ID inquire product information database, obtain with it is described
The corresponding target product voice data of target product ID and target product text data；
Video data obtains module, shows the target product text data for controlling the promotion terminal and plays the mesh
While marking product voice data, the video data of the potential customers of the camera captured in real-time of the promotion terminal, institute are obtained
Stating video data includes an at least frame images to be recognized；
Micro- Expression Recognition module obtains institute for identifying using micro- Expression Recognition model to each images to be recognized
State the corresponding micro- expression type of images to be recognized and related coefficient；
Investment willingness degree obtains module, for the micro- expression type and phase using investment willingness degree formula to the images to be recognized
Relationship number is calculated, and target investment willingness degree is obtained；
Target customer's determining module will be described potential if being greater than default wish degree threshold value for the target investment willingness degree
Client is determined as target customer, chooses target image from the corresponding images to be recognized of the target customer；
Customer information processing module, for obtaining target customer's letter based on the target investment willingness degree and the target image
Breath, and target customer's information is sent to user terminal.
9. a kind of computer equipment, including memory, processor and storage are in the memory and can be in the processor
The computer program of upper operation, which is characterized in that the processor realized when executing the computer program as claim 1 to
The step of aim client orientation method described in 7 any one based on micro- expression.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In target visitor of the realization as described in any one of claim 1 to 7 based on micro- expression when the computer program is executed by processor
The step of family localization method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910043485.5A CN109858958A (en) | 2019-01-17 | 2019-01-17 | Aim client orientation method, apparatus, equipment and storage medium based on micro- expression |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910043485.5A CN109858958A (en) | 2019-01-17 | 2019-01-17 | Aim client orientation method, apparatus, equipment and storage medium based on micro- expression |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109858958A true CN109858958A (en) | 2019-06-07 |
Family
ID=66894952
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910043485.5A Pending CN109858958A (en) | 2019-01-17 | 2019-01-17 | Aim client orientation method, apparatus, equipment and storage medium based on micro- expression |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109858958A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110400563A (en) * | 2019-07-18 | 2019-11-01 | 平安科技（深圳）有限公司 | Vehicle-mounted voice instruction identification method, device, computer equipment and storage medium |
| CN110415015A (en) * | 2019-06-19 | 2019-11-05 | 深圳壹账通智能科技有限公司 | Product degree of recognition analysis method, device, terminal and computer readable storage medium |
| CN110532971A (en) * | 2019-09-02 | 2019-12-03 | 京东方科技集团股份有限公司 | Image procossing and device, training method and computer readable storage medium |
| CN111062332A (en) * | 2019-12-18 | 2020-04-24 | 秒针信息技术有限公司 | Information pushing method and device |
| CN111144241A (en) * | 2019-12-13 | 2020-05-12 | 深圳奇迹智慧网络有限公司 | Target identification method and device based on image verification and computer equipment |
| CN111382691A (en) * | 2020-03-05 | 2020-07-07 | 甄十信息科技（上海）有限公司 | Screen content page turning method and mobile terminal |
| CN111476095A (en) * | 2020-03-05 | 2020-07-31 | 平安科技（深圳）有限公司 | Expression attribute identification method and device, computer equipment and storage medium |
| CN113487356A (en) * | 2021-07-08 | 2021-10-08 | 上海叮铃铃信息技术有限公司 | Customer management method and system based on face recognition |
| CN113762156A (en) * | 2021-09-08 | 2021-12-07 | 北京优酷科技有限公司 | Viewing data processing method, device and storage medium |
| CN113920564A (en) * | 2021-10-29 | 2022-01-11 | 中国平安财产保险股份有限公司 | Client mining method based on artificial intelligence and related equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779292A (en) * | 2016-11-18 | 2017-05-31 | 浪潮（苏州）金融技术服务有限公司 | A kind of is the method and device of offering customers service based on recognition of face |
| CN108009465A (en) * | 2016-10-31 | 2018-05-08 | 杭州海康威视数字技术股份有限公司 | A kind of face identification method and device |
| CN109165982A (en) * | 2018-08-28 | 2019-01-08 | 百度在线网络技术（北京）有限公司 | The determination method and apparatus of user's purchase information |
| CN109168052A (en) * | 2018-10-31 | 2019-01-08 | 杭州比智科技有限公司 | The determination method, apparatus and calculating equipment of service satisfaction |
-
2019
- 2019-01-17 CN CN201910043485.5A patent/CN109858958A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108009465A (en) * | 2016-10-31 | 2018-05-08 | 杭州海康威视数字技术股份有限公司 | A kind of face identification method and device |
| CN106779292A (en) * | 2016-11-18 | 2017-05-31 | 浪潮（苏州）金融技术服务有限公司 | A kind of is the method and device of offering customers service based on recognition of face |
| CN109165982A (en) * | 2018-08-28 | 2019-01-08 | 百度在线网络技术（北京）有限公司 | The determination method and apparatus of user's purchase information |
| CN109168052A (en) * | 2018-10-31 | 2019-01-08 | 杭州比智科技有限公司 | The determination method, apparatus and calculating equipment of service satisfaction |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110415015A (en) * | 2019-06-19 | 2019-11-05 | 深圳壹账通智能科技有限公司 | Product degree of recognition analysis method, device, terminal and computer readable storage medium |
| WO2020253363A1 (en) * | 2019-06-19 | 2020-12-24 | 深圳壹账通智能科技有限公司 | Product approval degree analysis method and device, terminal and computer readable storage medium |
| CN110400563A (en) * | 2019-07-18 | 2019-11-01 | 平安科技（深圳）有限公司 | Vehicle-mounted voice instruction identification method, device, computer equipment and storage medium |
| CN110532971B (en) * | 2019-09-02 | 2023-04-28 | 京东方科技集团股份有限公司 | Image processing apparatus, training method, and computer-readable storage medium |
| CN110532971A (en) * | 2019-09-02 | 2019-12-03 | 京东方科技集团股份有限公司 | Image procossing and device, training method and computer readable storage medium |
| US11961327B2 (en) | 2019-09-02 | 2024-04-16 | Boe Technology Group Co., Ltd. | Image processing method and device, classifier training method, and readable storage medium |
| CN111144241A (en) * | 2019-12-13 | 2020-05-12 | 深圳奇迹智慧网络有限公司 | Target identification method and device based on image verification and computer equipment |
| CN111062332A (en) * | 2019-12-18 | 2020-04-24 | 秒针信息技术有限公司 | Information pushing method and device |
| CN111476095A (en) * | 2020-03-05 | 2020-07-31 | 平安科技（深圳）有限公司 | Expression attribute identification method and device, computer equipment and storage medium |
| CN111382691A (en) * | 2020-03-05 | 2020-07-07 | 甄十信息科技（上海）有限公司 | Screen content page turning method and mobile terminal |
| CN113487356B (en) * | 2021-07-08 | 2022-04-05 | 上海叮铃铃信息技术有限公司 | Customer management method and system based on face recognition |
| CN113487356A (en) * | 2021-07-08 | 2021-10-08 | 上海叮铃铃信息技术有限公司 | Customer management method and system based on face recognition |
| CN113762156A (en) * | 2021-09-08 | 2021-12-07 | 北京优酷科技有限公司 | Viewing data processing method, device and storage medium |
| CN113762156B (en) * | 2021-09-08 | 2023-10-24 | 北京优酷科技有限公司 | Video data processing method, device and storage medium |
| CN113920564A (en) * | 2021-10-29 | 2022-01-11 | 中国平安财产保险股份有限公司 | Client mining method based on artificial intelligence and related equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109858958A (en) | Aim client orientation method, apparatus, equipment and storage medium based on micro- expression | |
| JP7500689B2 (en) | Technique for identifying skin color in images under uncontrolled lighting conditions | |
| Leng et al. | Person re-identification with content and context re-ranking | |
| WO2020143156A1 (en) | Hotspot video annotation processing method and apparatus, computer device and storage medium | |
| CN109783730A (en) | Products Show method, apparatus, computer equipment and storage medium | |
| WO2020119497A1 (en) | Product pushing method and apparatus, computer device and storage medium | |
| CN109345302A (en) | Machine learning model training method, device, storage medium and computer equipment | |
| CN110175298B (en) | User matching method | |
| WO2020253360A1 (en) | Content display method and apparatus for application, storage medium, and computer device | |
| JP2023531264A (en) | Systems and methods for improved facial attribute classification and its use | |
| CN104298983A (en) | Tongue fur image acquisition and analysis system with distributed user terminals | |
| CN110134806A (en) | The method and system of context user profile photo selection | |
| CN109376598A (en) | Facial expression image processing method, device, computer equipment and storage medium | |
| CN109447729A (en) | A kind of recommended method of product, terminal device and computer readable storage medium | |
| CN110111136A (en) | Video data handling procedure, device, computer equipment and storage medium | |
| US11625754B2 (en) | Method for providing text-reading based reward-type advertisement service and user terminal for executing same | |
| CN110178129A (en) | The system and method for generating the semantic expressiveness of the target image for image processing operations | |
| CN110717509A (en) | Data sample analysis method and device based on tree splitting algorithm | |
| Zhang et al. | Visual aesthetic understanding: Sample-specific aesthetic classification and deep activation map visualization | |
| KR102594093B1 (en) | Dermatologic treatment recommendation system using deep learning model and method thereof | |
| Neumann et al. | Body language and gender stereotypes in campaign video | |
| CN113821574A (en) | User behavior classification method and device and storage medium | |
| CN114253990A (en) | Database query method and device, computer equipment and storage medium | |
| CN113850416A (en) | Advertisement promotion cooperation object determining method and device | |
| CN111274476B (en) | House source matching method, device, equipment and storage medium based on face recognition |
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 |