15results about How to "Improve generation accuracy" patented technology

This invention relates to the field of image generation technology, and more particularly to a method for generating bird's-eye view images of reservoir landscapes based on a Stable Diffusion large model. The method includes the following steps: preprocessing acquired reservoir landscape images and semantically annotating them based on water conservancy terminology to construct an image-text pairing dataset; training a pre-trained Stable Diffusion model, incorporating LoRA adaptation parameters into its UNet attention layer, using the image-text pairing dataset to generate a target feature image; and updating the LoRA adaptation parameters based on the difference between the target feature image and the reservoir landscape image to obtain a LoRA model for the reservoir landscape bird's-eye view image. This invention achieves structurally controllable and semantically consistent reservoir landscape generation by fusing structural reference constraints and semantically weighted alignment and introducing LoRA modulation.

The application discloses a museum and art exhibition intelligent operation and visitor experience optimization system and method, relates to the technical field of intelligent management of museums and art exhibitions, and comprises the following steps: obtaining interactive behavior data and identity data of each target object; performing group projection on the interactive behavior data and the identity data, generating group structure data, and performing hierarchical attribution division on the group structure data from an identity dimension, to generate exhibition participation data of each group; and performing scheduling analysis on the correlation degree of the target objects contained in the corresponding group according to the exhibition participation data, to generate behavior-identity deviation data. Through the hierarchical allocation and dynamic reallocation mechanism, the scheme can accurately reflect the resource demand differences of different group levels, and solves the problem of insufficient display strategy adaptability caused by ignoring the behavior differences between group levels in the prior art.

Embodiments of the present disclosure disclose a process page display method and device, electronic equipment and a computer readable medium. A specific embodiment of the method comprises: in response to receiving a process page rendering request, obtaining code information and a set of business processing criteria information; generating initial full business process information; verifying the execution relationship between the initial full business process information, obtaining full business process information; rendering the full business process information and rendering and displaying it on the process page; in response to selecting selection information, generating a matching business batch as a first business batch; and displaying the corresponding business process processing tree on the process page. This embodiment can accurately render the corresponding business process information on the process page after receiving the business data related to the business process generation uploaded to the process page, and can set various page filtering operations for the business process information, greatly improving the rendering efficiency of the business process.

The invention discloses an energy storage energy management method, device and system and a medium, the energy storage energy management system comprises a local energy management system, a local energy management system and a local energy system, the local energy management system obtains operation data of each device in the local energy system, the model prediction precision of a locally deployed energy prediction model is improved; based on the data type of the operation fusion data, performing data fusion processing on all the operation data to obtain the operation fusion data; calling an energy prediction model to perform energy prediction on the operation fusion data to obtain a strategy prediction result; when the model prediction precision reaches the target precision, obtaining a corresponding local pre-stored strategy according to the operation fusion data, and adjusting the local pre-stored strategy according to a strategy prediction result to obtain a target control strategy; according to the invention, the accuracy of energy management and control can be improved, so that the reliability of an energy storage energy management system is improved.

The present application relates to radar detection technology field, especially to a kind of signal processing method of Doppler frequency offset interference of height-finding radar, the method is by monitoring the maneuverability of height-finding radar and dynamically adjusting frequency modulation period and frequency modulation bandwidth of symmetrical frequency modulation signal;Based on the frequency modulation period and frequency modulation bandwidth after adjustment generate positive slope frequency modulation signal and negative slope frequency modulation signal and carry out periodic alternate emission, collect and process and difference calculation first echo signal and second echo signal after reflection, obtain after eliminating Doppler frequency offset interference ranging frequency difference, further avoid the height-finding error caused by Doppler shift amount, to obtain accurate target height based on ranging frequency difference calculation.This application is by using the Doppler shift polarity opposite characteristics of positive and negative slope frequency modulation signal, by frequency spectrum line pairing and difference calculation offset Doppler shift component, finally based on the target height of corrected frequency information inversion, to realize high-precision height-finding.

The application relates to the field of image processing, in particular to a spunlace non-woven fabric defect intelligent identification method and system, which comprises the following steps: dividing a standard non-defect spunlace non-woven fabric image into multiple local image blocks according to spatial positions; constructing a local spectrum prior atlas corresponding to the image spatial positions; performing block processing on a to-be-detected image to obtain multiple original image blocks, and dynamically generating a personalized frequency domain mask for each original image block based on the local spectrum prior atlas to suppress the frequency band corresponding to normal textures; constructing a preliminary defect residual graph and a spectrum graph thereof; in response to the proportion of the medium-low frequency region energy in the total frequency domain energy in the spectrum graph exceeding a preset threshold, updating the mask parameters, applying the updated mask parameters to the personalized frequency domain mask, and obtaining a final defect residual graph; performing threshold segmentation and / or connected domain analysis on the final defect residual graph, and outputting a defect position. The application has the effect of improving defect detection accuracy.

The embodiment of the invention provides an audio generation mouth shape expression method of a large model based on emotion reference and discrimination, and relates to the technical field of computers, and the method comprises the steps: obtaining audio data and corresponding text data, emotion category labels and emotion reference videos; extracting audio features from the audio data through the large voice model, and extracting text features from the text data through the large voice model; extracting an emotion feature and an emotion reference feature from the emotion category label and the emotion reference video; performing facial parameter prediction based on the audio features, the text features, the emotion features and the emotion reference features by using a facial parameter generation model to obtain predicted facial parameters; the face parameter generation model is a large model obtained by performing emotion discrimination and content discrimination on the basis of historically predicted face parameters and performing model training according to a discrimination result; and generating a facial mouth shape expression according to the predicted facial parameters. According to the method, the expression generation accuracy in the complex emotional scene can be improved.

The present application relates to the field of image processing, and more particularly to a kind of image defogging method and system.The present application carries out blocking to the image to be defogged, for each block region, the RGB vector of all target pixel points in each block region is fitted, and the initial fitting matrix of each block region is obtained;Based on the inner product of the RGB vector of each pixel point in each block region and the initial fitting matrix of the block region, the effective pixel point of each block region is screened;The effective pixel point of each block region is fitted, and the optimization direction matrix of each block region is obtained;The luminance average value, optimization direction matrix of each block region are combined to obtain the fusion feature matrix of each block region;Based on the fusion feature matrix of each block region, the atmospheric light value of each block region is obtained;Based on the atmospheric light value and the transmissivity of each block region, the image after defogging is generated.The present application effectively improves the precision of defogging image.

The application proposes a large-scale MIMO system channel graph construction method based on a latent space diffusion model (LDM). The method uses the spatial consistency of the large-scale MIMO system channel to learn the conditional probability distribution of the statistical channel state information (sCSI) about the user terminal position to construct the channel graph. An improved variational autoencoder (VAE) is used to learn the low-dimensional representation of high-dimensional sCSI, and the LDM is improved to improve the generation accuracy of the low-dimensional representation of sCSI, and then a channel graph containing position and sCSI low-dimensional representation data is constructed. The storage and transmission cost of the channel graph is greatly reduced compared with directly using high-dimensional sCSI, and the sCSI low-dimensional representation in the channel graph can be accurately restored to high-dimensional sCSI using VAE, and the generation process does not depend on the pilot signal, which greatly saves the pilot overhead while accurately generating sCSI.

This invention discloses a multi-factor driven logical sequence generation method and apparatus, comprising: a feature input module, a multi-factor fusion module, a dynamic constraint module, a logical generation module, a reordering and evaluation module, and a feedback learning module. The method involves receiving text information from multiple sources, performing structured parsing and normalization to obtain a standardized set of text feature vectors; extracting multi-factors and constructing multi-factor constraints; generating a multi-factor driven logical sequence based on the set of multi-factor constraints; performing dynamic information updates and output constraint adjustments; generating candidate logical sequences; and designing multiple-dimensional indicators for re-evaluation. This invention significantly improves the accuracy, adaptability, and optimality of logical text sequences by addressing the problems of insufficient input information processing, lack of dynamic adaptability, and insufficient post-optimization mechanisms in the logical text sequence generation process.

The application discloses a kind of steel rail defect three-dimensional generation reconstruction methods based on ultrasonic sound field guidance, the B scanning data of steel rail is converted into defect cross section drawing by the cyclic generation of the adversarial network of ultrasonic sound field guidance, then defect cross section aperture and distance and other information are extracted from defect cross section drawing, then the transverse size of defect is calculated in combination with the amplitude of B scanning data, and the position of steel rail defect is converted from B scanning coordinate system to steel rail coordinate system by three-dimensional space transformation, to obtain the spatial position of steel rail defect in steel rail;Finally, the aperture of defect, transverse size and spatial position are transmitted into ultrasonic detection three-dimensional platform, to output the three-dimensional model of steel rail defect, complete the three-dimensional reconstruction of steel rail defect.

The application discloses a four-dimensional flow field magnetic resonance data synthesis method and system based on CT images, and is applied to the technical field of medical images, and the method comprises the steps of: acquiring a CTA image; performing image segmentation and three-dimensional reconstruction processing on the CTA image to obtain a target three-dimensional model of a thoracic aorta system; performing fluid dynamics simulation processing based on the target three-dimensional model to obtain blood flow field simulation parameters of the thoracic aorta system; and performing four-dimensional flow field magnetic resonance data synthesis processing based on the blood flow field simulation parameters to obtain magnetic resonance synthesis data corresponding to the CTA image. The application can generate 4D Flow MRI data from a CT image, and effectively improves the generation accuracy of the 4D Flow MRI data.

The application discloses a network security threat event report generation method and system, and belongs to the network security field.The method comprises the following steps: acquiring a network security threat event log, pre-processing the log, extracting key information by using a pre-trained ERNIE-Gram model, and generating structured knowledge points; integrating the knowledge points into a feature vector, inputting the feature vector into a KAN model for nonlinear feature extraction and multi-level feature fusion, and generating fusion feature data; inputting the fusion feature data into an OverLoCK model, performing multi-stage feature extraction and aggregation through a dynamic context guiding mechanism of the model, and generating summary semantic analysis information; standardizing the semantic analysis information, inputting the semantic analysis information into a pre-trained federal decision tree model, matching a decision path of the federal decision tree model according to a global split rule, and outputting an interpretable security event report containing a confidence distribution.The application can improve the accuracy and interpretability of threat event report generation.

The application discloses a kind of real-time industrial dynamic event clustering method and intelligence system based on industrial knowledge graph.The method comprises the following steps: continuously obtaining near real-time updated industrial information documents from field industrial information data source;For each document, use a general large language model to identify event type and event feature argument, form a unique event feature description;Entity linking is carried out in the industrial knowledge graph, and the event subject name is linked to the enterprise standard name, and is reconstructed into a feature-enhanced event description text and vectorization;Calculate its time decay semantic similarity with the existing events in the dynamic event library, determine whether it is a new event by a judgment model, and update the event library accordingly.The present application extracts fine-grained event elements and disambiguates knowledge graph entities to generate consistent and unique event expressions, effectively improving the accuracy and robustness of event clustering, can process incremental information in near real time, realize real-time monitoring and clustering of dynamic events, and has the advantages of high timeliness and strong interpretability.

The invention relates to the field of data processing, in particular to a data set construction method and system based on big data. The method comprises the following steps: constructing an initial data set containing normal samples and fault samples; decomposing a time sequence in the initial data set to obtain a plurality of frequency band sequences, and extracting time-frequency domain features; calculating the frequency band sensitivity degree of the fault sample and the normal sample on each frequency band based on the time-frequency domain characteristics; constructing an incidence matrix, and combining the frequency band sensitivity degree and the incidence matrix to calculate a frequency band comprehensive weight; constructing a conditional generative adversarial network, constructing a loss function of a generator by using the frequency band comprehensive weight, and training the conditional generative adversarial network based on the loss function; and generating a synthetic frequency band sequence by using the trained conditional generative adversarial network, performing inverse transformation reconstruction on the synthetic frequency band sequence to obtain a synthetic sample, and mixing the synthetic sample with a fault sample in the initial data set to construct an enhanced data set. The method and the device have the effect of improving the prediction accuracy of the LSTM model by using the enhanced data.