1066 results about "Reconstruction error" patented technology

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of financial science and technology, medical health and the like, and discloses an intelligent anomaly recognition and intervention processing method, device, equipment and medium. The method comprises the following steps: carrying out feature fusion by using a gating fusion network and generating a preliminary abnormal score, determining a reconstruction error through an auto-encoder and triggering abnormal early warning, calculating a causal effect value of key features in combination with a causal graph model and anti-factual reasoning, and calibrating the abnormal score to generate a final abnormal score and an intervention instruction. And executing an intervention action and recording a result. According to the method, the multi-dimensional feature information and the causal reasoning mechanism are fused, the self-encoder reconstruction error is combined to carry out anomaly judgment, the intervention instruction is generated and executed, closed-loop control of anomaly detection, reasoning analysis and intervention execution is achieved, and the recognition accuracy of complex events and the system response capacity are improved.

The invention discloses an adaptive mask medical image segmentation method based on a self-supervised mask and deep reinforcement learning, and the method comprises the steps: employing a classic encoder-decoder architecture for a self-supervised mask reconstruction network, fusing a Swin Transform encoder, and carrying out the feature fusion of local image blocks through a self-attention mechanism; according to the self-adaptive mask model, a PPO deep reinforcement learning algorithm is adopted, a strategy network and a value network are constructed, mask actions are dynamically regulated and controlled, reconstruction errors are gradually reduced, a mask strategy is continuously optimized in multiple times of strategy updating for self-adaptive optimization, and high-quality reconstruction of a medical image influenced by missing information is achieved; according to the method, high-quality feature representation can be obtained in an unlabeled data environment, and relatively high precision and accuracy are presented on a public data set.

The invention discloses an industrial image anomaly detection method based on deep learning, and particularly relates to the technical field of industrial visual detection. The problems of high false alarm rate, fuzzy fine defect positioning, insufficient real-time response capability, difficulty in model increment updating and the like caused by data distribution drift in an industrial scene are solved. According to the method, robust features are extracted through a multi-scale feature fusion auto-encoder, and a dynamic memory bank is constructed to update a normal sample prototype online; a dual-path detection mechanism is adopted to cooperate with a pixel-level reconstruction error and attention weighted feature matching deviation; efficient edge reasoning is realized in combination with block parallel processing and model compiling optimization; and designing an elastic incremental learning framework to prevent disastrous forgetting. And finally, false alarms caused by environmental changes are reduced, accurate positioning of pixel-level defects is realized, millisecond-level detection requirements of high-resolution images are met, safe and efficient model online evolution is supported, and adaptability and reliability of an industrial quality inspection system are comprehensively improved.

The invention relates to an industrial Internet of Things anomaly detection method based on time sequence and text joint modeling, and belongs to the technical field of industrial Internet of Things anomaly detection. The method comprises the following steps: constructing text prompt information based on collected industrial Internet of Things time sequence data, and respectively taking the text prompt information as inputs of a time sequence channel and a text prompt channel; a sensor association graph is constructed by using a multi-hop GCN, and on the basis of the association graph, time feature modeling from local to global is completed by using multi-scale expansion convolution and combining a differential attention mechanism; performing word segmentation processing on the text prompt information through a word segmentation device, and encoding the text prompt information into vector representation; and calculating attention weight between time sequence embedding and text prompt embedding, fusing to obtain joint embedding representation, enhancing the joint embedding representation, inputting the enhanced joint embedding representation into MLP for reconstruction, calculating an abnormal score through a reconstruction error, and carrying out industrial Internet of Things anomaly detection according to the abnormal score. The method is high in anomaly detection accuracy, and can improve the equipment anomaly perception and risk early warning capability.

The invention discloses a battery abnormal data prediction and state evaluation method, and relates to the technical field of battery state prediction, and the method mainly comprises the steps: carrying out the preprocessing of an experiment data set, and obtaining multi-dimensional time series data; a combined feature encoder, a pre-response encoder and a memory analysis module are constructed to realize a battery abnormal data fault prediction model; training the model by using the multi-dimensional time sequence data to obtain a trained model, and predicting the to-be-predicted data to obtain a prediction result; and calculating a reconstruction error between a prediction result and original data, constructing an AUROC evaluation model, and evaluating the battery abnormal data fault prediction model. By implementing the battery abnormal data prediction and state evaluation method provided by the invention, the feature extraction efficiency, the abnormal recognition precision, the detection stability and the generalization ability can be improved.

The invention relates to a wind power equipment blade fault detection technology, discloses an unsupervised wind power equipment blade fault detection method based on a phase perception parallel attention mechanism, and solves the problems that an existing wind power equipment blade fault detection method is high in dependence on labeled data, insufficient in generalization ability under strong noise and variable working conditions and high in fault detection efficiency. And a weak transient fault signal and a dynamic change characteristic are difficult to capture robustly. According to the scheme of the invention, the method comprises the steps: collecting a blade operation audio signal, and extracting a dual-channel time-frequency feature containing an amplitude spectrum and a phase spectrum through improved short-time Fourier transform; a deep adversarial auto-encoder is constructed by using an encoder containing a phase perception parallel attention module, a decoder and an auxiliary encoder, and normal working condition feature distribution is learned by reconstructing an error loss, potential representation consistency loss, adversarial loss and phase consistency loss optimization model during off-line training; in the reasoning stage, the fault is judged based on the feature distance score and the reconstruction error score.

The invention discloses a single-person abnormal behavior identification method and system based on multi-modal skeleton feature fusion, and the method comprises the steps: S1, collecting continuous RGB images and infrared thermal imaging images in a monitoring video, carrying out the human body detection and key point estimation of visible light and infrared images through employing a multi-modal fusion model of YOLOv12 in combination with Transform, and constructing a single-person posture time series data set; s2, key point speed vectors are calculated for the continuous skeleton frame sequence of each target person, skeleton key point information and speed information are fused, and an action feature sequence is formed; s3, inputting the motion feature sequence into an MPED-RNN model, decomposing skeleton motion into a global displacement component and a local attitude deformation component, and performing joint coding, decoding and prediction through a dual-channel GRU network; and S4, calculating a prediction error and a reconstruction error according to a reconstruction result and a future skeleton key point prediction result, evaluating whether the current behavior deviates from a normal trajectory, and judging whether the current behavior is in an abnormal state. According to the invention, real-time identification of abnormal behaviors of a single person in a complex scene is realized.

The invention provides a hydropower station AI supervision system and method based on a multi-modal large model, and relates to the technical field of intelligent hydropower. The system comprises a multi-modal data acquisition module, a cross-modal space-time alignment module, a multi-modal feature extraction module, a multi-modal large model processing module and an intelligent reasoning and decision module. A neural differential equation model is introduced to carry out space-time alignment on asynchronous sensing data, networks such as Vision Transformer, MelCNN, TCN and the like are utilized to extract multi-modal features, cross-modal fusion analysis is realized by combining a local and global attention mechanism and dynamic weight distribution, and equipment abnormality is further reasoned based on a reconstruction error, a mahalanobis distance and a knowledge graph and a maintenance strategy is generated. According to the method, high-precision anomaly detection, fault root cause positioning and dynamic maintenance optimization of key equipment of the hydropower station are realized, diagnosis errors caused by traditional manual inspection and data splitting are avoided, and the operation and maintenance intelligence level and the equipment operation reliability are improved.

The invention provides an offshore wind turbine generator fault diagnosis method and system based on multi-source sensor data fusion. The method comprises the steps that a vibration signal from at least one component of a wind turbine generator is acquired through a vibration sensor; performing time-frequency conversion on the vibration signal by applying synchronous compression wavelet transform to obtain time-frequency representation of the vibration signal; when the reconstruction error exceeds a preset threshold value, it is judged that an abnormal event exists in the vibration signal; obtaining the position of a part corresponding to the abnormal event; starting an image sensor and an acoustic sensor according to the position of the component, and acquiring an image signal and a sound signal of the component according to the image sensor and the acoustic sensor; according to the DS evidence theory, the vibration signal, the image signal and the sound signal, obtaining the confidence of the fault type; the fault type of the component is judged according to the maximum confidence allocation principle, high-resolution time-frequency analysis can be achieved through synchronous compression wavelet transform (SST), and the fault feature identification degree is improved in combination with the self-encoding neural network and the D-S evidence theory.

The invention discloses an industrial image change anomaly detection method and system based on artificial intelligence, and relates to the technical field of image recognition, and the method comprises the steps: collecting a dual-light-source industrial image, employing frequency domain saliency to guide fusion, and carrying out visual enhancement processing through color mapping and edge enhancement; inputting the enhanced image into a CNN convolutional network to generate a multi-scale feature map, extracting a multi-scale high-pass residual image through two-dimensional fast Fourier transform and a high-pass filtering template, and splicing all scales and coding to generate a token sequence through local attention guide fusion; constructing a self-induction visual model, and performing feature reconstruction on the token sequence to generate a reconstructed feature map; and calculating and reconstructing an error scoring graph by adopting double error indexes, sampling to obtain an abnormal smooth graph, and segmenting an abnormal region based on the abnormal smooth graph. And finally, a dual anomaly detection system of image-level judgment and region-level identification is constructed.

The invention discloses an intelligent automobile interpretable abnormity diagnosis method and system, and relates to the technical field of intelligent traffic. The method comprises the steps of collecting multi-dimensional sensor data based on an intelligent automobile test platform, and constructing a directed causal graph and a causal adjacency matrix which are used for describing a causal relationship between the sensor data; designing a causal constrained graph attention mechanism based on the causal adjacency matrix, and constructing a causal constraint enhanced graph attention anomaly diagnosis model; and based on the directed causal graph and the graph attention anomaly diagnosis model, constructing a hierarchical anomaly diagnosis strategy integrating a feature reconstruction error, a variable causal relationship and a graph attention network weight, positioning an anomaly root cause and identifying a propagation path of the anomaly in the sensor network. According to the invention, the problems of false correlation and lack of exception explanation ability of graph attention network learning in the prior art can be overcome, and reliable exception detection and root cause diagnosis of intelligent automobile multi-sensor data are realized.

The invention discloses an electronic component early fault detection method based on infrared thermal imaging, and the method comprises the following steps: collecting an infrared thermal imaging image, synchronously collecting current and voltage data, and obtaining an infrared thermal imaging sequence; distortion correction, noise suppression and radiance correction are carried out on the sequence, and metal reflection false high-temperature points are removed; executing time sequence difference to extract a continuous heating area, and generating a time sequence temperature feature vector; inputting the feature vector into an improved Times Net model, and outputting time sequence feature representation; calculating a reconstruction error, a prediction residual error and a frequency spectrum deviation, and fusing into an abnormal score; comparing the abnormal score with a dynamic threshold value, and outputting an electronic component identifier, a potential fault type and an abnormal type; and an early fault trend report is generated by combining historical data comparison. The method realizes early detection and trend prediction of electronic component faults, and is suitable for state monitoring and reliability guarantee of a complex circuit system.

The invention discloses a dynamic scene incremental reconstruction and rendering method based on 3DGS, and belongs to the field of specific computer models, and the method comprises the steps: constructing a 3DGS model at an initial moment based on an original image set; obtaining any visual angle image at the moment t in the dynamic scene, and determining a first updating area through semantic segmentation and target recognition; generating an increment updating region based on the luminosity error, the local similarity and the global semantic feature; generating a second update region, modeling in the second update region, and minimizing region reconstruction errors to generate an optimized Gaussian point set; and fusing and optimizing the real-time model at the previous moment to obtain a real-time model Gt at the moment t for real-time rendering. According to the method, geometric and texture information of a scene can be effectively coded, accurate detection and local increment updating of a dynamic region are supported, Gaussian point parameters are optimized to improve the continuity and visual quality of a model, low-delay real-time rendering is realized, and the efficiency and quality of three-dimensional scene processing in a dynamic environment are effectively improved.

The invention discloses a space-time sequence interpolation method and device oriented to heterogeneous deletion, and belongs to the technical field of space-time data processing. Aiming at random or continuous loss of the sensor network caused by faults and communication interruption, the method comprises the following steps: setting static space experts, dynamic space experts, short-term experts and long-term period experts in parallel in the same framework, and respectively capturing fixed geographical adjacency, time-varying space correlation, local continuous trend and long-period rules; spatial features are extracted through high-order diffusion diagram convolution and bidirectional gating circulation, time features are extracted through multi-layer space-time attention, a memory attention gating network is introduced to dynamically weight and fuse output of experts according to reconstruction errors, and node-level and time-step-level self-adaptive interpolation is achieved. Experiments show that compared with the prior art, the method has the advantages that under various real data sets and heterogeneous missing scenes, the precision and robustness are remarkably improved, and the method can be widely applied to scenes needing high-integrity spatio-temporal data, such as intelligent transportation, air quality monitoring and energy internet of things.

The invention discloses an efficient storage and transmission method for monitoring data of a low-voltage power distribution network based on lossy / lossless hybrid compression, and relates to the technical field of data storage and transmission, comprising the following steps: completing data denoising correction at an edge node, and setting a plurality of compression strategies and layering mechanisms; judging whether the data is abnormal or not based on the mahalanobis distance, and selecting a proper compression mode through reconstruction error and bandwidth adaptation; the compressed data is subjected to importance labeling and FEC optimization and then sent to a receiving end, the receiving end evaluates the decoding quality and the packet loss rate, and finally a feedback result is used for online updating of the auto-encoder. According to the method, the differential FEC redundancy rate is allocated, so that the lossless fidelity of a key fault waveform and the high compression ratio of a common periodic signal are considered; and meanwhile, closed-loop self-adaption of compression discrimination, coding and model optimization is realized by utilizing online updating of the variable auto-encoder, so that the storage and transmission efficiency is improved, and the reliability and the real-time performance of the system in sudden failure and network fluctuation scenes are enhanced.

The invention relates to the technical field of image processing, in particular to a picking ingot coating defect online detection method and system based on machine vision, and the method comprises the steps: obtaining a multi-view image, carrying out the feature extraction of an image data stream after preprocessing, and carrying out the matching or clustering with a preset defect dictionary; the method comprises the following steps: preliminarily identifying a potential defect area, triggering refined multi-angle image acquisition, applying a multi-task CNN model fused with a polarization perception convolution kernel, jointly optimizing pixel-level segmentation loss and boundary prediction loss, outputting pixel-level semantic segmentation and an accurate boundary of the defect area, and calculating a reconstruction error to carry out defect classification. And finally, the system also has the functions of defect root cause analysis and process flow adjustment, so that defect tracing and production optimization are realized. According to the system, advanced visual technology, deep learning and multi-modal information fusion are integrated, and online, accurate and automatic detection of picking ingot coating defects is realized.

The invention relates to the technical field of artificial intelligence, and discloses an account bill abnormal transaction identification and analysis method based on deep learning. The method comprises the following steps: fusing user historical account bill time series data and a static portrait, and constructing a multi-modal feature vector; using a converter-based conditional variation auto-encoder to learn personalized normal behavior distribution; calculating an abnormal score through the reconstruction error; dynamically generating a judgment threshold in combination with an extreme value theory to identify an abnormal transaction; and carrying out feature attribution analysis on the abnormity. The method does not need abnormal sample training, and has the advantages of strong generalization ability, low false alarm rate and high interpretability.

The invention relates to the technical field of traffic detection, in particular to a traffic abnormal event cooperative detection method and system based on vehicle-road cooperation, and the method comprises the steps: collecting vehicle end data and road end data from a vehicle end and a road end respectively, and carrying out the timestamp alignment, coordinate transformation, noise filtering and missing value supplementation of the vehicle end data and the road end data; establishing target state estimation based on a state space motion model, performing recursive estimation on a target, and identifying abnormal candidates based on observation residual errors; constructing a space-time diagram based on the vehicle end data and the road end data, reconstructing node features by adopting a time sequence diagram neural network, generating an anomaly score according to a reconstruction error, and outputting an anomaly candidate; and according to the state space motion model and the anomaly candidates of the time sequence diagram neural network, confidence fusion is carried out according to confidence, and whether an anomaly alarm is triggered and whether an anomaly type and positioning information are output are judged based on a fusion result. And through confidence fusion, false alarms triggered by isolated noise can be effectively suppressed.

The invention discloses a blocked pipe section identification method of a drainage pipe network system based on dynamic characteristics, and relates to the technical field of drainage pipe network monitoring. Real-time hydraulic association between pipe network nodes is quantified through a dynamic adjacency matrix to generate a spatial topology matrix, and a time sequence is divided based on a sliding window; a space-time fusion model based on GCN and Transform is constructed to carry out multi-scale dynamic coding, and then a decoder is utilized to reconstruct normal working condition data of a pipe network. By calculating the deviation degree of the pipe section level reconstruction error and the threshold value, the blocked pipe section is accurately recognized, end-to-end modeling from'pipe network topology-drainage time sequence data-external rainfall 'multi-source data to blocked pipe section recognition is achieved, and the technical difficulties of a traditional method in the aspects of dynamic topology modeling, long time sequence dependence capture and multi-modal feature fusion are solved.

The invention provides a multi-modal neural causal inference micro-service fault positioning method and system, and the method comprises the steps: accessing observability data in a service operation process, and representing the tracking information of each request as a directed acyclic graph of a multi-modal feature; performing multi-modal feature coding and graph self-coding anomaly detection on the calling graph, and identifying an abnormal node through a reconstruction error; based on service topology prior, learning a sparse causal relationship graph between services by adopting a multi-scale neural causal inference method; calculating a node root cause score according to the causal relationship graph and the abnormal score, and executing causal path search to generate a fault propagation path; and marking the potential root cause according to the path weight of the propagation graph and the node popularity, and outputting a visual diagnosis result. According to the method, the system operation state is comprehensively described by fusing three kinds of micro-service system multi-modal data of logs, indexes and Trace in the micro-service system, and the structure-perceived causal diagram is constructed, so that accurate and explainable root cause positioning is realized.

The invention belongs to the technical field of intelligent early warning, and particularly relates to an intelligent early warning method for a DMF waste liquid purification and recovery control platform, and the method comprises the steps: carrying out the principal component analysis of long-period historical data, and constructing a principal component transformation matrix of a static reference model; for a moment to be diagnosed, calculating a reconstruction value by using the static reference model to obtain a residual vector, carrying out eigenvalue decomposition on a covariance matrix of a residual matrix of a sliding time window, calculating a drift coherence index according to the distribution of drift eigenvalues, modulating a drift principal component vector of the sliding time window by combining the residual vector, and carrying out diagnosis on the moment to be diagnosed; and obtaining a drift compensation vector, superposing the drift compensation vector with a reconstruction value of a real-time data vector at a to-be-diagnosed moment to obtain an adaptive reconstruction value at the to-be-diagnosed moment, calculating a reconstruction error, comparing the reconstruction error with a fault alarm threshold, judging whether a fault exists at the to-be-diagnosed moment, and performing early warning. According to the invention, the early warning accuracy and robustness are improved.

The invention relates to the technical field of intelligent diagnosis, and discloses a micro-grid fault diagnosis method and system based on data driving and unsupervised learning. The method comprises the following steps: collecting current, voltage, temperature and power data of a micro-grid and constructing a time sequence matrix; inputting a time sequence prediction network and a time sequence reconstruction network, and performing parallel processing to obtain a prediction error and a reconstruction error; carrying out weighted fusion on the two errors and constructing a two-dimensional error space to judge normal fluctuation and fault abnormity; and extracting a state variable to generate a dynamic threshold to judge a fault. The false alarm rate and the missing report rate of fault diagnosis are reduced.

The invention discloses an equipment health degree dynamic assessment method and system based on an auto-encoder, and relates to the technical field of health assessment, and the method comprises the steps: reading operation time sequence data; data features are extracted after cleaning and screening; taking the data features as training data, performing dynamic weight constraint learning on the LSTM auto-encoder, and constructing a prediction model; receiving real-time data by using the model, calculating a reconstruction error, and establishing a first score; configuring a sequence prediction branch, identifying gradual deterioration through short-term prediction error alignment, and establishing a second score; mutational points are identified according to reconstruction errors, and a third score is established in combination with multi-scale sliding window analysis; and integrating the three grades to realize dynamic evaluation of the equipment health degree. The technical problems of high cost, poor adaptability and insufficient early warning capability caused by dependence on an intrusive sensor, a static rule and the like in the prior art are solved, and the technical effect of improving the accuracy and foresight of equipment health assessment is achieved.

The invention provides a VAE-ALSTM-based bridge structure state anomaly detection method and system, and the method comprises the steps: carrying out the standardization and sliding window segmentation of vibration monitoring data, extracting potential features through a variational auto-encoder, and carrying out the time sequence prediction and reconstruction in combination with an attention-enhanced long-short-term memory network, thereby achieving the detection of the abnormal state of a bridge structure. Fusing the prediction error and the reconstruction error to generate an abnormal score; a threshold value is automatically set through quartile distance statistics, and self-adaptive judgment under different bridge types is achieved; when the score crosses the boundary, real-time alarm is triggered; the system correspondingly comprises a data preprocessing and sample construction module, a VAE-ALSTM model construction module, a training stage prediction and reconstruction module, an error calculation and anomaly scoring module, a threshold setting module and an anomaly judgment and alarm module. The method and the system do not need manual threshold parameter adjustment, are high in precision and good in real-time performance, and can be widely applied to the fields of bridge health monitoring, operation and maintenance early warning and the like.

The invention belongs to the technical field of artificial intelligence, and relates to a distributed photovoltaic system abnormity monitoring and intelligent early warning method. Key parameters of the distributed photovoltaic system are collected at high frequency, noise filtering and signal normalization processing are carried out on data, then multi-domain feature extraction is carried out, anomaly detection is carried out through auto-encoder reconstruction error analysis, convolutional neural network feature classification and a single-class support vector machine, and an anomaly detection result is output. Establishing a multi-dimensional anomaly evaluation model, and synthesizing obtained evaluation results of all dimensions to obtain a final evaluation result; the method comprises the following steps: carrying out hierarchical classification to identify specific anomaly types, carrying out diagnosis of different types of anomalies, identification of typical anomalies and evaluation of fault severity, outputting the specific anomaly types and severity, and providing detailed diagnosis information for early warning decision making. The system state is comprehensively evaluated, corresponding measures are taken according to the grading early warning mechanism and the abnormity severity degree, and complete distributed photovoltaic system abnormity monitoring and intelligent early warning are formed.

The invention relates to the technical field of model transfer learning, and discloses a model adaptive optimization method based on transfer learning. The method comprises the steps that source domain model structure parameters and target domain task initial data distribution are obtained, the feature mapping relation of all levels of a source domain model is extracted, and a cross-domain feature migration reference topological framework is generated; dividing a migratable feature layer and a to-be-reconstructed feature layer according to a target domain data distribution difference, and dynamically adjusting a migration priority in combination with a sample distribution density; freezing and unfreezing the transferable feature layer layer by layer based on the priority, synchronously constructing a local feature reconstructor, and optimizing domain offset through iterative feature alignment; collecting a feature reconstruction error and a migration feature retention degree in each iteration, and calculating a dynamic balance coefficient to adjust a freezing proportion and reconstruction intensity; and fusing the two types of features through a global model integrator, and generating mixed feature representation to drive end-to-end training of a target domain task.

The invention discloses a reinforcement learning-based spring steel wire drawing control method, which comprises the following steps of: acquiring real-time process parameters to form a process state data sequence; inputting the process state data sequence into the state space model, and constructing a virtual working condition sample; based on the virtual working condition sample, pre-training a reinforcement learning controller and outputting an initial control strategy; inputting the initial control strategy into a lower-layer strategy network, and outputting a wire drawing speed adjusting instruction; the current process state and the wire drawing speed adjusting instruction serve as synchronous input, and an implicit context vector is generated; extracting control strategy characteristics in the edge controller, performing compressed encoding and forming control strategy representation; uploading the control strategy representation to a cloud server, and outputting a unified global control strategy model; and carrying out anomaly detection on the current process state, and if a detected state reconstruction error exceeds an anomaly judgment threshold, triggering safety control logic. The spring steel wire drawing control device realizes spring steel wire drawing control.

The invention discloses an abnormal radio signal monitoring method and system based on artificial intelligence. The method comprises the following steps: S1, acquiring a radio signal in real time and preprocessing the radio signal; s2, noise is added to the radio signal through a diffusion model, and a noise-added radio signal is generated; s3, de-noising the radio signal after noise addition, and recovering the de-noised radio signal; s4, calculating an original reconstruction error, setting a dynamic threshold value, and judging whether the radio signal is abnormal or not; s5, classifying the detected abnormal radio signals; s6, when an abnormal radio signal is detected, triggering real-time alarm and performing root cause analysis; and S7, performing incremental updating on the diffusion model, and adaptively adjusting a dynamic threshold. According to the method, the diffusion model and the de-noising network technology are combined, intelligent monitoring and recovery of abnormal radio signals are achieved, the method has the advantages of being high in precision and robustness, and the anomaly detection accuracy of the radio signals in the complex environment is remarkably improved.

The invention discloses a retina image unsupervised anomaly detection method for early screening of diabetes mellitus. The method comprises the following steps: carrying out registration and multi-scale attention-guided blood vessel segmentation on a longitudinal time sequence retina image of a patient; extracting a vascular skeleton and constructing a time sequence vascular topological graph, calculating geometric morphology and hemodynamic attributes of each vascular segment, identifying vascular morphology evolution characteristics by comparing topological graphs of adjacent time points, and calculating hemodynamic characteristics such as wall shear stress through simulation; the evolution and hemodynamic characteristics are jointly input into a time sequence encoder for unsupervised learning, and an early diabetes risk score is comprehensively generated by analyzing a reconstruction error, an abnormal score based on density estimation and a time sequence trajectory deviation degree of a potential space; the scheme of the invention does not depend on lesion labels, and can sensitively detect the tiny anomalies at the early stage of pathology from multi-dimensional dynamic changes, thereby providing an objective and quantitative new way for early screening and intervention of diabetes.

The invention belongs to the technical field of brain-computer interfaces, multi-modal feature alignment and information retrieval, and discloses a visual content retrieval method based on electroencephalogram signals. Cross-subject standardized electroencephalogram samples are obtained and input into an electroencephalogram encoder to extract low-dimensional electroencephalogram signal feature vectors; an image encoder is adopted to process the corresponding retrieval images to extract visual feature vectors; the low-dimensional electroencephalogram signal feature vector and the visual feature vector are jointly input into a prototype attention enhancement module to form a dynamic prototype pool, and prototype enhanced electroencephalogram signal representation is obtained through processing; the method comprises the following steps: performing classification training according to existing prototype enhanced electroencephalogram signal representation and image data pairs; for a plurality of time slices of the to-be-queried electroencephalogram signal sample, calculating a reconstruction error or a signal-to-noise ratio of each time slice to obtain a confidence coefficient; and adopting a weighted average or voting mechanism to fuse a plurality of time slice results, and outputting a stable and robust final visual retrieval classification result.