43results about How to "Training accurately" patented technology

The invention provides a time sequence data missing value filling method based on a diffusion model. The method comprises the following steps: constructing a filling model, training the filling model, and performing data missing value filling by utilizing the trained filling model. The filling model comprises a denoising diffusion module and an evaluation filling module; the de-noising diffusion module comprises an input data processing module, a residual connection block and an output integration module; the evaluation filling module comprises an incremental filling module and a sampling result integration module; according to the method, based on the current most advanced diffusion model architecture DiffWave, a self-supervised training mode is adopted, a filling target is generated through data of a Mask non-missing part, and meanwhile, an observable part information is used for assisting a diffusion model to predict noise, so that efficient and accurate training of the diffusion model is realized; according to the method, the time sequence missing value filling effect can be effectively improved, the problem of data missing in the time sequence is solved, and the data quality and value are improved.

The invention relates to a training method of an image generation model and an image generation method and device.The method comprises the steps that a first sample noise adding image and a first sample description text are obtained, the first sample description text is input into a preset text encoder for text encoding, and a first global semantic feature and a first character semantic feature are obtained; generating a first sample fusion condition feature based on the first global semantic feature and the sample time step; inputting the first character semantic feature, the first sample fusion condition feature and the first sample noise adding image into a preset noise predictor for noise prediction processing to obtain first prediction noise; performing denoising processing on the first sample noise-added image based on a denoising network to obtain a predicted image; and according to the difference between the first prediction noise and the sample noise label and the difference between the prediction image and the first sample image label, training the preset image generation model to obtain the target image generation model, thereby improving the semantic following capability and accuracy of the image.

The invention discloses a monocular video face reconstruction and driving method based on Gaussian splashing and a 3DMM model. The monocular video face reconstruction and driving method comprises the following steps: S1, converting each frame of a face RGB video into an explicit expression face grid based on the 3DMM model; s2, constructing a geometric enhancement module, and carrying out geometric alignment on eye and tooth regions by adopting an improved iterative nearest point algorithm; s3, initializing a 3D Gaussian ball, and driving the dynamic change of the 3D Gaussian ball according to the point change of the face grid under different expressions; s4, based on the characteristics of different face areas, adaptive optimization learning strategy coefficients are distributed; and S5, increasing confidence information for each 3D Gaussian ball, generating a confidence image through rendering, multiplying the confidence image by a rendered image, calculating high-dimensional feature similarity with the input image, and optimizing attribute parameters and rendering effects of the 3D Gaussian balls. According to the method, Gaussian splashing and a 3DMM model are combined, and high-precision reconstruction and driving of a monocular video face are realized.

The invention discloses a marine organism detection method, device and equipment and a storage medium, and relates to the technical field of target detection, and the marine organism detection method comprises the steps: obtaining a marine organism data set, and dividing the marine organism data set into a training set and a test set; introducing a Swin Transform network and an efficient multi-scale attention mechanism to improve the initial marine organism detection model, and obtaining a reference marine organism detection model; training and testing the reference marine organism detection model by using the training set and the test set to obtain a target marine organism detection model; and detecting the marine organisms in the marine organism image by using the target marine organism detection model. The target marine organism detection model obtained by the invention can more accurately detect the marine organisms and automatically identify and position the marine organisms, so that the detection efficiency is improved, and reliable data support is provided for scientific research and environmental protection.

The invention provides a computing power scheduling method and device for model training, a computer storage medium and electronic equipment, and relates to the technical field of computers. Comprising the following steps: in response to a received training task request for a to-be-trained model, dividing the to-be-trained model into a plurality of first model subtasks; aiming at the target model subtask, determining a heterogeneous computing power unit of a target resource type corresponding to the target model subtask from a heterogeneous computing power architecture according to a pre-configured computing power mapping table; the computing power mapping table comprises a one-to-one correspondence relationship between each computing task and the heterogeneous computing power unit of the corresponding resource type; and determining a target heterogeneous computing power unit from heterogeneous computing power units of the target resource type of each computing node in the heterogeneous computing power architecture, and allocating the target model subtask to the target heterogeneous computing power unit, so as to execute the target model subtask based on the target heterogeneous computing power unit. The method can improve the model training efficiency and the accuracy of the model prediction result.

The invention relates to the technical field of computers. The invention specifically discloses a vehicle insurance claim settlement amount prediction method and device, a medium and equipment. The method comprises the following steps: based on a plurality of local sample claim settlement data, using a predetermined initial prediction model to carry out federated model training in this round, and obtaining a first current model parameter; receiving a second current model parameter broadcasted by each non-target client; performing parameter aggregation processing based on the first current model parameter and each second current model parameter to obtain a current aggregation model parameter; at least based on the current aggregation model parameters, determining whether a preset training stop condition is met or not, if the training stop condition is not met, carrying out next-round federated model training until the preset training stop condition is met, and obtaining a target prediction model based on the current aggregation model parameters corresponding to the training round number; and predicting the claim settlement data of the target vehicle based on the target prediction model to obtain a target claim settlement amount. The accuracy of the prediction result can be improved.

The invention relates to the technical field of medical image analysis, in particular to a training method and system for a cataract recognition model and a storage medium. The method comprises the following steps: acquiring cataract eye image samples and non-cataract eye image samples in different stages; lesion region segmentation and image synchronous simulation enhancement are carried out on the cataract eye image samples and the non-cataract eye image samples in different stages to obtain cataract lesion enhanced image samples and non-cataract enhanced image samples; extracting local detail features of the lesion and extracting structural features of a mask pair, and constructing a multi-task convolutional neural network learning framework to train a cataract recognition model and optimize focus weighted loss; and obtaining a cataract related key feature set and carrying out online incremental learning updating to generate a cataract recognition online incremental updating model. According to the invention, the accuracy and robustness of cataract recognition model training can be improved.

The invention provides an intelligent feeder terminal fault on-line discrimination method, which comprises the following steps: firstly, collecting electric power parameters, switch states and environment data of each feeder terminal, converting the data into feature vectors through a feature extraction algorithm, and then utilizing a pre-trained abnormal probability identification model to identify the abnormal probability of each feeder terminal. Generating an abnormal probability value and calculating a confidence interval of the abnormal probability value, if the lower limit of the confidence interval exceeds a preset threshold value, judging that the terminal is abnormal, for the abnormal terminal, determining an adjacent terminal of the abnormal terminal according to a topological structure of a power grid, and then classifying feature vectors of the abnormal terminal and the adjacent terminal by using a pre-trained fault classification model; and finally, the method comprehensively considers the fault classification data of the abnormal terminal and the fault classification data of the adjacent terminal, and generates a fault comprehensive classification result of the abnormal terminal, so that intelligent online discrimination of the feeder terminal fault is realized, and improvement of the accuracy and efficiency of power grid fault diagnosis is facilitated.

The invention provides an unsupervised learning method applied to a millimeter-wave radar sensing task, electronic equipment and a storage medium, and can be applied to the technical field of millimeter-wave radars. The method comprises the following steps: acquiring a radar echo signal by using a target antenna array combination, and performing signal processing on the radar echo signal to obtain a positive sample pair; performing feature extraction on the positive sample pair by using an encoder and a momentum encoder to obtain a positive sample feature pair, and projecting the positive sample feature pair to obtain a positive sample projection pair; processing the positive sample projection pair and the negative sample projection pair stored in the queue memory bank by using a predefined comparison loss function to obtain a comparison loss value; performing parameter updating on the encoder based on the comparison loss value, performing parameter updating on the momentum encoder based on the encoder after parameter updating, and storing the positive sample projection pair into a queue memory bank; and iteratively carrying out the above operations in an unsupervised manner until a preset training condition is satisfied, and obtaining a pre-trained encoder.

The invention discloses a power grid operation weak point identification method and system based on a transmission and distribution network topological structure, and relates to the technical field of power grid operation, and the method effectively reduces the calculation cost, reduces the model complexity through a graph theory and a network simplification technology, optimizes the use of calculation resources, improves the data processing efficiency, and reduces the calculation cost. Therefore, accurate identification under limited calculation performance is realized; data quality indexes are read and evaluated, and data enhancement and dimensionless processing are performed, so that the integrity and accuracy of data are improved, and the stability of the system during data quality fluctuation is ensured; a deep learning algorithm is adopted, a model training process is optimized through cross validation and hyper-parameter tuning, and supervised learning is performed by using fault and operation historical data, so that dependence on a large amount of labeled data and computing resources is reduced, and efficient and accurate power grid operation weak point identification is realized.

The invention provides a motion enhancing device utilizing an exoskeleton robot, which comprises a hip exoskeleton, first drivers are respectively mounted on the left side and the right side of the hip exoskeleton, the output ends of the first drivers are connected with hip joint connecting plates, second drivers are mounted at the lower ends of the hip joint connecting plates, and the output ends of the second drivers are connected with leg exoskeletons. The first driver is used for controlling the hip joint connecting plate to swing in the left-right direction, and the second driver is used for controlling the leg exoskeleton to swing in the front-back direction; the leg exoskeleton comprises a thigh joint, the upper end of the thigh joint is connected to the output end of the second driver, a third driver is installed at the lower end of the thigh joint, the output end of the third driver is connected with a shank joint, and leg wearing pieces are installed on the thigh joint and the shank joint correspondingly. The second driver is used for controlling the thigh joint to swing in the front-back direction, and the third driver is used for controlling the shank joint to swing in the front-back direction.

The invention discloses a recommended verbal skill generation method and device, computer equipment and a storage medium, and belongs to the technical field of artificial intelligence. According to the method, the pre-training model comprising the adapter layer is constructed, flexible application of a general language model and a specific scene strategy is effectively combined, the adapter layer is used as a bridge for connecting the network layer of the pre-training model, and rapid training of the specific scene model is realized; first, a general language model is trained by using first training data, and then iterative training is performed by using second training data for a specific scene to obtain a verbal skill recommendation model. The invention further relates to the technical field of block chains, and training data is stored on a block chain network. According to the method, the pre-training model structure is adopted, the universal language model is firstly trained, then the corresponding verbal skill recommendation models can be respectively trained for various specific scenes, and verbal skill recommendation can be accurately and efficiently generated.

The invention belongs to the technical field of regional power consumption prediction, and aims to solve the problem that detailed power consumption data of each region can be directly exposed by an existing regional power consumption prediction method. The distributed calculation-based large-area power consumption prediction method comprises the following steps of: dividing a large area of which the power consumption is to be predicted into a plurality of small areas, and performing secondary division on the divided small areas into a plurality of areas; constructing a power consumption prediction model of each district, and updating the power consumption prediction model of the core district by using the encryption loss data and gradient data of the other districts as the power consumption prediction model of the small district to which the core district belongs; the core small area receives encryption model parameters of the other small areas and generates model gradients corresponding to the small areas; and constructing a power consumption prediction model of the whole large area through encrypted data transmission and model updating iteration between the core small area and the other small areas. According to the method, the leakage problem in data transmission can be avoided, and the prediction accuracy and safety are improved.

The invention relates to the technical field of image recognition, in particular to a roadway crack recognition method and system based on image recognition. The method comprises the following steps: acquiring roadway structure data; performing regional camera deployment according to the roadway structure data to obtain roadway camera deployment data; performing roadway image acquisition according to the roadway camera deployment data to generate a roadway time sequence image set; performing roadway form change analysis on the roadway time sequence image set to generate roadway time sequence change data; performing surface crack area identification on the roadway time sequence image set through the roadway time sequence change data to generate a roadway surface crack area image; and performing surface crack trend analysis on the roadway surface crack area image to generate surface crack texture trend data. According to the method, through dynamic change detection, hidden crack identification, crack classification and analysis and crack age deduction, the identification accuracy and comprehensiveness are improved.

The invention provides a speech synthesis model training method, a speech synthesis method, a speech synthesis device and speech synthesis equipment, and belongs to the technical field of computers. The first speech synthesis model comprises a language network and a diffusion network; the method comprises: processing a sample text and a sample audio through a language network to obtain a first audio feature, the sample text and the sample audio being consistent in content, the first audio feature comprising a content feature of the sample text and an acoustic feature of the sample audio; performing noise prediction on the sample audio added with the reference noise based on the first audio feature through a diffusion network to obtain predicted noise; determining a first training loss based on the reference noise and the predicted noise; and updating parameters of a language network and a diffusion network in the first speech synthesis model based on the first training loss to obtain a second speech synthesis model. According to the technical scheme, the purpose of end-to-end training of the speech synthesis model is achieved, the training steps are simplified, and the training efficiency of the model is improved.

The invention provides a training method and device for an anomaly detection model, and relates to the computer technology. According to one specific embodiment, the method comprises the steps that package data of historical packages are collected, the package data comprise the category and weight information of each article in the historical packages, and the historical packages comprise single-category packages and multi-category combined packages; identifying abnormal categories and / or abnormal multi-category combinations from the package data; and determining a negative sample and a positive sample according to the abnormal category and / or the abnormal multi-category combination so as to obtain an anomaly detection model. According to the embodiment of the invention, the method is based on the constructed positive sample and negative sample of the article category dimension, and the constructed positive sample and negative sample are from the package data of the single-category package and the multi-category combined package, thereby achieving the accurate and supporting model training of the single-category package and the multi-category combined package. And a reliable, efficient and high-applicability detection method is provided for subsequent abnormal package detection based on the model.

The invention relates to the technical field of electric power operation safety, and discloses an electric power operation risk early warning system, which comprises a plan acquisition module used for inputting a work plan of pre-execution operation and obtaining a main line switch related to the work plan; the text processing module is used for performing word vector processing on the pre-execution work plan and the historical work implementation content to obtain a first word vector related to the historical work implementation content, a second word vector related to the pre-execution work implementation content and a third word vector related to the route and the section; the judgment module is used for analyzing and judging a corresponding high-risk point according to the second word vector; the matching module is used for matching corresponding prompt words, new energy user information and involved transformer area information in the violation library according to the high-risk points; and the communication module is used for sending the prompt words, the new energy user information and the transformer area information to a worker. According to the invention, high-efficiency management is realized under the condition that the burden of workers is not increased.

The present application provides an intelligent online feeder terminal fault identification method, which first collects power parameters, switch states and environmental data of each feeder terminal, converts these data into feature vectors through a feature extraction algorithm, and then uses a pre-trained abnormal probability recognition model to generate an abnormal probability value and calculate its confidence interval. If the lower limit of the confidence interval exceeds a preset threshold, the terminal is judged to be abnormal. For the abnormal terminal, the method determines its adjacent terminals according to the power grid topology structure, and then uses a pre-trained fault classification model to classify the feature vectors of the abnormal terminal and the adjacent terminal respectively to obtain their respective fault classification data. Finally, the method comprehensively considers the fault classification data of the abnormal terminal and the adjacent terminal, generates a comprehensive fault classification result of the abnormal terminal, and realizes intelligent online identification of feeder terminal faults, which helps to improve the accuracy and efficiency of power grid fault diagnosis.

The invention discloses a training method and system for a medical image recognition model, and particularly relates to the technical field of image processing, and the method comprises the following steps: feature extractor training: obtaining a training set sample image, and inputting the training sample image into a network model; calculating an error between the output of the output layer and the target label through forward propagation of the network model; judging whether the network model is converged or not, and if the network model is converged, obtaining a network model trained by the feature extractor; and classifier training: inputting the training image into the feature extractor after feature extraction training is completed. The classifier adopted by the invention starts from the probability density function, the regression value of the dependent variable to the independent variable is directly calculated by using the training sample, iterative operation is not needed, the consumed time is short, the calculation efficiency is ideal, the classification processing capability of the classification model is ensured, the convergence speed of the model training process is high, and the method is suitable for large-scale popularization and application. And the recognition precision, the real-time performance and the accuracy of the recognition model are improved.

The invention discloses a dynamic modeling analysis system for combustion atmosphere data of a biomass oxygen-enriched combustion boiler, and relates to the technical field of dynamic modeling analysis. The system comprises a working condition self-adaptive feature extraction module, a time sequence maintaining model training module, an online self-adaptive correction module and an incremental model updating module. Through a working condition adaptive feature extraction module, a time sequence maintenance model training module, an online adaptive correction module and an incremental model updating module, adaptive feature vectors are generated through feature extraction, a basic model is obtained through training, and online correction and incremental updating are performed by combining feature vector verification threshold judgment. And finally, iterative upgrading is completed through an incremental learning fine tuning model, the reliability of dynamic modeling analysis of the combustion atmosphere data of the biomass oxygen-enriched combustion boiler is improved, and the problem that in the prior art, the reliability of dynamic modeling analysis of the combustion atmosphere data of the biomass oxygen-enriched combustion boiler is low is solved.

The invention relates to the technical field of face recognition, and discloses a personnel information automatic recognition system based on a support vector machine and a black-wing plinergic algorithm, and the system comprises an acquisition module which is used for obtaining an initial image of a to-be-detected person; the preprocessing module is used for preprocessing the initial image to obtain a face image of the to-be-detected person; the comparison module is used for inputting the face image into a support vector machine trained based on a black wing plinualism algorithm for classification so as to obtain face information of the to-be-detected person; the mapping module is used for correspondingly finding out personnel information of the to-be-detected personnel according to the face information; the permission judgment module is used for judging whether the personnel information of the to-be-detected personnel has access permission or not; and the alarm module is used for executing corresponding alarm operation when the permission judgment module judges that the personnel information does not have the access permission. By means of the recognition system, checking of safety inspectors on the access information of the constructors is reduced.

The invention relates to a data processing method and device, electronic equipment and a storage medium, and the data processing method comprises the steps: obtaining first putting accumulated data corresponding to first user equipment, and the first putting accumulated data is determined according to resources accumulated by first type information put to the first user equipment; determining whether the first release accumulated data reaches a first preset threshold value or not; if the first release accumulated data reaches a first preset threshold value, performing channel attribution on the first user equipment; and if the channel attribution is successful, generating return data for channel return based on the user related information of the first user corresponding to the first user equipment and the release related information of the first type of information. According to the embodiment of the invention, the type and quantity of the returned data can be enriched, user groups which are sensitive to prices but willing to watch advertisements are prevented from being filtered out, the diversity of the returned data is improved, and the finally determined user related information of the first user and the release related information can be more accurate.

The invention belongs to the technical field of rehabilitation assistive devices, and particularly relates to a limb rehabilitation assisting device and a limb rehabilitation training mechanism, the limb rehabilitation assisting device comprises a driver, an angle measuring part and a force measuring part, the driver can be bent under the action of external driving force, auxiliary training can be conducted on limbs, and the force measuring part is used for measuring the angle of the limb. The bending angle of the driver can be controlled by adjusting the driving force; the angle measuring part is used for accurately measuring bending data of each limb joint by arranging a strain gauge or an attitude sensor on a base; and the force measuring part measures the force data of each limb region by measuring the pressure of the air bag group corresponding to each limb region. The whole device is simple in structure, comprehensive and accurate in measurement of movement data of all limbs, good in comfort and capable of adapting to various rehabilitation strategies, the rehabilitation effect is improved, and precise rehabilitation is achieved.

The present disclosure provides a parameter prediction model training method, a parameter prediction method and device, comprising: obtaining first target data, the first target data comprising: historical application program interface (API) request parameter data and attribute data corresponding to the historical API request parameter data, the attribute data being used to describe parameter declaration of the historical API request parameter data, the parameter declaration comprising: parameter name definition and parameter value; determining a pre-training model based on the first target data, the pre-training model being used to perform first parameter prediction on an API request; obtaining second target data, the second target data being used to describe question and answer data pairs corresponding to a plurality of user questions, each user question being derived from the API request parameter data and the attribute data of the API request parameter data; and determining a parameter prediction model based on the second target data and the pre-training model. Thus, API parameter prediction is accurately performed.

The present invention discloses a method and system for image compression sensing reconstruction based on PIPO‑Net. The present invention obtains a two-dimensional image; preprocesses the two-dimensional image; inputs the preprocessed two-dimensional image into a penalty-based independent parameter optimization deep expansion network PIPO‑Net to obtain a compression sensing measurement value of the two-dimensional image, and finally outputs a reconstructed image. PIPO‑Net can obtain the average value of the original image and the image after de-averaging by a de-averaging sampling method. The sampling values ​​of the image after de-averaging are used for training, thereby improving the performance and stability of the network and better capturing the features of the image. In order to solve the problem of information loss, the present invention designs a high-frequency information supplement block and uses a wavelet high-frequency loss function for training to better retain the details and texture information of the image.

The invention discloses an intelligent periodontal exploration and diagnosis training device. The device comprises a tooth oral cavity model, a periodontal probe and a processing module, the tooth oral cavity model is provided with a gingival model, a tooth model is arranged at the periodontal pocket position of the gingival model, the gingival model is connected with the top of the tooth model, the peripheral veneers of the gingival model and the tooth model are separated, periodontitis characteristics exist around the periodontal pocket of the tooth model, and a first micro sensor is arranged in the tooth model; a second micro sensor and a pressure sensor are arranged at the tip of the working end of the periodontal probe; the processing module is connected with the tooth oral cavity model and the periodontal probe and is used for receiving and processing signals of the first micro sensor, the second micro sensor and the pressure sensor; by calculating the relative position and the relative angle of the first microsensor and the second microsensor, the correct measurement site, the probing depth and the probing angle of the periodontal probe around teeth are determined, measurement data are obtained according to signals of the pressure sensor, and the use force of the periodontal probe is determined. The result is displayed to the user in real time, so that the user can know the correct site, depth, angle and use force of periodontal exploration.

The utility model discloses a fitness equipment structure with an outer rotor motor. The fitness equipment structure comprises a base frame; the force application unit is fixed on the base frame; the motor comprises an inner stator and an outer rotor, and the outer rotor is rotatably arranged on the base frame and surrounds the inner stator; the transmission part is connected with the force application unit and the outer rotor; a code disc rotates along with the outer rotor; the coding unit is fixed on the inner stator or the base frame and is used for reading rotation information of the coded disc; the control unit is connected with the motor and the coding unit and is used for receiving the rotation information; the control unit controls an electromagnetic field of the inner stator to enable the outer rotor to have rotation resistance, and the control unit can selectively change the intensity of the electromagnetic field according to a preset program and the rotation information to enable the rotation resistance to be changed.

The invention relates to a meta-learning rapid adaptation method and device, equipment, a medium and a program product, and relates to the technical field of artificial intelligence. The method comprises the following steps: identifying similarity among tasks by analyzing task characteristics, and performing task grouping and specific initialization of model parameters according to the similarity so as to realize rapid adaptation to new tasks. The adaptive speed and the learning efficiency are remarkably improved while the generalization ability of the model is maintained. In addition, in-depth understanding of the rapid adaptive capacity of the meta-learning algorithm is provided theoretically, and high efficiency and reliability of the meta-learning algorithm are shown in practical application. According to the invention, the training of the multi-modal meta-learning model becomes faster and more accurate, and the application range and practical value of the meta-learning technology are greatly expanded.

The invention discloses a honeycomb-free large-scale MIMO unmanned aerial vehicle associated power control method and equipment based on reinforcement learning, and a medium, and relates to the technical field of wireless communication. The method comprises the following steps: acquiring parameters of a communication system containing multiple unmanned aerial vehicles and access points; a cellular-free large-scale MIMO transmission model is constructed; the method comprises the following steps: determining a state space, an action space and a reward function based on a cellular-free large-scale MIMO transmission model, and respectively constructing intelligent agents for an uplink power control coefficient optimization problem of an unmanned aerial vehicle in a system and an access point unmanned aerial vehicle clustering optimization problem; training the intelligent agent by using a depth deterministic strategy gradient reinforcement learning algorithm; and obtaining an uplink power control coefficient of the unmanned aerial vehicle in the system and an unmanned aerial vehicle clustering result of each access point by using the intelligent agent based on an optimized uplink power distribution strategy and an access point unmanned aerial vehicle clustering strategy. According to the method, efficient joint optimization of unmanned aerial vehicle access point association and uplink power control can be realized under a rapidly changing channel condition, and the method has relatively high practical practicability under a low-altitude economic background.

The invention relates to a model training method, a measurement method, a device, equipment and a program product. The model training method comprises the steps of performing edge detection on a to-be-recognized image based on an edge detection model to obtain an edge detection result; the to-be-identified image is a sample image to be labeled with a label; generating an initial label of the to-be-recognized image according to the edge detection result; adjusting the initial label according to the real edge condition of the to-be-recognized image to obtain a target label of the to-be-recognized image; and training a candidate detection model based on the sample image and the target label of the sample image until a training condition is reached, and obtaining a target detection model for edge detection. By adopting the method, the label generation efficiency and accuracy can be improved.