1011 results about "Adversarial network" patented technology

The invention discloses a mechanical arm positioning and grabbing method based on machine vision, and relates to the technical field of machine vision and mechanical arm control, the method comprises the following steps: synchronously acquiring RGB-D images of a target scene through a multi-view camera array, and generating three-dimensional point cloud data through data fusion; an improved LSD algorithm and a PnP algorithm are adopted to calculate the initial pose of the target object, illumination distortion is eliminated in combination with the generative adversarial network, and three-dimensional coordinates are output; a mechanical arm motion error transfer model is constructed based on Monte Carlo simulation, and a candidate grabbing scheme set is generated through reinforcement learning; and an optimal grabbing scheme is screened through a preset priority evaluation rule, and a mechanical arm joint movement track and a control instruction set are generated. Through multi-modal data fusion and a nonlinear optimization algorithm, the technical problems of large target positioning deviation and sensitive illumination interference in a complex environment are solved, and the grabbing precision and robustness of the mechanical arm are improved.

The invention provides an oil extraction equipment fault monitoring system and method, and belongs to the technical field of oil extraction equipment fault monitoring. The method comprises the following steps: acquiring operation data of oil extraction equipment, and performing feature extraction on the acquired operation data to obtain a target feature vector; fusing the obtained target feature vector with a historical fault case library and an oil extraction equipment physical constraint equation, and constructing a dynamically updated knowledge graph; based on the space-time causal adversarial network, analyzing the distribution offset of the target feature vector in the space-time dimension, detecting an abnormal event and outputting an abnormal type label; and according to the output abnormity type label, combining with a knowledge graph, tracing a propagation path of an abnormal event, and calculating a fault probability of a root cause component through a Bayesian network so as to carry out monitoring and early warning on the oil extraction equipment. According to the method, accurate fault detection and root cause positioning are realized through multi-modal data fusion and the dynamic causal knowledge graph, and the equipment shutdown risk and the operation and maintenance cost are remarkably reduced.

The invention discloses an artificial intelligence-based advertisement effect evaluation method and system, and the method comprises the steps: synchronously obtaining multi-source data containing a user behavior data flow and an advertisement putting index flow through a distributed collection engine, and generating a time-space synchronous multi-dimensional data cube; performing feature decoupling on the multi-dimensional data cube, and outputting a dynamic feature topology network with a weight; inputting the dynamic feature topology network into an adversarial training framework, and finally outputting an advertisement conversion probability space-time distribution diagram; based on the advertisement conversion probability space-time distribution map, deploying an attribution calculation unit for real-time feedback, and generating an incremental attribution map with a confidence interval; and inputting the incremental attribution atlas into a strategy generation adversarial network, and outputting an adversarial optimization advertisement putting strategy set meeting Pareto optimum. According to the embodiment of the invention, the accuracy and real-time performance of advertisement effect evaluation can be improved.

The invention discloses an automatic geological disaster recognition system and method based on multi-source remote sensing data, and particularly relates to the field of geological disaster recognition, and the system comprises a multi-modal remote sensing data acquisition module, a cross-domain physical fusion module, a spatio-temporal evolution decision module, a multi-cascade early warning decision module, an optimization control module and a visualization module. According to the geological disaster automatic identification system and method based on the multi-source remote sensing data, virtual features are generated through a cross-domain physical fusion module by using a domain adversarial network, the model generalization ability during cross-domain application is improved, physical association among the multi-source remote sensing data is deeply mined, and dependence on manual design rules is eliminated; through a three-layer processing chain technology composed of a spatial-temporal feature extraction layer, a dynamic graph evolution layer and a critical recognition layer, the capability of capturing disaster features in a complex geological environment is effectively improved, especially the recognition precision of precursor tiny deformation is improved, and the risk of missing report is reduced.

The invention provides a multi-image forgery detection method and system based on a cross-modal vision large language model, and relates to the technical field of image processing and computer vision, and the method comprises the steps: constructing a data set, and carrying out the preprocessing of the data set; according to the preprocessed data, respectively extracting visual features and language features through a pre-trained visual Transform and a language model so as to obtain cross-modal features; according to the cross-modal features, the visual features and the language features are clustered, cross-modal similarity is calculated, a matching relation is established, fusion is carried out, and fused multi-modal features are obtained; and according to the fused multi-modal features, performing adversarial training through a generator and a discriminator to generate an adversarial network. According to the method, high-precision detection and effective detection of multiple image counterfeiting types such as positioning splicing, copying and pasting, AIGC generation and the like are realized.

The invention relates to the technical field of cultural heritage digital protection, in particular to a super-volume historic building three-dimensional simulation modeling method based on multi-source heterogeneous data, and the method comprises the steps: firstly collecting node multi-source heterogeneous data such as laser point cloud, images, structural mechanical parameters and historical repair records, and then carrying out node feature enhancement through a node feature enhancement module; using an improved generative adversarial network to strengthen node edge features, adopting an adaptive threshold segmentation algorithm to extract surface texture features, converting mechanics and size data into a three-dimensional constraint condition parameter matrix, then using a topological relation verification algorithm, using a graph neural network to traverse and verify a component connection relation, and obtaining a three-dimensional confrontation model; and a re-calibration mechanism is triggered when the deviation exceeds the limit, the weight is adjusted based on a Bayesian optimization algorithm, fusion verification is carried out again, finally, hierarchical grid division is adopted to construct high-precision sub-models, and the sub-models are spliced into an integral three-dimensional model, so that the model precision and reliability are improved, and reliable digital support is provided for ancient building protection.

The invention, which relates to the technical field of payment security, discloses a payment-scene-oriented interaction intention identification and error correction system comprising an input analysis module, an intention simulation module, a dynamic decision module, a biological verification module, an audit evidence storage module, and a cross-scene knowledge migration module. According to the method, multi-modal data such as voice, texts, images and touch tracks are integrated, structured feature vectors are generated through a cross-modal attention network, the problem of incomplete single-modal coverage is solved, cross-modal data consistency verification is achieved based on a unified semantic tag system, and the reliability of input sources is graded by combining equipment fingerprints and geographic positions, so that the reliability of the input sources is improved. A high-risk transaction protection capability is enhanced, a generative adversarial network is utilized to construct a virtual attack sample library, attacks such as tampering with characters similar in shape and AI faking voiceprints are simulated, unknown threats are actively defended through cosine similarity matching, a user historical behavior statistical model is integrated, and known risks such as high-frequency small-amount transfer are passively intercepted. And a closed-loop incremental learning continuous optimization model is supported.

The invention relates to the technical field of image processing, and discloses an image recognition and analysis system based on AI. The system comprises a data acquisition module, a feature extraction module, a model training module, a multi-modal fusion module, a dynamic optimization module and the like. The method comprises the steps of collecting real-time image data by a multi-source sensor, extracting features by a cascade convolutional neural network, generating an adversarial network training model, integrating multi-source data by multi-modal fusion, optimizing feature vectors by an improved genetic algorithm, and constructing a classification decision tree. In addition, an anomaly detection module, a real-time reasoning module, a data enhancement module and a visualization module are further arranged. The system can accurately identify and analyze images, improve the model performance and generalization ability, meet the real-time requirement of edge computing equipment, generate an interpretable report to assist decision making, and have wide application prospects in the fields of security, medical treatment, automatic driving and the like.

The invention provides a bearing fault identification method based on a dynamic generative adversarial network and expert feedback, and relates to the field of bearing fault diagnosis, and the method comprises the steps: generating a high-fidelity fault vibration signal through employing a condition generator and a triple discriminator generative adversarial network; verifying and generating sample quality through a 1D residual verification network and adding the sample quality into a training set; segmenting the vibration signals passing the test by using layered adaptive sampling, and keeping high-frequency impact characteristics in the vibration signals; a dynamic sparse attention mechanism is adopted to reduce unnecessary attention calculation and improve calculation efficiency, and different types of faults are accurately recognized in combination with a hybrid expert system classifier; and detecting the confidence of the diagnosis result, and triggering a feedback mechanism to regenerate a sample to complete autonomous iterative optimization when the confidence is low. According to the method, a generative adversarial network, a fault diagnosis model and a feedback mechanism are fused, accurate diagnosis of bearing faults is achieved through multi-level data enhancement and screening feedback, the diagnosis precision is continuously improved in continuous iteration, and the method is suitable for solving the problem that a traditional method is poor in performance under data scarcity and noise interference. The innovative closed-loop evolutionary logic of generation-diagnosis-feedback is provided, and the robustness and accuracy of fault recognition are remarkably improved.

The invention discloses a dynamic time sequence event equipment system adversarial network construction and recommendation method and device, and relates to the field of killing network design, and the method comprises the steps: constructing an initial detection-command and control-strike warning network; a dynamic time sequence event on the battlefield is continuously monitored, when the event type of the dynamic time sequence event is an equipment state event, the initial detection-command-strike warning network is updated, and when the event type of the dynamic time sequence event is a chained event, a closed detection-command-strike link set containing an enemy target is generated; the closed detection-command-strike link set comprises a plurality of killing chains; based on the multi-dimensional evaluation index system, evaluating each killing chain to obtain an evaluation result corresponding to each killing chain; and according to the evaluation results corresponding to all the killing chains, recommending an optimal interception scheme of an air defense and anti-guide interception action. The method overcomes the problems of large calculation dimension and slow response speed of an existing method.

The invention discloses a bronze ware ornamentation and pattern digital restoration method based on an image enhancement technology, and relates to the technical field of image restoration, and the method comprises the steps: building a space mapping matrix; extracting multi-modal data features by using the surface state of the topological insulator, and registering a joint data volume; forming a super-resolution image through super-resolution reconstruction; obtaining a material degradation coefficient by using a wavelet finite element method and a graph neural network; generating an adversarial network by utilizing physical constraints, and generating an embarrassment repairing result; a material sensing three-dimensional model is constructed by adopting a wavelet packet decomposition and neural radiation field fusion technology, texture mapping is dynamically adjusted based on a graphene Moire effect, and virtual-real fusion is performed through holographic waveguide AR; by combining advanced technologies such as a metamaterial lens, micro-distance laser scanning, a topological insulator film, a graphene heterojunction and a nerve radiation field, high-precision three-dimensional digital restoration and repair of bronze cultural relics are realized, and immersive augmented reality display experience is provided.

The invention discloses an artificial intelligence rice water and fertilizer real-time monitoring method and system, and relates to the technical field of agricultural intelligent decision making, and the method comprises the steps: inputting a farmland feature data set into a soil thermodynamic diagram generation model, carrying out the high-resolution reconstruction of the farmland feature data set through a GAN adversarial network, and generating a whole-field high-precision soil thermodynamic diagram; based on the whole-field high-precision soil thermodynamic diagram, a collision relation between the fertilization amount and historical farming data is detected according to an FCL causal algorithm, a preliminary causal diagram is generated, and a causal diagram structure of the fertilization amount and the historical farming data is constructed by adopting a multiple linear regression method; based on a causal diagram structure, the multi-order causal effect of the fertilization amount, the soil parameters and the historical yield is analyzed through a dynamic allocation algorithm, and a water and fertilizer regulation and control strategy is formulated in combination with a multi-objective optimization algorithm. According to the method, the soil thermodynamic diagram generation model is constructed, so that the fuzzy problem of the edge of the field and the salinization area is solved, a high-fidelity soil space state substrate is provided for water and fertilizer regulation and control, and invalid irrigation is reduced.

The invention belongs to the technical field of optical detection, and particularly relates to a global power wave aberration prediction-based Ritchey-Common detection method and system, and the method comprises the steps: firstly building a three-dimensional optical model through the parameters of a standard spherical mirror, a to-be-detected plane mirror and an interferometer, and combining a preset Ritchey angle and an out-of-focus / astigmatism coefficient, injecting vibration and temperature noise by using Monte Carlo simulation to generate an enhanced training data space; secondly, constructing a phase error adjustment model based on an adversarial network, and decoupling a mapping relation between an adjustment error and wave aberration distortion through a multi-scale attention mechanism and a radial basis function network to realize adaptive correction of sub-aperture phase data; and finally, splicing sub-aperture phases by using a graph neural network and fusing multi-angle full-aperture measurement data to accurately reconstruct the surface shape of the plane mirror to be measured. According to the invention, through confrontation training and dynamic error compensation of physical constraints, the detection efficiency and reliability of the large-aperture optical element are significantly improved.

The invention discloses an optical cable intelligent label full life cycle management method and system, and the method comprises the steps: extracting structured data through an XML analysis engine based on a configuration file, carrying out the semantic analysis of an unstructured text, carrying out the regression quantification of text features through combining symbol quantiles, and constructing a knowledge graph; according to the knowledge graph, dynamically adjusting the character size, the two-dimensional code position and the error-tolerant rate by adopting a generative adversarial network technology according to the optical cable type and the pasting scene, and generating optical cable intelligent labels adaptive to different scenes; collecting a label image, and converting the label image into structured label identification data including an optical cable type, a connection relation and a transmission signal type by adopting a computer vision technology; the identification data and the knowledge graph are compared, consistency is checked, potential abnormity is analyzed, influences are evaluated, and graded early warning is generated; a micro-service architecture system is constructed, data analysis, AI generation, an intelligent recognition engine and knowledge graph service are integrated, and full-life-cycle management is achieved. The intelligent level of full-life-cycle management of the optical cable label is improved.

The invention relates to a multi-modal image threshold segmentation preprocessing method based on a convolutional neural network, and the method comprises the steps: unifying an image into a standard space, carrying out the pixel value mapping, carrying out the resampling, generating high and low frequency sub-bands, carrying out the soft threshold denoising of the high frequency sub-bands, enhancing the contrast of the low frequency sub-bands, and carrying out the fusion; an optimized VGGnet framework is constructed; a noise adversarial network is generated to carry out active learning loop training on a convolutional neural network model; the image is input into the model for prediction; a local entropy and a gradient magnitude are calculated based on a prediction result; optimal segmentation is realized by setting a double-layer matrix of a feature tag-segmentation method; grey matter Dice calculation is carried out on the segmented images, and preprocessing parameters of unqualified images are optimized through a dynamic parameter adjusting module based on a Gaussian process regression model. The segmentation precision and the processing efficiency of the multi-modal image are effectively improved, and the adaptability of the model to a complex image is enhanced.

The invention discloses a seabed target sonar detection method based on an adversarial network, and the method comprises the steps: S1, collecting deep sea sonar original data, and carrying out the preprocessing of the data, and obtaining a standardized sonar tensor; s2, taking the standardized sonar tensor as training input, and training an improved StyleGAN-3 model in combination with an environmental parameter vector; s3, calling the improved StyleGAN-3 model, and outputting a weak target feature tensor; s4, constructing a sonar detection network; s5, outputting an optimal detection parameter set; and S6, loading the optimal detection parameter set to the sonar detection network, performing target judgment on the weak target feature tensor, and outputting a target confidence value, a three-dimensional space positioning result and a target motion track. Compared with a traditional method, the method has the advantages that in a deep sea scene, the positioning error and the target trajectory loss rate are greatly reduced, the method has excellent online tracking and dynamic situation awareness capabilities, and a solid technical support is provided for ocean monitoring, security and emergency command application.

The invention relates to a power system digital twin model generation method and device, computer equipment, a readable storage medium and a program product. The method comprises the steps of obtaining multi-source heterogeneous data of a transformer substation, preprocessing the multi-source heterogeneous data to obtain preprocessed data, performing fuzzification processing based on the preprocessed data, extracting a fuzzy logic rule, constructing a to-be-trained adversarial network model, and training the to-be-trained adversarial network model based on the fuzzy logic rule. And when it is determined that the model verification conditions are met, training is stopped, and the digital twin model of the power system is obtained. In the modeling process, a model combining fuzzy logic and a generative adversarial network is introduced, and through dynamic rule base management and self-adaptive updating, it is ensured that the system can adapt to the complex operation environment of a power system in real time, so that the accuracy of automatic generation of the substation operation order is greatly improved, and the automation degree of the substation operation order is improved. And meanwhile, the rule compliance and the compatibility of the power system standard are ensured.

The invention discloses a landslide risk assessment method based on an extreme rainfall and geology coupling model, and relates to the technical field of geological disasters. Comprising the following steps: S1, constructing a three-dimensional probability density field of a fracture network and a non-Gaussian random field model of a permeability coefficient tensor; s2, setting a physical kernel layer according to the non-Gaussian random field model, setting a data driving layer through space-time Transform coding, and constructing a graph attention network model; s3, generating an adversarial network through physical information, constructing extreme rainfall coupling data, and updating the non-Gaussian permeability coefficient random field model according to the graph attention network model; and S4, acquiring an entropy generation rate according to the mechanical field data, the seepage field data and the temperature field data, and determining a risk level. Physical interpretability grading early warning of landslide risks is realized, and meanwhile, risk space distribution can be visually displayed through a sliding surface probability cloud picture, so that accurate decision support is provided for disaster prevention and control.

The invention belongs to the technical field of data processing, and particularly relates to a hidden ore body evaluation and positioning method based on multi-source data processing. The method mainly aims at the problems of incompleteness and isomerism of multi-source geological data in acquisition, fusion and modeling. Comprising the following steps: acquiring hyperspectral, geochemical and magnetic anomaly multi-source data of an evaluation area; intelligently complementing missing modal data by using a generative adversarial network based on geological constraints and modal outburst to form a complete multi-source data set; an unsupervised clustering algorithm combining geological correlation and entropy weight analysis is adopted to construct high-confidence-coefficient pseudo-label data, and knowledge mining of unlabeled samples is achieved; feature purification and dimension reduction are carried out through multi-modal feature fusion and hierarchical principal component analysis, and key feature vectors representing the existence of the ore body are extracted; and finally realizing space prediction of the concealed ore body by utilizing the classification model. According to the method, a high-quality data basis and a unified processing framework are provided for intelligent recognition of the hidden ore body, and efficient and accurate positioning of the hidden ore body is achieved.

The invention discloses a gravel pile three-dimensional point cloud volume determination method, device and equipment based on a small rail car, and relates to the technical field of bulk material measurement. The method comprises the following steps: preprocessing original point cloud data of a gravel pile, and carrying out regional division on the preprocessed point cloud data by adopting a curvature clustering segmentation algorithm; performing region restoration on the divided point cloud data by adopting an adversarial network, performing spatial division on the restored point cloud data based on an octree spatial topological index to obtain point cloud data of different hierarchies, and determining a topological connection relationship between the different hierarchies to obtain point cloud subsets; carrying out integral projection hierarchical calculation by adopting a convex hull-voxelization mixed strategy, and then carrying out collaborative optimization calibration on the initial volume of the gravel pile by adopting a time sequence point cloud training graph convolutional network to obtain a calibrated volume. The invention aims to quickly and accurately determine the three-dimensional point cloud volume of the gravel pile.

The invention provides a cross-scale nesting and fusion modeling method for a multi-scale geological model, which belongs to the field of geology and comprises four steps of multi-scale data preprocessing, data fusion, cross-scale nesting and fusion and dynamic coupling. According to the method, multi-source heterogeneous data can be integrated, data fusion is realized through a joint probability space, a variational assimilation framework, a fuzzy logic conflict factor and an alternating direction multiplier method, and the stability and precision of the model are improved by using a deep learning auxiliary fusion technology, such as a multi-scale convolutional adversarial network. Meanwhile, model parameters are determined through a volume average upscaling algorithm and random field condition simulation, dynamic adjustment is carried out through set Kalman filtering and localized set transformation Kalman filtering, and consistency, bidirectional feedback and iterative optimization of the parameters between the models are achieved. And finally, multi-scale convergence is realized through a dual grid strategy and restrictive interpolation, and convergence and dynamic balance of the model on the scale from kilometer to micrometer are ensured.

The invention relates to an internal threat detection method, system and device based on behavior analysis and a medium, and the method comprises the steps: obtaining multi-modal data of a user operation environment, carrying out the time sequence alignment of the multi-modal data based on a time window mechanism, generating a multi-modal feature data set, and constructing a dynamic behavior model in combination with the context information of user behaviors. Performing secondary modeling on the dynamic behavior model by using a graph neural network to generate a user behavior graph; generating an adversarial network based on the user behavior graph so as to generate simulated threat behavior data, and dynamically generating a behavior anomaly detection model by using a reinforcement learning method in combination with the simulated threat behavior data and the user behavior graph; performing threat detection according to the user behavior anomaly detection model, identifying an abnormal behavior, and generating a threat detection result; and calculating a trust score according to the threat detection result, and generating a response priority strategy based on the trust score to execute a response operation on the abnormal behavior. The method has the effect of improving the internal threat detection efficiency.

The invention discloses an electric energy metering data anomaly detection method based on a generative adversarial network and multi-task optimization, which solves the problem of data imbalance in an electric energy metering data anomaly detection scene by utilizing a dynamic resampling strategy, and trains the generative adversarial network in combination with electric energy metering characteristics. Pseudo samples consistent with real distribution are generated through adversarial training of a generator and a discriminator, and an abnormal sample set is expanded, so that the detection capability of the model on abnormal data is enhanced. By constructing a multi-task learning framework, sharing a feature extraction module and jointly optimizing an anomaly detection task and a load prediction task, the accuracy of anomaly detection and the precision of load prediction are remarkably improved. The method specifically comprises the following steps: a data preprocessing step, a dynamic resampling step, a generative adversarial network training step, a multi-task joint optimization step, and an anomaly detection and load prediction step.

The invention discloses a generative defense method and system for resisting attacks, relates to the technical field of network security, and aims to solve the problems that an existing defense scheme is high in calculation overhead and poor in real-time performance, static defense is easy to bypass, and robustness and accuracy are difficult to balance. The method comprises the following steps: constructing an adversarial network model by taking a pre-trained target model as a discriminator and a generative model as a generator; constructing a total loss function by combining a defensive loss function and an accuracy loss function, generating a defensive benign sample by adding defensive disturbance into a benign sample, generating a defensive confrontation sample by adding defensive disturbance after generating a confrontation sample based on the benign sample, and inputting the three types of samples into a target model to obtain total loss; and training the generative model to convergence by using a back propagation algorithm to obtain a trained adversarial network model for classification of defense disturbance samples. Defense generation network training is completed in the training stage, only defense disturbance needs to be overlaid in the reasoning stage, the real-time requirement is met, the robustness and accuracy of the model can be balanced, and the method is suitable for various attack scenes.

The invention discloses an image generation method based on a generative adversarial network, and the method comprises the following steps: 1, collecting data; the method comprises the steps that CT, MRI, ultrasound and other medical image data are collected and cover different disease types and stages, then professional doctors mark the data (such as focus positions and disease types), normalization, cutting and enhancement are carried out, and therefore a marked high-quality medical image data set is obtained; 2, model design: designing a generator and a discriminator; 3, defining a loss function to resist loss; step 4, adversarial training; and 5, generating an image. The defects in the prior art are overcome, and high-quality and diversified synthetic medical images can be generated based on the generative adversarial network technology; the image generated by the GAN can accurately simulate the texture features of a real medical image, such as the roughness of a tumor surface, the microstructure of a blood vessel wall and the like, so that a more real visual reference is provided for doctors.

The invention relates to the technical field of cable asset health management and predictive maintenance, and discloses a cable discharge signal blind separation and enhancement processing method based on an adversarial network, and the method comprises the steps: building a multi-modal monitoring data set through collecting mixed signals and environment data in cable operation; blind separation of discharge signals is realized by using the generative adversarial network, and prior information is not needed; identifying the number of potential signal sources through covariance analysis and double-criterion estimation; iterative optimization and signal enhancement are carried out in combination with a graph neural network and variational reasoning; and finally, through multiple cross validation and quality correction, an enhanced signal with high reliability is output. According to the method, the signal processing technology is deeply fused with asset management, risk prediction and operation and maintenance decision, weak discharge signals can be effectively separated and enhanced under the condition of low signal-to-noise ratio, the accuracy and reliability of cable early fault diagnosis are improved, and credible data support is provided for cable asset health state assessment, risk prediction and operation and maintenance decision.

The invention discloses an oil storage tank oil-water interface measurement system and method and a storage medium, and relates to the technical field of oil-water interface measurement, and the system comprises a data acquisition module, a data processing module, a decision support module and a data security module. According to the oil-water interface measuring system and method for the oil storage tank and the storage medium, a three-dimensional sensing network (distributed sensor network) is constructed by adopting a pressure sensor array and phased array ultrasonic scanning technology and combining dielectric property detection of a microwave dielectric constant sensor; the pressure, the dielectric constant, the liquid level and the vibration data in the tank body are obtained, basic data are provided for subsequent processing, the comprehensiveness of data obtaining can be improved through the sensing network, the pressure data are dynamically calibrated through Takagi-Sugeno fuzzy logic, the influence of temperature drift is effectively eliminated, and the accuracy of data obtaining is improved. The ultrasonic signals are subjected to feature enhancement through the generative adversarial network, and the emulsion layer boundary is accurately recognized.

The embodiment of the invention relates to the technical field of safety event deduction, in particular to an industrial production-oriented safety event generative adversarial deduction method, which comprises the following steps of: determining target equipment, and constructing a generative adversarial network-based deduction model adaptive to the target equipment; an equipment fault tree and a personnel behavior decision tree are obtained and coded into a directed acyclic graph as a causal graph constraint of the deduction model; a physical generation engine generates physical parameters under extreme working conditions based on the causal graph constraint, and then a language behavior generation engine generates an operator behavior sequence based on the causal graph constraint; performing data consistency verification by a multi-modal discriminator based on a cross-modal attention mechanism, and under the condition that the verification is passed, approving generation; and performing iterative training on the deduction model, deploying the deduction model after the training is completed, synchronizing the real digital twin model of the target equipment into the deduction model, and obtaining a deduction result of the security event in the future preset time output by the deduction model.

The invention discloses a wafer defect detection method based on a dual twin network, and the method carries out the training and feature extraction of a defect-free wafer image and a defect wafer image on the surface of a wafer through a dual neural network, can obtain more feature information, and improves the wafer intermode detection capability based on deep learning. The method comprises the following specific steps: acquiring an image of a wafer to be detected by a camera, and preprocessing the acquired image; generating a pseudo-defect image on the defect-free wafer image by using an adversarial network, and performing various transformations on original data through data enhancement to generate a new twin network training sample; and reconstructing and repairing the discriminant features to finally obtain a predicted image. A data set of a twin network is manufactured for training, and a trained model is used for detecting wafer defects; and inputting a wafer image to be detected into the twin network, and generating a prediction image by calculating a difference value between an input feature set Fm1 and an output feature set FAE1 of the twin network.

The invention provides a neural rehabilitation action detection method based on a domain generalization neural network, and the method comprises the steps: obtaining the neural rehabilitation action video frame data and descriptive text data of a patient in real time, inputting the preprocessed standard input data into a neural rehabilitation action detection model obtained through training, and outputting a motion quality evaluation result score; the newly designed neural rehabilitation action detection model comprises a feature extraction network and a multi-mode neural rehabilitation action detection network. The feature extraction network is used for extracting spatio-temporal features in a video and combining semantic information in text description to form efficient multi-modal feature representation, and the multi-modal neural rehabilitation action detection network comprehensively integrates key information of the video and the text and outputs a motion quality evaluation result score; meanwhile, a domain generalization generative adversarial network training strategy is designed for model training; the problem of distribution difference between the source domain data and the target domain data is effectively solved, and the domain generalization ability of the model under different training scenes and equipment conditions is remarkably improved.