Self-Sampling for Neural Point Cloud Consolidation

Metzer, Gal; Hanocka, Rana; Giryes, Raja; Cohen–Or, Daniel

doi:10.1145/3470645

Cited by 15 publications

(19 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To demonstrate the effectiveness our PCS-Net, we now compare it against various competitors, including a traditional resampling method (EAR [14]) and the state-of-the-art downsampling networks (i.e., S-Net [8], SampleNet [18], Self-sampling [25], CP-Net [26]). Most of the downsampling methods do not focus on the surface reconstruction task [8,18,26].…”

Section: Comparisonsmentioning

confidence: 99%

See 1 more Smart Citation

Deep Point Cloud Simplification for High-quality Surface Reconstruction

Li¹,

Guo²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

The growing size of point clouds enlarges consumptions of storage, transmission, and computation of 3D scenes. Raw data is redundant, noisy, and non-uniform. Therefore, simplifying point clouds for achieving compact, clean, and uniform points is becoming increasingly important for 3D vision and graphics tasks. Previous learning based methods aim to generate fewer points for scene understanding, regardless of the quality of surface reconstruction, leading to results with low reconstruction accuracy and bad point distribution. In this paper, we propose a novel point cloud simplification network (PCS-Net) dedicated to high-quality surface mesh reconstruction while maintaining geometric fidelity. We first learn a sampling matrix in a feature-aware simplification module to reduce the number of points. Then we propose a novel double-scale resampling module to refine the positions of the sampled points, to achieve a uniform distribution. To further retain important shape features, an adaptive sampling strategy with a novel saliency loss is designed. With our PCS-Net, the input non-uniform and noisy point cloud can be simplified in a feature-aware manner, i.e., points near salient features are consolidated but still with uniform distribution locally. Experiments demonstrate the effectiveness of our method and show that we outperform previous simplification or reconstruction-oriented upsampling methods.

show abstract

Section: Comparisonsmentioning

confidence: 99%

“…To be specific, SampleNet [18] is skilled in maintaining semantic information but not geometry. Self-sampling [25] can consolidate points near edges but produce noise (e.g. the decoration of 'Happy buddha') and large non-uniform regions (e.g.…”

Section: Comparisonsmentioning

confidence: 99%

Deep Point Cloud Simplification for High-quality Surface Reconstruction

Li¹,

Guo²,

Yang³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Liu et al [21] proposed a progressive geometry upsampling method with adversarial learning. Metzer et al [28] proposed a point cloud consolidation method (Self-Net) with self-supervision learning [29], [30], which uses the curvature and density of the point cloud to generate sharp edge points or points in sparse regions. Although both our method and Self-Net use HF points, the optimization objectives are different.…”

Section: Related Workmentioning

confidence: 99%

PUFA-GAN: A Frequency-Aware Generative Adversarial Network for 3D Point Cloud Upsampling

Líu¹,

Yuan²,

Hou³

et al. 2022

Preprint

View full text Add to dashboard Cite

We propose a generative adversarial network for point cloud upsampling, which can not only make the upsampled points evenly distributed on the underlying surface but also efficiently generate clean high frequency regions. The generator of our network includes a dynamic graph hierarchical residual aggregation unit and a hierarchical residual aggregation unit for point feature extraction and upsampling, respectively. The former extracts multiscale point-wise descriptive features, while the latter captures rich feature details with hierarchical residuals. To generate neat edges, our discriminator uses a graph filter to extract and retain high frequency points. The generated high resolution point cloud and corresponding high frequency points help the discriminator learn the global and high frequency properties of the point cloud. We also propose an identity distribution loss function to make sure that the upsampled points remain on the underlying surface of the input low resolution point cloud. To assess the regularity of the upsampled points in high frequency regions, we introduce two evaluation metrics. Objective and subjective results demonstrate that the visual quality of the upsampled point clouds generated by our method is better than that of the state-of-the-art methods.

show abstract

“…Text-to-Image (T2I) generative models [6,36,37,39] have greatly expanded the ways we can create and edit visual content. Recent works [23,27,33,44] have demonstrated high-quality Text-to-3D generation by optimizing neural radiance fields (NeRFs) [28] using the T2I diffusion models. Such automatic 3D asset creation with input text prompts alone has applications in a wide range of areas, such as graphics, VR, movies, and gaming.…”

Section: Introductionmentioning

confidence: 99%