101 results about "Graph generator" patented technology

The invention proposes a hard negative sampling comparison recommendation method based on a graph generation diffusion model, and the method comprises the following steps: coding a user-article interaction graph through a graph convolutional neural network, obtaining a node embedded vector, and carrying out the inner product of the node embedded vector, and obtaining a recommendation probability; the model parameters of the graph convolutional neural network are obtained through the following training steps: S1, generating a comparison view 1 and a comparison view 2 from the node embedding vector through an inline diffusion model view generator and a parameter topological noise reduction view generator; the inline diffusion model view generator comprises an inline diffusion model, and the parameter topological noise reduction view generator comprises a parameter topological noise reduction network; and S2, carrying out comparative learning on the comparative view 1 and the comparative view 2. According to the method, a signal alignment strategy is combined, a double-view generator comparison learning framework is built, and the problems that the model generation capacity is insufficient and the influence of data noise is large are solved, so that the performance of a recommendation system is improved.

The invention relates to an automatic code activity graph generation system and method, and the system comprises a plurality of modules. The source code analysis module extracts source codes from the code file and generates an abstract syntax tree corresponding to the source codes. The model conversion engine is based on query view conversion specification operation, analyzes a method calling relation and an execution sequence in the abstract syntax tree by utilizing a group of analysis rules, and converts the source code into a programming language abstract model. And the model warehouse is used for storing the abstract model, identifying the language type through the extension name of the source code file, and matching the corresponding programming language model. The activity graph generator then generates an activity graph according to the abstract model, including an activity graph reflecting code logic and an activity graph generated based on code annotations. The layout module determines the position, the size and the shape of each node in the activity graph through an analysis method calling relation and an execution path, and generates a connecting line between the nodes. The system can visually display the code process and improve the efficiency of developers.

Methods, systems, and apparatus for automatic detection of software bugs are disclosed. An example apparatus includes a comparator to compare reference code to input code to detect a source code error in the input code; a graph generator to generate a graphical representation of the reference code or the input code, the graphical representation to identify non-overlapping code regions; and a root cause determiner to determine a root cause of the source code error in the input code, the root cause based on the non-overlapping code regions.

One or more information maps are obtained by an agricultural work machine. The one or more information maps map one or more agricultural characteristic values at different geographic locations of a field. An in-situ sensor on the agricultural work machine senses an agricultural characteristic as the agricultural work machine moves through the field. A predictive map generator generates a predictive map that predicts a predictive agricultural characteristic at different locations in the field based on a relationship between the values in the one or more information maps and the agricultural characteristic sensed by the in-situ sensor. The predictive map can be output and used in automated machine control.

The invention relates to a traffic flow prediction method and system based on a dynamic perception expert network, and the method comprises the steps: obtaining traffic observation data, constructing a traffic network diagram, carrying out the feature embedding, and generating an initial spatial-temporal feature vector; inputting the initial spatial-temporal feature vector into a double-path time encoder, and respectively extracting a personalized time sequence feature vector and a time sequence dynamic feature vector through parallel channel independent paths and channel mixed paths; carrying out vector fusion through a gating mechanism to obtain a time context feature vector, inputting the time context feature vector into a multi-scale hybrid expert model, activating a plurality of most relevant time scale experts, and generating a corresponding routing weight; and generating a spatial dependency graph through a scale condition dynamic graph generator, performing graph convolution to extract a plurality of spatio-temporal feature vectors, performing weighted aggregation, sending the spatio-temporal feature vectors into a prediction model, and generating a traffic prediction value. Compared with the prior art, the model constructed by the method is relatively high in prediction precision and relatively high in robustness in a complex traffic scene.

A system and method employed by an infrastructure-as-code (IaC) tool that is configured for defining IaC configurations comprising containers of interconnected components of a cloud computing platform. A system includes a graph generator configured to generate a graph of dependencies between the interconnected components of the cloud computing platform based at least in part on user-defined input-output relationships between the interconnected components. The IaC tool is configured to use the graph of dependencies to automatically determine a sequence of the provisioning or deployment of the interconnected components of the cloud computing platform. The system further includes a remote procedure call interface configured to enable communications between the IaC tool and the cloud computing platform, the remote procedure call interface further configured to enable the provisioning or deployment of the interconnected components of the cloud computing platform, in the determined sequence.

The invention discloses a lateral movement detection method and device for deploying a time sequence perception graph neural network based on edge optimization, and the method comprises the steps: S1, capturing a host authentication log in real time through a log collection agent deployed on edge equipment, carrying out the event filtering and key field extraction through a rule matching engine, and forming a standardized entity data set; s2, designing an authentication graph structure by adopting an ontology graph modeling language, and definitely defining node types of a user, a host, a process and the like and a time sequence edge type containing a timestamp and an authentication type attribute; s3, mapping the entities into heterogeneous nodes based on a dynamic graph modeling technology, and constructing a complete transverse movement heterogeneous authentication graph through weighted time sequence edges; s4, importing the constructed authentication graph into a lightweight graph database Neo4j Embedded, and realizing dynamic topology maintenance by adopting an incremental PageRank algorithm; s5, defining each identity verification event as a multi-tuple structure, designing a time perception sub-graph generator, performing optimized and accelerated local neighborhood sampling by setting a time window, and generating a time sequence sub-graph sample required by training; s6, constructing a multi-scale attention coding framework, and integrating a local graph attention network and a global Transform encoder to realize multi-level feature fusion; and S7, aiming at the resource limitation of the edge equipment, implementing model cutting, a hierarchical quantification strategy and a streaming sub-graph loading mechanism, and ensuring the efficient operation of the system in a resource-limited environment.

The invention belongs to the technical field of metabonomics large model construction, and relates to a metabonomics large model construction method based on multi-dimensional database fusion, a global unique ID is allocated for metabolites, and multi-source heterogeneous data is uniformly coded into a multi-modal feature matrix, so that information loss caused by data isomerism in a traditional method is avoided; a dynamic graph generator is adopted, a multi-modal feature matrix and a prior path adjacency matrix are combined, dynamic edge weights are generated through similarity and learnable parameters to construct a metabolic relationship graph, a multi-modal graph attention mechanism is utilized to distribute adaptive weights for different modal features, graph convolution message passing is executed, and the metabolic relationship graph is obtained. Generating a node embedding matrix fused with multi-modal information and an optimized dynamic metabolism relation graph; besides, through a meta-learning method, each database is split into meta-tasks, a dynamic graph is used for training a task exclusive classifier in an inner cycle, a dynamic graph generator and meta-model parameters are jointly optimized through query set loss in an outer cycle, and the cross-database generalization ability of the model is remarkably improved.

The invention belongs to the technical field of artificial intelligence and computer vision, particularly relates to a visual large model emergency management and control system for panoramic situation awareness, and aims to solve the problems of weak multi-source visual data fusion, low risk deduction precision, lagging command decision and the like in urban public safety. The system comprises a global visual perception layer, a space-time alignment engine, a visual large model reasoning center, a dynamic situation map generator, a multi-granularity emergency decision module and an instruction distribution feedback unit. By constructing a high-precision space-time grid and a visual large model for joint reasoning, conversion from pixels to semantics is realized, and a time sequence evolution knowledge graph is generated and dynamically updated; and in combination with a hierarchical response mechanism and a cloud edge collaborative architecture, automatic disposal and man-machine collaborative decision making are supported, and the emergency response efficiency and the system robustness are improved.

The embodiment of the invention provides a face forgery detection method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a pre-training detection model and a to-be-detected face image, and inputting the to-be-detected face image into the pre-training detection model, so that a face key point detector obtains a face key point graph according to the input to-be-detected face image; a semantic segmentation graph generator generates a semantic segmentation graph according to the face key point graph; the feature extractor performs basic feature extraction according to the to-be-detected face image to obtain basic features; the global feature extraction branch performs global feature extraction according to the basic features to obtain a global feature map; the feature learning branch obtains a region-level forgery confidence map according to the basic features and the semantic segmentation map; the feature fusion device carries out fusion processing according to the global feature map and the region-level forged confidence map to obtain classification features; and the classifier obtains a face forgery detection result according to the classification features. According to the invention, the accuracy of face forgery detection is improved.

The invention discloses a palm image recognition method and system based on fusion of a convolutional neural network and an image neural network, and belongs to the technical field of crossing of biological feature recognition and artificial intelligence. The problems that in the prior art, local texture features and a global topological relation are separated, dynamic scene adaptability is poor, and multi-modal features are not fully fused are solved. The method comprises the following steps: 1) designing a dual-path hybrid network architecture (DPHGCN), extracting multi-scale local texture features through a lightweight CNN, constructing an adaptive graph structure by using a dynamic graph generator (DGG), and modeling a topological relation between key points in combination with a graph attention network; 2) developing a cross-modal feature alignment module (GIPL), and realizing dimension mapping and semantic alignment of CNN feature graph and graph node embedding; and 3) providing a layered attention fusion mechanism, and combining space attention, topological attention and cross-modal attention to dynamically weight and fuse heterogeneous features. According to the method, 98.7% of recognition accuracy is achieved on a public data set, the reasoning speed reaches 32 ms / frame, 76.2% of robustness is still kept in a 50% shielding scene, and the method can be widely applied to the fields of financial payment identity authentication, medical palmprint pathological analysis, real-time security verification and the like.

The invention relates to the technical field of data processing, in particular to a graph neural network-based large-scale knowledge graph construction system, which comprises a node increment receiver, a historical embedded memory, a dynamic neighbor sampler, a local sub-graph generator, a quantum state aggregator, a spatio-temporal topology checker, an embedded propagation controller and a graph updater. A time sequence control closed loop adopts a time window to process new nodes in batches, a historical embedding memory aligns version snapshots and multiplexes historical embedding, and a dynamic neighbor sampler optimizes sampling resources according to time sequence density; a precision adjustment closed loop triggers relation weight correction and quantum line parameter updating, and an embedded propagation controller freezes an inactive node gradient to focus key changes; resource optimization chain adjustment reduces neighborhood complexity, time attenuation coefficient compresses adjacency matrix scale, and historical embedded memory pointer call eliminates vector replication overhead. According to the method, computing resource consumption in the dynamic updating process is remarkably reduced, and knowledge graph representation quality is maintained.

The invention provides a multi-source heterogeneous data fusion processing method and system based on deep learning, and the method comprises the steps: obtaining a multi-source sensor data stream of an Internet of Things terminal, extracting a modal description vector of the multi-source sensor data stream, generating a meta-feature, and independently packaging each data source into a data block; processing paths are dynamically distributed according to element feature vectors in the data blocks, and parallel feature extraction branches are input into the data blocks of different modalities to perform modal feature extraction; inputting the extracted different modal features into a cross-modal semantic graph generator, extracting semantic similarity measurement values among the features through comparative learning, and obtaining a dynamically updated cross-modal knowledge graph; performing multilayer graph convolution calculation on the modal features based on a cross-modal knowledge graph to generate a joint feature tensor; according to the method, the problems of low fusion efficiency, poor adaptability, poor consistency and the like caused by multi-source, heterogeneous and strong time-varying characteristics of sensor data in an existing Internet of Things system can be effectively solved.

Systems and methods are described for training a large language model to operate as a completeness graph generator to automatically generate completeness graphs in response to queries based on instructions including forms, rules, and regulations. A dataset is obtained that includes instructions and associated ground truth completeness graphs, previously generated manually by domain experts. An active large language model is trained configured to produce a generated completeness graph in response to a query that is evaluated with a reward model based on validity of the generated completeness graph and semantic similarity of the generated completeness graph and the associated ground truth completeness graph. The active large language model is re-trained based at least partially on the reward.

The embodiment of the invention provides a layout congestion prediction method and a layout congestion prediction device based on Regression VisionTransform. The method comprises the following steps: acquiring a target data set containing real data; training a GAN model through the target data set to obtain a feature graph generator and a tag graph generator; generating a source data set containing virtual data by adopting the feature graph generator and the tag graph generator; pre-training of an R-ViT model is carried out through the source data set, transfer learning is carried out on the pre-trained R-ViT model, and a fine-tuned R-ViT model is obtained; and inputting the target data set into a fine-tuned R-ViT model to obtain predicted wiring information, realizing congestion prediction, building a prediction model by using a Transform architecture in combination with a convolution module, and training a prediction model with equivalent performance and accurate prediction in a relatively short training time through GAN data enhancement, transfer learning and an early stop method.

A global view generator for low-altitude economic unmanned aerial vehicle monitoring comprises a grouping strategy setting module, an unmanned aerial vehicle clustering and grouping module, a local view generation module and a view aggregation module which are connected in sequence, flexible definition and real-time adjustment of grouping logic are achieved through modular strategy design and dynamic expansion interfaces, and the overall view generator is high in practicability and easy to popularize. And the scene adaptation capability of the system is obviously improved. The unmanned aerial vehicle clustering and grouping module adopts a dynamic graph clustering model and an event triggering optimization mechanism, only performs local adjustment on related parts of incremental data, and ensures grouping efficiency and real-time response capability. And the local view generation module generates grouping boundaries by using a dynamic three-dimensional alpha-shape algorithm, constructs a local view by combining state feature mapping, and provides real-time data support for grouping and monitoring. The view aggregation module effectively solves the problem of boundary conflict between groups through a global optimization technology, ensures the consistency of global views in spatial relationship, grouping range and state characteristics, and provides reliable support for unmanned aerial vehicle monitoring and task coordination in a complex task scene.

The present invention discloses a multi-modal and multi-scale feature branch fusion classification method for keratitis anterior chamber images. The method comprises: multi-modal input and pre-processing of the input data; the image data is subjected to a multi-scale self-attention fusion feature map generator to obtain fusion feature maps of two image modal data at different scales; the extracted feature maps are input to different branches for processing, and the feature maps obtained respectively are adaptively pooled to form respective feature vectors; finally, the two image data feature vectors and the vector of text data are weighted summed, and input into a fully connected layer and a ReLU activation function for calculation, and the final vector is obtained for result prediction. The present invention makes full use of medical imaging data and text data, mines and utilizes the correlation information between modalities, enhances the feature expression capability, improves the ability to recognize and judge diverse images, and comprehensively and accurately improves the classification performance.

A method for acquiring skills through imitation learning by employing a meta imitation learning framework with structured skill discovery (MILD) is presented. The method includes learning behaviors or tasks, by an agent, from demonstrations: by learning to decompose the demonstrations into segments, via a segmentation component, the segments corresponding to skills that are transferrable across different tasks, learning relationships between the skills that are transferrable across the different tasks, employing, via a graph generator, a graph neural network for learning implicit structures of the skills from the demonstrations to define structured skills, and generating policies from the structured skills to allow the agent to acquire the structured skills for application to one or more target tasks.

A communication analysis system includes: an information receiver that receives information indicating analysis target communication performed by a monitoring target; an information obtainer that obtains past communication information indicating communication by the monitoring target; a WL determiner that determines, using the obtained analysis target communication and the whitelist, that non-WL communication has been established in the analysis target communication; a similar terminal extractor that extracts one or more terminals similar to destination and source terminals included in a non-WL communication link determined; a primary similar communication link extractor that extracts a past communication link similar to the non-WL communication link as a primary similar communication link from the obtained past communication information using the extracted similar terminals; and a NW graph creator that creates a NW graph for analysis as graph information for analyzing the non-whitelist communication using the extracted primary similar communication link and the obtained past communication information.

The present invention discloses a method for constructing a traffic flow prediction model and a traffic flow prediction method, comprising the following steps: S100, describing the relationship between entities in a network with a graph structure, and constructing a feature matrix between nodes at the same time; S200, constructing a traffic flow prediction model based on the graph structure and the feature matrix using a convolutional neural network framework combined with a spatiotemporal attention mechanism. The present invention establishes a framework consisting of three core components: a graph network generation layer, a graph convolution module, and a multi-scale temporal convolution module. A dynamic adjacency matrix is ​​generated by an adaptive dynamic graph generator and a heterogeneous adjacency relationship attention mechanism, and multiple types of adjacency matrices are aggregated, thereby better utilizing the traffic characteristics in traffic nodes, and capturing and extracting the spatial characteristics implicit in traffic characteristics, overcoming the limitations of traditional methods, and effectively solving the problems of the traditional traffic flow prediction method in the intelligent transportation system, such as the low prediction accuracy and the inability to reflect the impact of the traffic network.

A system and method for project management and scheduling across a communication network are provided. Namely, the system generates tasks within received project parameters and generates a timeline for the project and for each of the tasks. The system may assign each of the tasks to a vendor user having at least one qualification and transmit any of the project parameters, including the timeline, associated with the task to the vendor user. A view generator provides a project view that indicates a link between the vendor user, the task, and the timeline. Specific tasks in the project may be dependent on another of the tasks in the project, and the system may order the tasks in the timeline according to their dependencies.

A travel map creating apparatus includes a position sensor obtaining the positional relationship of an adjacent object relative to the apparatus; a floor map creator creating a floor map based on the positional relationship; a self-position calculator calculating the self-position on the floor map; a marker identifier identifying a marker; a marker position calculator calculating the relative position of the marker to the apparatus; a mode switcher switching the mode of the apparatus between a floor map creation mode and a marker identification mode; a restricted access information generator, based on the floor map, the self-position, and the relative position of the marker, defining a boundary of a restricted area to which the entry of an autonomous mobile robot is prohibited and generating restricted access information including the boundary; and a travel map creator creating a travel map including the defined restricted area based on the restricted access information.

An apparatus for diagnosing an autism spectrum disorder (ASD) based on a graph neural network includes a preprocessor configured to acquire brain image data, designate a region of interest (ROI) of the brain, and generate preprocessed data for neural network input, a spatial feature extractor configured to extract spatial features from the preprocessed data, a temporal feature extractor configured to analyze changes in brain activity over time and extract attention-based temporal features, a spatial and temporal convergence feature unit configured to analyze spatiotemporal correlations by combining the spatial features and the temporal features, a graph generator configured to convert connectivity between regions of interest into a graph structure and implement the same as nodes and edges, and a graph classifier configured to analyze a spatiotemporal pattern of the connectivity through the graph structure and classify whether or not there is an ASD.

The present application relates to the technical field of data processing, and more particularly to a large-scale knowledge graph construction system based on a graph neural network, comprising a node incremental receiver, a historical embedding memory, a dynamic neighbor sampler, a local subgraph generator, a quantum state aggregator, a space-time topology verifier, an embedding propagation controller and a graph updater. The time sequence control closed loop adopts time window batch processing of new nodes, the historical embedding memory aligns version snapshots and reuses historical embeddings, and the dynamic neighbor sampler optimizes sampling resources according to time sequence density; the precision adjustment closed loop triggers relationship weight correction and quantum circuit parameter updating, the embedding propagation controller freezes the gradient of non-activated nodes to focus on key changes; the resource optimization chain adjusts to reduce neighborhood complexity, the time decay coefficient compresses the size of the adjacency matrix, and the historical embedding memory pointer call eliminates vector copy overhead. The present application significantly reduces the computational resource consumption of the dynamic updating process and maintains the quality of the knowledge graph representation.

The present disclosure provides an apparatus and method of guided neural network model for image processing. An apparatus may comprise a guidance map generator, a synthesis network and an accelerator. The guidance map generator may receive a first image as a content image and a second image as a style image, and generate a first plurality of guidance maps and a second plurality of guidance maps, respectively from the first image and the second image. The synthesis network may synthesize the first plurality of guidance maps and the second plurality of guidance maps to determine guidance information. The accelerator may generate an output image by applying the style of the second image to the first image based on the guidance information.

The apparatus comprises a receiver 201 that receives a first depth map for a first video frame of the plurality of video frames. A divider 203 divides the first depth map into two-dimensional geometric primitives / polygons larger than two pixels. A map generator 207 generates a vertex position map indicating vertex positions of the geometric primitives in the first depth map. A depth map generator 205 generates a second depth map including the depth values ​​of the first depth map at the vertex positions, and a bitstream generator 209 generates a video bitstream to include the second depth map and the vertex position map for the first video frame. The division depends on the depth values ​​of the depth map. A complementary apparatus renders an image using a mesh generated from the bitstream by determining original vertex positions based on the received vertex position map and projecting them into a three-dimensional space.

One or more information maps are obtained by an agricultural system. The one or more information maps map one or more characteristic values at different geographic locations in a worksite. An in-situ sensor detects a mass flow value as a mobile machine operates at the worksite. A predictive map generator generates a predictive map that maps predictive mass flow values or predictive yield values at different geographic locations in the worksite based on a relationship between the values in the one or more information maps and the mass flow value detected by the in-situ sensor or the yield value based on the detected mass flow value. The predictive map can be output and used in automated machine control.

A system and method employed by an infrastructure-as-code (IaC) tool that is configured for defining IaC configurations comprising containers of interconnected components of a cloud computing platform. A system includes a graph generator configured to generate a graph of dependencies between the interconnected components of the cloud computing platform based at least in part on user-defined input-output relationships between the interconnected components. The IaC tool is configured to use the graph of dependencies to automatically determine a sequence of the provisioning or deployment of the interconnected components of the cloud computing platform. The system further includes a remote procedure call interface configured to enable communications between the IaC tool and the cloud computing platform, the remote procedure call interface further configured to enable the provisioning or deployment of the interconnected components of the cloud computing platform, in the determined sequence.

A system and method employed by an infrastructure-as-code (IaC) tool that is configured for defining IaC configurations comprising containers of interconnected components of a cloud computing platform. A system includes a graph generator configured to generate a graph of dependencies between the interconnected components of the cloud computing platform based at least in part on user-defined input-output relationships between the interconnected components. The IaC tool is configured to use the graph of dependencies to automatically determine a sequence of the provisioning or deployment of the interconnected components of the cloud computing platform. The system further includes a remote procedure call interface configured to enable communications between the IaC tool and the cloud computing platform, the remote procedure call interface further configured to enable the provisioning or deployment of the interconnected components of the cloud computing platform, in the determined sequence.

The invention relates to a composition scheme generation method, electronic equipment and a storage medium. The method comprises the following steps: acquiring an original scene image and composition preference information; determining an image feature tag set based on the original scene image, and determining a preference feature tag set based on the composition preference information; determining at least one style fine tuning model based on the image feature tag set and the preference feature tag set; loading the at least one style fine tuning model to a preset composition generator to obtain a composition generation model; and generating a composition scheme based on the original scene image, the preference feature tag set and the composition generation model. According to the scheme provided by the invention, the scene graph can be fully analyzed to generate the composition scheme meeting the photographing requirements of the user, and the aesthetic level of the composition scheme is improved, so that the photo photographing quality and the user experience are improved.