Search

Abstract Parsing of human images is a fundamental task for determining semantic parts such as the face, arms, and legs, as well as a hat or a dress. Recent deep-learning-based methods have achieved significant improvements, but collecting training datasets with pixel-wise annotations is labor-intensive. In this paper, we propose two solutions to cope with limited datasets. Firstly, to handle various poses, we incorporate a pose estimation network into an end-to-end human-image parsing network, in order to transfer common features across the domains. The pose estimation network can be trained using rich datasets and can feed valuable features to the human-image parsing network. Secondly, to handle complicated backgrounds, we increase the variation in image backgrounds automatically by replacing the original backgrounds of human images with others obtained from large-scale scenery image datasets. Individually, each solution is versatile and beneficial to human-image parsing, while their combination yields further improvement. We demonstrate the effectiveness of our approach through comparisons and various applications such as garment recoloring, garment texture transfer, and visualization for fashion analysis.

In this paper, we compare the performance of three different folding models when they are applied to three different map folding settings. Precisely, the three folding models include the simple folding model, the simple folding–unfolding model, and the general folding model. The different map folding settings are discussed by comparing different folded states, i.e., different overlapping orders on the set of the squares of 1 × n maps, the squares of m × n maps, and the squares lying on the boundary of m × n maps. These folding models are abstracts of manual works and robotics. We clarify the relationship between their reachable final folded states under different settings and give proof of all the inclusion relationships between every two of these sets. In addition, there are nine distinct problems with the three folding models applied to three folding settings. We give the optimal linear time solutions to all the unsolved ones: the valid total overlapping order problems of 1 × n maps, m × n maps, as well as the valid boundary overlapping order problems of m × n maps with the three different folding models. Our work gives the conclusion of the research field where the folding models and the overlapping orders of map folding are concerned.

Three-dimensional (3D) origami, which can generate a structure through folding a crease pattern on a flat sheet of paper, has received considerable attention in art, mathematics, and engineering. With consideration of symmetry, the user can efficiently generate a rational crease pattern and make the fabricated shape stable. In this paper, we focus on a category of axisymmetric origami consisting of triangular facets and edit the origami in 3D space for expanding its variations. However, it is difficult to retain the developability, which requires the sum of the angles around each interior vertex needing to equal 360 degrees, for designing origami. Intersections occur between crease lines when such a value is larger than 360 degrees. On the other hand, blank spaces (unfolded areas) emerge in the crease pattern when the value is less than 360 degrees. The former case is difficult to generate a realizable shape due to the crease lines are intersected with each other. For the latter case, however, blank spaces can be filled with crease lines and become a part of the origami through tucking. Here, we propose a computational method to add flaps or tucks on the 3D shape, which contains non-developable interior vertices, for achieving the resulting origami. Finally, on the application side, we describe a load-bearing experiment on a stool shape-like origami to demonstrate the potential usage.

In the study of origami, various parametric methods have been proposed to design crease patterns under geometric conditions for flat-folding. Each design method contributes to finding a desirable crease pattern, e.g., one with superior engineering properties, by manipulating parameters. On the other hand, to continuously deform other crease patterns, it is necessary to recreate it once with such a parametric method; however, this inverse problem is less studied. This paper is basic research to solve this problem and to allow parametric deformation of flat-foldable crease patterns. Given crease patterns are interpreted as networks consisting of twistfolding patterns that can be generated by an existing parametric method named twist-based design method. Then, by manipulating the parameters, the crease pattern is deformed. Importantly, any flat-foldable crease pattern having no crease line connecting two points on the boundary can be targeted, and it is locally guaranteed that deformed crease patterns have non-intersecting crease lines and are flat-foldable. The proposed method contributes to increased origami variations by deformations of existing crease patterns. [ABSTRACT FROM AUTHOR]

Origami tessellations have shown the potential to be utilized for approximating curved target surfaces by folding a set of elemental units. As those units are in partially folded states, the origami approximation captures the outline of the target while demonstrating a corrugated surface. In this paper, we focus on the cylindrical surface and propose a method for constructing its discrete version via square or rectangular units with crease patterns inspired by thewaterbomb tessellation.We unify the unit size through optimization to realize cost-efficient constructions. Three-dimensional curved structures are deployed by folding collinear creases between adjacent rows of units and the other creases in each row are remained unfolded. The deployed structure can still be folded to a compact state by folding all creases. To flexibly approximate the outline of the target surface, we analyse relationships between the width and height of the rectangular unit and utilize variations of the waterbomb tessellation to control the region of the origami structure when fully folded. The proposed method provides a novel solution to the so-called inverse-origami-design problem for generating foldable cylindrical surfaces using unfolded waterbomb origami units. [ABSTRACT FROM AUTHOR]

Origami tessellations are geometric pieces folded from a single sheet of paper with flatly overlapped facets. Most existing origami tessellations are constructed by first marking a grid of crease lines on the paper and then arranging repeating patterns along the grid. However, this design method is limited because it cannot design origami tessellations with patterns that cannot be represented on a grid, such as a regular pentagon. This paper proposes a new construction method for origami tessellations that solves this problem and enriches these varieties. In the proposed method, a boundary of an origami tessellation is determined first, and then patterns called triangle twist fold patterns are placed inside the boundary. A similar approach is known as a hole problem, although in this paper, the problem is redefined and discussed in a form suitable for origami tessellations. As a result, a grid-independent construction method was proposed, and new origami tessellations were obtained by using software that implements the method. [ABSTRACT FROM AUTHOR]