CN109800703A - Risk checking method, device, computer equipment and storage medium based on micro- expression - Google Patents
Risk checking method, device, computer equipment and storage medium based on micro- expression Download PDFInfo
- Publication number
- CN109800703A CN109800703A CN201910042599.8A CN201910042599A CN109800703A CN 109800703 A CN109800703 A CN 109800703A CN 201910042599 A CN201910042599 A CN 201910042599A CN 109800703 A CN109800703 A CN 109800703A
- Authority
- CN
- China
- Prior art keywords
- micro
- video data
- risk
- expression
- pending video
- 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
- 230000014509 gene expression Effects 0.000 title claims abstract description 197
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000003860 storage Methods 0.000 title claims abstract description 17
- 238000012550 audit Methods 0.000 claims abstract description 67
- 230000008921 facial expression Effects 0.000 claims abstract description 25
- 241001269238 Data Species 0.000 claims abstract description 22
- 238000012545 processing Methods 0.000 claims abstract description 12
- 235000021167 banquet Nutrition 0.000 claims description 34
- 238000004590 computer program Methods 0.000 claims description 17
- 239000000284 extract Substances 0.000 claims description 5
- 238000009432 framing Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 description 10
- 238000013075 data extraction Methods 0.000 description 4
- 238000009826 distribution Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012552 review Methods 0.000 description 3
- 208000019901 Anxiety disease Diseases 0.000 description 2
- 230000036506 anxiety Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000036651 mood Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000005266 casting Methods 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses the risk checking methods based on micro- expression, first M pending video datas of acquisition.Sub-frame processing is carried out to each pending video data, obtains N number of micro- facial expression image of each pending video data.N number of micro- facial expression image of each pending video data is input in micro- Expression Recognition model and is identified, N number of micro- Expression Recognition unit of each pending video data is obtained.Micro- Expression Recognition unit of each pending video data is evaluated according to micro- expression unit is preset, obtains the msu message of each pending video data.Risk audit report is finally generated according to the msu message of each pending video data and problem label.In ensure that audit simultaneously efficiency with guaranteeing the validity audited to pending video data.The invention also discloses a kind of, and the risk based on micro- expression audits device and associated computer device and storage medium.
Description
Technical field
The present invention relates to micro- Expression Recognition field more particularly to it is a kind of by the risk checking method of micro- expression, device, based on
Calculate machine equipment and storage medium.
Background technique
With the development of science and technology, existing many enterprises interview or face examine in link can introduce micro- expression come to employee or
Person client carries out Risk Evaluation.For example, the business personnel of financial industry needs when examining with client's progress face in exchange
Information screened.Therefore, lie detection system is begun to appear in these scenes, and existing lie detection system on the market, big portion
Dividing all includes specific physiological data collection equipment, but in financial business scene and is not suitable for.Many enterprises examine ring in face
Also micro- Expression Recognition technology can be introduced in section and do auxiliary reference to examine link for face, however currently use micro- expression as auxiliary
In the application of means, all it is only simply to remind special or abnormal micro- expression, link review efficiency is examined for face
Raising it is very limited.
Summary of the invention
The embodiment of the present invention provides a kind of risk checking method based on micro- expression, device, computer equipment and storage and is situated between
Matter, to solve the problems, such as that it is not high that link review efficiency is examined in face.
A kind of risk checking method based on micro- expression, comprising:
M pending video datas are obtained, each pending video data includes problem label, and M is positive integer；
Sub-frame processing is carried out to each pending video data, obtains the N number of of each pending video data
Micro- facial expression image, N are positive integer；
N number of micro- facial expression image of each pending video data is input in micro- Expression Recognition model and is known
Not, N number of micro- Expression Recognition unit of each pending video data is obtained；
Micro- Expression Recognition unit of each pending video data is matched according to micro- expression unit is preset, is obtained
To the msu message of each pending video data；
Risk audit report is generated according to the msu message of each pending video data and problem label.
A kind of risk audit device based on micro- expression, comprising:
Pending video data obtains module, for obtaining M pending video datas, each pending video counts
According to including problem label, M is positive integer；
Video data framing module obtains each institute for carrying out sub-frame processing to each pending video data
N number of micro- facial expression image of pending video data is stated, N is positive integer；
Micro- Expression Recognition module, for N number of micro- facial expression image of each pending video data to be input to micro- table
It is identified in feelings identification model, obtains N number of micro- Expression Recognition unit of each pending video data；
Micro- expression matching module presets micro- expression unit to micro- expression of each pending video data for basis
Recognition unit is matched, and the msu message of each pending video data is obtained；
Risk audit report generation module, for the msu message and problem mark according to each pending video data
Label generate risk audit report.
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 that the above-mentioned risk based on micro- expression is examined when executing the computer program
Kernel method.
A kind of computer readable storage medium, the computer-readable recording medium storage have computer program, the meter
Calculation machine program realizes the above-mentioned risk checking method based on micro- expression when being executed by processor.
In the above-mentioned risk checking method based on micro- expression, device, computer equipment and storage medium, acquisition M first
Pending video data, each pending video data include problem label；Each pending video data is carried out at framing
Reason, obtains N number of micro- facial expression image of each pending video data；By N number of micro- facial expression image of each pending video data
It is input in micro- Expression Recognition model and is identified, obtain N number of micro- Expression Recognition unit of each pending video data；According to
It presets micro- expression unit to evaluate micro- Expression Recognition unit of each pending video data, obtains each pending video
The msu message of data；Risk audit report is finally generated according to the msu message of each pending video data and problem label
It accuses.It is next further according to the result of micro- Expression Recognition after corresponding image carries out micro- Expression Recognition from being extracted in pending video data
Risk-warning is carried out to pending video data, and is intuitively presented by way of risk audit report, is being guaranteed
It ensure that the validity audit to pending video data the efficiency of audit simultaneously.
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 risk checking method in one embodiment of the invention based on micro- expression；
Fig. 2 is an exemplary diagram of the risk checking method in one embodiment of the invention based on micro- expression；
Fig. 3 is another exemplary diagram of the risk checking method in one embodiment of the invention based on micro- expression；
Fig. 4 is another exemplary diagram of the risk checking method in one embodiment of the invention based on micro- expression；
Fig. 5 is another exemplary diagram of the risk checking method in one embodiment of the invention based on micro- expression；
Fig. 6 is another exemplary diagram of the risk checking method in one embodiment of the invention based on micro- expression；
Fig. 7 is a functional block diagram of the risk audit device in one embodiment of the invention based on micro- expression；
Fig. 8 is another functional block diagram of the risk audit device in one embodiment of the invention based on micro- expression；
Fig. 9 is another functional block diagram of the risk audit device in one embodiment of the invention based on micro- expression；
Figure 10 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.
Risk checking method provided in an embodiment of the present invention based on micro- expression, can be applicable in the application environment such as Fig. 1,
Wherein, client (computer equipment) is communicated by network with server-side.Client sends pending video data to clothes
It is engaged in end, after server-side gets M pending video datas, carrying out sub-frame processing to each pending video data, obtaining
N number of micro- facial expression image of each pending video data；N number of micro- facial expression image of each pending video data is input to micro-
It is identified in Expression Recognition model, obtains N number of micro- Expression Recognition unit of each pending video data；According to presetting micro- table
Feelings unit evaluates micro- Expression Recognition unit of each pending video data, obtains examining for each pending video data
Nuclear information；Risk audit report is generated according to the msu message of each pending video data and problem label.Wherein, client
(computer equipment) can be, but not limited to various personal computers, laptop, smart phone, tablet computer and it is portable can
Wearable device.Server-side can be realized with the server cluster of the either multiple server compositions of independent server.
In one embodiment, it as shown in Fig. 2, providing a kind of risk checking method based on micro- expression, applies in this way
It is illustrated, includes the following steps: for server-side in Fig. 1
S10: M pending video datas are obtained, each pending video data includes problem label, and M is positive integer.
Wherein, pending video data refer to include user's face portion video data, specifically, pending video
Data further comprise the voice data of user.It is which is asked that problem label, which is for distinguishing each pending video data answer,
The indicating label of topic.The problem label can be made of at least one in number, letter, symbol or text.It is understood that
Ground, pending video data and problem label are one-to-one, the i.e. corresponding problem labels of a pending video data.
For example, can for following problems distribute different problems label: " which your kinsfolk has? ", " your income situation is such as
What? " " how is your physical condition? ", be the problem of these three problems are distributed respectively label it is 001,002 and illustratively
003.And pending video data the problem of including label then shows the video data is which correspondence problem user is answering
When video data.For example, problem label is then that in answer, " your kinsfolk has user for 001 pending video data
Which? " the video data of this problem.
In a specific embodiment, client perhaps attends a banquet to issue voice and carry out problem casting or client and exist
Problem is shown by text information on interface.User answers according to each problem, and M pending video counts
According to being the video data recorded to user to the answer of each problem.Specifically, the pending video data
It can also completely be recorded by once by being recorded to obtain respectively by the answer to different problems respectively
Video segmentation is carried out after system or interception obtains, and the number of pending video data is M, and M is positive integer, it is possible to understand that ground, M
Numerical value and the quantity of the problems in review process be consistent.
S20: sub-frame processing is carried out to each pending video data, obtains N number of micro- table of each pending video data
Feelings image, N are positive integer.
In this step, sub-frame processing is carried out to each pending video data to get each pending video data is arrived
N number of micro- facial expression image, N is positive integer.Specifically, after presetting certain frame per second or frame number, according to preset frame
Rate and frame number carry out sub-frame processing to each pending video data to get the N number of of corresponding each pending video data is arrived
Micro- facial expression image.Frame per second or frame number are higher, the quantity of micro- facial expression image is more (i.e. the numerical value of N is bigger), subsequent to obtain
Msu message is more accurate, and correspondingly, calculating to server-side consumption can be higher, and whole efficiency can reduce.It therefore, can be according to reality
Border precision and efficiency need to set frame per second and frame number.
S30: N number of micro- facial expression image of each pending video data is input in micro- Expression Recognition model and is known
Not, N number of micro- Expression Recognition unit of each pending video data is obtained.
Wherein, micro- Expression Recognition model is the network model that preparatory training obtains, for pending video data
Micro- facial expression image identified, and export a recognition result, i.e., micro- Expression Recognition unit.Illustratively, micro- Expression Recognition
Unit is happy, sad, frightened, angry, surprised, detest or contempt.Micro- Expression Recognition model may determine that the picture number of input
Correspond to the probability value of preset a variety of micro- expressions according to middle face, if the probability value of certain micro- expression is more than corresponding default threshold
Value, then obtaining the corresponding micro- expression of the measurement information to be checked is micro- Expression Recognition unit.For example, in the present embodiment, can incite somebody to action
Mood in micro- Expression Recognition model is set as 7 kinds of happy, sad, frightened, angry, surprised, detest and contempt.Specifically, may be used
It is labeled with the great amount of images data that preparatory acquisition respectively represents this 7 kinds of moods, forms image data set, then selection corresponds to
Neural network model or classifier be trained, finally obtain micro- Expression Recognition model.
After obtaining N number of micro- facial expression image of each pending video data, by the N number of of each pending video data
Micro- facial expression image, which is input to, to be identified in micro- Expression Recognition model to get N number of micro- expression of each pending video data is arrived
Recognition unit.Wherein, micro- Expression Recognition unit is used to indicate the maximum probability which kind of micro- expression corresponding micro- facial expression image belongs to.
S40: micro- Expression Recognition unit of each pending video data is matched according to micro- expression unit is preset, is obtained
To the msu message of each pending video data.
Wherein, presetting micro- expression unit is preset specific micro- expression unit, for example, setting is uneasy, nervous, anxiety
Micro- expression with the representatives such as fear is as presetting micro- expression unit.And according to presetting micro- expression unit to each pending video counts
According to micro- Expression Recognition unit carry out matching be whether to be deposited in the micro- Expression Recognition unit for judge each pending video data
With preset the consistent micro- Expression Recognition unit of micro- expression unit, different msu messages is obtained further according to judging result.It examines
Nuclear information is used to reflect an auditing result of corresponding pending video data, and illustratively, msu message may include depositing
In risk and risk is not present.If existing in micro- Expression Recognition unit of pending video data consistent with expression unit is preset as
Micro- Expression Recognition unit, then the msu message of the pending video data be there are risks.If pending video data is micro-
It is not present and is preset as the consistent micro- Expression Recognition unit of expression unit in Expression Recognition unit, then the pending video data
Msu message is that there is no risks.It further, can also be according to micro- Expression Recognition unit in pending video data and pre-
Quantity or the ratio of the consistent micro- Expression Recognition unit of expression unit are set as to divide risk class.In pending video data
Micro- Expression Recognition unit and be preset as the consistent micro- Expression Recognition unit of expression unit quantity it is more, risk class is got over
Height, preferably to make accurate instruction.
S50: risk audit report is generated according to the msu message of each pending video data and problem label.
Risk audit report is the presentation to the whole audit situation of M pending video datas.It specifically, can basis
The difference of the msu message of each pending video data generates different risk audit reports, for example, by statistics M to
Auditing msu message in video data is that there are the quantity of risk to export different risk audit reports, when there are the numbers of risk
When amount reaches certain threshold value, the unacceptable risk audit report of audit can be exported.When there are the quantity of risk to be less than another threshold
When value, the risk audit report that audit passes through can be exported.Furthermore, it is possible to classify to corresponding pending video data,
The msu message which pending video data is indicated according to the difference of msu message is which pending view there are risk
The msu message of frequency evidence is that there is no risks.It is possible to further only export, there are the pending video datas of risk to correspond to
The problem of label.Risk audit report finally is generated based on above- mentioned information, i.e. risk audit report includes to M pending videos
It is that whether the whole audit of data passes through as a result, further including problem label corresponding to different msu messages or msu message is to deposit
In problem label corresponding to the pending video data of risk.
In the present embodiment, M pending video datas are obtained first, and each pending video data includes problem mark
Label；Sub-frame processing is carried out to each pending video data, obtains N number of micro- facial expression image of each pending video data；It will
N number of micro- facial expression image of each pending video data, which is input in micro- Expression Recognition model, to be identified, is obtained each pending
N number of micro- Expression Recognition unit of core video data；According to presetting micro- expression unit to micro- expression of each pending video data
Recognition unit is evaluated, and the msu message of each pending video data is obtained；Finally according to each pending video data
Msu message and problem label generate risk audit report.Corresponding image is extracted from pending video data carries out micro- table
Result after feelings identification further according to micro- Expression Recognition to carry out pending video data Risk-warning, and is examined by risk
The form of core report is intuitively presented, and ensure that audit simultaneously with guaranteeing the validity audited to pending video data
Efficiency.
In one embodiment, as shown in figure 3, according to the micro- expression unit of preset risk to each pending video data
Micro- Expression Recognition unit matched, obtain the msu message of each pending video data, specifically comprise the following steps:
S41: N number of micro- Expression Recognition unit of each pending video data and the micro- expression unit of preset risk are carried out
Matching.
Wherein, the micro- expression unit of risk be it is preset indicate that there are micro- expression units of risk, illustratively, can be with
Micro- expression unit such as uneasy, nervous, anxiety and fear is the micro- expression unit of risk.It is to be appreciated that the example above is only one
Illustrative explanation, the specific micro- expression unit of risk can be set according to different actual needs or scene.
In this step, by N number of micro- Expression Recognition unit of each pending video data and the micro- expression of preset risk
Unit is matched, to obtain the msu message of each pending video data according to matching result.
S42: if the micro- Expression Recognition unit of any one in each pending video data and the micro- expression list of preset risk
First successful match, then the msu message of corresponding pending video data is that there are risks.
N number of micro- Expression Recognition unit of each pending video data is carried out with the micro- expression unit of preset risk one by one
Matching, and micro- Expression Recognition unit and the micro- expression units match of preset risk successfully refer in micro- Expression Recognition unit and preset
The micro- expression unit of risk in any one is identical.If the micro- Expression Recognition unit of any one in pending video data and pre-
If risk micro- expression units match success, then the msu message of corresponding pending video data is that there are risks.Specifically,
It can be by way of string matching or regular expression by micro- Expression Recognition unit and the micro- expression unit of preset risk
It is matched.
S43: if micro- Expression Recognition unit in each pending video data is and the micro- expression unit of preset risk
With failure, then the msu message of corresponding pending video data is that there is no risks.
Micro- Expression Recognition unit and the micro- expression units match of preset risk unsuccessfully refer to micro- Expression Recognition unit and own
The micro- expression unit of preset risk be different from.Micro- Expression Recognition unit in pending video data is and preset risk
Micro- expression units match be unsuccessfully the micro- Expression Recognition unit of each in the pending video data and the micro- expression of risk not
Together.The msu message of pending video data corresponding at this time is that there is no risks.
In the present embodiment, by N number of micro- Expression Recognition unit of each pending video data and the micro- table of preset risk
Feelings unit is matched.If the micro- Expression Recognition unit of any one in each pending video data and the micro- expression of preset risk
Units match success, then the msu message of the corresponding pending video data is that there are risks.If each pending video
Micro- Expression Recognition unit in data fails with the micro- expression units match of preset risk, then the corresponding pending video
The msu message of data is that there is no risks.Guarantee the accuracy of msu message through the above way.
In one embodiment, as shown in figure 4, it is raw according to the msu message of each pending video data and problem label
At risk audit report, specifically comprise the following steps:
S51: msu message is the risk number of the pending video data there are risk in M pending video datas of statistics
Data bulk obtains risk ratio according to risk data quantity.
It is the risk number of the pending video data there are risk by msu message in M pending video datas of statistics
Data bulk, and risk ratio is obtained according to the risk data quantity.Specifically, the risk data quantity obtained after statistics is removed
With the quantity M of pending video data to get arrive risk ratio.
S52: auditing result is obtained according to risk ratio.
After obtaining risk ratio, auditing result is obtained according to the difference of risk ratio.Optionally, auditing result can be with
Pass through and audit including audit and does not pass through.Preferably, auditing result further includes wait check.Specifically, difference can be preset
Ratio section, each ratio section correspond to an auditing result.Then judge that the risk ratio is located at which section, and root
Corresponding auditing result is obtained according to the corresponding ratio section.
It in a specific embodiment, can also be by way of setting ratio threshold value, by by risk ratio and ratio
Example threshold value is compared, to obtain different auditing results.Wherein, the quantity of proportion threshold value is at least one.
S53: being that the label that there is a problem of that the pending video data of risk is corresponding is determined as risk label by msu message.
Risk label is to be used to indicate the pending video data of correspondence problem there are the labels of risk.Specifically, will
Msu message is that the label that there is a problem of that the pending video data of risk is corresponding is determined as risk label.
S54: risk audit report is generated according to auditing result and risk label.
After obtaining auditing result and risk label, risk audit report is generated according to the two.Specifically, risk is examined
The auditing result of the secondary audit is designated in core report, and includes risk label.
In the present embodiment, counting msu message in M pending video datas first is the pending view there are risk
The risk data quantity of frequency evidence obtains risk ratio according to risk data quantity.And auditing result is obtained according to risk ratio.
It is in turn that the label that there is a problem of that the pending video data of risk is corresponding is determined as risk label by msu message.Last basis
Auditing result and the risk label generate risk audit report, guarantee the accuracy of risk audit report output with this, and
Risk auditing result is more intuitively presented.
In one embodiment, as shown in figure 5, obtaining auditing result according to risk ratio, specifically comprise the following steps:
S521: obtaining preset ratio threshold value, and preset ratio threshold value includes the first proportion threshold value and the second proportion threshold value,
In, the second proportion threshold value is greater than the first proportion threshold value.
Wherein, preset ratio threshold value is preset proportional numerical value, for measuring to risk ratio, to obtain pair
The auditing result answered.Preset ratio threshold value includes the first proportion threshold value and the second proportion threshold value, wherein the second proportion threshold value is greater than
First proportion threshold value.Three different auditing results are marked off according to two different proportion threshold values.It is to be appreciated that second
Proportion threshold value is greater than the first proportion threshold value.
S522: if risk ratio is greater than or equal to the second proportion threshold value, then auditing result is that audit does not pass through.
S523: if risk ratio is greater than or equal to the first proportion threshold value and less than the second proportion threshold value, then auditing result is
Wait check.
S524: if risk ratio is less than first proportion threshold value, then auditing result is that audit passes through.
By the way that risk ratio and two proportion threshold values to be compared, to obtain corresponding auditing result.Wherein, wait check
Refer to that needs are further audited.
In the present embodiment, by presetting two preset ratio threshold values of the first proportion threshold value and the second proportion threshold value,
To measure to risk ratio, corresponding auditing result is obtained.It ensure that the efficiency that auditing result obtains.
In one embodiment, as shown in fig. 6, before the step of obtaining M pending video datas, micro- table should be based on
The risk checking method of feelings further includes following steps:
S11: original video data is obtained, audio data is extracted from original video data.
Wherein, original video data is a complete recorded video, includes user to answer video of all the problems
Data also include voice broadcast data.Wherein, audio data refers to the acoustic information in original video data, the audio number
According to the sound of the sound and voice broadcast that include user.Specifically, it can be realized using FFmpeg from from original video number
According to middle extraction audio data.FFmpeg be it is a set of can be used to record, converted digital audio, video, and stream can be translated into
Open source computer program.Illustratively, it can be realized using following sentence:
ffmpeg-i ABC.mp4-f s16le-ar 16000test.wav；
Wherein, ABC.mp4 is original video data, and test.wav is the title for extracting obtained audio data.-ar
16000 refer to that sample rate is 16k, and s16le represents data as 16.It is to be appreciated that above-mentioned suffix and specific data are only one
A illustrative explanation, can specifically be adjusted according to actual needs.
S12: extracting M attend a banquet audio data and M audio user data respectively from audio data, wherein, it is each
Audio data of attending a banquet includes corresponding time interval of attending a banquet, and each audio user data include user time section.
It include the phonological component of phonological component that user issues and sending of attending a banquet in audio data.From audio data
This two parts audio data is extracted respectively, and is split, M attend a banquet audio data and M audio user numbers are obtained
According to.
Specifically, if the voice of the voice for part of attending a banquet and User Part passes through different audio tracks respectively and recorded
System, then can both separate both speech regions according to audio track, and the audio data of each audio track is all interval
A Duan Jingyin.Therefore both voices can be divided by the mute of this part interval respectively, obtains M and attends a banquet
Audio data and M audio user data.
If two kinds of voices are recorded by the same audio track, Application on Voiceprint Recognition or speech recognition area can be first passed through in advance
Separate attend a banquet phonological component and user speech part.And this two parts voice is existing for alternating, therefore natural can distinguish
Obtain M attend a banquet audio data and M audio user data.M attend a banquet audio data and M audio user numbers are obtained in segmentation
When, the corresponding time interval of audio data and audio user data in audio data of attending a banquet further is recorded, that is, is had
The body corresponding period.
S13: it is a pending that M is extracted from original video data according to the user time section of each audio user data
Video data.
After the user time section for obtaining each audio user data in step s 12, according to each user time area
Between M pending video datas are extracted from original video data.I.e. according to each section of user time section from original video
Interception obtains M pending video datas in the correspondence period in data.
In a specific embodiment, the video intercepting to original video data can be realized using FFmpeg.Its
In, FFmpeg be it is a set of can be used to record, converted digital audio, video, and the open source computer journey of stream can be translated into
Sequence.
It specifically, can be using the video intercepting realized to original video data of such as issuing orders:
Order: ffmpeg-i<input>-ss 0-t 10-y<output>
Wherein, it is initial time that ss is corresponding, and t is the duration, the order meaning for intercepted since 0 second 10 seconds when
Between.Can obtain initial time SS according to the left end point in user time section, further according to time interval length obtain continue when
Between, then the two numerical value are input in mentioned order, to execute the video intercepting to original video data.
S14: distributing a problem label for each audio data of attending a banquet, will according to user time section and time interval of attending a banquet
Problem label and pending video data are associated.
A problem label is distributed to each audio data of attending a banquet, for being identified to audio data of attending a banquet.Thereafter according to
Problem label and pending video data are associated by user time section and time interval of attending a banquet.Specifically, from first
It attends a banquet and chooses in time interval and the immediate user time section of the time interval of attending a banquet, and the two time intervals is corresponding
The problem of label and pending video data be associated.Next problem label and pending video data are sequentially carried out, directly
To the association for completing all problem label and pending video data.
In the present embodiment, original video data is obtained first, and audio data is extracted from original video data；Again from sound
Frequency extracts M attend a banquet audio data and M audio user data respectively in, wherein, each audio data of attending a banquet includes
Corresponding time interval of attending a banquet, each audio user data include user time section；According to the use of each audio user data
Family time interval extracts M pending video datas from original video data.It is asked for each audio data distribution one of attending a banquet
Label is inscribed, is associated problem label and pending video data according to user time section and time interval of attending a banquet.Thus
Guarantee the accuracy of obtained M pending video data.
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 risk audit device based on micro- expression is provided, should be audited based on the risk of micro- expression
Risk checking method in device and above-described embodiment based on micro- expression corresponds.As shown in fig. 7, should the wind based on micro- expression
Danger audit device includes that pending video data obtains module 10, video data framing module 20, micro- Expression Recognition module 30, micro-
Expression matching module 40 and risk audit report generation module 50.Detailed description are as follows for each functional module:
Pending video data obtains module 10, for obtaining M pending video datas, each pending video data
Including problem label, M is positive integer；
Video data framing module 20 obtains each pending for carrying out sub-frame processing to each pending video data
N number of micro- facial expression image of core video data, N is positive integer；
Micro- Expression Recognition module 30, for N number of micro- facial expression image of each pending video data to be input to micro- expression
It is identified in identification model, obtains N number of micro- Expression Recognition unit of each pending video data；
Micro- expression matching module 40 is preset micro- expression unit for basis and is known to micro- expression of each pending video data
Other unit is matched, and the msu message of each pending video data is obtained；
Risk audit report generation module 50, for the msu message and problem label according to each pending video data
Generate risk audit report.
Preferably, as shown in figure 8, micro- expression matching module 40 includes micro- expression matching unit 41 and matching result feedback
Member 42.
Micro- expression matching unit 41, for by N number of micro- Expression Recognition unit of each pending video data and preset
The micro- expression unit of risk is matched；
Matching result feedback unit 42, in each pending video data any one micro- Expression Recognition unit and
When the preset micro- expression units match of risk succeeds, then the msu message of corresponding pending video data is that there are risks；?
When micro- Expression Recognition unit in each pending video data is and the micro- expression units match of preset risk fails, then correspond to
Pending video data msu message be there is no risks.
Preferably, as shown in figure 9, risk audit report generation module 50 includes risk ratio acquiring unit 51, audit knot
Fruit acquiring unit 52, risk tag determination unit 53 and risk audit report generation unit 54.
Risk ratio acquiring unit 51, for count msu message in M pending video datas be there are risk to
The risk data quantity for auditing video data, obtains risk ratio according to risk data quantity；
Auditing result acquiring unit 52, for obtaining auditing result according to risk ratio；
Risk tag determination unit 53, for being there is a problem of that the pending video data of risk is corresponding by msu message
Label is determined as risk label；
Risk audit report generation unit 54, for generating risk audit report according to auditing result and risk label.
Preferably, auditing result acquiring unit 52 includes the first ratio for obtaining preset ratio threshold value, preset ratio threshold value
Example threshold value and the second proportion threshold value, wherein the second proportion threshold value is greater than the first proportion threshold value；If risk ratio is greater than or equal to the
Two proportion threshold values, then auditing result is that audit does not pass through；If risk ratio is greater than or equal to the first proportion threshold value and less than second
Proportion threshold value, then auditing result is wait check；If risk ratio is less than first proportion threshold value, then auditing result is that audit is logical
It crosses.
It preferably, should further include obtaining original video data, audio data extraction based on the risk audit device of micro- expression
Module, pending video data extraction module and problem label distribution module.
Original video data is obtained, audio data is extracted from original video data；
Audio data extraction module, for extracting M attend a banquet audio data and M user's sounds respectively from audio data
Frequency evidence, wherein, each audio data of attending a banquet includes corresponding time interval of attending a banquet, when each audio user data include user
Between section；
Pending video data extraction module, for according to the user time sections of each audio user data from original view
Frequency extracts M pending video datas in；
Problem label distribution module, for distributing a problem label for each audio data of attending a banquet, according to user time area
Between and time interval of attending a banquet problem label and pending video data are associated.
Specific restriction about the risk audit device based on micro- expression may refer to above for based on micro- expression
The restriction of risk checking method, details are not described herein.Modules in the above-mentioned risk audit device based on micro- expression can be complete
Portion or part are realized by software, hardware and combinations thereof.Above-mentioned each module can be embedded in the form of hardware or independently of calculating
In processor in machine equipment, it can also be stored in a software form in the memory in computer equipment, in order to processor
It calls and executes the corresponding operation of the above modules.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 10.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 for the data that the above-mentioned risk checking method based on micro- expression uses.The net of the computer equipment
Network interface is used to communicate with external terminal by network connection.To realize a kind of base when the computer program is executed by processor
In the risk checking method of micro- expression.
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 risk checking method of feelings.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program realizes the risk checking method in above-described embodiment based on micro- expression when being executed by processor.
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 risk checking method based on micro- expression characterized by comprising
M pending video datas are obtained, each pending video data includes problem label, and M is positive integer；
Sub-frame processing is carried out to each pending video data, obtains N number of micro- table of each pending video data
Feelings image, N are positive integer；
N number of micro- facial expression image of each pending video data is input in micro- Expression Recognition model and is identified, is obtained
To N number of micro- Expression Recognition unit of each pending video data；
Micro- Expression Recognition unit of each pending video data is matched according to micro- expression unit is preset, is obtained every
The msu message of the one pending video data；
Risk audit report is generated according to the msu message of each pending video data and problem label.
2. as described in claim 1 based on the risk checking method of micro- expression, which is characterized in that the basis presets micro- expression
Unit matches micro- Expression Recognition unit of each pending video data, obtains each pending video counts
According to msu message, specifically include:
By N number of micro- Expression Recognition unit of each pending video data and the micro- expression unit progress of preset risk
Match；
If any one of micro- Expression Recognition unit and the micro- expression list of preset risk in each pending video data
First successful match, then the msu message of the corresponding pending video data is that there are risks；
If micro- Expression Recognition unit in each pending video data is and the micro- expression unit of preset risk
With failure, then the msu message of the corresponding pending video data is that there is no risks.
3. as claimed in claim 2 based on the risk checking method of micro- expression, which is characterized in that it is described according to it is each it is described to
The msu message and problem label for auditing video data generate risk audit report, specifically include:
Msu message is the risk data number of the pending video data there are risk in M pending video datas of statistics
Amount, obtains risk ratio according to the risk data quantity；
Auditing result is obtained according to the risk ratio；
It is that the label that there is a problem of that the pending video data of risk is corresponding is determined as risk label by the msu message；
Risk audit report is generated according to the auditing result and the risk label.
4. as claimed in claim 3 based on the risk checking method of micro- expression, which is characterized in that described according to the Hazard ratio
Example obtains auditing result, specifically includes:
Preset ratio threshold value is obtained, the preset ratio threshold value includes the first proportion threshold value and the second proportion threshold value, wherein second
Proportion threshold value is greater than the first proportion threshold value；
If the risk ratio is greater than or equal to second proportion threshold value, the auditing result is that audit does not pass through；
If the risk ratio is greater than or equal to first proportion threshold value and less than the second proportion threshold value, the auditing result
For wait check；
If the risk ratio is less than first proportion threshold value, the auditing result is that audit passes through.
5. as described in claim 1 based on the risk checking method of micro- expression, which is characterized in that pending at the acquisition M
Before the step of core video data, the risk checking method based on micro- expression further include:
Original video data is obtained, extracts audio data from the original video data；
Extract M attend a banquet audio data and M audio user data respectively from the audio data, wherein each described
Audio data of attending a banquet includes corresponding time interval of attending a banquet, and each audio user data include user time section；
It is a pending that M is extracted from the original video data according to the user time section of each audio user data
Video data；
A problem label is distributed for each audio data of attending a banquet, according to user time section and time interval of attending a banquet by problem
Label and pending video data are associated.
6. a kind of risk based on micro- expression audits device characterized by comprising
Pending video data obtains module, for obtaining M pending video datas, each pending video data packet
Problem label is included, M is positive integer；
Video data framing module, for carrying out sub-frame processing to each pending video data, obtain it is each it is described to
N number of micro- facial expression image of video data is audited, N is positive integer；
Micro- Expression Recognition module is known for N number of micro- facial expression image of each pending video data to be input to micro- expression
It is identified in other model, obtains N number of micro- Expression Recognition unit of each pending video data；
Micro- expression matching module presets micro- expression unit to micro- Expression Recognition of each pending video data for basis
Unit is matched, and the msu message of each pending video data is obtained；
Risk audit report generation module, for raw according to the msu message and problem label of each pending video data
At risk audit report.
7. the risk based on micro- expression audits device as claimed in claim 6, which is characterized in that micro- expression matching module
Include:
Micro- expression matching unit, for by the N number of micro- Expression Recognition unit and preset wind of each pending video data
The micro- expression unit in danger is matched；
Matching result feedback unit, for any one of micro- Expression Recognition unit in each pending video data
When expression units match success micro- with preset risk, then the msu message of the corresponding pending video data is that there are wind
Danger；Micro- Expression Recognition unit in each pending video data is and the micro- expression units match of preset risk
When failure, then the msu message of the corresponding pending video data is that there is no risks.
8. the risk based on micro- expression audits device as claimed in claim 7, which is characterized in that the risk audit report is raw
Include: at module
Risk ratio acquiring unit is that there are the pending of risk for counting msu message in the M pending video datas
The risk data quantity of core video data obtains risk ratio according to the risk data quantity；
Auditing result acquiring unit, for obtaining auditing result according to the risk ratio；
Risk tag determination unit, for being the mark that there is a problem of that the pending video data of risk is corresponding by the msu message
Label are determined as risk label；
Risk audit report generation unit, for generating risk audit report according to the auditing result and the risk label.
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
Based on the risk checking method of micro- expression described in 5 any one.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In risk of the realization based on micro- expression as described in any one of claim 1 to 5 is examined when the computer program is executed by processor
Kernel method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910042599.8A CN109800703A (en) | 2019-01-17 | 2019-01-17 | Risk checking method, device, computer equipment and storage medium based on micro- expression |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910042599.8A CN109800703A (en) | 2019-01-17 | 2019-01-17 | Risk checking method, device, computer equipment and storage medium based on micro- expression |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109800703A true CN109800703A (en) | 2019-05-24 |
Family
ID=66559451
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910042599.8A Pending CN109800703A (en) | 2019-01-17 | 2019-01-17 | Risk checking method, device, computer equipment and storage medium based on micro- expression |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109800703A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110415108A (en) * | 2019-07-29 | 2019-11-05 | 中国工商银行股份有限公司 | Method for processing business and device, electronic equipment and computer readable storage medium |
| CN111897998A (en) * | 2020-07-10 | 2020-11-06 | 苏州跃盟信息科技有限公司 | Video processing method and processing system |
| CN114282882A (en) * | 2021-11-15 | 2022-04-05 | 深圳市东信时代信息技术有限公司 | Material work order auditing method, device and equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160306870A1 (en) * | 2015-04-14 | 2016-10-20 | Algoscent | System and method for capture, classification and dimensioning of micro-expression temporal dynamic data into personal expression-relevant profile |
| CN108537160A (en) * | 2018-03-30 | 2018-09-14 | 平安科技（深圳）有限公司 | Risk Identification Method, device, equipment based on micro- expression and medium |
| 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 |
| CN108765131A (en) * | 2018-05-18 | 2018-11-06 | 深圳壹账通智能科技有限公司 | Credit authorization method, apparatus, terminal and readable storage medium storing program for executing based on micro- expression |
| CN109165608A (en) * | 2018-08-30 | 2019-01-08 | 深圳壹账通智能科技有限公司 | The micro- expression recognition method of multi-angle of view, device, storage medium and computer equipment |
-
2019
- 2019-01-17 CN CN201910042599.8A patent/CN109800703A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160306870A1 (en) * | 2015-04-14 | 2016-10-20 | Algoscent | System and method for capture, classification and dimensioning of micro-expression temporal dynamic data into personal expression-relevant profile |
| CN108537160A (en) * | 2018-03-30 | 2018-09-14 | 平安科技（深圳）有限公司 | Risk Identification Method, device, equipment based on micro- expression and medium |
| CN108765131A (en) * | 2018-05-18 | 2018-11-06 | 深圳壹账通智能科技有限公司 | Credit authorization method, apparatus, terminal and readable storage medium storing program for executing based on micro- expression |
| 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 |
| CN109165608A (en) * | 2018-08-30 | 2019-01-08 | 深圳壹账通智能科技有限公司 | The micro- expression recognition method of multi-angle of view, device, storage medium and computer equipment |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110415108A (en) * | 2019-07-29 | 2019-11-05 | 中国工商银行股份有限公司 | Method for processing business and device, electronic equipment and computer readable storage medium |
| CN111897998A (en) * | 2020-07-10 | 2020-11-06 | 苏州跃盟信息科技有限公司 | Video processing method and processing system |
| CN114282882A (en) * | 2021-11-15 | 2022-04-05 | 深圳市东信时代信息技术有限公司 | Material work order auditing method, device and equipment and storage medium |
| CN114282882B (en) * | 2021-11-15 | 2023-04-14 | 深圳市东信时代信息技术有限公司 | Material work order auditing method, device and equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112333549B (en) | Video processing method and device, electronic equipment and storage medium | |
| CN111741356B (en) | Quality inspection method, device and equipment for double-recording video and readable storage medium | |
| WO2021000678A1 (en) | Business credit review method, apparatus, and device, and computer-readable storage medium | |
| CN112346567B (en) | Virtual interaction model generation method and device based on AI (Artificial Intelligence) and computer equipment | |
| CN110572601B (en) | Double-recording video recording system with real-time checking function | |
| CN109801105A (en) | Service methods of marking, device, equipment and storage medium based on artificial intelligence | |
| CN110619568A (en) | Risk assessment report generation method, device, equipment and storage medium | |
| CN110533288A (en) | Business handling process detection method, device, computer equipment and storage medium | |
| CN109800703A (en) | Risk checking method, device, computer equipment and storage medium based on micro- expression | |
| CN111861732A (en) | Risk assessment system and method | |
| CN109360550A (en) | Test method, device, equipment and the storage medium of voice interactive system | |
| CN112233690B (en) | Double recording method, device, terminal and storage medium | |
| CN109299094A (en) | Tables of data processing method, device, computer equipment and storage medium | |
| CN109766859A (en) | Campus monitoring method, device, equipment and storage medium based on micro- expression | |
| US20230177799A1 (en) | Relationship modeling and adjustment based on video data | |
| CN110598008B (en) | Method and device for detecting quality of recorded data and storage medium | |
| CN110473049A (en) | Finance product recommended method, device, equipment and computer readable storage medium | |
| CN109509088A (en) | Loan checking method, device, equipment and medium based on micro- Expression Recognition | |
| CN111783432A (en) | Generation method and device of credit certificate examination key point list | |
| CN109816141A (en) | Product data prediction technique, device, computer equipment based on data analysis | |
| CN109766772A (en) | Risk control method, device, computer equipment and storage medium | |
| CN108509041A (en) | Method and apparatus for executing operation | |
| CN114566159A (en) | Electric ticket circulation method and device based on checking operation | |
| CN110378587A (en) | Intelligent quality detecting method, system, medium and equipment | |
| CN114639152A (en) | Multi-modal voice interaction method, device, equipment and 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 |