455 results about "Multi feature" patented technology

The invention relates to the technical field of intelligent wearable health management, in particular to a health management method and device based on multi-modal data, equipment and a storage medium. Comprising the steps of obtaining multi-feature-dimension original data to construct an original data set; performing data preprocessing and feature extraction on the original data set to generate a feature vector set; performing trend capture on the feature vector set according to a time sequence to obtain a time sequence feature vector of each modal; calculating a cross-modal association weight according to each modal time sequence feature vector, outputting a joint feature vector based on the cross-modal association weight, and generating a health index value of each evaluation target dimension according to the joint feature vector; and forming a state vector by the health index value and the historical behavior data of the user, calculating action probability distribution according to the state vector, and sampling and outputting a health intervention suggestion based on the probability distribution. The problem of health assessment deviation caused by insufficient multi-modal data fusion and insufficient time sequence association mining in the prior art can be solved.

The invention discloses a DeepSORT pedestrian tracking method based on multi-feature space-time cooperative interaction, and belongs to the field of computer vision and intelligent video analysis. The method comprises the following steps: acquiring and processing pedestrian data, constructing a target detection and feature extraction model, detecting a test set after training to generate a candidate box, extracting appearance features to construct a cost matrix, matching and updating a trajectory by using a Hungary algorithm, and finally outputting a visual tracking result. In the detection stage, a small target feature enhancement pyramid is designed to improve the small target detection precision, PSConv, Triplet Attention and DyHead are fused to construct a multi-dimensional feature interaction mechanism, and the scale adaptability and the anti-shielding capability are enhanced; in the tracking stage, an IAU module is embedded into an Re-ID branch of DeepSORT, feature discrimination is enhanced through space-time and channel feature dynamic modeling, and ID Switch is reduced. The method effectively improves the perception recognition capability of a multi-scale and strong-shielding target, guarantees the detection accuracy and tracking robustness in a complex environment, and has a good application deployment value.

The invention provides a multi-feature trajectory prediction and cross-defense early warning method for edge and coast defense control objects, relates to the technical field of trajectory prediction and intelligent control, and solves various limitation problems of poor trajectory continuity, lack of environmental constraints and the like of an existing trajectory prediction technology in edge and coast defense scenes. The method comprises the following steps: firstly, acquiring multi-channel monitoring video data, performing environment adaptation preprocessing, and detecting each target and category; based on a detection result, splicing the track fragments in the single-path monitoring range, and obtaining global track data of the target after preset constraint inspection; extracting features of the global trajectory data from a plurality of preset dimensions, and performing feature fusion to obtain high-dimensional vector representation of the target trajectory; and constructing a multi-dimensional feature constraint trajectory prediction model, respectively modeling through time dependence and space interaction, and outputting a subsequent trajectory of the target under a preset constraint. The method can effectively realize prediction and cross-defense risk identification of future trajectories of key control objects, and is applied to the field of public security.

The invention discloses a dynamic scene robust vision SLAM (Simultaneous Localization and Mapping) method based on multi-feature collaborative optimization, which comprises the following steps of: acquiring an image sequence, carrying out dynamic target detection and segmentation through an instance segmentation network, generating a segmentation mask containing a dynamic region mark, and identifying and separating a dynamic object and a static background; removing feature points corresponding to the dynamic object based on the segmentation mask to obtain static feature points; carrying out pose estimation based on the static feature points, and for the key frame, carrying out feature matching with the previous key frame by minimizing a re-projection error, and solving to obtain the camera pose of each key frame; for non-key frames, performing camera pose tracking and data association on the previous frame by adopting an optical flow algorithm, and accumulating solving results to obtain pose tracks of all the non-key frames; the key frames and the non-key frames are subjected to differential processing by fusing feature matching and an optical flow algorithm, so that the calculation efficiency is remarkably improved while the positioning precision is ensured, and the real-time performance is improved.

The invention relates to an unmanned aerial vehicle track prediction method based on multi-feature LSTM, and belongs to the technical field of data processing, and the method comprises the following steps: 1, constructing a six-degree-of-freedom global motion state matrix of a target unmanned aerial vehicle; 2, performing joint estimation to obtain a three-dimensional wind field vector acting on the target unmanned aerial vehicle, and extracting the real-time speed direction of the target unmanned aerial vehicle; 3, fusing the historical state sequence, the three-dimensional wind field vector and the speed direction unit vector, and constructing a multi-dimensional time sequence feature vector; 4, the LSTM neural network learns a nonlinear maneuvering mode of the target unmanned aerial vehicle under the influence of the wind field through a door control mechanism of a forgetting door, an input door and an output door; and step 5, outputting the position probability distribution of the target unmanned aerial vehicle at a plurality of time points in the future through the LSTM neural network. The method has the advantages that the space-time cone representing the future trajectory is formed, and trajectory prediction of the target unmanned aerial vehicle is completed.

The invention provides a ship noise multi-feature classifier data enhancement method and system based on a multi-fine-grained conditional diffusion model. And compressing a waveform to a potential space through VQ-VAE, extracting a ship type / ship name cross semantic vector by using ResNet, and optimizing clustering in combination with a loss function. And a one-dimensional U-Net conditional diffusion model is constructed, unconditional / conditional model output is dynamically weighted and fused, and the weight is adaptively adjusted according to training loss. In the generation stage, a semantic prototype is constructed by using a high-fine-granularity label, parameters are determined by using low / medium-granularity mean value sampling and Bayesian optimization, and fine-granularity controllable waveform generation is realized. After the generated data is converted into multiple features such as MFCC and Lofar, the generated data and original data are combined to train a classifier, and a virtual class strategy relieves class imbalance. Experiments show that the MSE of generated data and real data is reduced, the classification accuracy is improved, the data diversity and the model generalization ability are remarkably enhanced, and the method is suitable for scenes such as underwater target recognition.

The invention discloses a navel orange grabbing pose estimation method based on multi-feature segmentation and a visual hedgehog algorithm, and relates to the technical field of computer vision, and the navel orange grabbing pose estimation method comprises the steps: collecting a color image and a depth image of a target navel orange, carrying out the spatial registration, and generating a three-dimensional point cloud; performing sphere fitting on the independent navel orange instance point cloud, outputting sphere center coordinates and radius parameters of each navel orange, and taking the sphere center coordinates and radius parameters of each navel orange as geometric data of the navel oranges; and collision detection and visibility analysis are conducted on the grabbing pose candidate set, collision cost and visibility cost are generated, comprehensive optimization is conducted on the collision cost and the visibility cost through a multi-target optimization function, the optimal grabbing pose is generated, and the mechanical arm is driven to execute grabbing operation based on the optimal grabbing pose. According to the method, the multi-feature segmentation and the visual hedgehog algorithm are combined, so that accurate calculation of the navel orange grabbing pose is realized.

The invention discloses a voice emotion recognition method based on multiple scales and multiple features, and belongs to the technical field of artificial intelligence. The method comprises the following steps: firstly, preprocessing an audio signal and extracting a spectrogram and a Mel-frequency cepstral coefficient; then, a residual network, a bidirectional long-short-term memory network and a HuBERT pre-training model are respectively utilized to extract spectrogram high-order spatial features, time sequence context features and voice semantic embedding features; secondly, inputting the first two features into a multi-dimensional multi-scale feature extraction module to extract richer time-frequency features, performing deep fusion by using a multi-layer cross attention mechanism, and performing weighted fusion with speech semantic embedded features; and finally, all the advanced features are spliced, and a final emotion category is recognized through a full-connection classifier. According to the invention, through combination of multi-scale feature extraction and an advanced fusion mechanism, the problem of insufficient complex emotion modeling ability in the prior art is effectively overcome, and the accuracy and robustness of voice emotion recognition are significantly improved.

A multi-feature three-dimensional reconstruction 3D Gaussian method based on laser vision comprises the steps that laser radar point cloud and camera images are aligned through space-time calibration, and a unified coordinate system is established; extracting geometric features by using point cloud data acquired by the Lidar point cloud, and initializing a Gaussian ellipsoid according to the Lidar point cloud; optimizing the brightness, the contrast ratio and the structural similarity of the rendered image and the real image by combining the mean absolute error L1 and the structural similarity SSIM; the curvatures of Gaussian ellipsoids of K-nearest neighbors are forced to be consistent, and long and short axes and line and surface features of the Gaussian ellipsoids are aligned to reduce geometric distortion; the distribution density of 3D Gaussian is dynamically adjusted through line / surface features and visual structure information extracted by Lidar, and balance between geometric detail enhancement and calculation efficiency is achieved. According to the method, the position, the scale and the rotation parameters of Gaussian are uniformly optimized, and the details and the calculation efficiency of the model are balanced while the consistency of the model structure is improved.

The invention relates to an industrial robot fault diagnosis method, device and equipment and a computer readable medium. The method comprises the steps that multiple pieces of operation state data of the industrial robot are acquired; preprocessing the operation state data to obtain enhanced data; performing feature extraction on the enhanced data, and constructing an operation feature vector of the industrial robot based on the extracted features; calculating the evidence support degree of each feature subset for different faults in the operation feature vector; and fusing all the evidence support degrees to obtain a fusion support degree, and determining that the industrial robot has a fault under the condition that the fusion support degree is greater than a preset alarm threshold value. According to the method, the multi-source operation state data is collected, preprocessing, feature vector construction and multi-feature subset evidence support degree calculation and fusion are performed, the multi-dimensional operation state of the robot is covered, the risk of misjudgment of a single feature is reduced, accurate fault recognition is achieved, and the technical problem that the accuracy of robot fault prediction is low is solved.

The invention relates to the technical field of entity relationships, and discloses a multi-feature fused entity relationship extraction optimization model construction method and system, and the method comprises the steps: obtaining an input text sequence and a relationship label set, and carrying out the coding processing, and obtaining a text embedding vector and a relationship embedding vector; calculating the similarity between the text embedding vector and the relation embedding vector, screening in combination with a similarity threshold to obtain a candidate relation subset, and fusing the relation embedding vector and the text embedding vector in the candidate relation subset to generate a relation enhancement vector; processing the relation enhancement vector through a mixed attention mechanism to obtain a semantic enhancement vector; constructing a three-dimensional marking matrix based on the semantic enhancement vector, and marking the three-dimensional marking matrix by adopting a diagonal marking strategy; and carrying out decoding processing on the three-dimensional mark matrix, and generating a structured entity relationship triple by adopting a constraint decoding strategy. The interactive fusion of the text and the relation label can be accurately processed in a unified coding space.

The invention relates to the technical field of computer vision, in particular to a multi-feature fusion visible light-infrared pedestrian re-identification method, which comprises the following steps of: respectively extracting low-layer specific features of two modes of an input image through independent visible light and infrared branches, and inputting the low-layer specific features into a shared deep network; a wavelet enhancement multi-feature generation module WMFGM and a coordinate attention fusion module CAFblock are introduced into the modal sharing network; a statistical normalized attention prototype module SNAP is introduced; the statistical normalization attention prototype module strengthens response of a target area through a statistical significance attention mechanism and performs prototype aggregation in a normalized semantic similarity space, so that local feature information with consistent semantics is extracted from visible light and infrared modalities; local feature information and global feature information are used for pedestrian distinguishing, various loss functions are used for guidance during model training, and cross-modal and intra-modal differences are effectively relieved.

The invention relates to the technical field of fault prediction, and particularly discloses a fault prediction method and system for a high-voltage switch operating mechanism. According to the method, early-stage accurate prediction of the fault of the high-voltage switch operating mechanism is realized through multi-feature extraction and fusion in combination with a lightweight Transformer model, and compared with a traditional single current signal analysis method, the method fully considers the coupling effect of electromagnetic and mechanical states, captures electromagnetic abnormality through current-electromagnetic coordination node features, and improves the accuracy of the fault diagnosis of the high-voltage switch operating mechanism. According to the method, mechanical vibration faults are recognized through vibration energy features, motion discordance is detected through electromagnetic driving force-motion acceleration cooperative features, so that possible fault modes of the mechanism are comprehensively covered, the accuracy and the real-time performance of evaluation are further improved through efficient fusion of the lightweight Transform model, the early fault missing report rate is effectively reduced, and the reliability of the mechanism is improved. And reliable guarantee is provided for safe operation of a power system.

The invention relates to the technical field of pipeline magnetic flux leakage detection, and provides a double-wheel magnetic flux leakage detection defect alignment method and device based on multi-feature fusion, and the method comprises the steps: obtaining a magnetic flux leakage detection data set, and carrying out the multi-feature vector extraction of the data set, and obtaining a multi-feature vector set; then, generating a first Gaussian pseudo-color image corresponding to the first multi-feature vector and a second Gaussian pseudo-color image corresponding to the second multi-feature vector; defect mileage positioning is carried out based on the two images, and a common defect interval is determined; in the common defect interval, a matching degree function is constructed based on the first multi-feature vector and the second multi-feature vector, and the optimal offset of the second round of magnetic flux leakage detection data relative to the first round of data is calculated; and according to the optimal offset, performing alignment fusion on the first round of magnetic flux leakage detection data and the second round of magnetic flux leakage detection data to obtain target fusion defect data. According to the embodiment, the accuracy and stability of defect alignment can be effectively improved, so that the reliability of pipeline safety management is effectively improved.

The invention relates to the technical field of satellite network routing, in particular to a dynamic satellite network routing selection method and system based on multi-feature learning, and the method comprises the steps: obtaining satellite node information, constructing a contact graph based on the satellite node information, building a dual depth value network according to the contact graph, training the dual depth value network, and obtaining a multi-feature-learning-based dynamic satellite network routing selection result. The method comprises the steps of obtaining a to-be-transmitted dual depth value network, transmitting the to-be-transmitted dual depth value network to a pre-constructed target base station cluster to obtain a pre-training satellite transmission model, and determining a simulated optimal path and a simulated suboptimal path based on a target base station and the pre-training satellite transmission model, and determining a target task transmission path based on the simulated optimal path and the simulated suboptimal path. According to the method, selection of the satellite network route is realized through fusion of the multi-feature data, and the satellite network route is adjusted when sudden interference influence factors exist, so that a more adaptive satellite network route is obtained.

The invention discloses a non-motor vehicle violation and illegal behavior intelligent identification method based on multiple features, relates to the technical field of non-motor vehicle behavior identification, and solves the problem that in the prior art, non-motor vehicle violation behavior identification systems mostly depend on a single feature for identification, so that misjudgment or missed judgment is easily caused in a complex traffic scene. According to the method, the global semantic understanding capability of a large model is introduced, so that deep fusion of various characteristics such as appearance, movement, behaviors and environment is realized. Therefore, the system can understand the traffic scene more comprehensively, so that the traffic violation behavior of the non-motor vehicle can be identified more accurately. In addition, the method further improves the recognition capability of the key information through the attention mechanism and region enhanced feature extraction. The multi-feature fusion mode not only improves the recognition accuracy, but also enhances the robustness and generalization ability of the system, so that the system can better adapt to different traffic scenes and environmental conditions.

The invention discloses a face image super-resolution reconstruction method and system based on multi-scale multi-feature interactive learning, and the method comprises the steps: firstly, extracting the spatial shallow features and frequency domain shallow features of an input low-resolution face image through a spatial branch and a frequency domain branch respectively; then, the features are input into a cascaded deep interaction module, and multi-scale feature extraction, space-frequency domain feature interaction based on a cross-modal attention mechanism and multi-scale feature aggregation are performed in sequence to capture deep structures and detail features. And then, fusing the shallow and deep features, and respectively outputting frequency domain and spatial domain super-resolution images through a reconstruction module. And a combined loss function combining space and frequency domain pixel loss and frequency domain loss is adopted for supervision. According to the method, structural consistency is kept and high-frequency details are recovered through a frequency domain, fuzzy artifacts are reduced through a multi-scale module, efficient feature fusion is achieved through cross-modal interaction, and the definition, the reality sense and the overall visual quality of a reconstructed image are remarkably improved.

The invention belongs to the technical field of data processing, and particularly discloses a multi-feature code matching method, a multi-feature code matching device, electronic equipment, a medium and a program product. Single feature code original information including identification information, matching attribute information, user association information and matching state information is constructed; judging the rule type according to the anchor point flag bit, if the rule is a common rule, querying a combination type and a serial number through a combination identifier, and executing hierarchical multi-feature code matching; and if the rule is the anchor point rule, executing pure anchor point feature code matching based on the anchor point type and the anchor point identifier. According to the method, differentiated processing is carried out between a common rule and an anchor point rule, so that the problems of high complexity, table item expansion and large resource occupation of a multi-feature code combination rule are effectively solved, efficient identification and matching of multi-dimensional message features are realized, and the method has the advantages of high processing performance, low resource consumption and good expansibility.

The invention discloses a power equipment partial discharge signal classification method and system based on multi-feature extraction, relates to the technical field of power equipment state monitoring and fault diagnosis, and is suitable for online monitoring of high-voltage equipment such as transformers, reactors, power cables and gas insulated switchgear. According to the method, through multiple feature extraction technologies such as wavelet denoising, variational mode decomposition, multi-scale dispersion entropy, mutual information and correlation entropy, and in combination with principal component analysis dimension reduction and support vector machine, random forest and K-nearest neighbor multi-model fusion classification, high-precision classification of partial discharge signals in a complex electromagnetic interference environment is realized. According to the method, the partial discharge type can be quickly identified and classified, and the accuracy and robustness of signal classification are improved. Compared with a traditional classification method, the method can effectively improve the recognition accuracy, improves the online monitoring precision of the partial discharge signal, and provides a scientific basis for state evaluation and intelligent maintenance of power equipment.

The invention discloses an on-line detection method for surface defects of an EVA (Ethylene Vinyl Acetate) adhesive film based on machine vision, which belongs to the technical field of industrial automatic quality detection and comprises the following steps: acquiring multiband image data of the surface of the EVA adhesive film acquired by a preset multi-angle self-adaptive imaging system; performing background difference processing on the multiband image data to generate a difference image; performing adaptive contrast enhancement processing on the difference image to obtain an enhanced image; extracting texture feature data and morphological feature data from the enhanced image to generate a multi-feature data set; performing dynamic weighted fusion processing on the multi-feature data set to generate a defect probability graph; the defect probability graph is input into the layered detection network for processing, a defect detection result is generated, the technical scheme that multi-angle self-adaptive imaging, multi-feature dynamic fusion and the layered detection network are combined is adopted, the defect imaging quality and feature significance can be improved in a self-adaptive mode, and high-speed accurate detection and dynamic optimization of the surface defects of the EVA adhesive film are achieved.

The invention discloses a dam upstream navigation management prediction method based on multi-feature factor research, and relates to the technical field of dam upstream navigation management. According to the method, multi-dimensional data of ship foundation, traffic flow, scheduling management and environment are collected, and key feature factors are extracted after preprocessing; constructing a navigation efficiency evaluation model fusing the cargo load, the ex-warehouse flow and the waiting time, and determining a weight coefficient by adopting a maximum variance method; a navigation state prediction model is constructed based on a random forest algorithm, and prediction of traffic flow density, collision risk and navigation efficiency is realized; and finally, generating an air route planning, scheduling adjustment and emergency management optimization scheme. According to the method, the limitation of single factor consideration of a traditional method is overcome, the comprehensiveness of navigation efficiency evaluation and the accuracy of navigation state prediction are improved, scientific decision support can be provided for navigation management of the upstream water area of the dam, the collision risk is effectively reduced, and the navigation efficiency is improved.

The invention relates to a pointer instrument reading method and system combining image correction and multi-feature recognition, and the method comprises the steps: obtaining an original image, positioning and cutting a dial region, and obtaining a to-be-processed image; calling the reference dial image, performing feature point identification, and performing perspective transformation on the to-be-corrected image to obtain a standard image; carrying out feature recognition on the standard image, wherein necessary categories comprise a dial plate center point, a half pointer, scale marks and numbers; obtaining pointer tip coordinates based on the half pointer and the dial center point; clustering the identified digits to obtain a plurality of scale values; correlating corresponding scale marks and angle information for each scale value, and performing ascending order arrangement; and screening out two scale marks closest to the tip of the pointer, calculating the relative angle between the tip of the pointer and the two scale marks, and carrying out interpolation to obtain the instrument reading. According to the pointer type instrument reading method and system, the accuracy of the instrument at different inclination angles can be effectively improved.

The invention relates to a neural electrophysiological signal analysis technology in the field of brain cognition calculation, in particular to a cross-individual EEG driving fatigue detection method based on multi-feature comparative learning. The method comprises the steps that PSD features and DE features are obtained based on electroencephalogram signal sample data; extracting a shallow PSD semantic feature, a shallow DE semantic feature, a deep PSD semantic feature and a deep DE semantic feature; obtaining a PSD prediction classification label and a DE prediction classification label based on the deep semantic features; defining a positive sample pair, a negative sample pair and a semi-positive sample pair, and constructing a contrast loss function; splicing the shallow semantic features to obtain a preliminary fusion representation, and obtaining a fusion feature vector based on the preliminary fusion representation, the deep PSD semantic features and the deep DE semantic features; and inputting the fused feature vector into a multi-feature decision maker to obtain a fatigue classification label of each sample. According to the method, the adaptability of the model to individual differences is improved, and the generalization ability of the model and the detection robustness are enhanced.

The invention provides an image multi-feature matching, tracking and positioning method, system and device for a rear axle assembly line and a medium, and belongs to the technical field of augmented reality processing of automobile rear axle assembly lines, and the method specifically comprises the steps: carrying out the image collection of a target at the automobile rear axle assembly line; calculating gradients of the image in the horizontal and vertical directions by using a Sobel operator to obtain gradient amplitudes and directions; matching the video frame with an offline collected template, and determining the initial position of the target in the video frame according to the similarity of the gradient descriptors; matching the ORB feature vector of the current image with an ORB feature vector of a pre-stored template image, and when the Hamming distance is smaller than a set threshold value, determining that the current image is a matched feature point pair; and the position of the target object is updated by minimizing the error of the feature points in the adjacent video frames. According to the method, the target can be accurately identified and tracked, and high-precision target position information is provided for quality detection, assembly guidance, process monitoring and other links in industrial automatic production.

The invention relates to the technical field of information security, in particular to an AI video detection method based on multi-feature branch fusion, and the method comprises the following steps: extracting a frame extraction color image from a video, and obtaining a standardized frame sequence after preprocessing; calculating an inter-frame differential volume; obtaining a time sequence spectrum volume according to the standardized frame sequence; performing bilateral filtering decomposition on the standardized frame sequence to obtain an illumination consistency volume; respectively inputting the three types of volume features into a deep convolutional neural network, and after feature extraction and fusion, outputting an AI forgery probability through a classifier; and comparing the AI forgery probability with a decision threshold to generate a video category label, and calculating an index. According to the method, the multi-dimensional features of the time-space domain, the frequency domain and the physical illumination domain are fused, the potential traces of the deeply-forged video are effectively captured, the stable detection performance is kept under various video quality conditions, the accuracy is improved, and reliable technical guarantee is provided for media information security.

The invention relates to the technical field of power system operation and maintenance, and discloses a grounding grid corrosion fault diagnosis method and system, a storage medium and equipment, and the method comprises the steps: arranging a plurality of types of sensors at key points of a grounding grid, collecting the electrical signal and environmental parameter data of the grounding grid in real time, carrying out the preprocessing of the collected original data, and carrying out the calculation of the data; a high-quality multi-source data set is generated; feature extraction is carried out on the multi-source data, normalization and weighted fusion are carried out on extracted multi-source features, and a comprehensive fault feature vector is generated; constructing a diagnosis model suitable for multi-feature input, performing adaptive optimization on parameters and feature fusion weights of the diagnosis model by using a whale optimization algorithm WOA, training the diagnosis model by using a comprehensive fault feature vector, and optimizing the model; and inputting the comprehensive fault feature vector into the optimized diagnosis model, outputting a corrosion fault diagnosis result, and generating a fault report and alarm information, thereby realizing accurate identification, positioning and early warning of grounding grid corrosion.

The invention provides a multi-branch sequence recommendation method based on dynamic channel fusion. The method aims at solving the problems that in sequence recommendation, data have a large amount of noise and are sparse, and existing model prediction is smooth. The method comprises the following steps: on one hand, enabling a user sequence to pass through an embedding layer, introducing position information by utilizing RoPE rotation position coding, and extracting long-term dependency of a user in the sequence through a multi-feature channel feature network to obtain a new embedded Eglobal, a short-term interest and long-term dependency interactive embedded Ecross, an embedded Eema after exponential smoothing filtering and an embedded Efreq after frequency domain modulation; and on the other hand, in order to fully fuse the characteristics of different channels, a traditional hyper-parameter fusion mode is abandoned, and a pooling network layered adaptive fusion mode is adopted. Firstly, four channel features are divided into two groups including a self-attention group and a trend group, features of different channels are learned through Squeeze average pooling, learned parameters are activated through Excitation by adopting sigmoid to obtain channel weights, and feature representations of different groups are sent into a gating network for weighted summation loss calculation.

The invention relates to a multi-feature-value multi-model fusion atrial fibrillation recognition method, and relates to the technical field of medical signal processing, and the method comprises the steps: obtaining an RR interval sequence; in parallel, extracting a multi-dimensional short-time feature set containing the RR interval fuzzy measure entropy from the RR interval sequence of the preset short time period, and generating a short-time recognition result through a short-time model; meanwhile, a multi-dimensional long-time feature set is extracted from the RR interval sequence of the preset long time period, and a long-time recognition result is generated through a long-time model; and the model result fusion module outputs a final atrial fibrillation recognition result based on the long-time recognition result by combining the proportion of the atrial fibrillation conclusions in the plurality of short-time recognition results and applying dual threshold judgment logic. According to the method, non-linear dynamic characteristics of different time scales are combined, and an intelligent fusion strategy of hierarchical decision is adopted, so that the detection sensitivity of paroxysmal atrial fibrillation and the suppression capability of interference noise can be effectively considered, and the comprehensive accuracy of automatic identification of atrial fibrillation is improved.

This invention discloses a tracking method based on multi-scale spatial constraints to combat occlusion. It extracts HOG, CN, and grayscale features from candidate target regions and uses PCA to reduce the dimensionality of HOG and CN features, accelerating computation. A channel weight fusion method is employed, training filters separately for each feature layer. Adaptive fusion weights enhance the response of effective feature layers, addressing the problem of insufficient multi-feature response fusion. A stepped spatial constraint method is proposed to optimize the color space constraint model, preventing model errors from obscuring target information and limiting the effectiveness of the spatial domain constraint model. Adaptive learning rate and diffusion search methods reduce irrelevant information learned by the filters and improve tracking accuracy when the target is occluded. A tritree scale acceleration method is proposed, introducing scale filters and transforming the parallel structure of the scale filters into a tritree classification structure.

The application discloses a method for integrated recognition of converter valve faults based on multi-feature fusion and dynamic graph modeling, and belongs to the field of converter valve fault recognition.The method comprises the following steps: S10, multi-source recording data preprocessing and sample construction; S20, multi-feature extraction based on a heterogeneous deep model; S30, feature semantic alignment based on an attention mechanism; and S40, feature fusion and fault recognition based on a dynamic graph convolution network.Through modeling of the multivariate coupling relationship and dynamic evolution law of multi-source recording data of the converter valve, the accuracy and robustness of the converter valve fault recognition under complex working conditions are improved.