Deep points consolidation

Wu, Shihao; Huang, Hui; Gong, Minglun; Zwicker, Matthias; Cohen–Or, Daniel

doi:10.1145/2816795.2818073

Cited by 121 publications

(74 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Huang et al [15] further developed a progressive method called EAR for edge-aware resampling of point sets. Later, Wu et al [33] proposed a consolidation method to fill large holes and complete missing regions by introducing a new deep point representation. Overall, these methods are not data-driven; they heavily rely on priors, e.g., the assumption of smooth surface, normal estimation, etc.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

PU-GAN: A Point Cloud Upsampling Adversarial Network

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

Self Cite

382

345

View full text Add to dashboard Cite

Point clouds acquired from range scans are often sparse, noisy, and non-uniform. This paper presents a new point cloud upsampling network called PU-GAN 1 , which is formulated based on a generative adversarial network (GAN), to learn a rich variety of point distributions from the latent space and upsample points over patches on object surfaces. To realize a working GAN network, we construct an up-down-up expansion unit in the generator for upsampling point features with error feedback and self-correction, and formulate a self-attention unit to enhance the feature integration. Further, we design a compound loss with adversarial, uniform and reconstruction terms, to encourage the discriminator to learn more latent patterns and enhance the output point distribution uniformity. Qualitative and quantitative evaluations demonstrate the quality of our results over the state-of-the-arts in terms of distribution uniformity, proximity-to-surface, and 3D reconstruction quality.

show abstract

Section: Related Workmentioning

confidence: 99%

“…Early methods [3,21,14,15,33] for the problem are optimization-based, where various shape priors are used to constrain the point cloud generation. Recently, deep neural networks brought the promise of data-driven approaches to the problem.…”

Section: Introductionmentioning

confidence: 99%

PU-GAN: A Point Cloud Upsampling Adversarial Network

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

Self Cite

382

345

View full text Add to dashboard Cite

show abstract

“…However, these approaches can easily fail in the presence of noisy normal estimates or high complexity shapes. As a result, they have has been extended significantly in recent years both through more robust estimates [HLZ*09], which have also been adapted to handle large missing regions through point skeleton estimation [WHG*15], and global techniques based on signed distance computations [MDGD*10], among others. Nevertheless, reliably estimating oriented normals remains challenging especially across different noise levels and shape structures.…”

Section: Related Workmentioning

confidence: 99%

PCPNet Learning Local Shape Properties from Raw Point Clouds

Guerrero

Kleiman

Ovsjanikov

et al. 2018

Computer Graphics Forum

288

251

View full text Add to dashboard Cite

In this paper, we propose PCPNET, a deep‐learning based approach for estimating local 3D shape properties in point clouds. In contrast to the majority of prior techniques that concentrate on global or mid‐level attributes, e.g., for shape classification or semantic labeling, we suggest a patch‐based learning method, in which a series of local patches at multiple scales around each point is encoded in a structured manner. Our approach is especially well‐adapted for estimating local shape properties such as normals (both unoriented and oriented) and curvature from raw point clouds in the presence of strong noise and multi‐scale features. Our main contributions include both a novel multi‐scale variant of the recently proposed PointNet architecture with emphasis on local shape information, and a series of novel applications in which we demonstrate how learning from training data arising from well‐structured triangle meshes, and applying the trained model to noisy point clouds can produce superior results compared to specialized state‐of‐the‐art techniques. Finally, we demonstrate the utility of our approach in the context of shape reconstruction, by showing how it can be used to extract normal orientation information from point clouds.

show abstract

“…Relevant work on geometry: Mousa et al [MCAG07] employ a kd‐tree in a similar way as we do, however theirs is aimed at efficient spherical harmonics representations of 3D objects rather than meshes. The process of deep points consolidation proposed by Wu et al [WHG*15] associates each surface point with a deep point which form a meso‐skeleton. The reconstructed surface is an improvement to naive solutions but involves a Poisson solver nevertheless.…”

Section: Related Workmentioning

confidence: 99%

Hierarchical Quad Meshing of 3D Scanned Surfaces

Bukenberger

Lensch

2018

Computer Graphics Forum

View full text Add to dashboard Cite

In this paper we present a novel method to reconstruct watertight quad meshes on scanned 3D geometry. There exist many different approaches to acquire 3D information from real world objects and sceneries. Resulting point clouds depict scanned surfaces as sparse sets of positional information. A common downside is the lack of normals, connectivity or topological adjacency data which makes it difficult to actually recover a meaningful surface. The concept described in this paper is designed to reconstruct a surface mesh despite all this missing information. Even when facing varying sample density, our algorithm is still guaranteed to produce watertight manifold meshes featuring quad faces only. The topology can be set‐up to follow superimposed regular structures or align naturally to the point cloud's shape. Our proposed approach is based on an initial divide and conquer subsampling procedure: Surface samples are clustered in meaningful neighborhoods as leafs of a kd‐tree. A representative sample of the surface neighborhood is determined for each leaf using a spherical surface approximation. The hierarchical structure of the binary tree is utilized to construct a basic set of loose tiles and to interconnect them. As a final step, missing parts of the now coherent tile structure are filled up with an incremental algorithm for locally optimal gap closure. Disfigured or concave faces in the resulting mesh can be removed with a constrained smoothing operator.

show abstract

Deep points consolidation

Cited by 121 publications

References 51 publications

PU-GAN: A Point Cloud Upsampling Adversarial Network

PU-GAN: A Point Cloud Upsampling Adversarial Network

PCPNet Learning Local Shape Properties from Raw Point Clouds

Hierarchical Quad Meshing of 3D Scanned Surfaces

Contact Info

Product

Resources

About