實體造型在各種設計和生產項目中起著關鍵的作用。他們作為虛擬雕刻，微結構設計和快速成型等等這些常常需要處理複雜形狀和拓撲模型的應用的基礎。隨著愈來愈複雜性的工業模型和需要執行重複性的操作，我們必須要有一個有效率的處理系統。然而，因為許多基本的幾何運算是計算密集型的，一般商用的幾何內核（例如Parasolid和ACIS），尤其是對自由曲面和高度詳細複雜的模型，計算成本的要求是非常高，甚至有時不能使用。本研究的目的是開發一個以自由曲面模型為對象能夠完全在GPU 上運行的實體建模，這將大大提高建模的效率。 / 引致操作中高計算強度的一個關鍵原因是因為他們使用了邊界表示法（B-REP）。我們引入了一個新的表述 - 名為多層深度法向圖像（LDNI）來直接在GPU 上描述對象。LDNI 能夠滿足一定的幾何和拓撲的要求，同時保持簡單和易於並行實現，提供了一個比邊界表示法更可行的形式。 / 在基於B-REP 的建模核心內執行基本操作（如布爾運算，閔可夫斯基和等）往往是耗時和容易出現偶爾失敗。我們已經開發出一些GPU 加速算法，包括布爾和一般閔可夫斯基和，它們的執行速度遠遠超過傳統的方法並且沒有顯著的穩定性問題出現。這些算法是基於圖像表示法來進行的，可以讓特定的任務在每個像素上獨立運行。 / 我們還開發了在GPU 上由B-REP 轉換成LDNI 和由LDNI 轉換成B-REP 的算法。我們為了高效率而使用光栅化來獲得圖像並利用GPGPU 庫來將圖像轉換成B-REP。有了這些算法，我們的建模框架可以與其他常見的CAD 內核進行交換操作。這框架還支持模型的局部細化網格，這有助減少內存消耗，並同時保持曲面G¹ 連續性。 / 最後，本研究的目的是提供一個全面處理複雜形狀的建模問題的框架。我們展示幾個使用了以GPU 為基礎的實體建模應用例子，相比現有的以B-REP 為基礎的系統，我們的框架在效率上得到極大的改善。 / Solid modeling plays a key role in a variety of design and manufacturing activities. They serve as the foundation for applications like virtual sculpting, microstructure design and rapid prototyping, which usually deal with complex shape and topology model. The fast growing complexity of industrial models and the need to perform some operations repeatedly thereby necessitate an efficient processing system. However, since many fundamental geometric operations are compute-intensive, most commercial geometric kernels (e.g. Parasolid and ACIS), especially for freeform and highly-detailed complex models, require very high computational cost and do not work on such complex models. The purpose of this research is to exploit a solid modeler for freeform objects completely running on GPUs, which will greatly improve the efficiency of shape modeling. / A critical reason for such high computing intensity in operations is because they are performed on Boundary Representation (B-Rep). A new representation named Layered Depth Normal Images (LDNI) is introduced in this thesis to describe an object directly on GPU. LDNI is an extension of ray-representation which inherits the good properties of Boolean simplicity, localization, domain decoupling and the ease of parallel implementation, therefore provides a more feasible form than usual boundary representations. / Performing fundamental operations (such as Boolean operations, Minkowski sum, etc) are often time-consuming and prone to numerical problems when being applied on B-Rep model. This work develops several GPU-accelerated algorithms, including Boolean and general Minkowski sum. Our algorithms are carried out on image-based representation and thus are able to perform much faster than conventional approaches without arise any significant instability or robustness issues. / Algorithms have also been developed for converting B-Rep to LDNI and vice versa on GPU. Our sampling method makes use of rasterization to get the images for high efficiency and GPGPU libraries to convert them back to B-Rep. With these algorithms, the representations of models can be converted between this framework and other common CAD kernels. The framework also supports local refinement of coarse mesh, which is useful for reducing memory consumption while maintaining G¹ continuity. / Finally, the objective of this research is to provide a comprehensive framework for complex shape modeling problem. Several applications are demonstrated in using the GPU-based solid modeler, which shows great improvement in the efficiency compared with existing B-rep based systems. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Leung, Yuen Shan. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 148-159). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese. / Abstract --- p.i / Chinese Abstract --- p.iii / Acknowledgements --- p.iv / List of Figures --- p.ix / List of Tables --- p.xi / List of Publications --- p.xii / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Motivation --- p.1 / Chapter 1.2 --- Contribution --- p.3 / Chapter 1.3 --- Organization --- p.4 / Chapter 2 --- Literature Review --- p.6 / Chapter 2.1 --- Exact Solid Modeling --- p.6 / Chapter 2.1.1 --- Solid Modeling Based on B-rep --- p.6 / Chapter 2.1.1.1 --- Data Structure --- p.7 / Chapter 2.1.1.2 --- Boolean Operation --- p.10 / Chapter 2.1.1.3 --- Problems --- p.11 / Chapter 2.1.2 --- Solid Modeling based on CSG --- p.11 / Chapter 2.1.2.1 --- Operations --- p.12 / Chapter 2.1.2.2 --- Comparison --- p.13 / Chapter 2.1.3 --- Boundary Space Partition Trees --- p.14 / Chapter 2.2 --- Approximate Solid Modeling --- p.16 / Chapter 2.2.1 --- Voxel representation --- p.16 / Chapter 2.2.2 --- Oct-tree representation --- p.18 / Chapter 2.2.3 --- Distance-Field --- p.20 / Chapter 2.2.4 --- Unorganized Point-Cloud --- p.22 / Chapter 2.3 --- Computational Intensive Operations --- p.23 / Chapter 2.3.1 --- Minkowski Sum --- p.23 / Chapter 2.3.2 --- Offset --- p.26 / Chapter 2.3.3 --- Sweeping --- p.29 / Chapter 2.4 --- Volumetric Representation to B-Rep Conversion --- p.32 / Chapter 2.5 --- Curved B-rep Generation on GPU --- p.33 / Chapter 3 --- Data Structure of Solid Modeling on GPU --- p.36 / Chapter 3.1 --- Layered Depth Normal Images --- p.36 / Chapter 3.2 --- Data Structure on GPU --- p.39 / Chapter 3.2.1 --- Encode Normal Vector --- p.41 / Chapter 3.2.2 --- Compactly Stored Data Structure --- p.42 / Chapter 4 --- Model Sampling in Image Space --- p.46 / Chapter 4.1 --- Introduction --- p.46 / Chapter 4.2 --- Prior Sampling Approaches --- p.48 / Chapter 4.2.1 --- Stencil Buffer --- p.50 / Chapter 4.2.2 --- Freepipe --- p.52 / Chapter 4.3 --- GPU-Accelerated Conservative Sampling --- p.53 / Chapter 4.3.1 --- Algorithm Overview --- p.54 / Chapter 4.3.2 --- Voxel Classification --- p.57 / Chapter 4.3.3 --- Fast Bit Checking --- p.60 / Chapter 4.3.4 --- MAX Blending for Depth Value Retrieval --- p.61 / Chapter 4.3.5 --- Implementation Detail --- p.62 / Chapter 4.4 --- Results and Discussion --- p.64 / Chapter 4.5 --- Applications --- p.67 / Chapter 5 --- Boolean Operations in Images Space --- p.69 / Chapter 5.1 --- Introduction --- p.69 / Chapter 5.2 --- GPU-Accelerated Boolean Operations --- p.70 / Chapter 5.2.1 --- Boolean Operations on shader --- p.70 / Chapter 5.2.2 --- Boolean Operations on CUDA --- p.71 / Chapter 5.3 --- Results and Discussion --- p.75 / Chapter 6 --- Minkowski Sum in Image Space --- p.77 / Chapter 6.1 --- Introduction --- p.77 / Chapter 6.2 --- Algorithm Based on Exterior Boundary Extraction --- p.78 / Chapter 6.3 --- GPU-Accelerated Minkowski Sum --- p.79 / Chapter 6.3.1 --- Minkowski Sums in Image Space Without Voids --- p.79 / Chapter 6.3.1.1 --- Algorithm Overview --- p.79 / Chapter 6.3.1.2 --- Implementation Issue --- p.80 / Chapter 6.3.1.3 --- Results and Performance --- p.81 / Chapter 6.3.2 --- Minkowski Sums with Inner Voids --- p.82 / Chapter 6.3.2.1 --- Algorithm Overview --- p.83 / Chapter 6.3.2.2 --- Convex Decomposition --- p.83 / Chapter 6.3.2.3 --- Pairwise Minkowski Sum --- p.85 / Chapter 6.3.2.4 --- Super-union in Image Space --- p.86 / Chapter 6.3.2.5 --- Results and Discussion --- p.90 / Chapter 6.3.2.6 --- Application --- p.93 / Chapter 7 --- Localized Construction of Curved Surfaces --- p.96 / Chapter 7.1 --- Local Construction of Curved Polygon --- p.98 / Chapter 7.1.1 --- Basic Construction Scheme --- p.98 / Chapter 7.1.1.1 --- Gregory Patch Interpolation --- p.98 / Chapter 7.1.1.2 --- Silhouette Curve Interpolation --- p.100 / Chapter 7.1.1.3 --- Cross-Tangent Function --- p.101 / Chapter 7.1.2 --- Property Analysis --- p.103 / Chapter 7.1.3 --- Discussions --- p.106 / Chapter 7.1.4 --- Algorithm for Generating N-Sided Polygons --- p.109 / Chapter 7.2 --- Flexible Shape Control --- p.111 / Chapter 7.2.1 --- Possibility of Extending the Basic Scheme --- p.111 / Chapter 7.2.2 --- Reorienting Silhouette Curve (Extended Scheme 1) --- p.111 / Chapter 7.2.3 --- Straight Silhouette (Extended Scheme 2) --- p.112 / Chapter 7.2.4 --- Crease (Extended Scheme 3) --- p.113 / Chapter 7.2.5 --- Silhouette Adjacent to Crease (Extended Scheme 4) --- p.113 / Chapter 7.2.6 --- Smooth Patch Bulging (Extended Scheme 5) --- p.114 / Chapter 7.2.7 --- Flat Surface (Extended Scheme 6) --- p.115 / Chapter 7.3 --- Implementation Details on GPU --- p.115 / Chapter 7.3.1 --- Refinement --- p.116 / Chapter 7.3.2 --- Position Mapping --- p.118 / Chapter 7.3.3 --- Normal Computation for Shading --- p.118 / Chapter 7.4 --- Results and Comparison --- p.119 / Chapter 8 --- B-rep Reconstruction on GPU --- p.124 / Chapter 8.1 --- Introduction --- p.124 / Chapter 8.2 --- B-rep Representation by Dual-Contouring --- p.125 / Chapter 8.3 --- GPU-Accelerated Dual-Contouring --- p.126 / Chapter 8.3.1 --- Algorithm Overview --- p.126 / Chapter 8.3.2 --- Implementation Issue --- p.129 / Chapter 8.4 --- Results and Discussion --- p.130 / Chapter 9 --- Applications --- p.133 / Chapter 9.1 --- Bone Graft --- p.133 / Chapter 9.2 --- Personalized Medical Treatment --- p.137 / Chapter 10 --- Conclusion --- p.139 / Chapter 10.1 --- Summary --- p.139 / Chapter 10.2 --- Future Work --- p.141 / Chapter A --- Appendix A: Examples and Results --- p.143 / Bibliography --- p.148

Yim-Pan Chui. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2001. / Includes bibliographical references (leaves 126-132). / Abstracts in English and Chinese. / Abstract --- p.ii / Acknowledgments --- p.v / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Introduction to Collaborative Virtual Environments --- p.1 / Chapter 1.2 --- Barriers to Resource Management and Optimization --- p.3 / Chapter 1.3 --- Thesis Contributions --- p.5 / Chapter 1.4 --- Application of this Research Work --- p.6 / Chapter 1.5 --- Thesis Organization --- p.6 / Chapter 2 --- Resource Optimization - Intelligent Server Partitioning --- p.9 / Chapter 2.1 --- Introduction --- p.9 / Chapter 2.2 --- Server Partitioning --- p.13 / Chapter 2.2.1 --- Related Works --- p.15 / Chapter 2.2.2 --- Global Optimization Approaches --- p.17 / Chapter 2.3 --- Hybrid Genetic Algorithm Paradigm --- p.17 / Chapter 2.3.1 --- Drawbacks of traditional GA --- p.18 / Chapter 2.3.2 --- Problem Modeling --- p.19 / Chapter 2.3.3 --- Discussion --- p.24 / Chapter 2.4 --- Results --- p.25 / Chapter 2.5 --- Concluding Remarks --- p.28 / Chapter 3 --- Dynamic State Management - Dead Reckoning of Attitude --- p.32 / Chapter 3.1 --- Introduction to Dynamic State Management --- p.32 / Chapter 3.2 --- The Dead Reckoning Approach --- p.35 / Chapter 3.3 --- Attitude Dead Reckoning by Quaternion --- p.37 / Chapter 3.3.1 --- Modeling of the Paradigm --- p.38 / Chapter 3.3.2 --- Prediction Step --- p.39 / Chapter 3.3.3 --- Convergence Step --- p.40 / Chapter 3.3.4 --- Overall Algorithm --- p.46 / Chapter 3.4 --- Results --- p.47 / Chapter 3.5 --- Conclusion --- p.51 / Chapter 4 --- Polynomial Attitude Extrapolation --- p.52 / Chapter 4.1 --- Introduction --- p.52 / Chapter 4.2 --- Related Works on Kalman Filtering --- p.53 / Chapter 4.3 --- Historical Propagation of Quaternion --- p.54 / Chapter 4.3.1 --- Cumulative Extrapolation --- p.54 / Chapter 4.3.2 --- Method I. Vandemonde Approach --- p.55 / Chapter 4.3.3 --- Method II. Lagrangian Approach --- p.58 / Chapter 4.4 --- History-Based Attitude Management --- p.60 / Chapter 4.4.1 --- Multi-order Prediction --- p.60 / Chapter 4.4.2 --- Adaptive Attitude Convergence --- p.63 / Chapter 4.4.3 --- Overall Algorithm --- p.67 / Chapter 4.5 --- Results --- p.69 / Chapter 4.6 --- Conclusion --- p.77 / Chapter 5 --- Forward Difference Approach on State Estimation --- p.78 / Chapter 5.1 --- Introduction --- p.78 / Chapter 5.2 --- Positional Forward Differencing --- p.79 / Chapter 5.3 --- Forward Difference on Quaternion Space --- p.80 / Chapter 5.3.1 --- Attitude Forward Differencing --- p.83 / Chapter 5.3.2 --- Trajectory Blending --- p.84 / Chapter 5.4 --- State Estimation --- p.86 / Chapter 5.5 --- Computational Efficiency --- p.87 / Chapter 5.6 --- Results --- p.88 / Chapter 5.7 --- Conclusion --- p.96 / Chapter 6 --- Predictive Multibody Kinematics --- p.98 / Chapter 6.1 --- Introduction --- p.98 / Chapter 6.2 --- Dynamic Management of Multibody System --- p.100 / Chapter 6.2.1 --- Multibody Representation --- p.100 / Chapter 6.2.2 --- Paradigm Overview --- p.101 / Chapter 6.3 --- Motion Estimation by Joint Extrapolation --- p.102 / Chapter 6.3.1 --- Individual Joint Extrapolation --- p.102 / Chapter 6.3.2 --- Forward Propagation of Joint State --- p.104 / Chapter 6.3.3 --- Pose Correction --- p.107 / Chapter 6.4 --- Limitations on Predictive Articulated State Management --- p.108 / Chapter 6.5 --- Implementation and Results --- p.109 / Chapter 6.6 --- Conclusion --- p.112 / Chapter 7 --- Complete System Architecture --- p.113 / Chapter 7.1 --- Server Cluster Model --- p.113 / Chapter 7.1.1 --- Peer-Server Systems --- p.114 / Chapter 7.1.2 --- Server Hierarchies --- p.114 / Chapter 7.2 --- Multi-Level Resource Management --- p.115 / Chapter 7.3 --- Aggregation of State Updates --- p.116 / Chapter 7.4 --- Implementation Issues --- p.117 / Chapter 7.4.1 --- Medical Visualization --- p.117 / Chapter 7.4.2 --- Virtual Walkthrough Application --- p.118 / Chapter 7.5 --- Conclusion --- p.119 / Chapter 8 --- Conclusions and Future directions --- p.121 / Chapter 8.1 --- Conclusion --- p.121 / Chapter 8.2 --- Future Research Directions --- p.122 / Chapter A --- Quaternion Basis --- p.124 / Chapter A.1 --- Basic Quaternion Mathematics --- p.124 / Chapter A.2 --- The Exponential and Logarithmic Maps --- p.125 / Bibliography --- p.126

Lee Kwok Wai. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2001. / Includes bibliographical references (leaves 144-148). / Text in English; abstracts in English and Chinese. / Lee Kwok Wai. / Abstract --- p.i / 摘要 --- p.ii / Acknowledgements --- p.iii / Table of Content --- p.iv / List of Figures & Tables --- p.viii / Chapter Chapter 1 --- Introduction --- p.1 / Chapter Chapter 2 --- Motion Symmetry and its Application in Classification of Articulated Motions --- p.5 / Chapter 2.1 --- Introduction --- p.7 / Chapter 2.1.1 --- Motivation & Related Works --- p.7 / Chapter 2.1.2 --- Transformation Matrix for a Rigid Motion --- p.8 / Chapter 2.2 --- Review of Motion Estimation --- p.9 / Chapter 2.2.1 --- Motion Estimation & Motion Fields --- p.9 / Chapter 2.2.2 --- Motion Field Construction from Optical Flow --- p.10 / Chapter 2.2.3 --- Motion Field Construction from Image Matching --- p.12 / Chapter 2.3 --- Motion Symmetry --- p.13 / Chapter 2.3.1 --- Problem Definition --- p.13 / Chapter 2.3.2 --- Definitions of Transformation Symmetry & Anti- symmetry --- p.14 / Chapter 2.3.2.1 --- Translation Symmetry --- p.15 / Chapter 2.3.2.2 --- Translation Anti-symmetry --- p.16 / Chapter 2.3.2.3 --- Rotation Symmetry --- p.17 / Chapter 2.3.2.4 --- Rotation Anti-symmetry --- p.18 / Chapter 2.3.2.5 --- Scaling Symmetry … --- p.18 / Chapter 2.3.2.6 --- Scaling Anti-symmetry --- p.19 / Chapter 2.3.3 --- Transformation Quasi-symmetry & Quasi-anti- symmetry --- p.19 / Chapter 2.3.4 --- Symmetric Transform of a Transformation --- p.19 / Chapter 2.3.5 --- Symmetric Motions & Periodic Symmetric Motions --- p.20 / Chapter 2.3.6 --- Transformation Vector Fields of Symmetric Motions --- p.20 / Chapter 2.4 --- Detection of Motion Symmetry --- p.23 / Chapter 2.4.1 --- Model-based Motion Parameter Analysis --- p.24 / Chapter 2.4.2 --- Transformation Matrices Analysis --- p.25 / Chapter 2.4.3 --- Simultaneous Resultant Transformation Matrix Analysis --- p.31 / Chapter 2.4.4 --- Motion Symmetry as a Continuous Feature --- p.38 / Chapter 2.5 --- Illustrations & Results … --- p.39 / Chapter 2.5.1 --- Experiment 1: Randomly Generated Data --- p.39 / Chapter 2.5.2 --- Experiment 2: Symmetry Axis for a 3D object --- p.41 / Chapter 2.6 --- Summary & Discussion --- p.44 / Chapter 2.7 --- Appendices --- p.47 / Chapter 2.7.1 --- Appendix 1: Reflection of a Point about a Line --- p.47 / Chapter 2.7.2 --- Appendix 2: Symmetric Transform of a Transformation --- p.49 / Chapter Chapter 3 --- Motion Representation by Feedforward Neural Networks --- p.53 / Chapter 3.1 --- Introduction --- p.54 / Chapter 3.2 --- Motion Modeling in Animation --- p.57 / Chapter 3.2.1 --- Parameterized Motion Representation --- p.58 / Chapter 3.2.2 --- Problems of Motion Analysis --- p.62 / Chapter 3.3 --- Multi-value Regression by Feedforward Neural Networks --- p.66 / Chapter 3.3.1 --- Review of Multi-value Regression Methods --- p.66 / Chapter 3.3.2 --- Problem Definition --- p.68 / Chapter 3.3.3 --- Proposed Methods --- p.69 / Chapter 3.3.3.1 --- Modular Networks with Verification Module --- p.69 / Chapter (a) --- Validation by Decoding --- p.70 / Chapter (b) --- Validation by inverse mapping --- p.71 / Chapter 3.3.3.2 --- Partition Algorithm --- p.73 / Chapter 3.4 --- Illustration & Results --- p.76 / Chapter 3.4.1 --- Cylindrical Spiral Function --- p.76 / Chapter 3.4.2 --- Elongated Cylindrical Spiral Function --- p.79 / Chapter 3.4.3 --- Cylindrical Spiral Surface --- p.83 / Chapter 3.4.4 --- S-curve Data --- p.87 / Chapter 3.4.5 --- Inverse Sine function --- p.89 / Chapter 3.5 --- Motion Analysis … --- p.91 / Chapter 3.6 --- Summary & Discussion --- p.94 / Chapter Chapter 4 --- Motion Representation by Recurrent Neural Networks --- p.98 / Chapter 4.1 --- Introduction --- p.99 / Chapter 4.1.1 --- Recurrent Neural Networks (RNNs) --- p.99 / Chapter 4.1.2 --- Fully & Partially Recurrent Neural Networks --- p.101 / Chapter 4.1.3 --- Back-propagation Training Algorithm --- p.105 / Chapter 4.2 --- Sequence Encoding by Recurrent Neural Networks --- p.106 / Chapter 4.2.1 --- Random Binary Sequence --- p.107 / Chapter 4.2.2 --- Angular Positions of Clock Needles --- p.108 / Chapter 4.2.3 --- Absolute Positions of Clock Needles´ةTips --- p.109 / Chapter 4.2.4 --- Henon Time Series --- p.111 / Chapter 4.2.5 --- Ikeda Time Series … --- p.112 / Chapter 4.2.6 --- Single-Input-Single-Output (SISO) Non-linear System --- p.114 / Chapter 4.2.7 --- Circular Trajectory --- p.115 / Chapter 4.2.8 --- Number Trajectories --- p.118 / Chapter 4.3 --- Animation Generation by Recurrent Neural Networks --- p.123 / Chapter 4.3.1 --- Storage & Generation of Animations --- p.127 / Chapter 4.3.2 --- Interpolation between two Motion Segments --- p.127 / Chapter 4.4 --- Motion Analysis by Recurrent Neural Networks --- p.129 / Chapter 4.5 --- Experimental Results --- p.130 / Chapter 4.5.1 --- Network Training --- p.130 / Chapter 4.5.2 --- Motion Interpolation --- p.134 / Chapter 4.5.3 --- Motion Recognition --- p.135 / Chapter 4.6 --- Summary & Discussion --- p.138 / Chapter Chapter 5 --- Conclusion --- p.141 / References --- p.144

Thesis (M.S.V.S.)--Massachusetts Institute of Technology, Dept. of Architecture, 1981. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND ROTCH. / Includes bibliographical references (leaf 50). / Sound inherently has a spatial quality, an ability to be localized in three dimensions. This is the essence of 3-D, or spatial, sound. A system capable of recording sounds as digitized samples and playing them back in a localized fashion was developed in the course of this research. This sound system combines special hardware and interactive software to create a system more flexible and powerful than previous systems. The spatial qualities of 3-D sound contribute to man's ability to interact with sound as data. An application which capitalized on these qualities was developed, allowing the user to interact with 3-D sound in a spatial environment. This application, called the Spatial Audio Notemaker, was not unlike a bulletin board, where the paper notes were recorded messages and the bulletin board was the user's environment. Using the Spatial Audio Notemaker, exploration into the manipulation of 3-D sound and the necessary interaction (using voice and gesture) and feedback (both visual and audio) to aid in this manipulation was accomplished. / by Mark Aaron Vershel. / M.S.V.S.

Thesis (M.S.V.S.)--Massachusetts Institute of Technology, Dept. of Architecture, 1982. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND ROTCH. / Bibliography: leaves 111-116. / The human face is a highly significant visual display which we are able to remember and recognize easily despite the fact that we are exposed to thousands of faces which may be metrically very similar. caricature is a graphical coding of facial features which seeks to be more like the face than the face itself: selected information is exaggerated, noise is reduced, and the processes involved in recognition are exploited. After studying the methods of caricaturists, examining perceptual phenomena regarding individuating features, and surveying automatic and man-machine systems which represent and manipulate the face, some heuristics for caricature are defined . An algorithm is implemented to amplify the nuance of a human face in a computer- generated caricature. This is done by comparing the face to a norm and then distorting the face even further away from that norm . Issues of style, context and animation are discussed. The applications of the caricature generator in the areas of teleconferencing, games, and interactive graphic interfaces are explored. / by Susan Elise Brennan. / M.S.V.S.

This thesis describes a procedural hair generator that is able to generate hair from just a few hairs, called hair guides, which are directly modeled by a 3d artist. The procedural hair generator is a part of Stubble project -- a tool for hair modeling in Autodesk Maya. The procedural hair generator can generate hair during rendering, thus avoiding storage of hair geometry in a scene file, which makes the rendering process very efficient. Furthermore, hair can be generated interactively and displayed by OpenGL during modeling in Maya. Generated hair geometry is mainly defined by interpolation from the mentioned hair guides; however it is also influenced by many hair properties. These properties can change hair geometry using noise functions, define hair color, width and more. To determine hair root positions on a given triangular mesh I use my own mesh sampling algorithm that generates random samples on a triangular mesh according to a density defined by a 2-dimensional texture. My sampling algorithm uses an innovative way of sampling from a discrete probability distribution, which can be used in other applications than mesh sampling.

Xue, Tianfan. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (p. 52-55). / Abstracts in English and Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Previous Approaches on Face Identification --- p.3 / Chapter 1.1.1 --- Face Identification --- p.3 / Chapter 1.1.2 --- General Objects --- p.4 / Chapter 1.1.3 --- Manifold Objects --- p.7 / Chapter 1.2 --- Previous Approaches on 3D Reconstruction --- p.9 / Chapter 1.3 --- Our approach for Face Identification --- p.11 / Chapter 1.4 --- Our approach for 3D Reconstruction --- p.13 / Chapter 2 --- Face Detection --- p.14 / Chapter 2.1 --- GAFI and its Face Identification Results --- p.15 / Chapter 2.2 --- Our Face Identification Approach --- p.17 / Chapter 2.2.1 --- Real Face Detection --- p.18 / Chapter 2.2.2 --- The Weak Face Adjacency Theorem --- p.20 / Chapter 2.2.3 --- Searching for Type 1 Lost Faces --- p.22 / Chapter 2.2.4 --- Searching for Type 2 Lost Faces --- p.23 / Chapter 2.3 --- Experimental Results --- p.25 / Chapter 3 3 --- D Reconstruction --- p.30 / Chapter 3.1 --- Assumption and Terminology --- p.30 / Chapter 3.2 --- Finding Cuts from a Line Drawing --- p.34 / Chapter 3.2.1 --- Propositions for Finding Cuts --- p.34 / Chapter 3.2.2 --- Searching for Good Cuts --- p.35 / Chapter 3.3 --- Separation of a Line Drawing from Cuts --- p.38 / Chapter 3.4 3 --- D Reconstruction from a Line Drawing --- p.45 / Chapter 3.5 --- Experiments --- p.45 / Chapter 4 --- Conclusion --- p.50

Xie, Yongming. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2008. / Includes bibliographical references (leaves 77-85). / Abstracts in English and Chinese. / Abstract --- p.i / Acknowledgement --- p.ii / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Isosurfaces --- p.1 / Chapter 1.2 --- Graphics Processing Unit --- p.2 / Chapter 1.3 --- Objective --- p.3 / Chapter 1.4 --- Contribution --- p.3 / Chapter 1.5 --- Thesis Organization --- p.4 / Chapter 2 --- Marching Cubes --- p.5 / Chapter 2.1 --- Introduction --- p.5 / Chapter 2.2 --- Marching Cubes Algorithm --- p.7 / Chapter 2.3 --- Triangulated Cube Configuration Table --- p.12 / Chapter 2.4 --- Summary --- p.16 / Chapter 3 --- Graphics Processing Unit --- p.18 / Chapter 3.1 --- Introduction --- p.18 / Chapter 3.2 --- History of Graphics Processing Unit --- p.19 / Chapter 3.2.1 --- First Generation GPU --- p.20 / Chapter 3.2.2 --- Second Generation GPU --- p.20 / Chapter 3.2.3 --- Third Generation GPU --- p.20 / Chapter 3.2.4 --- Fourth Generation GPU --- p.21 / Chapter 3.3 --- The Graphics Pipelining --- p.21 / Chapter 3.3.1 --- Standard Graphics Pipeline --- p.21 / Chapter 3.3.2 --- Programmable Graphics Pipeline --- p.23 / Chapter 3.3.3 --- Vertex Processors --- p.25 / Chapter 3.3.4 --- Fragment Processors --- p.26 / Chapter 3.3.5 --- Frame Buffer Operations --- p.28 / Chapter 3.4 --- GPU CPU Analogy --- p.31 / Chapter 3.4.1 --- Memory Architecture --- p.31 / Chapter 3.4.2 --- Processing Model --- p.32 / Chapter 3.4.3 --- Limitation of GPU --- p.33 / Chapter 3.4.4 --- Input and Output --- p.34 / Chapter 3.4.5 --- Data Readback --- p.34 / Chapter 3.4.6 --- FramebufFer --- p.34 / Chapter 3.5 --- Summary --- p.35 / Chapter 4 --- Volume Rendering --- p.37 / Chapter 4.1 --- Introduction --- p.37 / Chapter 4.2 --- History of Volume Rendering --- p.38 / Chapter 4.3 --- Hardware Accelerated Volume Rendering --- p.40 / Chapter 4.3.1 --- Hardware Acceleration Volume Rendering Methods --- p.41 / Chapter 4.3.2 --- Proxy Geometry --- p.42 / Chapter 4.3.3 --- Object-Aligned Slicing --- p.43 / Chapter 4.3.4 --- View-Aligned Slicing --- p.45 / Chapter 4.4 --- Summary --- p.48 / Chapter 5 --- GPU-Friendly Marching Cubes --- p.49 / Chapter 5.1 --- Introduction --- p.49 / Chapter 5.2 --- Previous Work --- p.50 / Chapter 5.3 --- Traditional Method --- p.52 / Chapter 5.3.1 --- Scalar Volume Data --- p.53 / Chapter 5.3.2 --- Isosurface Extraction --- p.53 / Chapter 5.3.3 --- Flow Chart --- p.54 / Chapter 5.3.4 --- Transparent Isosurfaces --- p.56 / Chapter 5.4 --- Our Method --- p.56 / Chapter 5.4.1 --- Cell Selection --- p.59 / Chapter 5.4.2 --- Vertex Labeling --- p.61 / Chapter 5.4.3 --- Cell Indexing --- p.62 / Chapter 5.4.4 --- Interpolation --- p.65 / Chapter 5.5 --- Rendering Translucent Isosurfaces --- p.67 / Chapter 5.6 --- Implementation and Results --- p.69 / Chapter 5.7 --- Summary --- p.74 / Chapter 6 --- Conclusion --- p.76 / Bibliography --- p.77

Orientador: João Paulo Papa / Banca: André Carlos Ponce de Leon Ferreira de Carvalho / Banca: Aparecido Nilceu Marana / Resumo: Arquiteturas de aprendizado em profundidade têm sido amplamente estudadas nos últimos anos, principalmente pelo seu alto poder discriminativo em muitos problemas considerados essenciais na área de visão computacional. Entretanto, um problema destes modelos diz res- peito ao grande número de parâmetros a serem ajustados, que podem chegar a milhares. Um outro ponto crítico está relacionado à necessidade de grandes bases de dados para treinar essas técnicas de aprendizado em profundidade, bem como a sua alta propensão ao chamado super-ajuste dos dados. Recentemente, a simplista ideia de desconectar neurônios ou conexões de uma rede, técnicas denominadas de Dropout e Dropconnect, respectivamente, tem se demonstrado muito eficazes e primordiais ao processo de aprendizado, embora ainda necessitem de uma escolha adequada de parâmetros. O presente projeto pretende identificar possíveis soluções para o problema mencionado por meio de técnicas de otimização meta-heurística, objetivando encontrar o número adequado do limiar de desligamento dos neurônios e conexões. Diferentes abordagens de aprendizado em profundidade, tais como, Máquinas de Boltzmann Restritas, Máquinas de Boltzmann em Profundidade, Redes de Crença em Profundidade, Redes Neurais Convolucionais; e diferentes meta-heurísticas, tais como, Algoritmo do Morcego, Algoritmo do Vagalume, Busca do Cuckoo, Otimização por Enxame de Partículas, foram utilizadas a fim de tentar solucionar este problema. Os resultados apresentados... / Abstract: Deep learning architectures have been extensively studied in the last years, mainly due to their discriminative power in many crucial problems in computer vision. However, one problem related to these models concerns with their number of parameters, which can easily reach thousands of hundreds. Another drawback is related to the need for large datasets for train- ing purposes, as well as their high probability of overfitting, mainly because of their complex architecture. Recently, a naïve idea of disconnecting neurones or connections from a network, known as Dropout or Dropconnect, respectively, has shown to be a promising solution to this problem. Nevertheless, it still requires an adequate parameter setting. This project aims to iden- tify possible solutions to the depicted problem by means of meta-heuristic optimization, trying to find the most suitable drop rate. Several machine learning approaches, such as, Restricted Boltzmann Machines, Deep Boltzmann Machines, Deep Belief Networks, Convolutional Neural Networks and several meta-heuristic techniques, such as, Particle Swarm Optimization, Bat Algorithm, Firefly Algorithm, Cuckoo Search, were employed in the context. The presented results show a possible trend in using meta-heuristic optimization to find suitable parameters in a wide range of applications, helping the learning process and improving the network's architecture / Mestre

Typescript (photocopy). / Digitized by Kansas Correctional Industries