310 results about "Reference dataset" patented technology

The invention relates to a computer implemented method for defect detection comprising: obtaining an imaging dataset and a reference dataset of an object comprising integrated circuit patterns; and detecting defects in the imaging dataset using the imaging dataset and the reference dataset, wherein a machine learning model for defect highlighting is applied to the imaging dataset as input and generates a highlighted defect dataset as output, and wherein the machine learning model for defect highlighting comprises at least one attention mechanism. The invention also relates to computer programs, computer-readable media and corresponding systems.

An example method includes (a) obtaining an object detection from a perception system that describes an object in an environment of the autonomous vehicle; (b) obtaining, from a reference dataset, a label that describes a reference position of the object in the environment; (c) determining a plurality of component divergence values respectively for a plurality of divergence metrics, wherein a respective divergence value characterizes a respective difference between the object detection and the label; (d) providing the plurality of component divergence values to a machine-learned model to generate a score that indicates an aggregate divergence between the object detection and the label, wherein the machine-learned model includes a plurality of learned parameters defining an influence of the plurality of component divergence values on the score; (e) evaluating a quality of a match between the object detection and the label based on the score.

The invention belongs to the field of data quality analysis, and particularly discloses a data set quality analysis and evaluation method, which performs integrity evaluation on a current data set, starts data confidence analysis only after the number of data groups reaches the standard, ensures that confidence evaluation is established on the basis of data with a complete structure, avoids erroneous judgment caused by data missing, and improves the quality of the data set. Meanwhile, during data confidence analysis, dynamic comparison between current data and a local time sequence neighborhood is implemented by constructing an adjacent data set of each data group; meanwhile, a historical reference data set is combined, longitudinal comparison with a historical fluctuation range is carried out, and dual verification of local consistency and historical comparability is fused, so that the anomaly detection sensitivity and the evaluation accuracy are improved to a great extent; and finally, establishing a closed-loop feedback path with an acquisition end when availability judgment is carried out based on a data set integrity and confidence evaluation result, and realizing self-adaptive adjustment of an acquisition strategy according to a risk positioning result.

System and methods are disclosed for matching extracted text data based on one or more similarity scores. The method may include receiving one or more documents from a plurality of data sources, utilizing an optical character recognition algorithm for extracting text data from the one or more documents, comparing, utilizing a fuzzy matching algorithm, the extracted text data to reference dataset(s) to determine one or more matches between the extracted text data and at least one of the reference dataset(s), wherein the one or more matches are based on at least one similarity score, inputting the determined one or more matches and the at least one similarity score into a trained machine-learning model to refine the one or more matches, and outputting a representation of the refined one or more matches and the at least one similarity score to a graphical user interface of a device.

The invention relates to a computer implemented method for defect detection in an object comprising integrated circuit patterns comprising: obtaining an imaging dataset and a reference dataset of the object; generating an input representation of a subset of the imaging dataset and a reference representation of a corresponding subset of the reference dataset in a feature space; and detecting defects in the object by comparing the input representation to the reference representation in the feature space. The invention also relates to a corresponding computer-readable medium, computer program product and system for defect detection.

Embodiments of the present disclosure relate to a method, an electronic device, and a computer program product for dataset updating. The method includes: acquiring a reference dataset comprising a plurality of reference images, wherein each reference image has a soft label, and the soft label indicates a probability that the each reference image belongs to each target class in a plurality of target classes. The method may further include: determining, according to a similarity between an input image received and the each reference image, an image soft label for the input image. The method further includes: adding the input image and the image soft label for the input image to the reference dataset to update the reference dataset.

The invention relates to a method for defect detection comprising: acquiring an imaging dataset of an object comprising integrated circuit patterns using an imaging system; obtaining a reference dataset corresponding to the acquired imaging dataset; generating an enhanced imaging dataset by filtering the acquired imaging dataset with one or more learned filters, wherein the one or more learned filters are obtained by solving an optimization problem comprising the deviation of the enhanced imaging dataset from the reference dataset; and detecting defects in the acquired imaging dataset by comparing the enhanced imaging dataset to the corresponding reference dataset. The invention also relates to a corresponding computer program, a computer-readable medium and a system for defect detection in objects comprising integrated circuit patterns.

A method for generating a segmentation mask of at least one image comprising generating a plurality of superpixels for the at least one image, automatically generating labels for the generated plurality of the superpixels, wherein the automatically generating labels comprises generating the labels for training a semantic segmentation model based on a plurality of segmentation masks by generating a label for each superpixel by identifying the most similar reference superpixel from a reference data set of reference superpixels, wherein each reference superpixel is associated with a class for supervised training of the semantic segmentation model, wherein the labeled superpixels form the segmentation mask of the at least one image, and computer-aided checking of the generated segmentation mask for correctness, wherein the checking for correctness comprises labeling unlabeled superpixels and correcting labels of incorrectly labeled superpixels by assigning the incorrectly labeled superpixels to the correct class.

The invention discloses a construction method of a TiH binary system machine learning potential. The construction method comprises the following steps: constructing a reference electronic structure data set by adopting first principle calculation software VASP; training and testing an artificial neural network potential-ANN potential model on the reference data set by using the artificial neural network model; vASP and ANN potential obtained through training are used for calculating and comparing a Dimer curve, a lattice constant, an elastic modulus and the like for performance testing; and performing targeted improvement on data with large comparison error between the ANN model and the DFT result in model test and performance test results by increasing related configuration weights until an ANN potential model conforming to the precision is obtained. The machine learning potential model constructed by the invention can better fit the interatomic interaction potential energy surface of the TiH binary system, solves the problems of small time-space scale and low precision of the traditional empirical potential in the calculation of the first principle, and provides a technical basis for the high-precision and large-size simulation and research of the TiH binary system.

The invention provides a power transmission line hardware service life evaluation method and system based on an aeolian vibration finite element, and relates to the technical field of power transmission line health evaluation, and the method comprises the steps: carrying out the element division of a fastening hardware finite element model based on a element division standard; applying the aeolian vibration pre-test scheme to the finite element model of the fastening metal, and obtaining a pre-test reference data set; obtaining a pre-test reference result based on the pre-test reference data set; a life diagnosis breeze characteristic limit stored in a database is matched based on a pre-test reference result; aeolian vibration environment parameters of the fastening fitting are obtained, aeolian vibration environment characteristic values are calculated, and a fastening fitting service life diagnosis curve is matched based on the service life diagnosis aeolian characteristic limit and the aeolian vibration environment characteristic values; acquiring stress characteristic parameters of each unit element based on the finite element model of the fastening fitting, acquiring a stress characteristic value of the unit element based on the stress characteristic parameters, and acquiring a service life diagnostic value of the fastening fitting based on the stress characteristic value of the unit element and the service life diagnostic curve of the fastening fitting.

The invention discloses an intelligent war game deduction decision-making method based on sample migration, and belongs to the technical field of artificial intelligence and military deduction games. The method comprises the following steps: determining environment information of wargame deduction migration, and constructing a reference data set; constructing a migration network model based on a CycleGAN network framework, training the migration network model by using the reference data set, and generating migration data by using the trained migration network model; constructing a soft constraint condition according to the migration data; under the constraint of the soft constraint condition, the migration data and exploration data in the current combat scene state are utilized to train the agent model, and a trained agent model is obtained; the trained agent model is used for intelligent wargame deduction decision making. The method has higher sample migration similarity and knowledge transitivity, meanwhile, the problem that an existing method is poor in training effect under non-ideal data is solved, and the robustness of an algorithm is improved.

The invention discloses an animal single cell data cell type annotation method, which is characterized by comprising the following steps: collecting Bulk RNA-seq data of purified cell types in various tissues and organs of a specific animal, and integrating the Bulk RNA-seq data into a reference data set after preprocessing; the method comprises the following steps: acquiring single-cell RNA-seq original data, screening the single-cell RNA-seq original data to obtain high-quality cells, screening high-variation genes from the high-quality cells, processing the high-variation genes, extracting principal components, performing dimensionality reduction on the principal components, and performing cell clustering based on a dimensionality reduction result to obtain cell clusters; and calculating expression similarity between the cell clusters and the reference data set based on the high-variation genes, determining initial cell types of the cell clusters according to a similarity result, and carrying out iterative tuning on the initial cell types with similar scores to obtain a final cell type annotation result. The method provides efficient and accurate technical support for animal single cell research.

Various embodiments of the present disclosure provide prompt engineering and iterative, feedback-based generative techniques that improve traditional LLM technology, including extractive LLM techniques. The techniques may include selecting one or more simple annotated question-answer pairs for an input data object comprising an input question and an input document from a reference dataset. The techniques may include selecting one or more complex annotated question-answer pairs from the reference dataset. The techniques may include generating a few-shot prompt based on the one or more simple annotated question-answer pairs and the one or more complex annotated question-answer pairs. The techniques may include providing the few-shot prompt to a large language model (LLM) to receive a predictive output for the input question.

The invention discloses a multi-dimensional large model data quality evaluation method and system, and the method comprises the steps: extracting data features in a data set and a reference data set which meet the lowest inherent quality requirements, and carrying out the evaluation of the quality of data based on the information field effect between data samples in the data set; and feature updating and context semantic association enhancement processing are carried out on each data sample, so that the semantic feature expression ability of each data sample is improved, and more accurate data quality evaluation is realized. Therefore, by introducing the context association information of the data set, the semantic information of each data sample can be more accurately understood, so that the accuracy of overall data quality evaluation is improved.

Methods and systems are provided for implementing training of learning models, including obtaining a pre-trained weight set for a learning model on a sample dataset and on a first loss function; selecting at least two tasks having heterogeneous features to be computed by a reference model; obtaining a reference dataset for the at least two tasks; designating a second loss function for feature embedding between the heterogeneous features of the at least two tasks; training the learning model on the first loss function and training the reference model on the second loss function, in turn; and updating the weight set based on a feature embedding learned by the learning model and a feature embedding learned by the reference model, in turn. Methods and systems of the present disclosure may alleviate computational overhead incurred by executing the learning model and loading different weight sets at a central network of the model.

Aspects of the subject disclosure may include, for example, obtaining a reference data set comprising a plurality of reference data records, wherein the reference data set is associated with a cloud infrastructure, and wherein each of the reference data records comprises a reference network name associated with one of a plurality of reference networks that is configured within the cloud infrastructure and a corresponding reference Virtual Local Area Network (VLAN) identifier for the one of the plurality of reference networks that is configured within the cloud infrastructure; obtaining an input data set comprising a plurality of input data records, wherein the input data set is associated with the cloud infrastructure, and wherein each of the input data records comprises a network name associated with one of a plurality of networks that is being considered for configuration within the cloud infrastructure and a corresponding VLAN identifier for the one of the plurality of networks that is being considered for configuration within the cloud infrastructure; analyzing the reference data records and the input data records in order to identify a set of one or more errors; and tagging one or more records in the input data set with each corresponding occurrence of each error. Other embodiments are disclosed.

A simulated environment is generated and adapted through interaction with a user. Prior performance metrics indicating behavior of a student in performing a prior learning task are obtained. A student profile is updated based on the prior performance metrics. Lesson parameters indicating content to be included in an interactive lesson are obtained. Rules for the interactive lesson are then generated based on the student profile and the lesson parameters. The rules are applied to a classifier trained via a reference data set representing prior performance of a student population. Via the classifier, instructions for generating a simulated environment encompassing the interactive lesson are generated. A representation of the simulated environment is then generated based on the instructions.

A method and system for quantifying semantic variance between neural network representations is provided. Two neural network representations to be compared are first extracted, the weight of each filter in an intermediate layer corresponding to each semantic concept is learned on a reference dataset using the Net2Vec method, then set IoU of each representation for all semantic concepts in the reference dataset are calculated, and finally variance between the set IoU of the two representations for all the semantic concepts are integrated to obtain semantic variance between the two neural network representations. The method solves the problem of lack of accurate measurement on the variance between neural network representations on a semantic information level, and has an accurate measurement effect.

A method includes obtaining matches between target records in a target dataset and a reference records in a reference dataset, each match of the matches comprising a corresponding confidence level of the match, categorizing the target records into review level categories according to the corresponding confidence level, and presenting a graphical user interface (GUI). The GUI includes a first section for a first review level category showing a first subset of the target records assigned to the first review level category, the first subset comprising target records related, in the GUI, to at least one matching reference record. The GUI includes a second section for a second review level category, wherein the second section shows a second subset of the target records assigned to the second review level category, the second subset comprising target records related, in the GUI, to at least one matching reference record.

The invention relates to a method for predicting the lifespan of an individual, comprising I. providing biological input data of the subject, ii. providing a list of confounding variables, iii. predicting an age-related mortality or disease, preferably the time to death and / or the mortality risk and / or the risk of age-related disease, for the subject by analyzing the data with an algorithm, wherein the algorithm comprises a neural network, which is trained on at least one reference dataset comprising biological data of at least one reference subject by applying: a) a selector layer to filter input data of i., and b) an adversarial learning framework, that removes from the input data the information related to the confounding variables, wherein preferably the adversarial learning framework, is a neural network comprising three elements: a feature extractor (FE), a predictor (P), and a confounder predictor (C). The invention further relates to a device or system comprising means for carrying out the steps of the method according to the invention, a computer program and a computer-readable storage medium.

The invention relates to a computer implemented method for defect detection comprising: obtaining an imaging dataset of an object comprising integrated circuit patterns; obtaining a reference dataset of the object; registering the imaging dataset and the reference dataset by obtaining at least one transformation field pair comprising an input transformation field and a corresponding reference transformation field, wherein the input transformation field or the reference transformation field can be zero; and detecting defects in the imaging dataset using the at least one obtained transformation field pair. The invention also relates to a computer-readable medium, a computer program product and a system for detecting defects.

Systems and methods are provided to generate improved sets of reference data that are ML model-agnostic. The system initiates an imbalance analysis on a training dataset (e.g., text, image, time series, etc.) that includes determining a set of classes in the data. Using the set of classes, the system processes mutual information (MI) across the data segments to generate a set of matrices from extracted partition-level mutual information. In some examples, the system may generate baseline reference data from the set of matrices and provide the baseline reference data for implementation with anomaly detection or model explainability in external machine learning (ML) models.

The invention discloses a building component autonomous inspection system and method based on a BIM (Building Information Modeling) and a VLM (Visual Language Model), and relates to the technical field of intelligent inspection of building engineering. Generating a spatial reference data set, a semantic tag set and a component semantic embedding vector; the method comprises the following steps: collecting multi-modal visual and spatial data, carrying out semantic segmentation and description generation by utilizing a visual language large model, calculating a semantic ambiguity coefficient and carrying out confidence evaluation; through BIM and visual semantic embedding vector comparison and spatial deviation analysis, calculating a cross-modal consistency index and generating a deviation report; calculating a comprehensive path cost coefficient and performing path optimization; and in combination with semantic recognition, cross-modal consistency and path cost, obtaining inspection performance evaluation indexes, realizing adaptive optimization through reinforcement learning, and writing back to a BIM database to realize continuous improvement of data and decision ability.

The invention relates to a method for diagnosing damage to a vehicle component generating airborne and / or structure-borne noise emissions, wherein data about these emissions are used for diagnosis, in which in a first method step (M1) - the measured airborne and / or structure-borne noise emissions are converted into airborne and / or structure-borne noise data, - at the same time, the current operating situation of the vehicle is determined by recording operating situation data, - and the detected data is stored in a computer, in which in a second method step (M2) - the airborne and / or structure-borne sound data is transformed using mathematical methods in such a way that the first transformed airborne and / or structure-borne sound data generated in this way can be analysed and evaluated, and in which in a third method step (M3) a normalisation process is carried out, - in which a reference data set associated with the current operating situation data is determined, which contains such transformed reference airborne and / or structure-borne noise data that would be formed in this operating situation for a vehicle without damage, - in which this reference data set is compared with the first transformed airborne and / or structure-borne sound data to generate normalised transformed airborne and / or structure-borne sound data, and - in which a detected difference gives an indication of airborne and / or structure-borne noise emitted by a defective vehicle component. The invention also relates to a diagnosis system, with which the method can be operated.

The invention provides a fault judgment method of a wind driven generator and computer equipment. The fault judgment method comprises the following steps: acquiring a first sample data set of the wind driven generator of a to-be-judged wind driven generator set; based on the data distribution of each piece of first sample data in the first sample data set, screening the first sample data set to obtain a second sample data set; comparing the second sample data set with a reference data set to obtain a fault discrimination condition for discriminating the fault of the wind driven generator; and based on the operation data of the wind driven generator and the fault discrimination condition, carrying out fault discrimination on the wind driven generator. The problem that efficient and accurate fault judgment of a lightweight algorithm is difficult to realize is solved, fault judgment can be carried out on the wind driven generator by screening and optimizing sample data and flexibly determining fault judgment conditions, and efficient and accurate fault judgment can be carried out through the lightweight algorithm.

A three-dimensional (3D) mapping system can be configured to generate a 3D map of a real-world environment using annotation of large image data sets, in which terrestrial imagery can be programmatically labeled with accurate labels using remotely sensed overhead image data. The 3D mapping system can implement photogrammetry to create a point cloud. Each pixel in the point cloud can be classified based on a consensus of each frame. The point cloud can be co-registered to a remotely sensed reference dataset to provide precise spatial coordinates for each pixel. Different patches of point clouds can be stitched together to provide a complete 3D map for a given area, such as a downtown area of a city.

An example method includes (a) obtaining an object detection from a perception system that describes an object in an environment of the autonomous vehicle; (b) obtaining, from a reference dataset, a label that describes a reference position of the object in the environment; (c) determining a plurality of component divergence values respectively for a plurality of divergence metrics, wherein a respective divergence value characterizes a respective difference between the object detection and the label; (d) providing the plurality of component divergence values to a machine-learned model to generate a score that indicates an aggregate divergence between the object detection and the label, wherein the machine-learned model includes a plurality of learned parameters defining an influence of the plurality of component divergence values on the score; (e) evaluating a quality of a match between the object detection and the label based on the score.

The invention relates to the field of nuclear power safe operation, in particular to a lightning arrester resistive current data classification method and a lightning arrester working state judgment method. The classification method comprises the following steps: acquiring humidity data and resistive current data of the lightning arrester to form an original data set; improving a bandwidth formula, generating B resistive current reference data sets according to a KDE kernel density estimation function, and forming a reference data set by the resistive current reference data sets and the humidity data; calculating the sum of squares of classification distance errors of an original data set of all data sets and a reference data set generated after processing under different clustering center numbers k, calculating Gap statistics, and selecting an optimal Kbest value; and performing K-means clustering analysis on the original data set based on the optimal Kbest value, and dividing the original data set into a rain class and a sunny class. The method is accurate in classification, effectively improves the online monitoring precision of the lightning arrester, avoids the misjudgment of the working state of the lightning arrester, and achieves the stable operation of a power system.

The invention relates to the technical field of power distribution, in particular to a power distribution method of a machine room power supply system, which comprises the following steps: acquiring a reference data set of parameter items of each acquisition moment in an operation data time sequence of bus connection equipment, the parameter items comprising current and / or voltage; calculating the tendency and prediction deviation of the parameter item at each acquisition moment, and normalizing the product of the prediction deviation of the parameter item, a preset error factor and the tendency to obtain an error factor of the parameter item; in the HTFE algorithm, the parameter item predicted values at the next acquisition moment are obtained by using the error factors, and the power required by the equipment at the next acquisition moment is obtained through the product of the parameter item predicted values, so that power distribution is realized, and the power distribution accuracy of the machine room power supply system is effectively improved.

The invention relates to the technical field of riding route planning, and particularly provides a riding route AI planning method based on user demands, which comprises the following steps: receiving user riding preference information and basic travel information, and carrying out integration quantitative preprocessing to form a structured user demand data set; acquiring multi-source reference data such as preset route database data and real-time dynamic data, and performing fusion preprocessing to generate an adaptive reference data set; inputting the two types of data sets into a trained AI planning model to calculate an optimal path, and embedding adaptive point locations based on preference constraints to generate an initial path; and performing multi-dimensional adaptability optimization by combining basic constraints, generating a personalized scheme with multiple versions, setting an output priority, completing route optimization according to a user interaction instruction, and outputting a final executable riding route. The method deeply adapts to multiple demands of the user, improves the planning accuracy and flexibility, and provides a personalized riding route service giving consideration to experience and practicability for the user.