44 results about "Global consistency" patented technology

This application discloses a method and system for consolidating evidence of network fraud through multimodal semantic consistency verification, comprising: collecting and modally identifying multimodal electronic evidence and generating original data fingerprints; extracting semantic information from each modal of evidence across five dimensions: time, amount, identity, behavior description, and geographical location; constructing a cross-modal evidence semantic association graph with a heterogeneous graph structure including evidence nodes, element nodes, and relational edges, and establishing attribution edges and cross-modal association edges based on element categories using differentiated semantic similarity calculation methods; generating initial feature vectors for each node, inputting the semantic association graph into a graph attention network model for consistency inference, performing consistency, contradiction, and uncertainty classification on cross-modal association edges, and obtaining a global consistency score; and performing differentiated on-chain consolidation operations based on the score to generate consolidation certificates. This invention can automatically discover semantic contradictions between different modalities of evidence, improving the quality of electronic evidence consolidation.

This application belongs to the field of industrial vision intelligence, specifically disclosing a multi-view 3D point cloud global registration method and system. Through this application, a multi-source observation data fusion mechanism is constructed. This mechanism utilizes relative pose observation to ensure the global consistency of the spatial topology of the multi-view point cloud; it utilizes point-to-surface distance observation to improve stitching accuracy and solve the point cloud layering problem; and it utilizes geometric feature observation to enhance registration robustness. During the iterative optimization of the objective function, this application introduces a dynamic weight assignment strategy based on the importance of multi-source errors at different stages, enabling adaptive weight adjustment of different observations during the optimization process and accelerating the algorithm's convergence. Before constructing the observation data, this application performs smoothing processing on the original multi-view point cloud, smoothing noise while preserving point cloud boundaries, thus increasing the accuracy of the observation results.

The present application relates to the technical field of multi-core heterogeneous computing, and discloses a storage coherent hub chip based on core integration, a storage coherent arbitration device and an adaptive control method, aiming to solve the technical problems of high cross-core memory access latency, low coupling degree of cache coherence maintenance and memory scheduling, and slow response of operation strategy adjustment under the existing multi-core architecture. The present application takes an independently packaged storage coherent core as a globally consistent unique maintenance node, integrates a single-cycle static addressing architecture, adopts a cache coherence state machine and a memory scheduling controller with deep fusion of logic layers, realizes automatic switching between robust mode and aggressive mode through a pure hardware MHM monitoring unit, and the atomic withdrawal process is transparent to the upper layer. The present application can reduce the cross-core memory access latency by more than 40%, improve the storage access throughput by 25%, while guaranteeing 99.999% operation reliability, adapting to various heterogeneous interconnection protocols, and being applicable to various application scenarios such as servers, high-frequency financial transactions, AR / VR wearable devices, edge computing, etc.

The application discloses a kind of anti-degradation multi-sensor fusion SLAM method and system, belong to robot synchronous positioning and mapping technical field.The method includes obtaining original data from camera, laser radar, inertial measurement unit and global navigation satellite system, based on VIG tight coupling optimization unit, original data is preprocessed and initial estimate in front end, obtain global pose, based on LIG tight coupling optimization unit, motion distortion correction and feature extraction processing are carried out by original data and global pose, based on two-stage factor graph optimizer, laser feature and GNSS observation information are fused in two stages, global optimization is executed, and high-precision, globally consistent pose and map are output.The application overcomes the limitation of single sensor, suppresses cumulative error, realizes high-precision, high robustness and global consistency positioning and mapping in large-scale complex scene outdoors.

This disclosure presents a method, device, and medium for heterogeneous sensor collaborative localization and map building. The method includes: generating a weighted fusion global descriptor for each platform; filtering cross-platform loop closure candidates through similarity calculation and maximizing algebraic connectivity; performing local point cloud registration on selected loops and employing a minimum vertex coverage algorithm to reduce communication overhead; dynamically electing and optimizing nodes; optimizing the decentralized pose graph using an asymptotic nonconvexity algorithm and fusing prior pose prediction with GNSS adaptive constraints; and unifying the global reference system through dynamic anchor point selection and iterative alignment. This disclosure achieves complementary advantages between backpack and UAV systems, significantly reducing communication and computational burdens while ensuring compliant operation in no-fly zones, and improving the accuracy and global consistency of collaborative localization and mapping in complex environments.

This invention discloses a composite measurement method, device, system, medium, and program product. By integrating global multi-view visual positioning and local fringe projection measurement technology, it accurately corrects the motion error of an industrial robot using externally measured pose. First, it completes multi-dimensional accurate calibration of local measurement, global positioning module, and hand-eye coordination, constructing a high-precision unified global coordinate system covering the entire measurement space. Then, through hardware triggering, it achieves strict spatiotemporal synchronization of local 3D topography data and robot end-effector global pose data, and combines intelligent path planning to complete a full-coverage scan of large-size structures, calculating the high-precision six-DOF global pose of the robot end-effector detached from its own sensing system. Subsequently, it accurately converts the local point cloud to the global coordinate system and completes seamless stitching and fusion of noise reduction, coarse registration, and fine registration, realizing high-precision, high-efficiency, and automated 3D topography measurement of large-size structures, ensuring the global consistency and accuracy reliability of the measurement results.

A method, apparatus, electronic device, and storage medium for map determination are disclosed, relating to the field of point cloud map technology. The method includes: acquiring raw point cloud data and inertial measurement data, and determining multiple keyframe poses based on the raw point cloud data and inertial measurement data; determining a first pose map, and determining edge constraints of the first pose map based on the pose estimation of each keyframe pose and the global pose observations of each keyframe pose; performing global optimization on the first pose map based on the edge constraints to determine an initial trajectory; determining candidate keyframe pose pairs based on the initial trajectory, and determining loop closure constraints based on the candidate keyframe pose pairs; determining a second pose map, and performing global optimization on the second pose map based on the loop closure constraints and edge constraints to determine the target keyframe pose; and stitching and fusing the raw point cloud corresponding to the target keyframe pose to obtain the target map. This application can effectively eliminate accumulated errors and obtain a map with optimal global consistency and higher accuracy.

This invention discloses an attention-integrated multimodal federated learning method and system for risk monitoring. On the client side, an attention-integrated hybrid early fusion module dynamically injects multimodal features into globally shared latent variables through attention mechanisms and entropy-based gating mechanisms. On the server side, a plug-and-play dual-track federated aggregation strategy decouples state aggregation from model aggregation. The state track employs robust filtering based on the geometric median, while the model track supports algorithm-independent parameter updates. Between communication rounds, a neighborhood-weighted personalized reference mechanism constructs a gradient-based semantic topology to balance global consistency and local task preferences. This invention simultaneously addresses the technical challenges of modality incompleteness, data statistical heterogeneity, and Byzantine adversarial robustness in multimodal federated learning, thus providing an efficient and robust collaborative learning solution for disaster risk perception under privacy-preserving conditions.

The application relates to the field of computer vision and mobile robot navigation, and discloses an object-level semantic visual SLAM method and system based on a hybrid attention mechanism target detection network and an ellipsoid model. The system comprises five parallel thread modules of semantic perception, visual tracking and repositioning, local mapping and fusion, loop detection and global consistency optimization. The application extracts local and global features of an image in parallel through a target detection network, and outputs high-precision semantic observations; when tracking is lost, a 2D inscribed ellipse-3D object ellipsoid dual geometric constraint is used to realize fast repositioning in cooperation with a P3P algorithm and an IoU cost function. In the mapping process, a Gaussian-Wasserstein distance is introduced to measure semantic re-projection errors, and a joint objective function containing map point geometric errors and object semantic errors is constructed to perform bundle adjustment. The application effectively solves the problem of feature extraction failure in motion blur and weak texture scenes, and significantly improves the construction accuracy of a semantic map and the survival ability of the system.

This invention relates to the field of natural language processing technology and discloses a method for resolving referential inconsistencies across long texts in cross-dialogue contexts. The method includes: segmenting the long text into blocks and vectorizing it to form a queryable vector index set; constructing a hybrid label and semantic index by combining a business entity dictionary and historical weakly supervised signals to generate a candidate entity set; inputting the candidate entity set and block summaries into a dual-channel scoring model, where the first channel uses a zero-shot common-reference classifier to generate a probability score, and the second channel is calibrated based on business rules; weighting and fusing the two to generate the final matching score for candidate entity pairs; constructing a weighted entity graph and applying the maximum spanning tree algorithm to solve for the global preliminary referential chain structure; correcting and filling in conflicting entities appearing in the global preliminary referential chain structure by combining entity type, block summary, and context vector information to form a global referential chain set. This invention has the advantages of improving global consistency and accuracy.

The application provides a federated instruction optimization method and system based on data quality control, comprising: using global initial model parameters to score instruction-response consistency of local data of each client, evaluating the quality of each sample; filtering low-quality data according to a preset quality threshold, and retaining data samples with consistency scores higher than the threshold; sorting the filtered sample data from high to low according to the consistency scores, and dividing them into multiple levels, each level containing the same number of data samples; each client uses the data samples in turn for local training according to the level order; after each round of training, the server receives local update parameters of all participating clients, and updates the global model parameters through weighted averaging; repeat the above steps until the preset round is reached. Through the implementation of global consistency and efficient data quality evaluation, combined with the hierarchical training method, the robustness of the model to noise can be improved, and the model training effect and performance can be improved.

The application discloses a multispectral target detection method based on multimodal interaction and fusion, and relates to the technical field of automatic driving multimodal information processing. The method can optimize the fusion effect in real time according to target characteristics and environmental changes, significantly improve the detection accuracy of the algorithm, meet the application requirements of real-time processing and resource limitation, supplement the missing information of each mode through interaction between the modes before fusion, and significantly improve the quality of cross-modal feature fusion. Fine-grained feature integration is adopted, which not only provides more accurate and rich local features, but also gradually establishes and enhances the long-range dependency relationship between modes, improves the global consistency and representation ability of multimodal features. Finally, the light and efficient interaction and fusion module designed by the method has stronger practicability, is suitable for current target detection tasks, can be popularized to multispectral target detection in other fields such as remote sensing, and has wide applicability and expansibility.

The application provides an automatic mapping method for complex terrain environment, and relates to the technical field of automatic mapping. The method comprises the following steps: firstly, acquiring multi-modal sensing data and performing pose calculation to construct an initial pose graph and a local mapping model; extracting spatial geometry and visual texture features to calculate an information matrix, and solving a spatial feature degradation index and a degradation type identifier; when the degradation index exceeds a preset threshold, determining a feature degradation area, calculating a dynamic observation weight matrix and extracting a local credibility topological boundary based on the feature degradation area; planning an active compensation survey trajectory based on the boundary, controlling a mapping platform to perform active compensation survey to acquire supplementary sensing data, and generating a compensation constraint edge by correlation; finally, injecting the weight matrix and the compensation constraint edge into the pose graph to perform global graph optimization solving, and outputting a globally consistent mapping product. The application solves the problems of divergence of pose solving caused by environmental feature degradation and lack of active compensation repair mechanism in the prior art.

The application relates to the technical field of military simulation data engineering, and discloses a multi-dimensional semantic knowledge graph framework construction method for military simulation, which comprises the following steps: analyzing an object model extension component of a high-level architecture to construct a simulation object model incremental state sequence, and mapping a multi-dimensional coordinate of data distribution management into a discretized grid data routing domain; analyzing a federal protocol session layer log to determine a member connection reliability index, and combining the grid data routing domain to calibrate a data reachability index of each grid unit; generating an extended semantic type identifier for a semantic assertion, constructing a semantic missing tendency evaluation model associated with the extended semantic type identifier and the data reachability index, and calculating a missing risk measurement value for an incremental semantic field; and generating a global consistency compensation factor based on the missing risk measurement value and fusing the semantic triple.

This invention relates to the field of semantic recognition technology and discloses an intelligent review system for legal application documents based on natural language processing. The system includes: extracting semantic propositional units from the legal documents to be tested, and constructing a semantic association graph based on logical polarity labels of legal subjects and behavioral predicates; using a linear superposition operator to calculate the directional decay distribution of initial weight scores along the association edges of the graph, identifying polarity break nodes in the graph, and combining weighted mapping of cross-paragraph association edges to complete global consistency verification and output the review results. This invention utilizes a legal hierarchy weight transmission model and a cross-paragraph logical anchor mechanism to discover implicit legal conflicts in long texts and eliminate semantic recognition biases caused by long-distance contextual transformations.

The application provides a multi-terminal cooperative digital media data processing system and method, and relates to the technical field of multi-terminal cooperation. Each member terminal generates a local media processing checkpoint data packet in a local audio and video processing process. Each member terminal sends the generated checkpoint data packet to a master terminal. The master terminal generates a global consistency checkpoint mark of a checkpoint frame interval after confirming that each member terminal has submitted the checkpoint data packet of the same checkpoint frame interval. After offline reconnection, the member terminal inquires the master terminal about the last valid global consistency checkpoint mark. The reconnected member terminal recovers the state configuration of the audio and video encoder at the checkpoint time and the local rendering buffer according to the global consistency checkpoint mark. The application can realize consistent recovery of the digital media processing state after multi-terminal offline reconnection, thereby ensuring the uniformity of the media content time sequence of cooperative processing.

The application discloses a kind of field of view angle parameter self-adapting calibration method based on edge computing, specifically includes: collection magnification parameter, optical parameter, image center coordinate, rotation angle parameter, environmental parameter and target pixel coordinate, form input parameter set;Performing internal parameter calibration, external parameter calibration and distortion correction, obtain correction parameter;Split local variable, establish global consistency variable and multiplier variable, execute alternate update to generate local rectification variable and residual value;According to residual value, generate penalty parameter;Call penalty parameter to complete variable correction and coordinated update, obtain rectification result;Boundary projection is executed to rectification result, generate constraint result;Constraint result is written into coordinate conversion process, and target geographic coordinate is output.The application realizes the linkage control of imaging parameter rectification result and coordinate conversion process, realizes the high-precision output of target pixel coordinate to target geographic coordinate.

The application relates to the technical field of image processing, and discloses a target detection method, a target detection device, equipment and a medium. The method comprises the following steps: acquiring a panoramic image to be detected; based on key image features in the panoramic image to be detected, adaptively slicing the panoramic image to be detected to obtain a plurality of original slice images; supplementing neighborhood information of the original slice images to obtain a plurality of supplemented slice images; performing target detection on the supplemented slice images to generate corresponding target detection boxes; and performing global consistency checking on the target detection boxes to eliminate target detection boxes that do not meet preset consistency conditions, thereby obtaining a target detection result corresponding to the panoramic image to be detected. The embodiment of the application can adapt to various panoramic image slicing scenes and improve the quality of panoramic image slicing.

The application discloses an education training scoring and optimization method based on multi-agent process collaboration training, and relates to the technical fields of intelligent education and multi-agent collaboration. The method constructs a collaborative system composed of a training guidance agent, a special evaluation agent group, an adversarial feedback agent and a global coordination agent. Through subtask decomposition of the education training task, combined with multi-modal data such as code, voice and video, the training task is evaluated in a process and multi-dimensional manner. An adversarial feedback mechanism is introduced to correct the credibility of the special evaluation results and improve the scoring robustness. Through local continuity constraints and global consistency optimization, dynamic balance between subtask scoring and overall objectives is achieved. The method can realize real-time, traceability and iterative optimization of education training scoring, improve the objectivity, fairness and generalization ability of the evaluation results, and is suitable for various online education and vocational training scenarios.

The application discloses an intelligent video summary system based on cross-modal alignment, relates to the technical fields of cross-modal alignment and video processing, and aims to solve the technical problems that in the prior art, a fine cross-modal alignment mechanism is lacked during multi-modal summary generation, the time sequence offset among frames, phonemes and words cannot be solved, and global consistency cannot be realized at the event level and the semantic paragraph level, specifically, an improved hierarchical cross-modal alignment algorithm and a dynamic weight distribution mechanism are introduced to solve the problems of multi-modal feature time sequence misplacement, semantic inconsistency and rigid weight distribution in the video summary, fine alignment of frames, phonemes and words is realized through soft dynamic time warping and sparse optimal transmission, and event level and semantic paragraph level representations are generated in combination with time sequence change detection and hierarchical attention, so that the accuracy of cross-modal fusion is improved.

This invention belongs to the field of gene sequence alignment technology, specifically relating to a haplotype-aware sequence-to-graph alignment method, system, device, and medium. This method directly embeds haplotype information into the edge association structure of a pan-genome graph, enabling the graph structure itself to express haplotype paths. Then, an improved haplotype-aware partial order alignment algorithm is used to map the sequence onto this graph, allowing it to identify and utilize the embedded haplotype information, ensuring global consistency of the alignment results. Finally, a path reuse backtracking based on global caching is used to obtain the optimal alignment path, significantly improving algorithm efficiency while maintaining accuracy.

This invention provides a method, apparatus, computer device, and storage medium for optimizing point cloud maps, belonging to the field of map optimization. The method includes: extracting a global feature map and real-time scanned current features from the point cloud map to be optimized; merging the current features and the global feature map to obtain merged features; and then updating the point cloud map to be optimized based on the merged features. This approach first ensures consistency in subsequent point cloud map optimization by extracting the global feature map, then ensures real-time performance by extracting features from the real-time point cloud, and finally unifies the positional relationship between the current features and the global feature map before merging the features. This helps reduce feature errors and improve the accuracy of the merging results, thereby constructing a more complete and accurate global feature map set. Finally, the point cloud map to be optimized is optimized based on the merged features, ensuring global consistency and real-time performance of the optimization results.

The application discloses a kind of multi-view clustering methods based on global and local anchor point learning, belong to data processing technical field, including the following steps: S1, generate target function for all views;S2, based on target function, anchor graph structure learning and global consistency structure learning are carried out, and double-layer anchor graph optimization target function is constructed;S3, based on double-layer anchor graph optimization target function, global and local anchor point learning multi-view clustering model is constructed using mapping matrix;S4, global and local anchor point learning multi-view clustering model is solved, and classification result is obtained.The application learns the local anchor point of specific view under the guidance of cross-view global anchor point selection, so as to effectively capture the local information of specific view and the cross-view global consistency of multi-view data.

This invention relates to a multimodal data screening method and system based on contrastive learning and multi-granular semantic alignment. The method includes: receiving raw multimodal data samples; calculating the cosine similarity between the visual feature vectors and text feature vectors of each sample to obtain a global consistency score; extracting key regions corresponding to cue words in the visual data, outputting multiple key region candidate boxes, extracting entity nouns in the text, and calculating the correspondence between key region candidate boxes and entity nouns to obtain a local region alignment score; performing part-of-speech tagging on the text, extracting detection targets, constructing an attention heatmap based on the word vectors of both, and calculating the overlap of attention points as a fine-grained score; fusing the global consistency score, local region alignment score, and fine-grained score to obtain a quality confidence score, and retaining or deleting corresponding samples based on the quality confidence score to complete the screening. Compared with existing technologies, this invention significantly reduces the risk of large model illusions and improves data quality.

This application provides a multi-view collaborative optimization Gaussian sputtering method, apparatus, device, and storage medium. The multi-view collaborative optimization Gaussian sputtering method includes: acquiring scene images from multiple different viewpoints and constructing a scene representation containing multiple 3D Gaussian distributions; parallel computing the rendering results of the 3D Gaussian distributions under each viewpoint, performing backpropagation to obtain multi-view gradients, fusing the multi-view gradients, and updating the parameters of the 3D Gaussian distributions; calculating the cross-view comprehensive contribution and cross-view consistency parameters of each 3D Gaussian distribution; and adaptively adjusting the 3D Gaussian distributions. This method, through multi-view parallel rendering and a viewpoint weight fusion mechanism based on the negative correlation of rendering errors, can effectively reduce the interference of local abnormal gradients generated by outliers such as shooting noise, local reflections, or accidental occlusion on parameter optimization at the source of gradient updates, thereby improving the robustness of 3D Gaussian distribution parameter updates and the global consistency of scene representation.

The application discloses a local AI inference and multi-user management system based on a protected execution environment, and relates to the technical field of artificial intelligence terminals and data security. The system comprises: a protected execution environment for storing keys, identity certificates and privacy data; a configurable-scale AI inference engine deployed in the environment; a multi-user management module for generating independent digital identities for each user and assigning independent AI instances with hardware-level isolation; a privacy data hierarchical management module for partition storage and access control according to multiple privacy levels; and a terminal cooperative scheduling module for managing multi-device AI inference requests to ensure global consistency of contexts. The system also supports end-cloud collaborative inference, with the cloud only processing desensitized intermediate feature data, deleting all data immediately after completion, and the local lightweight model automatically backing up in case of failure. In addition, the system also comprises privacy level visualization, guest mode, tamper-proofing audit, hardware feature binding, local computing power scheduling and other privacy compliance and home office auxiliary functions. The application is suitable for multi-user AI scenarios that require local processing of sensitive data.

The embodiment of the application provides a tree wired network topology inference method and related equipment, and belongs to the wired network topology sensing technical field. The method comprises the following steps: obtaining a reflection event set and an inter-port direct propagation distance from a limited number of measurement ports; constructing a tree initial connected structure based on the direct distance; determining and adding an intermediate connection point through consistency verification of cross-port reflection events, forming an updated connected structure; simulating the structure, matching and eliminating the explained part of the measured and simulated reflection events, and obtaining a differential reflection event set; taking the differential event as a driving force, generating a candidate distance set at a growth point, and iteratively updating the topology structure through verification formula growth and global consistency evaluation screening until a termination condition is met and an inference result is output. The application only needs to measure part of the ports, effectively reduces the uncertainty and computational complexity of the topology inference through the differential driving and verification formula growth mechanism, and improves the inference stability and reliability.

The application relates to an enterprise video authentication consistency verification method and system based on multi-dimensional data fusion, which comprises the following steps: issuing a dynamic visual challenge code and an acoustic interaction instruction, guiding a user to perform multi-modal interaction with an authentication physical object, a challenge code and an environment in a real physical space; after synchronously receiving an audio and video stream, a time sequence feature sequence is extracted in real time, continuous frames are abstracted into a multi-element space-time relationship graph sequence, a time sequence graph neural network is used to deeply analyze the space adjacency, motion cooperation and semantic correlation relationship among elements, and a feature vector representing global consistency is output. On this basis, four-dimensional verification is performed in parallel, and finally an adaptive decision model is used to fuse the scores of the four dimensions to output an authentication result. The application realizes a leap from single face feature verification to panoramic relationship verification, can effectively prevent the occurrence of virtual camera injection, pre-recorded video and deep fake scene, and greatly improves the authenticity of video authentication.

The application discloses an image generation method based on entity region guided diffusion, which introduces an entity region guided diffusion (ERGD for short) framework, combines a depth-aware regional attention (DARA for short) mechanism, a Gaussian soft mask, an overlap constraint and a foreground-background semantic separation strategy, and realizes efficient and fine entity-level controllable image generation; while maintaining global consistency, the method can accurately control the shape, position and interaction relationship of multiple entities, and solves the technical problems of boundary artifacts, semantic entanglement and high computational cost existing in the prior art. The application is particularly suitable for automatic driving, virtual reality, image editing and other scenes requiring multiple entity scene generation.