CN107463879A - Human bodys' response method based on deep learning - Google Patents
Human bodys' response method based on deep learning Download PDFInfo
- Publication number
- CN107463879A CN107463879A CN201710544873.2A CN201710544873A CN107463879A CN 107463879 A CN107463879 A CN 107463879A CN 201710544873 A CN201710544873 A CN 201710544873A CN 107463879 A CN107463879 A CN 107463879A
- Authority
- CN
- China
- Prior art keywords
- sample
- mrow
- msub
- track
- msup
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The present invention relates to captcha fields, Human bodys' response method more particularly to based on deep learning, the inventive method realizes the judgement of current page operating main body using the neutral net comprising LSTM, in structure machine behavior sample based on human body behavior sample, employ 4 kinds of generating modes, including random generation, positive sample track is cut into some subsegments at random again to be spliced the subsegment after cutting at random, the proportional zoom, disturbance etc. of trajectory parameters is carried out on the basis of positive sample;For such negative sample producing method based on positive sample, to have higher similitude with positive sample, confusion is stronger, and training sample, which trains the neutral net come, has higher resolution capability.This method realizes the identification of the current page operating main body based on LSTM networks first, and the inventive method is particularly suitable for the checking in dragging image hotpoint operating main body, the business scenario judged operating main body.
Description
Technical field
The present invention relates to captcha fields, the Human bodys' response method more particularly to based on deep learning.
Background technology
Network today technology is more and more flourishing, and network application is more and more, as various websites, Email, blog, electronics
Government websites etc. have turned into the necessity of everybody daily life.But along with fast development internet, network security also into
The problem of increasingly being highlighted for one.Particularly the automatic registering and logging of rogue program, maliciously pour water, with specific program Brute Force
The network security attacks such as account and password.That currently register or access the webpage to avoid these generations from being identified with regard to needs is people
Or program.Most common captcha (Completely Automated Public Turing Test to Tell
Computers and Humans Apart automatically distinguish the abbreviation of computer and the turing test of the mankind), it is in 1997
Invented by Mark D.Lillibridge et al..At present because of its huge meaning, each website is widely used in.Very simultaneously
More academic institutions and commercial company are also studied to it.
Captcha common at present has based on computer vision, including character recognition and object identification.These identifications are asked
Topic is often fairly simple for people but more difficult for computer, so as to distinguish people or program.Also it is based on human body
Behavior, including tap keyboard and mobile mouse.These behaviors of human body possess certain characteristic rule, can be special by these
It is people or program to levy distinguish operation computer.With the development of deep learning, computer vision also leads to a leap formula
It is progressive, either character recognition or object identification, the recognition accuracy more and more higher of computer.Most common base before causing
It is gradually reduced in the captcha of computer vision defence capability.And the captcha based on human body behavior starts to occur.Using
Captcha network security mean of defenses based on human body behavior start to be widely used, and net is being carried out using this kind of means
When network protects, the recognition capability of human body behavior or machine behavior is just needed to have first, but based on traditional machine learning side
The judging nicety rate that method is first classified again to human body behavior extraction feature is not high.Because some of human body behavior are characterized in profound level
Feature, it is difficult to extract this feature by the rule artificially formulated.
The content of the invention
It is an object of the invention to overcome the above-mentioned deficiency in the presence of prior art, there is provided the human body based on deep learning
Activity recognition method, developed on the basis of human body behavior and then generate substantial amounts of machine Behavioral training sample so that machine
Device Behavioral training sample it is higher with the similarity of human body behavior, the coverage mode of training sample is more comprehensively.Pass through the people of magnanimity
Training sample of both body behavior and machine behavior carrys out the nerve that training package contains LSTM (Long Short-Term Memory)
Network model;And the neural network model by training judges whether new input behavior is human body behavior.Methods described is special
Suitable for drag figure realize checking identifying code identification system operating main body judgement application scenarios.
In order to realize foregoing invention purpose, based on the Human bodys' response method of deep learning, using human body behavior and
The mouse drag movement locus training sample of production of machinery carrys out the neural network model of training package network containing LSTM, and uses training
Neural network model afterwards judges whether the operating main body of current page is human body behavior.
Methods described includes implemented below step:
(1) structure includes the neural network model of LSTM networks;
(2) magnanimity human body behavior sample and the sample of machine behavior are obtained, the machine behavior sample is in human body behavior sample
Generated on the basis of this, human body behavior sample is as positive sample, and machine behavior sample is as negative sample;
(3) neural network model built is trained using positive and negative samples;
(4) it is people or machine by the operating main body of the neural network model current page to judge trained.
Specifically, in the step (2), the single order difference (dx of mouse track information is usedi, dyi, dti) describe to operate
Behavior, wherein dxi=xi-xi-1, dyi=yi-yi-1, dti=ti-ti-1;Wherein xiFor the abscissa in screen position of mouse, yi
For ordinate of the mouse in screen position, tiFor time information.
The mouse movement trace information that the positive sample of training recorded from people when browsing webpage.
Further, in the step (2), negative sample is generated by machine, and the mode of generation is as follows:
A, in the range of the maximum occurrences of setting, path length is randomly generated, it is random to generate (dxi, dyi, dti);
B, concentrated in positive sample and extract N bars track, the trajectory random being drawn into is divided into n subsegment, then will be divided into
Subsegment random groups be spliced into new track;
C, concentrated in positive sample and extract M bars track;Calculate the transverse shifting of each track always distance sum (dxi), longitudinal direction move
Dynamic always distance sum (dyi) and mobile total time sum (dti);It is random to generate transverse shifting always distance sum (dxi) ', vertically move
Always distance sum (dyi) ' and mobile total time sum (dti) ', the characterising parameter of new movement locus is generated using below equation:
Wherein dxi′、dyi′、dti' be respectively new track lateral coordinates, longitudinal coordinate, the single order difference of time;
D, K bars track is extracted in positive sample;To dxi, dyi, dtiThe disturbance of [- 0.5,0.5] again is randomly generated respectively, is obtained
Obtain the characterising parameter of movement locus newly;
The sample that a, b, c, d mode are generated is mixed, as negative sample collection.
Further, in the step (3), in the training process of the neural network model, the onwards transmission mistake of signal
Journey is as follows:
By (the dx of training samplei, dyi, dti) signal input embedding layers, by embedding layers by dxi, dyi, dti
M dimensional vectors corresponding to changing into respectively, and by dxi, dyi, dtiCorresponding m dimensional vectors are spliced into the vector of a 3m dimension;Will be to
Amount is arranged in the LSTM neutral nets of 3m dimensions according to the sequential input hidden layer latitude of movement locus, is exported by LSTM neutral nets
3m*L vector into full articulamentum, and by softmax layers export this track whether be human body behavior judged result;Its
Middle L is the path length set.
Further, before neural network model training is carried out, including data are optimized with the process of processing:The place
Reason includes:
In the step (2), preceding 100 (x are usedi, yi, ti) it is used as the characterising parameter of mouse movement track;Work as mouse
It is shorter to mark displacement, when movable information is inadequate 100, with (0,0,0) polishing to 100;That is L=100.
In the step (2), by dxi, dyiValue be converted into integer between [- 50,50], by dtiValue conversion
For (0,200] between integer, when actual value beyond set span when, replaced using boundary value.
Further, the sample size included in the positive sample collection is no less than 5000, concentrates and selects in positive sample
80% sample selects 20% sample as test sample as training sample.
Further, the sample size included in the negative sample collection is no less than 5000, concentrates and selects in negative sample
80% sample selects 20% sample as test sample as training sample.
Compared with prior art, beneficial effects of the present invention:The present invention provides the Human bodys' response based on deep learning
Method, the inventive method realize the judgement of current page operating main body, the inventive method using the neutral net comprising LSTM
When building machine behavior sample based on human body behavior sample, 4 kinds of generating modes, including random generation are employed, will just
Sample trace is cut into some subsegments and is again spliced the subsegment after cutting at random at random, is entered on the basis of positive sample
The proportional zoom of row trajectory parameters, disturbance etc.;Such negative sample producing method, it is based on positive sample, has with positive sample
Higher similitude, confusion is stronger, and the neutral net come is trained using such training sample has higher explanation energy
Power.The inventive method realizes the identification of the current page operating main body based on LSTM networks first, and LSTM networks are a kind of
Time recurrent neural network, it is suitable for being spaced in processing and predicted time sequence and postponing relatively long critical event, this hair
Bright method is particularly suitable for the checking in dragging image hotpoint operating main body, the business scenario judged operating main body.
Brief description of the drawings:
Fig. 1 is to realize step schematic diagram based on the Human bodys' response method of deep learning.
Embodiment
With reference to test example and embodiment, the present invention is described in further detail.But this should not be understood
Following embodiment is only limitted to for the scope of the above-mentioned theme of the present invention, it is all that this is belonged to based on the technology that present invention is realized
The scope of invention.
It is an object of the invention to overcome the above-mentioned deficiency in the presence of prior art, there is provided the human body based on deep learning
Activity recognition method, the mouse drag movement locus training sample using human body behavior and production of machinery carry out training package net containing LSTM
The neural network model of network, and judge whether the operating main body of current page is human body using the neural network model after training
Behavior.
Methods described includes implemented below step as shown in Figure 1:
(1) structure includes the neural network model of LSTM networks;The neural network model includes:Embedding layers,
LSTM, full articulamentum and softmax layers.Wherein embedding layers will input discrete signal therein and change into continuous reality
Number vector, the vector after the conversion of embedding layers is input in LSTM according to sequential, operation behavior will be described by LSTM
Several time series vectors be spliced into a high dimension vector after be input in full articulamentum, and by after full articulamentum dimensionality reduction to
Amount is input in softmax layers;The LSTM neutral nets that the inventive method uses are a kind of time recurrent neural networks, are suitable for
It is spaced in processing and predicted time sequence and postpones relatively long critical event.LSTM is different from RNN place, essentially consists in
One is added in algorithm and judges that whether useful information is " processor ", the structure of this processor effect is referred to as cell.One
Three fan doors have been placed among individual cell, has been called input gate respectively, forgets door and out gate.When an information enters LSTM net
Among network, it can be judged whether according to rule useful.Only meeting the information of algorithm certification can just leave, and the information not being inconsistent is then
Passed into silence by forgeing door, selectivity is more embodied when being handled in information, treatment effeciency is higher, solves RNN neutral nets not
The long sequence Dependence Problem that can be realized very well.
(2) magnanimity human body behavior sample and the sample of machine behavior are obtained, human body behavior sample is as positive sample, machine row
It is sample as negative sample;Sample size included in the positive sample collection is no less than 5000, concentrates and selects in positive sample
80% sample selects 20% sample as test sample as training sample.Sample included in the negative sample collection
Quantity is no less than 5000, concentrates the sample of selection 80% to select 20% sample to be used as training sample and survey in negative sample
Sample sheet.
(3) neural network model built is trained using positive and negative samples;The training of neural network model is using forward
Backward algorithm.After in the rate of accuracy reached in test sample collection to the threshold value set, it is possible to think neural network model
Training is completed.
(4) it is people or machine by the operating main body of the neural network model current page to judge trained.
Specifically, in the step (2), the single order difference (dx of mouse movement trace information is usedi, dyi, dti) describe
Operation behavior, wherein dxi=xi-xi-1, dyi=yi-yi-1, dti=ti-ti-1;Wherein xiFor the horizontal seat in screen position of mouse
Mark, yiFor ordinate of the mouse in screen position, tiFor time information.The positive sample of training is remembered from people when browsing webpage
The mouse movement trace information recorded.Mouse movement track can very easily be collected by information by network front end function,
Position and time information of the cursor of mouse in screen can be returned to during mouse is dragged by the function, can be with
(x1, y1, t1)、(x2, y2, t2)、(x3, y3, t3)…(xn, yn, tn) form return, the present invention use mouse movement trace information
Single order difference (dxi, dyi, dti) corresponding operation behavior is described, it can reflect mouse of the mouse in moving process right
In each small period answered, in the translational speed of transverse direction, and lengthwise travel rate, and mobile lateral displacement and length travel feelings
Condition, embody fine feature when operator is operated.
When carrying out neural metwork training, it is necessary to the training sample of magnanimity, if human body Behavioral training sample is using manual
If exclusively carrying out collection, substantial amounts of manpower will be expended, and by setting picture to verify to gather people in web terminal in the existing stage
Body behavior, volunteers drag these identifying codes and realize checking, such accelerated accumulation sample money when logging in or browsing webpage
Source, improve operational efficiency.
Further, in the step (2), negative sample is generated by machine, and the mode of generation is as follows:
A, in the range of the maximum occurrences of setting, path length is randomly generated, it is random to generate (dxi, dyi, dti);
B, N bars tracks, such as 2000 are extracted in positive sample, the trajectory random being drawn into is divided into n (such as 3-
10) subsegment, then the thousands of subsegment random groups formed after segmentation are spliced into new track;
C, M bars track is extracted in positive sample;Calculate the transverse shifting of track always distance sum (dxi), vertically move it is total
Distance sum (dyi) and mobile total time sum (dti);It is random to generate transverse shifting always distance sum (dxi) ', vertically move always away from
From sum (dyi) ' and mobile total time sum (dti) ', the characterising parameter of new movement locus is generated using below equation:
Wherein dxi′、dyi′、dti' be respectively new track lateral coordinates, longitudinal coordinate, the single order difference of time;
D, K bars track is extracted in positive sample;To dxi, dyi, dtiThe disturbance of [- 0.5,0.5] again is randomly generated respectively, is obtained
Obtain the characterising parameter of movement locus newly;
The inventive method uses the negative sample that a, b, c, d mode generate, and fully positive sample feature is combined, compared to direct
The sample generated at random, has a simulation degree higher to human body behavior, thus by such negative sample train come god
There is higher identification capability through network.
The sample that a, b, c, d mode are generated respectively selection 2500, forms negative sample collection.Produced relative to single mode
Raw pays sample, and negative sample concentrates the sample generated comprising 4 kinds of modes so that negative sample collection has larger coverage.
Further, before neural network model training is carried out, including data are optimized with the process of processing:The place
Reason includes:
In the step (2), preceding 100 (x are usedi, yi, ti) it is used as the characterising parameter of mouse movement track;Work as mouse
It is shorter to mark displacement, when movable information is inadequate 100, with (0,0,0) polishing to 100.The every track manually dragged
Length may be different, the motion track information (x of mouse1, y1, t1)、(x2, y2, t2)、(x3, y3, t3)…(xn, yn, tn), middle n can
It can be more than or < 100, uniform length will be arranged to for the track trained and classified, meet the need that neural network parameter is set
Will.The form of every track is [(dx1, dy1, dt1), (dx2, dy2, dt2) ..., (dx100, dy100, dt100)], if track is grown
Inadequate 100 are spent, then with 0 filling, i.e. [(dx1, dy1, dt1), (dx2, dy2, dt2) ..., (0,0,0)].
For the ease of analyzing dxi, dyiValue be converted into integer between [- 49,50], by dtiValue be converted to
(0,200] between integer, when actual value beyond set span when, replaced using boundary value, by dxi, dyi, dti
Value to switch into integer be to calculate for convenience, such as (dxi, dyi, dti) actual value for (60.0, -75.3,
300.3), the numerical value change after integer processing and span limit is (50, -50,200).
Further, in the step (3), in the training process of the neural network model, the onwards transmission mistake of signal
Journey is as follows:By (the dx of training samplei, dyi, dti) signal input embedding layers, discrete (dx, dy, dt) is passed through
Embedding changes into continuous real number vector.Implementation process is:By dx=dx+50, dy=dy+50, dx and dy scope are put down
Move on to [1,100], embedding process is first dx, and dy, dt change into the one-hot vectors of 101 dimensions respectively, then multiply respectively
With the lookup_table matrixes of [101,10], by dx, dy, dt distinguish embedding into the vectors of 10 dimensions, then by these three to
Amount splicing obtains input of 30 dimensional vectors as lstm.
The vector of 30 dimensions is arranged in the LSTM neutral nets of 30 dimensions according to sequential input hidden layer dimension, by LSTM nerves
The expression vector of this movement locus of the dimension of network output 3000 exports this rail into full articulamentum, and by softmax layers
Mark whether be human body behavior judged result.
The mark result of the preceding backward algorithm of the training process use classics of neutral net, judged result and training sample has
During deviation, signal successively adjusts weight coefficient according to loss function back-propagation.
The inventive method, which uses, intersects entropy loss as loss function.Loss function is carried out using stochastic gradient descent method
Optimization.Final mask classification accuracy on checking sample set reaches 95%.
Claims (10)
1. the Human bodys' response method based on deep learning, it is characterised in that produced using human body behavior and machine behavior
Mouse movement track training sample carry out the neural network model of training package network containing LSTM, and use the neutral net after training
Model judges whether the operating main body of current page is human body behavior.
2. the method as described in claim 1, it is characterised in that include implemented below step:
(1) structure includes the neural network model of LSTM networks;
(2) obtain magnanimity human body behavior sample and machine behavior sample, wherein machine behavior sample are made a living with human body behavior sample
Into basis;Human body behavior sample is as positive sample, and machine behavior sample is as negative sample;
(3) neural network model built is trained using positive and negative samples;
(4) judge that the operating main body of current page is people or machine by the neural network model trained.
3. the method as described in claim 1, it is characterised in that in the step (2), use a jump of mouse track information
It is worth (dxi, dyi, dti) operation behavior, wherein dx describedi=xi-xi-1, dyi=yi-yi-1, dti=ti-ti-1, xiFor mouse
Abscissa value in screen position, yiFor ordinate value of the mouse in screen position, tiFor time information.
4. method as claimed in claim 3, it is characterised in that in the step (2), negative sample is generated by machine, generation
Mode is as follows:
A, in the range of the maximum occurrences of setting, path length is randomly generated, it is random to generate (dxi, dyi, dti);
B, concentrated in positive sample and extract N bars track, the trajectory random being drawn into is divided into n subsegment, the subsegment that will be divided into
Random groups are spliced into new track;
C, M bars track is extracted in positive sample;Calculate the transverse shifting of corresponding track always distance sum (dxi), vertically move it is total
Distance sum (dyi) and mobile total time sum (dti);It is random to generate transverse shifting always distance sum (dxi) ', vertically move always away from
From sum (dyi) ' and mobile total time sum (dti) ', the characterising parameter (dx ' of new movement locus is generated using below equationi,
dy′i, dti′):
<mrow>
<msup>
<msub>
<mi>dx</mi>
<mi>i</mi>
</msub>
<mo>&prime;</mo>
</msup>
<mo>=</mo>
<msub>
<mi>dx</mi>
<mi>i</mi>
</msub>
<mfrac>
<mrow>
<mi>s</mi>
<mi>u</mi>
<mi>m</mi>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>dx</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>,</mo>
</msup>
</mrow>
<mrow>
<mi>s</mi>
<mi>u</mi>
<mi>m</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>dx</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
<mrow>
<msup>
<msub>
<mi>dy</mi>
<mi>i</mi>
</msub>
<mo>&prime;</mo>
</msup>
<mo>=</mo>
<msub>
<mi>dy</mi>
<mi>i</mi>
</msub>
<mfrac>
<mrow>
<mi>s</mi>
<mi>u</mi>
<mi>m</mi>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>dy</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>,</mo>
</msup>
</mrow>
<mrow>
<mi>s</mi>
<mi>u</mi>
<mi>m</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>dy</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
<mrow>
<msup>
<msub>
<mi>dt</mi>
<mi>i</mi>
</msub>
<mo>&prime;</mo>
</msup>
<mo>=</mo>
<msub>
<mi>dt</mi>
<mi>i</mi>
</msub>
<mfrac>
<mrow>
<mi>s</mi>
<mi>u</mi>
<mi>m</mi>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>dt</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>,</mo>
</msup>
</mrow>
<mrow>
<mi>s</mi>
<mi>u</mi>
<mi>m</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>dxt</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
D, K bars track is extracted in positive sample;Dx in corresponding tracki, dyi, dtiOn the basis of randomly generate respectively [- 0.5,
0.5] disturbance again, the characterising parameter of new movement locus is obtained.
5. method as claimed in claim 4, it is characterised in that the neutral net employed in the step (1) includes following
Structure:Embedding layers, LSTM, full articulamentum and softmax layers.
6. method as claimed in claim 5, it is characterised in that in the step (3), the training of the neural network model or
In prediction, the onwards transmission process of signal is as follows:
By (the dx of training samplei, dyi, dti) signal input embedding layers, by the embedding layers by dxi、dyi、dti
The vector of m dimensions corresponding to changing into respectively, and by dxi、dyi、dtiCorresponding m dimensional vectors are spliced into the vector of a 3m dimension;Will
The vector of 3m dimensions is inputted in LSTM neutral nets according to sequential, by the expression of this track of LSTM neutral nets output 3m*L dimensions
Vector into full articulamentum, and by softmax layers export this track whether be human body behavior judged result.
7. method as claimed in claim 6, it is characterised in that in the step (2), use preceding 100 (xi, yi, ti) come
Characterising parameter as mouse movement track;When mouse moving distance is shorter, and trace information is inadequate 100, mended with (0,0,0)
Together to 100.
8. method as claimed in claim 7, it is characterised in that in the step (2), by dxi, dyiValue be converted into [-
49,50] integer between, by dtiValue be converted to (0,200] between integer.
9. method as claimed in claim 8, it is characterised in that the sample size included in the positive sample collection is no less than
5000, the sample of selection 80% is concentrated as training sample in positive sample, selects 20% sample as test sample.
10. method as claimed in claim 9, it is characterised in that the sample size included in the negative sample collection is no less than
5000, the sample of selection 80% is concentrated as training sample in negative sample, selects 20% sample as test sample.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710544873.2A CN107463879A (en) | 2017-07-05 | 2017-07-05 | Human bodys' response method based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710544873.2A CN107463879A (en) | 2017-07-05 | 2017-07-05 | Human bodys' response method based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107463879A true CN107463879A (en) | 2017-12-12 |
Family
ID=60546071
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710544873.2A Pending CN107463879A (en) | 2017-07-05 | 2017-07-05 | Human bodys' response method based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107463879A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108647518A (en) * | 2018-03-16 | 2018-10-12 | 广东工业大学 | A kind of Android platform malware detection method based on deep learning |
CN109063456A (en) * | 2018-08-02 | 2018-12-21 | 浙江大学 | The safety detecting method and system of image-type identifying code |
CN109522454A (en) * | 2018-11-20 | 2019-03-26 | 四川长虹电器股份有限公司 | The method for automatically generating web sample data |
CN109756494A (en) * | 2018-12-29 | 2019-05-14 | 中国银联股份有限公司 | A kind of negative sample transform method and device |
WO2019158015A1 (en) * | 2018-02-13 | 2019-08-22 | 京东方科技集团股份有限公司 | Sample acquisition method, target detection model generation method, target detection method |
CN110472251A (en) * | 2018-05-10 | 2019-11-19 | 腾讯科技(深圳)有限公司 | Method, the method for statement translation, equipment and the storage medium of translation model training |
CN115630365A (en) * | 2022-10-27 | 2023-01-20 | 云景技术有限公司 | Verification code malicious input detection method based on deep learning |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040009479A1 (en) * | 2001-06-08 | 2004-01-15 | Jay Wohlgemuth | Methods and compositions for diagnosing or monitoring auto immune and chronic inflammatory diseases |
CN101923433A (en) * | 2010-08-17 | 2010-12-22 | 北京航空航天大学 | Man-computer interaction mode based on hand shadow identification |
CN103809894A (en) * | 2012-11-15 | 2014-05-21 | 华为终端有限公司 | Gesture recognition method and electronic equipment |
CN106155298A (en) * | 2015-04-21 | 2016-11-23 | 阿里巴巴集团控股有限公司 | Man-machine recognition methods and device, the acquisition method of behavior characteristics data and device |
CN106407889A (en) * | 2016-08-26 | 2017-02-15 | 上海交通大学 | Video human body interaction motion identification method based on optical flow graph depth learning model |
CN106845351A (en) * | 2016-05-13 | 2017-06-13 | 苏州大学 | It is a kind of for Activity recognition method of the video based on two-way length mnemon in short-term |
-
2017
- 2017-07-05 CN CN201710544873.2A patent/CN107463879A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040009479A1 (en) * | 2001-06-08 | 2004-01-15 | Jay Wohlgemuth | Methods and compositions for diagnosing or monitoring auto immune and chronic inflammatory diseases |
CN101923433A (en) * | 2010-08-17 | 2010-12-22 | 北京航空航天大学 | Man-computer interaction mode based on hand shadow identification |
CN103809894A (en) * | 2012-11-15 | 2014-05-21 | 华为终端有限公司 | Gesture recognition method and electronic equipment |
CN106155298A (en) * | 2015-04-21 | 2016-11-23 | 阿里巴巴集团控股有限公司 | Man-machine recognition methods and device, the acquisition method of behavior characteristics data and device |
CN106845351A (en) * | 2016-05-13 | 2017-06-13 | 苏州大学 | It is a kind of for Activity recognition method of the video based on two-way length mnemon in short-term |
CN106407889A (en) * | 2016-08-26 | 2017-02-15 | 上海交通大学 | Video human body interaction motion identification method based on optical flow graph depth learning model |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2019158015A1 (en) * | 2018-02-13 | 2019-08-22 | 京东方科技集团股份有限公司 | Sample acquisition method, target detection model generation method, target detection method |
US11238296B2 (en) | 2018-02-13 | 2022-02-01 | Boe Technology Group Co., Ltd. | Sample acquisition method, target detection model generation method, target detection method, computing device and computer readable medium |
CN108647518A (en) * | 2018-03-16 | 2018-10-12 | 广东工业大学 | A kind of Android platform malware detection method based on deep learning |
CN110472251A (en) * | 2018-05-10 | 2019-11-19 | 腾讯科技(深圳)有限公司 | Method, the method for statement translation, equipment and the storage medium of translation model training |
CN110472251B (en) * | 2018-05-10 | 2023-05-30 | 腾讯科技(深圳)有限公司 | Translation model training method, sentence translation equipment and storage medium |
CN109063456A (en) * | 2018-08-02 | 2018-12-21 | 浙江大学 | The safety detecting method and system of image-type identifying code |
CN109063456B (en) * | 2018-08-02 | 2021-10-08 | 浙江大学 | Security detection method and system for image type verification code |
CN109522454A (en) * | 2018-11-20 | 2019-03-26 | 四川长虹电器股份有限公司 | The method for automatically generating web sample data |
CN109522454B (en) * | 2018-11-20 | 2022-06-03 | 四川长虹电器股份有限公司 | Method for automatically generating web sample data |
CN109756494A (en) * | 2018-12-29 | 2019-05-14 | 中国银联股份有限公司 | A kind of negative sample transform method and device |
CN109756494B (en) * | 2018-12-29 | 2021-04-16 | 中国银联股份有限公司 | Negative sample transformation method and device |
CN115630365A (en) * | 2022-10-27 | 2023-01-20 | 云景技术有限公司 | Verification code malicious input detection method based on deep learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107463878A (en) | Human bodys' response system based on deep learning | |
CN107463879A (en) | Human bodys' response method based on deep learning | |
CN109978893B (en) | Training method, device, equipment and storage medium of image semantic segmentation network | |
CN108416198A (en) | Man-machine identification model establishes device, method and computer readable storage medium | |
GB2578341A (en) | Method and apparatus for automatically recognizing electrical imaging well logging facies | |
CN108664893A (en) | A kind of method for detecting human face and storage medium | |
US8090538B2 (en) | System and method for interpretation of well data | |
CN103927550B (en) | A kind of Handwritten Numeral Recognition Method and system | |
CN111709244B (en) | Deep learning method for identifying cause and effect relationship of contradictory dispute | |
CN109242149A (en) | A kind of student performance early warning method and system excavated based on educational data | |
CN106777402B (en) | A kind of image retrieval text method based on sparse neural network | |
CN110781406A (en) | Social network user multi-attribute inference method based on variational automatic encoder | |
CN112329816A (en) | Data classification method and device, electronic equipment and readable storage medium | |
CN108549817A (en) | A kind of software security flaw prediction technique based on text deep learning | |
WO2020101128A1 (en) | Method for predicting shale gas production by using deep learning | |
CN109359815A (en) | Based on the smart grid deep learning training sample generation method for generating confrontation network | |
Ahmed et al. | Discovering diverse, high quality design ideas from a large corpus | |
CN110889450A (en) | Method and device for super-parameter tuning and model building | |
CN106354852A (en) | Search method and device based on artificial intelligence | |
Kumar Pandey et al. | Metaheuristic algorithm integrated neural networks for well-test analyses of petroleum reservoirs | |
CN110532372A (en) | A kind of accurate method for pushing of text object excavating further feature based on neural collaborative filtering | |
CN110070131A (en) | A kind of Active Learning Method of data-oriented driving modeling | |
CN110399690B (en) | Subway station pedestrian simulation method and device, electronic equipment and storage medium | |
CN109325125A (en) | A kind of social networks rumour method based on CNN optimization | |
CN104731937A (en) | User behavior data processing method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171212 |
|
WD01 | Invention patent application deemed withdrawn after publication |