48 results about "Dual graph" patented technology

The present invention smoothes a spherical graph signal composed of spherical signal points associated with graph vertices of a graph producing a smoothed spherical graph signal composed of smoothed spherical signal points. Each smoothed spherical signal point is computed by multiplying a vertex rotation matrix by the corresponding spherical signal point. The vertex rotation matrix is computed as a weighted average of neighbor rotation matrices using a local parameterization of the group of rotations. The present invention also filters anisotropically a graph signal composed signal points associated with graph vertices of a graph producing a filtered graph signal composed of filtered signal points. Each filtered signal point is computed as a weighted average of signal points corresponding to the corresponding graph vertices and neighbor graph vertices with neighbor weight matrices. The present invention also denoises the vertex positions of a polygon mesh without tangential drift. The face normals are smoothed on the dual graph of the polygon mesh. The smoothed face normals are used to construct neighbor weight matrices on the primal graph of the polygon mesh. The vertex positions are anisotropically filtered on the primal graph of the polygon mesh. The present invention also filters the vertex positions and face normals of a polygon mesh with interpolatory vertex positions and face normal constraints.

Techniques are disclosed for determining if the decomposition of layout design data is feasible, and for optimizing the segmentation of polygons in decomposable layout design data. Layout design data is analyzed to identify the edges of polygons that should be imaged by separate lithographic masks. In addition, proposed cut paths are generated to cut the polygons in the layout design data into a plurality of polygon segments. Once the separated edges and cut paths have been selected, a conflict graph is constructed that reflects these relationships. Next, a dual of the conflict graph is constructed. This dual graph will have a corresponding separation dual graph edge for each separated polygon edge pair in the layout design data. The dual graph also will have a corresponding cut path dual graph edge for each proposed cut path generated for the layout design data. After the dual graph has been constructed, it is analyzed to determine which of the proposed cut paths should be kept and which should be discarded. The layout design data is then modified to include the cut paths corresponding to the selected cut path dual graph edges. Alternately or additionally, separate sets of layout design data may be decomposed from original layout design data using the selected cut paths.

Techniques are disclosed for performing topologically planar routing of System in Packages (SiPs). A routing graph can be represented by a particle-insertion-based constraint Delaunay triangulation (PCDT) and its dual. A dynamic search routing may be performed using a DS* routing algorithm to determine the shortest path on the dual graph between a start point and an end point. Based on a dynamic pushing technique, net ordering problems may be solved. A first wire can be topologically routed. Dynamic search routing of a second wire may be performed. The first wire may be pushed or detoured in response to the dynamic searching routing of a second wire.

Embodiments of the present invention include techniques for reducing artifacts in rendered images. In one embodiment, a dual UV engine generates a dual of the graph defined by an initial UV set associated with a 3D model. The dual UV engine then uses existing flattening and layout engines to generate a dual UV set from this dual graph. Using the dual graph to define the dual UV set ensures that the seams corresponding to the initial UV set and the dual UV set minimally intersect.

Techniques are disclosed for determining if the decomposition of layout design data is feasible, and for optimizing the segmentation of polygons in decomposable layout design data. Layout design data is analyzed to identify the edges of polygons that should be imaged by separate lithographic masks. In addition, proposed cut paths are generated to cut the polygons in the layout design data into a plurality of polygon segments. Once the separated edges and cut paths have been selected, a conflict graph is constructed that reflects these relationships. Next, a dual of the conflict graph is constructed. This dual graph will have a corresponding separation dual graph edge for each separated polygon edge pair in the layout design data. The dual graph also will have a corresponding cut path dual graph edge for each proposed cut path generated for the layout design data. After the dual graph has been constructed, it is analyzed to determine which of the proposed cut paths should be kept and which should be discarded. The layout design data is then modified to include the cut paths corresponding to the selected cut path dual graph edges. Alternately or additionally, separate sets of layout design data may be decomposed from original layout design data using the selected cut paths.

The invention provides an indoor position service-oriented navigation network construction method, which comprises the following steps of S1, importing a BIM building model in a target building IFC format, and obtaining an original model capable of being applied to indoor position service through three-dimensional visualization; S2, extracting semantic information in IFC through semantic filtering, and constructing an ontology model in a formalized mode; S3, in the form of a dual graph, converting the topological relation in the original model into a graph model; S4, extracting geometric information of structural components such as rooms, columns, walls and stairs and indoor facilities such as furniture in the BIM model, constructing constraint boundaries, performing spatial subdivision through a limited Delaunay triangulation refinement algorithm, and constructing a geometric network model; and S5, integrating the ontology model, the graph model and the geometric model data to form anavigation network for indoor position service. Based on BIM model data, a dual model is established by extracting semantic information, geometric information and a topological relation, and a construction method considering a real indoor environment navigation geometric network is explored.

Embodiments of the present invention include techniques for reducing artifacts in rendered images. In one embodiment, a dual UV engine generates a dual of the graph defined by an initial UV set associated with a 3D model. The dual UV engine then uses existing flattening and layout engines to generate a dual UV set from this dual graph. Using the dual graph to define the dual UV set ensures that the seams corresponding to the initial UV set and the dual UV set minimally intersect.

The invention provides a latent congested road intersection prediction method and device, relating to the intelligent traffic field, and can conveniently, accurately, rapidly, reliably and stably predict road intersections having a higher possibility of congestion with low cost. The method comprises: generating SHP data of roads and intersections of an object area; based on the SHP data, using road segments as peaks, and topological nodes as connecting sides to form a road network topological dual graph representing the connection relation among road segments; based on the road network topological dual graph, performing community-based agglomeration on the road segments to obtain a plurality of communities; and assessing the entropy of intersections in a community greater than a filtering threshold, and determining latent congested road intersections based on assessed entropy. The latent congested road intersection prediction method and device are suitable for conveniently, accurately and rapidly predict congestion of road intersections.

The invention discloses a full traffic prediction method based on a dual graph framework, and the method comprises the steps: (1) expressing a road network structure as a topological graph, taking anintersection as a node, and taking a road segment connected with the intersection as an edge; preparing historical edge and node data and future edge and node data; (2) constructing a historical information encoder, inputting historical data into the encoder, realizing information transmission between edges and nodes through multi-layer dual mapping, and splicing outputs of the multi-layer dual mapping into a historical feature tensor; (3) constructing a future prediction decoder, decoding the historical feature tensor into future spatial-temporal features, and outputting a future prediction result; (4) taking an error between the prediction result and the actual data as a loss function to perform model training until the loss function converges; and (5) carrying out model test by using the trained model, and carrying out application after the test is finished. According to the invention, the prediction result can obtain the total complete description of the future traffic condition, and the prediction accuracy is high.

The invention belongs to the technical field of communication and relates to a host-standby machine duplicated hot-backup system and method based on a Rapid IO bus technique. A dual-communication computer, a dual-graph production computer and a Rapid IO switchboard are in cascade connection to build the host-standby machine duplicated hot-backup system, a self-adaption dynamic recursion ergodic Rapid IO network method which combines with a grading exchange port in a depth-first mode to conduct sequential processing is designed, a host-standby machine duplicated communication hot-backup and communication route redundant function of a communication computer of a cabin displayer is achieved. A Rapid IO data communication protocol (highest rate can achieve 3.125Gbps) replaces a communication mode such as dual-opening RAM, a 429 bus and a 1394 bus to achieve stable and reliable transmission of inter-plate high speed large volume data, communication rate is improved, and communication reliability and system stability are strengthened.

The invention discloses a method for identifying a vulnerable line and preventing a cascading fault of a power system. The method comprises the following steps of firstly, converting a complicated power grid to a bidirectional weighing dual graph; secondly, determining node constraint coefficients in the dual graph by active power coupling between power transmission lines and a dependence relation; thirdly, sequencing the vulnerable lines according to a sequence of constraint coefficients from small to large, and completing vulnerable line identification; and finally, improving a typical OPA model according to a vulnerable power transmission line identification result, and proposing a vulnerable power transmission line identification-based cascading fault prevention method. With the methodproposed by the invention, the vulnerable power transmission line which probably causes high risk can be effectively and accurately identified, and the power failure risk and the cascading fault power-failure probability of the system can be reduced.

The invention discloses an entity alignment method based on an edge type attention mechanism. The entity alignment method comprises the following steps: constructing a type dual graph of an electric power knowledge graph; training node characteristics of the type dual graph by adopting an attribute attention mechanism and obtaining embedding of each node in the type dual graph; converting nodes inthe type dual graph into attention coefficients of relationships in the knowledge graph, gathering neighbor entities and relationship information by the nodes in the knowledge graph according to theattention coefficients of the relationships, and obtaining a final structured embedding result; calculating the value of the loss function and aligning entities in different knowledge graphs to obtainentity pairs aligned by entities in different knowledge graphs. The invention also discloses a method for realizing the entity alignment method based on the edge type attention mechanism. According to the method, the thought of relationship information mining is broadened, the structural characteristics of entities are mined, and the method is high in accuracy, good in practicability, simple andreliable.

Method for detecting hotspots in a circuit layout includes constructing a layout graph having nodes, corner edges and proximity edges from the circuit layout, converting the layout graph to a corresponding dual graph, and iteratively selecting edges and nodes having weights greater than a predetermined threshold value at each iteration as hotspots.

The invention relates to the technical field of microelectronics, in particular to a grid electrode LELE dual graph forming method based on a DARC mask structure. A medium anti-reflection layer hard mask structure based on an advanced graph film layer is formed through double exposure techniques, so that an APF is finally used as a mask for a polycrystalline silicon etching technique. In addition, in the second etching process of a dual graph forming technique, a DARC hard mask is used for replacing a traditional silicon oxide hard mask, a bottom layer structure ODL based on spin coating and a middle layer structure SHB, so that the technical process where the APF is used as the mask in a relatively mature 40nm technology node is continued. Therefore, cost is reduced, and the maturity and stability of a 22-namometer technology node technique and technology node techniques below 22 namometers are further improved at the same time.

The invention provides a large-scale group performance animation synthesis method. The method comprises the steps of: firstly confirming original states of to-be-synthesized large-scale group performance animations and acquiring instance data of small-scale group performance animations; and then establishing group motion dual graphs, and synthesizing large-scale group performance animations according to the small-scale group performance animations. According to the large-scale group performance animation synthesis method, the small-scale group performance animations and the large-scale group performance animations which have apparently similar motion features are related with the large-scale group performance animations through the group motion dual graphs; by virtue of the group motion dual graphs, a user only needs to manually design or obtain the small-scale group performance animations by utilizing various known group motion capturing methods so as to automatically generate the large-scale group performance animations; and the purposes of editing and controlling the large-scale group performance animations by a designer can be realized by indirectly editing and designing the small-scale group performance animations.

The invention discloses a method, device and equipment for denoising a triangular mesh based on a dual graph neural network and a storage medium thereof, and the method comprises the steps: constructing a graph data structure for vertexes and triangular patches in the triangular mesh at the same time to form a dual graph, and inputting a vertex graph and a patch graph into a corresponding graph neural network; the graph structure established based on the vertexes is used for pre-denoising noise vertexes in the neural network, and the graph structure established based on the patches is used for denoising the normal direction of the patches in the neural network; and updating the pre-denoised vertexes according to the normal direction of the denoised patch so as to output a final triangular mesh. According to the scheme, the topological structure of the triangular mesh is better utilized, the denoising effect is better, geometric details are better kept, and in quantization errors, the average error of the normal angle of the patch and the average error of the vertex distance are smaller.

The invention provides a method and device for obtaining a cell relation model and a method and device for recommending a cell handover guidance parameter, and relates to the field of artificial intelligence (AI) and the technical field of data communication. The method for obtaining a cell relation model comprises the following steps: separately carrying out feature extraction on a topological graph representing historical traffic statistic data of N cells in a specified area and a dual graph corresponding to the topological graph through a GCN (graph convolutional neural network) to obtain first feature information of the topological graph and second feature information of the dual graph; based on the first feature information and the second feature information, determining fusion feature information of the N cells; fitting the fusion feature information of the N cells and the historical handover guidance parameters of the N cells to obtain a cell relation model used for indicating the function relation among the network state parameters, the handover guidance parameters and the performance indexes of the N cells in the specified area; and finally, based on the cell relation model, determining a target handover guidance parameter through a gradient descent algorithm. The network performance obtained by the method is better in effect.

The invention provides a method of increasing high-aspect ratio photoetching process windows through a dual graph technology. The method comprises the following steps: photoetching process requirements are divided to a combination of twice photoetching; a first photomask is used to carry out first photoetching processing on a process graph; a second photomask is used to carry out second photoetching processing on the process graph, and wherein the process window graph formed by the combination of the first photomask and the second photomask has a pixel size meeting the process requirements andare 1D groove graph photomasks. According to the method of increasing the high-aspect ratio photoetching process windows through the dual graph technology, in view of the high-aspect ratio layer, when the pixel size is reduced gradually and a target linewidth is reduced, the dual graph technology is adopted to split a one-layer 2D point graph to a two-layer 1D groove graph, and in combination ofthe PSM photomask process mode, the high-aspect ratio layer process windows are increased.

The invention discloses a method for expressing a topological relation of a polygonal scene. The method is used for expressing a topological structure of the polygonal scene. The method comprises the following steps of A, performing topological homeomorphism transformation on the polygonal scene and generating a topological homeomorphism graph; B, generating a dual graph for the topological homeomorphism graph generated in the step A, and then extracting a neighborhood relation between polygons in the dual graph; and C, establishing a polygonal shape set according to the neighborhood relation between the polygons obtained in the step B, performing sorting on the polygons by adopting k-order neighbor, and generating k-order neighbor coded character strings of the polygons to express the topological structure of the polygonal scene. According to the method for expressing the topological relation of the polygonal scene, provided by the invention, the topological relation of the polygonal scene can be expressed by using simple character strings, so that the efficiency of judging topological homeomorphism of the polygonal scene is improved.

The invention relates to the technical field of high-performance computing, in particular to a DCU cluster-oriented large-scale finite element grid parallel partitioning method and device. The method comprises the steps: discretizing a solving area into grids, and describing the distribution situation of the grids obtained through discretization through a weighted dual graph, wherein vertexes in the weighted dual graph represent the grids, edges in the weighted dual graph represent the connection relation between grids, and the dividing speed of the recursive spectral dichotomy can be greatly improved through parallelism of sub-graph dividing tasks; carrying out parallel calculation of hot spots of the spectral dichotomy on the basis of parallelism of sub-graph dividing, so the processing time of finite element grid dividing is further shortened, the problem that the time demand for dividing the large-scale grids is too long in a serial recursive spectrum dichotomy method is solved, and the efficiency of solving the problem is greatly improved.

The invention discloses an obstacle avoidance shortest path planning method based on a Voronoi diagram, and the method comprises the steps: collecting the information of an obstacle, including the position of a starting point, the position of a destination point, the size of the obstacle, and the position of a central point of the obstacle; according to the positions of the center points of the obstacles, regarding the center points of the multiple obstacles as a point set S on a plane, constructing optimal triangulation for the point set S through adoption of a Delaunay triangulation method, and obtaining a triangulation straight line dual graph, namely a Voronoi graph corresponding to the point set S and the center points of the obstacles, through construction; expanding each edge of the Voronoi diagram in the direction in which the two obstacles determining the edge are located to obtain a simple polygon; and solving an Euclidean shortest path of two points in the simple polygon to obtain a final obstacle avoidance shortest path. The method is low in complexity, and has high self-adaptability and certain flexibility for obstacles with different threat degrees.

A cross-view gait recognition method based on subspace learning of joint hierarchical selection comprises the following steps: firstly, dividing gait samples of a target view and a registered view into a training set and a test set, simultaneously carrying out hierarchical block division on gait data of the two views, respectively vectorizing multi-level gait energy image blocks, then carrying out feature selection, and carrying out cascade connection; secondly, projecting registration view angle data and target view angle data to a public subspace at the same time, enhancing the relation between the registration view angle data and the target view angle data in a mode of constructing a cross-view-angle dual graph, performing effective feature selection and effective gait energy graph block selection in the projection process, removing redundancy, forming a registration sample set in the public subspace, and projecting test target visual angle data into the public subspace through the trained target visual angle projection matrix to form target sample sets in the public subspace, and performing gait recognition on the two sample sets by adopting a nearest neighbor mode of Euclidean distance. According to the method, the registered view gait data is introduced into the target view field, and the cross-view gait recognition effect is enhanced.

The invention provides a grid segmentation method based on a graph convolution network, and the method takes the surface of a grid as a basic unit, and carries out the graph convolution operation in a dual graph formed based on the adjacency relation of the surface, so as to obtain the feature representation of the surface. In the feature acquisition stage, static and dynamic edge convolution is utilized at the same time, and the capability of learning information from potential relations between surfaces is also obtained while the actual adjacency relation is utilized. In addition, the features are further enhanced by utilizing the idea of feature embedding in instance segmentation, and finally, all parts of the grid are segmented by utilizing the enhanced features. According to the method, a better result is obtained on a data set segmented by a plurality of parts.