Computer Science

We present study results from two experiments to empirically validate that separable bivariate pairs for univariate representations of large-magnitude-range vectors are more efficient than integral pairs. The first experiment with 20 participants compared: one integral pair, three separable pairs, and one redundant pair, which is a mix of the integral and separable features. Participants performed three local tasks requiring reading numerical values, estimating ratio, and comparing two points. The second 18-participant study compared three separable pairs using three global tasks when participants must look at the entire field to get an answer: find a specific target in 20 seconds, find the maximum magnitude in 20 seconds, and estimate the total number of vector exponents within 2 seconds. Our results also reveal the following: separable pairs led to the most accurate answers and the shortest task execution time, while integral dimensions were among the least accurate; it achieved high performance only when a pop-out separable feature (here color) was added. To reconcile this finding with the existing literature, our second experiment suggests that the higher the separability, the higher the accuracy; the reason is probably that the emergent global scene created by the separable pairs reduces the subsequent search space.

Querying and interacting with models of massive material micro-structure requires localized on-demand generation of the micro-structure since the full-scale storing and retrieving is cost prohibitive. When the micro-structure is efficiently represented as the image of a canonical structure under a non-linear space deformation to allow it to conform to curved shape, the additional challenge is to relate the query of the mapped micro-structure back to its canonical structure. This paper presents an efficient algorithm to pull back a mapped micro-structure to a partition of the canonical domain structure into boxes and only activates boxes whose image is likely intersected by a plane. The active boxes are organized into a forest whose trees are traversed depth first to generate mapped micro-structure only of the active boxes. The traversal supports, for example, 3D print slice generation in additive manufacturing.

Point cloud super-resolution is a fundamental problem for 3D reconstruction and 3D data understanding. It takes a low-resolution (LR) point cloud as input and generates a high-resolution (HR) point cloud with rich details. In this paper, we present a data-driven method for point cloud super-resolution based on graph networks and adversarial losses. The key idea of the proposed network is to exploit the local similarity of point cloud and the analogy between LR input and HR output. For the former, we design a deep network with graph convolution. For the latter, we propose to add residual connections into graph convolution and introduce a skip connection between input and output. The proposed network is trained with a novel loss function, which combines Chamfer Distance (CD) and graph adversarial loss. Such a loss function captures the characteristics of HR point cloud automatically without manual design. We conduct a series of experiments to evaluate our method and validate the superiority over other methods. Results show that the proposed method achieves the state-of-the-art performance and have a good generalization ability to unseen data.

In this paper, we introduce Point2Mesh, a technique for reconstructing a surface mesh from an input point cloud. Instead of explicitly specifying a prior that encodes the expected shape properties, the prior is defined automatically using the input point cloud, which we refer to as a self-prior. The self-prior encapsulates reoccurring geometric repetitions from a single shape within the weights of a deep neural network. We optimize the network weights to deform an initial mesh to shrink-wrap a single input point cloud. This explicitly considers the entire reconstructed shape, since shared local kernels are calculated to fit the overall object. The convolutional kernels are optimized globally across the entire shape, which inherently encourages local-scale geometric self-similarity across the shape surface. We show that shrink-wrapping a point cloud with a self-prior converges to a desirable solution; compared to a prescribed smoothness prior, which often becomes trapped in undesirable local minima. While the performance of traditional reconstruction approaches degrades in non-ideal conditions that are often present in real world scanning, i.e., unoriented normals, noise and missing (low density) parts, Point2Mesh is robust to non-ideal conditions. We demonstrate the performance of Point2Mesh on a large variety of shapes with varying complexity.

Point clouds obtained with 3D scanners or by image-based reconstruction techniques are often corrupted with significant amount of noise and outliers. Traditional methods for point cloud denoising largely rely on local surface fitting (e.g., jets or MLS surfaces), local or non-local averaging, or on statistical assumptions about the underlying noise model. In contrast, we develop a simple data-driven method for removing outliers and reducing noise in unordered point clouds. We base our approach on a deep learning architecture adapted from PCPNet, which was recently proposed for estimating local 3D shape properties in point clouds. Our method first classifies and discards outlier samples, and then estimates correction vectors that project noisy points onto the original clean surfaces. The approach is efficient and robust to varying amounts of noise and outliers, while being able to handle large densely-sampled point clouds. In our extensive evaluation, both on synthesic and real data, we show an increased robustness to strong noise levels compared to various state-of-the-art methods, enabling accurate surface reconstruction from extremely noisy real data obtained by range scans. Finally, the simplicity and universality of our approach makes it very easy to integrate in any existing geometry processing pipeline.

Point cloud filtering is a fundamental problem in geometry modeling and processing. Despite of significant advancement in recent years, the existing methods still suffer from two issues: 1) they are either designed without preserving sharp features or less robust in feature preservation; and 2) they usually have many parameters and require tedious parameter tuning. In this paper, we propose a novel deep learning approach that automatically and robustly filters point clouds by removing noise and preserving their sharp features. Our point-wise learning architecture consists of an encoder and a decoder. The encoder directly takes points (a point and its neighbors) as input, and learns a latent representation vector which goes through the decoder to relate the ground-truth position with a displacement vector. The trained neural network can automatically generate a set of clean points from a noisy input. Extensive experiments show that our approach outperforms the state-of-the-art deep learning techniques in terms of both visual quality and quantitative error metrics. The source code and dataset can be found at this https URL.

Stroking and filling are the two basic rendering operations on paths in vector graphics. The theory of filling a path is well-understood in terms of contour integrals and winding numbers, but when path rendering standards specify stroking, they resort to the analogy of painting pixels with a brush that traces the outline of the path. This means important standards such as PDF, SVG, and PostScript lack a rigorous way to say what samples are inside or outside a stroked path. Our work fills this gap with a principled theory of stroking. Guided by our theory, we develop a novel polar stroking method to render stroked paths robustly with an intuitive way to bound the tessellation error without needing recursion. Because polar stroking guarantees small uniform steps in tangent angle, it provides an efficient way to accumulate arc length along a path for texturing or dashing. While this paper focuses on developing the theory of our polar stroking method, we have successfully implemented our methods on modern programmable GPUs.

Polygon meshes are an efficient representation of 3D geometry, and are of central importance in computer graphics, robotics and games development. Existing learning-based approaches have avoided the challenges of working with 3D meshes, instead using alternative object representations that are more compatible with neural architectures and training approaches. We present an approach which models the mesh directly, predicting mesh vertices and faces sequentially using a Transformer-based architecture. Our model can condition on a range of inputs, including object classes, voxels, and images, and because the model is probabilistic it can produce samples that capture uncertainty in ambiguous scenarios. We show that the model is capable of producing high-quality, usable meshes, and establish log-likelihood benchmarks for the mesh-modelling task. We also evaluate the conditional models on surface reconstruction metrics against alternative methods, and demonstrate competitive performance despite not training directly on this task.

The design of functional seating furniture is a complicated process which often requires extensive manual design effort and empirical evaluation. We propose a computational design framework for pose-driven automated generation of body-supports which are optimized for comfort of sitting. Given a human body in a specified pose as input, our method computes an approximate pressure distribution that also takes frictional forces and body torques into consideration which serves as an objective measure of comfort. Utilizing this information to find out where the body needs to be supported in order to maintain comfort of sitting, our algorithm can create a supporting mesh suited for a person in that specific pose. This is done in an automated fitting process, using a template model capable of supporting a large variety of sitting poses. The results can be used directly or can be considered as a starting point for further interactive design.

In this paper, we present a fully-customized AR display design that considers the user's prescription, interpupillary distance, and taste of fashion. A free-form image combiner embedded inside the prescription lens provides augmented images onto the vision-corrected real world. We establish a prescription-embedded AR display optical design method as well as the customization method for individual users. Our design can cover myopia, hyperopia, astigmatism, and presbyopia, and allows the eye-contact interaction with privacy protection. A 169 g dynamic prototype showed a 40 ∘ × 20 ∘ virtual image with a 23 cpd resolution at center field and 6 mm × 4 mm eye box, with the vision-correction and varifocal (0.5-3 m ) capability.