1026 results about "Small sample" patented technology

The invention discloses a cross-modal adaptive interaction-based few-sample industrial anomaly detection method, which comprises the following steps of: calculating total loss according to aligned final features, normal semantic projection, abnormal semantic projection, optimized visual features and optimized text features, and updating parameters of an industrial anomaly detection model by using the total loss. A trained industrial anomaly detection model is obtained; and inputting a to-be-detected industrial image into the trained industrial anomaly detection model to obtain the aligned final feature, the anomaly semantic projection, the optimized visual feature and the optimized text feature for judging the anomaly condition of the to-be-detected industrial image. According to the method, under the condition that only a small number of normal samples are needed, the distinguishing capacity of the model for normal and abnormal features can be remarkably enhanced, dependence on labeled data is reduced, and an industrial anomaly detection solution which is high in precision, low in cost and capable of being rapidly deployed is provided for intelligent manufacturing.

The invention relates to a copper-clad plate layering defect diagnosis method based on pattern recognition, and the method comprises the steps: forming a rich input sample set through spatial multi-channel imaging, variable illumination and time sequence synchronous sampling in order to solve the problems that the recognition capability of tiny layering defects is weak, and the recognition is easily influenced by noise and environment interference in the existing detection; according to the method, an attention mechanism and small sample discrimination are combined, high-confidence-coefficient defect identification is realized, region self-adaptive re-detection is further triggered in real time to reduce the missing detection risk, and finally defect distribution mapping is automatically generated and data archiving and traceability analysis are supported.

The invention discloses a small sample industrial defect detection system based on a multi-stage diffusion model. The small sample industrial defect detection system comprises a multi-stage diffusion generation module and a defect detection module based on multi-scale attention and physical constraint. The multi-stage diffusion generation module divides the diffusion generation process into three stages of global structure reconstruction, local detail refinement and texture feature synthesis through a stage control mechanism, and gradually guides feature evolution and improves the generation effect aiming at the quality and diversity problems of defect image generation under the small sample condition; the defect detection module based on multi-scale attention and physical constraint adopts a coding structure fusing local window attention and global attention, through multi-scale feature extraction and fusion, significant features of a defect area are effectively captured, gradient smoothing loss and edge energy consistency loss are introduced at a decoder end, and the defect detection accuracy is improved. Therefore, high-quality reconstruction of a normal area and effective suppression of an abnormal area are realized.

The invention relates to the technical field of reservoir evaluation, in particular to a deep learning-based reservoir evaluation model construction method and system. The method comprises the following steps: acquiring logging data and core analysis data of a target work area, and respectively carrying out abnormal value processing and missing value filling to obtain complete logging data and complete core data; calculating reservoir pore structure parameters based on the complete core data; and inverting the macroscopic porosity of the reservoir based on the complete logging data. According to the method, the accuracy, the stability and the robustness of reservoir evaluation are remarkably improved by fusing the core data and the logging data and combining the pore structure parameters, the macroscopic porosity and the geological attention mechanism, the problems of data missing and small samples are solved, and the interpretability and the geological consistency of the model are enhanced.

The invention discloses a small-sample high-density crowd detection method based on text cross-modal guidance, and belongs to the field of computer vision and cross-modal intelligent detection. The method aims at solving the problems that an existing method is high in labeling cost, few in sample, poor in generalization, lack of semantic guidance and insufficient in shielding scene adaptation. According to the method, the semantic guiding capacity of a CLIP text encoder and the feature extraction capacity of DINOv2 and SAM pre-training visual models are fused, three light-weight trainable modules are constructed, and text and visual feature fine-grained alignment is achieved. According to the technical scheme, detection is completed through data acquisition and labeling, feature preprocessing, model initialization, cross-modal feature fusion, mask generation and screening, module optimization and performance iterative optimization. According to the method, high-precision detection can be realized only by a small number of labeled samples, the labeling cost is remarkably reduced, the detection robustness of the high-density shielding scene is improved, and the method has the advantages of flexible scene adaptability and lightweight deployment, and is suitable for scenes such as monitoring video analysis and public safety early warning.

The invention discloses a radiation source open set identification method and device based on a dynamic group isovariant network and a generative adversarial network. The method comprises the following steps: preprocessing a plurality of radiation source individual signals of known categories to obtain a data set for closed set identification; a dynamic group isovariant network is constructed, the network is trained on the obtained known category data set, and optimal model parameters are found; simulating virtual unknown category data on the basis of the known category data by using the generative adversarial network, and retraining the constructed recognition model by combining the unknown category data with the known category data to obtain a model capable of being used for open set recognition; and combining actually acquired unknown category data with known category data to serve as a test data set, extracting features through a retrained model, inputting the features into an OpenMax layer, judging whether the category is a known category or an unknown category according to threshold distribution, and detecting the recognition performance of the retrained network open set. The training efficiency is high in small sample recognition task, the classification recognition performance is good, and the robustness is high in open set recognition task.

According to the small sample target detection method based on prototype network feature enhancement, firstly, a foreground feature aggregation module is adopted, redundant background information is removed in the category prototype construction process, and purer category features are extracted; and a condition information coupling module is utilized, and the category prototype is dynamically adjusted in combination with the characteristics of the query image, so that the adaptability is better. And then, dynamically supplementing the most similar support sample through a support sample expansion module in the training process, improving the expression ability of the category prototype, and enhancing the generalization performance of the model. Finally, an optimization strategy based on meta learning is adopted, and a joint optimization mode of classification loss, regression loss and meta loss is combined, so that the matching precision of the category prototypes is improved. The method has the advantages that the detection precision and generalization ability of small sample target detection are improved under the condition that extra data labeling cost is not increased, and the method is suitable for various application scenes.

The invention discloses a fault diagnosis method for an aviation hydraulic pump, belongs to the field of aviation hydraulic pumps, and aims to solve the problems of lack of feature information and low diagnosis precision caused by insufficient samples in an existing method. The method comprises the following steps: step 1, constructing a double-source feature extraction module, and respectively extracting time-frequency domain prior features and high-dimensional features of vibration signals; the extracted features are used for calculating a prior feature prototype and a high-dimensional feature prototype of a fault category; step 2, constructing a fault prototype strengthening module, and based on a self-attention fusion mechanism, performing weighted fusion on the prior feature prototype and the high-dimensional feature prototype to generate a fault strengthening prototype; and step 3, constructing a small sample fault classification module, measuring the distance between the test sample and the fault enhancement prototype through cosine similarity, outputting a fault category corresponding to the maximum similarity, and completing fault classification.

The invention provides a small sample cross-domain fault diagnosis method based on multistage feature alignment, and the method comprises the steps: collecting rotating machinery monitoring data in a source domain and a target domain, and constructing a cross-domain small sample fault diagnosis task data set with scarce target domain samples; constructing a diffusion model combining a multi-head self-attention mechanism and a normalization strategy, and generating an enhanced sample with target domain feature distribution; a shared feature extraction network fusing a source domain, a target domain and a generated sample is built, and unified modeling of diagnosis features is achieved; and designing a joint loss function which comprises source domain classification loss, dual-MMD distribution alignment loss, cross-domain triple loss and enhanced consistency regular terms, constructing a multi-stage feature alignment mechanism, and realizing fault feature migration and accurate diagnosis under the condition of small samples of a target domain. According to the method, source domain diagnosis knowledge can be effectively migrated under the condition of target domain data scarcity, the fault feature discrimination capability and the cross-domain adaptability of the model are enhanced, and the method is suitable for intelligent diagnosis tasks of the rotating machinery under complex working conditions.

The invention provides an unknown encrypted traffic identification method and system based on small sample incremental learning, and a storage medium, and the method comprises the steps: 1, carrying out the fine-grained classification of known traffic: extracting the features of each level of original encrypted traffic, building a respective variational auto-encoder for each type of known traffic, generating a potential representation, and carrying out the fine-grained classification of the known traffic; inputting into a classifier to classify known attacks in a fine-grained manner; step 2, specific label distribution of unknown traffic: judging whether the sample is a drift sample or an unknown sample by adopting a scoring function, and performing hierarchical clustering on the samples according to each hierarchical feature to realize label distribution of the unknown traffic; and step 3, dynamically updating the classification model: training a new classifier by adopting a new sample, connecting other classifiers to form a classification graph, and updating nodes of the classifiers by adopting a graph attention network to realize small sample incremental learning. The method has the beneficial effects that low-sample incremental modeling of a new class is effectively supported, and the fine-grained recognition capability and the model generalization adaptability are remarkably improved.

The invention discloses a power transmission line small sample abnormal data generation method and system based on a diffusion model and feature migration, and belongs to the technical field of abnormal detection. The problem that stable and high-fidelity abnormal data can be generated only when the demand quantity of abnormal samples is much larger than that of actually obtained abnormal samples is solved. The method comprises the steps that an abnormal defect library and an equipment main body library are constructed based on defect reference images in a data set and corresponding labels, and the data set at least comprises one defect reference image of each type; training an abnormal defect identification baseline model based on the data set and verifying to obtain a low-quality image; comparing the abnormal defect library and the equipment main body library with low-quality images by using an image comparison technology to obtain an abnormal defect reference image and a base image to be generated; generating a to-be-migrated region mask based on the to-be-generated base map; and based on the to-be-generated base map and the to-be-migrated region mask, using a diffusion model to migrate the abnormal region features of the low-quality image to the to-be-generated base map for anomaly generation. The method is used for generating the small sample abnormal data of the power transmission line.

The invention discloses a small sample modeling stability evaluation method and system based on Bootstrap resampling, and particularly relates to the technical field of computers and data intellectualization, the method comprises the following steps: completing data preprocessing and stable stage identification under a unified time base, and forming a segment set; an average block length is used as a decision quantity, an SBB is optimized to construct a sample space, and an optimal block length is determined in combination with variance consistency and nominal coverage rate consistency criteria; carrying out re-sampling training around short / medium / long scales, and collecting layered indexes such as prediction, parameters and features; stability calculation and empirical coverage rate calibration are completed based on intra-scale statistics and inter-scale weighted mixing, and a pseudo-stationary diagnosis score is output; and finally, engineering judgment and backspacing optimization are carried out according to the coverage rate, the correlation maintenance and the risk threshold. The system records random seeds, block metadata and model version generation evidence pointers, has the characteristics of traceability and reverifiability, and is suitable for scenes of production lines, network traffic, finance and the like.

The invention provides an industrial defect sample generation method based on small sample fine tuning and a controllable diffusion model, and the method comprises the steps: employing a surface defect image small sample data set, and screening out a model fine tuning data set through employing a pre-constructed image distribution evaluation module; finely adjusting the weight of the pre-trained stable diffusion model; an industrial defect sample conforming to the specified defect feature is generated by combining a trained space condition control neural network with a preset identifier as a cue word; and calculating a distribution difference value between the industrial defect generation sample and the original small sample data set, selecting images with the distribution difference value smaller than a preset threshold value, and adding the selected images into the original small sample data set to form an expanded training data set. According to the method, the small sample data set is utilized, the diffusion model weight is finely adjusted, the cue word containing the preset identifier, the space condition and the space condition control neural network are combined, the training samples are expanded, the calculation cost is reduced, and the application range is wide.

The invention discloses a transformer online life prediction method under an unbalanced multi-source small sample condition, which comprises the following steps of: eliminating noise signals from collected transformer vibration signal fault sample data, and expanding the number of samples on the premise of retaining minority class sample data characteristics; performing standardization processing on the acquired transformer fault sample data, dividing the data into a training set, a test set and a verification set, and respectively extracting features; a CNN transformer fault diagnosis model is constructed, and a fault diagnosis result 1 is obtained through the CNN model of the signals processed by the RP; constructing a transformer fault diagnosis model, and adjusting model hyper-parameters by using the verification set to obtain a fault diagnosis result 2; and sending the obtained fault diagnosis result 1 and the fault diagnosis result 2 into an FCNN model to adjust the FCNN model, and then obtaining a fault diagnosis result through the test set. According to the method, the problems of misjudgment and missed judgment of minority class fault samples can be effectively solved, and the diagnosis precision of the model is improved.

The invention discloses a noise-driven small sample distribution external detection method and system, and the method comprises the following steps: 1, obtaining a data set which comprises a small sample training set and a test set; 2, constructing a detection model, wherein the detection model comprises a data coding module, a noise extraction module and a fusion detection module; 3, performing noise enhancement on the noise region set, constructing a loss function in combination with the clean region set, training the detection model by adopting a gradient back propagation algorithm according to the loss function, and testing the detection model by adopting the test set; 4, performing small sample distribution external detection according to the tested detection model; according to the method, the problem of poor noise robustness of a current small sample distribution external detection method is solved, and the ID classification and OOD detection performance of a small sample learning model is improved.

The invention discloses a grain crop grain classification detection method and system based on small samples, and relates to the technical field of grain quality detection.The method comprises the steps that hyperspectral images of grain crop grains are obtained, quality categories are marked, and then multi-modal features are extracted to construct a classification data set; the data set is used for training a multi-modal fusion classification model, a feature fusion network carries out dynamic weighted fusion on multi-modal features by means of an attention mechanism, a high-dimensional fusion feature vector is generated, and a classifier detects a quality category according to the high-dimensional fusion feature vector. During detection, the multi-modal features of the to-be-detected grains are input into the trained model, and then the quality category can be obtained. The method effectively solves the problems that a single modal method is difficult to deal with complex differences among grain crop grain types, high in hyperspectral data dimension, few in samples, easy to over-fit and the like under the condition of small samples, dynamic focusing of key information is realized by introducing an attention mechanism to optimize modal fusion weight distribution, and the method is suitable for popularization and application. And the classification accuracy and the model robustness are improved.

The invention discloses a zero sample / small sample industrial defect detection method and system, and the method comprises the steps: freezing the parameters of a CLIP image encoder in a segmentation branch, extracting image features, projecting the corrected features, carrying out the one-by-one product operation of the image features of different layers and corresponding text features, and splicing to generate an image-text similarity graph; in the classification branch, the extracted final image features are multiplied by the text features, and a positive / abnormal classification result of the image is obtained after processing through a learnable adapter; the parameters of the adapter are optimized through the difference between the output result of the minimization segmentation branch and the classification branch and the real label, and the zero sample detection mode directly generates an abnormal graph through the matrix multiplication of the image and the text features among the layers; a reference sample is introduced in a small sample detection mode, the cosine similarity between each layer of features and input sample features is calculated, and an obtained similarity graph and an anomaly graph generated by a zero sample are added to form a final anomaly thermodynamic diagram.

The invention discloses a target prediction method for small sample data, and the method comprises the steps: extracting a target small sample time sequence data set, taking moments as groups, constructing a phase distribution sequence, and constructing a phase distribution transfer matrix; constructing a position generation method according to the phase distribution sequence and the small sample time sequence data set; then generating random phase distribution information by the system, and rapidly obtaining a large-scale model prediction training data set in combination with a phase distribution transfer matrix and a position generation method; finally, according to the model prediction training data set, an intelligent prediction model is constructed, model parameters are updated by pre-training the intelligent prediction model, and target intelligent prediction oriented to small sample data is achieved. The method can effectively solve the problem of state transition between adjacent moments, further generates large-scale high-quality training data, remarkably improves the accuracy of an intelligent prediction model, and solves the problem of audio and video generation in a real scene such as a small sample scene.

The invention provides a bearing fault variable working condition diagnosis method based on a multi-scale convolutional network and MAML, and relates to the technical field of equipment fault diagnosis. The method comprises the following steps: firstly, introducing fast Fourier transform to pre-process an original time domain vibration signal; secondly, a fault diagnosis model based on a multi-scale convolutional network and MAML is applied; and then, an internal and external circulation updating method based on model-independent element learning is adopted, so that the model can quickly adapt to a new task, and a relatively good prediction effect can be achieved only through a small amount of fine adjustment. According to the method, multi-scale feature extraction and meta-learning are creatively combined, the generalization ability of the model under variable working conditions is remarkably improved, the problem that a traditional fault diagnosis method depends on a single working condition and a large sample size is effectively solved, and the method is particularly suitable for small-sample and multi-working-condition bearing fault diagnosis scenes on an industrial site.

The invention discloses a transformer substation operation state evaluation method and system based on a comparative learning algorithm, and the method comprises the steps: obtaining the operation state historical data of transformer substation equipment, and constructing a transformer substation historical operation state data set; taking the state characteristic index of the primary equipment as a substation state evaluation index, and constructing an operation state characteristic set; determining operation state levels of all samples; establishing an initial contrast learning model based on a neural network model, extracting feature vectors of all samples, generating positive and negative sample pairs from the samples of the same and different operation state levels, constructing a contrast loss function according to the distance of the feature vectors and the positive and negative of the sample pairs, and training a substation state evaluation contrast learning model; and monitoring the operation state of primary equipment in the transformer substation in real time, inputting the trained model, and outputting the operation state grade of the transformer substation. Limited samples can be fully utilized, good evaluation performance is shown under the conditions of small samples and unbalanced samples, and the operation and maintenance work efficiency of the transformer substation can be effectively improved.

The invention relates to the technical field of data analysis, and discloses a time series data anomaly detection method and system based on convergence trajectory feature analysis, and the method comprises the steps: collecting and preprocessing multi-dimensional time series data of at least one industrial device belonging to the same type, and obtaining a standardized single device feature sequence of each industrial device; training a baseline auto-encoder model based on the standardized feature tensor to obtain baseline parameters and baseline features of the baseline auto-encoder model; performing disturbance training on the baseline auto-encoder model by taking the standardized single device feature sequence of each industrial device as a single sample based on the baseline parameters to obtain convergence trajectory features of the baseline auto-encoder model, and obtaining a standard digital portrait library based on all industrial devices; and carrying out feature attribution and root cause analysis on the digital portrait. According to the method, the sensitivity and the detection accuracy of weak signals, small samples and complex behavior anomalies are remarkably improved.

The invention discloses a method for identifying boundary features of a three-dimensional model of a closed cavity of an airplane, and provides a method for automatically and intelligently detecting the boundary features of the closed cavity in order to solve the problems that the extraction precision of the boundary features of the closed cavity is poor and the efficiency is low. Providing a minimum allowable gap recommendation mechanism based on a decision tree, and automatically extracting boundary part information; analyzing and screening an accurate boundary surface through a geometric distance and a normal vector included angle; and constructing a part attribute adjacency graph, and identifying the gap type of the closed cavity based on a GNN enhanced VF3 sub-graph isomorphism algorithm to realize automatic identification and marking of boundary parts and gaps of the closed cavity in the three-dimensional model. The method solves the problems of low manual detection efficiency and large error in a complex assembly environment, remarkably improves the precision and automation level of closed cavity boundary feature recognition, and is suitable for assembly area modeling scenes with few samples and high precision requirements in the field of aeronautical manufacturing.

The invention discloses a new energy power station inspection method based on a multi-modal large model small sample open set, and the method constructs a multivariate text learnable prompt, and guides a detection model to form a finer-grained category decision boundary through the text information related to an aggregation task. Due to the lack of real unknown class samples in the training process, the mining of unknown class pseudo samples is regarded as a bipartite graph matching task for the first time, and the model is optimized to form a compact unknown class decision boundary by adding unknown class virtual nodes, constructing a cost matrix and mining the unknown class pseudo samples. In order to solve the problem that known classes and unknown classes of small samples are prone to confusion, the method proposes unknown class optimization loss based on cost perception, considers the classification and positioning quality of the unknown classes, and improves the open set detection performance of the model. According to the method, transformation from a single-mode vision small model to a multi-mode vision large model is realized, and good generalized small sample open set target detection performance can be obtained only by a small amount of training data.

The invention provides a bearing small sample data expansion method and system based on a variational auto-encoder, and belongs to the field of deep learning and data enhancement. The problems that a traditional generation model has limitation in bearing small sample data, feature fuzziness and distortion are prone to occurring, and the data set quality is poor are solved. According to the method, a deep VAE framework is constructed, and a dimension reduction module, a data expansion module and a dimension raising module are used in a potential space; dimensionality reduction is performed on high-dimensional data by adopting a UMAP algorithm, so that the topological structure of the data is effectively reserved, and the extraction efficiency and quality of data features are improved; a Gaussian mixture model combining regularization and particle swarm optimization optimization is used for fitting distribution of scattered small sample data, new data with fusion features are expanded through sampling, and data diversity is increased; a radial basis function is used for nonlinear data dimension raising, new data can be ensured to be accurately mapped back to a high-dimensional space, meanwhile, the relation between features is reserved, and defect data with fusion features is reconstructed through a decoder.

The invention discloses a semi-supervised efficient video SAR shadow tracking method based on small samples, and belongs to the field of radar target recognition. The method provided by the invention solves the limitation of a traditional shadow tracking technology, and improves the detection precision and robustness of the shadow region through an improved segmentation model and a semi-supervised strategy; on the basis of the designed new mixed training strategy, labeled data and unlabeled data are effectively integrated, and the learning efficiency and generalization ability of the model are remarkably improved. The deep learning technology and the semi-supervised learning strategy are combined, limited annotation data and rich non-annotation data are fully utilized, the defects of a traditional method under the small sample condition are overcome, and the detection and tracking precision of the shadow area in the SAR image is effectively improved.

The present invention discloses a small sample fault diagnosis method based on an improved TimeGAN model, comprising: collecting elevator operation data, including normal samples and fault samples; establishing an improved TimeGAN model; inputting fault samples into the improved TimeGAN model to generate fault data; merging the elevator operation data and the generated fault data into a sample set, and extracting the time-frequency features of the sample set by using wavelet packet decomposition; dividing the extracted time-frequency features into a training set and a test set, and inputting the training set into a CNN model for training; inputting the test set into the trained CNN model, obtaining fault diagnosis results and verifying the fault diagnosis results. The method generates fault samples with time series features, reduces the imbalance of original data distribution, and fully extracts the time-frequency features of samples to obtain efficient and accurate fault diagnosis results, thereby realizing safe maintenance of intelligent equipment.

The invention discloses an object surface defect segmentation method based on small samples, and the method specifically comprises the steps: collecting images of different types of object surface defects and corresponding segmentation masks, constructing a small sample data set, and dividing the small sample data set into a training set and a test set; constructing training and testing tasks, and respectively extracting a corresponding support set and a query set for each task; constructing and training a defect segmentation network, wherein the defect segmentation network comprises a query prior mask generator, a prior-guided bidirectional feature interaction module, a prototype-loss compensation module and a context-aware attention-guided feature aggregation decoding module; and after training is completed, inputting the support set and the to-be-tested query image into the converged defect segmentation network, and outputting a query image prediction result. Experiments prove that the method realizes remarkable performance improvement on an FSSD-12 data set, and shows that the method has obvious advantages in a small sample defect segmentation task.

The invention discloses a special vehicle target detection algorithm under extremely small samples and small targets. The method is improved on the basis of a baseline algorithm YOLOv11n, and a special vehicle target detection method YOLOv11-SV is provided. The method comprises the following steps: firstly, autonomously constructing a minimum sample data set of a special vehicle; secondly, designing a new attention mechanism module WT-SE to improve the attention of the model to a vehicle target; then, a minimum target detection head is added in the feature fusion stage to adapt to detection of a smaller-scale special vehicle target; and finally, replacing a model loss function with EIoU Loss so as to improve the convergence speed of the model and further improve the detection accuracy. Through contrast experiments, the improved algorithm YOLOv11-SV has a good detection effect, and the improved network model has a certain generalization ability.

The invention provides a production control method and system based on SPC alarm linkage process locking, and relates to the technical field of quality management and control, and the method comprises the steps: deploying an SPC sensing node used for collecting technological parameters at a cutting edge position where a machining tool of production equipment is in contact with a workpiece, or at a contact point where a probe of online measurement equipment is in contact with the surface of the workpiece; the method comprises the following steps: acquiring process parameter data of a key production process in real time through an SPC sensing node deployed at a cutting edge position or a contact point; inputting the process parameter data as real-time data or small sample expansion data into an SPC judgment engine; and performing statistical process control analysis on the real-time data or the small sample expansion data through an SPC judgment engine so as to model data clusters which are densely distributed and have stable parameters in the data into a virtual reference curved surface representing a normal state domain. According to the invention, the accuracy of abnormity determination can be improved, and timely management and control of quality abnormity and minimization of production loss can be realized.

The invention discloses a defect identification model construction method and device based on electric power inspection small samples, a storage medium and computer equipment, and the method comprises the steps: obtaining an electric power inspection picture sample set, and obtaining a defect identification model based on the defect tag carrying condition of each electric power inspection picture sample; dividing the electric power inspection picture sample set into a labeled subset and a non-labeled subset, wherein the number of samples of the labeled subset is significantly smaller than the number of samples of the non-labeled subset; performing fine tuning on the teacher model pre-trained based on the general image data set through the labeled subset to obtain a fine-tuned teacher model; based on the non-label subset, through the fine-tuned teacher model, outputting defect category probability distribution of each electric power inspection picture sample in the non-label subset, and generating a soft label of the electric power inspection picture sample based on the defect category probability distribution; and performing model training on the initial recognition model based on the labeled subset, the unlabeled subset after the soft label is generated and the target loss function to obtain a defect recognition model.