109 results about "Data aided" patented technology

The invention provides a PC (Personal Computer) end quantitative model reasoning method. The method comprises the following steps: S1, loading input information; s2, deserializing the model; s3, operator calculation is carried out; and S4, storing the data, and selecting a dump data result. The quantitative model reasoning on the PC end is used for simply reasoning the model, verifying the result of the model and obtaining the data of the network layer of each stage reasoned in the model, and the PC end reasoning is used for correcting the data, assisting the quantitative model, reasoning the float model, post-quantitative model and training the quantitative model and verifying the correctness of the data. According to the method, model reasoning is convenient and effective, model data verification is facilitated, and data information of a quantitative model is provided.

The invention discloses a remote auxiliary diagnosis method and system based on multi-modal medical data, and relates to the technical field of auxiliary data processing, and the method comprises the steps: obtaining a customized remote auxiliary diagnosis task containing sample multi-modal medical data and corresponding label data, and then carrying out the remote auxiliary diagnosis according to the customized remote auxiliary diagnosis task, the target deep learning model is optimized by adopting a multi-strategy hybrid optimization algorithm, the optimized target deep learning model is deployed on a cloud server, and finally, the deployed target deep learning model is called to recognize the to-be-recognized multi-modal medical data uploaded by the medical staff to obtain a data auxiliary processing result. According to the invention, the method can achieve the custom remote auxiliary diagnosis task of a medical worker, assists the medical worker in improving the processing efficiency of multi-modal medical data, employs a multi-strategy hybrid optimization algorithm to optimize the target deep learning model, and can improve the accuracy of processing the custom remote auxiliary diagnosis task.

The invention provides an atlas updating method and device based on a large model, electronic equipment and a medium, information is extracted from large-scale text data by utilizing a natural language processing model, and the method mainly comprises the steps of data collection and preprocessing, entity extraction, relation extraction, atlas updating, model evaluation and iteration and the like. According to the method, an automatic crawler technology, multi-language text processing, transfer learning, visual data auxiliary entity extraction, relation extraction based on a pre-training language model and other technologies are comprehensively applied, and multiple beneficial effects of graph timeliness, multi-language adaptability, model generalization ability improvement, extraction accuracy improvement and the like are achieved. The atlas updating and continuous optimization process is optimized by means of an incremental updating strategy, a dynamic learning system, user feedback data, a conflict resolution strategy and the like. In conclusion, according to the method, the accuracy, the integrity and the user friendliness of the atlas are remarkably improved, meanwhile, the dependence on the annotation data is reduced, and the overall calculation efficiency and the model generalization ability are improved.

The invention relates to the technical field of government affair data processing and intelligent decision making, and particularly discloses an index management-based data aided analysis, prediction, research and judgment system, which is characterized in that a multi-level index management system is constructed, an association relationship between indexes is defined, and a dynamic update threshold value is configured; the method comprises the steps of collecting original operation data from multiple government affair data systems, and generating a standardized data set through quality verification and content integration; establishing a multi-index cooperative monitoring network by performing correlation analysis and abnormal feature recognition on the data set; deducing a plurality of development paths in a future time period based on state transition analysis of a historical sequence, and generating a prediction result including trend pre-judgment and risk level; and finally, constructing a decision scheme tree, generating a visual analysis report through multi-dimensional scoring, continuously optimizing parameters based on implementation feedback, and forming a closed loop of analysis prediction and decision support. According to the invention, deep fusion and intelligent research and judgment of cross-domain government affair data are realized.

A compute server of a distributed cloud computing network receives an inference request. The inference request triggers execution of code that is related to an AI application that interacts with the inference request and causes its input to be run through an AI model. Helper code executing on the compute server transmits an inference request to the AI model, where this inference request includes a prompt and one or more function definitions, and where the AI model is executing in the same datacenter as the compute server. The helper code receives a result of the AI model executing the inference request, where it includes structured data for calling a function. The helper code calls the function using the structured data. The helper code receives a response to the called function. The code processes a response to the inference request based at least on the received response to the called function.

Methods, device and data stream are provided to encode unaligned multi-view plus depth images associated with assistance metadata for generating depth maps aligned with the unaligned color views. At the encoding stage, for an unaligned color view, the contribution of each depth map is evaluated, and the objects represented are mapped. Picture patches are cut to be homogeneous in terms of objects and contributing views. Metadata describing this homogeneous cut are encoded together with the atlas. At the decoding stage, to generate a depth map aligned with the unaligned color view, only the depth maps referenced in the assistance metadata are warped.

The invention discloses a sea wave height inversion algorithm based on C-band SAR, and relates to the field of sea wave height inversion, and the algorithm comprises the steps: selecting mode data in offshore satellite data; eRA5 wind field data and wave height data are selected, mode data are processed after time matching, SAR image sub-images are selected, and then space matching is carried out on the SAR image sub-images and the ERA5 wind field data and the wave height data; obtaining training data by using the SAR image sub-image after space matching, the SAR data auxiliary information file, the ERA5 wind field data and the wave height data to construct a training data set; and constructing an inversion model, inputting the training data into the inversion model, and finally obtaining the SAR sea wave height. Ocean wave height inversion is carried out by adopting complete polarization data, the precision is high, the application threshold is high, and the application range is small; meanwhile, the problems of inversion precision and applicability are comprehensively considered, more SAR information is used for sea wave inversion, and meanwhile the application range of an inversion model is guaranteed.

An assisted driving system includes: inputs configured to receive sensor data, telematics data, and driver response data; an assisted driving module configured to, based on the sensor data and telematics data generate first non-urgent notification messages; and a ranking module configured to rank the first non-urgent notification messages based on the driver response data. The assisted driving module is configured, based on the ranking of the first non-urgent notification messages, to output via at least one output device the first non-urgent notification messages such that saliency of each of the first non-urgent notification messages is based on the respective ranking of that non-urgent notification message.

The application belongs to the field of OFDM channel estimation, and relates to a data-assisted short-wave OFDM sparse channel Turbo iteration estimation method, which comprises the following steps: S1, receiving a signal transmitted through a channel and processing the signal to obtain a receiving end symbol matrix Y; S2, performing channel estimation and equalization on Y to obtain an estimated channel and an OFDM signal pair, and performing Turbo iteration on the estimated channel and the OFDM signal pair to obtain a symbol variance; S3, selecting a virtual pilot in the signal according to the symbol variance; S4, performing channel estimation and equalization on Y according to the virtual pilot to obtain a current estimated channel and an OFDM signal; S5, performing Turbo iteration on the signal to obtain a current symbol variance; and S6, if a preset maximum iteration number is reached, obtaining a final estimated channel; otherwise, returning to step S3 to select a virtual pilot in the signal according to the symbol variance; and the application selects a virtual pilot in the signal according to the symbol variance, and uses the virtual pilot to perform estimation again, so that the spectral utilization rate is improved and the channel estimation performance is improved.

The invention discloses a medical data aided decision-making method and device fusing multi-modal data, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining the original medical data of a to-be-predicted patient; aligning semantics among the data of various modalities, and generating personalized feature vectors of the patient to be predicted; determining an activation weight according to the feature information of the personalized feature vector; according to knowledge data in the original medical data, a clinical rule is converted into a fine-adjustable constraint condition, and according to the constraint condition and the activated identification expert layer, feature analysis is performed on the personalized feature vector to obtain an illness state analysis state vector; assembling cue words, and reasoning the cue words; and generating a plurality of candidate reasoning paths, determining an optimal reasoning path in the candidate reasoning paths, and generating an optimal reasoning result by using the optimal reasoning path. The invention provides an LLM for performing dynamic verification and adjustment by fusing multi-dimensional contrast signals in real time in the process of medical data analysis and auxiliary decision generation.

The application provides an electromagnetic safety restoring force digital twin training method and system, constructs a test twin space, determines a test equipment group based on electromagnetic test data, and associates real interference data; determines interference dimensions based on equipment properties of each test equipment in the test equipment group, generates false interference data according to the interference dimensions; updates the state of the test equipment according to the real interference data, generates actual data, and directionally replaces the running data of the corresponding dimensions based on the false interference data to generate false data; collects calling information of the training end to the auxiliary plug-in, generates error area guide data according to the comparison information of the calling information, the actual data and the false data, and the auxiliary plug-in one-to-one corresponds to the running data of each dimension, so that the real interference data and the false interference data can be fused, the training complexity is improved, and the disposal efficiency of an operation and maintenance personnel for an electromagnetic safety event is improved.

The invention discloses an over-the-horizon railway safety management method and system. The over-the-horizon railway safety management method comprises the following steps: collecting video data of each key road section through a ground video monitoring subsystem; when there is no line node, the vehicle-mounted terminal carries out sampling and calculates the distance between the corresponding line node and the key road section; when the line nodes are known, the vehicle-mounted terminal obtains the real-time coordinates of the train, and the line node closest to the real-time coordinates serves as the current position of the train; when the distance between the current position of the train and the starting point of the nearest key road section to be driven in is smaller than or equal to the preset approaching distance, image data stream transmission between the ground video monitoring subsystem and the vehicle-mounted terminal is started through the over-the-horizon train-ground safety monitoring platform; and when the terminal distance between the current position and the nearest driven-out key road section is greater than or equal to a preset far distance, stopping image data stream transmission. According to the invention, high-definition image data aided driving can be realized in the high-speed running process of the train.

The invention relates to the technical field of intelligent traffic, in particular to a lane positioning system and method for vehicle-road collaborative awareness, which comprises a data collection module, a data transmission module, a storage library module, a central processing module and an instruction output module, and is characterized in that the data collection module is in control connection with the data transmission module, and the data transmission module is in control connection with a preprocessing module; the automatic driving module assists the vehicle in automatic driving according to the positioned data, the system and the automatic driving function are linked, and convenience and reliability are achieved; the system runs safely and stably in the system normal monitoring module, and an offset correction system is used in the lane offset correction module to correct lane offset, so that lane offset early warning and automatic correction are realized, and potential safety hazards are eliminated; after offset correction in the repeated positioning module, confirmation is carried out again, and result elimination early warning is fed back in the system feedback module; the system locates a lane by means of a sensor and a camera, can be used in a place which cannot be covered by a positioning system, and has very high environmental adaptability.

The invention relates to the field of laser communication, and provides a parallel low-complexity frequency offset compensation method, device, equipment and medium, and the method comprises the steps: calculating the M-power variance of N M-PSK signals which are input in parallel, and carrying out the first frequency offset compensation of an original signal through the difference result of N points at intervals; the M-PSK signal is a PSK signal of which the modulation order is M; and solving an N-point M-power spectrum for the signal after the first frequency offset compensation, and performing second frequency offset compensation for the signal after the first frequency offset compensation by using the spectral line position of the M-power spectrum. The two-stage frequency offset compensation method provided by the invention has the characteristics of high dynamic range, low resource consumption, no need of data assistance and support of burst signal processing, and is particularly suitable for a frequency offset compensation scene in a high-speed laser communication link.

The invention relates to an electroencephalogram signal preprocessing method and a related device thereof.The method comprises the steps that initial electroencephalogram signals collected by a multi-channel electrode array and at least one kind of auxiliary data collected synchronously with the initial electroencephalogram signals are obtained, and the auxiliary data comprise electrode impedance values and / or physiological state data; wherein the resistance reactance value indicates an electrode impedance value corresponding to the acquisition channel associated with the initial electroencephalogram signal, and the physiological data at least comprises motion posture data and / or physiological index data; performing multiple filtering processing on the initial electroencephalogram signal to obtain a filtered electroencephalogram signal; inputting the filtered electroencephalogram signals and at least one kind of auxiliary data into a signal restoration model, and optimizing the candidate electroencephalogram signals to obtain optimized target electroencephalogram signals. According to the scheme, the electroencephalogram signals can be preprocessed by dynamically adjusting the filtering parameters and utilizing the impedance value and the auxiliary data, and the quality of the electroencephalogram signals is improved.

The utility model relates to a 3D curved surface data acquisition device. The device comprises a first linear motion platform; the supporting seat is arranged on the first linear motion platform in a sliding manner; the rotating platform is rotationally arranged on the supporting seat, a product with a 3D curved surface is arranged on the rotating platform, and the first linear motion platform drives the product to move to a detection position; the motion extension direction of the second linear motion platform is different from that of the first linear motion platform; the sliding support is arranged on the second linear motion platform in a sliding mode, and the laser is obliquely arranged on the sliding support; and the control unit is used for controlling the first linear motion platform, the rotating platform and the second linear motion platform to move, so that the laser obtains 3D curved surface data of the product. According to the utility model, the 3D curved surface product is automatically moved, the curved surface data is comprehensively acquired by laser, and the high-precision detection of the 3D curved surface is assisted.

The invention discloses a dynamic perception-based maintenance auxiliary equipment monitoring system and method, and relates to the technical field of auxiliary equipment monitoring, and the system comprises the steps: dividing a monitoring region according to a monitoring range, marking a generator set, collecting historical operation data, auxiliary tool data and fault type data, and building a fault database for the monitoring region; calling the fault information of the monitoring area, analyzing the demand condition of the fault type for the auxiliary tools based on historical data, predicting the demand condition of the auxiliary tools in a monitoring period, analyzing the storage condition of the auxiliary tools in the monitoring range, and prompting management personnel to supplement until the auxiliary tools in the monitoring range are sufficient. According to the method, the fault database is established, maintenance records of effective maintenance personnel are integrated, data interference of invalid personnel is eliminated, and the actual requirements of different fault types for various auxiliary tools are objectively judged.

This application discloses a digital car key, a vehicle communication system, and a method. The digital car key may include a power supply module, a sensor, and a first wireless communication module. The sensor is coupled to the power supply module and the first wireless communication module to collect movement data of the digital car key and send the movement data to the first wireless communication module. The movement data includes movement speed and movement direction. The first wireless communication module is used to scan a first wireless broadcast signal emitted by a target vehicle and pair with the target vehicle to send movement data to the target vehicle, thereby determining the current distance between the digital car key and the target vehicle. Implementing the method of this application, by using the movement data of the digital car key to assist in determining the distance between the digital car key and the vehicle, helps to solve the problem of inaccurate distance measurement caused by obstacles in the transmission path of the wireless broadcast signal or sudden changes in the measured distance within a short period of time.

The invention relates to an industrial internet and digital twin intelligent manufacturing experiment table, which comprises a material temporary storage table, a parallel robot, a material transportation mechanism, a carrying manipulator, a simulation processing unit, a mobile transmission trolley, a material storage unit, an electrical control system and an intelligent gateway data acquisition system, the electrical control system comprises a PLC (Programmable Logic Controller), and the PLC is connected with the parallel robot, the material transportation mechanism, the carrying manipulator, the simulation processing unit and the mobile transmission trolley through an I / O (Input / Output) module to realize motion control, time sequence scheduling and state monitoring of each execution mechanism; the intelligent gateway data acquisition system is in two-way communication with the PLC controller through an industrial communication protocol, and production data acquisition, control instruction forwarding and cloud collaboration are realized. Real-time monitoring and data analysis can be realized, production data can be visualized through a cloud platform, a user is assisted to quickly position faults, process parameters are optimized, and the production efficiency and the equipment comprehensive efficiency are improved.

The utility model discloses a monitor operation module based on a voice recognition technology, which comprises a vital sign monitoring module used for collecting vital sign data of a patient and provided with a data transmission interface, and a data processing and display module used for receiving the data of the vital sign monitoring module and displaying the data on a display screen; the interaction module comprises a voice interaction unit, a touch control unit and an audio player; and the response module is connected with the data processing and display module, transmits the vital sign data in real time, monitors data changes, transmits prompt data to the display screen and the audio player, and assists in reminding a user. According to the vital sign monitor, the usability and the response speed of the vital sign monitor are improved through the man-machine interaction technology, the vital sign monitor is compatible with a clinically applied monitor through the data transmission interface, the operation burden of a doctor is reduced, and meanwhile a patient is reminded of key vital sign data.

The invention discloses an auxiliary management method and system for rattan plaiting skill data based on image analysis, and relates to the technical field of rattan plaiting image processing.The method comprises the steps that an initial plaiting topological graph is constructed by obtaining an original rattan plaiting image, and high-confidence and interference topology first-level sub-graphs are divided through pixel quality evaluation; taking the adjacent high-confidence topology sub-graphs as reference, calculating the similarity between the adjacent boundary length and the topological structure and coding a structure prior vector, and selecting weighted fusion according to vector dispersion or dividing into interference topology secondary sub-graphs according to weighted distance so as to determine generation prior; topological sub-graphs are repaired through a conditional generation model, a complete woven topological graph is output through fusion, and precision can be improved through optimization of time sequence, symmetry and boundary consistency. According to the method, the topological structure of the interference area can be accurately repaired, and the integrity and reliability of the rattan-plaited woven topological graph are remarkably improved.

The disclosure relates to a 5G or 6G communication system for supporting a higher data transmission rate. A user equipment (UE) includes a transceiver configured to: transmit, to a base station (BS), UE capability information that indicates support for data-aided modulation constellations, wherein the data-aided modulation constellations support a first transmission mode utilizing data symbols for channel estimation, and receive, from the BS, a modulation and coding scheme (MCS) indication. The UE also includes a processor operably coupled to the transceiver and configured to: determine, based on the MCS indication, a first data-aided modulation constellation, and generate first modulation symbols from input bits according to the first determined data-aided modulation constellation. The transceiver is further configured to transmit, to the BS, the first modulation symbols.

The invention provides a method for performing athlete trajectory tracking and motion data analysis by using an unmanned aerial vehicle cluster, and belongs to the technical field of physical training. The leading unmanned aerial vehicle collects wide-angle trajectory data, the assisting unmanned aerial vehicle obtains action videos at a high frame rate, and the collection accuracy is cooperatively guaranteed in combination with a local area network; the global trajectory is fused through Kalman filtering, action parameters such as joint angles are obtained by means of CNN and multi-view geometry, and the problem that in the prior art, the data dimension is single is solved; a visual report containing historical and standard data comparison is generated through multi-layer calculation of a presentation layer, a physiological layer and a tactical layer subsequently, fatigue early warning can be pushed to a coach in real time, comprehensive and deep data analysis support is provided for training of athletes, and the training precision is improved.

The invention provides a deep learning DD channel estimation system and method combined with an adaptive CIR truncation mechanism, and the system comprises a pilot frequency estimation module which carries out the channel estimation according to a received pilot frequency sequence, and outputs a pilot frequency assisted channel estimation value; the deep learning prediction module is used for predicting the channel state of the next moment according to the historical channel estimation sequence; the data auxiliary channel estimation module is used for balancing and decoding the data symbols based on the channel state, and obtaining data auxiliary channel estimation after reconstruction; the self-adaptive CIR truncation module takes data-assisted channel estimation as input and outputs a channel estimation value after truncation; and the periodic pilot frequency correction module is used for triggering pilot frequency updating and correcting the pilot frequency assisted channel estimation value. According to the method, the influence of a data judgment error is actively considered, the tap weight is dynamically output by using the lightweight neural network, the historical sequence is intelligently cleaned and cut off, error propagation is effectively inhibited, and the stability and robustness of prediction are improved.

The invention belongs to the field of supply chain auxiliary decision-making, and particularly relates to a supply chain data auxiliary decision-making analysis system and method. The supply chain data aided decision analysis method comprises the following steps: S10, acquiring supplier nodes selling parts to a local node in a supply chain, marking the sold parts as main parts, and capturing a current selling price; s20, dividing the industries for purchasing the main parts into reference industries belonging to the same genera as the local nodes and contrast industries belonging to different genera from the local nodes; and based on historical sales data of all suppliers, calculating the total sales amount of the main parts as the total sales amount. According to the scheme, market risks of suppliers are accurately measured through industry residual analysis and supplier sales proportion mapping, and the problem that risk scores of the suppliers are inaccurate is solved; and then a financial tool is connected to realize integration of purchasing and sleeve protection, and parts are replaced through residual error cross-correlation identification, so that supply interruption is finally avoided, and synchronous amplification of the purchasing price and the capital occupation cost is prevented.

The invention relates to the technical field of financial data processing, and particularly discloses an AI-based financial text data auxiliary processing system, which comprises a data acquisition module, a data preprocessing module, an AI intelligent analysis module, a data verification module, an intelligent decision-making auxiliary module, a data storage and management module and a safety management and control module. The system can automatically collect multi-format and multi-source financial text data in batches, after the data quality is optimized through preprocessing, intelligent text classification, key information extraction and structured conversion are achieved through an improved deep learning algorithm, the accuracy and compliance of the data are guaranteed in combination with multi-dimensional verification, then decision support such as financial analysis and risk early warning is provided, and the accuracy and the reliability of the system are improved. And data management and security are guaranteed through distributed storage and an omnibearing security mechanism. According to the method, full-process automation of financial text processing is realized, the processing efficiency and data accuracy are improved, the problem of data dispersion of an existing system is solved, the intelligent level of financial decision is improved, and compliance requirements are met.

A wireless device (12) is configured for radio access technology, RAT, dependent positioning in a wireless communication network (10). The wireless device (12) receives, from a network node (18), assistance data (20) that assists the wireless device (12) with RAT-dependent positioning, e.g., downlink Observed Time Difference of Arrival. The assistance data (20) may be based on device-specific information. The device-specific information may for example include a subscription associated with the wireless device (12). The wireless device (12) may perform the RAT-dependent positioning based on assistance from the assistance data (20).

The application discloses a kind of semi-supervised image semantic segmentation methods based on similarity weighting multi-teacher network model, comprising the following steps: preparing training set and verification set data for semantic segmentation model training;Construct similarity weighting based multi-teacher network model SW-MMTNet;The unlabeled data of training set is respectively subjected to strong data enhancement and weak data enhancement, to realize the data disturbance of unlabeled data;The training of SW-MMTNet model is carried out using the labeled data and unlabeled data of training set;The effect of segmentation model is evaluated using verification set, and the final image semantic segmentation model is obtained.The method can integrate the prediction information of multiple teacher networks, generate high-quality and stable pseudo-labels for unlabeled data, thereby effectively utilizing unlabeled data to assist the training of semantic segmentation model and reducing the demand for large amounts of data for semantic segmentation.In addition, the method also enables the student network of SW-MMTNet to learn more information through data disturbance.

The invention relates to an electric power dispatching auxiliary system and method. The system comprises a login authentication module, a data collection module, an auxiliary dispatching operation ticket to-be-written module, an auxiliary terminal calling module, an operation progress supervision module and a data output module. The system realizes secure access through identity recognition and authentication; acquiring and transmitting a scheduling application form, an operation order and equipment running state data by using a data collection module; the auxiliary writing module determines a writing requirement through similarity calculation and generates a corresponding scheduling operation ticket; the auxiliary terminal calling module calls a related unit according to the operation order and records an unsuccessful contact condition; the operation progress supervision module counts timeout conditions caused by equipment defects or low operation speed and calculates timeout rates in a classified manner; and finally, the data output module uniformly outputs various results. According to the method, the scheduling operation ticket generation efficiency can be improved, a scheduling terminal notification and supervision mechanism is optimized, and the intelligence and safety of power scheduling work are enhanced.

It can sometimes be difficult to perform signal switching correctly. [Solution] The switching device includes a receiving unit that receives IP (Internet Protocol) signals from multiple sources, which include video data, audio data, ancillary data, and time information, and a switching unit that uses the time information contained in the IP signals to switch the video data, audio data, and ancillary data to be output from among the video data, audio data, and ancillary data corresponding to each received IP signal.