257results about How to "Improve discrimination ability" patented technology

The invention relates to the technical field of medical information, and discloses a pancreatic cancer risk prediction method based on machine learning and multi-modal data, and the method comprises the steps: obtaining the multi-modal data of a target user; performing time sequence deduction on the molecular biological detection data to generate a virtual molecular time sequence; time sequence signals are extracted from the virtual molecule time sequence and the time sequence behavior monitoring data; calculating the dynamic coupling strength between the two time sequence signals to obtain a space-time coupling coefficient; weighted fusion is carried out on the features, and unified multi-modal feature representation is constructed; carrying out multi-modal feature representation training to obtain a special risk prediction model for the target user; and obtaining a risk quantitative score, and identifying a key risk driving factor which contributes to the score most. According to the invention, through multi-modal time sequence fusion and personalized modeling, early-stage, dynamic and explainable and evaluable pancreatic cancer risks are realized.

The invention discloses a hyperspectral image classification method based on frequency domain denoising and element gradient correction, and the method comprises the following steps: carrying out the preprocessing of all hyperspectral image data, and dividing an overall training sample set formed by the processed hyperspectral images into a training set and a verification set; constructing a sample weighting model based on frequency domain denoising and element gradient correction; in the training process, a parameterization frequency spectrum gating sensing transformation module is utilized to map features to a frequency domain through discrete Fourier transform, a learnable frequency spectrum response function is utilized to adaptively suppress spectrum jitter noise, and finally pure features are reconstructed. And automatically constructing a high-confidence pseudo-clean verification set based on a Gaussian mixture model and time domain consistency. According to the method, a time domain momentum updating mechanism is introduced, the variance of statistical estimation is effectively smoothed, random interference caused by training fluctuation is resisted, and the accuracy of pseudo clean set construction and the convergence stability of overall model training are further improved.

The invention discloses a multi-task electromagnetic model based on a hybrid expert network, and belongs to a wireless communication technology. The model comprises a preprocessing module, a feature extraction module, a task output module and a pre-training-fine tuning learning strategy. The preprocessing module carries out standardization processing on the multi-source electromagnetic signals; the feature extraction module is based on a Transform structure, introduces a hybrid expert network to replace part of a traditional feedforward neural network, and dynamically selects an expert sub-network through a task specific routing mechanism; the task output module configures a special structure according to different task targets; in the pre-training stage, a mask auto-encoder is used for pre-training large-scale label-free data, and a downstream task is subjected to full-amount fine adjustment through small-scale label data. According to the method, multi-task collaborative learning and differential expression are realized, and the recognition performance, robustness and processing efficiency of the model in a complex electromagnetic environment are improved.

The invention relates to the technical field of semantic segmentation, in particular to a night semantic segmentation method based on low illumination enhancement and edge optimization, and the method comprises the steps: inputting an image into a low-light enhancement repair network based on the Retinex theory, local contrast enhancement and adaptive feature fusion, obtaining a denoised and enhanced intermediate image, and carrying out the edge optimization of the intermediate image; inputting the intermediate image into a semantic segmentation network to obtain a category distribution diagram of each pixel; inputting a discriminator embedded with a channel attention module according to the category distribution diagram of each pixel, and optimizing a generator composed of a low light enhancement repair network and a semantic segmentation network through a multi-task joint optimization loss function; and inputting a night image to be detected and segmented into the optimized generator, and outputting a segmentation result. By adopting the method, the low-light enhancement repair network is combined with a local contrast enhancement and channel feature fusion mechanism, so that the overall brightness of the image is improved, the details and edge structures of the image are reserved, and the perception capability and robustness of the model are improved.

The invention relates to the technical field of electric data processing, in particular to a data blood relationship automatic tracking method and system based on a table in AI data, and the method comprises the steps: obtaining metadata in the table in the AI data, and analyzing a task log of a calculation engine to construct an initial directed acyclic graph; the initial directed acyclic graph comprises data nodes and explicit blood edges connected with the data nodes; and calculating a flow entropy coupling index between any two data nodes. According to the method, a flow entropy coupling index based on update time correlation and logic evolution divergence based on key entity distribution similarity are constructed, a graph comparison learning model is combined, under the condition that an explicit SQL grammatical structure is lacked, the implicit dependency of upstream and downstream nodes is automatically identified, and the flow entropy coupling index and the logic evolution divergence are combined. And end-to-end full-link data consanguinity complete tracking in a complex AI feature engineering scene is successfully realized.

The invention discloses a multi-modal contrast learning fault diagnosis method for a small sample scene, and the method comprises the following steps: carrying out the enhancement of a one-dimensional signal and two-dimensional image fused fault data set through physical simulation for the small sample scene with scarce industrial fault data, and constructing a positive and negative sample pair through a plurality of data enhancement strategies; based on heterogeneous multi-modal fault data, designing a double-flow encoder architecture of a time sequence branch and an image branch, extracting depth features and mapping the depth features to a unified feature space through a projection head; performing supervised contrast learning pre-training based on intra-modal and inter-modal dual contrast loss; supervision fine tuning is carried out based on multiple loss functions such as physical guidance, so that accurate diagnosis of equipment faults is realized in a small sample scene. According to the fault diagnosis method under the unbalanced sample and limited labeling conditions, the problem that a traditional data driving model depends on large-scale labeling samples is effectively relieved through supervised comparative learning and cross-modal information alignment.

The invention relates to a BERT and multi-module fusion-based track circuit fault label classification method and device and a medium, and the method comprises the steps: obtaining a to-be-classified track circuit fault text, and carrying out the cleaning and standardization processing of the to-be-classified track circuit fault text; performing word segmentation and vectorization processing on the track circuit fault text to generate a lexical symbol sequence conforming to BERT encoder input specifications; extracting a context semantic vector representation of the input lexical symbol sequence through a BERT encoder; performing feature extraction through a TextCNN module and a self-attention module which are parallel to obtain local features and weighted global features, and splicing the local features and the weighted global features to obtain a comprehensive feature vector; and synchronously generating a multi-label classification result of a fault phenomenon, a fault reason and a solution measure corresponding to the current fault text through a multi-label classification head based on the comprehensive feature vector. Compared with the prior art, the method has the advantages of high discrimination, high accuracy, high practicability and the like.

The invention discloses a police target tracking system based on deep learning, and relates to the field of police video monitoring. The system comprises a video sensing and front-end processing unit, an edge intelligent tracking server and a cloud model optimization and command platform, the video sensing and front-end processing unit collects visible light and thermal infrared video streams, the edge intelligent tracking server processes the visible light and thermal infrared video streams, and a stable track with an identity label is generated; and the cloud model optimization and command platform aggregates edge data based on a federated learning framework, iteratively optimizes a global model and realizes command scheduling. According to the system, the tracking robustness, the real-time performance and the multi-target identity keeping accuracy in a complex environment are effectively improved, and the actual combat requirements of police affairs are met.

The invention discloses a cross-city traffic prediction method and system based on multi-modal fusion and spatial expert routing, relates to the technical field of traffic prediction, and provides an adaptive modal selection mechanism based on a signal-to-noise ratio for the problems of missing, noise and uneven quality of multi-modal data in different cities. And a low-quality mode is dynamically suppressed in combination with comparative learning, and robust multi-mode fusion is realized. In order to solve the problems of large space structure difference and weak generalization ability among cities, a multi-modal guided space expert routing architecture is designed: modal sharing experts and routing experts are activated by using a multi-modal context, and local space dependence is adaptively modeled for different functional regions. The method supports any modal combination input, city specific fine tuning is not needed, and the zero sample cross-city prediction performance is significantly improved.

The invention relates to the technical field of clinical medical treatment, and discloses a disease deterioration risk identification method and system based on depth time sequence clustering enhancement, and the method comprises the steps: 1, obtaining multi-modal clinical sequence data of a patient, including physiological indexes of a time sequence, a laboratory detection result, historical diseases and medication data; 2, performing feature extraction and classification on the patient sequences by adopting a knowledge enhanced sequence clustering method, and grouping the patient sequences according to future outcome distribution of the patient sequences; 3, enabling the model to quickly adapt to a prediction task of a new patient subgroup through a meta-training process; and step 4, based on the trained meta-model, carrying out rapid adaptation on the new patient subtype, and predicting the possibility that the new patient subtype has a deterioration event in a certain time window in the future. The method and the system can effectively learn the disease change mode of the patient under the condition of limited clinical data, improve the prediction accuracy of the new patient subgroup, and are especially suitable for clinical prediction scenes under the condition of small samples.

This invention discloses a face forgery detection method based on identity decoupling and adaptive cosine embedding loss. Addressing the problems of existing deepfake detection methods being susceptible to interference from identity information and lacking flexibility due to fixed boundary constraints, this invention constructs an identity-independent deepfake detection network. It decouples the input face image into forensic features containing forgery clues and content features containing identity information, eliminating identity interference by ignoring content features. Simultaneously, it designs an adaptive cosine embedding loss based on visual differences, using absolute error and the visual differences between real and fake image pairs as adaptive boundary values ​​to dynamically adjust feature spacing constraints. This invention significantly improves the model's generalization ability when facing unknown forgery types through the synergistic optimization of decoupled representation learning and adaptive metric learning.

This invention discloses a hyperspectral image restoration method and system based on dual-enhanced low-rank tensor constraints, belonging to the field of image processing technology. The steps are as follows: S1: Acquire hyperspectral image data of the target area; S2: Image preprocessing to generate a test image; S3: Input the test image into a dual-enhanced low-rank tensor constraint model to obtain the restored image; wherein the dual-enhanced low-rank tensor constraint model includes a global prior learning module, a nonlocal prior learning module, and a prior integration module. The test image is first input into the global prior learning module to generate a global learning result, then the test image and the global learning result are input into the nonlocal prior learning module to generate a nonlocal learning result, and finally, the prior integration module performs prior weighted integration of the global and nonlocal learning results to obtain the restored image. The effect is that its performance is superior to other hyperspectral remote sensing image restoration methods, and it has a greater advantage in accurately estimating Earth monitoring data.

This invention relates to the field of electricity meter operation monitoring technology, specifically to an intelligent analysis system for electricity meter operation faults based on sensor monitoring. The system includes: an operation status acquisition module, which collects the voltage, current, active power, reactive power, and power factor of the electricity meter and obtains the operation status at each moment; a periodic analysis module, which uses active power as a load parameter and obtains a basic period based on the load parameters within a preset time period, then segments the preset time period to obtain different time periods and obtains the period fit at each moment; and a fault monitoring module, which obtains the comprehensive rationality of the load linkage deviation at each moment; and uses the comprehensive rationality of the load linkage deviation at a moment and the period fit to correct the squared Mahalanobis distance corresponding to the data points composed of voltage, current, reactive power, and power factor at that moment, and uses the corrected squared Mahalanobis distance for fault monitoring. This application can effectively monitor the operation faults of electricity meters.

The invention discloses an unmanned aerial vehicle remote sensing tree species classification method and system fusing a cascade canopy attention mechanism, and the method comprises the steps: obtaining unmanned aerial vehicle RGB image data of a campus scene, carrying out the preprocessing of the data, and obtaining a standardized input image; constructing a classification model fusing the cascaded canopy attention mechanism, wherein the classification model comprises a background suppression module, a fine-grained feature guide module, a channel alignment neck module and a global-local fusion classification head; inputting the standardized input image into a classification model, and obtaining an enhanced canopy image through a background suppression module; inputting the enhanced canopy image into a module carrying fine-grained feature guidance, and extracting fine-grained canopy features; performing channel calibration and feature smoothing on the fine-grained canopy features through a channel alignment neck module to obtain feature mapping of a unified dimension; and fusing the global canopy semantics and the local texture structure in the feature mapping by using a global-local fusion classification head, and outputting a tree species classification result.

The application relates to the technical field of computer vision, and discloses a semantic segmentation method and device, equipment and a storage medium, which comprise the following steps: preprocessing a to-be-segmented image to obtain a first input image and a second input image; inputting the first input image into a visual language model to obtain category prior information of the first input image; performing pixel feature extraction on the second input image to obtain a pixel feature map; determining classifier parameters through a dynamic analysis network based on the category prior information and the pixel feature map, and constructing a pixel-level classifier according to the classifier parameters; and performing semantic category discrimination on the pixel feature map through the pixel-level classifier to obtain a semantic segmentation result. The visual language model is used to obtain category prior information at the existence level from the input image, the category prior information is added to the construction process of the pixel-level classifier, the parameters of the classifier can adaptively change along with the semantic category of the input image, and the discrimination capability is improved.

The application discloses a kind of mineral resources intelligent prediction method based on PINNs, it is related to mineral resources exploration technical field, including, acquisition mining area multimodal geoscience data, and standardization pretreatment is carried out;Based on mineral resources probability distribution map, spatial target area is automatically delineated, and output hierarchical target area sequence, based on hierarchical target area sequence and mineral resources probability distribution map, application geostatistics method simulates the distribution of different confidence under resource grade and tonnage, and outputs target area resource quantity evaluation report;Uncertainty quantification and visual display are carried out to hierarchical target area sequence and target area resource quantity evaluation report, and comprehensive prediction result drawing is generated.The application realizes the semantic level alignment of multi-source heterogeneous data by constructing geological ontology knowledge graph and using graph convolution network to learn geological concept embedding, and effectively captures the deep feature association related to mineralization in multi-source heterogeneous data.

The application discloses an emotion recognition method based on online cross-modal knowledge distillation, comprising the following steps: acquiring electroencephalogram and electrocardiogram original signals and windowing and cutting; constructing electroencephalogram and electrocardiogram student models, extracting intermediate features from each modal data segment through an encoder, and obtaining non-normalized prediction output through a classifier; constructing a teacher probability distribution through a joint encoder fusion; introducing adaptive contrast loss to align the cross-modal intermediate features, introducing distillation loss to constrain the prediction probability distribution of each modal to align with the teacher probability distribution; synchronously optimizing new student model parameters through online collaborative training; and performing actual inference prediction based on the student model after training. The application combines double modal signals to make up for the defects of single modal information, excavates the complementarity of modes, realizes dynamic generation of teacher supervision signals and real-time collaborative learning of modes through online distillation, does not increase test calculation overhead, effectively improves the recognition accuracy, model robustness and generalization ability, and has good application prospect.

The invention discloses a multi-label smell description prediction method, and relates to the field of compound smell prediction, and the method comprises the steps: obtaining compound identification information, molecular structure descriptors and smell label data, and constructing a multi-label smell data set; generating a molecular structure feature vector through a molecular fingerprint coding technology, and extracting a multi-dimensional descriptor reflecting the physicochemical properties of molecules; compressing the molecular fingerprint features to a low-dimensional space through a dimension reduction algorithm; performing unbalanced data processing on the training set, fusing the dimension-reduced molecular fingerprints with the molecular descriptors to form a joint feature matrix, and configuring a class weight balance mechanism and overfitting suppression parameters by adopting a multi-label classification architecture; independently optimizing a probability threshold for each odor label based on the verification set; and outputting a multi-odor label combination prediction result according to the target molecule identification information. According to the scheme, the multi-odor characteristics of the compound can be accurately depicted, and the combined recognition accuracy of the compound odor is remarkably improved.

A dynamic graph convolution-based adaptive multi-modal data fusion prediction method and system, the method comprising: extracting a text sentiment feature vector for representing sentiment semantics and a personality feature vector for representing individualized traits from the obtained text input, extracting an audio feature vector and a visual feature vector from the obtained audio and video input; after mapping the text sentiment feature vector and the personality feature vector to a shared space, performing semantic alignment by contrast learning constraint and personality weighted regression constraint to obtain an aligned text sentiment feature vector; taking the aligned text sentiment feature vector as a query vector, performing multi-head attention interaction on the audio feature vector and the visual feature vector respectively to obtain an audio enhanced feature vector and a visual enhanced feature vector; constructing a multi-node graph structure and realizing cross-modal feature fusion through dynamic graph convolution to obtain a fusion feature matrix; and generating a prediction result using the fusion feature matrix.

The invention discloses a blueberry tree branch segmentation method and system, and the method comprises the steps: carrying out the multi-scale feature extraction of a to-be-segmented blueberry tree image, and obtaining the feature representation containing low-layer details, middle-layer semantics and high-layer context; information complementation and semantic alignment of deep and shallow layer features are realized through a bidirectional cross-level information interaction module, and fusion features are optimized by using a space and channel attention mechanism to obtain enhanced features; and performing up-sampling and dynamic reconstruction on the enhanced features by adopting a frequency domain dynamic convolution module, refining branch edges and texture details, and outputting a high-resolution pixel-level segmentation result. The method effectively solves the problems of fuzzy boundary, fracture adhesion and background interference in the segmentation of the slender branches and trunks of the blueberries, and has good precision and generalization ability.

The application provides a dynamic causal reinforcement intelligent medical hallucination detection and correction method and system, and relates to the technical field of medical hallucination detection.The method comprises the following steps: acquiring relevant data sources, and constructing a dynamic medical causal graph according to the relevant data sources; determining a to-be-detected diagnosis result, a to-be-detected diagnosis result evidence source and a to-be-detected diagnosis result confidence; acquiring actual detection data; obtaining a first medical hallucination discrimination result; determining whether the to-be-detected text needs to be corrected; in the case that the to-be-detected text needs to be corrected, processing the to-be-detected text according to a medical hallucination correction model to obtain a first training generated text; obtaining a combination loss function; obtaining a trained medical hallucination discrimination model and a trained medical hallucination correction model; and obtaining a final medical hallucination discrimination result and a final generated text.According to the application, the efficiency and accuracy of medical hallucination detection and correction can be improved.

The application relates to the technical field of computer vision, in particular to a power safety monitoring image detection method and system, which comprises the following steps: inputting a to-be-detected power safety monitoring image and a generated image into a trained power safety monitoring image detection model respectively to obtain a detection result, the power safety monitoring image detection model comprises a feature extraction network structure, a feature distillation network structure and a multi-scale aggregation network structure in sequence, the feature extraction network structure is used for extracting local spatial features and global context information features from the to-be-detected power safety monitoring image and each generated image respectively, and fusing the two to obtain fused features corresponding to each image; the feature distillation network structure is used for extracting Value values and Key values from the fused features of each image, and obtaining splicing features based on the Value values and the Key values of each image; and the multi-scale aggregation network structure is used for processing the splicing features to obtain the detection result of the to-be-detected power safety monitoring image.

The application provides a satellite image transformer substation detection method based on a target detection frame center. In view of the low accuracy and incomplete detection of the current single remote sensing image transformer substation target detection result, the method adopts a one-time detection and twice detection combined detection mode, regenerates an image block to be detected remote sensing image based on the center position of the initial detection result detection frame for twice detection, so that the target corresponding to the detection frame is located at the center position of the image block, the detection target is avoided from being cut into different image blocks, the integrity of the detection result frame is ensured, meanwhile, the target located at the center position of the image block can capture more information around the target in the feature extraction process, the discrimination ability of the model is enhanced, more false results in one-time detection are removed, and the accuracy of the transformer substation detection result is improved.

The invention provides an intelligent sorting system for paper sheet-shaped articles, and relates to the technical field of intelligent sorting of paper sheet-shaped articles, and the system comprises the steps that a target image corresponding to a target paper sheet-shaped article is obtained; obtaining a target character in the target image; if the target character is the first preset type character, determining whether the target paper sheet article is abnormal or not according to the gray level uniformity of a character area and an area outside a pattern area in the target image; if the target character is a second preset type character, performing edge contour extraction on the target image to obtain an edge contour image corresponding to the character and the pattern; performing feature extraction on the edge contour image to obtain an edge contour feature vector corresponding to the target image; determining whether the target paper sheet article is normal or not; according to the method, the capability of judging the defects such as middle bending is remarkably enhanced, the reliability and the automation level of overall sorting are improved, and the defects that in the prior art, the detection means is single, and the adaptability is insufficient are overcome.

The application belongs to the field of medical image recognition and relates to an esophageal lesion image recognition method based on a deep neural network, which comprises the following steps: step 1, processing actual digestive endoscopy video data to construct training data; step 2, constructing a deep neural network model and training the deep neural network model through the training data to obtain an early esophageal cancer recognition model; step 3, inputting digestive endoscopy video data to be detected into the early esophageal cancer recognition model, and the early esophageal cancer recognition model judging whether there is a suspected esophageal cancer lesion and outputting a lesion judgment result; the lesion judgment result comprising a spatial position and a progression stage of the lesion; the method models the time sequence dynamics of lesion characteristics in the endoscopy recognition process, and combines a graph neural network to depict the feature correlation of early esophageal squamous cell carcinoma under multi-view conditions, so as to improve the recognition accuracy, consistency and robustness of the model in a real clinical application scenario.

The invention discloses a semantic consistent double contrast learning long tail identification method, and belongs to the technical field of computer vision. Comprising the following steps: step 1, designing a double contrast learning framework, and optimizing a feature space through fusion supervision contrast learning; and meanwhile, a self-supervised comparative learning improvement model is introduced to enhance the feature invariance learning ability of data enhancement transformation. Step 2, designing a prototype-contrast dynamic semantic alignment strategy, adaptively correcting label noise caused by inter-class imbalance by utilizing class prototypes and classification probability in a semantic space, and coordinating the prototype-contrast dynamic semantic alignment strategy of feature space consistency through constraint based on the prototype; and 3, designing a double-branch decoupling classifier, dynamically adjusting the weight of the classifier in combination with a periodic cumulative learning strategy, realizing attention dynamic migration and tail category enhancement, inhibiting a head category dominant effect, and enhancing tail category representation learning. And therefore, the recognition performance of the long tail category is further improved.

Embodiments of the present specification disclose a multi-granularity graph contrast learning method, device, equipment, medium and program product. The method comprises: using a granule enhancement method to adaptively refine an original graph to generate a plurality of target granules; the original graph is composed of a plurality of nodes and edges representing the interaction relationship between the nodes; the target granule represents a multi-granularity homogeneous region composed of at least Z nodes in the original graph; Z is a positive integer greater than 1; a positive sample pair and a negative sample pair are constructed based on a plurality of target granules; the positive sample pair is composed of nodes in the same target granule in the original graph; the negative sample pair is composed of nodes in different target granules in the original graph; a contrast loss is determined based on the positive sample pair and the negative sample pair; and an initial vector of each node in the original graph is optimized based on the contrast loss to obtain a target vector corresponding to each of the plurality of nodes.

The application provides a model-optimized target tracking method and relates to the field of computer vision. The application introduces time regularization to effectively deal with the boundary effect problem and improve the accuracy of target tracking. In view of the complexity of the objective function, the application converts the objective function into a frequency domain and adopts an alternating direction multiplier method (ADMM) to effectively optimize, so that each sub-problem has a corresponding optimal solution. In addition, in terms of feature representation, the application combines traditional manual features and deep convolution features to obtain more semantic information and improve the discrimination performance of the filter. A large number of experiments show that the algorithm proposed in the application performs well on many advanced trackers.

This invention relates to the field of crop identification technology, specifically to a UAV image recognition system for crops in karst landforms based on multi-scale features. The system includes: a feature database construction module for constructing a hierarchical crop feature database, wherein the crop feature database comprises multiple storage areas storing feature data of different crop types, each storage area containing multiple feature layers, and multiple feature layers located in the same storage area are associated with each other; an image acquisition and preprocessing module for acquiring remote sensing images of the area to be identified collected by the UAV, and performing geometric correction, spectral correction, and standardized slice preprocessing on the remote sensing images to obtain the image to be analyzed; and a feature extraction and matching module for communicating with the feature database construction module, thereby improving the efficiency and accuracy of crop type identification.

The invention relates to an optimization method and device of a fine arrangement model, a storage medium and electronic equipment. The method comprises the following steps: acquiring user behavior data, marking an object clicked by a target user in the user behavior data as a positive sample, and marking an object which is not clicked as a negative sample; obtaining a model prediction probability, an exposure frequency and a sample feature of each negative sample, and calculating a dynamic weight of each negative sample; inputting all browsing objects and object information thereof in the user behavior data into a fine arrangement model to perform recommendation probability prediction on all the browsing objects to obtain a prediction result, and calculating to obtain a basic loss function according to the prediction result and click information of all the browsing objects; and performing weighted calculation on the basic loss function according to the dynamic weight of each negative sample to obtain a weighted loss function, and iteratively updating parameters of the fine arrangement model according to the weighted loss function. The technical problems that an existing fine ranking model does not make full use of difficult-to-load samples, and the judgment and ranking precision is limited are solved.