CN110260925A - Detection method and its system, the intelligent recommendation method, electronic equipment of driver's stopping technical superiority and inferiority - Google Patents
Detection method and its system, the intelligent recommendation method, electronic equipment of driver's stopping technical superiority and inferiority Download PDFInfo
- Publication number
- CN110260925A CN110260925A CN201910631451.8A CN201910631451A CN110260925A CN 110260925 A CN110260925 A CN 110260925A CN 201910631451 A CN201910631451 A CN 201910631451A CN 110260925 A CN110260925 A CN 110260925A
- Authority
- CN
- China
- Prior art keywords
- driver
- inferiority
- time series
- stopping technical
- technical superiority
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- 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
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Traffic Control Systems (AREA)
Abstract
A kind of detection method of driver's stopping technical superiority and inferiority provided by the present invention is based on multiple sensors, obtains the action datas such as steering wheel, the brake in driver's docking process with time series approach, obtains multi-modal time series；And the neural network framework for being based further on convolution variation self-encoding encoder carries out feature extraction to multi-modal time series, and generates analog sample；Probability Detection is reconstructed based on analog sample, and driver's stopping technical superiority and inferiority is judged based on testing result.Intelligent recommendation method provided by the present invention, can be based on the judgement of the superiority and inferiority to driver's stopping technical, and then can recommend when needed automatic parking function or other executable functions to driver.The present invention also provides the detection system and electronic equipment of driver's stopping technical superiority and inferiority, the two has beneficial effect same as mentioned above.
Description
[technical field]
The present invention relates to the field of mobile data detection, in particular to a kind of detection methods of driver's stopping technical superiority and inferiority
And its system, intelligent recommendation method, electronic equipment.
[background technique]
With satellite positioning tech, vehicle-mounted sensing technology maturation and it is universal.Obtain the perfect number of vehicle operating and position
According to the intelligent level for improving vehicle becomes possible.For living in city space for people, stop in narrow space
Vehicle is almost a required skill.And automatic parking technology is the Premium Features that current many type automobiles are all equipped with, this
To many people --- it is particularly useful especially for driving new hand.
But with the raising of vehicle intellectualized level, vehicle mounted intelligentized equipment is more and more to become increasingly complex.People are often
At this moment the busy each single item function of carefully studying his/her bought vehicle is properly automatically waking up these functions, formed automatic
The vehicle-mounted function recommendation changed just seems very intimate practical.
[summary of the invention]
It is difficult to the technical issues of accurately confirming the superiority and inferiority of driver driving operation for solution is existing, the present invention provides a kind of department
Detection method and its system, the intelligent recommendation method, electronic equipment of machine stopping technical superiority and inferiority.
The present invention is in order to solve the above technical problems, provide a kind of the following technical solution: inspection of driver's stopping technical superiority and inferiority
Survey method comprising following steps: step S1, by the vehicle sensory data during driver's shut-down operation with time series approach
It is handled, to obtain multi-modal time series；Step S2, the neural network framework based on convolution variation self-encoding encoder is to multimode
State time series carries out feature extraction, and generates analog sample；And step S3, Probability Detection is reconstructed based on analog sample,
And driver's stopping technical superiority and inferiority is judged based on testing result.
Preferably, in above-mentioned steps S1, the vehicle sensory data include based on sensor to driver's shut-down operation mistake
The data that brake, gear shifting action, commutation action or GPS location in journey are obtained.
Preferably, in above-mentioned steps S1, it is described with time series approach carry out processing specifically include provide it is multiple continuous
Time point, and the data that multiple sensors obtain are recorded according to time dot sequency.
Preferably, above-mentioned steps S2 further includes steps of step S21, constructs the mind of convolution variation self-encoding encoder
Through the network architecture；Wherein, the neural network framework of the convolution variation self-encoding encoder includes coding module and decoder module；Step
S22, multi-modal time series described in the neural network framework using convolution variation self-encoding encoder carries out convolutional encoding, and extracts more
Feature in mode time series；Step S23, the feature based on extraction, map vector value simultaneously carries out characteristic functional, and generates hidden
Variable vector space；And step S24, by hidden variable vector space described in convolution decoder, to generate analog sample.
Preferably, step S231 is further included steps of in above-mentioned steps S23, by the multi-modal time of extraction
The Feature Mapping mean vector space and variance vectors space of sequence；And step S232, Gaussian noise data are added, linearly to give birth to
At corresponding hidden variable vector space.
Preferably, in step s 24 by hidden variable vector space described in convolution decoder, further include obtain mean vector and
Variance vectors, and corresponding analog sample is generated with this.
The present invention is in order to solve the above technical problems, to provide another technical solution as follows: a kind of driver's stopping technical superiority and inferiority
Detection system includes: multiple sensors, and the sensor obtains multiple groups vehicle sensory data for detecting driver operation；When
Between retrieval module, for by the multiple groups vehicle sensory data during driver's shut-down operation with time series approach at
Reason, to obtain multi-modal time series；Sample generation module, for the neural network framework pair based on convolution variation self-encoding encoder
Multi-modal time series carries out feature extraction, and generates analog sample；And reconstruct probabilistic module, for being carried out based on analog sample
Probability Detection is reconstructed, and driver's stopping technical superiority and inferiority is judged based on testing result.
Preferably, the time series acquisition module further comprises: network construction constructs module, becomes for constructing convolution
Divide the neural network framework of self-encoding encoder；Wherein, the neural network framework of the convolution variation self-encoding encoder includes coding module
And decoder module；Characteristic extracting module, for the multi-modal time described in the neural network framework using convolution variation self-encoding encoder
Sequence carries out convolutional encoding, and extracts the feature in multi-modal time series；Hidden variable obtains module, for the spy based on extraction
Sign, map vector value simultaneously carries out characteristic functional, and generates hidden variable vector space；And analog sample generation module, for passing through
Hidden variable vector space described in convolution decoder, to generate analog sample.
The present invention is in order to solve the above technical problems, to provide another technical solution as follows: a kind of intelligent recommendation method is based on
The detection method of driver's stopping technical superiority and inferiority as described above is based on stopping technical to obtain the stopping technical superiority and inferiority grade of driver
Superiority and inferiority hierarchical selection recommends automatic parking function.
The present invention is in order to solve the above technical problems, to provide another technical solution as follows: an electronic equipment includes storage unit
And processing unit, the storage unit is for storing computer program, and the processing unit by the storage unit for being deposited
The computer program of storage executes the detection method of driver's stopping technical superiority and inferiority as described above.
Compared with prior art, a kind of detection method of the driver's stopping technical superiority and inferiority provided by the present invention given and its it is
System, intelligent recommendation method, electronic equipment have it is following the utility model has the advantages that
The present invention provides a kind of detection method of driver's stopping technical superiority and inferiority, different from the method for existing behavioral value, this
Method provided by inventing can effectively solve the existing uncertainty due to driver operation behavior, and cause asking for detection error
Topic.Specifically in the present invention by the multiple groups vehicle sensory data during driver's shut-down operation with time series approach, obtain
Multi-modal time series；The neural network framework for being based further on convolution variation self-encoding encoder carries out spy to multi-modal time series
Sign is extracted, and generates analog sample, to can get dominant and recessive character, so that big data analysis can be realized, to obtain more
Accurate detection result；Probability Detection is reconstructed based on analog sample, and judges that driver's stopping technical is excellent based on testing result
It is bad.Method provided by the present invention is intended to be obtained steering wheel, brake in driver's docking process etc. based on multiple sensors and driven
Action data simultaneously carries out respective handling, to can get the judgement so as to realize the superiority and inferiority to driver's stopping technical, Jin Erke
It realizes and recommends automatic parking function or other executable functions when needed to driver.
In order to preferably obtain data from multi-modal time series, convolution variation self-encoding encoder is constructed in the present invention
Neural network framework includes coding module and decoder module, further utilizes the neural network framework institute of convolution variation self-encoding encoder
It states multi-modal time series and carries out convolutional encoding, and extract the feature in multi-modal time series；Feature based on extraction, mapping
Vector value simultaneously carries out characteristic functional, and generates hidden variable vector space；And pass through hidden variable vector space described in convolution decoder, with
Generate analog sample.Based on above-mentioned cataloged procedure and decoding process, the feature of multi-modal time series can be extracted, and can be real
Now to the extensive processing of feature, compared with existing abnormality detection technology, above-mentioned method can carry out multi-modal time series fast
Fast feature extraction, and can therefrom extract and obtain recessive feature.
Further, the detection method of driver's stopping technical superiority and inferiority further includes the multi-modal time by that will extract
The Feature Mapping of sequence obtains mean vector space and variance vectors space, can carry out feature convenient for the feature based on extraction and open up
Exhibition, and by addition Gaussian noise data, it can get the new feature data of more multiple coincidence original characteristic rule, and new based on these
Characteristic corresponds to hidden variable vector space with linear generating.Based on above-mentioned step, it can be achieved that on the basis of a small amount of sample
On, it can also realize that the multi-modal time series to various dimensions complexity carries out abnormality detection.
It further, further include obtaining mean vector in the present invention by hidden variable vector space described in convolution decoder
And variance vectors, and corresponding analog sample is generated with this.Based on above-mentioned steps, analog sample obtained can meet former multimode
The characteristic rule of state time series, the problem of data deviation can be reduced.
In the present invention, it is reconstructed in Probability Detection based on analog sample and uses the reconstruct from variation autocoder
Probability introduce method for detecting abnormality, analog sample is detected, can solve it is existing based on distance method, the method for density and
When the method for cluster carries out abnormality detection technology, the problem bad for multi-dimensional data detection effect.
Vehicle intelligent recommended method provided by the present invention uses the detection method of above-mentioned driver's stopping technical superiority and inferiority can
Overcome the problems, such as that existing technology is difficult to accurately evaluate driver's stopping technical, ancillary technique of parking is current permitted
The Premium Features that polymorphic type automobile is all equipped with, this is to many people --- and it is particularly useful especially for driving new hand.But with
The raising of vehicle intellectualized level, vehicle mounted intelligentized equipment is more and more to become increasingly complex.People often have no time and carefully grind
Study carefully each single item function of his/her bought vehicle, therefore, the detection method based on above-mentioned driver's stopping technical superiority and inferiority in the present invention
Can provide it is a kind of can the driving technology based on driver recommend the technical solution of automatic parking automatically.
The present invention also provides the detection systems and a kind of electronic equipment of a kind of driver's stopping technical superiority and inferiority, have and above-mentioned department
The identical beneficial effect of detection method of machine stopping technical superiority and inferiority obtains driver's docking process, it can be achieved that based on multiple sensors
In the action datas such as steering wheel, brake and carry out respective handling, to can realize that the superiority and inferiority to driver's stopping technical is sentenced
It is disconnected.
[Detailed description of the invention]
Fig. 1 is that the step process of the detection method of driver's stopping technical superiority and inferiority provided in first embodiment of the invention is shown
It is intended to.
Fig. 2 is sensor sample data with the distribution table of time series.
Fig. 3 is the specific steps flow diagram in step S2 shown in Fig. 1.
Fig. 4 is the schematic diagram of encoder convolutional network structure.
Fig. 5 is the schematic diagram of decoder convolutional network structure.
Fig. 6 is the specific steps flow diagram of step S23 shown in Fig. 3.
Fig. 7 is that the functional module of the detection system of driver's stopping technical superiority and inferiority provided in second embodiment of the invention is shown
It is intended to.
Fig. 8 is the specific functional units schematic diagram of address word segmentation module shown in Fig. 7.
Fig. 9 is the specific functional modules schematic diagram that hidden variable shown in Fig. 8 obtains module.
Figure 10 is the step flow diagram of intelligent recommendation method provided in third embodiment of the invention.
Figure 11 is the functional block diagram of electronic equipment provided in fourth embodiment of the invention.
Description of drawing identification:
20, the detection system of driver's stopping technical superiority and inferiority；21, sensor；22, time series obtains module；23, sample
Generation module；24, probabilistic module is reconstructed；231, network construction construction unit；232, feature extraction unit；233, hidden variable obtains
Take unit；234, analog sample generation unit；2331, Feature Mapping unit；2332, hidden variable unit is generated；
30, electronic equipment；31, storage unit；32, processing unit.
[specific embodiment]
In order to make the purpose of the present invention, technical solution and advantage are more clearly understood, below in conjunction with attached drawing and embodiment,
The present invention will be described in further detail.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention,
It is not intended to limit the present invention.
Referring to Fig. 1, the first embodiment of the present invention provides a kind of detection method 10 of driver's stopping technical superiority and inferiority, packet
Include following steps:
Step S1 obtains multimode by the multiple groups vehicle sensory data during driver's shut-down operation with time series approach
State time series；
Step S2, the neural network framework based on convolution variation self-encoding encoder carry out feature to multi-modal time series and mention
It takes, and generates analog sample；And
Step S3 is reconstructed Probability Detection based on analog sample, and judges that driver's stopping technical is excellent based on testing result
It is bad.
Specifically, the multiple groups vehicle sensory data described in above-mentioned steps S1 can be obtained based on the sensor of installation in the car
?.The sensor is mountable in car body, based on the sensor can to during driver's shut-down operation brake, change
The information such as gear movement, commutation action and GPS location are obtained, and in a manner of time series, to form the multi-modal time
Sequence.
The multiple sensors being specifically included under same time series in the multi-modal time series correspond to numerical value.
It is described with time series approach to carry out processing and refer to provide multiple continuous time points in above-mentioned steps S1, and
The data that multiple sensors obtain are recorded according to time dot sequency.
In the present embodiment, the related data of the sensor of acquisition and the driving operation behavior of driver are closely related.For example,
It in one embodiment, in the present invention, can be according to setting in order to preferably record the concrete operations of driver in the process of backing up
In the concrete operations of the intracorporal sensor record driver of vehicle.For example, being set at the beginning of sensor being acquired corresponding operating
Terminating to fall back 5 minutes from parking, wherein the frequency of the sensor sample is that sampling in 1 second is primary, as shown in Figure 2, horizontal axis table
Show time point t (be specifically represented by t1, t2 ... tn), the numerical value s that the longitudinal axis indicates that sensor is incuded (is specifically represented by
s1,s2……sn)；Wherein, in conjunction with shown in Fig. 2, t1, t2, t3 ... ... tn respectively indicate 5 points, 4 points 59 seconds, 4 points of 58 seconds, 4
Divide 57 seconds, and so on.And corresponding s1 is represented by the data of brake respective sensor acquisition, s2 is represented by shift and corresponds to
The data of sensor acquisition, s3 are represented by the data for turning to respective sensor and obtaining, s4 is represented by GPS positioning system sensing
The data etc. that device obtains.Each numerical value s be represented by respective sensor at the time point on response, for the ease of
Record can calculate, can by sensor detected numerical value be set to the real number value of 0-1, for example, rotation direction sensor can be obtained
Value after the angle value normalization obtained.It is appreciated that for ease of calculation, may be based on the data volume of acquisition, and match properly
Data processing mode.
Therefore, if selecting the Outlier Detection Algorithm of existing unsupervised (or semi-supervised) type, can lack specific
Instruct label.It is difficult to using artificial extraction feature in multisensor and needs preferably to be imitated under fine-time granularity operation
Fruit.
And the present invention as shown in Figure 3, further limits to preferably solve the problem above-mentioned in above-mentioned steps S2
Neural network framework based on convolution variation self-encoding encoder carries out feature extraction to multi-modal time series, and generates simulation sample
This, specifically comprises the following steps:
Step S21 constructs the neural network framework of convolution variation self-encoding encoder；Wherein, the convolution variation self-encoding encoder
Neural network framework include coding module and decoder module；
Step S22, multi-modal time series described in the neural network framework using convolution variation self-encoding encoder carry out convolution
Coding, and extract the feature in multi-modal time series；
Step S23, the feature based on extraction, map vector value simultaneously carries out characteristic functional, and generates hidden variable vector space
(Latent Variable Vector Space)；And
Step S24, by hidden variable vector space described in convolution decoder, to generate analog sample.
Based on above-mentioned step, the rule lain in after specific data in multi-modal time series can be obtained, thus
The data of more multiple coincidence rule can be obtained, to overcome in existing technology, are divided just for more data are repeated
Analysis, and ignore the analysis of the rule of the little data of repeatability.
In above-mentioned steps S21, encoder convolution in the specific neural network framework for constructing convolution variation self-encoding encoder
Network structure (as shown in Figure 4) and decoder convolutional network structure (as shown in Figure 5), wherein corresponding symbol " c ", " s "
And " u ", it is expressed as convolution, down-sampling and joint.The whole structure of the neural network framework of the convolution variation self-encoding encoder
Feature extractor can specifically be done based on MC-1D-CVAE (the one-dimensional convolution variation self-encoding encoder of multichannel) by, which making, judges algorithm with abnormal
Frame.Specifically, building process includes coding module and the decoding that convolution kernel deconvolution is placed in the variation self-encoding encoder
Module.
Further, in order to preferably construct the neural network framework of the convolution variation self-encoding encoder, in the present invention
In some specific embodiments, building and training for network model is carried out using deep learning frame BigDL.With certain data period
On the basis of (such as can the moon be data period), stop carry out technology judgement for each driver, finally beaten to corresponding driver every time
Upper label if stopping technical is excellent grade or stopping technical is of inferior quality grade, and then can be not yet done for vehicle intelligent recommender system
Practice parking car owner's progress automatic parking function and uses prompting.
Wherein, the self-encoding encoder principle is a kind of unsupervised learning model of neural network version, can be to no label sample
This training neural network.Network training principle may make network output in the case where as far as possible close to input sample, study to number
According to recessive character (being usually expressed as low dimensional, basic hidden unit characteristic).The variation self-encoding encoder is a kind of nerve
The probabilistic model of network, distribution of the variation self-encoding encoder by learning sample, the generation of Lai Shixian sample, and sample size is about
Abundant, then the sample generated then can more embody the feature of initial data.
As shown in Figure 6, the feature such as step S23 based on extraction, map vector value simultaneously carries out characteristic functional, and linearly gives birth to
At hidden variable vector space, it can further be subdivided into following step:
Step S231, by the Feature Mapping mean vector space and variance vectors space of the multi-modal time series of extraction；
And
Step S232 is added Gaussian noise data, corresponds to hidden variable vector space with linear generating.
In above-mentioned steps S231- step S232, after extracting feature, noise is added, characteristic functionalization can be made more stronger,
To produce more richer virtual samples, to enhance sample.
Further, the step S24 can be further subdivided into: by hidden variable vector space described in convolution decoder, from
And mean vector and variance vectors can be obtained, and corresponding analog sample is generated with this.
The hidden variable vector space described in above-mentioned steps can be understood as the set of hidden variable, this set is to hidden variable
Addition sum number to multiply be closed, that is, being also referred to as linear space to addition and the closed vector space of scale multiplication.
In above-mentioned steps S3, Probability Detection is reconstructed based on analog sample, and judge that driver stops based on testing result
The superiority and inferiority of driving skills art specifically includes the method for detecting abnormality introduced using the reconstruct probability from variation autocoder, to institute
Analog sample is stated to be detected.Wherein, reconstruct probability can combine variation autocoder by considering the concept of variability
Probability nature.
In the above embodiment of the present invention, the purpose carried out abnormality detection is to find significant difference automatically from data
Sample data or mode, usual this kind of sample only account for seldom a part in population sample.Traditional abnormality detection technology packet
The method based on distance is included, the method for method and cluster based on density can not carry out multi-dimensional data processing.And neural network
Abnormality detection is done, strategy is substantially the general characteristic with neural network learning to sample, and is detected compared with this general characteristic
There is the sample of significant difference.So self-encoding encoder is exactly a kind of naturally anomaly detector.In variation self-encoding encoder,
The objective function (also referred to as evidence lower bound) of optimization:
ELBOi(θ, φ)=Eqθ(z|xi)[logpφ(xi|z)]-KL(qθ(z|xi)||p(z))
Wherein, its expression of ELBO (Evidence Lower Bound) is in functional qθMiddle maximization evidence lower bound.Wherein, xi
Be it is fixed, we can define xiUnder the conditions of distribution q (z | xi).This allows us to select each x different p
(z), above formula first item is the probability that sample is reconstructed, can be with it come structural anomaly score value (Anomaly Score):
Anomaly Score=1-Eqθ(z|xi)[pφ(xi|z)]
As it can be seen that can be determined based on this abnormal score in the abnormality detecting process in above-mentioned steps S3.Its
In, abnormal score is big, then corresponding multi-modal time series is judged as exception.Reconstruct probability is a kind of consideration variable distribution change
Anisotropic probability metrics.Using the formation characteristic of variation autocoder, data reconstruction can be derived, to analyze abnormal root
This reason.
In the present embodiment, for the ease of judgement, an outlier threshold can also be set, when abnormal score is more than this abnormal threshold
When value, multi-modal time series is just determined as exception, and at this point, the corresponding parking driver behavior of driver is then judged as of inferior quality
Grade.
Based on the detection method of driver's stopping technical superiority and inferiority provided by the present invention, may be implemented with sensing data,
The movement of the shut-down operation of driver is identified, is driven so as to steering, the brake etc. to steering wheel in driver's docking process
The response of action data judges the stopping technical of driver, to judge the unskilled driver of stopping technical, and can be
Recommend automatic parking function under suitable situation.
Using the neural network framework of above-mentioned convolution variation self-encoding encoder, whole sample can be made when convolutional encoding
This amount can be more, richer.Other model frameworks are compared, and a small amount of sample, which can be used, can be realized feature, especially hidden
Property feature extraction, thus meet driver higher row be dimension driver behavior, also can be improved to driver's stopping technical superiority and inferiority
Fast and accurately judgement.
In the above-mentioned methods, it based on the acquisition to hidden variable vector space, can solve due to the certain driver behaviors of driver
The frequency for acting appearance is relatively low, and the problem of can not effectively be detected, the method provided by the present embodiment can correspond to will be recessive
Feature also detected.
It needs the sample of magnanimity just to can be carried out corresponding operation or simulate compared to existing other encoders to compare, this
The neural network framework of convolution variation self-encoding encoder provided in embodiment can with variation self-encoding encoder realize extract feature and
Generating sample can reach preferable abnormality detection effect after labelling to a small amount of sample.
Referring to Fig. 7, the second embodiment of the present invention provides the detection system 20 of driver's stopping technical superiority and inferiority, packet
It includes:
Multiple sensors 21, the sensor obtain multiple groups vehicle sensory data for detecting driver operation；The biography
Sensor can the action responses such as steering wheel rotation, brake, back-up speed to vehicle record.
Time series obtain module 22, for by the multiple groups vehicle sensory data during driver's shut-down operation with time sequence
Column mode obtains multi-modal time series；
Sample generation module 23, for the neural network framework based on convolution variation self-encoding encoder to multi-modal time series
Feature extraction is carried out, and generates analog sample；And
Probabilistic module 24 is reconstructed, for Probability Detection to be reconstructed based on analog sample, and is based on testing result judgement department
Machine stopping technical superiority and inferiority.
Wherein, the sensor 21 includes angular transducer, pressure sensor, flow sensor, velocity sensor, acceleration
Spend any several combination in sensor, levelness sensor, transmission sensor or range sensor.So as to car body direction
Body speed of vehicle, acceleration in disk steering angle, docking process etc. are detected.
Multi-modal time Series Processing is lacked in order to solve existing Outlier Detection Algorithm and the artificial method for extracting feature
It falls into, in the present embodiment, referring to Fig. 8, time series acquisition module 23 further comprises:
Network construction construction unit 231, for constructing the neural network framework of convolution variation self-encoding encoder；Wherein, described
The neural network framework of convolution variation self-encoding encoder includes coding module and decoder module；
Feature extraction unit 232, for the multi-modal time described in the neural network framework using convolution variation self-encoding encoder
Sequence carries out convolutional encoding, and extracts the feature in multi-modal time series；
Hidden variable acquiring unit 233, for the feature based on extraction, map vector value simultaneously carries out characteristic functional, and generates
Hidden variable vector space；And
Analog sample generation unit 234, for passing through hidden variable vector space described in convolution decoder, to generate simulation sample
This.
Referring to Fig. 9, in the present embodiment, the hidden variable obtains module 233 and further comprises:
Feature Mapping unit 2331, the Feature Mapping mean vector space and side of the multi-modal time series for that will extract
Difference vector space；And
It generates hidden variable unit 2332 and hidden variable vector space is corresponded to linear generating for Gaussian noise data to be added.
In the present embodiment, the particular content of the neural network framework in relation to convolution variation self-encoding encoder and above-mentioned first is in fact
The associated description applied in example is consistent, and details are not described herein.
Referring to Fig. 10, third embodiment of the invention provides an intelligent recommendation method P10 comprising following steps:
Step P01, the detection method based on driver's stopping technical superiority and inferiority as described in above-mentioned first embodiment is to obtain department
The stopping technical superiority and inferiority grade of machine；And
Step P02 recommends automatic parking function based on stopping technical superiority and inferiority hierarchical selection.
Further, the intelligent recommendation method 30 further includes that can recommend different services based on stopping technical superiority and inferiority grade
Project, such as it is easier to Parking Stall or parking lot recommendation, insurance purchase recommendation.
In above-mentioned steps P02, specifically, the detection based on driver's stopping technical superiority and inferiority described in above-mentioned first embodiment
Method stopping technical superiority and inferiority grade obtained can be divided into multiple grades, and different recommendation function can be matched based on different grades
Energy.
The definitions relevant of detection method in relation to driver's stopping technical superiority and inferiority is identical with above-mentioned first embodiment,
This is repeated no more.
Figure 11 is please referred to, the fourth embodiment of the present invention provides an electronic equipment 30, and the electronic equipment 30 includes storage
Unit 31 and processing unit 32, the storage unit 31 is for storing computer program, and the processing unit 32 is for passing through institute
The computer program for stating the storage of storage unit 31 executes the data checking method that exceptional value is examined described in above-mentioned first embodiment
Specific steps.
In some specific embodiments of the present invention, the electronic equipment 30 can be hardware, be also possible to software.Work as electricity
When sub- equipment 30 is hardware, the various electronic equipments of video playing are can be with display screen and supported, including but not limited to
Smart phone, tablet computer, E-book reader, MP3 player (Moving Picture Experts Group Audio
Layer III, dynamic image expert's compression standard audio level 3), MP4 (Moving Picture Experts Group
Audio Layer IV, dynamic image expert's compression standard audio level 4) player, pocket computer on knee and desk-top meter
Calculation machine etc..When electronic equipment 30 is software, may be mounted in above-mentioned cited electronic equipment.It may be implemented into more
A software or software module (such as providing multiple softwares of Distributed Services or software module), also may be implemented into single
Software or software module.It is not specifically limited herein.
The storage unit 31 includes the storage unit of read-only memory (ROM), random access storage device (RAM) and hard disk etc.
Point etc., the processing unit 32 according to the program being stored in the read-only memory (ROM) or can be loaded into random visit
It asks the program in memory (RAM) and executes various movements appropriate and processing.In random access storage device (RAM), also deposit
It contains the electronic equipment 30 and operates required various programs and data.
The electronic equipment 30 may also include the importation (not shown) of keyboard, mouse etc.；The electronic equipment 30 is also
Can further comprise cathode-ray tube (CRT), liquid crystal display (LCD) etc. and loudspeaker etc. output par, c (figure not
Show)；And the electronic equipment 30 can further comprise the communication unit of the network interface card of LAN card, modem etc.
Divide (not shown).The communications portion executes communication process via the network of such as internet.
Particularly, disclosed embodiment, the process described above with reference to flow chart may be implemented as counting according to the present invention
Calculation machine software program.For example, disclosed embodiment of this invention may include a kind of computer program product comprising be carried on meter
Computer program on calculation machine readable medium, the computer program include the program generation for method shown in execution flow chart
Code.In such embodiments, which can be downloaded and installed from network by communications portion.
When the computer program is executed by the processing unit 32, executes the described of the application and have anti-fraud functional mind
The above-mentioned function of being limited in training method through network model.It should be noted that computer-readable medium described herein
It can be computer-readable signal media or computer readable storage medium either the two any combination.Computer
Readable storage medium storing program for executing for example may be-but not limited to-the system of electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor, dress
It sets or device, or any above combination.The more specific example of computer readable storage medium can include but is not limited to:
Electrical connection, portable computer diskette, hard disk, random access storage device (RAM), read-only storage with one or more conducting wires
Device (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-
ROM), light storage device, magnetic memory device or above-mentioned any appropriate combination.
In this application, computer readable storage medium can also be any tangible medium for including or store program, should
Program can be commanded execution system, device or device use or in connection.And in this application, computer can
The signal media of reading may include in a base band or as the data-signal that carrier wave a part is propagated, wherein carrying computer
Readable program code.The data-signal of this propagation can take various forms, including but not limited to electromagnetic signal, optical signal
Or above-mentioned any appropriate combination.Computer-readable signal media can also be appointing other than computer readable storage medium
What computer-readable medium, the computer-readable medium can be sent, propagated or transmitted for by instruction execution system, device
Either device use or program in connection.The program code for including on computer-readable medium can be fitted with any
When medium transmission, including but not limited to: wireless, electric wire, optical cable, RF etc. or above-mentioned any appropriate combination.
One or more programming languages or combinations thereof can be used to write the calculating for executing operation of the invention
Machine program code, described program design language include object oriented program language -- such as Java, Smalltalk, C+
+, it further include conventional procedural programming language-such as " C " language or similar programming language.Program code can
Fully to execute, partly execute on the user computer on the user computer, be executed as an independent software package,
Part executes on the remote computer or executes on a remote computer or server completely on the user computer for part.
In situations involving remote computers, remote computer can pass through the network of any kind --- including local area network (LAN)
Or wide area network (WAN)-is connected to subscriber computer, or, it may be connected to outer computer (such as utilize Internet service
Provider is connected by internet).
Flow chart and block diagram in attached drawing of the invention illustrate the system according to the various embodiments of the application, method
With the architecture, function and operation in the cards of computer program product.In this regard, each of flowchart or block diagram
Box can represent a part of a module, program segment or code, and a part of the module, program segment or code includes one
A or multiple executable instructions for implementing the specified logical function.It should also be noted that in some realization sides as replacement
In case, function marked in the box may also be distinct from that the sequence marked in attached drawing occurs.For example, two succeedingly indicate
Box can actually be basically executed in parallel, they can also execute in the opposite order sometimes, herein based on being related to
Function and determine.It is significant to note that in each box and block diagram and or flow chart in block diagram and or flow chart
Box combination, can the dedicated hardware based systems of the functions or operations as defined in executing realize, or can be with
It realizes using a combination of dedicated hardware and computer instructions.
Involved unit can be realized by way of software in an embodiment of the present invention, can also pass through hardware
Mode realize.Described unit also can be set in the processor.
As on the other hand, the fifth embodiment of the present invention additionally provides a kind of computer-readable medium, which can
Reading medium can be included in device described in above-described embodiment；It is also possible to individualism, and without the supplying dress
In setting.Above-mentioned computer-readable medium carries one or more program, and described program specifically includes: by driver's shut-down operation
Multiple groups vehicle sensory data in the process obtain multi-modal time series with time series approach；It is encoded certainly based on convolution variation
The neural network framework of device carries out feature extraction to multi-modal time series, and generates analog sample；And based on analog sample into
Row reconstruct Probability Detection, and driver's stopping technical superiority and inferiority is judged based on testing result.
Compared with prior art, a kind of detection method of the driver's stopping technical superiority and inferiority provided by the present invention given and its it is
System, intelligent recommendation method, electronic equipment have it is following the utility model has the advantages that
The present invention provides a kind of detection method of driver's stopping technical superiority and inferiority, different from the method for existing behavioral value, this
Method provided by inventing can effectively solve the existing uncertainty due to driver operation behavior, and cause asking for detection error
Topic.Specifically in the present invention by the multiple groups vehicle sensory data during driver's shut-down operation with time series approach, obtain
Multi-modal time series；The neural network framework for being based further on convolution variation self-encoding encoder carries out spy to multi-modal time series
Sign is extracted, and generates analog sample, to can get dominant and recessive character, so that big data analysis can be realized, to obtain more
Accurate detection result；Probability Detection is reconstructed based on analog sample, and judges that driver's stopping technical is excellent based on testing result
It is bad.Method provided by the present invention is intended to be obtained steering wheel, brake in driver's docking process etc. based on multiple sensors and driven
Action data simultaneously carries out respective handling, to can get the judgement so as to realize the superiority and inferiority to driver's stopping technical, Jin Erke
It realizes and recommends automatic parking function or other executable functions when needed to driver.
In the present invention, it is reconstructed in Probability Detection based on analog sample and uses the reconstruct from variation autocoder
Probability introduce method for detecting abnormality, analog sample is detected, can solve it is existing based on distance method, the method for density and
When the method for cluster carries out abnormality detection technology, the problem bad for multi-dimensional data detection effect.
Vehicle intelligent recommended method provided by the present invention uses the detection method of above-mentioned driver's stopping technical superiority and inferiority can
Overcome the problems, such as that existing technology is difficult to accurately evaluate driver's stopping technical, ancillary technique of parking is current permitted
The Premium Features that polymorphic type automobile is all equipped with, this is to many people --- and it is particularly useful especially for driving new hand.But with
The raising of vehicle intellectualized level, vehicle mounted intelligentized equipment is more and more to become increasingly complex.People often have no time and carefully grind
Study carefully each single item function of his/her bought vehicle, therefore, the detection method based on above-mentioned driver's stopping technical superiority and inferiority in the present invention
Can provide it is a kind of can the driving technology based on driver recommend the technical solution of automatic parking automatically.
The present invention also provides the detection systems and a kind of electronic equipment of a kind of driver's stopping technical superiority and inferiority, have and above-mentioned department
The identical beneficial effect of detection method of machine stopping technical superiority and inferiority obtains driver's docking process, it can be achieved that based on multiple sensors
In the action datas such as steering wheel, brake and carry out respective handling, to can realize that the superiority and inferiority to driver's stopping technical is sentenced
It is disconnected.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in original of the invention
Made any modification within then, equivalent replacement and improvement etc. should all be comprising within protection scope of the present invention.
Claims (10)
1. a kind of detection method of driver's stopping technical superiority and inferiority, it is characterised in that: the detection side of driver's stopping technical superiority and inferiority
Method the following steps are included:
Step S1 handles the vehicle sensory data during driver's shut-down operation with time series approach, more to obtain
Mode time series；
Step S2, the neural network framework based on convolution variation self-encoding encoder carry out feature extraction to multi-modal time series, and
Generate analog sample；And
Step S3 is reconstructed Probability Detection based on analog sample, and judges driver's stopping technical superiority and inferiority based on testing result.
2. the detection method of driver's stopping technical superiority and inferiority as described in the appended claim 1, it is characterised in that: in above-mentioned steps S1,
The vehicle sensory data include dynamic to the brake during driver's shut-down operation, gear shifting action, commutation based on sensor
The data that work or GPS location are obtained.
3. the detection method of driver's stopping technical superiority and inferiority as described in the appended claim 1, it is characterised in that: in above-mentioned steps S1,
It is described with time series approach to carry out processing and specifically include to provide multiple continuous time points, and the number that multiple sensors are obtained
It is recorded according to according to time dot sequency.
4. the detection method of driver's stopping technical superiority and inferiority as described in the appended claim 1, it is characterised in that: in above-mentioned steps S2 into
One step the following steps are included:
Step S21 constructs the neural network framework of convolution variation self-encoding encoder；Wherein, the mind of the convolution variation self-encoding encoder
It include coding module and decoder module through the network architecture；
Step S22, multi-modal time series described in the neural network framework using convolution variation self-encoding encoder carry out convolutional encoding,
And extract the feature in multi-modal time series；
Step S23, the feature based on extraction, map vector value simultaneously carries out characteristic functional, and generates hidden variable vector space；And
Step S24, by hidden variable vector space described in convolution decoder, to generate analog sample.
5. the detection method of driver's stopping technical superiority and inferiority as claimed in claim 4, it is characterised in that: in above-mentioned steps S23
It further includes steps of
Step S231, by the Feature Mapping mean vector space and variance vectors space of the multi-modal time series of extraction；And
Step S232 is added Gaussian noise data, corresponds to hidden variable vector space with linear generating.
6. the detection method of driver's stopping technical superiority and inferiority as claimed in claim 4, it is characterised in that: pass through in step s 24
Hidden variable vector space described in convolution decoder further includes obtaining mean vector and variance vectors, and generate corresponding simulation with this
Sample.
7. a kind of detection system of driver's stopping technical superiority and inferiority, it is characterised in that: the detection system of driver's stopping technical superiority and inferiority
System includes:
Multiple sensors, the sensor obtain multiple groups vehicle sensory data for detecting driver operation；
Time series obtain module, for by the multiple groups vehicle sensory data during driver's shut-down operation with time series approach
It is handled, to obtain multi-modal time series；
Sample generation module carries out multi-modal time series for the neural network framework based on convolution variation self-encoding encoder special
Sign is extracted, and generates analog sample；And
Probabilistic module is reconstructed, for Probability Detection to be reconstructed based on analog sample, and judges that driver is stopped based on testing result
Technology superiority and inferiority.
8. the detection system of driver's stopping technical superiority and inferiority as recited in claim 7, it is characterised in that: the time series obtains
Module further comprises:
Network construction constructs module, for constructing the neural network framework of convolution variation self-encoding encoder；Wherein, the convolution variation
The neural network framework of self-encoding encoder includes coding module and decoder module；
Characteristic extracting module is carried out for multi-modal time series described in the neural network framework using convolution variation self-encoding encoder
Convolutional encoding, and extract the feature in multi-modal time series；
Hidden variable obtain module, for the feature based on extraction, map vector value simultaneously carries out characteristic functional, and generate hidden variable to
Quantity space；And
Analog sample generation module, for passing through hidden variable vector space described in convolution decoder, to generate analog sample.
9. a kind of intelligent recommendation method, it is characterised in that: based on driver's stopping technical superiority and inferiority described in any one of claim 1-6
Detection method to obtain the stopping technical superiority and inferiority grade of driver, and automatic parking is recommended based on stopping technical superiority and inferiority hierarchical selection
Function.
10. a kind of electronic equipment, it is characterised in that: the electronic equipment includes storage unit and processing unit, and the storage is single
Member is executed for storing computer program, the computer program that the processing unit is used to store by the storage unit as weighed
Benefit requires the detection method of driver's stopping technical superiority and inferiority described in any one of 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910631451.8A CN110260925B (en) | 2019-07-12 | 2019-07-12 | Method and system for detecting quality of driver parking technology, intelligent recommendation method and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910631451.8A CN110260925B (en) | 2019-07-12 | 2019-07-12 | Method and system for detecting quality of driver parking technology, intelligent recommendation method and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110260925A true CN110260925A (en) | 2019-09-20 |
| CN110260925B CN110260925B (en) | 2021-06-25 |
Family
ID=67925888
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910631451.8A Active CN110260925B (en) | 2019-07-12 | 2019-07-12 | Method and system for detecting quality of driver parking technology, intelligent recommendation method and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110260925B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112508058A (en) * | 2020-11-17 | 2021-03-16 | 安徽继远软件有限公司 | Transformer fault diagnosis method and device based on audio characteristic analysis |
Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6208958B1 (en) * | 1998-04-16 | 2001-03-27 | Samsung Electronics Co., Ltd. | Pitch determination apparatus and method using spectro-temporal autocorrelation |
| CN106710306A (en) * | 2016-12-09 | 2017-05-24 | 深圳市元征科技股份有限公司 | Vehicle driving behavior monitoring method and vehicle driving behavior monitoring device |
| CN107492251A (en) * | 2017-08-23 | 2017-12-19 | 武汉大学 | It is a kind of to be identified and driving condition supervision method based on the driver identity of machine learning and deep learning |
| CN108009520A (en) * | 2017-12-21 | 2018-05-08 | 东南大学 | A kind of finger vein identification method and system based on convolution variation self-encoding encoder neutral net |
| CN108182452A (en) * | 2017-12-29 | 2018-06-19 | 哈尔滨工业大学（威海） | Aero-engine fault detection method and system based on grouping convolution self-encoding encoder |
| CN108216252A (en) * | 2017-12-29 | 2018-06-29 | 中车工业研究院有限公司 | A kind of subway driver vehicle carried driving behavior analysis method, car-mounted terminal and system |
| CN108710865A (en) * | 2018-05-28 | 2018-10-26 | 电子科技大学 | A kind of driver's anomaly detection method based on neural network |
| CN108725437A (en) * | 2018-07-17 | 2018-11-02 | 广东工业大学 | A kind of automatic stopping storage method and device |
| CN108764064A (en) * | 2018-05-07 | 2018-11-06 | 西北工业大学 | SAR Target Recognition Algorithms based on Steerable filter device and self-encoding encoder |
| CN108881196A (en) * | 2018-06-07 | 2018-11-23 | 中国民航大学 | The semi-supervised intrusion detection method of model is generated based on depth |
| CN109034953A (en) * | 2018-07-02 | 2018-12-18 | 西南交通大学 | A kind of film recommended method |
| WO2019051356A1 (en) * | 2017-09-08 | 2019-03-14 | The General Hospital Corporation | A system and method for automated labeling and annotating unstructured medical datasets |
| CN109543100A (en) * | 2018-10-31 | 2019-03-29 | 上海交通大学 | User interest modeling method and system based on Cooperative Study |
| CN109559209A (en) * | 2019-01-18 | 2019-04-02 | 深圳创新奇智科技有限公司 | A kind of electric business clothes based on multi-modal information, which are worn, takes recommended method |
| CN109726795A (en) * | 2017-10-30 | 2019-05-07 | 罗伯特·博世有限公司 | Method for the central artificial intelligence module of training |
| CN109800767A (en) * | 2018-12-12 | 2019-05-24 | 天津津航技术物理研究所 | Object detection method based on HOG feature and self-encoding encoder |
| CN109800658A (en) * | 2018-12-26 | 2019-05-24 | 中汽研（天津）汽车工程研究院有限公司 | Parking position type online recognition neural network based and positioning system and method |
| CN109910901A (en) * | 2019-02-15 | 2019-06-21 | 苏州工业园区职业技术学院 | A kind of intelligent DAS (Driver Assistant System) with driving behavior research and application function |
| CN109919189A (en) * | 2019-01-29 | 2019-06-21 | 华南理工大学 | A kind of depth K mean cluster method towards time series data |
| JP2019109306A (en) * | 2017-12-15 | 2019-07-04 | 日本電信電話株式会社 | Voice conversion device, voice conversion method and program |
| CN109978217A (en) * | 2018-11-09 | 2019-07-05 | 国网浙江省电力有限公司绍兴供电公司 | Methods of electric load forecasting based on singular spectrum analysis and memory network |
-
2019
- 2019-07-12 CN CN201910631451.8A patent/CN110260925B/en active Active
Patent Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6208958B1 (en) * | 1998-04-16 | 2001-03-27 | Samsung Electronics Co., Ltd. | Pitch determination apparatus and method using spectro-temporal autocorrelation |
| CN106710306A (en) * | 2016-12-09 | 2017-05-24 | 深圳市元征科技股份有限公司 | Vehicle driving behavior monitoring method and vehicle driving behavior monitoring device |
| CN107492251A (en) * | 2017-08-23 | 2017-12-19 | 武汉大学 | It is a kind of to be identified and driving condition supervision method based on the driver identity of machine learning and deep learning |
| WO2019051356A1 (en) * | 2017-09-08 | 2019-03-14 | The General Hospital Corporation | A system and method for automated labeling and annotating unstructured medical datasets |
| CN109726795A (en) * | 2017-10-30 | 2019-05-07 | 罗伯特·博世有限公司 | Method for the central artificial intelligence module of training |
| JP2019109306A (en) * | 2017-12-15 | 2019-07-04 | 日本電信電話株式会社 | Voice conversion device, voice conversion method and program |
| CN108009520A (en) * | 2017-12-21 | 2018-05-08 | 东南大学 | A kind of finger vein identification method and system based on convolution variation self-encoding encoder neutral net |
| CN108182452A (en) * | 2017-12-29 | 2018-06-19 | 哈尔滨工业大学（威海） | Aero-engine fault detection method and system based on grouping convolution self-encoding encoder |
| CN108216252A (en) * | 2017-12-29 | 2018-06-29 | 中车工业研究院有限公司 | A kind of subway driver vehicle carried driving behavior analysis method, car-mounted terminal and system |
| CN108764064A (en) * | 2018-05-07 | 2018-11-06 | 西北工业大学 | SAR Target Recognition Algorithms based on Steerable filter device and self-encoding encoder |
| CN108710865A (en) * | 2018-05-28 | 2018-10-26 | 电子科技大学 | A kind of driver's anomaly detection method based on neural network |
| CN108881196A (en) * | 2018-06-07 | 2018-11-23 | 中国民航大学 | The semi-supervised intrusion detection method of model is generated based on depth |
| CN109034953A (en) * | 2018-07-02 | 2018-12-18 | 西南交通大学 | A kind of film recommended method |
| CN108725437A (en) * | 2018-07-17 | 2018-11-02 | 广东工业大学 | A kind of automatic stopping storage method and device |
| CN109543100A (en) * | 2018-10-31 | 2019-03-29 | 上海交通大学 | User interest modeling method and system based on Cooperative Study |
| CN109978217A (en) * | 2018-11-09 | 2019-07-05 | 国网浙江省电力有限公司绍兴供电公司 | Methods of electric load forecasting based on singular spectrum analysis and memory network |
| CN109800767A (en) * | 2018-12-12 | 2019-05-24 | 天津津航技术物理研究所 | Object detection method based on HOG feature and self-encoding encoder |
| CN109800658A (en) * | 2018-12-26 | 2019-05-24 | 中汽研（天津）汽车工程研究院有限公司 | Parking position type online recognition neural network based and positioning system and method |
| CN109559209A (en) * | 2019-01-18 | 2019-04-02 | 深圳创新奇智科技有限公司 | A kind of electric business clothes based on multi-modal information, which are worn, takes recommended method |
| CN109919189A (en) * | 2019-01-29 | 2019-06-21 | 华南理工大学 | A kind of depth K mean cluster method towards time series data |
| CN109910901A (en) * | 2019-02-15 | 2019-06-21 | 苏州工业园区职业技术学院 | A kind of intelligent DAS (Driver Assistant System) with driving behavior research and application function |
Non-Patent Citations (4)
| Title |
|---|
| CHANGYONG YOON 等: "Object tracking from image sequences using adaptive models in fuzzy particle filter", 《INFORMATION SCIENCES》 * |
| 白辰甲: "基于计算机视觉和深度学习的自动驾驶方法研究", 《中国优秀硕士论文全文数据库 工程科技II辑》 * |
| 翟正利 等: "变分自编码器模型综述", 《计算机工程与应用》 * |
| 贾硕 等: "商用车辆异常驾驶行为检测算法研究", 《公路交通科技》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112508058A (en) * | 2020-11-17 | 2021-03-16 | 安徽继远软件有限公司 | Transformer fault diagnosis method and device based on audio characteristic analysis |
| CN112508058B (en) * | 2020-11-17 | 2023-11-14 | 安徽继远软件有限公司 | Transformer fault diagnosis method and device based on audio feature analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110260925B (en) | 2021-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104737523B (en) | The situational model in mobile device is managed by assigning for the situation label of data clustering | |
| CN109508681A (en) | The method and apparatus for generating human body critical point detection model | |
| CN109460514A (en) | Method and apparatus for pushed information | |
| CN107862339B (en) | Method and apparatus for outputting information | |
| CN108804704A (en) | A kind of user's depth portrait method and device | |
| CN112200318B (en) | Target detection method, device, machine readable medium and equipment | |
| US10482491B2 (en) | Targeted marketing for user conversion | |
| CN112734873A (en) | Image attribute editing method, device, equipment and medium for resisting generation network | |
| CN105335595A (en) | Feeling-based multimedia processing | |
| CN110263255A (en) | Acquisition methods, system, server and the storage medium of customer attribute information | |
| CN105940435A (en) | Measuring and diagnosing noise in urban environment | |
| CN114417174A (en) | Content recommendation method, device, equipment and computer storage medium | |
| CN114820121A (en) | Insurance business pushing system and method | |
| CN111291715A (en) | Vehicle type identification method based on multi-scale convolutional neural network, electronic device and storage medium | |
| CN110260925A (en) | Detection method and its system, the intelligent recommendation method, electronic equipment of driver's stopping technical superiority and inferiority | |
| CN109711849A (en) | Ether house address portrait generation method and device, electronic equipment and storage medium | |
| CN117272328A (en) | Functional code evaluation method and device based on neural network | |
| CN116451125A (en) | New energy vehicle owner identification method, device, equipment and storage medium | |
| CN116628349A (en) | Information recommendation method, device, equipment, storage medium and program product | |
| EP4303800A1 (en) | Method, computer, and system | |
| CN113011919B (en) | Method and device for identifying object of interest, recommendation method, medium and electronic equipment | |
| CN114418370A (en) | Test driving information processing method and device, electronic equipment and readable storage medium | |
| CN113516315A (en) | Wind power generation power interval prediction method, device and medium | |
| CN112560958A (en) | Person reception method and device based on portrait recognition and electronic equipment | |
| CN112820111A (en) | Abnormal vehicle identification method and device and electronic equipment |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20210518 Address after: Room 306, 3rd floor, building 9, Chongqing enlightenment science park, 69 Huoju Avenue, Jiulongpo District, Chongqing 400000 Applicant after: Chongqing qisaidi Technology Co.,Ltd. Address before: Room 201, building 9, Chongqing Qidi science and Technology Park, 69 torch Avenue, Jiulongpo District, Chongqing 400000 Applicant before: AINNOVATION (CHONGQING) TECHNOLOGY Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |