CN115587181A - Vehicle detection data classification method, system, computer and readable storage medium - Google Patents
Vehicle detection data classification method, system, computer and readable storage medium Download PDFInfo
- Publication number
- CN115587181A CN115587181A CN202211136820.4A CN202211136820A CN115587181A CN 115587181 A CN115587181 A CN 115587181A CN 202211136820 A CN202211136820 A CN 202211136820A CN 115587181 A CN115587181 A CN 115587181A
- Authority
- CN
- China
- Prior art keywords
- detection data
- vehicle detection
- prediction
- classification
- input
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides a vehicle detection data classification method, a system, a computer and a readable storage medium, wherein the method comprises the steps of preprocessing vehicle detection data to generate an input text; inputting the input vector into an ERNIE model to convert the input vector into a first word vector, and performing sequence characterization processing on the input text to generate a second word vector; splicing the first word vector and the second word vector to generate a word vector matrix, and inputting the word vector matrix into a DPCNN model; optimizing the DPCNN model through the isometric convolution function, and performing maximum pooling on the isometric convolution function through the optimized DPCNN model to generate a maximum characteristic value; and outputting a prediction classification label corresponding to the input text according to the maximum characteristic value. By means of the method, the vehicle detection data can be rapidly classified, and therefore time consumed by vehicle detection data classification is greatly shortened.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a vehicle detection data classification method, a vehicle detection data classification system, a computer and a readable storage medium.
Background
With the progress of science and technology and the rapid development of productivity, automobiles are popularized in daily life of people, become one of indispensable transportation tools for people to go out, and are greatly convenient for the life of people.
In recent years, with the development of laws and regulations related to data security, classification and classification of vehicle detection data is a prerequisite for achieving vehicle data security, and thus, in order to reduce the cost of vehicle detection data security work such as horsepower, torque, and acceleration time and to improve vehicle production efficiency, it is urgently required to have intelligentization capability.
The essence of vehicle detection data classification is intelligent automatic classification of detection data, however, in the prior art, the time consumed for vehicle detection data classification is long, so that the classification efficiency of the detection data is low, and the vehicle detection data is not favorably stored.
Disclosure of Invention
Based on this, the present invention provides a method, a system, a computer and a readable storage medium for classifying vehicle detection data, so as to solve the problem that the classification efficiency of the vehicle detection data is low due to the long time consumed for classifying the vehicle detection data in the prior art.
The first aspect of the embodiment of the invention provides a vehicle detection data classification method, which comprises the following steps:
when vehicle detection data are acquired, preprocessing the vehicle detection data to generate a corresponding input text, wherein the input text comprises a plurality of input vectors;
inputting a plurality of input vectors into a preset ERNIE model so as to convert the input vectors into corresponding first word vectors, and performing sequence characterization processing on the input text so as to generate corresponding second word vectors;
splicing the first word vector and the second word vector to generate a corresponding word vector matrix, and inputting the word vector matrix into a preset DPCNN model;
optimizing the preset DPCNN model through an equal-length convolution function, and performing maximum pooling on the equal-length convolution function through the optimized DPCNN model after a characteristic diagram output by the equal-length convolution function meets preset requirements to generate a plurality of maximum characteristic values;
and outputting a prediction classification label corresponding to the input text according to the maximum feature values based on a preset algorithm.
The invention has the beneficial effects that: preprocessing vehicle detection data acquired in real time to generate a corresponding input text; further, inputting a plurality of input vectors into a preset ERNIE model to convert the current input vector into a corresponding first word vector, and performing sequence characterization processing on the input text to generate a corresponding second word vector; on the basis, splicing the first word vector and the second word vector to generate a corresponding word vector matrix, and inputting the current word vector matrix into a preset DPCNN model; meanwhile, optimizing the preset DPCNN model through an equal-length convolution function, and performing maximum pooling on the equal-length convolution function through the optimized DPCNN model after a characteristic diagram output by the equal-length convolution function meets preset requirements to generate a plurality of maximum characteristic values; and finally, outputting the prediction classification label corresponding to the input text according to the maximum characteristic values based on a preset algorithm. By means of the method, the classification of the vehicle detection data can be simply and quickly completed on the premise that the ERNIE model and the DPCNN model participate together, so that the time consumed by the classification of the vehicle detection data is greatly shortened, the classification efficiency of the vehicle detection data is greatly improved, and the method and the device are suitable for large-scale popularization and use.
Preferably, the step of inputting a plurality of the input vectors into a predetermined ERNIE model to convert the input vectors into corresponding first word vectors includes:
inputting the input vectors into a pre-training layer in the preset ERNIE model, and converting the input vectors one by one through a bidirectional Transformer encoder in the pre-training layer so as to convert the input vectors into corresponding first word vectors.
Preferably, the step of outputting the predictive classification label corresponding to the input text according to the maximum feature values based on a preset algorithm includes:
when a plurality of maximum characteristic values are obtained, inputting the plurality of maximum characteristic values into a normalized Softmax function so that the normalized Softmax function outputs corresponding prediction classification probabilities according to the plurality of maximum characteristic values;
and inputting the prediction classification probability into a prediction function so that the prediction function outputs a corresponding probability matrix containing various labels, and acquiring the prediction classification label corresponding to the input text according to the probability matrix.
Preferably, after the step of outputting the predicted classification label corresponding to the input text according to a plurality of maximum feature values based on a preset algorithm, the method further includes:
and labeling the prediction classification label with the prediction label probability, and judging the security level of the prediction classification label according to the prediction label probability.
Preferably, the expression of the equal-length convolution function is:
E＝f(KX+b)
wherein E represents the equal-length convolution function, K represents a convolution kernel, X represents the word vector matrix, and b represents a deviation.
A second aspect of an embodiment of the present invention provides a vehicle detection data classification system, where the system includes:
the vehicle detection device comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for preprocessing vehicle detection data to generate a corresponding input text when the vehicle detection data is acquired, and the input text comprises a plurality of input vectors;
the processing module is used for inputting the input vectors into a preset ERNIE model so as to convert the input vectors into corresponding first word vectors, and performing sequence characterization processing on the input text so as to generate corresponding second word vectors;
the splicing module is used for splicing the first word vector and the second word vector to generate a corresponding word vector matrix and inputting the word vector matrix into a preset DPCNN model;
the optimization module is used for optimizing the preset DPCNN model through an equal-length convolution function, and performing maximum pooling on the equal-length convolution function through the optimized DPCNN model after a feature map output by the equal-length convolution function meets preset requirements to generate a plurality of maximum feature values;
and the output module is used for outputting the prediction classification label corresponding to the input text according to the maximum characteristic values based on a preset algorithm.
In the vehicle detection data classification system, the processing module is specifically configured to:
inputting the input vectors into a pre-training layer in the preset ERNIE model, and converting the input vectors one by one through a bidirectional Transformer encoder in the pre-training layer so as to convert the input vectors into corresponding first word vectors.
In the vehicle detection data classification system, the output module is specifically configured to:
when a plurality of maximum characteristic values are obtained, inputting the plurality of maximum characteristic values into a normalized Softmax function so that the normalized Softmax function outputs corresponding prediction classification probabilities according to the plurality of maximum characteristic values;
and inputting the prediction classification probability into a prediction function so that the prediction function outputs a corresponding probability matrix containing various labels, and acquiring the prediction classification label corresponding to the input text according to the probability matrix.
In the vehicle detection data classification system, the vehicle detection data classification system further includes an evaluation module, and the evaluation module is specifically configured to:
and labeling the prediction classification label with the prediction label probability, and judging the security level of the prediction classification label according to the prediction label probability.
In the above vehicle detection data classification system, the expression of the equal-length convolution function is:
E＝f(KX+b)
wherein E represents the equal-length convolution function, K represents a convolution kernel, X represents the word vector matrix, and b represents a deviation.
A third aspect of an embodiment of the present invention provides a computer, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the vehicle detection data classification method as described above when executing the computer program.
A fourth aspect of the embodiments of the present invention proposes a readable storage medium having stored thereon a computer program which, when executed by a processor, implements the vehicle detection data classification method as described above.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
FIG. 1 is a flowchart of a vehicle detection data classification method according to a first embodiment of the present invention;
fig. 2 is a block diagram illustrating a vehicle detection data classification system according to a second embodiment of the present invention.
The following detailed description will further illustrate the invention in conjunction with the above-described figures.
Detailed Description
To facilitate an understanding of the invention, the invention will now be described more fully with reference to the accompanying drawings. Several embodiments of the invention are presented in the drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
It will be understood that when an element is referred to as being "secured to" another element, it can be directly on the other element or intervening elements may also be present. When an element is referred to as being "connected" to another element, it can be directly connected to the other element or intervening elements may also be present. The terms "vertical," "horizontal," "left," "right," and the like as used herein are for illustrative purposes only.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
In the prior art, the time consumed for classifying the vehicle detection data is long, so that the classification efficiency of the vehicle detection data is low, and the vehicle detection data is not favorably stored.
Referring to fig. 1, a vehicle detection data classification method according to a first embodiment of the present invention is shown, and the vehicle detection data classification method according to this embodiment can simply and quickly complete classification of vehicle detection data on the premise that an ERNIE model and a DPCNN model participate together, so as to greatly shorten time consumed by vehicle detection data classification, thereby greatly improving classification efficiency of vehicle detection data, and being suitable for wide popularization and use.
Specifically, the vehicle detection data classification method provided in this embodiment specifically includes the following steps:
step S10, when vehicle detection data are obtained, preprocessing the vehicle detection data to generate a corresponding input text, wherein the input text comprises a plurality of input vectors;
specifically, in the present embodiment, it should be noted that the vehicle detection data classification method provided in the present embodiment is mainly used for effectively classifying the vehicle detection data of the automobile manufacturing enterprise, and meanwhile, the probability prediction can be performed on the classification result to find the most accurate classification label in the database, where it should be noted that the vehicle detection data provided in the present embodiment may include detection data such as engine horsepower, engine torque, and vehicle acceleration time.
In addition, in the present embodiment, it should be noted that the vehicle detection data classification method provided by the present embodiment is implemented based on a classification server provided in the background. It should be noted that the ERNIE model, the DPCNN model, and the transform encoder are pre-installed in the classification server, and in addition, a plurality of processing algorithms are pre-programmed in the classification server, so as to improve the classification efficiency of the vehicle detection data and correspondingly shorten the time consumed for classifying the vehicle detection data.
Therefore, in this step, it should be noted that, when the classification server obtains vehicle detection data input from the outside, the current classification server immediately preprocesses the received vehicle detection data, specifically, the classification server provided in this embodiment immediately sequentially performs data splitting and file classification on the received vehicle detection data to generate a plurality of input texts, specifically, the input text provided in this embodiment includes a plurality of input vectors, and preferably, in this embodiment, the current plurality of input vectors are marked as: w = (W) 1 ，W 2 ，W 3 …，W n ) Wherein W is i (i =1,2,3, \8230; n) represents the ith of the input text described above.
Step S20, inputting a plurality of input vectors into a preset ERNIE model so as to convert the input vectors into corresponding first word vectors, and performing sequence characterization processing on the input text so as to generate corresponding second word vectors;
further, in this step, it should be noted that, after the classification server obtains the input text and the input vector corresponding to the current vehicle detection data, the current classification server immediately inputs the currently obtained input vector into an ERNIE model preset therein, so as to convert the currently obtained input vector into a corresponding first word vector through the ERNIE model, where it is noted that the ERNIE model can implement an enhanced language representation model based on a knowledge masking policy by performing mask processing on entities, words, phrases, and the like.
In this step, it should be noted that the step of inputting a plurality of input vectors into a predetermined ERNIE model to convert the input vectors into corresponding first word vectors includes:
inputting the input vectors into a pre-training layer in the preset ERNIE model, and converting the input vectors one by one through a bidirectional Transformer encoder in the pre-training layer so as to convert the input vectors into corresponding first word vectors.
It should be noted that the transform encoder is composed of a feedforward network layer and a self-attention layer, the encoder of the ERNIE model uses multiple layers of transforms, and the bidirectional representation of the pre-training language is realized through the joint adjustment of each layer of the transforms.
Therefore, in this step, the input vectors are input into a pre-training layer in the ERNIE model, and the bidirectional Transformer encoder in the pre-training layer performs a conversion process on the input vectors one by one, so as to convert the input vectors into corresponding first word vectors.
Meanwhile, the current classification server also performs sequence characterization processing on the input text to obtain a corresponding serialized text, wherein the serialized text is marked as X i ＝(X 1i ，X 2i ，…，X ji ) Wherein X is ji A word vector representing the jth word in the ith sentence, enabling the generation of a corresponding second word vector.
Step S30, splicing the first word vector and the second word vector to generate a corresponding word vector matrix, and inputting the word vector matrix into a preset DPCNN model;
specifically, in this step, it should be noted that, after the classification server obtains the first word vector and the second word vector respectively, the current classification server immediately performs a splicing process on the currently received first word vector and second word vector to generate a corresponding word vector matrix X, and the word vector matrix X = X 1 ⊕X 2 ⊕X 3 ⊕…⊕X n 。
Further, the current classification server will generate the word vector matrix X = X in real time 1 ⊕X 2 ⊕X 3 ⊕…⊕X n Inputting the real-time data into a DPCNN model preset in the real-time data input device.
Step S40, optimizing the preset DPCNN model through an equal-length convolution function, and after a characteristic diagram output by the equal-length convolution function meets preset requirements, performing maximum pooling on the equal-length convolution function through the optimized DPCNN model to generate a plurality of maximum characteristic values;
furthermore, in this step, it should be noted that, after the classification server inputs the word vector matrix to the DPCNN model, the current classification server further performs optimization processing on the current DPCNN model through a preset equal-length convolution function, and performs maximum pooling processing on the current equal-length convolution function through the optimized DPCNN model after a feature map output by the current equal-length convolution function meets a preset requirement, that is, after the size of the feature map output by the current equal-length convolution function is fixed, so as to generate a plurality of maximum feature values.
In this step, it should be noted that the maximum pooling processing provided in this embodiment is specifically set as: the pooling size is 3, the step length is 2, and the number of the characteristic graphs is fixed, so that the calculation time and the data size of each pooling layer of the DPCNN model are both shortened to 1/2 of the original value, the calculation efficiency is obviously improved, and the calculation time is correspondingly shortened.
And S50, outputting a prediction classification label corresponding to the input text according to the maximum feature values based on a preset algorithm.
Finally, in this step, it should be noted that after the classification server calculates the maximum feature values, the current classification server further outputs the predicted classification labels corresponding to the input text according to the maximum feature values based on an algorithm preset in the current classification server.
In this step, it should be noted that the step of outputting the prediction classification label corresponding to the input text according to the plurality of maximum feature values based on the preset algorithm includes:
when a plurality of maximum characteristic values are obtained, inputting the plurality of maximum characteristic values into a normalized Softmax function so that the normalized Softmax function outputs corresponding prediction classification probabilities according to the plurality of maximum characteristic values;
and inputting the prediction classification probability into a prediction function so that the prediction function outputs a corresponding probability matrix containing various labels, and acquiring the prediction classification label corresponding to the input text according to the probability matrix.
Wherein, the expression of the normalized Softmax function is as follows:
in addition, in this embodiment, it should be further noted that the above-mentioned prediction classification label includes a prediction label probability, the sum of the prediction label probabilities is 1, and after the step of outputting the prediction classification label corresponding to the input text according to a plurality of maximum feature values based on a preset algorithm, the method further includes:
and labeling the prediction classification label with the prediction label probability, and judging the security level of the prediction classification label according to the prediction label probability.
In this step, the probability of each predicted classification label is marked, so that the worker can visually observe the possibility of each classification label, and can accurately know the type of the input text, thereby being convenient for storage.
In this embodiment, it should be noted that, the expression of the equal-length convolution function is as follows:
E＝f(KX+b)
wherein E represents the equal-length convolution function, K represents a convolution kernel, X represents the word vector matrix, and b represents a deviation.
When the method is used, the vehicle detection data acquired in real time is preprocessed to generate a corresponding input text; further, inputting a plurality of input vectors into a preset ERNIE model to convert the current input vector into a corresponding first word vector, and performing sequence characterization processing on the input text to generate a corresponding second word vector; on the basis, splicing the first word vector and the second word vector to generate a corresponding word vector matrix, and inputting the current word vector matrix into a preset DPCNN model; meanwhile, optimizing the preset DPCNN model through an equal-length convolution function, and performing maximum pooling on the equal-length convolution function through the optimized DPCNN model after a characteristic diagram output by the equal-length convolution function meets preset requirements to generate a plurality of maximum characteristic values; and finally, outputting a prediction classification label corresponding to the input text according to the maximum characteristic values based on a preset algorithm. By means of the method, the classification of the vehicle detection data can be simply and rapidly completed on the premise that the ERNIE model and the DPCNN model participate together, so that the time consumed by the classification of the vehicle detection data is greatly shortened, the classification efficiency of the vehicle detection data is greatly improved, and the method is suitable for large-scale popularization and use.
It should be noted that the above implementation process is only for illustrating the applicability of the present application, but this does not represent that the vehicle detection data classification method of the present application has only the above-mentioned unique implementation flow, and on the contrary, the vehicle detection data classification method of the present application can be incorporated into the feasible embodiments of the present application as long as the vehicle detection data classification method of the present application can be implemented.
In summary, the vehicle detection data classification method provided by the above embodiment of the present invention can simply and rapidly complete the classification of the vehicle detection data on the premise that the ERNIE model and the DPCNN model participate together, thereby greatly shortening the time consumed by the classification of the vehicle detection data, further greatly improving the classification efficiency of the vehicle detection data, and being suitable for large-scale popularization and use.
Referring to fig. 2, a vehicle detection data classification system according to a second embodiment of the present invention is shown, the system includes:
the acquiring module 12 is configured to, when vehicle detection data is acquired, pre-process the vehicle detection data to generate a corresponding input text, where the input text includes a plurality of input vectors;
the processing module 22 is configured to input a plurality of input vectors into a preset ERNIE model, so as to convert the input vectors into corresponding first word vectors, and perform sequence characterization processing on the input text, so as to generate corresponding second word vectors;
the concatenation module 32 is configured to perform concatenation processing on the first word vector and the second word vector to generate a corresponding word vector matrix, and input the word vector matrix into a preset DPCNN model;
the optimization module 42 is configured to perform optimization processing on the preset DPCNN model through an equal-length convolution function, and perform maximum pooling processing on the equal-length convolution function through the optimized DPCNN model after a feature map output by the equal-length convolution function meets preset requirements, so as to generate a plurality of maximum feature values;
and the output module 52 is configured to output a prediction classification label corresponding to the input text according to the plurality of maximum feature values based on a preset algorithm.
In the vehicle detection data classification system, the processing module 22 is specifically configured to:
inputting the input vectors into a pre-training layer in the pre-set ERNIE model, and converting the input vectors one by one through a bidirectional Transformer encoder in the pre-training layer so as to convert the input vectors into corresponding first word vectors.
In the vehicle detection data classification system, the output module 52 is specifically configured to:
when a plurality of maximum characteristic values are obtained, inputting the plurality of maximum characteristic values into a normalized Softmax function so that the normalized Softmax function outputs corresponding prediction classification probabilities according to the plurality of maximum characteristic values;
and inputting the prediction classification probability into a prediction function so that the prediction function outputs a corresponding probability matrix containing various labels, and acquiring the prediction classification label corresponding to the input text according to the probability matrix.
In the vehicle detection data classification system, the vehicle detection data classification system further includes an evaluation module 62, and the evaluation module 62 is specifically configured to:
and labeling the prediction classification label with the prediction label probability, and judging the security level of the prediction classification label according to the prediction label probability.
In the vehicle detection data classification system, the expression of the equal-length convolution function is as follows:
E＝f(KX+b)
wherein E represents the equal-length convolution function, K represents a convolution kernel, X represents the word vector matrix, and b represents a deviation.
A third embodiment of the present invention provides a computer including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the vehicle detection data classification method provided as the first embodiment when executing the computer program.
A fourth embodiment of the present invention provides a readable storage medium on which a computer program is stored, which when executed by a processor, implements the vehicle detection data classification method provided as the first embodiment described above.
In summary, the vehicle detection data classification method, system, computer and readable storage medium provided in the embodiments of the present invention can simply and rapidly complete classification of vehicle detection data on the premise that the ERNIE model and the DPCNN model participate together, so as to greatly shorten the time consumed by vehicle detection data classification, thereby greatly improving the classification efficiency of vehicle detection data, and are suitable for wide popularization and use.
The above modules may be functional modules or program modules, and may be implemented by software or hardware. For a module implemented by hardware, the modules may be located in the same processor; or the modules may be located in different processors in any combination.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Further, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following technologies, which are well known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that various changes and modifications can be made by those skilled in the art without departing from the spirit of the invention, and these changes and modifications are all within the scope of the invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A vehicle detection data classification method, characterized in that the method comprises:
when vehicle detection data are acquired, preprocessing the vehicle detection data to generate a corresponding input text, wherein the input text comprises a plurality of input vectors;
inputting a plurality of input vectors into a preset ERNIE model so as to convert the input vectors into corresponding first word vectors, and performing sequence characterization processing on the input text so as to generate corresponding second word vectors;
splicing the first word vector and the second word vector to generate a corresponding word vector matrix, and inputting the word vector matrix into a preset DPCNN model;
optimizing the preset DPCNN model through an equal-length convolution function, and performing maximum pooling on the equal-length convolution function through the optimized DPCNN model after a characteristic diagram output by the equal-length convolution function meets preset requirements to generate a plurality of maximum characteristic values;
and outputting a prediction classification label corresponding to the input text according to the maximum feature values based on a preset algorithm.
2. The vehicle detection data classification method according to claim 1, characterized in that: the step of inputting a plurality of input vectors into a predetermined ERNIE model to convert the input vectors into corresponding first word vectors includes:
inputting the input vectors into a pre-training layer in the pre-set ERNIE model, and converting the input vectors one by one through a bidirectional Transformer encoder in the pre-training layer so as to convert the input vectors into corresponding first word vectors.
3. The vehicle detection data classification method according to claim 1, characterized in that: the step of outputting the prediction classification label corresponding to the input text according to the maximum feature values based on a preset algorithm comprises the following steps:
when a plurality of maximum characteristic values are obtained, inputting the plurality of maximum characteristic values into a normalized Softmax function so that the normalized Softmax function outputs corresponding prediction classification probabilities according to the plurality of maximum characteristic values;
and inputting the prediction classification probability into a prediction function so that the prediction function outputs a corresponding probability matrix containing various labels, and acquiring the prediction classification label corresponding to the input text according to the probability matrix.
4. The vehicle detection data classification method according to claim 1, characterized in that: the predicted classification label comprises a predicted label probability, the sum of the predicted label probabilities is 1, and after the step of outputting the predicted classification label corresponding to the input text according to the maximum feature values based on a preset algorithm, the method further comprises the following steps:
and labeling the prediction classification label with the prediction label probability, and judging the security level of the prediction classification label according to the prediction label probability.
5. The vehicle detection data classification method according to claim 1, characterized in that: the expression of the equal-length convolution function is as follows:
E＝f(KX+b)
wherein E represents the equal-length convolution function, K represents a convolution kernel, X represents the word vector matrix, and b represents a deviation.
6. A vehicle inspection data classification system, the system comprising:
the system comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for preprocessing vehicle detection data to generate a corresponding input text when the vehicle detection data are acquired, and the input text comprises a plurality of input vectors;
the processing module is used for inputting the input vectors into a preset ERNIE model so as to convert the input vectors into corresponding first word vectors, and performing sequence characterization processing on the input text so as to generate corresponding second word vectors;
the splicing module is used for splicing the first word vector and the second word vector to generate a corresponding word vector matrix and inputting the word vector matrix into a preset DPCNN model;
the optimization module is used for optimizing the preset DPCNN model through an equal-length convolution function, and performing maximum pooling on the equal-length convolution function through the optimized DPCNN model after a characteristic diagram output by the equal-length convolution function meets preset requirements to generate a plurality of maximum characteristic values;
and the output module is used for outputting the prediction classification label corresponding to the input text according to the maximum feature values based on a preset algorithm.
7. The vehicle detection data classification system according to claim 6, characterized in that: the processing module is specifically configured to:
inputting the input vectors into a pre-training layer in the pre-set ERNIE model, and converting the input vectors one by one through a bidirectional Transformer encoder in the pre-training layer so as to convert the input vectors into corresponding first word vectors.
8. The vehicle detection data classification system according to claim 6, characterized in that: the output module is specifically configured to:
when a plurality of maximum characteristic values are obtained, inputting the plurality of maximum characteristic values into a normalized Softmax function so that the normalized Softmax function outputs corresponding prediction classification probabilities according to the plurality of maximum characteristic values;
and inputting the prediction classification probability into a prediction function so that the prediction function outputs a corresponding probability matrix containing various labels, and acquiring the prediction classification label corresponding to the input text according to the probability matrix.
9. A computer comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the vehicle detection data classification method according to any one of claims 1 to 5 when executing the computer program.
10. A readable storage medium on which a computer program is stored, the program, when executed by a processor, implementing the vehicle detection data classification method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211136820.4A CN115587181A (en) | 2022-09-19 | 2022-09-19 | Vehicle detection data classification method, system, computer and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211136820.4A CN115587181A (en) | 2022-09-19 | 2022-09-19 | Vehicle detection data classification method, system, computer and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115587181A true CN115587181A (en) | 2023-01-10 |
Family
ID=84778750
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211136820.4A Pending CN115587181A (en) | 2022-09-19 | 2022-09-19 | Vehicle detection data classification method, system, computer and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115587181A (en) |
-
2022
- 2022-09-19 CN CN202211136820.4A patent/CN115587181A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110084274B (en) | Real-time image semantic segmentation method and system, readable storage medium and terminal | |
| CN111522916B (en) | Voice service quality detection method, model training method and device | |
| CN111126481A (en) | Training method and device of neural network model | |
| CN114155244A (en) | Defect detection method, device, equipment and storage medium | |
| CN111444986A (en) | Building drawing component classification method and device, electronic equipment and storage medium | |
| CN110688847A (en) | Technical contract determination method, device, computer equipment and storage medium | |
| CN111414732A (en) | Text style conversion method and device, electronic equipment and storage medium | |
| CN115964115B (en) | Numerical control machine tool interaction method based on pre-training reinforcement learning and related equipment | |
| CN115587181A (en) | Vehicle detection data classification method, system, computer and readable storage medium | |
| CN112989829A (en) | Named entity identification method, device, equipment and storage medium | |
| CN116842195A (en) | Knowledge graph and large model based automatic generation report generation method | |
| CN113469237B (en) | User intention recognition method, device, electronic equipment and storage medium | |
| CN111291748B (en) | Cascade distributed artificial intelligence case number identification system | |
| CN112434156A (en) | Power grid operation warning method and device based on mixed text classification model | |
| CN116431135B (en) | Method, system, computer and readable storage medium for writing automobile code | |
| CN111881266A (en) | Response method and device | |
| CN115329063B (en) | User intention identification method and system | |
| CN117194985B (en) | Multi-mode multi-task training system and multi-mode multi-task training method | |
| CN115357646B (en) | Bridge state monitoring method and system | |
| CN117993868B (en) | Dual-mode attention-based power grid engineering project audit early warning method and system | |
| CN115270645B (en) | Design method and system based on ERNIE-DPCNN model | |
| CN112307227B (en) | Data classification method | |
| CN117453910A (en) | Text classification method and device and automobile | |
| JPH08339360A (en) | Learning system applying neural network | |
| CN116303538A (en) | Training method of calculation engine selection model, calculation engine selection 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 |