CN110223158A - A kind of recognition methods of risk subscribers, device, storage medium and server - Google Patents
A kind of recognition methods of risk subscribers, device, storage medium and server Download PDFInfo
- Publication number
- CN110223158A CN110223158A CN201910423191.5A CN201910423191A CN110223158A CN 110223158 A CN110223158 A CN 110223158A CN 201910423191 A CN201910423191 A CN 201910423191A CN 110223158 A CN110223158 A CN 110223158A
- Authority
- CN
- China
- Prior art keywords
- micro
- expressive features
- target user
- face image
- multimedia file
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/174—Facial expression recognition
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Collating Specific Patterns (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to field of computer technology, propose recognition methods, device, storage medium and the server of a kind of risk subscribers.The recognition methods of the risk subscribers can play some specific multimedia file before carrying out credit face and examining for user, and acquire face image when user appreciates the multimedia file;After the multimedia file finishes, first micro- expressive features are identified from collected face image;Then, first micro- expressive features and the micro- expressive features of pre-stored benchmark are compared, and judge whether the user is potential risk user according to the result of the comparison;After examining face to face, is examined in the collected face image of process from face and identify second micro- expressive features;Finally, whether being potential risk user according to the user, different micro- expression fraud identification models is chosen respectively and carries out risk identification.Using the recognition methods of risk subscribers proposed by the present invention, the accuracy rate of risk subscribers identification can be improved.
Description
Technical field
The present invention relates to field of computer technology more particularly to a kind of recognition methods of risk subscribers, device, storage medium
And server.
Background technique
It is examined in process in the face of credit examination & approval, finance company can propose that various problems allow user to answer, and acquire user
Face image when answering a question detects the micro- expression generated in face image, the micro- expression input building in advance that will test
Micro- expression cheat identification model, to differentiate whether user lies.
However, certain illegal users can ferment lie before face is examined, the micro- expression generated when face is examined and answers a question
Smaller with the difference of normal users, model possibly can not identify that the tiny difference, such as some particular emotion occur at this time
Although number it is higher, not yet reach risk decision threshold, model can generate misjudgement at this time, and there are risks.
Summary of the invention
In view of this, the embodiment of the invention provides a kind of recognition methods of risk subscribers, device, storage medium and services
Device can be improved the accuracy rate of risk subscribers identification.
The embodiment of the present invention in a first aspect, providing a kind of recognition methods of risk subscribers, comprising:
When playing destination multimedia file, the first face image of target user is acquired, and scheme from first face
First micro- expressive features are identified as in, the destination multimedia file is determined according to the personal information of the target user, institute
Stating the first face image is each frame face image for including in collected video when playing destination multimedia file;
Described first micro- expressive features and the micro- expressive features of pre-stored benchmark are compared, the micro- expression of benchmark
Feature is micro- expressive features that normal users are generated when appreciating the destination multimedia file;
If the difference between first micro- expressive features and the micro- expressive features of the benchmark is more than preset threshold value, sentence
The fixed target user is potential risk user;
Second face image of the target user when face is examined is acquired, and identifies from second face image
Two micro- expressive features, each frame face image that second face image includes in collected video when examining for face;
If the target user is potential risk user, described second micro- expressive features are inputted into first constructed in advance
Micro- expression cheats identification model, and the target user according to the output result judgement of described first micro- expression fraud identification model
It whether is risk subscribers;
If the target user is not potential risk user, the described second micro- expressive features input is constructed in advance the
Two micro- expressions cheat identification model, and the target according to the output result judgement of described second micro- expression fraud identification model is used
Whether family is risk subscribers, and first micro- expression fraud identification model and second micro- expression fraud identification model have not
Same risk decision threshold.
The second aspect of the embodiment of the present invention provides a kind of identification device of risk subscribers, comprising:
First micro- expressive features identification module, for acquiring the first of target user when playing destination multimedia file
Face image, and first micro- expressive features are identified from first face image, the destination multimedia file is according to institute
The personal information for stating target user determines that first face image is when being broadcasting destination multimedia file in collected video
The each frame face image for including;
Micro- expressive features comparison module is used for described first micro- expressive features and the micro- expressive features of pre-stored benchmark
It is compared, the micro- expressive features of benchmark are that micro- expression that normal users are generated when appreciating the destination multimedia file is special
Sign;
Potential risk user's determination module, if between described first micro- expressive features and the micro- expressive features of the benchmark
Difference be more than preset threshold value, then determine the target user for potential risk user;
Second micro- expressive features identification module, for acquiring second face image of the target user when face is examined, and
Second micro- expressive features, second face image collected video when examining for face are identified from second face image
In include each frame face image;
First risk subscribers determination module, it is micro- by described second if being potential risk user for the target user
First micro- expression fraud identification model that expressive features input constructs in advance, and identification model is cheated according to described first micro- expression
Output result judgement described in target user whether be risk subscribers;
Second risk subscribers determination module, if not being potential risk user for the target user, by described second
Second micro- expression fraud identification model that micro- expressive features input constructs in advance, and according to described second micro- expression fraud identification mould
Whether target user described in the output result judgement of type is risk subscribers, the first micro- expression fraud identification model and described the
Two micro- expression fraud identification models have different risk decision thresholds.
The third aspect of the embodiment of the present invention, provides a kind of computer readable storage medium, described computer-readable to deposit
Storage media is stored with computer-readable instruction, and such as the embodiment of the present invention is realized when the computer-readable instruction is executed by processor
First aspect propose risk subscribers recognition methods the step of.
The fourth aspect of the embodiment of the present invention, provides a kind of server, including memory, processor and is stored in institute
The computer-readable instruction that can be run in memory and on the processor is stated, the processor executes described computer-readable
The step of recognition methods for the risk subscribers that the first aspect such as the embodiment of the present invention proposes is realized when instruction.
The recognition methods for the risk subscribers that the embodiment of the present invention proposes can play before carrying out credit face and examining for user
Some specific multimedia file, and acquire face image when user appreciates the multimedia file;When the multimedia file is broadcast
After putting, first micro- expressive features are identified from collected face image;Then, by first micro- expressive features and in advance
The micro- expressive features of the benchmark first stored are compared, and judge whether the user is potential risk user according to the result of the comparison;
After examining face to face, is examined in the collected face image of process from face and identify second micro- expressive features;Finally, according to the user
Whether it is potential risk user, chooses different micro- expression fraud identification models respectively and carry out risk identification.In actual operation,
The micro- expression fraud identification model and special micro- expression fraud identification model that a routine can be set, if user is
Potential risk user then chooses special micro- expression fraud identification model and carries out risk identification.It is arranged in this way, even if certain
Illegal user has fermented lie face is pre-trial, and the difference of the micro- expression and normal users that generate when face is examined is smaller, but due to
The illegal user can be determined as to (illegal user is usual this period before face is examined by potential risk user in operation before
The some lies of tissue can be fermented, will not be absorbed at this time appreciate the multimedia file, therefore the first micro- expressive features captured and
The micro- expressive features of benchmark can have greater difference), therefore what is chosen is that special micro- expression cheats other model, which has more special
Different risk decision threshold can recognize that tiny micro- expression difference, to prevent the generation of misjudgement, improves risk subscribers and knows
Other accuracy rate.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is a kind of flow chart of one embodiment of the recognition methods of risk subscribers provided in an embodiment of the present invention;
Fig. 2 is a kind of flow chart of second embodiment of the recognition methods of risk subscribers provided in an embodiment of the present invention;
Fig. 3 is a kind of structure chart of one embodiment of the identification device of risk subscribers provided in an embodiment of the present invention;
Fig. 4 is a kind of schematic diagram of server provided in an embodiment of the present invention.
Specific embodiment
The embodiment of the invention provides a kind of recognition methods of risk subscribers, device, storage medium and server, Neng Gouti
The accuracy rate of high risk user identification.
In order to make the invention's purpose, features and advantages of the invention more obvious and easy to understand, below in conjunction with the present invention
Attached drawing in embodiment, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that disclosed below
Embodiment be only a part of the embodiment of the present invention, and not all embodiment.Based on the embodiments of the present invention, this field
Those of ordinary skill's all other embodiment obtained without making creative work, belongs to protection of the present invention
Range.
Referring to Fig. 1, a kind of one embodiment of the recognition methods of risk subscribers includes: in the embodiment of the present invention
101, when playing destination multimedia file, the first face image of target user is acquired, and from first face
First micro- expressive features are identified in portion's image;
At the beginning of face is examined not yet, server can be to handle the target user of credit operation to play some specific mesh
Mark multimedia file.The multimedia file can be video file or audio file, for example can be a short-movie of making laughs, or
The first cavatine of person one.The destination multimedia file is determined according to the personal information of the target user, can specifically include:
(1) the identity card scanned copy of the target user is obtained;
(2) OCR identification is carried out to the identity card scanned copy, obtains age, gender, nationality and the ground of the target user
Area;
(3) it inquires and the age, gender, the type of the associated multimedia file of nationality and area;
(4) multimedia file of the type is chosen from the multimedia file library constructed in advance, it is more as the target
Media file.
By the identification to identity card scanned copy, the personal letter such as age, gender, nationality and area of target user is extracted
Then breath is played out according to the multimedia file that these information choose suitable type.System can construct various more matchmakers in advance
The age of body file type and user, gender, nationality, area between incidence relation, if than user be middle-aged male, then select
Take the film of the subject matters such as military affairs, current events;If user is young woman, popular idol play or variety show can be chosen;If with
Family is the personnel of some ethnic group, can choose characteristic song broadcasting of the ethnic group, etc..It is arranged in this way, energy
It is enough effectively to avoid being played to the uninterested multimedia file of user.If what is played is the uninterested multimedia file of user,
Then user may voluntarily play mobile phone or close mesh rest, can not grab the user effectively micro- expression at this time.
During playing destination multimedia file, the first face image of the target user is acquired, described first
Face image is each frame face image for including in collected video when playing destination multimedia file.Specifically, in mesh
When mark user appreciates the destination multimedia file, the face of the target user is acquired by the camera that specific position is arranged
Image, the face image of acquisition are the continuous video images of multiframe.After the multimedia file finishes, from described first
Identify first micro- expressive features in face image, micro- expression can be smile, frown, left eye upward, blink, beep mouth etc. it is all kinds of
Facial expressions and acts, micro- expressive features can be the number that each micro- expression occurs, and be also possible to the accounting that each micro- expression occurs.It is right
Collected each frame face image is analyzed, and can be counted the target user and be gone out in appreciating the destination multimedia file processes
The total quantity of existing micro- expression, wherein micro- expressive features such as the number of each micro- expression appearance and accounting.
Specifically, the micro- expression of some in image can be grabbed by following steps:
(1) facial image is extracted from video image using Face datection algorithm;
(2) human face characteristic point is extracted from facial image;
(3) dimension-reduction treatment, such as PCA dimensionality reduction are carried out to the human face characteristic point extracted;
(4) human face action feature is identified according to the human face characteristic point after dimensionality reduction.
Through the above steps, the micro- expression of each of the face appearance in video image can be grabbed, for example is blinked, eyeball
Towards upper right side etc..First micro- expressive features are essentially the target user in the process for appreciating the destination multimedia file
The number of each micro- expression of middle appearance or the statistical value of accounting, such as shown in table below 1 or table 2:
Table 1
Micro- expression classification | Frequency of occurrence |
It smiles | 5 |
Eyeball is towards upper right side | 2 |
It blinks | 46 |
… |
Table 2
Micro- expression classification | Accounting in the quantity of all micro- expressions of appearance |
It smiles | 27% |
Eyeball is towards upper right side | 10% |
… |
102, described first micro- expressive features and the micro- expressive features of pre-stored benchmark are compared;
After identifying first micro- expressive features, by described first micro- expressive features and the micro- expression of pre-stored benchmark
Feature is compared, and the micro- expressive features of benchmark are micro- table that normal users are generated when appreciating the destination multimedia file
Feelings feature.Specifically, the corresponding micro- expressive features of benchmark can be acquired respectively for the multimedia file of each storage in advance.Some
The micro- expressive features of the corresponding benchmark of multimedia file can obtain in the following manner: acquiring multiple general staff's viewings respectively should
Then the video image of multimedia file identifies micro- expressive features of each personnel respectively, finally seeking this, expression is special slightly
The average value of sign obtains the micro- expressive features of benchmark.
Specifically, step 102 may include:
(1) primary vector is constructed according to described first micro- expressive features, each element of the primary vector is respectively institute
State the number or accounting that each micro- expression that first micro- expressive features include occurs;
(2) secondary vector is constructed according to the micro- expressive features of the benchmark, each element of the secondary vector is respectively institute
State the number or accounting that each micro- expression that the micro- expressive features of benchmark include occurs;
(3) the distance between the primary vector and the secondary vector are calculated, as described first micro- expressive features and
Difference between the micro- expressive features of benchmark.
When comparing first micro- expressive features and the micro- expressive features of benchmark, micro- expressive features can be expressed as to the shape of vector
Formula, each element of vector are respectively the number or accounting for each micro- expression appearance that micro- expressive features include.Such as, it is assumed that the
One micro- expressive features are data shown in table 1, then these data can be expressed as to the vector form of (5,2,46 ...), i.e., the
One vector.Assuming that the micro- expressive features of benchmark are data shown in table 3, then these data can be expressed as (4,0,52 ...)
Vector form, i.e. secondary vector.It should be noted that the number for the element that primary vector and secondary vector have is identical, and to
The one-to-one correspondence that puts in order of each micro- expression element in amount, such as first element of two vectors indicate appearance of smiling
Number.The distance between the two vectors, the difference between as first micro- expressive features and the micro- expressive features of benchmark can be calculated
Not.
Table 3
Micro- expression classification | Frequency of occurrence |
It smiles | 4 |
Eyeball is towards upper right side | 0 |
It blinks | 52 |
… |
103, judge whether the difference between described first micro- expressive features and the micro- expressive features of the benchmark is more than default
Threshold value;
After the comparison, judge whether the difference between described first micro- expressive features and the micro- expressive features of the benchmark surpasses
Cross preset threshold value.If the difference between first micro- expressive features and the micro- expressive features of the benchmark is more than preset threshold
Value, thens follow the steps 104, no to then follow the steps 105.
104, determine that the target user is potential risk user;
If target user is to intend the illegal user to cheat loan that tells a lie, this period before credit face is examined, it will usually
The some lies of tissue are fermented, at this time when watching the multimedia file, will not be generally absorbed in completely, therefore the micro- table of first captured
Feelings feature can have greater difference with the micro- expressive features of benchmark.Therefore, if first micro- expressive features and the micro- expression of the benchmark
Difference between feature is more than preset threshold value, then determines that the target user for potential risk user, then executes step
106。
105, determine that the target user is general user;
If between first micro- expressive features and the micro- expressive features of the benchmark relatively, showing the target user
For general normal users, determine that the target user for general user, then executes step 106 at this time.
106, second face image of the target user when face is examined is acquired, and is identified from second face image
Second micro- expressive features out;
Next, the second face image, second face image can be acquired during the target user face is examined
The each frame face image for including in collected video when examining for face, identifies the target user from second face image
Micro- expressive features, i.e., second micro- expressive features.
107, judge whether the target user is potential risk user;
Then, judge whether the target user is potential risk user.If the target user is potential risk user,
Then follow the steps 108;If the target user is not potential risk user, 109 are thened follow the steps.
108, first micro- expression that described second micro- expressive features input constructs in advance is cheated into identification model, and according to institute
State whether target user described in the output result judgement of first micro- expression fraud identification model is risk subscribers;
The target user is potential risk user, and it is micro- that described second micro- expressive features are inputted to first constructed in advance at this time
Expression cheats identification model, and judges whether the target user is risk subscribers according to the output result of the model.First is micro-
Expression fraud identification model is more special micro- expression fraud identification model, and for the user of potential risk, model is set
The parameter (such as decision threshold) set is different with conventional identification model, for example for general user, the table of eyes upward occurs
It is determined as risk subscribers feelings 5 times or more;And for potential risk user, occur the expression of eyes upward 3 times or more and is determined as
Risk subscribers.
109, second micro- expression that described second micro- expressive features input constructs in advance is cheated into identification model, and according to institute
State whether target user described in the output result judgement of second micro- expression fraud identification model is risk subscribers.
The target user is general user, and described second micro- expressive features are inputted the second micro- expression constructed in advance at this time
Identification model is cheated, and judges whether the target user is risk subscribers according to the output result of the model.Second micro- expression
Fraud identification model is micro- expression fraud identification model an of routine, and for general user, the parameter of model setting (for example is sentenced
Certainly threshold value) on the basis of the test result of general user.Generally speaking, the described first micro- expression fraud identification model and described the
Two micro- expressions fraud identification models have different risk decision thresholds, are respectively suitable for potential risk user and general user
Risk identification.
The recognition methods for the risk subscribers that the embodiment of the present invention proposes can play before carrying out credit face and examining for user
Some specific multimedia file, and acquire face image when user appreciates the multimedia file;When the multimedia file is broadcast
After putting, first micro- expressive features are identified from collected face image;Then, by first micro- expressive features and in advance
The micro- expressive features of the benchmark first stored are compared, and judge whether the user is potential risk user according to the result of the comparison;
After examining face to face, is examined in the collected face image of process from face and identify second micro- expressive features;Finally, according to the user
Whether it is potential risk user, chooses different micro- expression fraud identification models respectively and carry out risk identification.In actual operation,
The micro- expression fraud identification model and special micro- expression fraud identification model that a routine can be set, if user is
Potential risk user then chooses special micro- expression fraud identification model and carries out risk identification.It is arranged in this way, even if certain
Illegal user has fermented lie face is pre-trial, and the difference of the micro- expression and normal users that generate when face is examined is smaller, but due to
The illegal user can be determined as to (illegal user is usual this period before face is examined by potential risk user in operation before
The some lies of tissue can be fermented, will not be absorbed at this time appreciate the multimedia file, therefore the first micro- expressive features captured and
The micro- expressive features of benchmark can have greater difference), therefore what is chosen is that special micro- expression cheats other model, which has more special
Different risk decision threshold can recognize that tiny micro- expression difference, to prevent the generation of misjudgement, improves risk subscribers and knows
Other accuracy rate.
Referring to Fig. 2, a kind of second embodiment of the recognition methods of risk subscribers includes: in the embodiment of the present invention
201, when playing destination multimedia file, the first face image of target user is acquired;
Step 201 is similar with step 101, specifically can refer to the related description of step 101.
202, after the multimedia file finishes, determine that the target user is glad according to first face image
Appreciate the time of the destination multimedia file;
After the destination multimedia file finishes, determine that the target user is glad according to first face image
Appreciate the time of the multimedia file.When differentiating potential risk user relatively absorbedly, it is desirable that user appreciates multimedia text
Part, it is therefore desirable to determine that the target user appreciates the time of the multimedia file.
Specifically, step 202 may include:
(1) the quantity accounting of face image in first face image is counted;
(2) target user is calculated according to the playing duration of the quantity accounting and the multimedia file to appreciate
The time of the destination multimedia file.
Specifically it can determine that it is more that the target user appreciates by accounting of the statistics face image in all face images
The time of media file, if the playing duration than multimedia file is 10 minutes, face image is in all face images
Accounting is 80%, then the target user can be calculated and appreciate the time of the destination multimedia file as 80%*10=8 points
Clock.
203, judge whether the time is greater than preset minimum time;
If the time is greater than preset minimum time, 204 are thened follow the steps, otherwise directly executes step 208.If described
Time is less than preset minimum time, show target user appreciate the multimedia file time it is very little, for example broadcast less than file
The half (minimum time of setting) for putting duration, then show that target user loses interest in the multimedia file, can not lead at this time
It crosses subsequent step and judges whether the target user is potential risk user, preset prompt information can be exported and give credit authorization people
Member, or directly determine that the target user for general user, i.e., directly executes step 208.If the time is greater than preset
Between lower limit, show target user be relatively absorbed in appreciate the multimedia file, can be judged at this time by subsequent step the target use
Whether family is potential risk user.
204, first micro- expressive features are identified from first face image;
205, described first micro- expressive features and the micro- expressive features of pre-stored benchmark are compared;
206, judge whether the difference between described first micro- expressive features and the micro- expressive features of the benchmark is more than default
Threshold value;
If the difference between first micro- expressive features and the micro- expressive features of the benchmark is more than preset threshold value, hold
Row step 207, it is no to then follow the steps 208.
207, determine that the target user is potential risk user;
208, determine that the target user is general user;
209, second face image of the target user when face is examined is acquired, and is identified from second face image
Second micro- expressive features out;
210, judge whether the target user is potential risk user;
If the target user is potential risk user, 211 are thened follow the steps;If the target user is not potential risk
User thens follow the steps 212.
211, first micro- expression that described second micro- expressive features input constructs in advance is cheated into identification model, and according to institute
State whether target user described in the output result judgement of first micro- expression fraud identification model is risk subscribers;
212, second micro- expression that described second micro- expressive features input constructs in advance is cheated into identification model, and according to institute
State whether target user described in the output result judgement of second micro- expression fraud identification model is risk subscribers.
Step 205-212 is identical as step 102-109, specifically can refer to the related description of step 102-109.
The recognition methods for the risk subscribers that the embodiment of the present invention proposes can play before carrying out credit face and examining for user
Some specific multimedia file, and acquire face image when user appreciates the multimedia file;When the multimedia file is broadcast
After putting, determine that the user appreciates the time of the multimedia file according to collected face image;When judging described
Between whether be greater than preset minimum time, if then identifying first micro- expressive features from collected face image, otherwise
Directly determine the user for general user;Then, by first micro- expressive features and the micro- expressive features of pre-stored benchmark into
Row compares, and judges whether the user is potential risk user according to the result of the comparison;After examining face to face, process is examined from face and is adopted
Second micro- expressive features are identified in the face image collected;Finally, whether being potential risk user according to the user, select respectively
It takes different micro- expressions to cheat identification model and carries out risk identification.Compared with one embodiment of the invention, the present embodiment exists
When differentiating potential risk user, it can consider that user appreciates the time of the multimedia file, only when the time is more than certain
Between lower limit when can just execute subsequent potential risk user and judge operation.It is arranged in this way, potential risk can be effectively improved
The accuracy rate of user's identification.
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.
Essentially describe a kind of recognition methods of risk subscribers above, below by the identification device to a kind of risk subscribers into
Row description.
Referring to Fig. 3, a kind of one embodiment of the identification device of risk subscribers includes: in the embodiment of the present invention
First micro- expressive features identification module 301 acquires the of target user for when playing destination multimedia file
One face image, and first micro- expressive features are identified from first face image, the destination multimedia file according to
The personal information of the target user determines that first face image is collected video when playing destination multimedia file
In include each frame face image;
Micro- expressive features comparison module 302 is used for described first micro- expressive features and the micro- expression of pre-stored benchmark
Feature is compared, and the micro- expressive features of benchmark are micro- table that normal users are generated when appreciating the destination multimedia file
Feelings feature;
Potential risk user determination module 303, if being used for described first micro- expressive features and the micro- expressive features of the benchmark
Between difference be more than preset threshold value, then determine the target user for potential risk user;
Second micro- expressive features identification module 304, for acquiring second face image of the target user when face is examined,
And second micro- expressive features are identified from second face image, second face image collected view when being examined for face
The each frame face image for including in frequency;
First risk subscribers determination module 305, if being potential risk user for the target user, by described second
First micro- expression fraud identification model that micro- expressive features input constructs in advance, and according to described first micro- expression fraud identification mould
Whether target user described in the output result judgement of type is risk subscribers;
Second risk subscribers determination module 306, if not being potential risk user for the target user, by described
Second micro- expression fraud identification model that two micro- expressive features inputs construct in advance, and according to described second micro- expression fraud identification
Whether target user described in the output result judgement of model is risk subscribers, the first micro- expression fraud identification model and described
Second micro- expression fraud identification model has different risk decision thresholds.
Further, described first micro- expressive features identification module may include:
Identity card scanned copy acquiring unit, for obtaining the identity card scanned copy of the target user;
OCR recognition unit, for the identity card scanned copy carry out OCR identification, obtain the target user age,
Gender, nationality and area;
File fingerprint query unit, for inquiring and the age, gender, the associated multimedia of nationality and area
The type of file;
Multimedia file selection unit, for choosing the multimedia of the type from the multimedia file library constructed in advance
File, as the destination multimedia file.
Further, the identification device of the risk subscribers can also include:
Time determination unit is appreciated, determines that the target user appreciates the more matchmakers of target according to first face image
The time of body file;
Time comparing unit is appreciated, if being greater than preset minimum time for the time, is executed from first face
The step of identifying first micro- expressive features in portion's image.
Further, the appreciation time determination unit may include:
Face image accounting counts subelement, for counting the quantity accounting of face image in first face image;
Time computation subunit is appreciated, for calculating according to the playing duration of the quantity accounting and the multimedia file
Obtain the time that the target user appreciates the destination multimedia file.
Further, micro- expressive features comparison module may include:
Primary vector construction unit, for constructing primary vector, the primary vector according to described first micro- expressive features
Each element be respectively number or accounting that described first micro- expressive features each micro- expression for including occurs;
Secondary vector construction unit, for constructing secondary vector, the secondary vector according to the micro- expressive features of the benchmark
Each element be respectively number or accounting that the micro- expressive features of the benchmark each micro- expression for including occurs;
Vector distance computing unit, for calculating the distance between the primary vector and the secondary vector, as institute
State the difference between first micro- expressive features and the micro- expressive features of the benchmark.
The embodiment of the present invention also provides a kind of computer readable storage medium, and the computer-readable recording medium storage has
Computer-readable instruction realizes any one wind indicated such as Fig. 1 or Fig. 2 when the computer-readable instruction is executed by processor
The step of recognition methods of dangerous user.
The embodiment of the present invention also provides a kind of server, including memory, processor and storage are in the memory
And the computer-readable instruction that can be run on the processor, the processor are realized when executing the computer-readable instruction
As Fig. 1 or Fig. 2 any one risk subscribers indicated recognition methods the step of.
Fig. 4 is the schematic diagram for the server that one embodiment of the invention provides.As shown in figure 4, the server 4 of the embodiment wraps
It includes: processor 40, memory 41 and being stored in the computer that can be run in the memory 41 and on the processor 40
Readable instruction 42.The processor 40 realizes the identification side of above-mentioned each risk subscribers when executing the computer-readable instruction 42
Step in method embodiment, such as step 101 shown in FIG. 1 is to 109.Alternatively, the processor 40 execute the computer can
The function of each module/unit in above-mentioned each Installation practice, such as the function of module 301 to 306 shown in Fig. 3 are realized when reading instruction 42
Energy.
Illustratively, the computer-readable instruction 42 can be divided into one or more module/units, one
Or multiple module/units are stored in the memory 41, and are executed by the processor 40, to complete the present invention.Institute
Stating one or more module/units can be the series of computation machine readable instruction section that can complete specific function, the instruction segment
For describing implementation procedure of the computer-readable instruction 42 in the server 4.
The server 4 can be smart phone, notebook, palm PC and cloud server etc. and calculate equipment.It is described
Server 4 may include, but be not limited only to, processor 40, memory 41.It will be understood by those skilled in the art that Fig. 4 is only to take
The example of business device 4, does not constitute the restriction to server 4, may include components more more or fewer than diagram, or combine certain
A little components or different components, such as the server 4 can also include input-output equipment, network access equipment, bus
Deng.
The processor 40 can be central processing unit (CentraL Processing Unit, CPU), can also be
Other general processors, digital signal processor (DigitaL SignaL Processor, DSP), specific integrated circuit
(AppLication Specific Integrated Circuit, ASIC), ready-made programmable gate array (FieLd-
ProgrammabLe Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng.
The memory 41 can be the internal storage unit of the server 4, such as the hard disk or memory of server 4.
The memory 41 is also possible to the External memory equipment of the server 4, such as the plug-in type being equipped on the server 4 is hard
Disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure DigitaL, SD) card, flash card
(FLash Card) etc..Further, the memory 41 can also both include the internal storage unit of the server 4 or wrap
Include External memory equipment.The memory 41 is for storing needed for the computer-readable instruction and the server other
Program and data.The memory 41 can be also used for temporarily storing the data that has exported or will export.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes the complete of each embodiment the method for the present invention
Portion or part steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-OnLy
Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can store journey
The medium of sequence code.
The above, the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;Although referring to before
Stating embodiment, invention is explained in detail, those skilled in the art should understand that: it still can be to preceding
Technical solution documented by each embodiment is stated to modify or equivalent replacement of some of the technical features;And these
It modifies or replaces, 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.
Claims (10)
1. a kind of recognition methods of risk subscribers characterized by comprising
When playing destination multimedia file, the first face image of target user is acquired, and from first face image
Identifying first micro- expressive features, the destination multimedia file is determined according to the personal information of the target user, described the
One face image is each frame face image for including in collected video when playing destination multimedia file;
Described first micro- expressive features and the micro- expressive features of pre-stored benchmark are compared, the micro- expressive features of benchmark
The micro- expressive features generated for normal users when appreciating the destination multimedia file;
If the difference between first micro- expressive features and the micro- expressive features of the benchmark is more than preset threshold value, institute is determined
Stating target user is potential risk user;
Second face image of the target user when face is examined is acquired, and identifies that second is micro- from second face image
Expressive features, each frame face image that second face image includes in collected video when examining for face;
If the target user is potential risk user, described second micro- expressive features are inputted into the first micro- table constructed in advance
Feelings cheat identification model, and whether cheat target user described in the output result judgement of identification model according to described first micro- expression
For risk subscribers;
If the target user is not potential risk user, it is micro- that described second micro- expressive features are inputted to second constructed in advance
Expression cheats identification model, and the target user according to the output result judgement of described second micro- expression fraud identification model is
No is risk subscribers, and first micro- expression fraud identification model and second micro- expression fraud identification model are with different
Risk decision threshold.
2. the recognition methods of risk subscribers according to claim 1, which is characterized in that the destination multimedia file according to
The personal information of the target user determines, specifically includes:
Obtain the identity card scanned copy of the target user;
OCR identification is carried out to the identity card scanned copy, obtains age, gender, nationality and the area of the target user;
Inquiry and the age, gender, the type of the associated multimedia file of nationality and area;
The multimedia file that the type is chosen from the multimedia file library constructed in advance, as the destination multimedia text
Part.
3. the recognition methods of risk subscribers according to claim 1, which is characterized in that when the destination multimedia file is broadcast
After putting, before identifying first micro- expressive features in first face image, further includes:
Determine that the target user appreciates the time of the destination multimedia file according to first face image;
If the time is greater than preset minimum time, execution identifies that first micro- expression is special from first face image
The step of sign.
4. the recognition methods of risk subscribers according to claim 3, which is characterized in that described to be schemed according to first face
As the time for determining that the target user appreciates the destination multimedia file includes:
Count the quantity accounting of face image in first face image;
The target user is calculated according to the playing duration of the quantity accounting and the multimedia file and appreciates the mesh
Mark the time of multimedia file.
5. the recognition methods of risk subscribers according to any one of claim 1 to 4, which is characterized in that it is described will be described
First micro- expressive features and the micro- expressive features of pre-stored benchmark are compared and include:
Primary vector is constructed according to described first micro- expressive features, each element of the primary vector is respectively described first micro-
The number or accounting that each micro- expression that expressive features include occurs;
Secondary vector is constructed according to the micro- expressive features of the benchmark, each element of the secondary vector is respectively that the benchmark is micro-
The number or accounting that each micro- expression that expressive features include occurs;
The distance between the primary vector and the secondary vector are calculated, as described first micro- expressive features and the benchmark
Difference between micro- expressive features.
6. a kind of identification device of risk subscribers characterized by comprising
First micro- expressive features identification module, for acquiring the first face of target user when playing destination multimedia file
Image, and first micro- expressive features are identified from first face image, the destination multimedia file is according to the mesh
The personal information for marking user determines that first face image is when playing destination multimedia file includes in collected video
Each frame face image;
Micro- expressive features comparison module, for carrying out described first micro- expressive features and the micro- expressive features of pre-stored benchmark
Compare, the micro- expressive features of benchmark are micro- expressive features that normal users are generated when appreciating the destination multimedia file;
Potential risk user's determination module, if for the difference between described first micro- expressive features and the micro- expressive features of the benchmark
Not Chao Guo preset threshold value, then determine the target user for potential risk user;
Second micro- expressive features identification module, for acquiring second face image of the target user when face is examined, and from institute
It states and identifies second micro- expressive features in the second face image, second face image wraps in collected video when examining for face
The each frame face image contained;
First risk subscribers determination module, if being potential risk user for the target user, by described second micro- expression
First micro- expression fraud identification model that feature input constructs in advance, and according to the defeated of described first micro- expression fraud identification model
Whether target user described in result judgement is risk subscribers out;
Second risk subscribers determination module, if not being potential risk user for the target user, by described second micro- table
Second micro- expression fraud identification model that the input of feelings feature constructs in advance, and according to described second micro- expression fraud identification model
Export whether target user described in result judgement is risk subscribers, the first micro- expression fraud identification model and described second micro-
Expression, which cheats identification model, has different risk decision thresholds.
7. a kind of computer readable storage medium, the computer-readable recording medium storage has computer-readable instruction, special
Sign is, realizes that the risk as described in any one of claims 1 to 5 is used when the computer-readable instruction is executed by processor
The step of recognition methods at family.
8. a kind of server, including memory, processor and storage can transport in the memory and on the processor
Capable computer-readable instruction, which is characterized in that the processor realizes following steps when executing the computer-readable instruction:
When playing destination multimedia file, the first face image of target user is acquired, and from first face image
Identifying first micro- expressive features, the destination multimedia file is determined according to the personal information of the target user, described the
One face image is each frame face image for including in collected video when playing destination multimedia file;
Described first micro- expressive features and the micro- expressive features of pre-stored benchmark are compared, the micro- expressive features of benchmark
The micro- expressive features generated for normal users when appreciating the destination multimedia file;
If the difference between first micro- expressive features and the micro- expressive features of the benchmark is more than preset threshold value, institute is determined
Stating target user is potential risk user;
Second face image of the target user when face is examined is acquired, and identifies that second is micro- from second face image
Expressive features, each frame face image that second face image includes in collected video when examining for face;
If the target user is potential risk user, described second micro- expressive features are inputted into the first micro- table constructed in advance
Feelings cheat identification model, and whether cheat target user described in the output result judgement of identification model according to described first micro- expression
For risk subscribers;
If the target user is not potential risk user, it is micro- that described second micro- expressive features are inputted to second constructed in advance
Expression cheats identification model, and the target user according to the output result judgement of described second micro- expression fraud identification model is
No is risk subscribers, and first micro- expression fraud identification model and second micro- expression fraud identification model are with different
Risk decision threshold.
9. server according to claim 8, which is characterized in that the destination multimedia file is according to the target user
Personal information determine, specifically include:
Obtain the identity card scanned copy of the target user;
OCR identification is carried out to the identity card scanned copy, obtains age, gender, nationality and the area of the target user;
Inquiry and the age, gender, the type of the associated multimedia file of nationality and area;
The multimedia file that the type is chosen from the multimedia file library constructed in advance, as the destination multimedia text
Part.
10. server according to claim 8 or claim 9, which is characterized in that described by described first micro- expressive features and preparatory
The micro- expressive features of the benchmark of storage, which are compared, includes:
Primary vector is constructed according to described first micro- expressive features, each element of the primary vector is respectively described first micro-
The number or accounting that each micro- expression that expressive features include occurs;
Secondary vector is constructed according to the micro- expressive features of the benchmark, each element of the secondary vector is respectively that the benchmark is micro-
The number or accounting that each micro- expression that expressive features include occurs;
The distance between the primary vector and the secondary vector are calculated, as described first micro- expressive features and the benchmark
Difference between micro- expressive features.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910423191.5A CN110223158B (en) | 2019-05-21 | 2019-05-21 | Risk user identification method and device, storage medium and server |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910423191.5A CN110223158B (en) | 2019-05-21 | 2019-05-21 | Risk user identification method and device, storage medium and server |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110223158A true CN110223158A (en) | 2019-09-10 |
CN110223158B CN110223158B (en) | 2023-08-18 |
Family
ID=67821504
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910423191.5A Active CN110223158B (en) | 2019-05-21 | 2019-05-21 | Risk user identification method and device, storage medium and server |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110223158B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111401197A (en) * | 2020-03-10 | 2020-07-10 | 支付宝(杭州)信息技术有限公司 | Picture risk identification method, device and equipment |
CN111768190A (en) * | 2020-05-29 | 2020-10-13 | 平安银行股份有限公司 | Safe use method and device of self-service equipment and computer equipment |
CN112541411A (en) * | 2020-11-30 | 2021-03-23 | 中国工商银行股份有限公司 | Online video anti-fraud identification method and device |
CN113243918A (en) * | 2021-06-11 | 2021-08-13 | 深圳般若计算机系统股份有限公司 | Risk detection method and device based on multi-mode hidden information test |
CN113407225A (en) * | 2021-06-23 | 2021-09-17 | 未鲲(上海)科技服务有限公司 | Code list generation method and device, computer equipment and storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108734570A (en) * | 2018-05-22 | 2018-11-02 | 深圳壹账通智能科技有限公司 | A kind of Risk Forecast Method, storage medium and server |
CN109697666A (en) * | 2018-12-18 | 2019-04-30 | 深圳壹账通智能科技有限公司 | Method for prewarning risk, device, computer equipment and storage medium |
-
2019
- 2019-05-21 CN CN201910423191.5A patent/CN110223158B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108734570A (en) * | 2018-05-22 | 2018-11-02 | 深圳壹账通智能科技有限公司 | A kind of Risk Forecast Method, storage medium and server |
CN109697666A (en) * | 2018-12-18 | 2019-04-30 | 深圳壹账通智能科技有限公司 | Method for prewarning risk, device, computer equipment and storage medium |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111401197A (en) * | 2020-03-10 | 2020-07-10 | 支付宝(杭州)信息技术有限公司 | Picture risk identification method, device and equipment |
CN111401197B (en) * | 2020-03-10 | 2023-08-15 | 支付宝(杭州)信息技术有限公司 | Picture risk identification method, device and equipment |
CN111768190A (en) * | 2020-05-29 | 2020-10-13 | 平安银行股份有限公司 | Safe use method and device of self-service equipment and computer equipment |
CN112541411A (en) * | 2020-11-30 | 2021-03-23 | 中国工商银行股份有限公司 | Online video anti-fraud identification method and device |
CN113243918A (en) * | 2021-06-11 | 2021-08-13 | 深圳般若计算机系统股份有限公司 | Risk detection method and device based on multi-mode hidden information test |
CN113407225A (en) * | 2021-06-23 | 2021-09-17 | 未鲲(上海)科技服务有限公司 | Code list generation method and device, computer equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN110223158B (en) | 2023-08-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110223158A (en) | A kind of recognition methods of risk subscribers, device, storage medium and server | |
TWI731297B (en) | Risk prediction method and apparatus, storage medium, and server | |
Dibeklioğlu et al. | Recognition of genuine smiles | |
CN109522815B (en) | Concentration degree evaluation method and device and electronic equipment | |
US10810409B2 (en) | Identifying facial expressions in acquired digital images | |
Fu et al. | Learning race from face: A survey | |
Ashraf et al. | The painful face: Pain expression recognition using active appearance models | |
US7912246B1 (en) | Method and system for determining the age category of people based on facial images | |
CN108846694A (en) | A kind of elevator card put-on method and device, computer readable storage medium | |
Li et al. | Stylet2i: Toward compositional and high-fidelity text-to-image synthesis | |
WO2014068567A1 (en) | Method and system for predicting personality traits, capabilities and suggested interactions from images of a person | |
CN105160299A (en) | Human face emotion identifying method based on Bayes fusion sparse representation classifier | |
Marcel et al. | Improving face verification using skin color information | |
CN111639584A (en) | Risk identification method and device based on multiple classifiers and computer equipment | |
Saeed | Facial micro-expressions as a soft biometric for person recognition | |
US10997609B1 (en) | Biometric based user identity verification | |
Song et al. | Dynamic facial models for video-based dimensional affect estimation | |
Mou et al. | Alone versus in-a-group: A multi-modal framework for automatic affect recognition | |
CN112199530B (en) | Multi-dimensional face library picture automatic updating method, system, equipment and medium | |
Bai et al. | Automatic long-term deception detection in group interaction videos | |
Hadid et al. | On the use of dynamic features in face biometrics: recent advances and challenges | |
CN114511912A (en) | Cross-library micro-expression recognition method and device based on double-current convolutional neural network | |
Minkov et al. | A comparison of different features for automatic eye blinking detection with an application to analysis of deceptive behavior | |
Raiman et al. | Move, and i will tell you who you are: detecting deceptive roles in low-quality data | |
Vesnicer et al. | Face recognition using simplified probabilistic linear discriminant analysis |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |