Revisiting Point Cloud Completion: Are We Ready For The Real-World?
- URL: http://arxiv.org/abs/2411.17580v1
- Date: Tue, 26 Nov 2024 16:46:47 GMT
- Title: Revisiting Point Cloud Completion: Are We Ready For The Real-World?
- Authors: Stuti Pathak, Prashant Kumar, Nicholus Mboga, Gunther Steenackers, Rudi Penne,
- Abstract summary: We show that current benchmark synthetic point clouds lack rich topological features that are important constituents of point clouds captured in realistic settings.
We contribute the first real-world industrial point cloud dataset for point cloud completion, RealPC.
We show how 0-dimensional $mathcalPH$ priors, which extract the global topology of a complete shape in the form of a 3-D skeleton, can assist a model in generating topologically-consistent complete shapes.
- Score: 1.982969884513013
- License:
- Abstract: Point clouds acquired in constrained and challenging real-world settings are incomplete, non-uniformly sparse, or both. These obstacles present acute challenges for a vital task - point cloud completion. Using tools from Algebraic Topology and Persistent Homology ($\mathcal{PH}$), we demonstrate that current benchmark synthetic point clouds lack rich topological features that are important constituents of point clouds captured in realistic settings. To facilitate research in this direction, we contribute the first real-world industrial point cloud dataset for point cloud completion, RealPC - a diverse set of rich and varied point clouds, consisting of $\sim$ 40,000 pairs across 21 categories of industrial structures in railway establishments. Our benchmark results on several strong baselines reveal a striking observation - the existing methods are tailored for synthetic datasets and fail miserably in real-world settings. Building on our observation that RealPC consists of several 0 and 1-dimensional $\mathcal{PH}$-based topological features, we demonstrate the potential of integrating Homology-based topological priors into existing works. More specifically, we present how 0-dimensional $\mathcal{PH}$ priors, which extract the global topology of a complete shape in the form of a 3-D skeleton, can assist a model in generating topologically-consistent complete shapes.
Related papers
- Bridging Domain Gap of Point Cloud Representations via Self-Supervised Geometric Augmentation [15.881442863961531]
We introduce a novel scheme for induced geometric invariance of point cloud representations across domains.
On one hand, a novel pretext task of predicting translation of distances of augmented samples is proposed to alleviate centroid shift of point clouds.
On the other hand, we pioneer an integration of the relational self-supervised learning on geometrically-augmented point clouds.
arXiv Detail & Related papers (2024-09-11T02:39:19Z) - Node-Level Topological Representation Learning on Point Clouds [5.079602839359521]
We propose a novel method to extract node-level topological features from complex point clouds.
We verify the effectiveness of these topological point features on both synthetic and real-world data.
arXiv Detail & Related papers (2024-06-04T13:29:12Z) - Point Cloud Compression with Implicit Neural Representations: A Unified Framework [54.119415852585306]
We present a pioneering point cloud compression framework capable of handling both geometry and attribute components.
Our framework utilizes two coordinate-based neural networks to implicitly represent a voxelized point cloud.
Our method exhibits high universality when contrasted with existing learning-based techniques.
arXiv Detail & Related papers (2024-05-19T09:19:40Z) - Zero-shot Point Cloud Completion Via 2D Priors [52.72867922938023]
3D point cloud completion is designed to recover complete shapes from partially observed point clouds.
We propose a zero-shot framework aimed at completing partially observed point clouds across any unseen categories.
arXiv Detail & Related papers (2024-04-10T08:02:17Z) - Variational Relational Point Completion Network for Robust 3D
Classification [59.80993960827833]
Vari point cloud completion methods tend to generate global shape skeletons hence lack fine local details.
This paper proposes a variational framework, point Completion Network (VRCNet) with two appealing properties.
VRCNet shows great generalizability and robustness on real-world point cloud scans.
arXiv Detail & Related papers (2023-04-18T17:03:20Z) - Spotlights: Probing Shapes from Spherical Viewpoints [25.824284796437652]
We propose a novel sampling model called Spotlights to represent a 3D shape as a compact 1D array of depth values.
It simulates the configuration of cameras evenly distributed on a sphere, where each virtual camera casts light rays from its principal point through sample points on a small concentric spherical cap to probe for the possible intersections with the object surrounded by the sphere.
arXiv Detail & Related papers (2022-05-25T08:23:18Z) - Learning a Structured Latent Space for Unsupervised Point Cloud
Completion [48.79411151132766]
We propose a novel framework, which learns a unified and structured latent space that encoding both partial and complete point clouds.
Our proposed method consistently outperforms state-of-the-art unsupervised methods on both synthetic ShapeNet and real-world KITTI, ScanNet, and Matterport3D datasets.
arXiv Detail & Related papers (2022-03-29T13:58:44Z) - Voxel-based Network for Shape Completion by Leveraging Edge Generation [76.23436070605348]
We develop a voxel-based network for point cloud completion by leveraging edge generation (VE-PCN)
We first embed point clouds into regular voxel grids, and then generate complete objects with the help of the hallucinated shape edges.
This decoupled architecture together with a multi-scale grid feature learning is able to generate more realistic on-surface details.
arXiv Detail & Related papers (2021-08-23T05:10:29Z) - PMP-Net: Point Cloud Completion by Learning Multi-step Point Moving
Paths [54.459879603473034]
We design a novel neural network, named PMP-Net, to mimic the behavior of an earth mover.
It moves each point of the incomplete input to complete the point cloud, where the total distance of point moving paths should be shortest.
It learns a strict and unique correspondence on point-level, which can capture the detailed topology and structure relationships between the incomplete shape and the complete target.
arXiv Detail & Related papers (2020-12-07T01:34:38Z) - TearingNet: Point Cloud Autoencoder to Learn Topology-Friendly
Representations [20.318695890515613]
We propose an autoencoder, TearingNet, which tackles the challenging task of representing point clouds using a fixed-length descriptor.
Our TearingNet is characterized by a proposed Tearing network module and a Folding network module interacting with each other iteratively.
Experimentation shows the superiority of our proposal in terms of reconstructing point clouds as well as generating more topology-friendly representations than benchmarks.
arXiv Detail & Related papers (2020-06-17T22:42:43Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.