1701 results about "Data detection" patented technology

The invention relates to the field of data detection, in particular to a network traffic anomaly detection method based on aggregated mimicry distillation. According to the method, protocol level analysis is carried out on traffic through a multi-branch feature extraction network, and a unified representation vector is generated through a cross-layer fusion mechanism; calculating a first abnormal score based on a protocol perception weighted confrontation soft contrast mechanism; a heterogeneous teacher model is constructed and dynamically weighted and aggregated, and the student model generates a second abnormal score through knowledge distillation learning; forming a heterogeneous redundant detection pool by the student model, the teacher model and the rule detector, dynamically selecting the detector and applying adaptive disturbance to obtain a third abnormal score; and dynamically fusing the three types of abnormal scores to output a detection result. According to the method, the problems of protocol semantic segmentation, weak boundary sample discrimination, knowledge migration simplification and defense path predictability are effectively solved, and the detection accuracy, robustness and dynamic defense capability are remarkably improved.

Techniques for determining a presence of an object, especially an object such as animal or debris, in a path of a vehicle, are discussed herein. For example, sensors of various modalities, which may include multispectral sensors, may capture data representing an environment the vehicle is traversing. In examples, one or more trained machine learned (ML) models, operating on a vehicle computing system, may detect and / or classify objects in the environment, based on input data of one or more modalities or spectral bands. The ML models may be pre-trained using training data including real sensor data, synthetic data, and / or augmented data, along with auto-generated annotations. In some examples, hyperspectral data may be used to identify materials associated with detected objects. A confidence score associated with the detection of the object may also be computed. The vehicle may be controlled based on detection of the object and its classification.

The invention discloses a multi-modal abnormal data detection and restoration method and system oriented to a power business scene. The method comprises the following steps: collecting multi-source heterogeneous power data, abstracting a power system into a weighted undirected graph, uniformly mapping the multi-source heterogeneous data into a graph signal, and preprocessing the collected data; extracting spatial features of nodes in a topological structure by adopting a graph convolutional network, and capturing a time dependency relationship in combination with a time sequence encoder; identifying various types of data abnormal points through an abnormal scoring function fusing the time sequence prediction error and the neighborhood consistency; a prediction-reconstruction combined repair strategy is adopted, time sequence prediction and neighborhood diffusion estimation are fused, and a preliminary repair value is generated; a lightweight parameter adapter is introduced, a scene feature vector is used as input, a repair weight and a regularization coefficient are dynamically generated, and a repair strategy is automatically adjusted; and performing physical consistency verification on a data result, wherein the physical consistency verification comprises power injection conservation constraint, voltage amplitude range constraint and time sequence continuity constraint.

The invention discloses an optical surface defect data detection method based on deep learning, and relates to the technical field of optical defect detection, and the method comprises the following steps: constructing an optical scattering physical model, inputting a collected optical surface image into the optical scattering physical model for multi-modal data synthesis, and generating multi-modal image data; constructing a deep learning feature extraction network, inputting multi-modal image data, and performing multi-scale feature fusion and enhancement through a bidirectional attention feedback mechanism to generate a deep feature map; performing spatial domain analysis on the depth feature map by using a deep learning region generation method, positioning coordinates of potential defect regions, and generating a candidate defect region coordinate set; through multi-modal data synthesis driven by an optical scattering physical model, the limitation of a single imaging mode is broken through, the scattering characteristics of defects under multi-physics field coupling are dynamically analyzed, the recognizable degree of weak defects in a complex scattering environment is enhanced, and the problem of defect missing detection is solved.

Techniques and apparatuses for enhanced text suggestions are described. An example method includes detecting a user gesture performed by a user of the computing system based on data from one or more neuromuscular sensors and identifying a set of text characters corresponding to the user gesture. The method further includes causing display of the set of text terms in a user interface and determining whether a cognitive load of the user meets one or more criteria. The method also includes providing a text suggestion to the user based on the set of text characters in accordance with a determination that the cognitive load of the user meets the one or more criteria, and forgoing providing the text suggestion to the user based on the set of text characters, in accordance with a determination that the cognitive load of the user does not meet the one or more criteria.

The present invention relates to the technical field of intelligent out-of-distribution fault detection for construction machinery, and discloses an out-of-distribution fault detection method and system based on energy propagation and graph learning, and the method includes: acquiring vibration acceleration signals in typical fault states, carrying out similarity calculation to obtain an adjacency matrix composed of the maximum mutual information coefficients, and taking the adjacency matrix as input in a graph neural network; carrying out feature extraction on the adjacency matrix through adopting a GraphSage graph convolution method, and generating each node representation; calculating an energy score of each node, and distinguishing between in-distribution data and out-of-distribution data; and enhancing out-of-distribution data confidence estimation for each node, and carrying out out-of-distribution data identification and out-of-distribution data detection under different working conditions of a rolling bearing.

The invention belongs to the technical field of power network inspection, and particularly relates to a power network security inspection system which comprises a multi-source sensing module, an edge calculation module, a zero-trust security verification module, a core analysis module, an execution module and a feedback optimization module. Through an edge-cloud collaborative architecture, sensors, network flow and video data are fused in real time, the detection efficiency is improved, a physical equipment state and zero-trust dynamic authority control are introduced, 'physical-digital 'dual verification is realized, real threats and false alarms are marked, algorithm parameter adjustment is guided, and a feedback optimization module continuously reduces the false alarm rate through a closed loop mechanism; full-dimension safety protection of the power network is realized through modular design, dynamic branch processing and a closed-loop optimization mechanism; the threat level and the risk assessment result are dynamically adjusted through the environmental risk correction factor, the risk misjudgment rate is reduced, and the fault prediction accuracy in extreme weather is improved.

The invention relates to the technical field of image recognition, and discloses a crop lodging image data detection method and system, and the method comprises the steps: carrying out the self-adaptive illumination correction processing of the original image data of crop lodging, and obtaining an illumination compensation image of the original image data; performing multi-scale super-pixel segmentation on the illumination compensation image to obtain a crop area image of the illumination compensation image; extracting texture features and shape features in the crop region image, and performing feature fusion on the texture features and the shape features to obtain a multi-modal fusion feature set of the crop region image; performing multi-dimensional feature collaborative analysis on the multi-modal fusion feature set to obtain an initial discrimination result of the multi-modal fusion feature set; performing confidence coefficient optimization on the initial judgment result in combination with a spatial context relationship to obtain a target lodging detection result of the crop region image; according to the invention, the efficiency of crop lodging image data detection can be improved.

The invention discloses a semiconductor data detection system and method based on artificial intelligence, and relates to the technical field of semiconductor detection.The method comprises the steps that parameterized revolving door operation is executed through a quantum circuit, a reformed group equation is constructed to generate a physical loss item, and meanwhile a generative adversarial network is constructed through a quantum discriminator and a quantum generator; generating an enhanced data set; constructing a physical topology adjacency matrix according to the enhanced data set, loading a source domain convolutional neural network model and carrying out feature transformation, and generating a physical constraint embedded analysis engine through a dynamic physical graph attention mechanism and lattice differential homeomorphic mapping; deploying a physical constraint embedded analysis engine, processing real-time wafer data, generating a lattice defect tensor and a time-space confidence field, obtaining a confidence score, triggering a process adjustment instruction when the confidence score meets a high confidence condition, and otherwise, storing data which does not meet the condition into a feedback queue; according to the invention, deep coupling of a quantum generation process and a semiconductor lattice physical rule is realized.

The invention provides an end side AI agent abnormal data detection method and device based on edge computing, and relates to the technical field of data processing.The method comprises the steps that real-time collection is conducted on an end side through an AI agent, and a plurality of time sequence data streams are obtained; a time sequence-frequency spectrum self-adaptive joint feature extraction module arranged in the AI intelligent body is used for dynamically analyzing the multiple time sequence data streams, edge calculation is used for carrying out anomaly analysis and recognition on the joint feature vector, an initial early warning signal is generated, inversion is carried out, an inversion result is generated for cyclic verification, and the result is fed back to the AI intelligent body for updating and optimization. And constructing an abnormal data detection report. The technical problem that in the prior art, due to the fact that abnormal data detection is low in reliability and low in response speed, real-time abnormal data is difficult to efficiently and accurately recognize is solved. The technical effect of efficiently and accurately detecting and identifying the abnormal data in a real-time environment is achieved.

The invention relates to the technical field of semiconductors, and discloses a wafer defect visual AI detection method, device and equipment and a storage medium, and the method comprises the steps: extracting a single target die image from a complete wafer image, forming a data set original image, carrying out the automatic marking of the data set original image, and generating a wafer defect instance segmentation marking data set. Calling a PyTorch framework to construct an initial deep learning wafer detection model, wherein the initial deep learning wafer detection model uses a Vue framework to realize front-end logic; training the initial deep learning wafer detection model pair according to the wafer defect instance segmentation labeling data set to obtain a trained deep learning wafer detection model, obtaining a to-be-detected wafer, detecting the to-be-detected wafer according to the trained deep learning wafer detection model to obtain wafer defect data corresponding to the to-be-detected wafer, and obtaining the wafer defect data corresponding to the to-be-detected wafer. The detection process is high in automation degree, high in accuracy, good in expansibility and wide in application range.

Systems and methods for automated unmanned aerial vehicle recognition. A multiplicity of receivers captures RF data and transmits the RF data to at least one node device. The at least one node device comprises a signal processing engine, a detection engine, a classification engine, and a direction finding engine. The at least one node device is configured with an artificial intelligence algorithm. The detection engine and classification engine are trained to detect and classify signals from unmanned vehicles and their controllers based on processed data from the signal processing engine. The direction finding engine is operable to provide lines of bearing for detected unmanned vehicles.

The invention discloses a power load prediction method based on a dynamic expert pool and a load balancing mechanism MoE, and belongs to the field of power load prediction, and the method comprises the following steps: collecting multi-dimensional input data needed by power load prediction, detecting and repairing an abnormal value in the input data, and obtaining a power load prediction result; constructing a time-feature matrix by using the repair data; projecting the time feature matrix into three subspaces of Q, K and V, calculating attention weights among feature dimensions, and obtaining enhanced features through attention weighting; the enhanced features are input into a routing layer of the MoE model, selection probability distribution of experts is calculated, the MoE model adopts a dynamic expert pool and introduces a load balancing mechanism, and a Top-2 expert selection strategy is randomly distributed in the reasoning stage; and when the MoE model is migrated to a new scene, updating the attention weight of the MoE model by adopting a meta-learning strategy. According to the method, complex factors influencing the power load can be comprehensively captured, and the power load prediction precision and the cross-scene adaptive capacity are remarkably improved.

The invention relates to the technical field of data detection, in particular to an unmanned aerial vehicle atmospheric data anomaly detection and correction system based on deep learning, and provides the following scheme: an air mass reference coordinate system is constructed based on the attitude of an unmanned aerial vehicle and a relative wind direction; re-projecting the original measurement data to obtain an atmosphere data sequence with a stable reference direction; then, generating a frequency mask dynamically changing along with time, wherein the frequency mask is used for constraining positioning of disturbance energy of the rotor wing in the time-frequency analysis process; performing multi-scale decomposition on the time-frequency coefficient matrix based on a multi-channel frequency mask, and respectively extracting an aerodynamic disturbance estimation signal, a background atmosphere signal and a turbulence structure signal; and inputting the three types of signals into a preset deep learning model to realize identification and corresponding correction of atmospheric data anomaly. According to the invention, rotor disturbance and a real atmospheric structure can be distinguished, and the quality and reliability of atmospheric observation data are improved.

The invention provides an Internet of Vehicles multi-protocol dynamic analysis method based on a protocol description model. The Internet of Vehicles multi-protocol dynamic analysis method comprises the following steps: constructing a multi-level PDM comprising a protocol metadata layer, a dynamic adaptation layer, a security verification layer and a resource scheduling layer; generating a dynamic analysis engine integrating protocol identification, parameter adjustment and result standardization functions based on PDM; after receiving multi-source data, calling an engine to analyze, distributing resources by combining a resource scheduling layer through PDM safety verification, and outputting standardized analysis data; when protocol updating is detected, PDM is updated through edge-cloud collaboration, and a new protocol can be adapted without reconstructing an engine. According to the invention, the analysis flexibility and precision are improved through the PDM dynamic adaptation to the Internet of Vehicles scene; double security check and Nash equilibrium resource scheduling guarantee security and real-time performance; edge-cloud collaborative updating gives consideration to efficiency and privacy, and is suitable for a multi-protocol dynamic analysis scene of the Internet of Vehicles.

Systems and methods for automated unmanned aerial vehicle recognition. A multiplicity of receivers captures RF data and transmits the RF data to at least one node device. The at least one node device comprises a signal processing engine, a detection engine, a classification engine, and a direction finding engine. The at least one node device is configured with an artificial intelligence algorithm. The detection engine and classification engine are trained to detect and classify signals from unmanned vehicles and their controllers based on processed data from the signal processing engine. The direction finding engine is operable to provide lines of bearing for detected unmanned vehicles.

The invention relates to a data synchronization method and system for intelligent dirty data detection and restoration based on DataX. According to the method, an intelligent dirty data processing engine is embedded between a Reader plug-in and a Writer plug-in of DataX, type matching, format verification, constraint conflict pre-detection and abnormal value recognition are completed online, and type conversion, format standardization, abnormal value replacement or flexible writing are carried out on dirty data according to a preset or self-adaptive strategy. And finally, the Writer executes insertion, updating, skipping or log recording only according to an instruction, and a traceable governance log is generated in the whole process. According to the method, the synchronization-cleaning capability is embedded into the DataX framework, so that the synchronization success rate and the data quality are remarkably improved, the external ETL dependence and the artificial script cost are reduced, and the method has the advantages of high intelligence, high throughput and flexible configuration.

A method of detecting anomalies using alarm data from a manufacturing network for manufacturing a product in batches includes obtaining alarm data, the alarm data indicates timing alarm state information of each alarm among one or more types of a plurality of alarms among a plurality of types of alarms during a plurality of batches. Each type of alarm is indicative of a different type of problem. The plurality of alarms includes a first alarm and a second alarm. For each batch, a probability that the second alarm will occur within a specified period of time after the first alarm is calculated based on the alarm data. A cumulative sum (CUSUM) of average deviations of probabilities calculated for each batch is stepwise calculated, and a batch having a maximum absolute value of the CUSUM is identified as a point-of-change batch exhibiting a potential anomaly indicative of a problem in the manufacturing network that requires investigation or repair.

The invention discloses a flight flow intelligent prediction and air traffic collaborative optimization system based on big data, and relates to the technical field of air traffic management. Comprising the steps that a data acquisition module acquires multi-source data from a civil aviation database and aligns the multi-source data to generate a fusion data matrix; the traffic prediction module extracts traffic feature vectors, generates a feature similarity matrix by calculating distribution differences, and generates a basic traffic prediction result based on migration prediction model parameters; the disturbance correction module detects a sudden disturbance event based on real-time weather and airspace state data and corrects a basic prediction result; the capacity adjusting module generates a capacity adjusting scheme when the capacity deviation value exceeds a threshold value according to the correction result and the airport operation capability data; and the collaborative optimization module finally generates a multi-airport collaborative scheduling scheme through a game equilibrium algorithm based on the correction result, the capacity scheme and the regional coordination data. According to the invention, the accuracy of flight flow prediction and the efficiency of multi-airport collaborative management are effectively improved.

The invention discloses a fruit and vegetable storage and fresh-keeping control management system and method based on full-period detection, and relates to the technical field of fruit and vegetable storage management, and the method comprises the following steps: dividing a fruit and vegetable storage period into an initial fruit and vegetable storage period, a stable storage period and a later fruit and vegetable storage period; setting a storage environment fluctuation detection mechanism for judging whether storage environment leakage occurs or not according to the environment fluctuation degree; setting a storage environment leakage risk analysis mechanism for dynamically adjusting the time interval of data detection; setting a fruit and vegetable storage and detection period adjusting mechanism for adaptively adjusting the period duration of a stable storage period and the time interval of storage environment detection; the timeliness and accuracy of abnormal fluctuation detection are improved, so that the damage of storage environment fluctuation to the physiological state of fruits and vegetables is reduced; the duration of the stable storage period and the time interval of storage environment detection are adaptively shortened, and the problem that the rotting rate of fruits and vegetables in the later period is too high due to the fact that a fixed storage period is selected is solved.

The invention provides a safety production data detection method and a safety production data detection system, which are applied to the technical field of lock processing, and can more accurately and comprehensively identify and cause potential abnormities in the production process by introducing comparative analysis of multi-dimensional current characteristics and combining standardized processing and micro-cutting strategies, so that the safety production efficiency is improved. The problems of inaccurate diagnosis and easy misjudgment in the prior art are effectively solved, so that the safety production level and efficiency in the precision manufacturing fields such as lock processing are remarkably improved, unnecessary production interruption and resource waste are avoided, and the credibility of an automatic diagnosis system is enhanced.

The application discloses a dual-channel parallel detection airborne phased array radar ground moving target extraction method, comprising the following steps: acquiring coherent pulse train data of an airborne phased array radar; forming sum-difference beam data according to preset sum-difference beam weighting coefficients; constructing a pulse compression matching function to perform pulse compression on the sum-difference beam data; performing platform motion compensation and pulse Doppler processing on the pulse compressed sum-difference beam data; performing high-speed channel determination according to amplitude statistical results of each Doppler unit of the sum beam, and performing CFAR detection; performing low-speed channel determination according to the amplitude statistical results of each Doppler unit of the sum beam, processing low-speed channel data by using sum-difference 3DT-STAP algorithm, and performing CFAR detection and amplitude normalization processing on the processed data; merging the CFAR detection results of the high-speed channel data and the CFAR detection and amplitude normalization processing results of the low-speed channel data, and extracting a point track.

The invention provides an intelligent driving system evaluation method and device, electronic equipment and a storage medium, and relates to the technical field of intelligent driving, detection data of an intelligent driving system is collected, and the detection data comprises environment data and driving behavior data of each vehicle; performing causal reasoning according to the environment data and the driving behavior data in combination with a preset confrontation scene, and updating a scene key factor of each driving scene; performing user portrait processing according to the environment data and the driving behavior data, and updating the weight of the evaluation dimension decision model; and according to the scene key factors and the weights, combining with the environment data and the driving behavior data, applying an evaluation dimension decision model to carry out personalized scoring on each driving scene, and obtaining a personalized scoring result for optimizing the intelligent driving system. Accurate, dynamic and demand-fitting adaptive evaluation of the intelligent driving system is realized, so that the intelligent driving system is optimized.

The invention belongs to the technical field of medical robots and minimally invasive surgery, and provides a robot control system and method for blue laser vaporization surgery of prostatic hyperplasia. Mapping the nuclear magnetic image volume data to a deformation field under an ultrasonic acquisition coordinate system, and deforming the preoperative nuclear magnetic image to an intra-operative ultrasonic space by using the deformation field to complete image registration; fusing the registered image with a stereoscopic vision system; the spatial depth of the surface of the target tissue is obtained from the endoscopic image so as to supplement navigation information; according to the utility model, the prostate deformation and probe posture change adaptive capacity in an operation is improved, the real-time visual closed-loop regulation and control capacity is improved, and the characteristics of small light spots, shallow heat diffusion, excellent hemostasis and the like of blue laser are combined, so that the vaporization and hemostasis precision in a tiny blood vessel dense area is ensured, and the problems of large tissue trauma and the like are avoided.

An optical multi-pass cell system, a gas absorption data detection method, and a terminal device. The optical multi-pass cell system comprises: an optical multi-pass cell (11), which comprises an input end (101) and an output end (102); and a driving mechanism (12), which is used for driving M sub-mirrors of the optical multi-pass cell (11) to perform periodic reciprocating motion in a Z-axis direction during each data collection process of a gas absorption data detection process, wherein the Z-axis direction is parallel to the optical axis of the optical multi-pass cell (11), M≥1, and the detection process comprises: when the optical multi-pass cell (11) is filled with a gas to be detected, a detection beam is inputted from the input end (101), is reflected multiple times in the optical multi-pass cell (11), and then forms an interference beam that is outputted from the output end (102). Optically noise-smoothed gas absorption data can be obtained, such that the sensitivity of detecting gas absorption data on the basis of an optical multi-pass cell system can be improved.

The invention relates to the technical field of machine vision detection, in particular to a flaw detection method and system for large mold production based on machine vision. The method comprises the following steps: positioning the surface and a key curvature area of a spherical acoustic focusing lens mold, and establishing a three-dimensional coordinate reference frame of the spherical acoustic focusing lens mold; the method comprises the following steps: acquiring an image of a spherical area of a whole spherical acoustic focusing lens mold, preprocessing the image, and acquiring point cloud data of the surface of the spherical acoustic focusing lens mold; and based on the preprocessed image, utilizing a convolutional neural network model to identify the surface defect of the spherical acoustic focusing lens mold, and combining point cloud data to detect an area with geometric anomaly. According to the method, a unified three-dimensional coordinate reference frame is established, and high-resolution image data and point cloud data generated by high-precision structured light / laser scanning are precisely registered in space.

The invention discloses an image detection model training method, an image detection method, a device and equipment, which are applied to the technical field of image processing, and the method comprises the steps: obtaining training samples from a training sample set, the training samples being image blocks in intracranial three-dimensional magnetic resonance angiography data, the label information of the training sample comprises a thermodynamic diagram label and an aneurysm quantity label, and the thermodynamic diagram label represents the central point position of the aneurysm in the training sample; inputting the training sample into a preset detection network model to obtain a predicted thermodynamic diagram and a predicted density diagram; calculating the comprehensive loss based on the predicted thermodynamic diagram and the thermodynamic diagram label, and based on the predicted density map and the aneurysm number label; and updating parameters of the preset detection network model by using the comprehensive loss until a training ending condition is met, and determining the preset detection network model with the updated current parameters as a target image detection model. In this way, the accuracy of intracranial three-dimensional magnetic resonance angiography data detection is improved.

The invention belongs to the technical field of data detection and processing, and provides a user electricity charge data anomaly detection method and system, and the method comprises the steps: building and employing a season sensitive baseline model to automatically learn an electricity utilization fluctuation rule of a specific season of a region based on historical electricity utilization data and meteorological data, introducing a season regulation factor into the season sensitive baseline model, and obtaining a user electricity charge anomaly detection result. And deviation correction is carried out in combination with the temperature data, so that reasonable difference of electricity utilization in different seasons is eliminated, temperature difference between south and north can be distinguished, and the false alarm rate of user electricity charge data anomaly detection is greatly reduced. Meanwhile, through year-on-year-on-year-on-year-on-year-on-year-on-year-on-year-on-year-on-year-on-year-on-year-on-year-on-year-on-year-

A smart cart system accounts for edge cases in user interactions by leveraging sensor data and machine-learning models of a smart cart system. For example, a smart cart system uses sensor data to detect when a user removes an item from the smart cart system and presents content to the user on a display of the smart cart system based on the removed item. The smart cart system captures images of the storage area and applies an item identification model to the images to identify the item removed from the storage area. The smart cart system identifies a set of candidate items based on location sensor data describing a location of the smart cart system when the item was removed and computes presentation scores for each of the set of candidate items based on item data for each item the removed item.

Systems and methods provide agentless security assessment for workloads and cloud posture control across multi-cloud environments. Discovery modules are configured with collection intervals to ingest posture control data including assets, identities, configurations, activities, network flows, and build-time artifacts. A multi-cloud configuration inventory maintains current and historical states and produces misconfiguration and identity-activity findings. For workload vulnerability evaluation, an external snapshot manager obtains point-in-time root-disk state without installing an in-workload agent. A file system data processor derives operating system and package metadata, and a detector matches the metadata against a vulnerability feed refreshed on a recurring basis to identify vulnerabilities. Identified vulnerabilities are correlated with misconfiguration and activity findings to generate prioritized risk exposures. Results are stored per workload and presented through graphical user interfaces that display risk levels, timelines, alerts, and guided remediation, enabling continuous, low-overhead security coverage for cloud workloads and configurations.