19results about How to "Increase training speed" patented technology

This invention discloses a target recognition method, docking system, and docking method in a space environment. By adjusting the output intensity of the sunlight simulator's light source and the position and angle of the simulated work equipment, different types of simulated target images are obtained, forming a target photo library. The simulated target images are processed into small target images and then placed into a small target image library. This small target image library is merged with the target photo library augmented by GAN data, and divided into training and testing libraries proportionally. An SA-YOLOX network is constructed. The SA-YOLOX network is trained using the training library to obtain an SA-YOLOX target detection model. The augmented target photo library is tested using the SA-YOLOX target detection model and the testing library to obtain test results. Based on the test results and model metrics, an evaluation result is output. This invention effectively overcomes the performance index problem of the YOLO series algorithms in solving the special situation of high-contrast image target recognition in complex space environments.

The application provides a traffic classification method and system based on federal semi-supervised learning. The method comprises the following steps: constructing an unlabeled traffic dataset and a labeled traffic dataset; a center server decomposes a global model into supervised learning parameters and unsupervised learning parameters and initializes the parameters; the parameters and an auxiliary agent are sent to each client; each client performs unsupervised training based on the supervised learning parameters, the unsupervised learning parameters and the auxiliary agent by using a local unlabeled traffic dataset, uploads the difference between the unsupervised learning parameters to the center server; the center server aggregates and updates each unsupervised learning parameter; supervised training is performed by using a local labeled traffic dataset, and the difference between the supervised learning parameters and the difference between the unsupervised learning parameters are sent to each client; a new auxiliary agent is obtained based on nearest neighbor search, and the new auxiliary agent is sent to each client when a set sending condition is met; the two steps are iteratively executed until a stop condition is met.

This invention discloses a method, system, device, and storage medium for predicting oil pipe corrosion rate based on a PCA-PSO-SVR hybrid model, specifically including the following steps: S1 Collecting corrosion detection data and operating condition parameters of the oil pipe to form a dataset; S2 Preprocessing the dataset; S3 Using the PCA model to perform dimensionality reduction on the dataset and extracting the main features affecting the corrosion rate; S4 Initializing the parameters of the PSO model; S5 Optimizing the parameters of the SVR model using the PSO model; S6 Constructing a corrosion rate prediction model based on the optimized SVR model; S7 Inputting the preprocessed dataset into the corrosion rate prediction model to obtain the prediction result and complete the prediction.

This application discloses a federated large language model training method, apparatus, and device. The method includes: constructing an initial mask matrix for each original model parameter in the initial large language model; splitting the initial mask matrix to obtain initial mask blocks based on the current pruning round; determining the current mask block corresponding to the initial mask block based on the mask reconstruction dataset, a predetermined sparsity rate, the original model parameter matrix corresponding to the initial mask block, and the inverse Hessian matrix of the original model parameter matrix, and determining the current model parameters corresponding to each non-zero, non-target element in the current mask block to obtain the current model parameter matrix; determining the pruned large language model based on the current model parameter matrix corresponding to each pruning round; and optimizing and training the pruned large language model based on a first optimized dataset to obtain the target large language model. This application can perform reasonable and accurate model pruning, simplify the model structure, and improve the model training speed.

The application discloses a quad-rotor formation obstacle avoidance control method based on an improved DDPG algorithm, and belongs to the technical field of unmanned aerial vehicle formation control. The method adopts an improved DDPG reinforcement learning algorithm to plan an obstacle avoidance path of the quad-rotor. When the improved DDPG reinforcement learning algorithm is executed, the priority weight of each quadruple experience in the experience replay pool is initialized. After the quadruple experience in the experience replay pool is sampled and trained, the priority weight of each quadruple experience is recalculated based on a TD error, and a Sum_Tree structure is updated. The method effectively alleviates the training instability caused by hyperparameter sensitivity, the misleading of policy updating caused by overestimation of Q values, and the problem that key experiences are not sufficiently learned, accelerates the convergence speed of the quad-rotor formation obstacle avoidance training, and improves the obstacle avoidance effect.

The application discloses an aircraft cluster cooperative non-cooperative target capturing method driven by entropy regularization, applies the entropy regularization thought to reinforcement learning, optimizes the updating method of the parameters of an evaluation network and a policy network, so that a UAV can obtain a more optimal maneuvering strategy, and the UAV performs actions according to the maneuvering strategy, and realizes the interception of an attacking aircraft. The aircraft cluster cooperative non-cooperative target capturing method driven by entropy regularization can not rely on a deterministic strategy in the interception process of multiple UAVs, but adopts a random strategy, avoids the training from falling into a local optimal point, and the actions of the aircraft are randomized as much as possible while the aircraft completes the task, so that the exploratory and robustness can be practically improved.

The application discloses a polar code construction method based on genetic algorithm acceleration convergence, which comprises the following steps: step 1, setting parameters; step 2, initial genetic algorithm population; step 3, selecting two parents from the population by using a roulette algorithm; step 4, crossing the two parent information bits to generate offspring; step 5, mutating the unlocked information bits of the offspring by using the non-locked frozen bits; step 6, adding the mutated offspring to the population, selecting the optimal one according to the fitness, updating the population and recording the population information; step 7, repeating steps 3 to 6 until the number of genetic iterations reaches a number of spans, judging whether the optimal fitness decreases to a limit or not, if yes, reducing the number of locked bits, temporarily releasing the channel, repeating steps 3 to 6 until the number of genetic iterations reaches a minimum number of spans, if the optimal fitness decreases and the temporarily released channel changes, updating the channel locking condition according to the temporarily released channel, if no, directly repeating steps 3 to 6; and step 8, ending when the genetic iteration stopping condition is met. The application can solve the problem of slow convergence speed by dynamically locking the channel.

This application belongs to the field of artificial intelligence and relates to a training method for a multilingual model, including: acquiring a preset monolingual pre-trained model; constructing a first initial multilingual model based on the monolingual pre-trained model; adjusting the first initial multilingual model based on the first parameters of the first encoder within the monolingual pre-trained model to obtain a second initial multilingual model; acquiring pre-collected training data; and finally training the second initial multilingual model using the training data to obtain a target multilingual model. This application also provides a training device, computer equipment, and storage medium for a multilingual model. Furthermore, this application relates to blockchain technology, allowing the target multilingual model to be stored in a blockchain. This application, based on the use of a monolingual pre-trained model, effectively reduces the training cost of the target multilingual model and improves the training and generation efficiency of the target multilingual model.

This invention discloses a grouped point cloud-driven arm-hand operating system. The arm-hand operating system includes a VR-based teleoperation data acquisition module, an environmental perception and encoding unit, and a two-layer feature fusion diffusion denoising model. Specifically: the VR-based teleoperation data acquisition module consists of a posture recognizer, a posture mapper, and a data storage module; the environmental perception and encoding unit consists of a point cloud grouper, a point cloud encoding module, and a state integrator; and the two-layer feature fusion diffusion denoising model consists of a two-layer feature fusion module and a diffusion denoising network. This arm-hand operating system can respond to environmental changes in real time, achieving comprehensive scene perception while quickly formulating task execution plans. This data acquisition-algorithm deployment collaborative framework effectively solves the challenge of balancing cost, speed, and accuracy in intelligent arm-hand operation, providing a cost-effective solution for embodied intelligence, robotic operations, and other fields.

This application discloses a training method and apparatus for a joint fraud detection network, a joint fraud detection method and apparatus, a joint fraud detection system, a computer device, a computer-readable storage medium, and a computer program product. The training method includes: acquiring unilateral encrypted data sent by each participant, and performing embedding calculations using each unilateral encrypted data in at least one graph embedding layer or in a graph embedding layer connected to a fraud classification sub-network to obtain joint embedding data corresponding to multiple participants; performing fraud detection using the fraud classification sub-network and the joint embedding data of multiple participants to obtain fraud detection results corresponding to multiple participants; and adjusting the network parameters of the fraud classification sub-network using preset fraud labels of multiple participants and fraud detection results of multiple participants until the training stop condition is met.

This invention discloses a low-light image enhancement method based on random region concealment and reconstruction. The network consists of two stages. In the first stage, the network is trained on a well-lit image dataset to reconstruct the noisy image after random regions are concealed, resulting in a denoising and reconstruction feature encoder for low-light noisy images. In the second stage, the network inherits the feature encoder model parameters from the first stage and inputs the extracted different features into a multi-scale feature fusion module guided by the reconstructed features. This allows the network to retain more image detail features while denoising. This invention utilizes a richer dataset of well-lit images, saving time and manpower by capturing a large number of real paired low-light images for network training. By fusing features extracted from different feature encoders, the network can quickly adapt to real low-light images under complex lighting conditions, achieving good low-light enhancement and denoising effects.

A multi-objective optimization method for a two-dimensional pump damping groove based on a GA-BP neural network includes the following steps: S1, establishing design variable parameters in the two-dimensional pump damping groove structure; S2, constructing a simulation model of a two-dimensional piston pump and using this model to simulate and obtain the performance parameters of the hydraulic pump; S3, constructing a GA-BP surrogate model, the process of which is as follows: S31, generating a sample dataset with good parameter space filling performance using Latin hypersquare sampling based on the design variable parameters; S32, establishing a BP neural network model and using a genetic algorithm to optimize and train the model based on the sample dataset, finally obtaining a BP neural network model optimized by the genetic algorithm; S4, using the GA-BP neural network surrogate model as the fitness function, performing multi-objective optimization of the damping groove structure parameter space using a fast non-dominated sorting genetic algorithm to obtain the optimal parameter solution set. This invention significantly improves the comprehensive fluid performance of two-dimensional pumps.

The application discloses a pathological image color restoration method and a scanner based on deep learning, and the method comprises the following steps: reducing the resolution of input data and target data based on a bilateral grid downsampling technology; training a neural network model for the first time by using the input data and the target data after the resolution is reduced; improving the resolution of the input data and the target data after the resolution is reduced based on a bilateral grid upsampling technology; and training the neural network model for the second time by using the input data and the target data after the resolution is improved, so that the training speed of the neural network model is improved, and the staining effect of the pathological image collected by a digital pathology scanner is improved.

This invention provides an apparatus and method for generating an entity relation extraction model. The apparatus receives text to be labeled and, based on multiple fields in the text and entity information and relation information in an entity relation database, generates at least one entity information to be labeled for each field and at least one relation information to be labeled for each field. The apparatus labels the at least one entity information and the at least one relation information to be labeled for each field according to a modified labeling format. The apparatus generates multiple combinations from the at least one labeled entity information and the at least one labeled relation information and stores them in the entity relation database. Based on a pre-trained language model, the apparatus inputs these combinations into the pre-trained language model to generate an entity relation extraction model.

This application discloses a method and apparatus for use in a node during model training in wireless communication. The node first receives at least a first dataset and a second dataset; then receives or sends a first signaling; and sends first reporting information; the first dataset is used for training a first model, and the second dataset is used for training a second model; the first reporting information depends on the second model; the first signaling is used to indicate whether the second model depends on the first model. This application improves the reliability of model training and reduces the time consumed by model training by optimizing the training method of AI / ML models, thereby improving transmission performance and spectral efficiency.

This invention introduces a deep learning method for EDR estimation based on aircraft response characteristics. By combining staged calculation and neural network learning, this method provides an accurate and reliable EDR estimation approach. Unlike traditional EDR estimation methods, the method described in this invention features clearer stage division and dynamic adjustment capabilities. The method includes the following steps: systematically integrating flight data collected from Quick Access Recorders (QARs) of all flights nationwide, constructing time windows, and performing intelligent hierarchical sampling; reading flight data from CSV files and dividing flight stages using a state machine; calculating takeoff and landing stages and cruise stages separately, using raw QAR data to calculate high-level features; finally, training the calculated high-level features using a ConvLSTM-ELM neural learning network model to obtain the optimal parameters of the model. After neural network training and learning, this method can accurately estimate the EDR value of the entire flight route, improving the accuracy and real-time performance of EDR calculation using recorded QAR data in the aviation field.

The application provides a cable-supported photovoltaic support vortex vibration response model construction method, and belongs to the technical field of cable-supported photovoltaic support detection. The application extracts modal parameters through fast Fourier transform, performs non-uniform sampling in a six-dimensional parameter space based on a sparse grid point distribution method, performs computational fluid dynamics numerical simulation on the sampling points to establish an aerodynamic force characteristic database, trains a vortex vibration response prediction model to output a vortex-induced lift coefficient, adopts a modal selection criterion based on energy proportion to retain key modes, adopts a symplectic geometry structure-preserving numerical integration algorithm to perform time-domain integration and solving, introduces a corotational coordinate system to perform geometric nonlinear correction, adopts an immersed boundary-lattice Boltzmann coupling method to update a fluid-structure coupling interface, adjusts an artificial intelligence model learning rate through an adaptive adjustment factor, judges convergence, and outputs vortex vibration response results, thereby solving the problem of excessive consumption of resources in vortex vibration response prediction and calculation of cable-supported photovoltaic supports.

Disclosed are a method for training a neural network, an optimizer for updating parameters of a neural network, an electronic device, and a computer-readable storage medium. Various embodiments of the present disclosure combine the advantages of first-order momentum information and second-order momentum information, fully utilize the derivative information and curvature information of a loss function at each time step in the process of training a neural network, and release the memory for storing the derivative information after obtaining the first-order momentum information, thereby being capable of improving the training speed of the neural network while reducing the memory overhead.