98results about How to "Enhance reasoning ability" patented technology

The invention relates to an intelligent semantic analysis knowledge graph system for multi-source heterogeneous data. The system comprises a data preprocessing module used for preprocessing to-be-processed multi-source heterogeneous data to obtain format-standardized text data; the semantic analysis module is used for performing natural language processing on the text data to obtain words, and then performing vocabulary layer, syntactic layer and semantic layer analysis on the text data and the words to obtain structured data; the structured data comprises a triple consisting of entities, relationships and events and a network diagram of the entities and the relationships; the model structure mapping module is used for mapping the structured data into a predefined structure of a knowledge graph to obtain structured graph data; and the knowledge graph module analyzes entities and relationships in the graph data, performs reasoning to obtain potential relationships in the graph data by utilizing a logic rule and a machine learning algorithm, realizes cross-data-source and cross-document deep semantic association, and improves the efficiency of large-scale data processing and knowledge management.

The invention belongs to the technical field of mechanical arm grabbing, and discloses a visual sense-force sense-based mechanical arm instruction grabbing method, which comprises the following steps of: acquiring an RGB-D image of a scene, performing cutting processing on the RGB-D image through OpenCV, and extracting visual features of each object in the RGB-D image; a text command input by a user is coded through BERT, and language features are output; inputting the language features and the visual features into a CLIP model for fusion and alignment, and performing further alignment by adopting a Co-TRM structure to generate bimodal features; generating a point cloud image according to the depth map of the RGB-D image, and based on the point cloud image, combining visual features of an object, and generating a grabbing attitude through Graspnet; the problem that in the prior art, when grabbing is conducted based on vision and instructions, self-adaptive adjustment on grabbing force is not convenient is solved.

The invention relates to the technical field of natural language processing, in particular to a model distillation method and system based on a teacher model and situational reasoning. By using a multi-teacher learning framework, the student model can draw knowledge from a plurality of large teacher models, and diversified learning paths are provided for the student model; due to the introduction of a context awareness and example generation mechanism, the understanding and reasoning ability of a student model to a complex situation is enhanced by considering context information of input data and an input example; the reasoning chain strategy can ensure that the student model can generate an accurate and deep reasoning process, and the decision process of the teacher model is simulated; a parameter fine adjustment technology is adopted, only key parameters in the student model are adjusted, efficient migration of teacher model knowledge is achieved, and the number of parameters needing to be adjusted in the distillation process of the student model is reduced; the resource consumption in the model deployment and updating process is remarkably reduced, so that the trained model keeps the performance similar to that of the model.

The invention provides a method and a system for intelligently generating a complete dispatch report for financial unfavorable assets, and relates to the technical field of data investigation and arrangement, comprising Sp1, complete dispatch demand acquisition, Sp2, intelligent reasoning planning, Sp3, data acquisition and processing, Sp4, public opinion retrieval and analysis, Sp5, legal information extraction and verification, Sp6, data integration and report generation, and Sp7, report modification and output, and the method comprises the steps of Sp1, complete dispatch demand acquisition, Sp2, intelligent reasoning planning, Sp3, data acquisition and processing, Sp4, public opinion retrieval and analysis, Sp5, legal information extraction and verification, Sp6, data integration and report generation, and Sp7, report modification and output. By integrating a large model, an online retrieval tool, a data table tool and a key point extraction tool, automation and intellectualization of a complete dispatching process are realized. The system can automatically acquire, process and integrate data, greatly improves the efficiency of data acquisition, analysis and report generation, and reduces the probability of manual intervention and errors.

The invention discloses a knowledge graph-based knowledge question and answer method and device and a storage medium, and belongs to the technical field of natural language processing. The method comprises the following steps: processing an obtained question statement input by a user to obtain a subject entity and a target question embedding vector; mapping the subject entities to the obtained target knowledge graph and the embedded representation thereof, and generating candidate answer entities according to the subject entities and the proximity relation thereof; performing similarity calculation on the subject entity embedding vector, the candidate answer entity embedding vector and the target question embedding vector to obtain a target candidate answer entity set; extracting a relation path set from the subject entity to the target candidate answer entity from the target knowledge graph, and obtaining a target relation path embedding vector based on the relation path set; and performing similarity calculation on the target question embedding vector and the target relation path embedding vector to complete answer reasoning. According to the method, the accuracy and efficiency of knowledge questions and answers are improved by combining a multi-hop reasoning technology.

The invention relates to a power grid business intelligent auditing method based on AI enhancement, and the method comprises the following steps: S1, collecting data from a power grid SCADA system and a financial system, integrating historical and real-time data, and constructing an auditing knowledge graph; s2, constructing an instruction recognition model by using RoBERTa to analyze a natural language instruction of an auditor, and combining knowledge graph information with a language model to improve the understanding and reasoning ability of an instruction background; s3, constructing a multi-task deep learning framework to realize processing and fusion of multiple auditing tasks; s4, on the basis of a natural language instruction input by an auditor, extracting query targets, conditions and parameters involved in the instruction according to the instruction recognition model, obtaining corresponding auditing data in the database, and obtaining intention classification; and S5, calling a corresponding auditing model based on the intention classification result and the auditing data, and obtaining an analysis result. According to the invention, efficiency and precision in power grid business auditing work are effectively improved.

The invention discloses a visual detection optimization method and system based on data set classification. The method comprises the steps of introducing a scene classification label and carrying out standardization processing on scene features in a data set labeling and preprocessing stage, and carrying out scene classification multi-task learning in a model selection and training stage, so that a model dynamically adjusts feature learning in different scenes, and the multi-scene adaptive capacity is improved. And in the model evaluation and verification stage, scene-based hierarchical evaluation is introduced and corresponding optimization is carried out. In the model deployment and reasoning stage, a dynamic scene is introduced to adapt to reasoning, and the reasoning path of the model is dynamically adjusted to adapt to the characteristics of the scene, so that the reasoning accuracy is improved. In the later optimization, scene feature clustering and transfer learning are used, and the multi-scene performance of the model is optimized. According to the method, under the condition that the model is not jointly trained through multiple data sets, the multi-scene adaptive capacity of the visual detection model is improved through data set classification.

The invention provides an interpretable AI inference method and system for self-supervised neural symbol fusion, and the method comprises the steps: S1, a self-supervised learning module: extracting feature information of unlabeled data, and obtaining a feature vector representing symbol knowledge; s2, a neural symbol fusion module: constructing a deep learning framework by using an AI intelligent learning strategy, and training feature vectors in combination with a reward function to obtain structured knowledge; s3, an explainable reasoning module: constructing an explainable prediction model based on neural symbol fusion, and obtaining a solved reasoning result; in the process of explaining the reasoning result, the complex reasoning process is simplified and displayed. Therefore, not only are the advantages of the neural network in processing complex modes and data association kept, but also the interpretability and logical reasoning ability of symbol reasoning are introduced, so that the performance and interpretability of the whole system are improved. Therefore, the model can optimize feature representation and logical reasoning at the same time in the training process.

The invention discloses an intelligent agent method for complex question and answer reasoning of a figure knowledge graph based on a large model. The method comprises the steps that on the basis of multiple recall modes, corresponding prompt examples are dynamically matched according to questions to be answered, and the multiple recall modes comprise any one of embedded vector similarity matching, character string fuzzy matching and fusion matching; key entity recognition, relation screening and multi-candidate entity recursion search of the question to be answered are carried out by using a large language model in a few-sample prompt mode, and a candidate entity set of each sub-question of the question to be answered is obtained; and on the basis of the to-be-selected entity set, thinking and actions are iteratively generated in an intelligent agent mode by utilizing a large language model, and through an observation result obtained through interaction with the knowledge graph, a final answer is obtained after multiple rounds of interaction. According to the method, the reasoning ability of the large language model is enhanced with high-quality prompts.

The invention relates to the technical field of natural language processing, and discloses a multi-round dialogue optimization method based on a large language model.The method comprises the steps that before the large language model outputs answers according to current questions input by a user, the current questions input by the user are subjected to question preference alignment, answer preference alignment and preferred item preference alignment in a memory list; and in combination with a self-attention reward iteration function and a model updating rule based on a strategy gradient, an optimal matching historical question and an optimal matching historical answer of the current question are found, an optimal historical question and answer pair of the current question is formed, context window limitation of a large language model can be overcome, accurate historical memory can be provided for the large language model, and the user experience is improved. The context understanding ability and reasoning ability of the large language model in multiple rounds of dialogues are greatly improved, so that the large language model can accurately output coherent and high-quality answers, the quality of the multiple rounds of dialogues is improved, and the user experience is ensured.

The invention discloses a calibration data selection method for training uniform quantization after a deep learning model. The method comprises four steps of clustering centroid initialization, clustering centroid updating, training set data sorting and calibration set data selection. Firstly, random data conforming to uniform distribution are sent to a network for forward reasoning, and a plurality of clustering centroids are determined; then, reasoning is performed through training set data, a knowledge distance is calculated, a clustering centroid is updated, and an activation distribution space accurately reflecting training set data distribution is constructed; then, training set data are sent to the network for reasoning again, knowledge distances are calculated, summation is carried out, and scores of all data samples are obtained and sorted; and finally, according to the size of a specified calibration set, selecting training set data with relatively high scores as the calibration set from high to low. According to the method, through reasonably selecting calibration data, the effect of training uniform quantification after the deep learning model can be improved.

The invention provides a cultural relic fragment splicing, repairing and complementing system and method, and the system comprises a multi-modal large model understanding and vectorization processing module which is used for carrying out the understanding of cultural relic fragments through a multi-modal large model, and generating a multi-modal semantic vector through cross-modal vectorization processing; the knowledge graph construction and reasoning module is used for constructing a knowledge graph of the cultural relic fragments and carrying out reasoning and verification on splicing of the cultural relic fragments; and the fragment matching and splicing module is used for performing reasoning by utilizing the multi-modal semantic vector and the knowledge graph enhanced historical information to complete splicing and repairing of fragments. According to the technical scheme, by fusing the multi-modal data and generating the unified semantic vector, the form, texture and historical background of the fragments can be considered at the same time, and the splicing and repairing accuracy and integrity are remarkably improved.

The invention discloses a Zynq-based convolutional neural network hardware acceleration system, which is realized based on a Zynq heterogeneous chip and comprises a PL end module, a PS end module and a storage module. The PL end module comprises a convolution acceleration module, a full connection layer acceleration module and a Softmax activation module and is used for completing convolution calculation, full connection operation and activation function processing; the PS end module comprises an input data processing module and a post-processing module and is responsible for loading and preprocessing input data and sorting and screening detection results. The storage module is composed of an on-chip memory and an off-chip memory, wherein the on-chip memory stores an intermediate processing result of input data, a convolution kernel and intermediate feature map data; and the off-chip memory stores the original image data, the weight parameter of the full connection layer and the final detection result. All the modules are connected through a 64-bit AXI bus, the PS end module is responsible for coordinating data flow, and the PL end module executes efficient parallel computing. The system makes full use of the high-performance characteristic of the Zynq heterogeneous architecture, is suitable for an efficient reasoning acceleration scene of the convolutional neural network, and particularly has a wide application prospect in embedded AI equipment.

The invention provides a large language model heterogeneous knowledge reasoning method and device based on an atomic knowledge operator, and the method comprises the steps: building an atomic reasoning tree based on an original to-be-answered question; wherein the atomic inference tree comprises father nodes and leaf nodes, root nodes in the father nodes are original to-be-answered questions, non-root nodes in the father nodes are different levels of sub-questions obtained by decomposing the original to-be-answered questions, the leaf nodes are atomic questions obtained by decomposing the original to-be-answered questions, and each leaf node corresponds to an atomic knowledge operator; the atomic knowledge operator comprises a search operator, a relation operator and a filtering operator; for leaf nodes, reasoning atomic problems according to retrieval-knowledge on the basis of a large language model and an atomic knowledge operator; for a father node, based on a large language model, according to child node reasoning answers, brother node reasoning answers or retrieval-knowledge, reasoning child questions; reasoning the atomic reasoning tree from a leaf node to a root node from bottom to top to obtain a reasoning result of the original to-be-answered question; wherein the retrieval-knowledge is obtained by executing dynamic knowledge retrieval from various heterogeneous knowledge sources. According to the method, an original to-be-answered question is decomposed into atomic-level fine granularity, a leaf node in an atomic reasoning tree corresponds to an atomic knowledge operator, a more accurate reasoning result is achieved, meanwhile, knowledge is allowed to be dynamically retrieved from various heterogeneous knowledge sources at each sub-question and atomic question node, and the reasoning efficiency is improved. Different types of queries can be flexibly coped with, richer, more accurate and more complementary information is provided, and the overall reasoning ability of the algorithm is enhanced.

The invention relates to the technical field of road traffic marking detection, and provides a road traffic marking detection method based on efficient visual transformation. According to the method, road traffic marking data sets are combined, screened, expanded and labeled; constructing a road traffic marking detection model composed of a feature extraction network, a multi-scale feature map enhancement network and a detection head; inputting into a feature extraction network for feature extraction; inputting into a multi-scale feature map enhancement network for feature fusion; and inputting into a detection head for detection and classification to obtain a final detection result. According to the scheme of the invention, the feature extraction network utilizes an efficient attention mechanism to improve the global context information capture capability and improve the complex road traffic marking detection precision; the multi-scale feature map enhancement network adopts a bottom-up path structure to enhance the capturing capability of small target features, and the recognition precision and robustness under a complex background are improved.

The invention relates to an insulator online spraying method and system based on an unmanned aerial vehicle, and the method comprises the following steps: controlling the unmanned aerial vehicle with a spray gun to be near an insulator to be sprayed, and enabling the unmanned aerial vehicle to collect the video data of the insulator to be sprayed in real time; constructing an insulator target detection model, and inputting the video data of the insulator to be sprayed into the insulator target detection model for insulator detection to obtain an insulator detection result; and an insulator target tracking model is constructed, a target insulator needing to be sprayed is selected from insulator detection results, the unmanned aerial vehicle is controlled to continuously track the target insulator through the insulator target tracking model, the unmanned aerial vehicle is kept at the optimal spraying position, and the insulator spraying task is completed.

The invention provides a binary network training method and device based on gradient quantization, and the method comprises the steps: carrying out the quantization of the weight and an input activation value of a full-precision model in a full-quantization training process of a full-precision model with a sample image as a training sample and a class label of the sample image as a training label, and obtaining a binary network training result; obtaining first quantized data with a target bit width; quantizing the activation value gradient to obtain second quantized data having the target bit width under the condition that the first quantized data having the target bit width is subjected to back propagation and the activation value gradient is output; and updating the activation value gradient and the weight gradient according to the first quantized data and the second quantized data to obtain a trained binary network. According to the method, binarization gradient training is achieved, the training power consumption of the model and the requirements for memory and computing power are effectively reduced, and then the training efficiency and reasoning performance of the classification model in a complex scene are improved.

The invention belongs to the technical field of circuit fault prediction, and discloses a PCB analog circuit component fault prediction method based on an improved large language model. Excitation is applied to the PCB analog circuit, and output voltages of a to-be-predicted component in the PCB analog circuit under different frequencies are extracted; performing data preprocessing according to a Pearson's correlation coefficient method; the improved large language model takes a general large language model as a base model, a linear layer, a cross attention layer and a full connection layer are added, and feature vectors after cross-modal alignment are obtained; and improving cue words of the general large language model, and adding and connecting the cue word vectorization of the general large language model and the feature vector after cross-modal alignment to serve as an input feature vector of the large model. According to the method provided by the invention, the loss of the features is reduced during feature extraction, so that the prediction result has generalization and reliability.

The invention discloses a cognitive diagnosis method based on hierarchical constraint perception. The method comprises the following steps: S1, inputting student information to be processed; s2, performing hierarchical division on historical scores in the student information; s3, performing convolution feature extraction on the student information of the same level, and performing attention network aggregation to obtain knowledge features in the level; s4, aggregating student information of different levels, and forming a level knowledge proficiency degree vector by combining knowledge features in the levels and among the levels; s5, combining the hierarchical knowledge proficiency vector with the personalized knowledge proficiency vector to form a comprehensive proficiency vector, inputting the comprehensive proficiency vector into the cognitive diagnosis model, and predicting a practice question result; according to the method, cognitive diagnosis is carried out according to student information, an attention mechanism of internal and external knowledge states is introduced, the diagnosis precision is improved, the interpretability and reasoning ability of diagnosis are enhanced, and the result is more accurate and can be easily applied to teaching practice.

The invention provides a knowledge graph question answering method and device based on an agent debate mechanism, and the method comprises the steps: determining a target semantic entity in a target question statement, determining a first semantic relation corresponding to the target semantic entity from a target knowledge graph database, and determining a knowledge graph question answering result based on the first semantic relation. A multi-agent debate mechanism is adopted to convert the target question statement into a first single-hop query statement, a second semantic relation matched with a first semantic entity in the first single-hop query statement is determined from the target knowledge graph database, and based on the target question statement and the second semantic relation, the first single-hop query statement is obtained; and generating the target answer statement corresponding to the target question statement, so that the reasoning extensibility of the target answer statement can be improved to a certain extent, and the matching degree between the target answer statement and the target question statement can be improved, thereby improving the reasoning ability of a large language model to a certain extent.

The present disclosure relates to a model deployment method and device, electronic equipment, storage medium and product. The method comprises: obtaining an image saliency segmentation model; the image saliency segmentation model is used for saliency segmentation of an input image, the image saliency segmentation model comprises a preset activation function, the output value of the preset activation function is in a preset range, the maximum value of the preset range is a first preset value; converting the model format of the image saliency segmentation model into a target model format supported by a target chip; deploying the image saliency segmentation model in the target model format on the target chip. The present disclosure can limit the size of the output value of the preset activation function, thereby improving the accuracy of the image saliency segmentation model. In addition, by converting the model format of the image saliency segmentation model into the target model format, the image saliency segmentation model can be efficiently run on the target chip, thereby improving the inference efficiency and inference performance of the image saliency segmentation model.

The invention relates to an artificial intelligence technology, and provides a to-be-financed invoice verification method, apparatus and device, and a medium, which perform table restoration on contract information based on a large language model, and perform subsequent processing after the contract information is restored into a table, so that the reasoning effect of the large language model can be improved. The target cue word is generated according to each piece of commodity information in the target invoice to be financed and the contract data table, and the target cue word is input to the large language model for reasoning, so that the large language model can be utilized to perform automatic matching on the invoice to be financed and the contract information, manual intervention is not needed, the matching efficiency is improved, and the error rate is low; according to the method, the reasoning result is post-processed to remove invalid information, so that the valid information is utilized to verify the to-be-financing invoice, and the calculation power is saved, so that the verification efficiency is improved.

The present invention discloses a camera intelligent operation and maintenance method and system based on deep learning, including: S1: collecting camera historical fault data and preprocessing to obtain preprocessed camera historical fault data; S2: according to the preprocessed camera historical fault data, counting the number of occurrences of each historical fault type and historical fault cause, and extracting the historical fault type feature vector and the historical fault cause feature vector; S3: generating camera operation and maintenance text and performing data enhancement, extracting the historical fault description feature vector, the historical maintenance measure description feature vector, the first relationship feature vector and the second relationship feature vector; S4: collecting camera current fault data, and calculating the current fault type, the first fault cause and the second fault cause; S5: generating maintenance suggestions and work orders, and selecting maintenance personnel to dispatch work orders. The present invention can solve the problem that the fault location of the traditional camera operation and maintenance method depends on manual experience.

This application discloses a data-aware intelligent analysis method, system, storage medium, and computer device. The method includes: receiving a target problem and initializing state information, wherein the state information includes key sub-conclusions, a reasoning state logic diagram, and key data descriptions; a language model-based agent iteratively performs the following operations: obtaining input information for the current step; updating the state information of the current step based on the action and environmental feedback information of the previous step, and making decisions based on the state information of the current step to obtain the action and tools to be invoked for the current step; invoking the tools to execute the action of the current step, and obtaining environmental feedback information for the current step after the action is completed; checking whether the target problem has been analyzed; if not, proceeding to the next loop; if completed, generating analysis results. This method can improve the reasoning ability, response consistency, and interpretability of deep reasoning models.

The present application discloses a data processing method, device, equipment, medium and program product in the field of computer technology. In the initialization stage of model reasoning, the present application creates a first cache area and a second cache area; if a model reasoning request is received, the context information of the preprocessing stage of the model reasoning request is stored in the first cache area; based on the context information in the first cache area, each decoding operation in the decoding stage of the model reasoning request is executed; wherein, the context information of the previous decoding operation is stored in the second cache area, and the next decoding operation of the previous decoding operation is executed based on the context information in the second cache area. This scheme is based on the principle of exchanging space for reasoning efficiency, and utilizes two cache areas to store the context information of the preprocessing stage and the decoding stage respectively, which improves the model reasoning performance and efficiency, and does not require changing the model structure, so that the model can be quickly put into use.

The invention relates to a knowledge-enhanced big language model reasoning method based on experience knowledge. Comprising the following steps: constructing a knowledge base; obtaining reasoning problems and corresponding problem solving experience; extracting the problem solving experience into structured knowledge entries; a reasoning structure of the problem is extracted in the structure abstraction process, and experience is refined into a reusable solution, judgment and correction mode in the mode generalization process; retrieval of knowledge: for a new problem, related items are retrieved by using a reasoning structure of the new problem; the diversity reordering step then selects a final set of entries, providing a strategically diverse context for reasoning. The problem-based core reasoning structure adopts a diversity perception retrieval mechanism, so that the LLM reasoning accuracy is improved, and the LLM can learn and evolve from practice under the condition that re-training is not needed.

The invention relates to a medical data processing method and system, computer equipment and a storage medium, and the method comprises the steps: obtaining input information and medical data information corresponding to the input information; based on a preset thinking chain template, processing the input information and the medical data information to obtain an initial output result; further, according to the initial output result, the thinking chain template is adjusted until the input information and the medical data information are processed based on the adjusted thinking chain template, and an output result meeting the rationality requirement is obtained; the rationality requirement is used for indicating that the output result needs to accord with preset medical knowledge information. Through the data processing method and device, the problem that the output result is inaccurate due to the lack of understanding of the deep meaning of the data is solved, the deep meaning of the data is understood in the data processing process, and the accuracy of the output result is improved.

The invention relates to a test case generation model training method and a test case generation method, and the method comprises the steps: carrying out the first-stage reinforcement learning training of a preset model through employing a first training data set, and obtaining a first target model; wherein the first training data set comprises a first function specification sample of a plurality of function points and a first standard test case, and the preset model is used for generating a first prediction test case according to the first function specification sample; in the first-stage reinforcement learning training process, a target reward value is adopted to guide a prediction model to optimize a generation strategy, and the target reward value is determined according to a correctness reward rule, a format reward rule and a coverage rate reward rule according to a first function specification sample, a first standard test case and a first prediction test case. On the basis, the test case generation model obtained through training can reverse multiple test scenes such as a test scene, a test scene corresponding to a market problem and a test scene corresponding to a boundary condition.

The invention discloses a public security event identification method based on small sample learning and a knowledge graph, which comprises the following steps: generating a virtual scene similar to a real event scene through the combination of small sample data and the knowledge graph, and dynamically activating nodes related to the event through a knowledge activation threshold value, freezing and storing the unrelated nodes, and dynamically releasing frozen nodes; identifying dynamic evolution of the event by monitoring node relation fluctuation in the knowledge graph; a multidimensional event analysis result is generated through mutual reinforcement of a graph fission mechanism and an interactive training mechanism; generating a risk path identification result through the event aggregation trigger point; the scheme has the characteristics of realizing efficient learning of small samples, enhancing semantic understanding ability, dynamically adapting to diversified scenes, improving event identification flexibility and constructing a global risk assessment framework.

This application proposes a data merging method, apparatus, device, and computing medium, comprising: obtaining the micro-rank matrix corresponding to the merging matrix of each model to be merged, and a first matrix and a second matrix corresponding to each micro-rank matrix, wherein the matrix parameters of each first matrix conform to a uniform distribution, and the matrix parameters of each second matrix conform to a Gaussian distribution; performing pruning processing on each first matrix and each second matrix to obtain corresponding first pruned matrices and second pruned matrices; obtaining a singular value scaling matrix for each first matrix based on each first matrix and the corresponding first pruned matrix; and obtaining the merging matrix of the merged model based on each singular value scaling function, each first pruned matrix, and each second pruned matrix. The embodiments of this application process the pruned micro-rank matrix using the singular value scaling matrix to adjust the singular values ​​of the merging matrix, thereby avoiding distribution shift during inference of the merging model and improving the inference performance of the merging model.