40 results about "Semantic context" patented technology

This invention relates to the field of natural language processing technology, specifically to a method and system for matching the intent of intelligent NPC dialogues in multimodal interaction. The method involves real-time acquisition of multimodal data, generating multimodal semantic features through submodal preprocessing; integrating the semantic features of each modality using a cross-modal fusion module based on semantic association, generating a candidate intent set based on semantic context representation and combined with a semantic parsing module and predefined intent templates; tracking changes in user intent in real time through a continuous semantic learning mechanism and an interactive memory module, combined with a Bayesian update method, dynamically adjusting the confidence level of each intent in the candidate intent set, and filtering the final intent; constructing an NPC semantic cognition model, and performing semantic consistency analysis to perform semantic checks on user input and NPC dialogue state; combining the final intent and a decision engine to generate a dialogue strategy and output synchronized response content; this invention improves the accuracy of dynamic matching of intelligent dialogue intents.

This invention discloses a method and apparatus for renewable energy power prediction based on time-series discrete labeling, belonging to the field of renewable energy power prediction; it includes: acquiring historical observation multivariate time series of target power plants; constructing and training a time-series labeling mapping, dividing the multivariate time series into blocks and encoding them to obtain a continuous latent representation; normalizing the continuous latent representation and the introduced learnable codebook respectively, and allocating discrete indices using nearest neighbor search to obtain a discrete time-series label matrix; expanding the discrete time-series labels output by the trained time-series labeling mapping and incorporating them into the unified vocabulary of a pre-trained language model to construct a conditional autoregressive generative model, and performing fine-tuning training while freezing the backbone network parameters of the pre-trained language model; given the environmental semantic context and the historical time-series label sequence obtained through the time-series labeling mapping, generating a future discrete label sequence based on the trained model in an autoregressive manner, and inputting the generated future discrete label sequence into the decoder of the time-series labeling mapping to reconstruct a prediction sequence in the continuous domain.

The present application belongs to the technical field of image processing, and specifically relates to a remote sensing image compression and reconstruction method based on feature perception and latent diffusion super-resolution. The method comprises an encoding and compression stage and a decoding and reconstruction stage. The encoding and compression stage performs intelligent analysis, selective compression and data packaging on the original high-resolution remote sensing image, intelligently identifies and separates the feature region and the homogeneous background region in the image through a convolutional neural network, and differentially processes the two regions, with the feature region being recorded at high precision and the homogeneous background region being aggressively down-sampled. The decoding and reconstruction stage recovers a high-resolution reconstructed image from the compressed data packet, and performs reconstruction through a latent diffusion super-resolution model. The model can utilize the semantic context of the image to generate visually natural textures, effectively avoiding overall blurring and other problems caused by traditional interpolation, and thus enabling the image to still well maintain key feature details under high compression ratio.

The application relates to the technical field of artificial intelligence, and provides a pedestrian attribute analysis method and device, a storage medium and an electronic device. The pedestrian attribute analysis method comprises the following steps: obtaining semantic context information of a pedestrian attribute in a pedestrian image; the semantic context information is obtained by analyzing the pedestrian image or associated information thereof, and comprises at least one of individual attribute association, group attribute information and space-time constraint information; and performing a pedestrian attribute analysis task by using the semantic context information, wherein the pedestrian attribute analysis task comprises at least one of a model training task, a model inference task and a post-processing task. When analyzing the pedestrian attribute, the semantic context information of the pedestrian attribute, such as the individual attribute association, the group attribute information and the space-time constraint information, is used in the training, inference and post-processing stages, so that the accuracy of the pedestrian attribute analysis can be significantly improved.

The application discloses a remote sensing image binary change detection network based on frequency domain feature interaction. The network is aimed at the pseudo change problem caused by light, season, sensor noise and complex background in high-resolution remote sensing images, and a binary change detection network composed of a hybrid encoder, a frequency domain feature interaction module, a difference feature enhancement module and a lightweight decoder is constructed. The hybrid encoder extracts multi-scale spatial detail features through a spatial encoder and extracts global semantic context through a semantic encoder, so as to give consideration to local boundary texture and advanced semantic information. The frequency domain feature interaction module converts the double-time-phase features into the frequency domain, generates a content-aware filter according to the semantic context, and applies asymmetric frequency domain filtering to the two time phases respectively, so as to suppress the pseudo change caused by the difference in imaging conditions. The difference feature enhancement module further extracts robust change difference features through multi-scale spatial interaction, frequency-aware double-gating and lightweight refinement, and finally outputs a binary change mask by the decoder. The application can improve the change region recognition accuracy, boundary positioning ability and robustness in complex remote sensing scenes.

This invention discloses a context-based copy-paste data augmentation method, system, and storage medium for multi-class remote sensing target detection, belonging to the field of image processing and target detection technology. The method is implemented through the following steps: First, prepare a multi-class remote sensing dataset and set hyperparameters; second, generate copy regions of spatial context, using the DBSCAN clustering method to preserve target spatial context information; next, generate paste regions of semantic context, using a ResNet50 network to extract background feature vectors and match candidate paste regions; finally, perform the paste operation and update the annotation information. This invention generates copy regions through clustering, preserves spatial context, and matches paste regions based on semantic context, making the augmented image closer to the real image. It effectively alleviates the long-tail problem of datasets, enhances the network's learning balance, improves detection performance in low-sample classes and complex scenes, and can be seamlessly integrated with various remote sensing target detectors.

The application discloses a kind of based on semantic context's maintenance information aggregation search method and system, it is related to natural language processing technical field;After text analysis to search request, based on the maintenance semantic context extraction of preset static maintenance semantic label library obtains dynamic context;After the structured organization construction of dynamic context and search request, multiple semantic equivalent rewriting is carried out to obtain multiple rewritten search requests;Based on the scheme maintenance result search of preset multi-source maintenance knowledge base to all rewritten search requests obtains search result set;Based on dynamic context, search result set is sorted, and the first preset search result is obtained as final output.Analysis user natural language request, then rely on static maintenance semantic label library to complete dynamic context, then by structured fusion and multi semantic rewriting search, aggregate multi-source knowledge without omission, to quickly obtain the accurate scheme of adaptation, improve search accuracy while also greatly improve search efficiency.

The application discloses a kind of untrained intelligent agent context self-adapting method and system for open world self-view action recognition, and constructs the collaborative reasoning intelligent agent framework supported by visual context engine, semantic context engine and reflection strategy library, wherein, visual context engine extracts robust hand-object interaction clues, semantic context engine provides context information such as object category, action mode and occurrence scene related to human action, and reflection strategy library provides suitable prompt words for large multi-modal model according to historical prompt strategy application effect and newly identified object and action.The application proposes an integrated reflection intelligent agent collaborative framework suitable for completely open world setting, and the method has dynamic self-evolution capability for open world new action not seen before, so that the recognition of human action type under the first view of open world can be completed without any model training.

This invention discloses a knowledge graph completion method and system based on path-structure fusion, relating to the field of data storage technology. The method includes: at the data preprocessing level, a path filtering strategy based on PMI (Progressive Mindset Analysis) is used to accurately eliminate redundant and noisy paths by quantifying the co-occurrence correlation between entities, providing high-quality semantic context for subsequent inference. Secondly, during path modeling, the Mamba state-space model is used for path sequence modeling, and multi-level attention is used to aggregate path information. Thirdly, at the decoding and scoring level, a dual-tower architecture of path-structure is constructed, adaptively fusing and scoring global path semantic features with local graph topological features. Finally, through a mechanism of attribute completion—similar node determination—bidirectional path search, high-scoring interpretable paths are automatically mined from known entities and their similar nodes, and a hybrid scoring formula is used for comprehensive evaluation, ultimately outputting the inference link.

This invention belongs to the field of natural language processing and dialogue emotion recognition, specifically involving an emotion classification method based on a hybrid expert model and a large-scale model collaboration. The method first encodes the dialogue text to construct basic features. Then, it extracts semantic, contextual, and knowledge-enhanced features through heterogeneous expert networks, and uses a dynamic routing gating network to weightedly fuse the features output by different experts, forming a unified emotion feature representation. Next, this feature representation is mapped to a soft cue vector, concatenated with the word embedding sequence of the original text, and input into a pre-trained large-scale model for emotion inference. During the training phase, cross-entropy loss is used as the target, updating only the parameters of the expert network, routing gating network, and soft cue mapping network, achieving efficient end-to-end parameter training. This invention significantly improves the accuracy of emotion classification by fusing multi-dimensional features and combining the common-sense reasoning ability of a large-scale model, while also achieving higher training efficiency.

This invention discloses a morphological adaptive and feature decoupling network method for medical image segmentation. Addressing the technical challenges of segmenting small medical lesions, this invention achieves high-precision and robust segmentation results through the collaborative design of a morphological adaptive ResNeXt module and a context-details decoupling module. The heterogeneous dual-stream architecture of the context-details decoupling module achieves explicit decoupling of semantic context and high-frequency details. Combined with large-kernel spatial gating and global residual connections, it effectively filters background noise, prevents weak features from being overwhelmed, and significantly improves the detection rate of small lesions. The morphological adaptive ResNeXt module's fully grouped geometric alignment strategy configures dedicated offsets and modulation scalars for each semantic subspace, suppressing deformation drift and allowing the receptive field to adaptively fit the complex, non-rigid boundaries of the lesion, eliminating edge jaggedness and providing reliable computer-aided support for clinical screening of small medical lesions.

The application relates to the field of artificial intelligence and natural language processing technology, and discloses an AI follow-up robot intelligent interaction method and system, wherein the method comprises the following steps: extracting a structured entity; constructing a semantic context vector based on the structured entity; performing emotion source classification according to the emotion polarity; outputting an individualized response expected time; judging whether the response is overdue and triggering a compensation reminder strategy. Compared with the follow-up interaction response mode in the prior art which depends on a rule template or a keyword trigger, especially in the unstructured scene where a patient expresses the drug experience, subjective feeling or non-timely response in free language, the technical problem that accurate recognition, emotion understanding and individualized compensation are difficult to realize is solved, and since the application introduces a structured label extraction mechanism, fuses an emotion discrimination method, and combines a response prediction model of historical behavior and emotion state, sensitive information recognition in doctor-patient interaction is realized, and the understanding ability and follow-up interaction efficiency of the AI follow-up robot are improved.

The invention relates to a computer-implemented method for determining a time and a type of a maintenance measure of a fuel cell system (1) having a plurality of components (4), comprising the following steps: - determining (S1) an abnormality of a component (4) in the fuel cell system (1); - if an abnormality is detected, recording (S3 - S6) operating variables and determining operating states of the fuel cell system (1) and the components (4) of the fuel cell system (1), - assigning (S7) operating variables and operating states to the component (4) of the fuel cell system (1) affected by the abnormality in a semantic context of a knowledge graph; - carrying out (S9) a state prediction on the basis of the operating variables and operating states using the knowledge graph in order to establish a fault suspicion; - signalling (S10) a necessary maintenance measure, in particular by defining a time for carrying out a maintenance measure and the type of maintenance measure to be carried out, depending on the fault suspicion or in particular a provided uncertainty of the fault suspicion.

This invention discloses a training and inference method for complex type inference at the Python repository level. It constructs a semantic graph of types; organizes type semantic contexts and constructs single-round type inference training samples, performing supervised fine-tuning on a large language model to obtain a single-round type inference base model; represents the predicted type and reference type as a type tree, constructing verifiable structured rewards; constructs multi-round refinement objectives according to the depth of the reference type, aggregating round-by-round refinement rewards, termination penalties, and final result rewards to obtain a hierarchical reinforcement learning model; constructs a repository-level scheduling order; for the multi-round refinement process, in each round, multiple candidate outputs are mapped to a depth-constrained abstract type space, and the stage decoding result is obtained through type space self-consistency aggregation, outputting the final predicted type. This invention can simultaneously enhance the cross-process semantic utilization capability, deep structure inference capability, and repository-level type consistency in complex type inference.

This application provides a network closed-loop control method, device, medium, and program product based on tail delay prediction. The method includes: acquiring flow identification information and training semantic context of the target set communication flow during the distributed synchronous training process of artificial intelligence; determining the current path and candidate paths based on the flow identification information and collecting corresponding telemetry data; predicting the tail delay index of the current path and each candidate path within a preset prediction window by combining the training semantic context and telemetry data; evaluating the path based on the tail delay index and path adjustment cost of the candidate paths, using the current path as a comparison benchmark, and determining the target path; adjusting the target set communication flow to the target path when the preset path control conditions are met and the suppression conditions are not hit; and executing hold or back control based on the operational feedback information after path adjustment to update the tail delay prediction process. This application can identify the risk of tail delay degradation of the set communication flow in advance, reducing false triggering and path oscillation.

A scene layout anomaly perception text-to-image large model hallucination detection method, system and device, the method is: acquiring scene layout anomaly detection data and preprocessing, obtaining object instance set; Construct a scene layout anomaly perception model for text-to-image hallucination detection; Including double flow graph module, cross-modal structure alignment module and anomaly sorting module; Train the cross-modal structure alignment module and the anomaly sorting module to obtain the model weight file of the training iteration; Read the model weight file and perform hallucination detection on the text-to-image large model; The invention decouples the detection of hallucination into two dimensions of semantic context perception and geometric structure modeling of object instance through the innovative semantic-geometric double flow interaction and cross-modal alignment framework, the consistency between the two is inferred through the cross-modal interaction Transformer network, the mismatch signal of semantic and geometry is accurately captured, and the fine-grained hallucination such as unreasonable object attribute and mismatched object relationship in the generated image is accurately positioned.

The application discloses a picture-text recognition and translation system based on semantic recognition, which comprises a document analysis module, an image local text recognition module, a context perception module, a translation instruction generation module and a picture-text reorganization module; the document analysis module can separate original picture-text mixed arrangement documents into continuous text sets and image region sets; the image local text recognition module can obtain local text segments and determine basic semantic vectors by processing the image region sets; the context perception module can calculate the semantic correlation degree S of the basic semantic vectors and each paragraph vector; the translation instruction generation module can generate translation input sequences; and the picture-text reorganization module can generate translated image sets through the visual attributes and position information of original image regions, and then obtain bilingual translation documents. The application can make full use of complete semantic contexts provided by document main texts by the context perception module to realize the translation of image local texts.

This invention relates to the field of computer vision technology and discloses an AI-synthesized video detection method and system based on multi-agent collaboration. The method includes: acquiring the video to be detected and sampling keyframes; understanding the keyframes based on a visual language model and generating a forgery hypothesis for the current round based on the generated global semantic context; generating a spatial-temporal routing query; performing target localization and segmentation on the video to be detected based on the spatial-temporal routing query; extracting structured physical evidence from the target region video; performing evidence-constrained reasoning to obtain a preliminary judgment result and confidence level; performing consistency checks on the execution trajectory to obtain a reliability score and feedback information, and outputting the final authenticity judgment result. This invention improves the reliability, generalization ability, and traceability of AI-synthesized video detection by combining contextual modeling, hypothesis-driven routing, multi-dimensional physical evidence collection, evidence-constrained reasoning, and closed-loop reflective verification.

This invention discloses a method and system for selecting personally identifiable information based on a dual-tower model architecture. First, a query tower integrates multi-source information to generate query vectors, encompassing privacy-sensitive features and semantic context information, avoiding excessive privacy protection or significant loss of text utility. Second, a candidate option tower independently encodes candidate options into vectors, fully preserving and accurately representing the features of each candidate option, meticulously capturing unique information, ensuring the vectors truly reflect their own characteristics, and accurately measuring similarity to the query vector. Third, by calculating the matching score between the query vector and candidate option vectors, the degree of matching between the candidate option and the original information is quantitatively evaluated. The similarity calculation considers multi-dimensional features, accurately reflecting their semantic and feature similarity. The highest-scoring option is selected, comprehensively balancing privacy protection and text utility, ensuring that the candidate option retains key features while maintaining a reasonable similarity to sensitive information, achieving a balance between the two.

The present application relates to the technical field of unmanned aerial vehicle three-dimensional path planning, and particularly relates to an unmanned aerial vehicle path planning method based on semantic context perception reinforcement learning. The present application comprises: fusing multi-source sensing data collected by the unmanned aerial vehicle to construct a multi-attribute three-dimensional map; constructing an envelope region between the current position and the target position of the unmanned aerial vehicle, and obtaining an environmental context vector; inputting the environmental context vector into a pre-trained reinforcement learning strategy network to output a set of path planning preference weights; obtaining path planning weights through path connectivity testing; constructing a comprehensive cost function based on the path planning weights; using a global path planning algorithm to search for a feasible path in the multi-attribute three-dimensional map according to the comprehensive cost function, and converting the feasible path into continuous flight instructions. The present application improves the environmental perception capability and navigation decision concentration, takes into account adaptability and planning safety, and has strong generalization capability and rapid migration characteristics.

This invention discloses a large model-assisted computer vision method and system for detecting defects in industrial products, relating to the field of industrial vision technology. For each candidate defect region, the system extracts its visual feature vector, semantic discriminant vector generated based on LLM, and knowledge association vector retrieved from a knowledge graph in parallel. Visual features provide pixel-level evidence, the semantic discriminant vector is the logical reasoning of visual evidence based on contextual reports by LLM, and the knowledge association vector incorporates historical experience and causal chains. A fully connected classifier performs a fusion decision on these three types of vectors, integrating visual evidence, semantic context, and historical experience. This enables the system to not only output high-confidence classification results but also generate explainable discriminant reasons, transforming defect detection from an isolated classification task into a comprehensive analysis process of perception, reasoning, and knowledge retrieval. This provides a data foundation for achieving truly intelligent quality diagnosis and process optimization.

The present disclosure is directed to a system and method that provides a unified context graph mapping semantically context providers to corresponding entities and / or attributes, each entity being related to one or more of the attributes, provides a first prompt comprising user selected entities to a language model (LM) to determine, based on the unified context graph and from among the context providers, a first set of relevant context providers, transforms an attribute of the selected entities into an attribute vector embedding semantically describing the attribute, compares the attribute vector embedding with vector embeddings associated with the context providers, to determine a second set of relevant context providers different from the first set of relevant context providers, generates, from the first and second sets of relevant context providers, context associated with the selected entities, and determines from a second prompt comprising the generated context a LM reply to provide to the user in response to the user input.

This invention relates to a semantic context preservation method, system, device, and medium in an AI-driven human-machine collaborative modeling system. The method includes: upon receiving an API request parsed from natural language instructions, performing semantically prioritized multi-level parsing verification on object references in the request to obtain verified object identifiers; converting the tool call request into an API request for structural analysis software based on the object identifiers, intercepting and queuing the API requests, and driving the structural analysis software to serially execute the API requests in an asynchronous execution environment; in response to a state synchronization signal, obtaining the latest object list from the structural analysis software, and fully updating the state domain based on the latest object list, while retaining the semantic mapping relationships pointing to the latest object list in the semantic domain. This invention achieves continuous and effective model state awareness and semantic references in human-machine alternating collaborative modeling scenarios through a two-layer decoupled storage, multi-level parsing, and asynchronous serialization adaptation mechanism.

This invention relates to the field of audio signal processing, and provides a method and device for semantic completion in edge-side far-field simultaneous interpretation, comprising: S1: acquiring the original noisy speech signal in the far-field environment through the microphone array of the edge device, and performing adaptive beamforming to lock the location of the target sound source; S2: performing time-frequency analysis on the residual speech spectrogram to detect low signal-to-noise ratio missing frame regions caused by far-field attenuation or background noise masking; S3: obtaining the semantic context vector of the translated text in the current session, and constructing a generative completion network based on semantic feedback; S4: using the residual speech spectrogram as the main input, and injecting the semantic context vector as conditional guidance information into the generative completion network; S5: inputting the reconstructed complete speech spectrogram into the edge-side translation model. This invention introduces "semantic-guided acoustic restoration," which can solve the problem of "unclear hearing and inaccurate translation" of weak far-field signals on edge devices, and improve the usability of simultaneous interpretation in noisy environments.

This invention relates to the field of artificial intelligence technology and discloses an automated review method for procurement and sales contracts based on a large-scale AI model. The method includes: first, based on the large-scale AI model, performing context-aware and deep semantic deconstruction on the target contract text to generate multi-level semantic expressions; then, automatically matching these expressions with a preset three-dimensional compliance risk map to obtain contextualized risk diagnosis results. Subsequently, combining the original contract semantic context, the large-scale model reconstructs the clauses to form optional revision schemes, determines the final decision through human-computer interaction, and outputs the finalized contract through global logic self-checking. Finally, the multi-level semantic expressions, risk diagnosis results, final decision, and finalized text are used as historical cases for targeted feedback optimization of the model, forming an adaptive review strategy and knowledge system. This invention can improve the efficiency of automated review of procurement and sales contracts.

This application provides an intelligent customer service system and a computer-readable medium. The intelligent customer service system employs GraphRAG (Graph Retrieval Augmentation) technology and related AI models to help the unmanned customer service system better perceive the semantic context of user questions and generate more accurate and natural answers. Simultaneously, leveraging the semantic matching capabilities of the AI ​​models, it automatically deduplicates newly entered question-and-answer data and when the intelligent customer service system restarts. The intelligent customer service system provided in this application utilizes the semantic understanding capabilities of AI models combined with the question matching capabilities of GraphRAG, effectively improving the intelligence, accuracy, and naturalness of the generated answers, and enhancing the system's ability to handle complex issues.

The application discloses a driver attention prediction method and system based on semantic guidance multi-layer spatio-temporal feature fusion. The method comprises the following steps: acquiring a continuous traffic video segment, and uniformly adjusting the continuous traffic video segment to a fixed size input into a trained driver attention prediction network; adopting a Video Swin Transformer as a backbone network to extract multi-layer spatio-temporal features, and acquiring shallow layer spatial details and deep layer semantic context; guiding deep layer semantic information to flow to the shallow layer and promoting semantic information transmission through multi-level residual connection and up-sampling operation; introducing a hierarchical feature reorganization module to adaptively reconstruct each layer feature after fusion, highlight key salient regions, suppress redundant irrelevant features, and enhance saliency expression; respectively deconvolving and decoding the hierarchical features of four independent branches to generate an intermediate saliency map, and finally splicing and fusing to generate a final prediction result. The method can accurately predict the driver attention distribution, and has important significance for the development of advanced auxiliary driving.

The present application relates to the technical field of digital technology grabbing graphics enhancement, and particularly relates to a digital technology grabbing graphics enhancement method and system, the method comprising the following steps: activating one or more semantic contexts of a scene object; for each activated semantic context, dynamically determining an importance degree value according to user attention, user interaction behavior and virtual tour narrative priority; for each activated semantic context, generating an independent enhancement rendering layer, the independent enhancement rendering layer being used to present visual performance of the semantic context; receiving the independent enhancement rendering layer and the importance degree value, fusing the independent enhancement rendering layer according to a semantic area where a pixel is located and the importance degree value, to obtain a final enhancement image; and rendering and outputting the final enhancement image in real time. The above improves the sense of reality of a virtual scene and the sense of immersion of a user.

This invention relates to a language decoding method, apparatus, and electronic device based on electroencephalogram (EEG) signals. The method includes: acquiring EEG signals from a subject and preprocessing them to obtain a neural feature time series; inputting the neural feature time series into a parallel decoding architecture, which includes at least two decoding branches for decoding sub-units of language with different orthogonal dimensions from the neural feature time series; obtaining the sub-unit probability sequence output by each decoding branch through the parallel decoding architecture; performing temporal probability accumulation and fusion on the sub-unit probability sequences of each decoding branch to obtain a stable sub-unit sequence; performing legality verification and combination of the stable sub-unit sequences according to phonological rules to generate a legal syllable sequence; and performing disambiguation processing on the syllable sequence based on semantic context to output continuous natural language text. This invention achieves effective modeling and generalized decoding of massive syllable categories under limited training data conditions.

Proposed is a novel machine learning system (1) for contract intelligence automation, and corresponding method for training the machine learning system for automated text analysis and for applying it. A plurality of contracts (2) with a plurality of clauses (22) is received by the system (1), wherein wordings of equivalent clauses (22) vary across contracts (2); contract text (21) is read into a data processing system (15); clause text chunks (121) are identified, that are equivalent across contracts (2), and assigned a contract term category (122). Considering a respective semantic context (131) of each of the contracts (2), a semantic meaning (132) of each of the clause text chunks (121) is determined and encoded in a clause embedding space (141) and stored in vector database (14); data automation tasks can be performed using entries of the vector database (14), particularly automatic consistency monitoring (197) and outlier detection across contracts (2) and a monitoring of clause (22) nuances across contracts (2) e.g. via a graphical representation (16).