Vehicle detection data classification method, system, computer and readable storage medium
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JIANGLING MOTORS
- Filing Date
- 2022-09-19
- Publication Date
- 2026-06-23
Smart Images
Figure CN115587181B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing technology, and in particular to a method, system, computer, and readable storage medium for classifying vehicle detection data. Background Technology
[0002] With the advancement of technology and the rapid development of productivity, automobiles have become widespread in people's daily lives and have become an indispensable means of transportation, greatly facilitating people's lives.
[0003] In recent years, with the gradual promulgation of laws and regulations related to data security, the classification and grading of vehicle inspection data has been clearly defined at the national level as a prerequisite for achieving vehicle data security. Therefore, in order to reduce the cost of vehicle inspection data security work, such as horsepower, torque, and acceleration time, and improve the efficiency of vehicle production, the classification and grading of vehicle inspection data urgently needs to have intelligent capabilities.
[0004] The essence of vehicle inspection data classification is the intelligent and automated classification of inspection data. However, existing technologies take a long time to classify vehicle inspection data, resulting in low classification efficiency and making it difficult to preserve the vehicle inspection data. Summary of the Invention
[0005] Based on this, the purpose of the present invention is to provide a vehicle detection data classification method, system, computer, and readable storage medium to solve the problem that the existing technology takes a long time to classify vehicle detection data, resulting in low classification efficiency.
[0006] The first aspect of this invention proposes a method for classifying vehicle detection data, the method comprising:
[0007] When vehicle detection data is acquired, the vehicle detection data is preprocessed to generate corresponding input text, which includes several input vectors.
[0008] Several input vectors are input into a preset ERNIE model to convert the input vectors into corresponding first word vectors, and the input text is subjected to sequence feature processing to generate corresponding second word vectors.
[0009] The first word vector and the second word vector are concatenated to generate a corresponding word vector matrix, and the word vector matrix is input into a preset DPCNN model;
[0010] The preset DPCNN model is optimized by using an equal-length convolution function. After the feature map output by the equal-length convolution function meets the preset requirements, the optimized DPCNN model is used to perform max pooling on the equal-length convolution function to generate several maximum feature values.
[0011] Based on a preset algorithm, a predicted classification label corresponding to the input text is output according to several maximum feature values.
[0012] The beneficial effects of this invention are as follows: Real-time vehicle detection data is preprocessed to generate corresponding input text; further, several input vectors are input into a preset ERNIE model to convert the current input vector into a corresponding first word vector, and the input text is subjected to sequence feature processing to generate a corresponding second word vector; based on this, the first and second word vectors are concatenated to generate a corresponding word vector matrix, and the current word vector matrix is input into a preset DPCNN model; simultaneously, the preset DPCNN model is optimized using an equal-length convolution function, and after the feature map output by the equal-length convolution function meets preset requirements, the optimized DPCNN model performs max pooling on the equal-length convolution function to generate several maximum feature values; finally, a predicted classification label corresponding to the input text is output based on the preset algorithm according to the several maximum feature values. The above method enables simple and rapid classification of vehicle detection data by combining the ERNIE and DPCNN models, thereby significantly reducing the time spent on vehicle detection data classification and greatly improving the classification efficiency. This method is suitable for widespread promotion and use.
[0013] Preferably, the step of inputting the plurality of input vectors into a preset ERNIE model to convert the input vectors into corresponding first word vectors includes:
[0014] Several input vectors are input into the pre-trained layer of the preset ERNIE model, and the bidirectional Transformer encoder in the pre-trained layer performs transformation processing on each of the input vectors to convert the input vectors into corresponding first word vectors.
[0015] Preferably, the step of outputting the predicted classification label corresponding to the input text based on a preset algorithm according to several maximum feature values includes:
[0016] When several maximum feature values are obtained, the several maximum feature values are input into the normalized Softmax function so that the normalized Softmax function outputs the corresponding predicted classification probability based on the several maximum feature values;
[0017] The predicted classification probability is input into the prediction function so that the prediction function outputs a probability matrix containing various labels, and the predicted classification label corresponding to the input text is obtained based on the probability matrix.
[0018] Preferably, the predicted classification label includes predicted label probabilities, the sum of which is 1. After the step of outputting the predicted classification label corresponding to the input text based on a preset algorithm according to several maximum feature values, the method further includes:
[0019] The predicted classification label is labeled with the predicted label probability, and the security level of the predicted classification label is judged based on the predicted label probability.
[0020] Preferably, the expression for the equal-length convolution function is:
[0021] E = f(KX + b)
[0022] Where E represents the equal-length convolution function, K represents the convolution kernel, X represents the word vector matrix, and b represents the bias.
[0023] A second aspect of this invention provides a vehicle detection data classification system, the system comprising:
[0024] The acquisition module is used to preprocess the vehicle detection data when it is acquired to generate corresponding input text, which includes several input vectors.
[0025] The processing module is used to input several input vectors into a preset ERNIE model to convert the input vectors into corresponding first word vectors, and to perform sequence feature processing on the input text to generate corresponding second word vectors.
[0026] The concatenation module is used to concatenate 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;
[0027] The optimization module is used to optimize the preset DPCNN model using an equal-length convolution function, and after the feature map output by the equal-length convolution function meets the preset requirements, to perform max pooling on the equal-length convolution function using the optimized DPCNN model to generate several maximum feature values.
[0028] The output module is used to output a predicted classification label corresponding to the input text based on a preset algorithm and several maximum feature values.
[0029] In the aforementioned vehicle detection data classification system, the processing module is specifically used for:
[0030] Several input vectors are input into the pre-trained layer of the preset ERNIE model, and the bidirectional Transformer encoder in the pre-trained layer performs transformation processing on each of the input vectors to convert the input vectors into corresponding first word vectors.
[0031] In the aforementioned vehicle detection data classification system, the output module is specifically used for:
[0032] When several maximum feature values are obtained, the several maximum feature values are input into the normalized Softmax function so that the normalized Softmax function outputs the corresponding predicted classification probability based on the several maximum feature values;
[0033] The predicted classification probability is input into the prediction function so that the prediction function outputs a probability matrix containing various labels, and the predicted classification label corresponding to the input text is obtained based on the probability matrix.
[0034] The vehicle inspection data classification system mentioned above further includes an evaluation module, which is specifically used for:
[0035] The predicted classification label is labeled with the predicted label probability, and the security level of the predicted classification label is judged based on the predicted label probability.
[0036] In the aforementioned vehicle detection data classification system, the expression for the equal-length convolution function is:
[0037] E = f(KX + b)
[0038] Where E represents the equal-length convolution function, K represents the convolution kernel, X represents the word vector matrix, and b represents the bias.
[0039] A third aspect of the present invention provides a computer, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the vehicle detection data classification method as described above.
[0040] A fourth aspect of the present invention provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle detection data classification method as described above.
[0041] 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. Attached Figure Description
[0042] Figure 1 A flowchart of the vehicle detection data classification method provided in the first embodiment of the present invention;
[0043] Figure 2 This is a structural block diagram of the vehicle detection data classification system provided in the second embodiment of the present invention.
[0044] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation
[0045] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
[0046] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.
[0047] Unless otherwise defined, 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 pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.
[0048] Existing technologies take a long time to classify vehicle inspection data, resulting in low classification efficiency and hindering the preservation of vehicle inspection data.
[0049] Please see Figure 1 The image shows a vehicle detection data classification method provided in the first embodiment of the present invention. The vehicle detection data classification method provided in this embodiment can easily and quickly classify vehicle detection data with the joint participation of the ERNIE model and the DPCNN model, thereby greatly shortening the time spent on vehicle detection data classification and thus greatly improving the classification efficiency of vehicle detection data. It is suitable for widespread promotion and use.
[0050] Specifically, the vehicle detection data classification method provided in this embodiment includes the following steps:
[0051] Step S10: When vehicle detection data is obtained, the vehicle detection data is preprocessed to generate corresponding input text, which includes several input vectors.
[0052] Specifically, in this embodiment, it should first be noted that the vehicle inspection data classification method provided in this embodiment is mainly used to effectively classify the vehicle inspection data of automobile manufacturing enterprises. At the same time, it can also perform probability prediction on the classification results to find the most accurate classification label in the database. It should be pointed out that the vehicle inspection data provided in this embodiment may include test data such as engine horsepower, engine torque, and vehicle acceleration time.
[0053] Furthermore, it should be noted that the vehicle detection data classification method provided in this embodiment is implemented based on a classification server set up in the background. Specifically, this classification server is pre-configured with an ERNIE model, a DPCNN model, and a Transformer encoder. Additionally, several processing algorithms are pre-written within the classification server to improve the classification efficiency of vehicle detection data and correspondingly shorten the time required for vehicle detection data classification.
[0054] Therefore, in this step, it should be noted that when the aforementioned classification server receives vehicle detection data from the outside, the current classification server will immediately preprocess the received vehicle detection data. Specifically, the classification server provided in this embodiment will immediately perform data splitting and file classification processing on the received vehicle detection data in sequence to generate several input texts. Specifically, the input texts provided in this embodiment include several input vectors. Preferably, in this embodiment, the current several input vectors are labeled as: W = (W1, W2, W3…, W…). n ), where W i (i = 1, 2, 3, ... n) represents the i-th element of the above input text.
[0055] Step S20: Input several input vectors into a preset ERNIE model to convert the input vectors into corresponding first word vectors, and perform sequence feature processing on the input text to generate corresponding second word vectors;
[0056] Furthermore, in this step, it should be noted that after the classification server obtains the input text and input vector corresponding to the current vehicle detection data, the current classification server will immediately input the currently obtained input vector into its internally pre-set ERNIE model, so that the current input vector can be converted into the corresponding first word vector through the ERNIE model. It should be pointed out that the above-mentioned ERNIE model can perform masking processing on entities, words, phrases, etc., thereby realizing an enhanced language representation model based on a knowledge masking strategy.
[0057] In this step, it should be noted that the step of inputting the aforementioned input vectors into a preset ERNIE model to convert the input vectors into corresponding first word vectors includes:
[0058] Several input vectors are input into the pre-trained layer of the preset ERNIE model, and the bidirectional Transformer encoder in the pre-trained layer performs transformation processing on each of the input vectors to convert the input vectors into corresponding first word vectors.
[0059] It should be noted that the Transformer encoder consists of a feedforward network layer and a self-attention layer. The ERNIE model's encoder uses a multi-layer Transformer, and achieves bidirectional representation of the pre-trained language through joint adjustment of each Transformer layer.
[0060] Therefore, this step involves inputting the above input vectors into the pre-training layer of the preset ERNIE model, and then using the bidirectional Transformer encoder in the pre-training layer to transform each of the above input vectors into the corresponding first word vectors.
[0061] At the same time, the current classification server will also perform sequence feature processing on the above input text to obtain the corresponding serialized text, which is labeled as X. i =(X 1i X 2i , ..., X ji ), where X ji This represents the word vector of the j-th word in the i-th sentence, which is used to generate the corresponding second word vector.
[0062] Step S30: The first word vector and the second word vector are concatenated to generate a corresponding word vector matrix, and the word vector matrix is input into a preset DPCNN model;
[0063] 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 will immediately concatenate the received first word vector and second word vector to generate the corresponding word vector matrix X, and the word vector matrix X = X1⊕X2⊕X3⊕…⊕X n .
[0064] Furthermore, the current classification server will generate a word vector matrix X = X1⊕X2⊕X3⊕…⊕X in real time. n The data is fed into its pre-configured DPCNN model in real time.
[0065] Step S40: Optimize the preset DPCNN model using an equal-length convolution function, and after the feature map output by the equal-length convolution function meets the preset requirements, perform max pooling on the equal-length convolution function using the optimized DPCNN model to generate several maximum feature values.
[0066] Furthermore, in this step, it should be noted that after the above-mentioned classification server inputs the word vector matrix into the DPCNN model, the current classification server will further optimize the current DPCNN model through a pre-set equal-length convolution function. After the feature map output by the current equal-length convolution function meets the preset requirements, that is, after the size of the feature map output by the current equal-length convolution function is fixed, the optimized DPCNN model will perform max pooling on the current equal-length convolution function to generate several maximum feature values.
[0067] In this step, it should be noted that the maximum pooling processing provided in this embodiment is specifically set as follows: the pooling size is 3, the stride is 2, and the number of feature maps is fixed. This reduces the computation time and data size of each pooling layer of the DPCNN model to half of the original, thereby significantly improving computational efficiency and correspondingly shortening the computation time.
[0068] Step S50: Based on a preset algorithm, output the predicted classification label corresponding to the input text according to several maximum feature values.
[0069] Finally, it should be noted in this step that after the above classification server calculates several maximum feature values, the current classification server will further output the predicted classification label corresponding to the above input text based on its internally pre-set algorithm according to the current several maximum feature values.
[0070] In this step, it should be noted that the step of outputting the predicted classification label corresponding to the input text based on a preset algorithm according to several maximum feature values includes:
[0071] When several maximum feature values are obtained, the several maximum feature values are input into the normalized Softmax function so that the normalized Softmax function outputs the corresponding predicted classification probability based on the several maximum feature values;
[0072] The predicted classification probability is input into the prediction function so that the prediction function outputs a probability matrix containing various labels, and the predicted classification label corresponding to the input text is obtained based on the probability matrix.
[0073] The expression for the normalized Softmax function is as follows:
[0074]
[0075]
[0076] Furthermore, in this embodiment, it should be noted that the predicted classification label includes predicted label probabilities, the sum of which is 1. After the step of outputting the predicted classification label corresponding to the input text based on a preset algorithm according to several maximum feature values, the method further includes:
[0077] The predicted classification label is labeled with the predicted label probability, and the security level of the predicted classification label is judged based on the predicted label probability.
[0078] In this step, by labeling the probability of each predicted category label, staff can intuitively observe the likelihood of each category label, thereby enabling them to accurately determine the type of the input text and facilitate its saving.
[0079] In this embodiment, it should be noted that the expression for the equal-length convolution function is as follows:
[0080] E = f(KX + b)
[0081] Where E represents the equal-length convolution function, K represents the convolution kernel, X represents the word vector matrix, and b represents the bias.
[0082] In practice, real-time vehicle detection data is preprocessed to generate corresponding input text. Further, several input vectors are fed into a pre-defined ERNIE model to convert the current input vectors into corresponding first word vectors. Sequence feature processing is then performed on the input text to generate corresponding second word vectors. Based on this, the first and second word vectors are concatenated to generate a corresponding word vector matrix, which is then input into a pre-defined DPCNN model. Simultaneously, the pre-defined DPCNN model is optimized using an equal-length convolution function. After the feature map output by the equal-length convolution function meets preset requirements, max pooling is performed on the equal-length convolution function using the optimized DPCNN model to generate several maximum eigenvalues. Finally, a predicted classification label corresponding to the input text is output based on the preset algorithm using these maximum eigenvalues. This method enables simple and rapid classification of vehicle detection data with the combined participation of the ERNIE and DPCNN models, significantly reducing the time required for vehicle detection data classification and thus greatly improving the classification efficiency. It is suitable for widespread promotion and use.
[0083] It should be noted that the above implementation process is only to illustrate the feasibility of this application, but it does not mean that the vehicle inspection data classification method of this application has only one implementation process. On the contrary, as long as the vehicle inspection data classification method of this application can be implemented, it can be included in the feasible implementation scheme of this application.
[0084] In summary, the vehicle detection data classification method provided by the above embodiments of the present invention can easily and quickly classify vehicle detection data with the joint participation of the ERNIE model and the DPCNN model, thereby significantly shortening the time spent on vehicle detection data classification and thus greatly improving the classification efficiency of vehicle detection data, making it suitable for widespread promotion and use.
[0085] Please see Figure 2 The figure shows a vehicle detection data classification system provided in the second embodiment of the present invention. The system includes:
[0086] The acquisition module 12 is used to preprocess the vehicle detection data when the vehicle detection data is acquired, so as to generate corresponding input text, the input text including several input vectors;
[0087] The processing module 22 is used to input several input vectors into a preset ERNIE model to convert the input vectors into corresponding first word vectors, and to perform sequence feature processing on the input text to generate corresponding second word vectors.
[0088] The concatenation module 32 is used to concatenate 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;
[0089] The optimization module 42 is used to optimize the preset DPCNN model through the equal-length convolution function, and after the feature map output by the equal-length convolution function meets the preset requirements, the optimized DPCNN model is used to perform max pooling on the equal-length convolution function to generate several maximum feature values.
[0090] The output module 52 is used to output a predicted classification label corresponding to the input text based on a preset algorithm and several maximum feature values.
[0091] In the aforementioned vehicle detection data classification system, the processing module 22 is specifically used for:
[0092] Several input vectors are input into the pre-trained layer of the preset ERNIE model, and the bidirectional Transformer encoder in the pre-trained layer performs transformation processing on each of the input vectors to convert the input vectors into corresponding first word vectors.
[0093] In the aforementioned vehicle detection data classification system, the output module 52 is specifically used for:
[0094] When several maximum feature values are obtained, the several maximum feature values are input into the normalized Softmax function so that the normalized Softmax function outputs the corresponding predicted classification probability based on the several maximum feature values;
[0095] The predicted classification probability is input into the prediction function so that the prediction function outputs a probability matrix containing various labels, and the predicted classification label corresponding to the input text is obtained based on the probability matrix.
[0096] The vehicle inspection data classification system described above further includes an evaluation module 62, which is specifically used for:
[0097] The predicted classification label is labeled with the predicted label probability, and the security level of the predicted classification label is judged based on the predicted label probability.
[0098] In the aforementioned vehicle detection data classification system, the expression for the equal-length convolution function is:
[0099] E = f(KX + b)
[0100] Where E represents the equal-length convolution function, K represents the convolution kernel, X represents the word vector matrix, and b represents the bias.
[0101] The third embodiment of the present invention provides a computer, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the vehicle detection data classification method provided in the first embodiment above.
[0102] The fourth embodiment of the present invention provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle detection data classification method provided in the first embodiment above.
[0103] In summary, the vehicle detection data classification method, system, computer, and readable storage medium provided in the above embodiments of the present invention can easily and quickly classify vehicle detection data with the joint participation of the ERNIE model and the DPCNN model, thereby significantly shortening the time spent on vehicle detection data classification and thus greatly improving the classification efficiency of vehicle detection data, making it suitable for widespread promotion and use.
[0104] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can reside in the same processor; or the above modules can be located in different processors in any combination.
[0105] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.
[0106] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.
[0107] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can 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, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0108] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0109] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.
Claims
1. A method for classifying vehicle inspection data, characterized in that, The method includes: When vehicle detection data is acquired, the vehicle detection data is preprocessed to generate corresponding input text, which includes several input vectors. Several input vectors are input into a preset ERNIE model to convert the input vectors into corresponding first word vectors, and the input text is subjected to sequence feature processing to generate corresponding second word vectors. The first word vector and the second word vector are concatenated to generate a corresponding word vector matrix, and the word vector matrix is input into a preset DPCNN model; The preset DPCNN model is optimized by using an equal-length convolution function. After the feature map output by the equal-length convolution function meets the preset requirements, the optimized DPCNN model is used to perform max pooling on the equal-length convolution function to generate several maximum feature values. Based on a preset algorithm, a predicted classification label corresponding to the input text is output according to several maximum feature values; The step of inputting the plurality of input vectors into a preset ERNIE model to convert the input vectors into corresponding first word vectors includes: Several input vectors are input into the pre-training layer of the preset ERNIE model, and the bidirectional Transformer encoder in the pre-training layer is used to transform each input vector one by one to convert the input vectors into the corresponding first word vectors. The step of outputting the predicted classification label corresponding to the input text based on a preset algorithm according to several maximum feature values includes: When several maximum feature values are obtained, the several maximum feature values are input into the normalized Softmax function so that the normalized Softmax function outputs the corresponding predicted classification probability based on the several maximum feature values; The predicted classification probability is input into the prediction function so that the prediction function outputs a probability matrix containing various labels, and the predicted classification label corresponding to the input text is obtained based on the probability matrix.
2. The vehicle detection data classification method according to claim 1, characterized in that: The predicted classification label includes predicted label probabilities, the sum of which is 1. After the step of outputting the predicted classification label corresponding to the input text based on a preset algorithm according to several maximum feature values, the method further includes: The predicted classification label is labeled with the predicted label probability, and the security level of the predicted classification label is judged based on the predicted label probability.
3. The vehicle detection data classification method according to claim 1, characterized in that: The expression for the equal-length convolution function is: Where E represents the equal-length convolution function, K represents the convolution kernel, X represents the word vector matrix, and b represents the bias.
4. A vehicle inspection data classification system, characterized in that, The system includes: The acquisition module is used to preprocess the vehicle detection data when it is acquired to generate corresponding input text, which includes several input vectors. The processing module is used to input several input vectors into a preset ERNIE model to convert the input vectors into corresponding first word vectors, and to perform sequence feature processing on the input text to generate corresponding second word vectors. The concatenation module is used to concatenate 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 is used to optimize the preset DPCNN model using an equal-length convolution function, and after the feature map output by the equal-length convolution function meets the preset requirements, to perform max pooling on the equal-length convolution function using the optimized DPCNN model to generate several maximum feature values. The output module is used to output a predicted classification label corresponding to the input text based on a preset algorithm and several maximum feature values. The processing module is specifically used for: Several input vectors are input into the pre-training layer of the preset ERNIE model, and the bidirectional Transformer encoder in the pre-training layer is used to transform each input vector one by one to convert the input vectors into the corresponding first word vectors. The output module is specifically used for: When several maximum feature values are obtained, the several maximum feature values are input into the normalized Softmax function so that the normalized Softmax function outputs the corresponding predicted classification probability based on the several maximum feature values; The predicted classification probability is input into the prediction function so that the prediction function outputs a probability matrix containing various labels, and the predicted classification label corresponding to the input text is obtained based on the probability matrix.
5. A computer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the vehicle detection data classification method as described in any one of claims 1 to 3.
6. A readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the vehicle detection data classification method as described in any one of claims 1 to 3.