298 results about "Sample Label" patented technology

The invention belongs to the technical field of aero-engine state monitoring and intelligent fault diagnosis, and discloses a multi-modal information fusion bearing fault diagnosis method based on self-supervised learning. The method comprises the following steps: firstly, through mask reconstruction self-supervision pre-training, extracting stable feature representation insensitive to mask disturbance from an unlabeled multi-modal signal, and dynamically updating each modal feature reference point by using an index moving average algorithm; in a downstream fault diagnosis task, a multi-modal joint decision model comprising a pre-training encoder, a single-modal classifier and a fusion classifier is constructed, and adaptive weighted fusion of multi-modal decision is realized through contribution degree calculation based on a cooperative game Shapley value in combination with a deviation degree of modal features and a reference point. According to the method, the dependence of the deep neural network on fault labeling data is effectively reduced, the accuracy and robustness of the diagnosis system in a multi-modal signal diagnosis scene are improved through a dynamic fusion mechanism, and the method is suitable for industrial scenes with limited sample label resources.

The embodiment of the invention provides a model training method, a facial expression recognition method and device and electronic equipment, and relates to the technical field of video processing. The model training method comprises the following steps: acquiring a sample video and a first sample label; extracting a time feature and a space feature of the sample video by using a time-space feature extraction network in the facial expression recognition model of the initial structure; calculating an attention weight representing the correlation between the spatial feature and the time feature by using a mapping network; performing weighted aggregation on the time features by using the attention weight to obtain fused spatio-temporal features; inputting the fused spatial-temporal features into a classification network to obtain a first recognition result; and performing model training based on the difference between the first recognition result and the first sample label to obtain a trained facial expression recognition model with higher accuracy.

The invention relates to an intelligent agent continuous training and effect evaluation closed-loop method and system fusing work order feedback, and a medium, and relates to the technical field of deep learning. The agent continuous training and effect evaluation closed-loop method comprises the following steps: receiving and analyzing work order feedback information, and recording the number of error information and information storage duration in combination with a preset structure sample label; dynamically judging the number of error information and the information storage duration, triggering a retraining rule of an original agent model, distributing candidate information weights, and screening a target sample set; directionally training and optimizing the original agent model according to a retraining rule in combination with the target sample set, and generating a new agent model; constructing a historical work order problem set, comparing and verifying a new agent model in combination with an original agent model, generating an agent evaluation table, and correcting a candidate information weight; by accurately focusing the weak link, the continuous and efficient enhancement of the model capability is realized, so that the answer accuracy of the intelligent agent is remarkably improved, and the AI illusion is effectively inhibited.

The invention relates to the technical field of crop remote sensing samples, and discloses a crop remote sensing image intelligent sample library construction method based on a knowledge graph. The method comprises the steps of collecting multi-source remote sensing image data, and generating an enhanced image data set through fusion and enhancement processing; and constructing a crop field knowledge graph integrating the crop growth model, the soil type and the climate condition. A knowledge graph is used as guidance, crop areas are recognized through semantic analysis, multi-scale features are extracted, and category attributes and associated contexts are deduced through a graph reasoning algorithm in combination with a relation path; and generating a sample label candidate set through deep learning of the graph attention network, screening and correcting the sample label candidate set by using an optimization algorithm constrained by the knowledge graph, and outputting high-quality sample labels. And integrating and storing labels and image data into a graph structure sample library, implementing a dynamic updating mechanism to adapt to new data, running a quality monitoring process driven by a knowledge graph, and adjusting a construction strategy according to result feedback to realize intelligent and efficient construction of the sample library.

The invention relates to the technical field of remote sensing information, and discloses a corn fine classification method and system based on multi-source time sequence remote sensing image feature fusion, and the method comprises the steps: obtaining a multi-source remote sensing image and digital elevation model data of a target region; arranging corn and non-corn sample points in the target area, calculating multi-temporal vegetation indexes, texture features and gradient features based on the preprocessed remote sensing image, and constructing time sequence features; carrying out feature selection on the constructed time sequence features by utilizing importance screening and high-correlation feature redundancy elimination to obtain a feature subset; fusing the final feature subset with the gradient features to form a final optimized feature vector of each sample point; and adopting a Boosting ensemble learning algorithm, taking the optimized feature vector and the sample label as input, training a corn classification model, and predicting classification. The method has the advantage that high-precision automatic identification and classification of corn crops in a large-scale farmland can be realized.

The present application relates to the technical field of sample pretreatment automation, and discloses a system of a full-automatic sampling tube cap screwing and pipetting workstation and a running method thereof.The system comprises a motion point definition module, a cap screwing adaptation identification module, a label scanning module, a pipetting planning and execution module, a cap screwing torque calling module and a real-time torque acquisition module.The system realizes continuous scanning and reliable identification of the cap profile in the motion state through a dynamic coordinate system and key motion points, dynamically calls matched torque parameters to execute cap screwing according to the identification result, and simultaneously acquires real-time torque data.The system synchronously completes sample label information analysis, pipetting path planning and pressure monitoring.The scheme realizes adaptive perception and closed-loop cap screwing control of the cap state, and improves the robustness of cap identification, the consistency of cap screwing action and the reliability of the overall process.

The invention relates to a method and a system for identifying a slow-dip-angle structural plane of a pumped storage upper reservoir, and the method comprises the following steps: selecting a plurality of surveyed pumped storage power station upper reservoir regions as sample regions, and obtaining the exploration sample data of the sample regions; gridding the sample area into a three-dimensional voxel grid, preprocessing exploration sample data, extracting exploration features of each voxel in the three-dimensional voxel grid, adding a sample label indicating whether the voxel belongs to a low-dip-angle structural plane, and putting the sample label into a training sample set; constructing a recognition model based on a machine learning model, and performing iterative training on the recognition model by using the training sample set to obtain a voxel low-dip-angle structural plane probability recognition model; and acquiring exploration data of a reservoir region on the pumped storage power station to be identified, acquiring a prediction probability of whether each voxel in the region belongs to a low-dip-angle structural plane or not by using the voxel low-dip-angle structural plane probability identification model, and further identifying the low-dip-angle structural plane in a target region.

The invention discloses an automatic garbage classification method and system based on machine learning, and relates to the technical field of environmental protection, and the method comprises the steps: carrying out the multi-physical field dynamic excitation and response data collection of garbage articles, obtaining multi-modal feature data, and carrying out the environment disturbance correction of the multi-modal feature data; inputting the corrected multi-modal feature data into a feature decoupling network constrained by a physical mechanism, and decoupling and outputting essential attribute feature vectors and representation attribute feature vectors; inputting the essential attribute feature vector into a classification decision unit, and outputting a classification result, a confidence score and an uncertain sample mark based on a multi-level classification decision of essential attributes; and for the samples indicated by the uncertain sample marks, starting a physical verification-oriented active knowledge acquisition mechanism, acquiring classification labels and updating a feature database. According to the invention, the accuracy and adaptive evolution capability of garbage classification in a complex real scene are improved.

The application belongs to the technical field of power grid transient stability evaluation, and specifically discloses a boundary sample enhancement method and system for power grid transient stability evaluation, which comprises the following steps: training a transient stability evaluation model by using an initial training sample set, predicting the initial training sample by using the trained transient stability evaluation model, and screening boundary samples based on the obtained prediction probability; calculating the local density of all non-boundary stable samples, and performing undersampling on the non-boundary samples; training a mask autoencoder generative adversarial network by using the screened boundary samples, generating new boundary samples without labels, obtaining sample labels by using a time domain simulation technology, and adding the new samples that pass the test to the training sample set after undersampling to realize boundary sample enhancement. The application can effectively identify the samples near the classification decision boundary of the transient stability evaluation model, and provides reliable training data basis for boundary sample generation and enhancement.

Embodiments of the present application provide a quantum neural network classifier training method and device, electronic equipment and medium. The scheme is as follows: obtaining a training data set and a to-be-trained classifier; for each training sample data, classifying the training sample data by using the to-be-trained classifier to obtain a first predicted label; calculating a first loss value of the to-be-trained classifier according to a sample label corresponding to each training sample data and the first predicted label; when the to-be-trained classifier has not converged, adjusting the classifier parameters based on the first loss value, and returning to execute the step of classifying each training sample data by using the to-be-trained classifier to obtain the first predicted label corresponding to the training sample data until the to-be-trained classifier converges at the current time. Through the technical scheme provided by the embodiments of the present application, the optimization of the quantum neural network classifier is realized, and the classification accuracy and attack resistance of the quantum neural network classifier are improved.

The invention discloses a training sample construction method and device for an abnormal transaction detection model, a medium and a product. The method can be applied to the fields of artificial intelligence, block chains and financial science and technology, and comprises the steps of obtaining a first transaction feature of a first transaction node and a second transaction feature of a second transaction node in a block chain transaction network; the first transaction node is a transaction node labeled with a sample label; the second transaction node is a transaction node which is not labeled with a sample label; generating a sample label of the second transaction node based on the sample label of the first transaction node according to the first transaction feature of the first transaction node and the second transaction feature of the second transaction node; and taking the first transaction node and the second transaction node marked with the sample labels as a sample training data set. According to the technical scheme of the embodiment of the invention, automatic labeling of the training sample labels of the abnormal transaction detection model is realized, and the sample label labeling efficiency and the label labeling accuracy are improved.

This application discloses a large-scale model memory retrieval method and related apparatus in the field of education, relating to the field of artificial intelligence technology. The method includes: acquiring the user's current dialogue information and a set of user memories to be retrieved; invoking a memory retrieval model; generating representative memories from the user memory set without relying on the current dialogue information; applying attention weighting to the current dialogue information based on the representative memories to obtain a goal-oriented enhanced query representation; and retrieving target memories related to the current dialogue information from the user memory set based on the goal-oriented enhanced query representation. The memory retrieval model is trained using the training user memory set and training dialogue information as training samples, and using related memories of the labeled training dialogue information as sample labels. The model of this application can generate representative memories without relying on the current dialogue information to semantically enhance the current dialogue information, and performs memory retrieval based on the enhanced query representation, thus improving the accuracy of memory retrieval.

The application discloses an energy model training method, a data security detection method and system. The energy model training method comprises the following steps: obtaining training sample data, sample labels and historical probability output results corresponding to historical data; inputting the training sample data into an energy model to be trained to obtain first probability output results corresponding to the training sample data; determining a target loss function based on the first probability output results, the sample labels and the historical probability output results, wherein the target loss function at least comprises a cross-entropy loss function and a contrast loss function; and adjusting model parameters of the energy model by using the target loss function to obtain a target energy model. The application solves the technical problem that the prediction result accuracy of a pre-trained natural language model used for data security detection is not high.

This application discloses an automated rendering optimization method based on visual UI. The method provides optimization for rendering images containing text content. To obtain a rendered image that achieves the desired effect in terms of graphic content, text content, and the fusion between the two, the fusion degree between the two is quantified during sample construction. This helps to convey information while expressing the aesthetic appeal of the image and restoring the designer's intention. Furthermore, the sample labels also indicate the position of the text content in the sample graph, allowing the image processing model to understand the relationship between the boundary and fusion degree between graphic and text content, thus better locating the text content. After the image processing model converges, it possesses the ability to locate the text content and distinguish it from the graphic content. This application provides the means to identify whether the rendering effect of an image needs further optimization.

A generative artificial intelligence-driven method for detecting inappropriate content relates to the field of inappropriate content detection technology, solving the problems of inaccurate detection and reliance on manual labor. The method includes: automatically collecting inappropriate content to obtain a first inappropriate content dataset, and performing detection through a content security detection platform to obtain a first detection result; extracting sample labels from the data in the first inappropriate content dataset to obtain undetected inappropriate features and their combination patterns; manually analyzing the data to obtain highly concealed inappropriate features and novel combination methods, generating prompt words for the AIGC model; constructing a multivariate, highly concealed inappropriate content dataset based on the inappropriate content generated by the AIGC model and not detected by the content security detection platform; and training a content security detection model using the labels and multimodal features of the above two inappropriate content datasets. This invention has high detection efficiency and accuracy, can detect highly concealed and diverse inappropriate content, and reduces reliance on manual labor.

The invention provides a learning motivation determination method and device, equipment and a medium. The method comprises the steps of obtaining a current learning behavior of a target student and at least one first feature related to a learning motivation in response to a learning motivation determination instruction of a user, determining the first feature according to the current learning behavior and a preset first feature type, and determining a learning motivation of the target student according to the at least one first feature and a first sample set; a first probability of each learning motivation category under the condition of all the first features is determined, the first sample set comprises a plurality of first samples, each first sample comprises at least one first feature of a sample student and a sample label, and the sample label is used for indicating that the learning motivation category of the sample student is an active learning category or a passive learning category; and determining the learning motivation category with the maximum first probability in each learning motivation category as the learning motivation category of the current learning behavior of the target student. According to the method, the accuracy of determining the learning motivation is improved.

The invention provides a small sample training method and device for an OCR model, electronic equipment and a medium, and the method comprises the steps: carrying out the data enhancement of a small sample training data set, generating a plurality of meta-tasks, and enabling each meta-task to comprise a plurality of image samples and corresponding task domain identifiers and text recognition result labels; training an initial OCR (Optical Character Recognition) model by taking the plurality of image samples of each meta-task and the corresponding task domain identifiers as training samples and taking the text recognition result labels as sample labels; the initial OCR model comprises a task adaptation layer, an initial text detection layer, an initial text recognition layer and an initial correction layer; the task adaptation layer takes the task domain identifier as input to generate task adaptation parameters, and the task adaptation parameters are used for adaptively updating the parameters of the initial text detection layer, the initial text recognition layer and the initial correction layer. According to the invention, model training can be carried out based on small samples, the training efficiency is high, the generalization ability of the model is improved, and the training cost is reduced.

This invention relates to a method for rice growth status perception and agricultural decision-making based on multimodal semantic guidance, belonging to the fields of smart agriculture, computer vision, and artificial intelligence. The method includes acquiring rice canopy images; constructing a dataset; utilizing a small number of samples labeled by agricultural experts and their corresponding descriptive semantic information; introducing a multimodal semantic-guided auxiliary labeling mechanism to generate growth status labels for the remaining samples, thereby reducing data labeling costs and improving labeling consistency; based on the labeled dataset, constructing a multi-task recognition model to collaboratively identify multi-dimensional growth states of rice, such as phenological stage, nutritional status, disease type, disease impact degree, and population growth, within a unified feature space. Based on the multi-task recognition results, rice growth status analysis information is generated, forming agricultural guidance strategies such as fertilization management and pest and disease control, which are then distributed to field execution equipment, realizing an intelligent management closed loop combining rice growth status perception and agricultural decision-making.

The invention discloses a directional pickup method for a conference machine, related equipment and a computer program product, and the method comprises the steps: obtaining an angle code of a designated spokesman, and transmitting the angle code and an original audio signal collected by a microphone module into a speech enhancement neural network model; the model takes an audio signal which carries an angle coding label and contains noise as a training sample in advance, and takes a clean audio signal of a corresponding angle of the angle coding label as a sample label for training. The model processes an input audio signal, extracts an audio intermediate feature, maps an angle code to an angle code feature matched with the audio intermediate feature in dimension, fuses the angle code feature and the audio intermediate feature, continues reasoning based on the fused feature, and outputs an enhanced audio signal of a target angle. Through the fusion of the neural network model and the angle information, the directional pickup effect can be improved by means of the input and output mapping capability learned by the neural network model on training data, and the adaptability to complex scenes is higher.

This invention discloses a method for establishing time-spectral sedimentary facies samples based on seismic forward modeling. The method includes: Step S1: Constructing typical sedimentary models and various sedimentary facies property templates for the actual work area based on core sedimentary characteristics; Step S2: Obtaining sedimentary facies geological profiles based on typical sedimentary models and sedimentary facies property templates; Step S3: Extracting actual seismic forward modeling parameters from actual seismic data; Step S4: Conducting forward modeling of the sedimentary facies geological model based on the actual seismic parameters; Step S5: Converting the forward modeled seismic data into time-spectral data through continuous wavelet transform; Step S6: Extracting typical seismic traces from the forward modeled seismic profiles, using the sedimentary facies in the geological model as labels and the time-spectral data as feature values ​​to establish sedimentary facies seismic samples. A large number of high-precision samples are generated under the guidance of typical models, and the sample labels are completely accurate, eliminating errors caused by time-depth calibration, thus meeting the needs of machine learning for sedimentary facies prediction.

The invention belongs to the related technical field of power electronic control, and discloses a Buck converter optimal phase shift angle prediction model training method, ripple optimization method and system, and the method comprises the steps: obtaining the training data of a converter under different working conditions, an input sample comprises the positive slope and negative slope of the output current of each phase of the converter, and the input sample comprises the positive slope and negative slope of the output current of each phase of the converter; the sample label is an optimal phase shift angle; and training an artificial neural network by using the training data set, so that the artificial neural network predicts the optimal phase shift angle of the converter based on the positive slope and the negative slope, and the trained artificial neural network is a prediction model of the optimal phase shift angle of the converter. Based on the trained neural network model, the optimal phase shift angle of the converter under different working conditions can be directly predicted, feedback adjustment is performed on a converter switching signal based on a prediction result, current ripples can be effectively suppressed, and the method for predicting the optimal phase shift angle by adopting the neural network model is not influenced by the state of the converter, so that the method is simple and convenient. The converter can be suitable for a converter in a symmetric or asymmetric state.

This specification provides a method for training a lane detection model. The method includes: determining lane line image samples and sample labels, wherein the sample labels are target lane lines in the lane line image samples; extracting features from the lane line image samples to obtain a first feature image and a second feature image; determining lane line clustering information and initial lane line key points based on the first feature image and the second feature image; performing clustering based on the lane line clustering information and the initial lane line key points to obtain predicted lane lines; and training a lane line detection model based on the predicted lane lines and the target lane lines. By integrating the clustering process of lane line key points into the entire lane line detection model training process, lane line key point clustering is performed simultaneously with network training, thereby improving the overall performance of the lane line detection model.

The invention provides a target detection model training method and a target detection method, and the training method comprises the steps: inputting a training sample with a to-be-detected target into a convolution fusion network of a Hough detection model, outputting a plurality of space fusion features, and enabling the training sample to have a corresponding Hough space label, the Hough space label is generated by processing a sample label of the training sample by using a Hough transform algorithm; inputting the plurality of spatial fusion features into a detection accumulation network of a Hough detection model so as to map the image spatial features to a discrete Hough space for target detection, and outputting prediction spatial information of a to-be-detected target, the prediction spatial information representing position size information of the to-be-detected target; and calculating a model updating gradient of the Hough detection model in the discrete Hough space based on the prediction space information and the Hough space label, mapping the model updating gradient to an image space to update model parameters of the Hough detection model, and obtaining a trained target detection model.

The application relates to a low-permeability reservoir classification method, which comprises the following steps: drilling a limited number of core samples distributed at different well depths, and performing physical property testing on each core sample to obtain a corresponding core classification result, which is used as a sample label; acquiring parameter values of each well logging curve in preferred well logging curves, collecting core sample labels and corresponding physical parameters at different depths as labeled sample data, and acquiring well logging curve parameter values at a set well depth according to each well logging parameter curve with a set step, and taking well logging curve parameter values at depths without core samples as unlabeled sample data; based on the labeled sample data and the unlabeled sample data, a semi-supervised machine learning model is trained; and based on the trained semi-supervised machine learning model, the core classification result at the set well depth of each well logging is predicted. The scheme improves the accuracy of reservoir classification.

The present application relates to a kind of based on reinforcement learning's unsupervised mismatch detection method, the present application first inputs the matching point without label to mismatch detection network, training detection network, by multiple iterations to obtain best network parameter, and using the final network model of saving best network parameter to the matching point set to be detected for detection, by generating N extract subset and evaluating its corresponding model, finally the maximum consistent set of model is regarded as correct matching point pair set, to complete mismatch detection and elimination.The present application has higher mismatch detection efficiency, precision and stability, and on the one hand, it overcomes the sample label problem;On the other hand, it is not limited by mismatch rate, and a large number of correct matching points can be obtained with fewer sampling times;And the method is a kind of unsupervised learning framework, which can be compatible with other classification networks, to solve the mismatch detection problem.

Embodiments of the present application provide a sensitive entity recognition model training method, a sensitive entity recognition method, a computing device, a computer readable storage medium and a computer program product. The sensitive entity recognition model training method comprises: performing word segmentation processing on unstructured text and generating a sample label sequence based on a preset rule annotation; inputting a word segmentation sequence into a bidirectional encoding layer to generate a semantic vector, so as to determine an emission score of each label corresponding to the word segmentation; inputting the emission score into a conditional random field layer to obtain a transition score between adjacent labels, and combining the emission score and the transition score to determine a joint probability of each label corresponding to the word segmentation sequence; outputting a predicted label sequence based on the joint probability; and finally training the bidirectional encoding layer and the conditional random field layer with the objective of minimizing the difference between the sample and the predicted label sequence. The technical solution provided by the embodiments of the present application realizes accurate sensitive entity recognition.

The application provides a sample generation method, device and equipment and a storage medium, wherein the method comprises: mapping an original sample to a first label system to obtain a sample label of the original sample, then mapping a business label according to the sample label, and classifying, slicing and filling the original sample based on the business label to obtain a final target sample. The application isolates the establishment and use of the sample from each other, and decouples the relationship, so as to overcome the problems of non-uniform labels and inaccurate sample label semantics, and realize real reuse of the sample.

Disclosed are a downlink channel prediction method, apparatus, device, medium and program product, applied to an access point (AP) in a wireless network, the wireless network including the AP and a site (STA) associated therewith, the method comprising: obtaining training data, the training data comprising first uplink data and a first sample tag, the first uplink data is used for determining first uplink channel state information UL CSI, and the first sample label is first real downlink channel state information DL CSI acquired through a channel detection process; training a basic model based on the training data to obtain a trained first target model, a loss function of the basic model being determined based on a first prediction DL CSI output by the basic model and a first sample label; acquiring second uplink data sent by the STA, wherein the second uplink data is used for determining second UL CSI; and inputting the second UL CSI into the first target model to perform downlink channel prediction processing, and generating a second predicted DL CSI, thereby improving the reliability and effectiveness of downlink channel state information acquisition.

The application relates to a water supply network leakage detection method and device, equipment and medium, the method comprising: replacing corresponding non-core wavelet time-frequency diagrams in wavelet time-frequency diagrams in training samples with equivalent wavelet time-frequency diagrams, and covering part of the core wavelet time-frequency diagrams in the wavelet time-frequency diagrams as training samples; taking the core wavelet time-frequency diagrams as supervision labels; pre-training a masking completion pre-training model of a water supply network leakage detection model to a convergence state; taking the wavelet time-frequency diagrams in the training samples as training samples, and taking sample labels corresponding to the wavelet time-frequency diagrams as supervision labels; fine-tuning a fine-tuning model to a convergence state to complete training of the water supply network leakage detection model; and inputting vibration sound signals of a water supply pipeline to be detected into the water supply network leakage detection model trained to the convergence state to determine whether the water supply pipeline to be detected is in a leakage state or a non-leakage state. The application can greatly reduce dependence on labeled data and improve generalization ability.

The invention belongs to the field of deep learning and remote sensing image processing, and particularly discloses a hyperspectral few-sample classification network construction method and a hyperspectral ground feature classification method, and the method comprises the steps: obtaining a hyperspectral remote sensing image and sample label data thereof; performing cross-domain reconstruction and spectral curve extraction, and filtering out common information among categories through eigenvalue decomposition to obtain a discretized spectral curve; extracting physical invariance features, and taking the physical invariance features as constraint rules to generate virtual samples; the reconstructed hyperspectral data and the spectrum self-supervision auxiliary information are input into a double-branch variational automatic encoder network, multi-loss constraint of cross reconstruction is carried out, and the hyperspectral data and the spectrum self-supervision auxiliary information of the same category show greater similarity in a potential space; and outputting a final surface feature prediction result based on the multinomial logistic regression classifier, and completing the construction of the hyperspectral few-sample classification network. According to the invention, high-precision and high-robustness hyperspectral ground feature classification can be realized under the condition of sample scarcity.