Advances in Visual Computing: 5th International Symposium, ISVC 2009, Las Vegas, NV, USA, November 30 - December 2, 2009, Proceedings, Part I (Lecture ... Vision, Pattern Recognition, and Graphics)

It is with greatpleasure that we present the proceedings of the 5th International Symposium on Visual Computing (ISVC 2009), which was held in Las Vegas, Nevada. ISVC o?ers a common umbrella for the four main areas of visual c- puting includingvision,graphics,visualization,andvirtualreality.Thegoalisto provide a forum for researchers, scientists, engineers, and practitioners throu- out the world to present their latest research?ndings, ideas, developments, and applications in the broader area of visual computing. This year, the program consisted of 16 oral sessions, one poster session, 7 special tracks, and 6 keynote presentations. Also, this year ISVC hosted the Third Semantic Robot Vision Challenge.The responseto the call for papers was verygood;wereceivedover320submissionsfor themainsymposiumfromwhich we accepted 97 papers for oral presentation and 63 papers for poster presen- tion. Special track papers were solicited separately through the Organizing and Program Committees of each track. A total of 40 papers were accepted for oral presentation and 15 papers for poster presentation in the special tracks. All papers were reviewed with an emphasis on potential to contribute to the state of the art in the?eld. Selection criteria included accuracy and originality of ideas, clarity and signi?cance of results, and presentation quality. The review process was quite rigorous, involving two to three independent blind reviews followed by several days of discussion. During the discussion period we tried to correct anomalies and errors that might have existed in the initial reviews. Front matter......Page 1 Introduction......Page 41 Shape Representations......Page 43 Shape Descriptors......Page 44 Fourier Descriptors......Page 45 Hu Moment Invariants, Φ......Page 46 Experimental Results......Page 47 References......Page 50 Introduction......Page 52 The Particle-Based Detector......Page 53 Redundant Multi-camera Setup: Sampling from the View-Volume Intersection......Page 55 Applications and Experimental Results......Page 57 Conclusion and Future Work......Page 60 Introduction......Page 62 Features and Matching......Page 63 Feature Matching......Page 65 Model Fitting......Page 66 Model Extraction......Page 68 Evaluation......Page 71 Conclusions......Page 72 Related Work......Page 74 The {\it Bag of Features} Concept......Page 75 Mathematical Foundations......Page 76 Algorithm......Page 78 Retrieval by Feature Histograms......Page 80 Conclusions and Outlook......Page 82 Introduction......Page 84 Efficient Pixel-Level Categorization......Page 85 Dense Features......Page 86 Codebook Generation......Page 87 Integral Linear Classifiers......Page 88 Experiments......Page 89 Parameters of the ERF......Page 90 Comparison between HKM and ERF......Page 91 Time Cost Evaluation......Page 92 Conclusions......Page 93 Introduction......Page 95 Related Work......Page 96 Lattice-Boltzmann Lighting......Page 97 Cinematic Relighting......Page 98 Dynamic Lighting......Page 99 Implementation......Page 100 Compression......Page 101 Results......Page 102 Conclusion......Page 104 Introduction......Page 107 Related Work......Page 108 Line Drawing......Page 109 Cartoon Water Shader......Page 110 Water Flow Shader......Page 113 Results and Analysis......Page 114 Evaluation......Page 115 Conclusion and Discussion......Page 116 Introduction......Page 119 Related Work......Page 120 Proximity Queries......Page 121 System Overview......Page 123 Experiments......Page 124 Conclusion......Page 126 Introduction......Page 129 The GPLVM......Page 130 The SGPLVM......Page 131 Building an Audio-Visual Corpus......Page 132 Speech Parameterisation......Page 133 Audio-Visual Mapping......Page 134 Experiments......Page 135 Results......Page 136 Conclusions and Future Work......Page 139 Introduction......Page 141 Previous Work on Ridge Extraction......Page 143 Tensor Product B-Spline Representation......Page 144 Characteristics of Ridges on Surfaces with Deficient Smoothness......Page 145 Dealing with Deficient Smoothness......Page 146 Results and Discussion......Page 147 Conclusions, Limitations and Future Work......Page 149 Introduction......Page 151 Related Work......Page 153 Shader Compiler......Page 154 Algorithm for Partitioning with Minimum Duplication (PMD)......Page 155 Comparison with the Naive Algorithm......Page 157 Selective Partitioning of Vertex Shader......Page 158 Simulation Framework......Page 159 Experiments and Results......Page 161 Conclusion......Page 162 Introduction......Page 165 Methods......Page 166 Overview Visualization......Page 167 Detailed Visualization......Page 168 Experiment......Page 170 Survey Results......Page 171 Conclusion and Future Work......Page 172 References......Page 173 Introduction......Page 175 Segmentation of the Source Image......Page 177 First Image Importance Map......Page 179 Second Image Importance Map......Page 181 Construction of the Image Importance Maps......Page 182 Conclusion......Page 185 Introduction......Page 187 Related Works......Page 189 Overview......Page 190 Building Speedlines......Page 191 Experimental Results......Page 192 Discussion......Page 193 Conclusions and Future Works......Page 195 Introduction......Page 197 Scale and Human Visual Attention......Page 198 Saliency and Compression......Page 199 Evaluation......Page 202 Fast Occlusion Sweeping......Page 207 Fast Sweeping......Page 208 Ambient Occlusion......Page 209 Occlusion Sweeping......Page 210 Implementation and Results......Page 211 Subsurface Scattering Effects......Page 215 Conclusion......Page 217 Introduction......Page 219 Simulated Bee Colony Algorithm Implementation......Page 222 Results......Page 224 Experiment #1 – Congressional Voting Data......Page 225 Experiment #2 – Synthetic Datasets......Page 226 References......Page 227 Introduction......Page 229 Related Work......Page 230 Pre-processing......Page 231 Feature Detection......Page 232 Feature Description......Page 233 Feature Matching and Outlier Removal......Page 234 Feature Integration......Page 236 Performance......Page 238 Conclusion......Page 239 Introduction......Page 241 Computation of the Features Orientation......Page 242 Generation of the Planar Textures......Page 244 Localization Using Planar Features......Page 246 Results......Page 247 Conclusion......Page 249 Fast and Accurate Structure and Motion Estimation......Page 251 Overview......Page 253 Five Point Pose Estimation......Page 254 Distance Constraint......Page 255 Evaluation......Page 256 Inlier Frequency Evaluation......Page 257 Precision Evaluation......Page 260 Conclusions......Page 261 Introduction......Page 263 Flow Standardization......Page 265 Learning Algorithm......Page 266 Results......Page 268 Conclusions......Page 271 Introduction......Page 273 Half Life 2 World......Page 275 Mapping......Page 276 Experimental Results......Page 278 Navigation and Path Planning......Page 279 Conclusion......Page 281 Introduction......Page 283 Gabor Filters......Page 285 Separability of Gabor Filters......Page 286 Integrating Interpolation and Convolution......Page 287 Experimental Results......Page 289 Conclusion......Page 291 Introduction......Page 293 Regions and Local Feature Descriptors......Page 294 Region Graphs and Region Spectra......Page 295 Restoring Information with Neumann Eigenvalues......Page 297 Pattern Detection without Segmentation......Page 298 Experiments......Page 300 MPEG-7 Results......Page 301 Skin Cell Results......Page 302 Conclusion......Page 303 Introduction......Page 305 Harris-Laplace and Harris Corner Measure......Page 306 Analysis of Multi-scale Product of Cornerness Measure......Page 307 Performance Evaluation of the Proposed Approach......Page 309 Experiment Results and Discussion......Page 311 Conclusion......Page 313 References......Page 314 Introduction......Page 315 2D and 3D Junction Extraction and Interpretation......Page 317 2D Interpretation......Page 319 3D Interpretation......Page 320 Temporal Matching......Page 321 Results......Page 322 Quantification on an Artificial Image Sequences......Page 323 Indoor and Outdoor Scenes......Page 324 Conclusion......Page 325 Introduction......Page 327 Mathematical Foundations......Page 328 Vectorial Harmonics......Page 329 Fast and Accurate Correlation in ${\mathcal VH}$......Page 331 Rotation Estimation......Page 333 Experiments......Page 334 References......Page 336 Introduction......Page 337 Problem Formulation......Page 339 Overview......Page 340 Category-Specific Contextual Visual Word......Page 341 Experimental Results......Page 342 Conclusion......Page 345 Introduction......Page 347 Edge-Preserving Multiscale Analysis......Page 348 Experimental Results......Page 352 Introduction......Page 357 Related Work......Page 358 Using Force for Tumor Detection......Page 359 Preprocessing......Page 360 Coarse Detection of Tumor......Page 361 Evaluation......Page 363 Comparison of Force with Ground Truth......Page 364 References......Page 365 Introduction......Page 367 Digital Deformable Model......Page 368 Top-Down Segmentation for Histology Images......Page 370 Experiments......Page 371 Conclusion......Page 375 Introduction......Page 377 Segmentation......Page 378 Tensor Weighted Distances......Page 379 Realistic Synthetic Cell Images......Page 380 Bright-Field Microscopy Images......Page 381 Evaluation......Page 382 Bright-Field Microscope Images......Page 383 Conclusions......Page 385 Introduction......Page 387 Pulmonary Nodule Definitions......Page 389 Nodule Simulation......Page 390 Statistical Nodule Modeling......Page 391 Performance Evaluation......Page 392 References......Page 395 Introduction......Page 397 Architecture Overview......Page 398 Reconstruction of Sub-volume Blocks......Page 400 Hybrid Multiresolution Volume Rendering Incorporating Sub-volume Blocks......Page 401 Communication and Data Exchange......Page 402 Results......Page 403 Conclusion and Future Work......Page 405 Introduction......Page 407 Voronoi Sampling......Page 409 Feature Space......Page 410 Tree Induction......Page 411 Multiple Object Detection......Page 412 Performance Measure......Page 413 Generating a Gold Standard......Page 414 Clear Cell Renal Cell Carcinoma (ccRCC)......Page 415 Proliferation in Murine Liver Tissue......Page 416 Conclusion......Page 417 Introduction......Page 419 Geometric Study of 3D Textures......Page 420 Statistical Measure for Contrast......Page 423 Psychological Experiments......Page 424 Comparison between Human and Computational Ranking......Page 425 Evaluation of Our Method of Texture Segmentation......Page 426 Segmentation Results of Volumetric Ultrasound Images......Page 428 Conclusion......Page 429 Introduction......Page 431 Unconstrained Reconstruction......Page 432 Detection......Page 434 Tracking......Page 435 Sparse Bundle Adjustment with Additional Camera Center Constraints......Page 436 Experiments with Synthetic Data......Page 438 Experiments with Real Scenes......Page 439 Conclusion......Page 440 Introduction......Page 443 Optical Flow Computation......Page 444 Multiresolution Expression......Page 445 Zooming and Optical Flow......Page 446 Mathematical Property of Algorithm......Page 447 Numerical Examples......Page 448 Optical Flow Computation......Page 449 Estimation of Zooming Operation......Page 451 Conclusions......Page 452 Introduction......Page 455 Segmentation of Motion with Affine Fundamental Matrix......Page 456 Monte Carlo Experiments......Page 460 Conclusions......Page 463 Motivation for Articulated Human Motion Analysis......Page 465 Learning Models for Multiple Poses and Viewpoints......Page 467 Training Data......Page 468 Pose and Viewpoint Estimation......Page 469 Experimental Results......Page 471 Conclusion......Page 473 Introduction......Page 475 Related Work......Page 476 Feature Extraction......Page 477 Motion Model Fitting......Page 478 Results with Noise-Free Synthetic Data......Page 480 Results with Real-World Video Sequences......Page 483 Conclusions......Page 485 Introduction......Page 487 ML Formulation......Page 488 Validation of Point Correspondences......Page 489 Algorithm......Page 491 Experiments......Page 492 Conclusion......Page 495 References......Page 496 Introduction......Page 497 Problem Formulation......Page 498 Linkage between Feature Correspondences and Models......Page 499 Eigenlink......Page 500 Adaptive Random Sampling......Page 501 Two View Image Matching......Page 505 Conclusion......Page 506 Introduction......Page 508 Related Works......Page 509 RiverLand Overview......Page 510 Creating the River Network......Page 511 Generating Terrain Height......Page 513 User-Defined Ridges......Page 514 Randomized Fractal Overlay......Page 515 Performance......Page 516 Conclusions......Page 517 Introduction......Page 520 Related Work......Page 521 System Architecture......Page 522 Calibration......Page 523 Reconstruction......Page 524 Pong......Page 525 Video Compositing......Page 527 Conclusion......Page 528 Introduction......Page 530 The VEE System......Page 531 Trail Recording......Page 532 Panoramic Video Stitching......Page 533 Trail Playback......Page 535 Locomotion Display......Page 536 Concluding Remarks......Page 538 Introduction......Page 540 Energy Function......Page 542 Likelihood Term......Page 543 Prior Term......Page 545 Simulated Environment......Page 546 Qualitative Analysis......Page 547 References......Page 549 Previous Work......Page 551 JanusVF......Page 552 Previous Work......Page 553 Offset Grids......Page 554 Quadtree Subdivision......Page 555 Comparison Tests......Page 556 Viewing Angle Tests......Page 557 Free Motion Tests......Page 558 Conclusion......Page 559 Introduction......Page 561 Related Work......Page 562 Unconstrained Poses......Page 563 Constrained Poses......Page 565 Conclusion and Future Work......Page 568 References......Page 569 Introduction and Background......Page 571 Image Pre-processing......Page 573 Seed Finding in Scale Space......Page 574 Nuclei Segmentation Based on Evolving Generalized Voronoi Diagram......Page 575 Experimental Results......Page 578 Discussions and Conclusions......Page 581 Introduction......Page 584 Energy Functional......Page 585 Minimization of the Energy Functional......Page 586 Benchmarks......Page 588 Application to Real Data......Page 590 Conclusions and Discussion......Page 592 Introduction......Page 594 Particle Filter Basis......Page 595 Image Registration Guided by Particle Filter......Page 596 Algorithm: Image Registration Guided by Particle Filter......Page 597 Experiments and Results......Page 598 Conclusions and Future Work......Page 602 Introduction......Page 604 Review of Orientation Fields and the OFT......Page 606 Generation of Orientation Weights......Page 609 Summary of the Algorithm......Page 611 Results......Page 612 Discussion......Page 613 Conclusion......Page 614 Introduction......Page 616 Segmentation of a Zygote Cell......Page 618 Detailed Recovery of the Cell Contours......Page 619 Segmentation of Pronuclei in a Zygote Image......Page 621 Experimental Results......Page 622 Segmentation of Pronuclei......Page 623 Conclusions......Page 624 Introduction......Page 626 User Input......Page 628 Fitting the Articulated Model to the Splines......Page 629 Optimising Vertebrae Location......Page 630 Evaluation Using Manually Identified Splines......Page 632 Discussion and Conclusion......Page 633 Introduction......Page 636 Related Work......Page 637 Trivariate Splines and the BB-Form......Page 638 Preliminaries and Basic Notation......Page 639 Trivariate Splines – GPU Visualization......Page 641 The CUDA Kernels......Page 642 Vertex Shader Computations......Page 643 Fragment Shader Computations......Page 644 Results and Discussion......Page 645 Conclusion......Page 646 Introduction......Page 648 Preliminaries......Page 650 Overview......Page 651 Counting the Number of Curve Arcs......Page 653 Identifying and Tracing Closely Located Arcs, Grid Perturbation......Page 654 Algorithm's Pseudocode......Page 655 Addressing Accuracy Problems and Relaxing CAD Assumption......Page 656 Results and Conclusion......Page 657 Introduction......Page 660 Overview......Page 662 Table Structure......Page 664 Plane Sweeping......Page 666 Single Cube......Page 667 Empty-Space Culling Using Octree......Page 668 Conclusion......Page 669 Introduction......Page 672 Related Work......Page 673 Statistical Modelling......Page 674 Illuminant Spectra Estimation......Page 676 Experimental Evaluation......Page 679 Experimental Results......Page 680 Conclusions......Page 682 Introduction......Page 684 Related Work......Page 686 Blending......Page 687 Selecting the Bricks......Page 689 Rendering......Page 690 Implementation and Results......Page 691 Conclusions and Future Work......Page 693 References......Page 694 Introduction......Page 696 Background on Tangent Vector Fields as One-Forms......Page 698 Texture Representation......Page 699 Pseudo-height and Tilt......Page 700 Texture Reconstruction......Page 701 Editing and Animation......Page 703 Discussion and Conclusion......Page 705 Introduction......Page 708 Solving the Images of the Circular Point......Page 710 Two Non-intersecting, Non-concentric Circles......Page 711 Synthetic Scene......Page 712 Real World Scene......Page 714 Conclusion......Page 716 Introduction......Page 718 Related Work......Page 719 Poisson Surface Reconstruction......Page 720 Shared Memory Implementation......Page 721 Distributed Memory Implementation......Page 722 Results......Page 725 Conclusion......Page 728 Introduction......Page 730 Stereo SLAM Using Edge Points......Page 732 Matching between 3D Map and Training Images......Page 734 Experiments......Page 736 Conclusions......Page 738 Introduction......Page 740 Image Formation......Page 741 Radiance Maps from Apollo Imagery......Page 742 Juggling Algorithm......Page 744 Experimental Results......Page 745 References......Page 748 Introduction......Page 750 The Ames Stereo Pipeline......Page 751 Bundle Adjustment......Page 752 Sub-pixel Stereo Correlation......Page 753 Bundle Adjustment......Page 756 Conclusions and Future Work......Page 758 Introduction......Page 760 Previous Work......Page 761 Factorization of Correspondence and Camera Pose Errors......Page 762 Signal-to-Noise Ratio (SNR) Analysis......Page 764 Aerial Imagery......Page 765 Artificial Data......Page 766 Signal-to-Noise Ratio......Page 767 Conclusion......Page 768 Introduction......Page 770 A Deformable 3D FaceModel......Page 771 Facial Expression Recognition......Page 772 Classification Results Using the CMU Data......Page 774 Cross-Check Validation Using the CMU Data......Page 776 Dynamic vs. Static Recognition on Non-aligned Videos......Page 777 References......Page 778 Introduction......Page 780 Facial Shape Recovery......Page 781 Experimental Results......Page 785 Conclusion......Page 788 Comparison to Other Methods......Page 790 Shape Model......Page 791 View-Based Local Texture Patches......Page 792 Locating the Features......Page 793 Comparison Results......Page 794 Pose Handling and Estimation......Page 795 Pose Results......Page 796 Tracking Video Sequences......Page 798 References......Page 800 Introduction......Page 802 Related Work......Page 803 Framework......Page 804 Local Modeling......Page 805 Stochastic Search in Local Modeling......Page 806 Experimental Results......Page 808 Conclusions and Future Work......Page 811 Introduction......Page 814 Stereo Matching-Based Reconstruction......Page 815 Experimental Results......Page 816 Validation with FRGC Database......Page 820 Conclusions and Future Work......Page 821 References......Page 822 Introduction......Page 824 Object Types......Page 826 Single Camera......Page 827 Simultaneous Use of Several Cameras......Page 828 Reconstruction Algorithm......Page 829 Surface Voxel Determination......Page 830 Memory Consumption......Page 831 Performance Tests......Page 832 References......Page 834 Introduction......Page 836 Data Term......Page 838 Euler-Lagrange Equation......Page 839 Full Approximation Scheme (FAS)......Page 840 V-Cycle and W-Cycle......Page 841 O-Cycle......Page 842 Multi-Level Adaptive Technique (MLAT)......Page 843 Regularizer Adaptation......Page 844 Future Work and Conclusion......Page 846 Introduction......Page 848 SSD Algorithm......Page 849 CSX......Page 850 Overlapping Communication and Computation......Page 851 Data Distribution......Page 852 Parallelized SSD......Page 854 Implementation and Performance Results......Page 856 Conclusion and Future Works......Page 857 Introduction......Page 859 Related Work......Page 860 Problem Statement......Page 861 Sampling the First Correspondence......Page 862 Sampling the Third Correspondence......Page 863 Setup......Page 866 Results......Page 867 Conclusion......Page 869 Introduction......Page 871 Projective Geometric Relations......Page 873 3D Estimation Using the Cross-Ratio......Page 874 Phase-Shifted Patterns......Page 875 Experiments......Page 876 Quantitative Evaluation (Parametric Surfaces)......Page 877 Qualitative Evaluation......Page 878 Quantitative Evaluation (Non-parametric Surfaces)......Page 879 Conclusion......Page 881 Introduction......Page 883 Plane Sweeping and Related Methods......Page 885 A Hybrid Method......Page 886 Labeling the Projector Pixels......Page 887 Recovering Structure and Projector Motion......Page 889 Calibrating the Setup......Page 890 Discussion......Page 891 Conclusions......Page 893 Introduction......Page 895 Proposed Method......Page 897 Numerical Scheme and Experimental Results......Page 899 Conclusion......Page 903 Introduction......Page 905 Channel Differential Structure......Page 907 Active Surfaces......Page 908 Internal Forces......Page 909 External Forces......Page 910 Experiments and Results......Page 912 Conclusion......Page 913 Introduction......Page 915 Facial Attribute Manipulation......Page 916 Regression Methods......Page 917 Evaluation......Page 919 Binary Attributes......Page 920 Measured Attributes......Page 921 Prior Work......Page 922 Conclusion......Page 924 Introduction......Page 926 Local Binary Fitting Model......Page 927 Contrast Constrained Local Binary Fitting Model......Page 929 Narrow Band Algorithm for Contrast Constrained Local Binary Fitting Model......Page 930 Experimental Results......Page 931 Results of Contrast Constrained Local Binary Model......Page 932 Comparisons with Local Binary Fitting Model and Piecewise Smooth Model......Page 933 Conclusion......Page 934 Introduction......Page 936 Level Sets......Page 937 Rendering Combustion......Page 938 Reconstructing Prior Work......Page 939 Extension 1: Wood Grain......Page 940 Extension 2: Moisture......Page 941 Results......Page 943 Conclusions and Future Work......Page 944 Introduction......Page 946 The Boundary Length Bias Problem......Page 947 Local Bias Correction......Page 949 Probabilistic Formulation......Page 950 Toy Example......Page 951 Contour Completion......Page 952 Medical Background......Page 954 Mathematical Model......Page 955 Phase I......Page 957 Phase II......Page 958 Numerical Result......Page 959 Conclusion......Page 962 Introduction......Page 964 3D Object Representation......Page 965 Viewing Modes......Page 966 Protocol Extension......Page 967 Web Viewer Prototype with WlzIIP......Page 970 Evaluation......Page 971 Discussion and Conclusions......Page 972 Related Work......Page 974 Tensor Field Differential Operators......Page 976 Methods......Page 977 Results......Page 979 Conclusions and Future Work......Page 982 Introduction......Page 984 Methods......Page 985 Filter Scale Selection......Page 986 Filter Orientation Selection......Page 988 Experimental Setup for Quantitative Validation......Page 990 Results......Page 991 References......Page 993 Literature Reviews......Page 995 Data......Page 996 Wavelet-Based Representation for Shapes......Page 998 Multiscale Curvature-Like Characterization......Page 999 Sparsity of the Representation......Page 1000 Non-parametric Tests......Page 1001 Discussion......Page 1002 Introduction......Page 1005 The Method......Page 1007 Video Signal......Page 1008 One Dimensional Signals......Page 1009 Data Processing......Page 1010 Experimental Results and Interpretation......Page 1012 Discussion......Page 1014 Introduction......Page 1015 Related Work......Page 1016 The DRONE Flow Graph......Page 1018 Single-Screen Rendering (AS1)......Page 1019 Multi-screen Rendering (AS2)......Page 1021 Remote Rendering (AS3)......Page 1022 A Simple Command Language......Page 1023 Performance Measurements......Page 1024 Conclusion and Future Work......Page 1025 Introduction......Page 1027 Fast Complete Rebuild of Acceleration Data Structures......Page 1028 Intelligent Local Update......Page 1032 Complete Rebuild......Page 1033 Partial Rebuild......Page 1034 Conclusion and Future Work......Page 1036 Augmented Reality......Page 1039 The Approach......Page 1040 Feature Extraction......Page 1041 Feature Synthesis......Page 1042 Feature Matching......Page 1043 Pose Hypothesis Generation......Page 1044 Visualization......Page 1045 Evaluation......Page 1046 Conclusion......Page 1047 Introduction......Page 1049 Previous Work......Page 1050 Design Approach......Page 1051 Interaction Flexibility......Page 1052 Multiple Display Settings......Page 1053 Technical Design......Page 1054 Interaction with the Simulation......Page 1055 Discussion......Page 1056 Future Work......Page 1057 Introduction......Page 1059 Related Work......Page 1060 Object Model......Page 1062 Extension of the Object Model......Page 1063 Surfaces and Lights......Page 1064 Calling Sequence......Page 1065 Implementation......Page 1066 Applications......Page 1067 Run-Time Costs......Page 1068 Conclusion and Future Work......Page 1069 Introduction......Page 1071 Framework Components......Page 1072 Asynchronous Operation and Forced Synchronicity......Page 1073 Arbitrary Viewing Parameters......Page 1074 Compression......Page 1075 System Environment......Page 1076 Latency......Page 1077 Dynamic Image Scaling......Page 1078 Future Work......Page 1080 Introduction and Related Work......Page 1083 Definition of the Model Scalar Field......Page 1085 Cost Function Definition......Page 1087 Overall Algorithm Description......Page 1088 Selecting a Leaf......Page 1089 Expanding the Tree......Page 1090 Experimental Results......Page 1091 Conclusions......Page 1094 Introduction......Page 1095 Reformulating the Multi-label MRF as an s-t Cut......Page 1096 Data Term Penalty: Severing T-Links and Intra-Links......Page 1097 Prior Term Penalty: Severing N-Links......Page 1098 Edge Weight Approximation with Least Squares......Page 1099 LS Error and Rank Deficiency Analysis......Page 1102 Image Segmentation Results......Page 1103 Conclusions......Page 1105 Introduction......Page 1107 Reduction of SDDs to Laplacians......Page 1108 Graphs as Electric Networks – Support Basics......Page 1109 Steiner Preconditioners......Page 1110 The Combinatorial Multigrid Solver......Page 1112 From Steiner Preconditioners to Multigrid......Page 1113 Experiments......Page 1114 SDD Linear Systems......Page 1115 Discussion......Page 1117 Introduction......Page 1119 Theory: Notations and Uncertainty in Segmentation......Page 1121 Optimal Energy Functional Weights......Page 1122 Method: Segmenting Novel Images......Page 1124 Experiments......Page 1125 Discussion......Page 1127 Introduction......Page 1129 Energy-Minimizing Segmentation......Page 1131 Non-contextual Globally Optimal Weights......Page 1132 Implementation Details......Page 1133 Results and Discussion......Page 1134 Conclusion......Page 1139 Introduction......Page 1141 Non-rigid Registration Using MRFs......Page 1143 Approximated Curvature Penalty......Page 1145 Experiments......Page 1146 Conclusion......Page 1148 Back matter......Page 1150