Geometric deep learning has sparked a rising interest in computer graphics to perform shape understanding tasks, such as classification and semantic segmentation. When the input is polygonal surface, one suffer from irregular mesh structure. Motivated by geometric spectral theory, we introduce <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Laplacian2Mesh</i> , novel flexible convolutional neural network (CNN) framework for coping with...

Identifying points of interest (POIs) on the surface 3D shapes is a significant challenge in geometric processing research. The complex connection between POIs and their descriptors, combined with small percentage shape, makes detecting any given shape highly challenging task. Existing methods directly detect from entire resulting low efficiency accuracy. Therefore, we propose novel multi-modal POI detection method using coarse-to-fine approach, key idea reducing data complexity enabling...

3D shape segmentation is a fundamental and crucial task in the field of image processing analysis. To segment shapes using data-driven methods, fully labeled dataset usually required. However, obtaining such can be daunting task, as manual face-level labeling both time-consuming labor-intensive. In this paper, we present semi-supervised framework for that uses small, set shapes, well weakly with sparse scribble labels. Our first employs an auxiliary network to generate initial labels...

Shape description and feature detection are fundamental problems in computer graphics geometric modeling. Among many existing techniques, those based on geodesic distance have proven effective providing intrinsic discriminative shape descriptors. In this article we introduce a new function for three-dimensional (3D) use it detection. Specifically, the girth (IGF) defined 2D closed surface. For point p surface, value of IGF at is length shortest nonzero path starting ending . The invariant...

Face enhancement aims to improve low-quality face images a higher-quality level. However, in real-world nighttime scenes, complex degradation factors often affect these images, making it challenging preserve important facial details. Existing image algorithms typically focus on independently conducting super-resolution and brightness enhancement, assuming fixed level based simulated training datasets. Nonetheless, real scenes involve processes, where dynamically variably manifest. Therefore,...

Detecting points of interest on 3D shapes is a fundamental research problem in geometry processing. Due to the complicated relationship between and their geometric features, detecting any given shape remains challenging. lack training data, previous data-driven methods for mainly focus utilizing hand-crafted features predict probabilities each point being POI, which greatly limits detection performance. In this paper, we propose novel algorithm by using projective neural networks. Our method...

This paper presents an improved algebraic pruning method for point projection Bezier curves. It first turns the into a root finding problem, and provides simple but easily overlooked to avoid invalid roots which is obviously irrelative closest point. The continued fraction its expansion are utilized strengthen robustness. Since NURBS curves can be turned form, new also works with Examples presented illustrate efficiency robustness of method.

This paper presents an improved algebraic pruning method for point projection Bezier curves. It first turns the into a root finding problem, and provides simple but easily overlooked to avoid invalid roots which is obviously irrelative closest point. The continued fraction its expansion are utilized strengthen robustness. Since NURBS curves can be turned form, new also works with Examples presented illustrate efficiency robustness of method.

The shape diameter function (SDF) is a scalar defined on closed manifold surface, measuring the neighborhood of object at each point. Due to its pose oblivious property, SDF widely used in analysis, segmentation and retrieval. However, computing computationally expensive since one has place an inverted cone point then average penetration distances for number rays inside cone. Furthermore, diameters are highly sensitive local geometric features as well normal vectors, hence diminishing their...

Detecting points of interest is a fundamental problem in 3D shape analysis and can be beneficial to various tasks multimedia processing. Traditional learning-based detection methods usually rely on each vertex’s geometric features discriminate from other vertices. Observing that are related not only themselves but also the surrounding vertices, we propose novel context-aware algorithm by adopting spatial attention mechanism this article. By designing context module, our approach presents...

In order to extract foreground objects from dense multi-view images precisely and automatically, a level set evolution segmentation method without user interaction is proposed. Firstly, we make statistical analysis of the straight lines in Epipolar Plane Image (EPI) EPI-lines corresponding object are converted into original image space get an initial contour. Then, design contour growing algorithm shorten gaps between broken edge segments morphological operation utilized obtain closed...

In this paper, we present a fast method for segmentation of triangular mesh surfaces into simple patches, where each patch is an approximation part spherical surface. Given surface, all triangles it cluster to user-specified number patches by energy minimization. All result meshes are as possible with our algorithm. Experimental results show that algorithm efficient and robust.

In order to segment interest objects from multi-view images accurately, an Epipolar Plane Image (EPI) analysis based object segmentation algorithm is proposed. Firstly, the straight lines in EPI are extracted by improved Hough transform. Then, according depth of object, corresponding converted into original image space get initial contour. Following that, we design a boundary growing shorten gaps between broken edge segments. Finally, contour linking applied obtain closed curve. Experimental...

Compact E-cash is the first scheme which can withdraw 2 l coins within (1) operations and then store them in () bits.Because of its high efficiency, a lot research has been carried out on basis, but no previous pay attention to privacy payees some cases, have same requirement as payers.We propose double-blind compact scheme, means that both payer payee keep anonymous while spending.In our bank cannot determine whether two different transactions are one connect with related him, this way,...