16 results about "Semantic interpretation" patented technology

This invention discloses a method for air-ground collaborative risk prediction and dynamic electronic fence construction based on a generative large-scale model, comprising: obtaining a standardized air-ground collaborative multimodal spatiotemporal dataset; generating a structured state representation for risk inference; outputting risk semantic interpretation, candidate risk links, and risk factor embedding representations; constructing a spatiotemporal risk field and performing forward prediction based on the risk factor embedding representation and the air-ground collaborative state map, obtaining a risk probability heatmap and uncertainty map within the future prediction time window; generating a three-dimensional dynamic electronic fence set, and assigning a confidence weight and an effective timestamp to each fence boundary; and performing event-triggered iterative updates to the air-ground collaborative state map, risk probability heatmap, and three-dimensional dynamic electronic fence set based on real-time new data. In dynamic scene simulation, this invention reduces the frequency of false fence triggering compared to periodic recalculation methods, while maintaining a significant decrease in risk omission rate.

This invention discloses an automatic mathematical formula conversion method and system integrating image recognition and semantic processing, relating to the field of data processing technology. The method includes: performing symbol-level recognition on mathematical formula images to obtain an original symbol data set; parsing the structured data to generate a data expression set containing symbol units, categories, positions, and interrelationships; constructing a candidate structure space for mathematical expressions representing multiple potential semantic interpretations; introducing a set of mathematical semantic constraints for consistency evaluation to determine the target mathematical expression structure; and performing structure-preserving formula conversion processing to generate the mathematical formula representation result. This invention solves the technical problems of insufficient structural ambiguity resolution, missing mathematical semantic information, and structural distortion during format conversion in existing technologies, leading to insufficient accuracy and reliability in mathematical formula recognition and conversion. It achieves the technical effect of accurate and unambiguous parsing of mathematical formulas and improves the accuracy and reliability of formula recognition and conversion.

A computer-implemented method is provided for converting a human intent of a user into an artificial intelligence (AI) prompt. The method includes receiving a textual statement from the user, the textual statement comprising the human intent; employing an interpretation generation engine to generate a plurality of semantic interpretations of the textual statement; receiving a selection of a first semantic interpretation of the plurality of semantic interpretations from the user, the first semantic interpretation comprising a clarified human intent; generating the AI prompt based at least in part on the selection such that the AI prompt is configured to elicit a response from an AI model based on the clarified human intent; and sending the AI prompt to the AI model in order to generate the response.

This invention discloses a motor fault diagnosis method based on a large language model, belonging to the field of motor fault diagnosis. The method includes: first, collecting three-phase current data; then, using dual-tree complex wavelet transform to extract the amplitude and phase information of multi-layer complex wavelet coefficients as signal fault features, avoiding manual feature selection; next, constructing a multi-layer feature extraction and hierarchical weighted domain adversarial collaborative training framework that integrates grouping and Browsing optimization to extract multi-scale features and enhance domain invariance, thereby improving cross-condition generalization ability; finally, aligning the extracted signal embedding and text embedding across modalities through category-level comparative learning, and fine-tuning the large language model to recognize signal modes. This invention improves fault classification accuracy while enhancing model generalization and achieving textual semantic interpretation of diagnostic results.

The application relates to the technical field of semantic analysis, in particular to a bid evaluation clause extraction system based on semantic analysis, which comprises a bid text analysis module, wherein according to a bid text segment in an extracted bid, concept node information entropy weight is generated, the concept node information entropy weight is calculated with an extracted node associated semantic vector, and a binary semantic feature fingerprint is generated. In the application, the maximum weight matching node set is subjected to position alignment splicing, numerical parameter and text string reconstruction, and an index label is established in combination with entity name content and time node value, finally forming a bid evaluation rule extraction dictionary, so that the output result is changed from pure semantic interpretation into a structured rule unit which is searchable, archivable, callable and reviewable, and the bid evaluation rule extraction scene presents stronger clause restoration capability, rule aggregation capability, heterogeneous expression compatibility and subsequent reuse capability.

The present application relates to the technical field of industrial automatic monitoring, in particular to a cooling pipeline component monitoring system based on multi-modal visual fusion, comprising visual extraction, linear feature modeling, feature fusion and visual semantic interpretation modules. The present application obtains visible light texture, infrared radiation and depth geometric features through neural network branches; constructs a cross-modal kernel attention network integrating a transformer and a B-spline basis function, models the heat conduction gradient using the basis function, and generates a dynamic diffusion spatio-temporal correlation vector; uses a spatial attention residual network to weight and fuse consistent features, generating a multi-modal semantic fusion feature map; finally, an average standard Euclidean distance damage index model is called to evaluate feature deviation and accurately locate the three-dimensional coordinates of the leakage point. The present application improves the accuracy and quantity-effect level of pipeline monitoring.

A multi-agent collaborative control method and system for constructing a control plane core closed loop, a logical global unique authoritative state structure is constructed on the server as a unified data benchmark, and concurrent requests are classified and routed before execution, which are divided into core governance requests and state writeback requests; the core governance request is decided by the pre-action decision layer and enters the unified change pipeline, and the state writeback request is written into the authoritative state structure after the execution result, state summary or diagnostic information is semantically interpreted and absorbed by the controlled writeback channel; the system performs controlled submission on the state application result and writeback mapping result, generates an immutable event record, and outputs incremental data projection according to the updated authoritative state structure, wherein the runtime capability plane and external capability services do not have the right to directly write state updates to the authoritative state structure.

The application belongs to the field of knowledge retrieval enhancement generation, and discloses a multi-path mixed knowledge retrieval enhancement generation method and system applied to a vertical field. Through collaborative optimization of query understanding, multi-path recall, result reordering and knowledge fusion generation, the whole process from query to content generation is enhanced. A query rewriting strategy based on a large language model is introduced to improve semantic analysis and intent recognition capability. A multi-path recall mechanism is adopted, combined with text feature retrieval, vector semantic retrieval and multi-hop retrieval mechanism based on brain inspiration, to realize high-coverage recall of deep semantic information. Through multi-stage reordering optimization by a multi-stage reordering model, the accuracy and robustness of the recall result are improved. Finally, in the knowledge fusion and generation stage, weighted average or graph attention mechanism is used to realize dynamic fusion of multi-source knowledge and context-aware generation, thereby significantly improving the knowledge accuracy, generation consistency and semantic interpretability in professional field tasks.

This invention discloses a knowledge-driven method for geochemical data analysis, processing, and report generation. The method includes: receiving raw geochemical tables and a first geological knowledge base; generating a first standard table through field mapping, cleaning, and standardization; performing density outlier screening on a first sample set to form a second and third sample set; generating a first symbiotic deviation table based on element symbiotic relationships; splitting the second sample set into a fourth and fifth sample set; merging the fifth sample set into the third sample set to form a sixth sample set; training a reconstruction model based on the sixth sample set to generate a first reconstructed data table; and then combining the first geological knowledge base to generate a spatial association table, a combination matching table, and an evidence table, forming a first semantic interpretation package and generating a geochemical report. This invention improves the reliability and traceability of geochemical analysis results through continuous processing of anomaly identification, data reconstruction, and evidence-based report generation.

A technique extends a file system infrastructure of a storage system to provide a custom namespace within a pathname of a logical construct configured to invoke semantically interpretative context as a command embedded in a data access protocol request issued by a client and directed to the logical construct served by the storage system, without alteration to the data access protocol. The extension includes a “plug-in” engine of a data access protocol server executing on a network protocol stack of the storage system. The engine operates to extract a pathname from the request to determine whether the custom namespace incorporating the command is present and directed to the logical construct. If so, the engine semantically interprets the command within a context of the custom namespace to essentially convert the command to one or more predefined operations directed to the logical construct. The storage system then performs the operations and returns the results to the client.

This invention discloses an intelligent and transparent management and control system and method for testing centers based on the Internet of Things (IoT). The system includes: constructing a testing center and platform architecture, establishing a correspondence between resource identifiers; parsing work order elements and distributing them to corresponding testing equipment; completing testing data acquisition and preprocessing to obtain standard testing data; constructing an improved Mask2Former model and outputting masking results; outputting semantic interpretation information based on zero-shot semantic segmentation; generating traceability records and testing reports, and completing security protection. This invention, by integrating the improved Mask2Former model and zero-shot semantic segmentation, achieves transparent supervision of the entire testing process, a closed-loop traceability system for data and video, and safe and reliable operation.

This application provides a method and system for harmful video recognition based on collaborative enhancement, belonging to the field of harmful video recognition technology. The method includes: obtaining effective knowledge through a bidirectional knowledge coupling processing approach combining RAR visual recognition enhancement, RAG knowledge enhancement, and semantic analysis; constructing an evidence chain based on the effective knowledge; performing multi-dimensional generative reasoning on the evidence chain to obtain a reasoning result; mapping the confidence level, semantic consistency score, and credibility of the reasoning result corresponding to candidate symbols to a multi-dimensional risk assessment index vector; and identifying harmful information in the video segment to be evaluated based on the multi-dimensional risk assessment index vector. This application transforms the identification process of the video segment to be evaluated from the traditional feature classification model into a closed-loop process of symbol recognition – semantic interpretation – multi-dimensional risk assessment, and also supports symbol semantic tracing, effectively improving the accuracy of harmful symbol recognition.

This invention provides a fine-grained method and system for detecting harmful memes based on information theory semantic decompression, belonging to the field of multimodal information processing and internet content security. It includes: extracting image and text features of the meme to be detected and fusing them into a joint vector; calculating the risk value of the meme based on the joint vector and classifying it into low-risk or high-risk memes; using the joint vector of the high-risk meme as a query vector to retrieve an external explicit semantic explanation from a pre-built semantic decompression knowledge base, and concatenating and fusing it with the original high-risk meme features as the semantic decompression context to construct an enhanced input prompt; and outputting the fine-grained category of the meme based on the original low-risk meme or the enhanced input prompt. This invention significantly improves the model's ability to understand metaphors and emerging hate symbols, enhances fine-grained detection accuracy, and significantly reduces the system's computational overhead and inference latency.