185 results about "Data generator" patented technology

A Generative Adversarial Network (GAN) is used to generate synthetic network traffic data, such as for use in training Machine Learning (ML) models. In one embodiment, a new NWDAF analytic “SyntheticData” is defined. The analytic receives as input from a requesting network function (NF) at least an amount of network traffic data requested and the type of network traffic data requested. The SyntheticData analytic uses a GAN model to generate realistic synthetic network traffic data based on actual network traffic collected in the wireless communication network. The analytic sends to the requesting NF the specified amount of synthetic network traffic data of the specified type. In one embodiment, the synthetic network traffic data generation is implemented as a new logical function of an NWDAF: the Data Generator Logical Function (DGLF).

The invention provides a three-dimensional visual model construction method and system for a mine slope, and relates to the technical field of data processing, and the method comprises the steps: obtaining the inclined image data of the mine slope; performing error correction on the inclined image data; performing live-action modeling according to the inclined image data after error correction to obtain a three-dimensional grid model and a point cloud data file; generating a digital terrain model according to the point cloud data file; extracting a contour line of the digital terrain model, and exporting a contour line file; generating a three-dimensional terrain curved surface by using a terrain data generator according to the contour line file; the three-dimensional terrain curved surface extends downwards by a preset depth, and a three-dimensional entity slope model is generated; physical parameters of side slope rock soil are set, and a Moire-Coulomb intensity criterion is given to the three-dimensional entity side slope model so as to simulate shear failure characteristics of the rock soil; on the basis of shear failure characteristics of rock and soil, grid division is carried out on the three-dimensional solid slope model; and generating a three-dimensional visual model of the mine slope.

The invention relates to the crossing field of industrial control system (ICS) safety and artificial intelligence, and particularly discloses a PLC high-interaction honeypot system based on multi-agent task splitting and RAG enhancement, which adopts a localized multi-agent collaborative architecture based on edge computing and is composed of a high-simulation equipment layer and an intelligent decision-making layer. The high-simulation equipment layer comprises a PLC dynamic mirror image, an HMI interface and a sensor data generator, and an active trapping environment is constructed through protocol fingerprint confusion and virtual and real data fusion technologies. The decision-making layer deploys a multi-agent task scheduling engine, integrates four kinds of agents including protocol analysis, behavior analysis, threat assessment and response generation, and realizes attack context perception and strategy dynamic generation based on a local RAG knowledge base. The load balancing agent dynamically allocates tasks according to equipment resources, and cooperates with offline knowledge update (USB flash disk encryption synchronization threat features) to form a closed-loop defense system, thereby ensuring physical isolation of an industrial network and realizing high-fidelity active defense.

The invention relates to the field of semiconductors, and discloses an FPGA-based time sequence generator, a time sequence generation method, terminal equipment and a readable storage medium, and the FPGA-based time sequence generator comprises a clock source which is used for providing a working frequency; the test period buffer is used for sequentially caching the data of each test period; the time sequence data cache is used for sequentially caching all time sequence data; the test period counter is used for starting counting every time one test period is entered; the time sequence edge generator is used for determining whether the time sequence clock is enabled or not according to the count of the test period data of the same test period and the integer part value of the time sequence data, and outputting the decimal part value of the time sequence clock which is not full of the integer multiple; the data generator is used for generating the received one-bit mode data into parallel data with a preset bit width when the time sequence clock is enabled; and the converter is used for converting the parallel data into serial data by using the preset frequency multiplication clock of the working frequency and then outputting the serial data. According to the scheme, the time resolution based on the time sequence generator can be improved.

The invention discloses a test data intelligent generation and privacy compliance desensitization method, which belongs to the technical field of electric digital data processing, and comprises the following steps: monitoring the structure change of a preset regulation update source and a target application system, and obtaining compliance dynamic information; analyzing the compliance dynamic information to generate a dynamic compliance rule base; based on the dynamic compliance rule base, generating an initial test data set through a preset intelligent data generator; desensitizing the initial test data set, and outputting a desensitized test data set; scanning the desensitized test data set in real time for compliance verification, and generating a compliance verification report; and dynamically optimizing the generation strategies of the dynamic compliance rule base and the intelligent data generator based on the compliance verification report. Dynamic rule base construction, intelligent separated data generation, closed-loop verification optimization and desensitization audit tracing technologies are adopted, laws and regulations and system changes can be automatically adapted, and the service validity, privacy security and law compliance of test data are synchronously guaranteed.

A data capture stage includes a frame at least partially surrounding a target object, a rotation device within the frame and configured to selectively rotate the target object, a plurality of cameras coupled to the frame and configured to capture images of the target object from different angles, a sensor coupled to the frame and configured to sense mapping data corresponding to the target object, and an augmentation data generator configured to control a rotation of the rotation device, to control operations of the plurality of cameras and the sensor, and to generate training data based on the images and the mapping data.

Control system (1) comprising: a sensor (44) that detects an acceleration based on a vibration of a robot (3); an interpolator (222) that interpolates a plurality of sensor data sets detected by the sensor (44); and a data generator (223) that calculates a correlation coefficient based on a plurality of interpolation data sets interpolated by the interpolator (222), combines the plurality of interpolation data sets based on the correlation coefficient, and averages composite data with fine sampling intervals resulting from the combination of the plurality of interpolation data sets.

The present disclosure provides a ship monitoring system capable of visualizing a risk of a collision or an approach in a width direction perpendicular to a heading of another ship. The ship monitoring system includes a first data generator, a second data generator, and processing circuitry. The first data generator generates first ship data indicative of a position and a velocity of a first ship. The second data generator generates second ship data indicative of a position and a velocity of a second ship. The processing circuitry specifies a risk range where a risk value indicative of a risk of a collision between the first ship and the second ship is above a threshold among an estimated course of the second ship, based on a position of the first ship and a position of the second ship at each timing, that are estimated from the first ship data and the second ship data, when assuming that the first ship changes a course to an arbitrary direction and crosses the estimated course of the second ship. The processing circuitry further display a polygonal OZT (Obstacle Zone by Target) having vertexes at least comprised of a rear end and a front end of the risk range, and a representative point of the first ship located at a position corresponding to the rear end of the risk range.

The invention discloses an uncertain knowledge graph reasoning method based on semi-supervised confidence distribution learning. The method comprises the following steps: converting triple confidence in uncertain knowledge graph training data into confidence distribution; learning embedding of the uncertainty knowledge graph on the marked data and the pseudo-marked data generated by the pseudo-marked data generator at the same time by using a relation learner based on confidence distribution learning; generating a high-quality pseudo-confidence distribution label for the unmarked data by using a pseudo-marked data generator; iteratively training a relation learner based on confidence distribution learning and a pseudo-mark data generator by utilizing element self-training until the relation learner and the pseudo-mark data generator converge; and transmitting data to be complemented into the trained relationship learner based on confidence distribution learning for reasoning to realize uncertainty knowledge graph complementation. According to the method, the supervision information of a small amount of confidence or no confidence in the marked data can be captured, and the method is suitable for the scene of unbalanced confidence distribution of the triple of the uncertainty knowledge graph.

Proposed is a calibration method for a camera and a lidar using a calibration board having a specific pattern. The method may include extracting one frame for an image captured by a camera and point cloud data acquired by a lidar, by a data generator, identifying a feature point for a calibration board included in one extracted frame for the image and the lidar, by the data generator, and performing calibration for the camera and the lidar based on the identified feature point, by the data generator. The present method is a technology developed with support from the Ministry of Trade, Industry and Energy / Korea Planning and Evaluation Institute of Industrial Technology (Project No. 20022003 / Project name-Automotive industry technology development project / Project name-Development of industrial autonomous driving and safety assurance technology based on longitudinal and lateral interconnection).

The invention relates to the technical field of wireless communication protection, in particular to a wireless communication self-adaptive safety protection method and system based on edge computing. Capturing physical layer channel state information when the Internet of Things terminal equipment accesses the network gateway for the first time, a functional area semantic identifier of the network gateway and a current timestamp, and generating an anonymous session voucher; tracking operation track data of the anonymous session voucher, and constructing a track operation fragment; obtaining a privacy risk assessment value based on trajectory uniqueness, spatial information sensitivity and behavior association analysis of the trajectory operation fragment; when the privacy risk assessment value exceeds a preset risk threshold value, triggering a track link breaking mechanism; obtaining privacy enhanced trajectory data based on privacy protection processing and a trajectory broken link mechanism; training a virtual data generation model through the privacy enhancement trajectory data; and after the virtual data generation model converges, driving a virtual data generator in the virtual data generation model to generate a virtual operation and maintenance data stream so as to realize self-adaptive protection of wireless communication data.

An arrangement has a random data generator for producing a crypto seed, and may use a secure transport mechanism. The arrangement may produce, through a first cryptographic operation, a cryptographic intermediate product that is deterministically dependent on both the crypto seed and a user's secret received through the secure transport mechanism. This cryptographic intermediate product constitutes a digital identity of the party. A second cryptographic operation uses the digital identity of the party to produce a cryptographic output including the crypto seed in encrypted form. The arrangement transmits at least part of the cryptographic output through the secure transport mechanism.

The invention discloses a photovoltaic-energy storage-electric vehicle collaborative V2G optimization scheduling method adaptive to a building scene, and relates to the technical field of building energy management and vehicle network interaction, and the method comprises the steps: training to obtain a data generator; outputting sample data of new energy output and building load scenes through a data generator, and generating a plurality of typical scenes by using a clustering algorithm and a Nash equilibrium model; in each typical scene, planning vehicle network interaction facility configuration, and carrying out cooperative scheduling on a source, a load, a storage and a vehicle; selecting participating automobiles, and scheduling the participating automobiles in combination with battery loss and a compensation mechanism; and comprehensive evaluation of economy and operation quality is carried out on the collaborative scheduling result of the source, the load, the storage and the vehicle, the planning result of vehicle network interaction facility configuration and the scheduling result of the participating vehicle. According to the method, the uncertainty scene serves as input, cost minimization is achieved through the two-stage model, and the economical efficiency and feasibility of the scheme are verified through evaluation.

An apparatus for sharing data according to one embodiment includes a first combined data generator configured to generate first combined data by combining target data to be shared and a random bit string, a data encryptor configured to generate a ciphertext for the first combined data, a key encryptor configured to generate one or more ciphertexts for a decryption key of the ciphertext for the first combined data using one or more public keys, a second combined data generator configured to generate second combined data by combining the ciphertext for the first combined data, the one or more ciphertexts for the decryption key, and the random bit string, and a data discloser configured to disclose the second combined data to one or more entities to share the target data to be shared.

A modular adversarial training data generator (“generator”) generates adversarial training data for a machine learning (ML) model to detect malicious protocol data units (PDUs). The adversarial training data mimics high volume cyberattacks by perturbing PDUs to bypass malicious detection systems. For HyperText Transfer Protocol (HTTP) PDUs, the generator rearranged, replaces, and grid searches values of HTTP header fields to generate the adversarial training data. The generator further biases grid search based on metrics for values of HTTP header fields that quantify impact of replacing the values on malicious verdicts by ML models trained on the adversarial training data.

The invention belongs to the technical field of artificial intelligence and distributed machine learning, and discloses a personalized federal learning data generation method and a related device. The personalized federal learning data generation method comprises the following steps: acquiring a local data similarity distribution vector of a selected client; acquiring a trained local model of the selected client and sending the trained local model to a server in federal learning; and obtaining a global data generator issued by the server, training based on the local sample similarity distribution vector to obtain a local data generator, and finally generating personalized federal learning data of the selected client by using the local data generator. The technical scheme disclosed by the invention specifically relates to a personalized federated learning data generation scheme based on adversarial training and similarity evaluation, and can effectively obtain a generator and generate personalized federated learning data with relatively high reliability.

The invention discloses a digital annotation quality automatic verification method and system, and relates to the technical field of data processing. The method comprises the following steps: identifying and classifying original annotation data to obtain an identification task level; task matching is carried out according to the identified task level and the multi-level processing path, and task import connection is established; obtaining a distribution verification target through task import connection, calling an adversarial data generator to create semantic and quality dimension variants, comparing the annotation consistency of original data and variant data, and obtaining annotation vulnerabilities and potential error information; and configuring an identification tolerance threshold according to the identification task level, configuring an identification constraint parameter based on the vulnerability and the potential error information, executing annotation verification in the matching processing path based on the identification constraint parameter, and outputting a verification result. The technical problems of low efficiency and insufficient accuracy caused by the fact that digital labeling quality verification depends on manual sampling inspection in the prior art are solved, and the technical effect of improving verification efficiency and accuracy is achieved.

A virtual reality (VR)-based tour route providing apparatus, which creates a tour route of a corresponding tourism area through a VR experience in which a tourism area of users is implemented according to one embodiment of the present invention, includes an area information receiver, a zone information receiver, a VR data generator, a tracking data generator, a zone visit information generator, a zone-of-interest selection unit, and a tour route information generator

A storage product manufactured as a standalone computer component and installed in a computing system to implement an internet application. The storage product includes a network interface, a host interface, computing circuits, and a local storage device having a storage capacity accessible via the network interface. A data generator is connected to the network interface. A local host system is connected to the host interface to control access, made via the network interface. The data generator can send bulk data to the network interface. The computing circuits can generate derived data from the bulk data and store the derived data and / or the bulk data in the local storage device. A central server and / or a user device can connect over internet via to the network interface of the storage product to access the derived data and / or the bulk data.

Corner table creation for graphics processing is described. An example of an apparatus includes corner data generator circuitry, including a circuit to generate a plurality of vertex-corner lists for a plurality of portions of a triangle mesh, wherein a vertex-corner list for a portion includes an index for a vertex, a count of corners for the vertex, and a list of corners associated with the vertex, and a circuit to receive the vertex-corner lists and generate one or more edge hash maps, each edge hash map including corner indices for edges of the triangle mesh.

Provided are a device and method for automatically generating source code. The device includes a training data generator that receives first training data consisting of a pair of natural language specification and source code and converts the first training data into second training data consisting of a pair of natural language specification and abstract syntax tree and a large language model trainer that trains a large language model based on the second training data, in which the abstract syntax tree includes structural information and semantic information of the source code.

The application provides a visual language large model adaptive bias evaluation method for evaluating bias performance of a target model on a specific task, and the method comprises the following steps: S1, obtaining an initial evaluation dataset, wherein the initial evaluation dataset comprises a plurality of evaluation data; S2, fine-tuning a preset large language model using the initial evaluation dataset in a preset fine-tuning manner to obtain an initial data generator; and S3, generating evaluation data based on the initial data generator to perform multiple rounds of bias evaluation on the target model, wherein in each round of bias evaluation, based on all evaluation data corresponding to the last round of bias evaluation, a data generator corresponding to the last round of bias evaluation is iteratively updated according to a preset objective function to obtain a current round data generator, and the current round data generator is used to generate a current round evaluation dataset to perform bias evaluation on the target model.

A virtual reality (VR)-based tour route providing apparatus, which creates a tour route of a corresponding tourism area through a VR experience in which a tourism area of users is implemented according to one embodiment of the present invention, includes an area information receiver, a zone information receiver, a VR data generator, a tracking data generator, a zone visit information generator, a zone-of-interest selection unit, and a tour route information generator.

The invention relates to a system (100) for optimizing the configuration of crash tests and simulations for vehicle development. The system (100) comprises a data preparation module (200) that splits a data set (210) containing crash test and simulation data into a first training data set (210), a validation data set (230), and a test data set (240); a training module (300) with a first AI model (350) for creating configurations of physical crash tests and virtual simulations, wherein the first AI model (350) is trained with the first training data set (220); and a data value module (400) that calculates data values ​​(φ(i)) for the data points (i) of the crash test and simulation data of the training data set (220) with respect to the validation data set (230).a model optimization module (500) comprising a second AI model (550) and a training data generator (570), wherein the training data generator (570) creates a second training data set (580) which is used to train the second AI model (550) to predict the relationship between input parameters (p) and the computed data values ​​(φ(i)); a parameter module (700) which uses the trained second AI model (550) to find crash test parameters with a maximum data value (φ(i)).

Proposed is a calibration method for a lidar and an IMU using features according to the shape of a specific object included in point cloud data. The method may include placing point cloud data acquired by a lidar mounted on a vehicle on a predefined world coordinate system, by a data generator, extracting a region to be used for calibration from the placed point cloud data, by the data generator, identifying at least one object included in the extracted region, by the data generator, and performing calibration on the point cloud data by fitting point cloud included in the at least one identified object to a pre-stored model, by the data generator. The present method is a technology developed with support from the Ministry of Land / KAIA (Project No. RS-2021-KA160637).

The invention discloses a wireless data transmission system and method using a compression technology, and the system comprises a data generator which is used for generating to-be-transmitted wireless data and transmitting the to-be-transmitted wireless data to a compression module, the compression module is used for compressing the to-be-transmitted wireless data into compressed data by using a self-adaptive differential pulse code modulation algorithm, and the compressed data is transmitted to the compression module through the self-adaptive differential pulse code modulation algorithm; the gateway end comprises a gateway protocol end and a transmission module, and the gateway protocol end is used for receiving the compressed data packet and sending the compressed data packet to the transmission module for transmission; and the server is used for receiving and decompressing the compressed data packet transmitted by the transmission module so as to realize the transmission of the wireless data to be transmitted. According to the invention, low-cost, high-efficiency and high-real-time wireless data transmission under a large field domain can be realized.

A modular adversarial training data generator (“generator”) generates adversarial training data for a machine learning (ML) model to detect malicious protocol data units (PDUs). The adversarial training data mimics high volume cyberattacks by perturbing PDUs to bypass malicious detection systems. For Hypertext Transfer Protocol (HTTP) PDUs, the generator rearranged, replaces, and grid searches values of HTTP header fields to generate the adversarial training data. The generator further biases grid search based on metrics for values of HTTP header fields that quantify impact of replacing the values on malicious verdicts by ML models trained on the adversarial training data.

Apparatus for generating an enhanced sound field description, including: a sound field generator for generating at least one sound field description indicating a sound field with respect to at least one reference location; and a meta data generator for generating meta data relating to spatial information of the sound field, wherein the at least one sound field description and the meta data constitute the enhanced sound field description. The at least one sound field description includes in a certain case a first sound field description related to the at least one reference location and a second sound field description related to a further reference location being different from the reference location, and the meta data relating to the spatial information indicates the reference location and the further reference location or one of both reference locations and a vector extending between both reference locations.

Proposed concepts aim to provide schemes, solutions, concepts, designs, methods, and systems pertaining to medical data augmentation. In particular, medical data is augmented (e.g., linked to or annotated with) generation distance data describing a distance between user(s) and the medical data generation device during generation of the medical data. Subsequently, at least a part of the medical data shown to viewer(s) of the medical data may be determined based on the generation distance data. Accordingly, appropriate medical data may be presented to the viewer without the need for manual configuration. That is, based on observations of the distance of users(s) (e.g., healthcare professionals) who interact with the data generator, and user(s) who watch or are close to the data generators, it is possible to anticipate what parts of the medical data may be needed for subsequent viewing by a user.