Point2Skeleton: Learning Skeletal Representations from Point Clouds

• URL: http://arxiv.org/abs/2012.00230v2
• Date: Wed, 7 Apr 2021 04:24:20 GMT
• Title: Point2Skeleton: Learning Skeletal Representations from Point Clouds
• Authors: Cheng Lin, Changjian Li, Yuan Liu, Nenglun Chen, Yi-King Choi, Wenping Wang
• Abstract summary: We introduce Point2Skeleton, an unsupervised method to learn skeletal representations from point clouds. We first predict a set of skeletal points by learning a geometric transformation, and then analyze the connectivity of the skeletal points to form skeletal mesh structures.
• Score: 36.62519847312199
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We introduce Point2Skeleton, an unsupervised method to learn skeletal representations from point clouds. Existing skeletonization methods are limited to tubular shapes and the stringent requirement of watertight input, while our method aims to produce more generalized skeletal representations for complex structures and handle point clouds. Our key idea is to use the insights of the medial axis transform (MAT) to capture the intrinsic geometric and topological natures of the original input points. We first predict a set of skeletal points by learning a geometric transformation, and then analyze the connectivity of the skeletal points to form skeletal mesh structures. Extensive evaluations and comparisons show our method has superior performance and robustness. The learned skeletal representation will benefit several unsupervised tasks for point clouds, such as surface reconstruction and segmentation.

Related papers

• SPEAL: Skeletal Prior Embedded Attention Learning for Cross-Source Point Cloud Registration [20.32815928095936]
  Point cloud registration has remained largely unexplored in cross-source point clouds and unstructured scenes. We propose a novel method termed SPEAL to leverage skeletal representations for effective learning of intrinsic topologies of point clouds. To the best of our knowledge, our approach is the first to facilitate point cloud registration with skeletal geometric priors.
  arXiv  Detail & Related papers  (2023-12-14T05:35:51Z)
• Hierarchical Skeleton Meta-Prototype Contrastive Learning with Hard Skeleton Mining for Unsupervised Person Re-Identification [70.90142717649785]
  This paper proposes a generic unsupervised Hierarchical skeleton Meta-Prototype Contrastive learning (Hi-MPC) approach with Hard Skeleton Mining (HSM) for person re-ID with unlabeled 3D skeletons. By converting original prototypes into meta-prototypes with multiple homogeneous transformations, we induce the model to learn the inherent consistency of prototypes to capture more effective skeleton features for person re-ID.
  arXiv  Detail & Related papers  (2023-07-24T16:18:22Z)
• SkeletonMAE: Graph-based Masked Autoencoder for Skeleton Sequence Pre-training [110.55093254677638]
  We propose an efficient skeleton sequence learning framework, named Skeleton Sequence Learning (SSL) In this paper, we build an asymmetric graph-based encoder-decoder pre-training architecture named SkeletonMAE. Our SSL generalizes well across different datasets and outperforms the state-of-the-art self-supervised skeleton-based action recognition methods.
  arXiv  Detail & Related papers  (2023-07-17T13:33:11Z)
• Skeletal Point Representations with Geometric Deep Learning [0.6696732597888386]
  We propose novel geometric terms for calculating skeletal structures of objects. Results are similar to traditional fitted s-reps but are produced much more quickly.
  arXiv  Detail & Related papers  (2023-03-03T18:08:12Z)
• LAKe-Net: Topology-Aware Point Cloud Completion by Localizing Aligned Keypoints [35.82520172874995]
  LAKe-Net is a novel point cloud completion model by localizing aligned keypoints. A new type of skeleton, named Surface-skeleton, is generated from keypoints based on geometric priors. Experimental results show that our method achieves the state-of-the-art performance on point cloud completion.
  arXiv  Detail & Related papers  (2022-03-31T03:14:48Z)
• Skeleton-bridged Point Completion: From Global Inference to Local Adjustment [48.2757171993437]
  We propose a skeleton-bridged point completion network (SK-PCN) for shape completion. Given a partial scan, our method first predicts its 3D skeleton to obtain the global structure. We decouple the shape completion into structure estimation and surface reconstruction, which eases the learning difficulty.
  arXiv  Detail & Related papers  (2020-10-14T22:49:30Z)
• SkeletonNet: A Topology-Preserving Solution for Learning Mesh Reconstruction of Object Surfaces from RGB Images [85.66560542483286]
  This paper focuses on the challenging task of learning 3D object surface reconstructions from RGB images. We propose two models, the Skeleton-Based GraphConvolutional Neural Network (SkeGCNN) and the Skeleton-Regularized Deep Implicit Surface Network (SkeDISN) We conduct thorough experiments that verify the efficacy of our proposed SkeletonNet.
  arXiv  Detail & Related papers  (2020-08-13T07:59:25Z)
• Skeleton-Aware Networks for Deep Motion Retargeting [83.65593033474384]
  We introduce a novel deep learning framework for data-driven motion between skeletons. Our approach learns how to retarget without requiring any explicit pairing between the motions in the training set.
  arXiv  Detail & Related papers  (2020-05-12T12:51:40Z)
• Skeleton Extraction from 3D Point Clouds by Decomposing the Object into Parts [0.0]
  We propose to extract curve skeletons, from unorganized point clouds, by decomposing the object into its parts. We use translational symmetry, the fundamental property of GCs, to extract parts from point clouds. A part based approach also provides a natural and intuitive interface for user interaction.
  arXiv  Detail & Related papers  (2019-12-26T20:52:57Z)

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.