98results about How to "Improve learning effect" patented technology

The application discloses an infrared image generation method and system based on a physical prior constraint generative adversarial network, relates to the field of infrared image generation, and aims to solve the problems of insufficient physical reality and lack of representation ability of infrared radiation characteristics of the existing infrared image generation method. The application first designs a physical prior knowledge learning network, which is used to learn the prior representation related to the infrared radiation characteristics from the infrared image, and realizes the prediction of the infrared radiation related parameters through a deep learning network. Secondly, a generative adversarial network based on the physical prior constraint is designed based on the proposed physical prior knowledge learning network, the physical prior features are introduced into the image generation process, and the learning ability of the generator to the overall distribution of the infrared radiation and the representation ability of the generator to the infrared radiation characteristics are enhanced. The method breaks through the limitation that the traditional generator only relies on shallow texture mapping. The application is suitable for the fields of visible light to infrared image generation, infrared vision algorithm research and development.

The application provides a method for training a blood pressure prediction model based on meta learning, which comprises the following steps: acquiring a training set, dividing the training set into a first training set, a second training set and a third training set; pre-training a blood pressure prediction model using the first training set to obtain a pre-trained blood pressure prediction model; initializing an initial meta learner using the parameters of the pre-trained blood pressure prediction model; training the initial meta learner using a plurality of training tasks in the second training set based on a meta learning algorithm to obtain a target meta learner; initializing the target meta learner for each patient in the third training set respectively to obtain an initial personalized blood pressure prediction model corresponding to each patient, and training the initial personalized blood pressure prediction model corresponding to each patient using the training data of the patient in the third training set to obtain a personalized blood pressure prediction model corresponding to the patient.

The application belongs to the technical field of medical image processing, and particularly relates to a pancreas image segmentation method based on complementary attention; the method comprises the following steps: acquiring a pancreas CT image, pre-processing the pancreas CT image, extracting features of the pre-processed pancreas CT image by using a residual dense module to obtain a first feature map; processing the first feature map by using a progressive pyramid pooling module to obtain a second feature map; processing the second feature map by using a main branch decoder and an edge branch decoder respectively to obtain a region feature map and an edge feature map; processing the region feature map and the edge feature map respectively by using a complementary attention mechanism to obtain a pancreas result map and an edge result map; and the application can solve the problems of over-segmentation and under-segmentation, thereby improving the accuracy of pancreas segmentation.

The embodiments of the present disclosure disclose a method and device for training and predicting a response duration prediction model, a medium and a product. The method comprises: obtaining a set of sample data from a sample set as a training set, wherein the sample data comprises feature information of a sample service request and a real response duration of the sample service request; training a plurality of initial multi-objective models based on the sample data in the training set to obtain a plurality of multi-objective models, wherein the learning objectives of the multi-objective models comprise a probability that a service request response duration is within a predetermined range and a predicted response duration of a service request whose response duration is within the predetermined range, and each multi-objective model corresponds to a different predetermined range. The technical solution can more accurately predict the response duration of a service request.

The present application provides a kind of night image halo removal method, including constructing prior knowledge, according to prior knowledge, construct initialization network and depth unfolding network, using initialization network to extract variable initial value set from night halo pollution image;Night halo pollution image and variable initial value set are input into depth unfolding network and are operated multiple iterations, and the final halo removal image is extracted from night halo pollution image;Wherein, depth unfolding network includes multiple proximal networks, and each proximal network participates in an iteration operation.The present application makes full use of prior knowledge, uses the depth unfolding network with multiple proximal networks to iterate mapping diagram, halo-free feature map, constraint variable and halo removal image, so as to extract the final halo removal image, to better preserve image texture and restore image details, under the premise of not increasing application cost, the halo and stripe artifact in image are preferably removed.

This invention discloses a hierarchical supervised super-resolution reconstruction method, system, device, and medium for remote sensing images. The method includes acquiring a high-resolution original remote sensing image, preprocessing it to obtain multiple sets of sub-pixel remote sensing images; constructing a sub-pixel remote sensing image reconstruction network, and pre-training it using a low-resolution remote sensing image and a set of sub-pixel remote sensing images as input and ground truth, respectively; constructing an upsampling-free remote sensing image super-resolution network, and optimizing it using multiple sets of sub-pixel images; constructing a multi-level supervised upsampling-free super-resolution network, and training it using an intermediate-resolution remote sensing image and a high-resolution original remote sensing image as ground truth, respectively; and performing super-resolution reconstruction based on the remote sensing image super-resolution reconstruction model, inputting the remote sensing image whose resolution needs to be improved, to obtain a high-resolution remote sensing image. This invention, by employing a hierarchical supervised and upsampling-free super-resolution reconstruction deep network, can improve the quality and detail of super-resolution reconstruction of remote sensing images.

The application discloses an elliptical multi-scale hidden state space method for intelligent rapid prediction of turbulent flow, and belongs to the cross field of machine learning and computational fluid dynamics. The method first encodes the velocity field, external force field, boundary information and Reynolds number in a multi-scale manner to construct a feature pyramid; then elliptical local attention and four types of physical bias are introduced, and physical consistent feature aggregation is realized through anti-symmetric weight construction; adjacent scale low-frequency bidirectional exchange is used to depict positive and inverse order series of turbulent flow energy; long time series evolution is completed with adaptive time step relying on physical perception hidden state space and system identification; the flow field is output through coarse-to-fine decoding, and autoregressive training is carried out by using joint loss. The application can accurately capture the characteristics of turbulent flow rotation, multi-scale coupling and long-time dynamics, significantly improve the prediction accuracy and stability of high Reynolds number instantaneous turbulent flow, suppress error accumulation, and is suitable for rapid intelligent solution of complex turbulent flow scenes, and has strong generalization ability and engineering practical value.

The application discloses a kind of marine unmanned equipment target detection methods based on multi-modal representation learning, belong to artificial intelligence, marine high-end equipment and so on field, comprising: S1: image data, acoustic data and numerical data are respectively preprocessed, and through space-time alignment, generate after synchronization multi-modal data;S2: based on the multi-modal data after synchronization, extract multi-modal feature vector, through projection head mapping to low-dimensional embedding space, based on contrast learning loss and sequence mask reconstruction loss, large-scale pre-training is carried out, and pre-training multi-modal model is generated;S3: based on pre-training multi-modal model, freeze the parameters of encoder, design prompt vector and multi-modal data are spliced, and through self-regularization constraint, parameter efficient fine-tuning is carried out, and classification model is generated;S4: based on pruning and quantization operation, the classification model is compressed, and deployed to the edge computing unit of marine unmanned equipment.The normal target detection capability of marine unmanned equipment is improved.

The application discloses a spine-pelvis joint segmentation method based on frequency-space cooperation and adaptive fusion, and belongs to the technical field of medical image processing. A joint segmentation dataset containing spine and pelvis structures is constructed, and standardization preprocessing is completed by supplementing pelvis annotation to public data and combining private clinical data; a multi-scale three-dimensional segmentation network is built, a frequency-space cooperation feature modeling HFSM module is introduced in the coding stage, local details and global semantic features are extracted through joint extraction in the spatial domain and the frequency domain; a selective cross adaptive fusion SCAF module is connected between the encoder and the decoder, boundary enhancement, structure perception and gate weight modulation are performed on different levels of coding features and decoding features, adaptive feature aggregation is realized, the network adopts an end-to-end training mode, a weighted combination loss function is used to optimize model parameters, and the joint segmentation result of the spine and the pelvis is output. The application can effectively improve the segmentation accuracy and integrity of the connection region, complex edge and small structure of the spine-pelvis.

The invention relates to the technical field of cutting, detecting, hole separating and plate arranging, in particular to an automatic cutting, detecting, hole separating and plate arranging equipment system which comprises a discharging module, a mold hole number detecting module, a cutting module, a visual detecting module, a code printing module, a code scanning module, a plate arranging and sorting module and a control module. Wherein the discharging module is used for automatically feeding a material belt, the mold cavity number detection module is used for identifying a product mold cavity number, the cutting module is used for cutting a product PIN, the visual detection module is used for PIN quality detection and positioning, the code printing module is used for product marking, and the code scanning module is used for reading and identifying code information. The wobble plate sorting module is used for classifying and wobbling products according to mold cavity numbers; the control module is used for coordinating the working process of each module and integrating intelligent algorithms to optimize production parameters; full-process automation is achieved, manual intervention is greatly reduced, the production takt is improved, the method is suitable for large-scale continuous production, and the production efficiency and the automation level are improved.

The application discloses an optimization method and system for water source conservation function evaluation based on self-supervised reconstruction, and relates to the technical field of ecological environment data processing. The method comprises the following steps: acquiring ecological index data of a target evaluation area, and dividing the ecological index data into multiple ecological factor groups according to ecological function types; tokenizing and encoding each ecological factor group to obtain a group token sequence, masking part of the tokens according to a preset factor group level self-supervised masking strategy, inputting the obtained masked token sequence into a token reconstruction network to complete self-supervised pre-training, removing a reconstruction output layer to obtain an initial representation model; based on evaluation unit data labeled with water source conservation function, fine-tuning the initial representation model and synchronously training an initial evaluation model to obtain a target representation model and a target evaluation model; inputting ecological factor group token sequences of a region to be evaluated, and outputting water source conservation function evaluation results; and the method is suitable for different regional ecological conditions, and improves the utilization efficiency of multi-source ecological data and the stability of evaluation results.

The invention relates to a language learning system, which comprises a learning system, a video player and a foreign language video material, and is characterized in that the output end of the learning system is electrically connected with the video player, and the foreign language video material is in data connection with the learning system and the video player through a data memory. By adopting the translation algorithm module and the subtitle display module, an English learning program is highly simplified, English learning becomes closer to a native language learning process, information of a foreign language video material is decomposed, extracted, translated and combined, a learner can comprehensively and accurately learn the foreign language video material, and meanwhile, the learning efficiency is improved. The translated subtitle information is transmitted to the video player in real time to be displayed, so that a learner can view the translated subtitle information while watching the video, and the learning efficiency and effect are improved.

This utility model discloses a photolithography principle demonstration device, comprising: an outer frame including a frame and multiple baffles; a lighting system including an ultraviolet lamp group and a lens group, the lens group being disposed below the ultraviolet lamp group; an imaging system including an optical lens and a photomask, the photomask being disposed between the optical lens and the lens group; a focusing system including a power module, a moving module, and a support platform, the support platform being connected to the moving module and located below the optical lens, the power module being connected to the moving module and capable of driving it to move up and down, thereby causing the support platform to move along the optical axis of the optical lens; a power supply module connected to the lighting system and the power module; and a control module connected to the lighting system, the power module, and the power supply module. This utility model is applicable to teaching demonstrations of photolithography processes.

This invention provides an enzyme turnover rate prediction method based on a dual-route hybrid expert mechanism, belonging to the field of bioinformatics technology. It solves the problems of low data utilization, insufficient information mining, and poor robustness in existing prediction methods due to missing values ​​and temperatures. The technical solution includes the following steps: S1: Constructing a unified standard enzyme turnover rate dataset; S2: Extracting multimodal embedding features; S3: Constructing intra- / inter-modal encoders; S4: Combining modalities into an expert hybrid module; S5: Designing an attention fusion mechanism. This invention can achieve high-precision prediction under complex in vitro environmental conditions.

Embodiments of the present application disclose a training method and device of a pre-training model applied to a three-dimensional image and equipment, and belong to the technical field of artificial intelligence. The method comprises the following steps: repeating the following steps until a preset condition is reached: determining a first feature vector of a to-be-processed mask image based on a first to-be-trained model, determining a second feature vector of a to-be-processed three-dimensional image; obtaining a third feature vector of each other three-dimensional image; updating the parameters of the first to-be-trained model according to the first feature vector, the second feature vector and the third feature vectors; and the first to-be-trained model after the preset condition is reached can extract a feature map of the three-dimensional image. The first feature vector and the second feature vector of the same to-be-processed three-dimensional image form a positive sample pair, the first feature vector of the to-be-processed three-dimensional image and the third feature vector of the other three-dimensional image form a negative sample pair, and the parameters of the first to-be-trained model are updated. The obtained model can accurately obtain a feature map of the three-dimensional image.

The invention discloses a deep counterfeiting detection method based on double-flow fusion and adaptive feature enhancement, and belongs to the technical field of computer vision and digital media security. According to the method, a double-flow feature extraction network is constructed, a high-low feature adaptive enhancement module (HLFAE) is adopted in a spatial flow to decompose multi-scale texture features, and micro texture expression of a forged area is enhanced in combination with expansion convolution and a channel attention mechanism; introducing a high-frequency sub-band capable of learning a discrete wavelet transform self-adaptive decomposition image into a frequency domain flow, and amplifying frequency domain artifact features through a convolutional network; a multi-modal enhanced feature attention module is designed to dynamically fuse space and frequency domain features, significant features are weighted through a multi-scale convolution kernel and a double attention mechanism, and feature interaction consistency is improved based on a cross-modal contrast enhancement (CMCE) module. According to the method, the highest AUC value of 99.63% is achieved on data sets such as FaceForce + + and Celeb-DFv2, the cross-domain generalization ability and the anti-disturbance robustness are remarkably improved, and the method is suitable for financial risk control and media content auditing.

This invention relates to the field of immersive educational interactive solutions designing using virtual reality technology, specifically to an immersive educational interactive method and system integrating virtual reality technology. The method includes: data acquisition: collecting students' learning behavior data through virtual reality devices; learning profile generation: generating students' learning profiles based on the learning behavior data using artificial intelligence algorithms; and dynamic adjustment of learning paths: dynamically adjusting the learning path according to the learning profile. The technical solution of this invention, by integrating virtual reality technology and artificial intelligence algorithms, achieves real-time monitoring of students' learning behavior data, dynamic adjustment of learning paths, and generation of personalized learning feedback, significantly improving learning effectiveness, learning interest, and learning persistence. This invention has a higher level of intelligence, stronger adaptability, and more significant educational value, and can provide innovative solutions for personalized learning in the field of education.

The application belongs to the technical field of image processing, and provides a rock FIB-SEM sequence image multi-phase registration segmentation method and system, a technical scheme of which is based on a constructed detail-preserving non-rigid registration subnetwork, and through affine correction and non-rigid registration, a registered image, a displacement vector field and a Jacobian determinant thereof are output; based on a constructed deformation-aware multi-phase segmentation subnetwork, the registered image is segmented to obtain a plurality of channel probability maps; in order to achieve registration accuracy and segmentation accuracy, the subnetworks are jointly trained by using a constructed loss function to obtain parameter-optimized subnetworks; and based on the parameter-optimized subnetworks, a target sequence image is processed to obtain three-dimensional image volume data and multi-phase semantic segmentation three-dimensional labels aligned therewith. The method can simultaneously complete high-precision image registration and semantic segmentation while preserving high-frequency details and noise features with geological significance in the original image.

The utility model relates to a desk type multi-simple-pendulum physical cycle demonstration instrument, belonging to the teaching demonstration equipment technical field, the desk type multi-simple-pendulum physical cycle demonstration instrument comprises a main body frame, a push plate, adjusting devices and pendulum balls, the push plate is arranged at the bottom of the main body frame, the plurality of adjusting devices are uniformly arranged at the top of the main body frame along the transverse direction at intervals, and the pendulum balls are arranged on the push plate. By adopting the structure, the pendulum length of each pendulum ball is accurately adjusted through the adjusting devices, and then the push plate is used for driving all the pendulum balls to synchronously start simple pendulum motion, so that the relationship between the simple pendulum period and the pendulum length can be intuitively understood; and the interaction and periodic change of the multi-simple-pendulum system in the swinging process provide powerful support for physics teaching, and the teaching quality and the learning effect of students are improved.

The application discloses a kind of brain-like reinforcement learning method and system based on layered experience playback, it is related to reinforcement learning and brain-like computing technical field.The method comprises the following steps: S1, collecting observation data and pre-processing;S2, initialize experience buffer pool and actor network, critic network and corresponding target network, parameter initialization;S3, initialize exploration noise and select action from actor network according to current state and execute, store the obtained experience sample to experience buffer pool;S4, obtain new sample from experience buffer pool, carry out short-term memory pool update;S5, use attention discrimination module to determine whether part of experience in short-term memory experience pool is transferred to long-term memory experience pool;S6, the parameters of actor network, critic network and corresponding target network are updated.The application uses the above method, improves the experience utilization rate of intelligent agent, improves the performance of reinforcement learning, and has wide application potential in many fields.

The application provides a dynamic contrast-enhanced magnetic resonance image classification method, system, device and medium, the method comprising: acquiring breast dynamic contrast-enhanced magnetic resonance images of different disease ages and preprocessing to generate a breast dynamic enhancement magnetic resonance image dataset; according to the breast dynamic enhancement magnetic resonance image dataset, an image classification prediction model comprising a multi-task classification network and a perception discrimination network connected in turn is constructed; the acquired breast dynamic contrast-enhanced magnetic resonance image to be classified is input into the image classification prediction model for multi-task classification, and an image classification result comprising the position, spatial structure and spatial signal of the tumor is obtained. The method can fully and accurately mine the rich multi-dimensional, multi-scale heterogeneous space, time and semantic correlation representation possessed by the breast contrast-enhanced magnetic resonance image data, and can supplement the missing different spatial dimension information, ensuring the reliability and accuracy of image classification prediction.

This invention relates to the field of virtual reality technology, and more particularly to a method and system for basic medical virtual experimental teaching. The method includes constructing a physiological digital twin model; achieving real-time bidirectional coupling of macroscopic and microscopic physiology through a dynamic simulation engine; compiling trainee operations based on force feedback devices and providing interaction; utilizing a knowledge graph to trace errors and create competency profiles for the operations, and generating personalized feedback. This invention significantly improves the physiological fidelity and teaching relevance of virtual experiments.

The application provides a kind of based on feature enhancement and composite model's ultra-short-term wind power prediction method, belongs to ultra-short-term wind power prediction technical field, this method includes: by calculating the mutual information of multiple meteorological characteristics and wind power output and normalizing, screening to obtain key features, and constructing wind power prediction fitness function, by minimizing wind power prediction fitness function, obtain enhanced input features;By introducing one-dimensional convolutional neural network and bidirectional long short gate recurrent unit, construct the residual network of bidirectional long short gate recurrent unit and convolutional neural network fusion, and introduce space-time attention mechanism, construct to obtain composite residual network ultra-short-term wind power prediction model;And use multiple meteorological characteristics and enhanced input features to predict to obtain ultra-short-term wind power prediction result;The application solves the problem that the influence of meteorological multi-scale fluctuation on output cannot be accurately grasped, prediction is disconnected, disturbance is not considered and prediction accuracy continues to decline with step increase.

This invention relates to a single-stage fine-grained target detection method and apparatus based on remote sensing images. The method includes: acquiring image data and label data; cropping and flipping the image data to construct training and testing sets; constructing a feature extraction network using a convolutional neural network as the detection model; extracting features from the image data using the feature extraction network; and outputting the location and category information of fine-grained targets in the image. A positive and negative sample allocation algorithm is used to divide the location information boxes into positive and negative samples, and the convolutional neural network is used to learn the allocation of positive and negative samples. Based on the PyTorch deep learning framework, the network structure is written in Python, and the convolutional neural network is trained using the image and label data in the training set to obtain the target detection model. The image data to be detected is used as input to the target detection model to output the category and location information of the target objects in the image data to be detected.

The application discloses a patent technology function extraction method for technical literature, analyzes the features of patent technology terms and function terms in the field of high-end equipment, and improves the accuracy of technology term and function term extraction in Chinese patents. The application constructs a deep learning model for technology term and function term extraction, combines the sentence pattern rules of patents, constructs the heuristic features of technology terms, positions the function sentences by constructing a function term feature dictionary, accelerates the training speed of the model, and improves the extraction accuracy. In order to reduce the cost of manual sample labeling and avoid the model overfitting phenomenon caused by too small data set, a self-training algorithm is used to realize weak supervision learning of the model. The technology terms most similar to the theme of the patent text are selected from the word clustering, the cosine similarity is used to combine the similar semantic terms, and a technology function matrix is constructed.

The invention relates to the field of wireless communication networks, and provides a self-adaptive sum-difference channel phase compensation method based on double-flow feature fusion, which comprises the following steps: receiving a non-cooperative target satellite signal, constructing a spatial network pre-training data set in combination with antenna beam pointing, and pre-training a spatial network model to output a static phase difference compensation predicted value; constructing a time network pre-training data set based on the historical phase deviation cached by the register and the temperature difference value of the environment sensor, and pre-training a time network model to output a phase difference change trend predicted value; and carrying out feature fusion on the static phase difference compensation predicted value and the phase difference change trend predicted value, carrying out training through a fusion network model, outputting a final phase compensation value of a sum-difference channel, and driving a digital phase shifter to carry out real-time phase correction. Through heterogeneous feature separation of a spatial stream and a time stream, the phase compensation precision and the environmental robustness are greatly improved, and particularly, accurate output of sum-difference channel phase compensation is ensured in a complex dynamic scene.

The application provides a kind of teaching system and method based on VR virtual classroom, it is related to virtual reality technical field, the teaching system, comprising: acquisition unit, for obtaining the limb node coordinates of user;First identification unit, for obtaining the VR teaching action corresponding to VR limb guide model according to limb node coordinates, VR limb guide model and action arrangement;Execution unit, for executing VR teaching action by VR limb guide model to carry out teaching guidance;Second identification unit, for obtaining the actual action formed by the change of limb node coordinates in the process of VR teaching action teaching according to limb node coordinates;Comparison unit, for comparing actual action and VR teaching action, and obtaining comparison result;Generation unit, for generating the next teaching plan corresponding to action arrangement according to comparison result.The teaching system and method of the application can improve the learning enthusiasm and learning ability of user under the premise of ensuring the quality of teaching.

The application discloses a high-speed rare thin flow field prediction method and system based on a mask self-encoder, and belongs to the technical field of cross hyper-sonic aerodynamics and artificial intelligence. The application mainly comprises the following steps: constructing a cross-flow multi-scale flow field database; constructing and pre-training a multi-scale mask self-encoder model, forcing the model to learn to reconstruct the full-field flow field from sparse information through a random mask strategy; introducing a physical information constraint to fine-tune the model, wherein the physical information constraint comprises a Chapman-Enskog distribution function constraint, a moment equation conservation residual and a wall Maxwell slip boundary condition; introducing a Knudsen number self-adaptive reasoning mechanism to automatically adjust the feature weight through a gate network; and using the trained model to predict the flow field. The application realizes cross-flow unified rapid prediction covering a wide Knudsen number range, guarantees the prediction accuracy, significantly improves the calculation efficiency, and has good physical consistency and generalization ability.

The application discloses an industrial control method and device based on model training, and belongs to the field of industrial control. The industrial control method based on model training comprises the following steps: using a target algorithm prediction model to perform operation processing on real-time industrial big data in a target industrial scene, so that a target control algorithm and a target independent variable for controlling a target controlled object in the target scene output by the target algorithm prediction model are obtained; and inputting the value of the target independent variable in the real-time industrial big data into a target controller in which the target control algorithm is deployed, so that the value of a target control variable is obtained, to control the target controlled object. The industrial control method based on model training has high precision and accuracy of the obtained target control algorithm, can significantly reduce the labor and time cost, improve the control efficiency and control effect, and has high universality.

The invention discloses a power transformer fault diagnosis method combining domain knowledge and a capsule network. The method comprises the steps of performing feature enhancement in combination with the domain knowledge; performing different normalization processing according to different characteristics of the original features and the knowledge features, and organizing the original features and the knowledge features into a two-dimensional matrix for convolution; and inputting a capsule network for training to obtain a trained network model, and diagnosing test data. According to the method, the domain knowledge is utilized to perform feature enhancement on the original data, and different normalization methods are adopted according to the characteristics of the original features and the created knowledge features, so that the interpretability and the diagnosis precision of the algorithm are improved; feature values are converted into a two-dimensional matrix, and the learning ability of the algorithm is improved through a hidden mode between convolutional layer mining features; aiming at the characteristic that each feature of the data of the dissolved gas in the transformer oil has specific significance, the capsule network sensitive to the vector is adopted, so that the relationship among the features can be reserved, and the method is very important for understanding and analyzing a complex mode in the data of the dissolved gas.