26results about How to "Good distinction" patented technology

The invention discloses a cigarette packet appearance defect detection method and system based on deep learning, and relates to the technical field of appearance detection.When the system runs, multi-angle and multi-light-source image acquisition is carried out through a high-definition camera and scanning equipment, it is ensured that appearance details of cigarette packets are comprehensively and accurately captured, and the detection accuracy of the appearance defects of the cigarette packets is improved. According to the method, firstly, the image quality and consistency are further improved through image cutting, graying processing, denoising processing, image enhancement and data standardization of the data preprocessing module, a solid foundation is laid for subsequent defect detection, secondly, the defect detection module carries out feature extraction through strong and weak data enhancement and a feature extractor, and the defect detection accuracy is improved. By calculating the similarity of positive and negative samples in a feature space and a total loss function # imgabs0 #, defect identification and optimization are performed in combination with MSE loss, the defect detection precision and robustness of the model are improved, the evaluation module evaluates appearance defects in real time by matching with a preset threshold, and a continuously optimized detection closed loop is formed.

The invention belongs to the technical field of data processing, and discloses an enterprise multi-type data labeling method and system based on feature engineering, and the method comprises the steps: collecting initial data of an enterprise internal information system, and dividing the initial data to obtain a text data set, a table data set and an image data set; feature vectors of different types of data are fused after features are extracted and screened, and a multi-dimensional feature space is constructed; constructing an annotation model according to the feature information of the feature space, constructing similar sample pairs as positive sample pairs and irrelevant sample pairs as negative sample pairs for unannotated new data, and comparing, learning and optimizing the similarity between the similar sample pairs; and carrying out label inference based on the feature vectors obtained by comparative learning, and generating a labeling result. According to the method, efficient and automatic labeling of various types of data can be achieved, the method is suitable for various heterogeneous data such as texts, tables and images, and high labeling precision and expansibility are achieved.

The invention provides application of a metabolite S-cysteine succinic acid (S-Cysteinosuccinic acid) as a molecular marker for diagnosis and prognosis of lung cancer brain metastasis tumors, and application of a reagent for detecting the metabolite in preparation of a reagent for diagnosis and prognosis of the lung cancer brain metastasis tumors. The invention further provides a method for evaluating the occurrence risk of the lung cancer brain metastasis tumor, and the method comprises the steps that subjects are divided into a high-risk group of the lung cancer brain metastasis tumor and a low-risk group of the lung cancer brain metastasis tumor according to the expression level of metabolites in samples of the subjects, and the metabolites comprise S-Cysteinosuccinic acid. The invention also relates to a kit or a chip containing the reagent for detecting the metabolite.

The invention relates to a multi-scale feature fusion remote sensing image water body extraction method based on a compressed pyramid gating unit, and belongs to the technical field of remote sensing. The method comprises the following steps: acquiring a water body remote sensing image data set, preprocessing the data set, performing data set division on the data set, and performing data enhancement processing; a CNN hybrid Transform encoder is adopted to extract feature maps of a plurality of scales from the image; adding the feature maps of different scales into a multi-scale jump connection module, and extracting feature information of multiple scales; fusing the feature information of the plurality of scales, and performing weight multiplication on the fused feature and the feature of each scale to obtain a final multi-scale adaptive receptive field selection result feature map; and training the process by adopting cross entropy loss, and performing remote sensing image water body extraction after training is completed. According to the method, the extraction capability of water bodies with different shapes and sizes and the distinguishing capability of water body information and non-water body information can be effectively improved.

The invention belongs to the technical field of medical diagnosis, and provides a methylation biomarker for metabolic diseases and application of the methylation biomarker. The methylation level of at least one part of chr1: 119708470-119708810, chr3: 49653374-49653573, chr3: 125029184-125029370, chr7: 75987317-75987513, chr10: 6145842-6146049, chr17: 47942423-47942577 and chr19: 46778792-46779015 is detected from a sample to be detected, and the methylation level of at least one part of the chr1: 119708470- The high methylation level of the biomarker provided by the invention is highly related to metabolic disorder, and the biomarker has good accuracy and distinguishing ability, can identify metabolic disorder crowds in the early stage of children, and has important clinical significance for early diagnosis and early treatment of related diseases.

The application discloses a method, system, chip, device and computer readable storage medium for rapidly identifying the producing area of Amomum villosum and evaluating the quality of Amomum villosum based on an electronic nose and an electronic tongue, and belongs to the technical field of traditional Chinese medicine identification. The technical problems to be solved are low detection efficiency and high cost of the existing Amomum villosum identification method. The technical solution points are as follows: a method for rapidly identifying the producing area of Amomum villosum and evaluating the quality of Amomum villosum based on an electronic nose and an electronic tongue, which comprises the following steps: electronic nose test sample preparation and electronic tongue test sample preparation; an electronic nose system adopts a direct headspace air suction method to collect signals of the electronic nose test sample; an electronic tongue system collects signals of the electronic tongue test sample; average values of response values are extracted to form an odor characteristic vector; data analysis and producing area identification / quality evaluation are performed according to the fusion characteristic vector, and finally, model application is performed.

The invention discloses a nitrogen-containing heterocyclic organic compound with room-temperature phosphorescence performance and a preparation method and application thereof, and belongs to the technical field of organic light-emitting materials, the general formula of the nitrogen-containing heterocyclic organic compound with room-temperature phosphorescence performance is # imgabs0 #, and R1, R2 and R3 are understood by referring to the specific description provided by the invention; according to the preparation method, different phenyl substituent groups are introduced through chemical modification, and a series of pyridine derivatives with different phosphorescence efficiencies and phosphorescence lifetime are obtained. The pure organic room-temperature phosphorescent material has wide application potential in the fields of safety anti-counterfeiting encryption, biological imaging and the like, and the organic phosphorescent material is high in phosphorescent efficiency, long in service life, easy to obtain, simple to prepare and fast to separate and purify, and is an efficient phosphorescent material.

The invention relates to the technical field of deep learning, in particular to an AI-based binary homologous sample analysis method. The method comprises the following steps: reading file content, identifying a packing type, and carrying out unpacking processing to obtain an original data file; disassembling the unshelled original data file; the assembly instruction code is converted into an intermediate code, the natural language large model converts the intermediate code into a high-level language for malicious code feature marking, and a malicious sample data set is made; preprocessing a part of malicious samples, and then carrying out label coding; constructing an LSTM model capable of enhancing malicious sample feature extraction capability and detection capability, and inputting a part of processed malicious samples into the LSTM model for training; and inputting the preprocessed new malicious sample into the trained LSTM model so as to obtain a final result. The method has the advantages that the feature relationship in the language is extracted more quickly, and the effect of performing family classification on malicious samples is remarkably improved.

The invention relates to the technical field of underwater target detection, and discloses an underwater target recognition method, device and equipment, comprising: collecting an acoustic target echo signal of a water area to be recognized; preprocessing the acoustic target echo signal to obtain a corresponding bright spot model map and a Mel spectrum map; inputting the bright spot model map into a first feature extractor, using a block segmentation module to segment the bright spot model map into a plurality of blocks, and then merging and position encoding to obtain sequence data; using a shift window transformer to extract features of the sequence data to obtain bright spot model map features; extracting features of the Mel spectrum map through a second feature extractor composed of a plurality of serially connected two-dimensional convolutional layers to obtain Mel spectrum map features; using a block embedding module to fuse the bright spot model map features with the Mel spectrum map features to obtain fused features; inputting the fused features into a classifier, outputting a classification result of whether there is an underwater target, and completing underwater target recognition.

The invention belongs to the field of cassava molecular markers, and particularly relates to a cassava KASP marker and application of the cassava KASP marker in cassava variety distinguishing. The 27 KASP markers provided by the invention have good distinguishing ability in cassava genetic analysis, and can effectively reveal genetic differences among different materials. The DNA fingerprint spectrum of 184 parts of cassava materials is constructed based on 27 core KASP markers, and the purpose of rapid variety identification can be achieved. The KASP marker disclosed by the invention can be used for effectively identifying false selfing offspring. In conclusion, the KASP marker provided by the invention provides a new method and thought for genetic improvement and variety identification of cassava, provides a powerful tool for subsequent cassava breeding and genetic improvement research, and is beneficial to breeding of new varieties with excellent characters.

This invention discloses an image-aware hashing method and system based on residual networks, belonging to the field of information security; it includes S1, using a residual network to perform a first-stage training on an image group to obtain a hash sequence; S2, reading the weights after the completion of the first-stage training, using these weights to perform a second-stage training, and applying a loss function L... f2 By adjusting the network to increase the distance between the original image and the approximate image, the network is ultimately able to distinguish between the original image and the approximate image, thus overcoming the shortcomings of existing technologies.

The present invention discloses a method and device for clustering people who died of cardiovascular and cerebrovascular diseases based on DNN and consensus clustering, which relates to the field of biological data processing, including: obtaining indicators for clustering, performing multivariate COX regression on the indicators, and screening out analysis indicators; constructing and training a population classification model based on a deep neural network to obtain a trained population classification model; inputting the analysis indicators of each individual in the population to be analyzed into the trained population classification model to obtain classification results to determine the surviving population and the population who died of cardiovascular and cerebrovascular diseases in the population to be analyzed; obtaining the analysis indicators of each individual in the population who died of cardiovascular and cerebrovascular diseases through the last hidden layer output of the trained population classification model, clustering the population who died of cardiovascular and cerebrovascular diseases using consensus clustering based on the advanced features, and obtaining clustering results. The present invention solves the problem of difficulty in identifying different subsets of people who died of cardiovascular and cerebrovascular diseases.

The invention provides a heavy metal ion detection method based on an aggregation-induced enhanced fluorescence sensing array. Three ligand-protected AuNCs (GSH-AuNCs, MPA-AuNCs and [2-MBA]-AuNCs) are firstly prepared and are embedded into ZIF-8 through a coordination reaction, different fluorescence intensity adjustment degrees of the three sensing elements by heavy metal ions are used as output signals, and the heavy metal ions are used for detecting the fluorescence intensity of the three sensing elements. According to the method, linear discriminant analysis (LDA) and hierarchical clustering analysis (HCA) are used for analysis to obtain unique spectrum fingerprint data of each heavy metal ion, a database for identifying eight heavy metal ions is successfully established, software SPSS Statistics 27 is used for processing the obtained data, the unique spectrum fingerprint data of each heavy metal ion is obtained, and the spectrum fingerprint data of each heavy metal ion is obtained. Therefore, accurate identification and sensitive quantitative detection of eight heavy metal ions (Ni < 2 + >, Cr < 2 + >, Co < 2 + >, Pb < 2 + >, Cd < 2 + >, Ag < + >, Cu < 2 + > and Zn < 2 + >) are realized.

The present invention discloses a sound data classification method, terminal and medium based on multi-dimensional feature weighted fusion, the method comprising the following steps: S1, pre-processing the input sound data to obtain sound segment data; S2, extracting static and dynamic combined features J from the sound segment data; S3, extracting initial input features K from the sound segment data; S4, processing the static and dynamic combined features J and the initial input features K through a fully connected neural network to obtain an attention weight vector H; S5, calculating the final input feature G, G=J*H+K*(1-H); S6, inputting the final input feature G into an LSTM network to obtain a category m of sound data classification. The present invention links the front and back frames together by adding dynamic differential features, and at the same time, redistributes the resources evenly distributed on different dimensions of the fused features according to the attention weight, so as to achieve the purpose of "learning from each other's strengths and making up for each other's weaknesses", making the fused features more effective and more distinguishable, thereby improving the accuracy of classification.

This invention relates to the field of biomedical technology and proposes RBP4 and Ferritin as biomarkers for evaluating the efficacy of hormone therapy for allergic rhinitis. The biomarkers include one or a combination of RBP4 and Ferritin. Studies have confirmed that both RBP4 and Ferritin are capable of distinguishing between hormone-sensitive and hormone-insensitive allergic rhinitis. RBP4 is expressed at higher levels in the hormone-sensitive allergic rhinitis group, while Ferritin is expressed at higher levels in the hormone-insensitive allergic rhinitis group. Compared to single biomarker detection, the combined use of RBP4 and Ferritin significantly improves the differential diagnostic efficacy, providing better differentiation between the two patient groups.

The present invention discloses a feature extraction method and device for multi-feature fusion of converter transformer vibration signals. The method selects the on-site measured signal of the converter transformer as the analysis object, and performs a collective empirical mode decomposition on it to extract the intrinsic mode. Then, the time domain statistical characteristic components, wavelet entropy and fractal dimension of each intrinsic mode component are calculated to form a characteristic parameter matrix. Finally, the matrix is ​​subjected to data dimension reduction through principal component analysis to form a multi-feature fusion vector, and the feature vectors of different operating conditions of the converter transformer are classified using fuzzy C-means clustering. This method overcomes the limitations of a single feature and has good discrimination performance under different operating conditions.

The invention provides a brocade pattern generation method and system based on a diffusion model, and relates to the technical field of pattern generation, and the method comprises the steps: collecting a brocade sample image, extracting the structural features of a brocade sample pattern, and building a multi-mode joint embedding space; analyzing a design intention input by a user into a multi-modal condition vector, and mapping the multi-modal condition vector to a multi-modal joint embedding space to obtain condition embedding features; introducing a structure perception attention mechanism into the noise prediction network of the diffusion model, generating a structure constraint feature map, and in the reverse denoising process of the diffusion model, carrying out hierarchical fusion on the condition embedding features and the structure constraint feature map to obtain a brocade generation result; and performing brocade process feasibility verification on the brocade generation result, and outputting a brocade pattern. According to the method, the intelligent level and the production efficiency of brocade pattern generation can be improved by combining a structure perception attention mechanism and a hierarchical fusion strategy.

The invention relates to the technical field of deep learning, and in particular to an AI-based binary homology sample analysis method; the method comprises reading file content, identifying a shell type, and then performing shell removal processing to obtain an original data file; disassembling the shelled original data file; converting assembly instruction codes into intermediate codes, and a natural language large model converting the intermediate codes into high-level languages ​​for malicious code feature marking to produce a malicious sample data set; preprocessing some malicious samples, and then performing label encoding; constructing an LSTM model capable of enhancing malicious sample feature extraction and detection capabilities, and inputting some processed malicious samples into the LSTM model for training; and inputting the preprocessed new malicious samples into the trained LSTM model to obtain a final result; the advantages of the invention are that feature relations in a language can be extracted more quickly, and the effect of performing family classification on malicious samples can be significantly improved.

This application discloses a method for identifying and classifying furniture surface defects based on image detection. The method includes: acquiring an original image of the furniture surface; converting the original image into luminance and chrominance components to obtain a standard luminance map; generating a candidate region set based on the standard luminance map and chrominance components; calculating the structure tensor based on the standard luminance map using overlapping local windows to determine the local principal direction and directional validity, and constructing a pixel-level principal direction field and directional validity map; further performing directional-guided background reconstruction on the standard luminance map based on the principal direction field and directional validity map to obtain a background reconstruction map and generate a residual response map; finally, extracting texture interpretation features and defect interpretation features for each candidate region, inputting them into a competition suppression model for discrimination, determining the defect category of the retained regions, and outputting the defect location and defect category. This invention is applicable to the identification and classification of furniture surface defects with directional textures.

The invention discloses a feature extraction method based on improved multi-scale fractional order approximate entropy, and relates to the technical field of nonlinear system feature extraction. Comprising the following steps: S1, data acquisition; s2, a data preprocessing step; s3, solving the improved multi-scale fractional order approximate entropy; and S4, a visualization step. According to the method, the multi-scale fractional order approximate entropy is improved, so that the accuracy of multi-scale entropy calculation is improved while the obtained information is more detailed.

The invention discloses a refrigerator oil wear resistance detection device, which comprises a pressure experiment cabin, the pressure experiment cabin is connected with one end of a three-way pipeline through a pipeline, and the other two ends of the three-way pipeline are respectively connected with a vacuum system and a coolant system. A vacuum valve, a refrigerant valve and a constant pressure valve are correspondingly arranged on the pipeline between the vacuum system, the coolant system and the pressure experiment bin and the three-way pipeline, and an agent inlet valve is arranged on the pipeline between the pressure experiment bin and the constant pressure valve. The method comprises the following steps: setting the vacuum degree and temperature of a pressure experiment bin, injecting a refrigerant into the pressure experiment bin, driving a ring block to rotate through a main shaft, enabling the ring block to be in contact friction with a test block, and further judging the lubricating performance of the test oil according to the width and depth of a grinding crack of the test block and the weight loss before and after friction. And the method has good discrimination.

The present invention relates to the field of environmental traceability technology, and specifically to an environmental traceability tracking system for a Juncao production process, comprising a data acquisition module, a data analysis module and a traceability tracking module. The data acquisition module is used to collect environmental data and quality data of Juncao. The data analysis module is used to construct an abnormal data analysis model of an isolated forest algorithm optimized by a wolf pack algorithm, and to construct an SVM-LSTM network model optimized by a dung beetle algorithm. The traceability tracking module is used to mark the Juncao production process, and to input abnormal data into the Juncao traceability analysis model, and to trace the source of the abnormal data. The present invention uses an isolated forest algorithm optimized by a wolf pack algorithm to analyze abnormal data, has better distinguishing ability, is more adaptable, and improves the accuracy of anomaly detection. At the same time, it combines the SVM-LSTM network model optimized by the dung beetle algorithm to perform environmental traceability tracking, which speeds up the convergence of the model, makes the training process more efficient, and improves the overall operating efficiency.

The invention relates to application of a diagnostic marker combination in construction of a prognosis risk assessment model of esophageal squamous carcinoma. The diagnostic marker combination comprises SLC2A1, KIF11, KCNMA1, BARHL2, CA9, TIMP1, MAGEC3 and PDZK1IP1. The prognosis of the esophageal squamous carcinoma patient is predicted on the basis of detecting the diagnostic marker combination and a corresponding esophageal squamous carcinoma prognosis risk assessment model, effective guidance is provided for a clinician to make a treatment decision on the esophageal squamous carcinoma patient, invalid treatment is reduced, the treatment cost and uncomfortable experience of the patient are reduced, and the clinical application prospect is wide. And accurate prediction and individualized comprehensive treatment of prognosis of the esophageal squamous cell carcinoma patient can be promoted.

The present invention belongs to the field of biomedicine technology, and specifically discloses a marker for screening for cardia cancer. The marker for screening for cardia cancer provided by the present invention is 2-Hydroxy-2-methylbutyric acid, and the detection reagent of the marker can be used to prepare a cardia cancer screening product. By detecting the expression level of 2-Hydroxy-2-methylbutyric acid in human serum, cardia cancer can be effectively detected, which can be used for large-scale screening of asymptomatic people in areas with a high incidence of cardia cancer, which is conducive to screening and early detection of asymptomatic high-risk people.

The invention discloses a rapid probucol crystal form identification method which comprises the following steps: step 1, taking potassium bromide, drying, and pressing a potassium bromide blank sheet; 2, adding probucol crystal powder to be detected, lightly grinding and uniformly mixing the probucol crystal powder and potassium bromide powder for not more than 2 minutes, and tabletting; step 3, acquiring an infrared spectrum; and comparing with a probucol crystal form infrared spectrum obtained in advance to obtain the crystal form of the probucol crystal to be detected. The inventor finds that infrared spectrum is adopted to detect the crystal form I or the crystal form II of a single probucol crystal, obtained spectrums are obviously different in three positions, and the two crystal forms can be distinguished. Potassium bromide and probucol crystals are mixed and ground, the crystal form I is partially converted into the crystal form II, and the infrared spectrum is different from that of the two crystals and can be distinguished. Based on the discovery, the invention provides two crystal forms of the probucol crystal and a standard infrared spectrum of a mixture of the two crystal forms, so that the probucol crystal forms can be quickly identified.