30results about How to "Reduce computing needs" patented technology

The invention relates to the technical field of prostate MRI (Magnetic Resonance Imaging) image segmentation, in particular to a prostate MRI image segmentation method based on Mamb-Unet, which realizes effective fusion of features of different scales between an encoder and a decoder, and fuses multi-scale information on a space axis by using a space attention bridge segment to fuse fusion features of different levels from the encoder, so that the segmentation accuracy of the prostate MRI image is improved. And then global and local information is captured through 1D convolution and full connection in the channel attention bridge segment, so that a channel attention map is generated, channel information is better integrated, and multi-level different-scale feature fusion is realized.

The invention provides a smoothed particle fluid dynamics simulation wave impact structure boundary force processing method, which comprises the following steps: S1, initializing the speed, density and position of particles by establishing a numerical model; s2, dividing fluid particles and boundary particles; s3, calculating and correcting the density and speed variation of the particles according to the hydrodynamic equation and the boundary rejection effect; s4, calculating the pressure and the speed of boundary particles through a Schelerd interpolation method, and reflecting the interaction between the fluid and the solid; s5, based on the pressure and the position of the boundary particles, calculating the pressure intensity, obtaining the stress and the torque of the structure, and updating the degree of freedom; and S6, updating the particle state and entering the solution of the next time step. According to the method, when the influence of the boundary force is determined, the influence of the relative speed of fluid particles and wall surface particles is considered, meanwhile, the maximum speed in a flow field is approximately replaced by the artificial sound velocity, and the purposes of accurately simulating the interaction process of the ocean wave-ocean engineering structure and ensuring the stability of numerical simulation calculation are achieved.

The invention belongs to the field of oil-gas exploration seismic data processing, and relates to a first-motion wave travel time chromatography velocity inversion method, system and device and a medium. According to the method, the first arrival travel time and the grid unit travel time are acquired, and the minimum grid unit travel time is selected; an initial velocity field closer to a real structure is established through micro-logging constraint so as to replace a conventionally used gradient velocity field. The initial velocity field and the first arrival travel time are combined to construct the target function, the slowness information and the regularization information are added in the target function, the multi-information constrained target function is obtained, and the inversion stability and the calculation precision are improved. And adopting a Gaussian-Newton method to minimize the target function of the multi-information constraint, solving the minimized target function of the multi-information constraint, and carrying out iterative updating on a solving result to obtain an inversion velocity field. Explicit calculation and sensitive kernel function storage are not needed, the method is more suitable for large-scale data processing, and the calculation time of speed inversion and the requirement for storage space are reduced.

The invention relates to the technical field of automatic control, in particular to an intelligent parking lot access control system and method, and the system comprises a reservation information processing module, a vehicle entrance detection module, a parking space intelligent distribution module, a navigation path calculation module, an indicator light control module and a parking duration calculation module. According to the invention, through automatic identity verification and parking space matching, manual intervention is reduced, the efficiency and accuracy of the reservation process are optimized, the magnetic field inductor is used for identifying vehicle entrance and starting the image acquisition equipment, and vehicle information is identified and recorded, so that the calculation demand and equipment energy consumption of image analysis are reduced, and the vehicle entrance speed and safety are optimized; through parking space distribution and navigation path calculation, in combination with indication lamp control, an optimal navigation route is provided for a driver, a vehicle is guided to accurately arrive at a designated parking space, the time for finding the parking space by the vehicle is shortened, and through a mode of automatically calculating parking fees and controlling an exit barrier gate, payment and departure processes are optimized, and the convenience of the parking process is improved.

The invention relates to a question and answer method and device, computer equipment and a readable storage medium. The method comprises the following steps: inputting an initial question text of a user into a quantized language model to obtain word slot parameters corresponding to the initial question text; the word slot parameter comprises description information related to the target question and answer scene; the quantized language model is a language model obtained by compressing model parameters of the large language model; inputting the word slot parameters and the initial question text into the quantized language model to obtain a standardized question text; the standardized question text is a text which conforms to a preset text structure and clearly reflects the intention of the user; the standardized problem text is input into a vector model, and a plurality of documents with the highest correlation degree with the standardized problem text are obtained; and inputting the standardized question text and the plurality of documents into the quantized language model to obtain an answer text generated by answering the initial question text. The method can improve the reasoning speed of the question answering system.

The invention discloses a rapid assessment method for seawater intrusion and sustainable freshwater exploitation in an phreatic aquifer considering hydraulic boundary changes. The method comprises the following steps: constructing a conceptual model based on specific working conditions; potential energy distribution in the phreatic aquifer is calculated; the seawater intrusion range is determined by searching an equipotential line equal to potential energy at a salt toe position in a potential energy field through interpolation and the height of a salt-fresh water interface; and calculating the sustainable fresh water exploitation rate: calculating the stagnation point position, wherein the water pumping rate when the stagnation point coincides with the salt toe position is the sustainable fresh water exploitation rate. According to the method, the evaluation precision and calculation efficiency of the seawater invasion range and the sustainable fresh water exploitation rate can be remarkably improved while the result accuracy is ensured; by adopting the conceptual model based on the working condition, the research area range and parameter setting can be flexibly adjusted according to actual requirements, and by optimizing the calculation model, the requirements of calculation resources and manpower and material resources are remarkably reduced under the condition that additional equipment or resource investment is not increased.

The invention discloses a method for rapidly evaluating seawater intrusion and sustainable freshwater exploitation in a confined aquifer by considering hydraulic boundary change. The method comprises the following steps: constructing a conceptual model based on specific working conditions; potential energy distribution in the confined aquifer is calculated; the seawater intrusion range is determined by searching an equipotential line equal to potential energy at a salt toe position in a potential energy field through interpolation and the height of a salt-fresh water interface; and calculating the sustainable fresh water exploitation rate: calculating the stagnation point position, wherein the water pumping rate when the stagnation point coincides with the salt toe position is the sustainable fresh water exploitation rate. According to the method, the evaluation precision and calculation efficiency of the seawater invasion range and the sustainable fresh water exploitation rate can be remarkably improved while the result accuracy is ensured; by adopting the conceptual model based on the working condition, the research area range and parameter setting can be flexibly adjusted according to actual requirements, and by optimizing the calculation model, the requirements of calculation resources and manpower and material resources are remarkably reduced under the condition that additional equipment or resource investment is not increased.

The invention discloses a multi-modal medical image classification method based on an improved DenseNet, a medical image classification network is constructed, the network comprises a preliminary feature extraction module, a feature refining module and a classification module, the feature refining module fuses the interlayer dense connection characteristic of the DenseNet and the efficient feature processing capability of a Mamba module, and the classification module is used for classifying the Mamba module. Extraction and fusion of complex image features are realized; and the classification module adopts a full connection layer and a Softmax function and is responsible for converting rich image feature representation into a specific image classification result. According to the invention, through the transition layer with the Mama module, the capturing capability of the model on the key features of the image is enhanced, so that the classification precision of the 3D medical image is further improved on the basis of maintaining the original advantages of the DenseNet, and powerful technical support is provided for the application of deep learning in medical image processing.

The invention discloses an Android malicious application detection method based on a multi-view deep neural network, and relates to the technical field of information. The invention provides an Android malicious application detection method based on a multi-view deep neural network in view of the technical problems of numerous attack means, complex and changeable semantics, difficulty in ensuring the coverage rate of attack behaviors by executing detection only from a single view and the like of the current Android malicious application. According to the method, multi-view static feature information is organically fused, the synergistic effect of static features under all views is fully exerted, a target object is comprehensively described, limitation and fuzziness caused by application analysis from a single view are avoided, and finally accurate detection of malicious applications is achieved.

The invention discloses a quantitative laser speckle blood flow imaging method based on a neuromorphic visual sensor, and relates to the technical field of biomedical engineering. The method solves the problem that static scattered light caused by physiological behaviors and electronic noise of imaging equipment have adverse effects on blood flow imaging quality in a traditional laser speckle contrast imaging method, and comprises the following steps: step 1, using a helium-neon laser as a light source, and irradiating a scattering medium with an expanded laser beam to form a speckle pattern; 2, a microfluidic injector is adopted, two parameters of the inner diameter and the flow of the injector are set, and the flow speed range is controlled; 3, asynchronously sensing the change of the light intensity of the speckle pattern by using an event camera, and shooting the speckle pattern by using a frame camera; and 4, acquiring event stream data by the event camera, and processing the data by using an algorithm. Real-time and high-resolution blood flow imaging and flow velocity calculation are achieved, and the inherent limitation of a traditional method is overcome.

The invention relates to the technical field of target detection and tracking, and provides a lightweight dual-dynamic vehicle detection and classification method and system in a complex road environment, and the method comprises the steps: carrying out the preprocessing of a vehicle image through an image collection module, and constructing a target data set; inputting the new vehicle image into a vehicle detection classification model for analysis so as to automatically output the type and position information of the vehicle; wherein the vehicle detection classification model comprises a trunk feature extractor, a multi-scale adaptive feature fusion device and a classification locator; the trunk feature extractor extracts multiple pieces of feature layer information from the vehicle image; and the multi-scale adaptive feature fusion device extracts and fuses the information of a plurality of feature maps and automatically outputs the category and position information of the vehicle through the classification locator. Compared with an existing advanced deep learning vehicle detection algorithm, the method has the advantages that the parameter quantity and the calculated quantity are smaller, the detection precision is higher, and the light-weight vehicle detection problem under the complex road background can be solved.

The invention discloses an encrypted traffic classification method based on feature fusion and a packet dilated convolutional network, which identifies network traffic with high precision and ensures privacy and computational efficiency at the same time, and comprises the following steps: data preprocessing: only retaining 26 bytes to protect privacy through segmentation, filtering and truncation; feature extraction: adopting a double-sliding window mechanism and a designed hash function, generating concrete and abstract features in cooperation with an embedded layer, and realizing joint feature representation; in the model design, firstly, lightweight convolutional network design is carried out, a receptive field is expanded through a lightweight grouping dilated convolutional structure, and channel mixing and point-by-point convolution are added to increase interchannel communication; and in a classification output stage, a cosine annealing mechanism is used to dynamically adjust a learning rate so as to improve a training effect and a convergence speed. The method is superior to an existing lightweight method in the aspects of accuracy, stability, processing speed and resource occupation, is suitable for efficient traffic classification application of resource-constrained equipment, and has practical and promotional values.

The invention discloses an intelligent robot end-to-end motion planning method. The method comprises the following steps: constructing a simulation environment for extraterrestrial detection; the simulation environment is used for simulating motion characteristics of the robot in the extraterrestrial detection environment and wheel soil contact dynamic characteristics, and outputting sensing information of the robot based on a received action instruction; an end-to-end motion planning model is constructed, and the motion planning model takes perception information of the robot as input and takes the action instruction as output; the motion planning model is provided with a reward function, and the reward function is fused with artificial priori knowledge and actual engineering constraints; training the motion planning model based on a simulation environment to obtain a trained model; and in response to deployment of the trained model to an upper computer of the extraterrestrial exploration robot, enabling the robot to output a motion planning result based on real sensing information measured by a sensor of the robot. According to the invention, computing resources are saved, and the end-to-end motion planning capability of the robot under the constraint of limited resources can be improved.

The invention discloses an attention learning mechanism-based inverter model control prediction method, system and device, and a storage medium, and the method comprises the steps: pre-constructing a three-phase inverter simulation model, carrying out the real-time sampling of a state signal through employing an MPC controller, and obtaining a signal characteristic set; performing feature selection on the signal feature set, and extracting feature data; setting a screening threshold to screen the feature data, and taking the screened feature data as optimal features; inputting the optimal feature into a pre-trained machine learning model module to obtain a prediction result of the control signal of the three-phase inverter; the control performance can be improved while more calculation amount and communication traffic are reduced.

The invention discloses a transcranial magnetic stimulation waveform design method, a transcranial magnetic stimulation waveform design device and electronic equipment, and belongs to the technical field of medical instrument design, and compared with a traditional commercial discharge circuit, the transcranial magnetic stimulation waveform design method selects a novel discharge circuit with controllable waveform parameters, and has higher flexibility. Taking a first preset circuit parameter and a second preset circuit parameter in the waveform parameter controllable transcranial magnetic discharge circuit as optimization indexes, taking a minimum selectivity index as an optimization target, and utilizing an optimization algorithm to find out an optimal first preset circuit parameter and a corresponding second preset circuit parameter; taking a waveform corresponding to the optimal first preset circuit parameter and the corresponding second preset circuit parameter as a target design waveform; the method can improve the selectivity of transcranial magnetic stimulation, the determined target design waveform can improve the controllable transcranial magnetic stimulation effect to a large extent, and meanwhile, the potential of waveform parameter controllable transcranial magnetic stimulation is exerted to the maximum extent.

The invention discloses a drop-out fuse state monitoring method based on multi-size features and deep learning, and the method comprises the steps: S1, forming a data set for obtained drop-out fuse pictures, and dividing the data set into a training set and a test set; s2, adding an improved receptive field block and a coordinate attention module to a YOLOx backbone network, adding an adaptive spatial feature fusion module to PANet, carrying out secondary fusion on features of different scales, then introducing a loss function of weighting loss and positioning loss fusion, and finally carrying out lightweight improvement, constructing a state monitoring model, and carrying out state monitoring. Performing training verification on the constructed state monitoring model through the training set and the test set; s3, identifying a drop-out fuse picture acquired in real time by adopting the trained and verified state monitoring model; according to the image data collected by the application, the construction of a special database is realized, the YOLOx is improved and lightweight operation is carried out, and the complexity of the model is reduced, so that the method is suitable for an embedded platform of an electric unmanned aerial vehicle, and the detection effect of the drop-out fuse is improved.

The invention discloses a method for aligning an imbricate spectrum hidden Markov sequence on a directed acyclic generic genome diagram in the technical field of sequence alignment, which comprises the following steps of: carrying out pre-alignment preprocessing on the directed acyclic generic genome diagram to obtain a Viterbi diagram, a training diagram and an alignment reference diagram; then taking the length of the longest comparison reference path in the comparison reference map as the matching state number of the shingled spectrum hidden Markov model, training on the training map after initializing the model parameters and the shingled width of each node to obtain the shingled spectrum hidden Markov model, and when multi-sequence comparison is needed, performing multi-sequence comparison on the shingled spectrum hidden Markov model to obtain the shingled spectrum hidden Markov model. A Viterbi decoding algorithm based on virtual node probability calculation is used on the model, the most probable state path of each sequence in the imbricate spectrum hidden Markov model is calculated on a Viterbi graph, the comparison relation of all positions in the sequences is determined according to the state paths, and therefore multi-sequence comparison is completed. According to the method, the calculation time and storage space requirements of sequence alignment are greatly reduced.

The invention provides an image fusion method and device, and belongs to the technical field of image fusion, and the method comprises the steps: importing a plurality of original high-resolution hyperspectral images, and carrying out the down-sampling analysis of the original high-resolution hyperspectral images, and obtaining an original low-space hyperspectral image and an original high-space multispectral image; performing fusion analysis on the original low-space hyperspectral image and the original high-space multispectral image through a training model to obtain a target high-resolution hyperspectral image; and updating and analyzing the training model according to the target high-resolution hyperspectral image and the original high-resolution hyperspectral image to obtain an image fusion model. According to the method, the fidelity of the space and the spectrum of the fused image is kept, the calculation requirement is reduced while the spectrum and the space information are recovered, the calculation complexity is also reduced, better algorithm interpretability is achieved, and therefore a more accurate fusion result is obtained.

The invention provides a system and a method for battery cell degradation detection. Methods and systems are provided for a high voltage multi-array battery for a plug-in vehicle. In one example, a method may include monitoring that a rate of change of a difference in maximum cell voltage deviations of at least one cell array of a high voltage multi-array battery exceeds a predetermined threshold. In response to the rate of change exceeding a predetermined threshold, plug-in charging is aborted and a diagnostic fault code (DTC) is recorded in a memory of the vehicle.

The invention discloses a vehicle outfield noise prediction method and device, electronic equipment and a storage medium. The vehicle outfield noise prediction method comprises the following steps: acquiring a first sound transfer function from each noise source on a vehicle to a sound source near field, near-field noise values of the vehicle under different operation conditions and a second sound transfer function from each noise source on the vehicle to a far-field observation point; determining the sound power of each noise source on the vehicle according to the first sound transfer function and the near-field noise value; and obtaining a far-field noise prediction value of each noise source on the vehicle according to the sound power of each noise source on the vehicle and the second sound transfer function. By adopting the method, accurate prediction of the vehicle outfield noise can be realized, the uncertainty of model simplification and environmental parameter setting can be avoided, so that the prediction precision and efficiency are improved, the cost is reduced, and the method has higher applicability and flexibility.

This invention discloses a method for constructing and predicting a relay health index based on Mahalanobis distance and multi-channel information fusion. The method includes: first, sampling the coil drive current signal sequence and contact voltage signal sequence during a single relay cycle, extracting the peak and trough time points and amplitudes of each sequence; then defining the first 10% of the relay's cycles as the healthy period, extracting the relative deviation between the current cycle feature value and the mean feature value of the healthy period, further calculating the Mahalanobis distance of the feature vector relative to the health status benchmark, and sequentially performing Box-Cox transformation, normalization, and monotonicity mapping to obtain the final health index. Using [variable name] as input and [variable name] as output, a random forest is selected as the regression prediction model for model training, and online prediction is achieved. The HI curve output by this invention avoids severe oscillations, greatly improving the stability of the prediction results and its engineering application value. Furthermore, the model is lightweight and suitable for edge deployment.

The invention relates to a distribution line detection method based on deep learning, and the method comprises the following steps: S1, collecting normal and fault images of a distribution line, collecting the images under different visual angles and different illumination conditions through employing an unmanned plane device, and expanding a data set through employing a data augmentation and synthesis technology; s2, performing preliminary feature learning on the expanded data set by applying a self-supervised learning method; s4, training the distribution line detection model based on the training data set, adjusting a learning rate by using One-Cycle Policy, quickly entering an optimal learning stage and preventing over-fitting, and optimizing a learning process of an initial stage by using a Warmup strategy, and S3, establishing a distribution line detection model based on the OfficitDet and Transformer, and S4, training the distribution line detection model based on the training data set, adjusting a learning rate by using One-Cycle Policy, quickly entering the optimal learning stage and preventing over-fitting, and optimizing the learning process of the initial stage by using a Warmup strategy; s5, adopting a model pruning and quantification technology to obtain a final distribution line detection model; and S6, deploying the final distribution line detection model on unmanned aerial vehicle equipment to realize on-site real-time fault detection and feedback. According to the invention, the detection efficiency and accuracy can be greatly improved, and diversified environmental changes and equipment limitations can be effectively dealt with.

The invention relates to the technical field of computer vision and space calculation, and discloses a feature processing method, a visual tracking method, a device, equipment and a medium, and the feature processing method comprises the steps: obtaining a transformation relation between a current frame image coordinate system and a current world coordinate system according to a current frame image captured by the equipment; according to the transformation relation, converting the world coordinates of the historical road sign points in the point cloud library into image coordinates in the current frame image, so as to determine corresponding projection points of the historical road sign points on the current frame image; dividing the current frame image into a plurality of grids, determining a first type of feature points in each grid, the first type of feature points being feature points associated with feature points of the current frame image and feature points of a previous frame image; if the number of the first type of feature points in the grids is less than the preset value, adding a corresponding number of projection points in the corresponding grids; and carrying out optical flow tracking on the corresponding number of projection points, and determining potential feature points of the corresponding number of projection points on the current frame image so as to improve the stability of front-end feature tracking.

The present application discloses a modal decision method and device for a multimodal mobile platform, which relates to the field of motion control; the method comprises: obtaining basic parameter data of the multimodal mobile platform; determining trajectory information data according to the basic parameter data; inputting state space data into a Markov decision model, and outputting modal decision information; the state space data comprises basic parameter data and trajectory information data; the Markov decision model comprises: a strategy neural network and a reward neural network connected to each other; and controlling the multimodal mobile platform to move according to the modal decision information. The present application can realize modal motion decision control of a multimodal mobile platform.

The invention discloses a large vehicle right blind area pedestrian anti-collision detection method based on an improved YOLOv8 model, a feature recombination up-sampling module in the model introduces a double-layer routing attention mechanism and a lightweight up-sampling operator, and the double-layer routing attention mechanism filters out non-critical value pairs in a coarse region of a feature map; the capability of extracting small target features in the feature map is improved; the lightweight upsampling operator performs block processing on an input feature map, performs feature recombination on each block by using convolution operation, and enhances a sensing range so that the upsampled feature map can be fused with more context information. The improved YOLOv8 model is tested and trained, so that the average accuracy of a detection result can be improved. A video set standard method is adopted to replace a manual method, and a photoelectric and video combined method is adopted to judge the response time, so that the response speed when an alarm occurs is improved, and traffic safety accidents are reduced.

The invention provides a lightweight target recognition device and method based on end-side deployment, and a computer storage medium, and the device comprises an image obtaining module which is used for obtaining an original image of a to-be-recognized target; the feature extraction module is used for extracting a first feature image from the original image; the analysis processing module is used for carrying out binarization inverse residual processing on the first feature image to obtain a second feature image; and the result output module is used for obtaining a target recognition result according to the second feature image and outputting the target recognition result. Binaryzation inverse residual processing is carried out through the analysis processing module, a binaryzation neural network, depth separable convolution and inverse residual are combined, the light weight of the model can be achieved on the premise that the depth and channel width of the model are not sacrificed, in other words, the model precision is improved under the condition that the extremely low model storage and calculation requirements are kept, and the calculation efficiency is improved. The problem that an existing lightweight target recognition model is poor in precision is solved.

An intermediate set of sensor fusion is disclosed. Computer-implemented methods and related aspects for generating a predicted output for an automated driving system of a vehicle are disclosed. A computer-implemented method includes: obtaining a first sensor data set originating from a first sensor; a second set of sensor data originating from a second sensor different from the first sensor is obtained. Further, the computer-implemented method includes: processing the first sensor data set using the first set of encoder networks; a second set of sensor data is processed using a second set of encoder networks. The computer-implemented method further includes fusing the one or more first sets of encoded features with the one or more second sets of encoded features using a fusion algorithm; and generating a prediction output based on the set of fused encoded features using a decoder network trained to address perception tasks or planning tasks of the vehicle based on the encoded sensor data features.

The invention relates to the field of data processing and artificial intelligence, in particular to an agricultural product purchase behavior prediction method and system based on a graph neural network. According to the technical scheme, the agricultural product purchase behavior prediction method and system based on the graph neural network comprise the agricultural product purchase behavior prediction method based on the graph neural network; compared with a traditional agricultural product purchase behavior prediction method based on a graph neural network, the method adopts a traditional graph neural network model to carry out feature extraction and relation modeling, and generally, training is directly carried out on a large-scale graph, so that the calculation and storage requirements are too high, the training efficiency is low, and the prediction efficiency is high. According to the agricultural product purchase behavior prediction method based on the graph neural network, a multi-scale variational graph auto-encoder is introduced, local and global structure features of the graph can be captured more comprehensively, node representation richness and diversity are improved, prediction accuracy is improved, calculation and storage requirements are reduced through a static sub-graph segmentation method, and the prediction efficiency is improved. And the training process is accelerated.

The invention relates to a low-code platform loop component rendering performance optimization method and system. The method comprises the following steps of 1, traversing and analyzing all style attributes of a component; 2, identifying a variable source and a use mode of the variable source in the style attribute; 3, analyzing whether the attribute value contains an expression or not based on the use mode of the variable, and dividing the attribute containing the variable into three types according to the dependency relationship: pure static attribute, dynamic attribute irrelevant to circulation and dynamic attribute relevant to the circulation; 4, analyzing and judging the relevance between the variable and the cyclic variable through the variable action range; step 5, extracting and processing the identified static pattern; 6, generating a unique class name for the extracted static style; and 7, constructing an independent style sheet file, grouping styles according to media query, and generating compressed CSS codes. According to the method, the style rendering performance is effectively optimized through static style extraction and dynamic style dependency analysis.