Search results

Title: Learning 3D model from 2D in-the-wild images
Creator: Tran, Luan Quoc
Date: 2020
Collection: Electronic Theses & Dissertations
Description: Understanding 3D world is one of computer vision's fundamental problems. While a human has no difficulty understanding the 3D structure of an object upon seeing its 2D image, such a 3D inferring task remains extremely challenging for computer vision systems. To better handle the ambiguity in this inverse problem, one must rely on additional prior assumptions such as constraining faces to lie in a restricted subspace from a 3D model. Conventional 3D models are learned from a set of 3D scans or...
Show moreUnderstanding 3D world is one of computer vision's fundamental problems. While a human has no difficulty understanding the 3D structure of an object upon seeing its 2D image, such a 3D inferring task remains extremely challenging for computer vision systems. To better handle the ambiguity in this inverse problem, one must rely on additional prior assumptions such as constraining faces to lie in a restricted subspace from a 3D model. Conventional 3D models are learned from a set of 3D scans or computer-aided design (CAD) models, and represented by two sets of PCA basis functions. Due to the type and amount of training data, as well as, the linear bases, the representation power of these model can be limited. To address these problems, this thesis proposes an innovative framework to learn a nonlinear 3D model from a large collection of in-the-wild images, without collecting 3D scans. Specifically, given an input image (of a face or an object), a network encoder estimates the projection, lighting, shape and albedo parameters. Two decoders serve as the nonlinear model to map from the shape and albedo parameters to the 3D shape and albedo, respectively. With the projection parameter, lighting, 3D shape, and albedo, a novel analytically differentiable rendering layer is designed to reconstruct the original input. The entire network is end-to-end trainable with only weak supervision. We demonstrate the superior representation power of our models on different domains (face, generic objects), and their contribution to many other applications on facial analysis and monocular 3D object reconstruction.
Show less

Title: A spatio-temporal model for white matter tractography in diffusion tensor imaging
Creator: Goo, Juna
Date: 2020
Collection: Electronic Theses & Dissertations
Description: This dissertation focuses on the theoretical and applied aspects of a spatio-temporal modeling for the reconstruction of in-vivo fiber tracts in white matter when a single brain is scanned with magnetic resonance imaging (MRI) on several occasions. The objective of this research is twofold: one is how to estimate the spatial trajectory of a nerve fiber bundle at a given time point in the presence of measurement noise and the other is how to incorporate a progressive deterioration of brain...
Show moreThis dissertation focuses on the theoretical and applied aspects of a spatio-temporal modeling for the reconstruction of in-vivo fiber tracts in white matter when a single brain is scanned with magnetic resonance imaging (MRI) on several occasions. The objective of this research is twofold: one is how to estimate the spatial trajectory of a nerve fiber bundle at a given time point in the presence of measurement noise and the other is how to incorporate a progressive deterioration of brain connectivity into a hypothesis test. This dissertation leverages the spatio-temporal behavior of water diffusion in a region of the brain where the estimation of fiber trajectories is made from smoothing the time-varying diffusion tensor field via the Nadaraya-Watson type kernel regression estimator to its eigenvector field. The estimated fiber pathway takes the form of confidence ellipsoids given the estimates of mean and covariance functions. Furthermore, this dissertation proposes a hypothesis test in which the null hypothesis states that true fiber trajectories remain the same over a certain time interval. This null hypothesis indicates no substantial pathological changes of fiber pathways in that region of the brain during the observed time period. The proposed test statistic is shown to follow the limiting chi-square distribution under the null hypothesis. The power of the test is illustrated via Monte Carlo simulations. Lastly, this dissertation demonstrates the test can also be applied to a real longitudinal DTI study of a single brain repeatedly measured across time.
Show less

Title: Unconstrained 3D face reconstruction from photo collections
Creator: Roth, Joseph (Software engineer)
Date: 2016
Collection: Electronic Theses & Dissertations
Description: This thesis presents a novel approach for 3D face reconstruction from unconstrained photo collections. An unconstrained photo collection is a set of face images captured under an unknown and diverse variation of poses, expressions, and illuminations. The output of the proposed algorithm is a true 3D face surface model represented as a watertight triangulated surface with albedo data colloquially referred to as texture information. Reconstructing a 3D understanding of a face based on 2D input...
Show moreThis thesis presents a novel approach for 3D face reconstruction from unconstrained photo collections. An unconstrained photo collection is a set of face images captured under an unknown and diverse variation of poses, expressions, and illuminations. The output of the proposed algorithm is a true 3D face surface model represented as a watertight triangulated surface with albedo data colloquially referred to as texture information. Reconstructing a 3D understanding of a face based on 2D input is a long-standing computer vision problem. Traditional photometric stereo-based reconstruction techniques work on aligned 2D images and produce a 2.5D depth map reconstruction. We extend face reconstruction to work with a true 3D model, allowing us to enjoy the benefits of using images from all poses, up to and including profiles. To use a 3D model, we propose a novel normal field-based Laplace editing technique which allows us to deform a triangulated mesh to match the observed surface normals. Unlike prior work that require large photo collections, we formulate an approach to adapt to photo collections with few images of potentially poor quality. We achieve this through incorporating prior knowledge about face shape by fitting a 3D Morphable Model to form a personalized template before using a novel analysis-by-synthesis photometric stereo formulation to complete the fine face details. A structural similarity-based quality measure allows evaluation in the absence of ground truth 3D scans. Superior large-scale experimental results are reported on Internet, synthetic, and personal photo collections.
Show less

Title: Additive manufacturing for electronic systems (AMES
Creator: Mohd Ghazali, Mohd Ifwat
Date: 2019
Collection: Electronic Theses & Dissertations
Description: Over the last few decades, a significant interest towards the manufacturing of complex three-dimensional (3D) structures for conceptual models have led to an incredible amount of research and development. 3D structures have been an integral part of physical models or functional end-products and are widely adapted as a miniaturizing technique in manufacturing industries, and in particular, the electronics industry. Advancement in electronics technology has lead to the need for fabricating...
Show moreOver the last few decades, a significant interest towards the manufacturing of complex three-dimensional (3D) structures for conceptual models have led to an incredible amount of research and development. 3D structures have been an integral part of physical models or functional end-products and are widely adapted as a miniaturizing technique in manufacturing industries, and in particular, the electronics industry. Advancement in electronics technology has lead to the need for fabricating consumable electronics within a smaller lattice space. To meet the challenge of high functional density integration, Additive Manufacturing (AM) techniques by 3D printing is a promising solution for satisfying the ever-increasing demand for a higher quality product with the ability to customize based on an individual customer needs. AM techniques allows the possibility of developing low cost, multifunctional, compact, lightweight, and miniaturized electronics that can be easily integrated with conventional systems or platforms. In this dissertation, approaches towards utilizing existing AM techniques for fabricating structures that are compatible to carry electrical functionality for RF applications is proposed. The end goal is to develop processes using AM technique as an alternate manufacturing approach to achieve a fully functional electronics system. Specifically, AM holds significant potential in realizing low-loss, high-performance, and light-weight RF components such as transmission lines, waveguides, resonators, filters, and antennas. In order to realize a complete RF system by AM, multiple processes are developed. First, to establish connection for allowing electrical functionality, conductive traces must be patterned on the substrate. Two different metal patterning techniques for selectively patterning conductive traces on the 3D printed substrate is developed. Next, to realize a compact system, a smaller form factor is a necessity and this can be achieved by utilizing the flexibility in the third dimension (z-axis) in designing non-planar RF structures. A number of non-planar RF structures are demonstrated showcasing the advantages of AM in fabricating compact designs. Moreover, for fabricating efficient RF circuits, the losses associated with the printed plastics should be minimized. The currently available printing polymers have high dielectric loss and hence an alternative process that utilizes air as a substrate is developed by using a LEGO-like self-alignment procedure in which the structure is printed in multiple parts and snapped together face to face to integrate the complete structure. Furthermore, a number of active and passive components must be integrated into the printed plastic to achieve a RF system. For this purpose, three different solder-free embedding processes are developed to embedded active devices such as diodes into the 3D printed plastics. Finally, a combination of the above-mentioned processes is utilized to achieve a fully 3D printed electronics system and a potential application of such multi-functional system is demonstrated. Overall, this work demonstrates that 3D printing can be adopted in the fabrication of microwave and millimeter wave high functional density circuits and systems.
Show less

Title: Geometric and topological modeling techniques for large and complex shapes
Creator: Feng, Xin
Date: 2014
Collection: Electronic Theses & Dissertations
Description: The past few decades have witnessed the incredible advancements in modeling, digitizing and visualizing techniques for three–dimensional shapes. Those advancements led to an explosion in the number of three–dimensional models being created for design, manufacture, architecture, medical imaging, etc. At the same time, the structure, function, stability, and dynamics of proteins, subcellular structures, organelles, and multiprotein complexes have emerged as a leading interest in...
Show moreThe past few decades have witnessed the incredible advancements in modeling, digitizing and visualizing techniques for three–dimensional shapes. Those advancements led to an explosion in the number of three–dimensional models being created for design, manufacture, architecture, medical imaging, etc. At the same time, the structure, function, stability, and dynamics of proteins, subcellular structures, organelles, and multiprotein complexes have emerged as a leading interest in structural biology, another major source of large and complex geometric models. Geometric modeling not only provides visualizations of shapes for large biomolecular complexes but also fills the gap between structural information and theoretical modeling, and enables the understanding of function, stability, and dynamics.We first propose, for tessellated volumes of arbitrary topology, a compact data structure that offers constant–time–complexity incidence queries among cells of any dimensions. Our data structure is simple to implement, easy to use, and allows for arbitrary, user–defined 3–cells such as prisms and hexahedra, while remaining highly efficient in memory usage compared to previous work. We also provide the analysis on its time complexity for commonly–used incidence and adjacency queries such as vertex and edge one–rings.We then introduce a suite of computational tools for volumetric data processing, information extraction, surface mesh rendering, geometric measurement, and curvature estimation for biomolecular complexes. Particular emphasis is given to the modeling of Electron Microscopy Data Bank (EMDB) data and Protein Data Bank (PDB) data. Lagrangian and Cartesian representations are discussed for the surface presentation. Based on these representations, practical algorithms are developed for surface area and surface–enclosed volume calculation, and curvature estimation. Methods for volumetric meshing have also been presented. Because the technological development in computer science and mathematics has led to a variety of choices at each stage of the geometric modeling, we discuss the rationales in the design and selection of various algorithms. Analytical test models are designed to verify the computational accuracy and convergence of proposed algorithms. We selected six EMDB data and six PDB data to demonstrate the efficacy of the proposed algorithms in handling biomolecular surfaces and explore their capability of geometric characterization of binding targets. Thus, our toolkit offers a comprehensive protocol for the geometric modeling of proteins, subcellular structures, organelles, and multiprotein complexes.Furthermore, we present a method for computing “choking” loops—a set of surface loops that describe the narrowing of the volumes inside/outside of the surface and extend the notion of surface homology and homotopy loops. The intuition behind their definition is that a choking loop represents the region where an offset of the original surface would get pinched. Our generalized loops naturally include the usual 2g handles/tunnels computed based on the topology of the genus–g surface, but also include loops that identify chokepoints or bottlenecks, i.e., boundaries of small membranes separating the inside or outside volume of the surface into disconnected regions. Our definition is based on persistent homology theory, which gives a measure to topological structures, thus providing resilience to noise and a well–defined way to determine topological feature size.Finally, we explore the application of persistent homology theory in protein folding analysis. The extremely complex process of protein folding brings challenges for both experimental study and theoretical modeling. The persistent homology approach studies the Euler characteristics of the protein conformations during the folding process. More precisely, the persistence is measured by the variation of van der Waals radius, which leads to the change of protein 3D structures and uncovers the inter–connectivity. Our results on fullerenes demonstrate the potential of our geometric and topological approach to protein stability analysis.
Show less

Title: Example-Based Parameterization of Linear Blend Skinning for Skinning Decomposition (EP-LBS
Creator: Hopkins, Kayra M.
Date: 2017
Collection: Electronic Theses & Dissertations
Description: This thesis presents Example-based Parameterization of Linear Blend Skinning for Skinning Decomposition (EP-LBS), a unified and robust method for using example data to simplify and improve the development and parameterization of high quality 3D models for animation. Animation and three-dimensional (3D) computer graphics have quickly become a popular medium for education, entertainment and scientific simulation. In addition to film, gaming and research applications, recent advancements in...
Show moreThis thesis presents Example-based Parameterization of Linear Blend Skinning for Skinning Decomposition (EP-LBS), a unified and robust method for using example data to simplify and improve the development and parameterization of high quality 3D models for animation. Animation and three-dimensional (3D) computer graphics have quickly become a popular medium for education, entertainment and scientific simulation. In addition to film, gaming and research applications, recent advancements in augmented reality (AR) and virtual reality (VR) are driving additional demand for 3D content. However, the success of graphics in these arenas depends greatly on the efficiency of model creation and the realism of the animation or 3D image.A common method for figure animation is skeletal animation using linear blend skinning (LBS). In this method, vertices are deformed based on a weighted sum of displacements due to an embedded skeleton. This research addresses the problem that LBS animation parameter computation, including determining the rig (the skeletal structure), identifying influence bones (which bones influence which vertices), and assigning skinning weights (amounts of influence a bone has on a vertex), is a tedious process that is difficult to get right. Even the most skilled animators must work tirelessly to design an effective character model and often find themselves repeatedly correcting flaws in the parameterization. Significant research, including the use of example-data, has focused on simplifying and automating individual components of the LBS deformation process and increasing the quality of resulting animations. However, constraints on LBS animation parameters makes automated analytic computation of the values equally as challenging as traditional 3D animation methods. Skinning decomposition is one such method of computing LBS animation LBS parameters from example data. Skinning decomposition challenges include constraint adherence and computationally efficient determination of LBS parameters.The EP-LBS method presented in this thesis utilizes example data as input to a least-squares non-linear optimization process. Given a model as a set of example poses captured from scan data or manually created, EP-LBS institutes a single optimization equation that allows for simultaneous computation of all animation parameters for the model. An iterative clustering methodology is used to construct an initial parameterization estimate for this model, which is then subjected to non-linear optimization to improve the fitting to the example data. Simultaneous optimization of weights and joint transformations is complicated by a wide range of differing constraints and parameter interdependencies. To address interdependent and conflicting constraints, parameter mapping solutions are presented that map the constraints to an alternative domain more suitable for nonlinear minimization. The presented research is a comprehensive, data-driven solution for automatically determining skeletal structure, influence bones and skinning weights from a set of example data. Results are presented for a range of models that demonstrate the effectiveness of the method.
Show less

MSU Libraries

Digital Repository

You are here