145 results about "Dummy data" patented technology

The invention discloses a network intrusion detection method based on a self-attention residual generative adversarial network, which belongs to the technical field of network intrusion detection and comprises the following steps: collecting network flow data, and preprocessing the network flow data to obtain real training data; a FARD-WGAN-GP model is constructed and trained, and hybrid pseudo data is generated based on the trained FARD-WGAN-GP model; constructing a time sequence attention detection model; training a time sequence attention detection model based on the real training data and the pseudo data; network intrusion real-time detection is carried out based on the trained time sequence attention detection model; on the basis of the real-time detection result, optimizing the FARD-WGAN-GP model, and on the basis of the optimized FARD-WGAN-GP model, optimizing the time sequence attention detection model; and carrying out network intrusion detection based on the optimized time sequence attention detection model. According to the method, the quality of the generated data is improved, and the detection accuracy is improved.

The invention discloses a logistics freight vehicle transportation supervision system and method based on the Internet of Things, and relates to the technical field of vehicle transportation supervision, the supervision system is constructed based on multi-source data fusion and hierarchical anomaly detection, and the problems of false data and signal interference which are difficult to identify in traditional supervision can be effectively solved; through multi-level comparison and block chain evidence storage, supervision blind areas caused by manual tampering or equipment faults are greatly reduced; meanwhile, the self-learning analysis module dynamically recognizes repeated or regular pseudo data, and by iteratively updating a threshold value and improving an early warning level, false alarms can be remarkably reduced, and accurate positioning of potential anomalies can be enhanced; the real-time performance and traceability of freight vehicle transportation supervision are improved, and the method has significant improvement value for the traditional mode which depends on a single data source and lacks a rapid verification mechanism.

The invention discloses a visual question-answering system based on multi-modal knowledge autonomous learning and a construction method, and belongs to the technical field of artificial intelligence. The invention provides a visual question and answer model construction method fusing image information, natural language questions and external knowledge. A knowledge dynamic screening and replacement strategy is constructed by introducing an autonomous evolution mechanism, and whether knowledge is inserted or updated can be judged according to a current training state and knowledge similarity, so that the problem that a traditional static knowledge base cannot adapt to task changes is solved. According to the method, through pseudo data construction and multi-modal triple extraction, structured and high-association-degree'image-question-answer 'knowledge representation is established, so that knowledge has better semantic expression ability, and redundancy and noise are effectively reduced. According to the method, rapid retrieval and active calling of knowledge are achieved by means of the vector indexing technology, key support information can still be obtained in a complex scene, and the expression of the model in a visual question-answer task based on knowledge is remarkably improved.

A first ECU and a plurality of second ECUs are connected to a communication bus. The first ECU repeatedly transmits a beacon signal through the communication bus. Each of the first ECU and the plurality of second ECUs transmits data according to an order set in advance when the beacon signal is transmitted. The first ECU or the second ECU transmits dummy data whose transmission destination is different from the remaining ECUs when there is no transmission data to be transmitted to one of the remaining ECUs other than itself.

The invention discloses a dam safety detection method and system based on multi-source detection data, and relates to the technical field of dam safety detection, and the method comprises the steps: cooperatively collecting dam multi-source data through multiple platforms and multiple sensors; the collected multi-source data are preprocessed; constructing a multi-scale teacher network, and performing high-precision feature learning and risk quantification by using labeled multi-source data; teacher network knowledge migration is carried out through a knowledge distillation technology, and a lightweight student network is trained in combination with cross-domain pseudo data; and deploying the trained lightweight student network to an unmanned aerial vehicle or a robot dog, and carrying out dam safety real-time detection. Through integration of multi-source data acquisition, cross-domain mutual training of teachers and students, lightweight model deployment and automatic early warning decision, pain points of single data, difficulty in cross-domain adaptation, insufficient precision, deployment limitation and low efficiency are solved, the system can be directly deployed on an unmanned aerial vehicle or a robot dog, dependence on cloud computing power is not needed, data acquisition and analysis time delay is reduced, and the system can be widely applied to unmanned aerial vehicles or robot dogs. And emergency scenes are quickly responded.

The present invention provides a method for controlling a flash memory module. The flash memory module includes a plurality of dies, each die includes a plurality of blocks, each block includes a plurality of pages, and the method includes the steps of: selecting a super block, wherein the super block includes a plurality of first blocks respectively located in the plurality of dies; for each first block in the super block, determining whether the first block is a full block or a blank block; and if the first block is not the full block or the blank block, writing dummy data to the first block so that the first block becomes the full block; and erasing the plurality of first blocks in the super block, so that the plurality of first blocks become a plurality of blank blocks.

The invention discloses a rapid yield analysis method based on xgboost proxy model truncation sampling, and the method comprises the following steps: generating a first random sample point set, and obtaining a training set through simulation; constructing a proxy simulator, and training the proxy simulator by using the training set; generating a second random sample point set, and inputting the second random sample point set into the proxy simulator to obtain a pseudo data set; establishing a filter, determining two hyperspherical surfaces based on the pseudo data set, dividing a process parameter space, evaluating a safety coefficient and the probability that a sample point enters a real simulator, and simulating the sample point of the process parameter space according to the probability to obtain a real sample set; iteratively optimizing the hyper-spherical surface and the proxy simulator by using the real sample set until convergence to obtain a final safety coefficient; and generating a third random sample point set, carrying out truncation sampling and failure discrimination by combining a real simulator and an agent simulator, and carrying out statistics on the final yield. According to the method, the xgboost network is adopted as an agent model to reduce the simulation cost, the sample size is reduced through truncation sampling, and rapid and accurate prediction of the yield of the integrated circuit is realized.

The invention relates to the technical field of deep learning, and particularly provides a joint classification method for hyperspectral images and laser radar data. The method comprises the following steps: firstly, extracting multi-scale features of multi-modal data through a decoupling three-branch encoder, and then carrying out attention guidance cross-modal fusion on three initial features of the same scale through an autonomously designed feature fusion module to obtain three multi-modal fusion features of different scales; inputting the deepest fusion feature into a weight sharing decoder, and obtaining pseudo data through a reconstruction task; and the real data and the pseudo data are identified by the structure consistency perception discriminator, and the discrimination loss is returned, so that the encoder parameters are updated, and iteration is performed to the optimal. And finally, inputting the three-scale fusion features obtained based on the optimal parameters into a multi-level feature fusion classifier to complete classification, thereby effectively improving the classification precision, and realizing the adaptive fusion and enhancement of the multi-modal features.

Systems and methods are disclosed to provide a security framework for the authentication of digital content and the prevention of unauthorized modifications, especially targeting the vulnerabilities introduced by deepfake technologies. It innovates by merging reverse engineering with expression manipulation detection to discern genuine from altered digital media. The process involves comparing historical data against synthetic or real-time content, generating a “video mesh” that enables precise manipulation identification. Enhanced by Smart Contracts for each content piece to record and verify changes, this system advances digital identity and transaction security significantly beyond current methodologies. Furthermore, it employs Generative AI within the Identity Intelligent Clip Reviewer to scrutinize blockchain-secured data for manipulation signs, issuing a Proof of Digital Manipulation (PODM) for verified authenticity. This comprehensive method ensures the integrity and trustworthiness of digital interactions across various platforms, marking a significant step forward in the protection against sophisticated cyber threats and unauthorized data alterations.

This provides a machine learning technique that can suppress overfitting of new knowledge when applying IFSL to SaB. [Solution] This disclosure provides a machine learning device 100 that performs continuous learning based on a small number of new class data compared to basic class data, comprising: a pre-training module 30 that learns the weights of a neural network (NN) using basic class data; a pseudo-continuous learning module 40 that learns the weights of the NN learned by the pre-training module 30 using pseudo-data generated based on the basic class data; and a new class learning module 50 that learns the weights of the NN learned by the pseudo-continuous learning module 40 using basic class and new class data.

The CDR in a reception device (sink device) at a subsequent stage can be prevented from being unlocked even when jitter-fluctuation occurs due to transmission via an unstable transmission path such as a network. The clock data recovery unit extracts a clock and data on the basis of the baseband video data of a first format corresponding to a predetermined transmission path input from the predetermined transmission path. The extracted data is processed by the processing unit on the basis of the extracted clock. In this case, in a case where the occurrence of the jitter-fluctuation is detected, dummy data capable of clock extraction at the subsequent stage is output. For example, the processing unit outputs baseband video data in a second format, and the baseband video data in the second format is baseband video data corresponding to a predetermined wired interface.

A display device includes a substrate, a plurality of scan lines disposed on the substrate, a plurality of data lines and a plurality of dummy data lines intersecting the scan lines, a plurality of first pixel circuit units connected to the scan lines and the plurality of data lines, a first light-emitting element connected to the first pixel circuit unit, a plurality of second pixel circuit units connected to the scan lines and the plurality of dummy data lines, and a second light-emitting element connected to the second pixel circuit unit, where at least one of a plurality of scan lines intersects another adjacent scan line.

According to some example embodiments an operating method of a memory controller includes buffering first user data received from a host in a first region of a buffer memory corresponding to a first memory region of a memory device, receiving a flush request from the host, duplicating the first user data in a second region of the buffer memory corresponding to a second memory region of the memory device, padding, as padded data, the first user data duplicated to the second region with dummy data, programming the padded data in the second memory region of the memory device, buffering second user data received from the host in the first region, and programming, in the first memory region of the memory device, the first user data buffered in the first region and the second user data buffered in the first region.

The invention discloses an artificial intelligence digital science and technology system based on big data analysis, and relates to the technical field of artificial intelligence, and the system comprises a backup early warning module which is used for collecting data in real time through a collector, and comparing whether the task attribute and task configuration of data backup are abnormal or not in real time; the pseudo data monitoring module is used for monitoring whether data artifacts exist in the data backup process or not in real time and extracting the monitored data artifacts in real time to mark pseudo data positions; and the hierarchical search module is used for receiving alarm signals of existence of the pseudo data in real time through the data receiver, searching the pseudo data in real time and tracking a crushing execution result. According to the artificial intelligence digital science and technology system based on big data analysis, it is ensured that whether backup data is abnormal or not can be judged in time and early warned during data backup, the situation of false data monitoring omission is avoided, and the accuracy of data backup of artificial intelligence digital science and technology is improved.

This invention relates to the field of deep learning technology, and in particular provides a joint classification method for hyperspectral images and LiDAR data. The method first extracts multi-scale features from multimodal data by decoupling a three-branch encoder. Then, a self-designed feature fusion module performs attention-guided cross-modal fusion on three initial features at the same scale, resulting in three multimodal fused features at different scales. The deepest fused feature is input to a weight-shared decoder, where pseudo-data is obtained through a reconstruction task. A structure consistency-aware discriminator identifies real data from pseudo-data and returns a discrimination loss, thereby updating the encoder parameters and iterating until optimal. Finally, the fused features at the three scales obtained based on the optimal parameters are input into a multi-level feature fusion classifier to complete the classification. This method effectively improves classification accuracy and achieves adaptive fusion and enhancement of multimodal features.

A storage system is provided. The storage system includes a storage device having a plurality of memory cells, and a storage controller configured to control the storage device to store write data in a first memory cell among the plurality of memory cells, identify a current charge amount of a first cell string including at least one of the first memory cell and a current charge amount of a second cell string adjacent to the first cell string, and store dummy data in at least one memory cell connected to the first cell string or the second cell string based on the current charge amount of the first cell string and the current charge amount of the second cell string.

An arithmetic processing apparatus according to the present disclosure includes a digital filter. The arithmetic processing apparatus further includes a dummy data input unit configured to input dummy data to the digital filter when there is no input data input to the digital filter. The arithmetic processing apparatus further includes a cancellation processing unit configured to perform, on output data output from the digital filter, arithmetic processing for canceling an output component caused by the dummy data.

A method and apparatus for wireless communications between a Human Interface Device (HID) and a host. The host receives a plurality of virtual endpoints for a single HID, and transmits multiple polling tokens to the virtual endpoints in each period. The HID responds either with sensor data, or dummy data when there is no new sensor data. The dummy data keeps the polling proceeding, allowing time for more sensor data to appear in the same period.

The application provides a vehicle trajectory positioning method, device and electronic equipment in a no-positioning-signal scenario. The method comprises: acquiring a multi-modal data stream of a vehicle in a target time period, the target time period being a time period from the vehicle driving out of the no-positioning-signal scenario to a positioning-signal scenario; performing a mirror flipping operation on the multi-modal data stream with the end timestamp of the multi-modal data stream as a center point to obtain an initial pseudo data stream of a preset time period of the vehicle; performing an inverse inversion processing on data values in the initial pseudo data stream to obtain a pseudo data stream, and inputting the pseudo data stream into a Kalman filter to perform fusion positioning processing on the pseudo data stream by the Kalman filter to obtain a driving trajectory of the vehicle in the target time period. The problem that the prior art needs to increase a reverse filter to realize vehicle positioning in a no-signal scenario by a forward and reverse filter and has a high no-signal positioning cost is solved.

According to one aspect of the present disclosure, a memory device is provided. The memory device includes a memory cell array and a peripheral circuit coupled to the memory cell array. The peripheral circuit is configured to receive a first instruction, in response to the first instruction, generate a dummy data and write the dummy data in the memory cell array, receive a second instruction indicating to perform a read operation in the memory cell array, and send a returned data based on the read operation. In response to a read data of the read operation comprising the dummy data, the returned data is set feature information indicating that the returned data comprises the dummy data.

An example electronic device may include communication circuitry, memory including storing instructions, and at least one processor including processing circuitry. The instructions, when executed by the at least one processor individually and / or collectively, may cause the electronic device to detect an event for an emergency situation, transmit a registration request message including on-demand S-NSSAI to a mobility management entity, receive a registration accept message including the on-demand S-NSSAI, transmit an establishment request message for a PDU session corresponding to the on-demand S-NSSAI to the mobility management entity, receive an establishment accept message for the PDU session from the mobility management entity, and in response to receiving the establishment accept message, transmit dummy data to a user data entity via the PDU session until an end of the event to maintain the PDU session.

The present disclosure relates to a communication apparatus and method for efficiently carrying out digital pre-distortion (DPD) on a transmission signal of a wireless communication system according to various operating conditions. The communication apparatus that carries out DPD on a transmission signal of a wireless communication system, according to an embodiment of the present disclosure, may comprise a power amplifier that amplifies the transmission signal. The communication apparatus may comprise a DPD apparatus that is connected to an input terminal of the power amplifier and carries out, on the basis of operating condition data of the communication apparatus, a DPD operation by using fake data generated for the transmission signal. The DPD apparatus may include a generator based on a conditional GAN. The generator may be configured to receive, as inputs, the operating condition data and input / output signals collected from the power amplifier and generate, on the basis of data acquired through repeated training of the conditional GAN using the collected input / output signals and the operating condition data, the fake data learned to compensate for nonlinear distortion at the output of the power amplifier.

The application discloses a tunnel scene LoRa-Mesh self-organizing network Beidou positioning system, comprising: a tunnel scene modeling module, which is used for generating a tunnel three-dimensional structure model and outputting structure deviation correction parameters; a deployment parameter generation module, which is used for generating a LoRa-Mesh backbone node deployment coordinate set, a Beidou positioning signal transmitting unit deployment coordinate set and a coverage parameter set; a LoRa-Mesh backbone node self-organizing network module, which is used for completing deployment and constructing a LoRa-Mesh network; a Beidou positioning signal transmitting module, which is used for completing deployment and reporting, via LoRa-Mesh, the deployment coordinates and working state data of a transmitting unit; a Beidou positioning signal simulation generation module, which is used for generating adaptive pseudo data one by one in combination with Beidou satellite original data; and a pseudo data issuing and transmitting control module, which is used for issuing pseudo data and controlling a transmitting unit to transmit Beidou radio frequency positioning signals. The application realizes stable coverage and continuous transmission of Beidou positioning signals in a tunnel, and improves positioning precision, system reliability and tunnel scene adaptation capability.

Methods, devices, and systems for managing memory devices are provided. In one aspect, a memory system can include a memory device including memory pages, a memory controller coupled to the memory device. The memory controller is configured to send, to the memory device, a first command indicating to identify dummy data, and in response to receiving, from the memory device, a response indicating that a first memory page is a zero page and without receiving data of the first memory page, determine data comprised in the first memory page as dummy data.

The present invention discloses a solder defect detection method and system based on small target recognition. The method collects solder images, performs image enhancement on the solder images, obtains a pseudo data set of the solder images, constructs a fast deformation convolution kernel, and based on the fast deformation convolution kernel, adopts a YOLOv5 neural network to construct a solder defect detection network and adopts a pseudo data set of the solder images to train the solder defect detection network. The trained solder defect detection network is used to identify solder defects, thereby solving the technical problem of low defect detection accuracy of micron-level solders. Solder defects can be identified quickly and accurately through the constructed solder defect detection network.

This invention provides a continuous learning image classification method based on generative playback, relating to the fields of continuous learning and computer vision. This method uses an image classification network to continuously learn the image classification task; it uses an image generation network to learn the distribution of images for the current task, and when a new task arrives, it generates images of the learned task as playback data, mitigating the catastrophic forgetting problem of the image classification network during continuous learning. The method employs a Teacher-Student architecture, using the image classification network for the (t-1)th task as the Teacher model, and initializing task T with its parameters. t An image classification network is used as the Student model; the Teacher model remains fixed during subsequent training to assist the Student model in remembering old knowledge. When a new task arrives, an image generation network generates pseudo-data from past tasks, enabling the model to distinguish between old and new task knowledge when learning a new task.

The invention relates to a multi-modal artificial intelligence generation content unified identification method and system. For multi-modal to-be-identified data, directly identifying the multi-modal to-be-identified data by adopting a fast thinking mode or performing interpretability identification by adopting a slow thinking mode based on the scene demand identification large model; the large identification model is trained by adopting a fast and slow thinking dual-mode two-stage training framework, an interpretable authentic identification data set is constructed, and standardized labeling from evidence extraction, space-time positioning to thinking chain construction is carried out. In a dual-mode supervision fine tuning stage, for a fast thinking mode, supervision fine tuning is utilized to realize fast discrimination and multi-classification traceability in a high-concurrency and low-delay scene; for a slow thinking mode, learning a complete interpretable reasoning chain from evidence extraction to criterion induction to conclusion generation; then, the model output is optimized through a preference alignment stage. Compared with the prior art, the method can provide a unified and reliable technical scheme for authenticity verification and interpretable review in a complex content scene.

A driving device is provided. The driving device includes a driver and a timing controller (TCON). The driver is coupled to a pixel circuit and a backlight circuit of a display panel. The TCON is coupled to the driver. In a frame cycle, after a delayed interval, the TCON outputs output image data to the driver, such that the driver drives the pixel circuit and the backlight circuit according to the output image data and backlight image data at the same time. Alternatively, when the driver drives the pixel circuit, after an interval based on dummy data of the backlight image data, the driver drives the backlight circuit according to available data of the backlight image data.

The present invention discloses a multimodal visual language understanding and positioning method, device, terminal and medium, comprising: constructing a visual language understanding and positioning model; processing unlabeled image data to generate multiple pseudo-label sources; learning a visual language understanding and positioning model for each source respectively; selecting all pseudo data sources step by step from easy to difficult according to the average number of entities in the expression text of each source based on a multi-source synchronous adaptive algorithm; performing sample instance-level reliability calculation on each paired data of the image and the pseudo-label based on the multi-source synchronous adaptive algorithm, obtaining an optimal sample subset through reliability evaluator selection, reliability threshold selection and a greedy algorithm, and learning an optimal visual language understanding and positioning model based on the optimal sample subset; the present invention proposes an unsupervised visual language understanding and positioning method, which can adapt a multimodal basic large model CLIP based on multi-source pseudo-labels to solve the problem of unsupervised visual positioning.

The embodiment of the invention provides a normalized structure reconstruction and feature sharpness perception online adaptation method and related equipment, and belongs to the technical field of transfer learning. The method comprises the following steps: a normalization layer reconstruction stage: obtaining a batch normalization model; generating pseudo data matched with the source domain data distribution by using a data-free distillation technology; replacing a batch normalization layer in the batch normalization model with a grouping normalization layer or a layer normalization layer to form an initial student model; performing knowledge distillation on the initial student model by using the pseudo data to migrate knowledge of the batch normalization model to the reconstructed student model to obtain a normalization layer reconstruction model; an online adaptation stage: obtaining a test sample; screening the prediction confidence of the test sample based on the normalization layer reconstruction model to obtain a reliable sample subset; calculating feature sharpness loss based on the reliable sample subset; and with the purpose of minimizing the feature sharpness loss, updating parameters of the normalization layer reconstruction model to complete online adaptation.