493 results about "Paired Data" patented technology

A target detection method, system and device based on cross-modal fusion and an attention guiding mechanism and a medium, the target detection method comprises the following steps: obtaining a visible light-infrared image pairing data set, and carrying out data processing and division on the visible light-infrared image pairing data set to obtain a training set, a verification set and a test set; constructing a multi-modal target detection network; setting network training parameters; training and optimizing the multi-modal target detection network by using the training set, outputting a training weight file after the training is finished, verifying the trained weight file by using the verification set, and selecting the weight file with the highest precision as an optimal weight file; loading the test set and the optimal weight file into a multi-modal target detection network, and performing target detection on the test set to obtain a target detection result; the system, the device and the medium are used for bearing and implementing the method. The method has lower false detection and false detection conditions, and improves the robustness and accuracy of target detection.

The invention discloses a low-dose CT reconstruction method and system based on global optimization iteration deep learning, and the method comprises the steps: collecting CT cross section data of a normal dose, and generating low-dose CT data; the method comprises the following steps: establishing a dual-domain analysis compression iteration model, constructing maximum posteriori estimation based on a CT projection chordal graph data composite Poisson noise generation mechanism, subdividing CT projection data into a projection domain and a chordal graph domain, sequentially updating the projection data, the chordal graph data and image data, and constructing an iteration algorithm to realize CT global optimization modeling. Expanding an iterative algorithm into a trained reconstruction network; inputting the paired data into the expanded network for training, and storing a model with the minimum output result loss; a sub-network and an attention mechanism are added for CT data features, jump connection weights are finely adjusted, middle layer neural network parameters are finely adjusted and updated through middle supervision, and a global optimization iteration deep learning model NGACI-Net is obtained. According to the method, the problems that the low-dose CT image quality is poor, the reconstruction efficiency is low, and generalization and interpretability are lacked are solved.

The invention discloses a marine ecological disaster early warning method based on multi-source heterogeneous big data fusion. The marine ecological disaster early warning method comprises the steps that remote sensing image data and actual measurement environment data of a target sea area are acquired respectively; sequentially performing boundary segmentation and feature fusion on the remote sensing image data based on the image segmentation network to obtain multispectral feature data; performing space-time alignment on the multispectral feature data and the environmental data to obtain a space-time alignment data set; mapping each sample pair data in the time-space alignment data set into graph node data, and analyzing a connection relationship between any two adjacent graph node data to construct an adjacency relationship matrix; performing neighborhood aggregation processing on the adjacency relation matrix to generate a fusion feature vector; and inputting the fusion feature vector into a pre-constructed marine disaster prediction model to obtain a multi-dimensional prediction result, and carrying out ecological disaster real-time early warning based on the multi-dimensional prediction result. According to the method, the real-time performance of early warning of the marine ecological disasters is improved.

The invention provides a cache management system during large model reasoning, and relates to the technical field of large-scale artificial intelligence reasoning optimization. The system comprises a cache module, a cache manager, a retrieval enhancement generation controller, a reasoning time controller, a large language model reasoning engine, a process reward model reasoning engine, a result reward model reasoning engine, a Monte Carlo tree search scheduler and a runtime scheduler, the cache module is used for caching required key value pair data in a pre-filling stage, and the cache manager is used for storing the key value pair data in the pre-filling stage. The key-value pair data is used for storing context information of the preorder lexical elements, and the key-value pair data is also used for being reused in the decoding stage. According to the embodiment of the invention, in the pre-filling stage, the required key value pair data is generated and cached, and the cached key value pair data is used for storing the context information of the preorder lexical elements. In the decoding stage, the caches are reused, repeated calculation is avoided, and therefore the GPU memory is saved.

The invention relates to the technical field of information retrieval and generation, in particular to a retrieval enhancement method based on multi-modal data fusion and modal perception. According to the method, firstly, a dual-channel architecture is adopted to perform feature extraction and coding on a text and an image respectively, and mutually independent embedded representation spaces are constructed, so that high-quality collaboration and matching of cross-modal representation are realized; and a pseudo-pairing generation mechanism is introduced to effectively mine and reconstruct the existing non-paired data in the knowledge base. And designing a query modal perception and dynamic weighting mechanism for accurately controlling the fusion proportion of the image-text bimodal information in the retrieval stage so as to match the modal demand difference of different query contents. And further executing aggregation retrieval and reordering of the cross-modal information by using dynamic weighted fusion retrieval to generate a candidate set of multi-modal responses. According to the method, accurate matching and dynamic weight adjustment of the image-text content are realized, and the accuracy and expression integrity of the generated content are improved.

The invention provides a space projection error analysis-combined 4D millimeter wave radar and visual camera automatic calibration method. The method comprises the steps of constructing a sensor internal reference model; designing a composite multifunctional calibration board and a three-dimensional adjustable holder, performing rigid geometrical relationship modeling, and automatically associating physics-features-coordinates; collecting multi-frame synchronization frame pair data, and performing time sequence fine grit alignment and interpolation compensation to obtain a space-time alignment frame pair sequence; extracting cross-modal spatial constraints of the features and automatically screening and matching point pairs to obtain a high-confidence point pair set; performing point-point constraint and point-point error, point-plane error and pixel reprojection error minimization to obtain a standardized error term; and constructing an overall objective function, and performing spatial constraint and time deviation estimation to obtain corrected external parameters and synchronization deviation. According to the invention, through a full-automatic standardized process, the probability of artificial participation and error occurrence is greatly reduced; and multi-modal and multi-scene high-precision calibration is supported, and the flexibility and robustness are greatly improved.

The invention discloses a three-dimensional medical image visual language model pre-training method for reconstructing a hybrid strategy, which belongs to the technical field of medical image calculation and comprises the steps of constructing a medical image text pair data set, a language text mask reconstruction strategy, a visual image mask reconstruction strategy, a semantic perception fusion strategy and multi-task joint learning. According to the method, the large language model is finely adjusted, the diagnosis and attribute information in the medical report is extracted by using the finely-adjusted large language model, the efficient prompt is generated, and the large language model has relatively high generalization ability, so that the cost of manual annotation is greatly saved; according to the semantic perception fusion strategy, the text feature obtained by the text encoder and the image feature obtained by the image encoder are combined to obtain the new text feature, so that the text perceives the diagnosis and attribute information of the image in advance, the alignment of the image and the text in the embedding space is further optimized, and the pre-training efficiency is improved.

The invention discloses a precipitation field space downscaling method, and relates to the technical field of meteorological data processing, and the method comprises the steps: obtaining high-resolution topographic information data and precipitation pairing data composed of high-resolution and low-resolution precipitation data, and carrying out the normalization processing of the high-resolution topographic information data and the precipitation pairing data; inputting the normalized precipitation pairing data and the high-resolution terrain information data into a pre-trained precipitation field space downscaling model, and outputting to obtain a high-resolution precipitation field; wherein the precipitation field space downscaling model comprises a convolutional neural network and a de-noising diffusion probability model, and the de-noising diffusion probability model is constructed based on residual calculation, forward noise addition, a multi-scale feature extraction fusion network and reverse de-noising. According to the method, the high-resolution precipitation field can be reconstructed from the low-resolution precipitation field.

The invention discloses a video highlight detection method based on a weak supervision multi-mode large model, and the method comprises the steps: constructing the pairing data of a video clip and text description based on a video language large model, and configuring a highlight detection module for carrying out the significance score calculation of a video frame according to a highlight detection instruction, a highlight detection model is obtained through training in a multi-stage progressive training mode and is used for highlight detection of a video. According to the method, the precision and efficiency of highlight detection are remarkably improved by generating a high-quality pseudo tag and designing a time-sequence-sensitive highlight detection module and a progressive multi-stage training strategy, and the method has wide practical application value in scenes such as content optimization and personalized recommendation, video editing and manufacturing, social media, education and training and the like.

The invention provides a database knowledge body construction method and a structured query language generation method, a relational database knowledge body comprises a plurality of knowledge body units, and the plurality of knowledge body units are mutually connected through associated fields to form a knowledge network; each knowledge body unit corresponds to one data table, each knowledge body unit comprises multi-layer node data, and the multi-layer node data comprises table information, field information, field vectors, field enumeration value information and field enumeration value vector representation of the data tables. The construction method comprises the steps of obtaining first data table information based on data table metadata; reasoning and extracting the text code pairing data set through a large language model to obtain second data table information; and fusing the first data table information and the second data table information to construct a relational database knowledge body. Refined, low-redundancy, accurate, comprehensive and rich dependency data is provided for the TEXT2SQL task of the large language model, and the TEXT2SQL generation capacity of the large language model is improved.

The invention discloses a question and answer generation method based on a knowledge graph, and relates to the technical field of artificial intelligence and natural language process.The method comprises the steps that a bidirectional semantic encoder is adopted to process input text corpus, entity words and relational words are extracted, a first semantic vector is formed, and the first semantic vector is converted into a basic knowledge structure through a triple decomposer; aligning the basic knowledge structure with deep semantic information in a preset knowledge graph by using a hierarchical knowledge fusion network to generate a multi-dimensional knowledge feature matrix, calculating association strength among knowledge nodes by using a dynamic weight distributor, and constructing a hierarchical question and answer path according to a sequence of the association strength from high to low; and inputting the hierarchical question and answer path into a constructed attention model, and generating a question and answer pair data set according to a question and answer template library. According to the method, the knowledge extraction accuracy and the knowledge fusion effect are improved, the question and answer path construction process is optimized, and a question and answer pair data set with a complete structure is formed.

The invention relates to a knowledge-intensive visual question and answer automatic data generation method and device, and the method comprises the steps: constructing an original visual data set containing the professional knowledge of a target domain according to a static image, a video stream and multimedia content; extracting a representative frame sequence, converting the audio information into text information, and extracting character information in the static image to construct a structured visual instance database; according to the prompt text meeting the preset professional depth condition, establishing a three-level prompt system containing domain knowledge, an evaluation standard and a generation specification; generating a corresponding visual question and answer pair data set according to the dynamic cooperation of the main agent and the domain expert agent; generating a multi-agent quality evaluation system according to the quality evaluation result; and designing a difficulty grading mechanism according to the negative example sample. According to the method, the professionality, the accuracy and the diversity of the visual question and answer data are remarkably improved, and reliable data support is provided for training and evaluation of a multi-modal large model.

The invention discloses a vertical field large model-oriented high-quality data set construction method and system, and relates to the technical field of text processing, and the method comprises the steps: selecting seed text data, carrying out long text processing, segmenting the seed text data into a plurality of text segments, processing the text segments by using a large model tool to generate seed question and answer pairs, and constructing a seed question and answer pair; auditing and optimizing are carried out according to a preset auditing standard; making an instruction generation data set for the cue word, the text fragment and the seed question and answer pair according to a preset organization rule, performing fine adjustment on a pre-selected base model by using the instruction generation data set, and performing iterative optimization on the large model by testing the performance of the large model after fine adjustment and adjusting and optimizing parameters, thereby obtaining an intelligent extraction large model IELM; and based on the intelligent extraction big model, processing the text data by utilizing a preset script code, automatically generating knowledge question and answer pair data corresponding to the text data, and performing format conversion processing to obtain an instruction data set which is used for a training data set of the vertical domain knowledge big model.

The invention relates to the technical field of large language models, in particular to a complex multi-field guided teaching question and answer generation method, device and equipment and a computer storage medium. According to the complex multi-field guided teaching question and answer generation method, aiming at the problem that the model is single in application field, the corresponding multi-round question and answer pair data set is processed by using a multi-field data source, and the model is finely adjusted based on the multi-round question and answer pair data set, so that the problem that the related field is single is solved, and the applicability and practicability of the teaching question and answer model are enhanced; aiming at the problem that the model lacks guiding ability, the input of the teaching question and answer model is further processed through thinking chain-based question extraction and disassembly and prompt engineering and context learning mechanism-based processing, so that the model question guiding ability is improved; in order to solve the problem that teaching effect evaluation indexes are single, the invention provides a new evaluation system, and a plurality of evaluation indexes can more comprehensively and accurately evaluate and optimize model teaching effects.

The invention provides an improved retrieval enhancement generation method based on Embedded-FineTuning, which is used for preprocessing and cleaning large-scale multi-source heterogeneous data of a power grid to ensure the consistency and quality of the data, and dynamically dividing the processed text data based on a LumberChunker technology, dividing text blocks according to semantic turning points, and generating text blocks suitable for fine tuning training; then, performing question and answer pair generation of multi-round questions and answers on the basis of the text block by adopting a T5 model to form a question and answer pair data set for fine tuning; the method comprises the following steps of: designing a customized prompt template for an actual application scene in the field of an electric power grid to guide a model generation process; embedded-FineTuning fine tuning training is carried out on a question and answer pair data set, and model parameters are optimized, so that the embedded representation performance and accuracy of the model in the field of electric power grids are improved.

The invention relates to the technical field of meteorological analysis, and more specifically relates to a satellite-borne and multiband ground-based radar reflectivity factor joint verification method and a satellite-borne and multiband ground-based radar reflectivity factor joint verification system. The scheme comprises the following steps: setting data sources of a spaceborne radar and a ground-based weather radar; performing data quality control according to the data sources of the spaceborne radar and the ground-based weather radar to form a corrected data source; matching and comparing the reflectivity factors of the spaceborne radar and the ground-based weather radar according to the corrected data source to form optimal pairing data; establishing a fused networking radar consistency evaluation algorithm to determine an alarm list, and forming a problem level; standard comparison is carried out on the ground-based radar reflectivity factors to form a matching result of the standard database and the novel spaceborne radar FY3-G / PMR; evaluation classification is set, and evaluation results under different classifications of the matching result of the standard database and the novel spaceborne radar FY3-G / PMR are formed and stored.

The invention provides a multi-modal style migration method, system and equipment based on knowledge distillation, a teacher-student model is designed by introducing a knowledge distillation technology, and multi-modal guided style migration can be realized only by using an image data set to carry out one-time training by using a CLIP text image space alignment property. Compared with CLIPStyler, the speed and efficiency are greatly improved, and compared with LDAST, large-scale text image pairing data is not needed, and manpower and material resources are saved. Besides, compared with the fact that both CLIPStyler and LDAST are trained by designing fixed models, a spatial expansion mapping module is designed in the invention, so that the method can be inserted into all traditional image style migration networks to become a multi-modal style migration network, plug-and-play is realized, and the method has excellent adaptability.

The invention belongs to the field of computer vision, and particularly relates to a consistent identity picture generation method based on a diffusion model, which comprises the following steps of: generating optimal paired data by adopting a multi-modal large language model, the paired data comprising text data and corresponding image data; carrying out encoding processing on the text data by adopting a text encoder to obtain a text embedding vector; processing the image data by adopting an image encoder to obtain an image embedding vector; inputting the text embedding vector and the image embedding vector into an MLP module for fusion to obtain a fusion embedding vector; stacking all the fusion embedding vectors, and processing the stacked fusion embedding vectors by adopting a cross attention layer to obtain deep-level features; performing multi-stage adjustment on the deep feature to obtain a consistent identity picture; according to the invention, a coarse-to-fine multi-stage fine adjustment method is adopted, and the visual quality of the generated image is gradually optimized.

Provided is a noisy channel generative model of two sequences, for example text and speech, which enables uncovering the associations between the two modalities when limited paired data is available. To address the intractability of the exact model under a realistic data set-up, example aspects of the present disclosure include a variational inference approximation. To train this variational model with categorical data, a KL encoder loss approach is proposed which has connections to the wake-sleep algorithm.

The invention discloses a tendering and bidding field question-answering method and system based on a retrieval enhancement generation architecture, and relates to the technical field of tendering and bidding. According to the method, a tendering and bidding data set is collected and constructed, and fragment texts are vectorized and stored in a vector database through an embedded model; therefore, efficient retrieval of related information can be realized according to similarity retrieval. When a user asks a question, the system retrieves relative content according to the similarity and fills the content into a standardized verbal skill template to ensure the accuracy and specialty of question and answer, and then a large language model base is finely adjusted through question and answer pair data, so that the model has higher understanding and answering ability in the field. Compared with a question answering system based on deep learning, the accuracy and the breadth are greatly improved, and the professionality of system answering is ensured. According to the method, the problems of non-standard user question expression, incomplete question and answer knowledge coverage and low question and answer generation quality of a question and answer method in the bidding and tendering field are solved.

The invention discloses a severe weather image sharpening method based on continuous learning, and the method comprises the steps: 1, carrying out the preprocessing of an image data set under various severe weather conditions, including image pairing and multi-task division; 2, dividing the image restoration network into two parts of feature projection and output projection, designing a degradation sensing module for extracting degradation specific information output by the feature projection under different weather conditions, and establishing background supervision and degradation supervision by adopting a multi-teacher model for guiding learning of a student model; and 3, establishing respective optimization targets for the student model and the teacher model, inputting pairing data of different weather removal tasks, and optimizing the student network by a multilateral distillation normal form, thereby obtaining an optimal model with integrated image severe weather removal capability. According to the method, different types of severe weather images continuously acquired in a real scene can be dealt with, and the implementation effect of an image restoration network in a resource-limited scene is improved.

The invention discloses an image artifact restoration method and system based on an anti-fact diffusion model, and relates to the technical field of image processing, and the method comprises the steps: obtaining multi-modal data, and carrying out the data screening of the multi-modal data; preprocessing the screened data, and constructing a standardized data set; constructing an image artifact repairing model based on an anti-fact diffusion model; training the image artifact repair model through a standardized data set; and carrying out artifact restoration on the image polluted by the artifact through the trained image artifact restoration model. Only artifact-free data is used in the training stage, and the limitation of the lack of paired data on the current artifact repair technology is solved; the method is not limited to a single part of a single mode during restoration, and when images of different modes are processed, the quality of the restored images can still be ensured, and the high resolution and details of the images are kept; the image quality can be effectively improved, key details in the image are reserved, and the accuracy of the image is improved.

The invention discloses a marine target detection method and system based on progressive light thunder information deep fusion, and the method is realized based on an X-band radar and a visible light camera, and the method comprises the steps: receiving a radar echo signal and a visible light visual image of a to-be-detected marine target, and forming a plurality of paired data; preprocessing the radar echo signal to obtain radar echo data after matched filtering, and obtaining radar point cloud data through a constant false alarm algorithm; carrying out nonlinear modal calibration on the preprocessed radar echo signal and radar point cloud data to obtain a radar pseudo image, and carrying out weighted addition on the radar pseudo image and a visible light visual image to carry out first-stage fusion; inputting the paired radar pseudo image, visible light visual image and weighted fusion image into a pre-trained target detection model to obtain the category and position information of the marine target contained in the to-be-detected area; the target detection model adopts a two-stage fusion strategy and combines a gating attention mechanism to realize deep fusion of light thunder information and improve the detection precision.

The invention discloses a cross-view image geographic positioning method based on dynamic threshold pseudo tag self-training learning, and the method specifically comprises the following steps: introducing a difficult sample feature mining method, dynamically adjusting the loss weight of a sample according to the change of similarity, and building a dynamic difficult sample triple loss model; the method comprises the following steps: dynamically adjusting a confidence threshold value of a sample by adopting an index moving average weighting method, iteratively training and screening an unlabeled sample, namely a pseudo label, establishing a pseudo label self-training mechanism of a dynamic threshold value, mining and utilizing non-paired data, and solving the problem of high manual labeling cost; a reference image most similar to a query image is found through image retrieval, and the offset of a query position is predicted. Experiments on CVUSA and CVACT data sets show that as the distance threshold increases, the accuracy of the cross-view image geographic positioning method based on dynamic threshold pseudo tag self-training learning presents a stable rising trend, and the cross-view image geographic positioning method based on dynamic threshold pseudo tag self-training learning is superior to other methods under the same threshold condition.

The invention discloses a protein generation model optimization method based on deep learning, and relates to the technical field of protein generation. The optimization method comprises the following steps: generating a candidate protein electron density distribution diagram based on the structural characteristics of a target spot by adopting a pre-trained diffusion model; converting the candidate protein electron density distribution diagram into a corresponding amino acid sequence; determining a functional index value corresponding to the amino acid sequence, and constructing a feedback data set; setting a reward threshold value, and marking the feedback data set as a positive sample data set and a negative sample data set according to the reward threshold value; constructing a utility function taking the reward threshold as a reference point, and respectively calculating utility values of the positive sample data set and the negative sample data set; and performing iterative optimization on the diffusion model according to the utility value. By adopting the technology provided by the invention, the dependence on preference paired data of large-scale and high-quality protein sequences can be avoided, and the performance of the protein generation model can be effectively and continuously improved.

The invention discloses a time sequence remote sensing composite under-cloud reconstruction method and system based on observation modeling and a radiation reconstruction network, and belongs to the technical field of remote sensing, and the method comprises the steps: carrying out the mask processing of a composite cloud time sequence remote sensing image based on observation, and obtaining a real degraded sequence after mask; inputting the real degraded sequence into the trained radiation reconstruction network to obtain a cloudless surface reflectance sequence, and completing composite under-cloud reconstruction; wherein the training of the radiation reconstruction network comprises the following steps: constructing a simulated degradation-cloudless pairing data set; building a radiation reconstruction network; and training the constructed radiation reconstruction network by using the constructed simulated degradation-cloudless pairing data set to obtain a trained radiation reconstruction network. According to the method, joint processing of the space-time spectrum data is realized, the method can be effectively applied to joint processing of thin cloud radiation deviation correction and thick cloud information loss reconstruction, and finally reconstruction of the surface reflectance sequence under the composite cloud is realized.

The invention relates to a depth-of-field expansion method for an optical surgical microscope. The method comprises the following steps: collecting an open-source high-definition high-frequency image database; acquiring point spread functions corresponding to different focus depths of the operating microscope; constructing a random depth surface type database; simulating a partial out-of-focus image of the non-flat sample imaged by the optical microscope; simulating a perfect focusing image of the flat sample imaged by the optical microscope; combining a large number of paired data partial out-of-focus images-perfect focusing images; and constructing an encoder-decoder structure network model from the image end to the image end to carry out model training and application. The problem that high-quality pairing training data conforming to imaging of the operating microscope is difficult to obtain for model training is solved, partial defocusing / focusing simulation data of the operating microscope conforming to a real physical rule generation effect can be synthesized in batches, a high-quality training data set is constructed at low cost, the number, diversity and randomness of the training data are ensured, and the training efficiency is improved. And defocus information is efficiently learned and recovered while focusing information is kept.

An audio translation labeling and analysis method based on a multi-agent cooperation mechanism belongs to the field of information technology and artificial intelligence, and is characterized by comprising the following steps: S1, translating an audio file into a text file; s2, outputting a dialogue set of text structure arrangement for the text file through a model in which domain knowledge and language style learning are introduced; s3, extracting and structurally expressing a time expression, a place name and an abstract paragraph by adding a large language model of a prompt project, and associating an event background with a session context to obtain an information set; s4, splicing the dialogue set and the information set to obtain a document set; s5, forming a text-label pair data set by a series of texts and the corresponding labels with inclusion relations, and constructing the text-label pair data set into an instruction data set conforming to instruction fine tuning format specifications; s6, performing fine adjustment on the large language model base according to the instruction data set to obtain a fine-adjusted model; and S7, inputting the document set into the fine-tuned model, and labeling all levels of labels to obtain the document set with the labels, which is used for full-process automatic processing from audio transcription to text optimization, information extraction, labeling and analysis report generation.

The invention relates to a territorial space planning data enhancement method based on machine learning, and belongs to the field of territorial space planning. The method comprises the following steps: data acquisition and processing: acquiring remote sensing image data from a plurality of data sources, and generating a low-resolution (LR) image pair and a high-resolution (HR) image pair; multi-scale feature extraction: extracting features of different scales, and outputting a multi-scale convolution feature image set; performing multi-scale feature fusion: performing weighted average fusion on the feature image set, and obtaining a fused feature image through residual connection; image enhancement: performing up-sampling of the feature map through a sub-pixel convolution layer, redistributing channel information to a spatial dimension, and obtaining an enhanced image through residual connection; training and optimizing the model: training the model by using the paired data set, fusing the mean square error loss and the perception loss by using a loss function, and optimizing the model by using an Adam optimizer. According to the method, more comprehensive data support is provided for space-time dynamic analysis in territorial space planning through data enhancement.

The invention relates to the technical field of computer graphics and computer vision, and particularly provides a monocular relief depth estimation model construction method based on a progressive fine tuning strategy. The method comprises the following steps: constructing a pseudo real relief image data set by using a text-based image generation model; generating a depth pseudo label according to the pseudo real relief image data set; obtaining an initial relief depth estimation model based on the pseudo real relief image data set and the depth pseudo label thereof; collecting a multi-view real relief image, and generating a real relief image and depth label pairing data set in combination with three-dimensional reconstruction and geometric detail optimization; according to the method, the initial relief depth estimation model is subjected to lightweight fine adjustment by adopting low-rank adaptation LoRA, a final monocular relief depth estimation model is obtained, high-quality training data is effectively constructed in the absence of real labels, and the model generalization ability and the prediction precision are considered at the same time.