We propose a new fast algorithm for solving the Maximum Common Subgraph (MCS) problem. MCS is an NP-complete problem. In this paper, we focus on a special class of graphs, i.e. Planar Triangulation graphs, which are c...
详细信息
the recognition of unconstrained handwriting images is usually based on vectorial representation and statistical classification. Despite their high representational power, graphs are rarely used in this field due to a...
详细信息
the proceedings contain 34 papers. the topics discussed include: a global method for reducing multidimensional size graphs;graph descriptors from B-matrix representation;dimensionality reduction for graph of words emb...
ISBN:
(纸本)9783642208430
the proceedings contain 34 papers. the topics discussed include: a global method for reducing multidimensional size graphs;graph descriptors from B-matrix representation;dimensionality reduction for graph of words embedding;entropy versus heterogeneity for graphs;learning generative graph prototypes using simplified von Neumann entropy;information-geometric graph indexing from bags of partial node coverages;maximum likelihood for Gaussians on graphs;towards performance evaluation of graph-based representation;measuring the distance of generalized maps;aggregated search in graph databases: preliminary results;speeding up graph edit distance computation through fast bipartite matching;two new graph kernels and applications to chemoinformatics;parallel graduated assignment algorithm for multiple graph matching based on a common labeling;and smooth simultaneous structural graph matching and point-set registration.
In this paper, we investigate the Max-Cut problem and propose a probabilistic heuristic to address its classic and weighted version. Our approach is based on the Estimation of Distribution Algorithm (EDA) that creates...
详细信息
the proceedings contain 34 papers. the special focus in this conference is on graph-basedrepresentations in patternrecognition. the topics include: Speeding Up graph Edit Distance Computation through Fast Bipartite ...
ISBN:
(纸本)9783642208430
the proceedings contain 34 papers. the special focus in this conference is on graph-basedrepresentations in patternrecognition. the topics include: Speeding Up graph Edit Distance Computation through Fast Bipartite Matching;two New graph Kernels and Applications to Chemoinformatics;generalized Learning graph Quantization;parallel Graduated Assignment Algorithm for Multiple graph Matching based on a Common Labelling;smooth Simultaneous Structural graph Matching and Point-Set Registration;automatic Learning of Edit Costs based on Interactive and Adaptive graphrecognition;exploration of the Labelling Space Given graph Edit Distance Costs;graph Matching based on Dot Product Representation of graphs;indexing with Well-Founded Total Order for Faster Subgraph Isomorphism Detection;graph Descriptors from B-Matrix Representation;graph Transduction as a Non-cooperative Game;a graph-based Approach to Feature Selection;spatio-Temporal Extraction of Articulated Models in a graph Pyramid;semi-supervised Segmentation of 3D Surfaces Using a Weighted graph Representation;convexity Grouping of Salient Contours;hierarchical Interactive Image Segmentation Using Irregular Pyramids;tiled Top-Down Pyramids and Segmentation of Large Histological Images;segmentation of Similar Images Using graph Matching and Community Detection;automatic Street graph Construction in Sketch Maps;people Re-identification by graph Kernels Methods;dimensionality Reduction for graph of Words Embedding;automatic Labeling of Handwritten Mathematical Symbols via Expression Matching;structure-based Evaluation Methodology for Curvilinear Structure Detection Algorithms;keygraphs for Sign Detection in Indoor Environments by Mobile Phones;classification of graph Sequences Utilizing the Eigenvalues of the Distance Matrices and Hidden Markov Models;using Kernels on Hierarchical graphs in Automatic Classification of Designs;entropy versus Heterogeneity for graphs.
A novel method is proposed to match functional connectivity patterns represented by graphs for spatial registration of fMRI data. Different from existing functional connectivity patternbased registration methods that...
详细信息
ISBN:
(纸本)9783642408427;9783642408434
A novel method is proposed to match functional connectivity patterns represented by graphs for spatial registration of fMRI data. Different from existing functional connectivity patternbased registration methods that detect corresponding functional units across different subjects by minimizing their difference in functional connectivity strength, our method adopts a graph representation to characterize functional connectivity information among all voxels in fMRI data of each subject, then detects spatial correspondence between subjects using graph matching. To integrate information of both functional connectivity strength and spatial relations, the graph representation of functional connectivity information of fMRI data models each voxel as one graph node and connects each pair of graph nodes with an edge weighted by their functional connectivity strength measure, estimated as correlation coefficient between their functional signals. To make the graph matching computationally feasible, an iterative matching strategy with stochastic resampling is proposed to match graphs of spatially distributed local functional connectivity patterns and subsequently to drive the image registration iteratively. the proposed method has been validated by registering resting state fMRI data of 20 healthy subjects. the validation experiment results have demonstrated that our method can achieve improved inter-subject functional consistency. A comparison experiment result has further indicated that the proposed method can achieve better performance than existing methods.
Multi-view human action recognition has gained a lot of attention in recent years for its superior performance as compared to the single view recognition. In this paper, we propose algorithms for the real-time realiza...
详细信息
ISBN:
(纸本)9781450317726
Multi-view human action recognition has gained a lot of attention in recent years for its superior performance as compared to the single view recognition. In this paper, we propose algorithms for the real-time realization of human action recognition in distributed camera networks (DCNs). We first present a new method for fast calculation of motion information by Motion Local Ternary pattern (Mltp) that is tolerant to illumination change, robust in homogeneous region and computationally efficient. Next, we combine the local interest point detector with Mltp to generate 3D patches containing motion information and introduce two feature descriptors for the extracted 3D patches. Taking advantage of the proposed Mltp, 3D patches generated from background can be further removed automatically and thus the foreground patches can be highlighted. Finally, the histogram representationsbased on Bag-of-Words modeling, are transmitted from local cameras to the base station for classification. At the base station, a probability model is produced to fuse the information from various views and a class label is assigned accordingly. Compared to the existing algorithms, the proposed methods have three advantages: 1) no preprocessing is required;2) communication among cameras is unnecessary;and 3) positions and orientations of cameras do not need to be fixed. We further evaluate both descriptors on the most popular multi-view action dataset IXMAS. Experimental results indicate that our approaches repeatedly achieve state-of-the-art results when various numbers of views are tested. In addition, our approaches are tolerant to the various combination of views and benefit from introducing more views at the testing stage.
While finding clones in source code has drawn considerable attention, there has been only very little work in finding similar fragments in binary code and intermediate languages, such as Java bytecode. Some recent stu...
详细信息
While finding clones in source code has drawn considerable attention, there has been only very little work in finding similar fragments in binary code and intermediate languages, such as Java bytecode. Some recent studies showed that it is possible to find distinct sets of clone pairs in bytecode representation of source code, which are not always detectable at source code-level. In this paper, we present a bytecode clone detection approach, called SeByte, which exploits the benefits of compilers (the bytecode representation) for detecting a specific type of semantic clones in Java bytecode. SeByte is a hybrid metric-based approach that takes advantage of both, Semantic Web technologies and Set theory. We use a two-step analysis process: (1) pattern matching via Semantic Web querying and reasoning, and (2) Content matching, using Jaccard coefficient for set similarity measurement. Semantic Web-basedpattern matching helps us to find method blocks which share similar patterns even in case of extreme dissimilarity (e.g., numerous repetitions or large gaps). Although it leads to high recall, it gives high false positive rate. We thus use the content matching (via Jaccard) to reduce false positive rate by focusing on content semantic resemblance. Our evaluation of four Java systems and five other tools shows that SeByte can detect a large number of semantic clones that are either not detected or supported by source code based clone detectors.
In the field of structural patternrecognitiongraphs constitute a very common and powerful way of representing objects. the main drawback of graphrepresentations is that the computation of various graph similarity m...
详细信息
ISBN:
(纸本)9783642208447
In the field of structural patternrecognitiongraphs constitute a very common and powerful way of representing objects. the main drawback of graphrepresentations is that the computation of various graph similarity measures is exponential in the number of involved nodes. Hence, such computations are feasible for rather small graphs only. One of the most flexible graph similarity measures is graph edit distance. In this paper we propose a novel approach for the efficient computation of graph edit distance based on bipartite graph matching by means of the Volgenant-Jonker assignment algorithm. Our proposed algorithm provides only suboptimal edit distances, but runs in polynomial time. the reason for its sub-optimality is that edge information is taken into account only in a limited fashion during the process of finding the optimal node assignment between two graphs. In experiments on diverse graphrepresentations we demonstrate a high speed up of our proposed method over a traditional algorithm for graph edit distance computation and over two other sub-optimal approaches that use the Hungarian and Munkres algorithm. Also, we show that classification accuracy remains nearly unaffected by the suboptimal nature of the algorithm.
暂无评论