• Corpus ID: 244527380

Density-aware Chamfer Distance as a Comprehensive Metric for Point Cloud Completion

  title={Density-aware Chamfer Distance as a Comprehensive Metric for Point Cloud Completion},
  author={Tong Wu and Liang Pan and Junzhe Zhang and Tai Wang and Ziwei Liu and Dahua Lin},
Chamfer Distance (CD) and Earth Mover’s Distance (EMD) are two broadly adopted metrics for measuring the similarity between two point sets. However, CD is usually insensitive to mismatched local density, and EMD is usually dominated by global distribution while overlooks the fidelity of detailed structures. Besides, their unbounded value range induces a heavy influence from the outliers. These defects prevent them from providing a consistent evaluation. To tackle these problems, we propose a… 

Figures and Tables from this paper

Completing Partial Point Clouds with Outliers by Collaborative Completion and Segmentation

This paper proposes an end-toend network, named CS-Net, to complete the point clouds contaminated by noises or containing outliers, and designs a novel completion decoder which harnesses the labels obtained by segmentation together with FPS to purify the point cloud and leverages KNN-grouping for better generation.

Multi-View Partial (MVP) Point Cloud Challenge 2021 on Completion and Registration: Methods and Results

Methods and results in the Multi-View Partial Point Cloud Challenge 2021 on Completion and Registration are reported, and the top-ranked solutions will be analyzed and the future research directions are discussed.

Shape Completion with Points in the Shadow

Inspired by the classic shadow volume technique in computer graphics, a new method is proposed to reduce the solution space effectively and outperforms state-of-the-art methods qualitatively and quantitatively on MVP datasets.

PU-MFA: Point Cloud Up-Sampling via Multi-Scale Features Attention

Inspired by prior studies that reported good performance at generating high-quality dense point set using the multi-scale features or attention mechanisms, PU-MFA merges the two through a U-Net structure and adaptively uses multi- scale features to refine the global features effectively.

CD2: Fine-grained 3D Mesh Reconstruction with Twice Chamfer Distance

This paper proposes a fine-grained reconstruction method CD 2, which directly generates well-structured meshes and outperforms others by alleviating VC and IT problems and demonstrates that point-vertex mapping relation is more suitable for 3D mesh reconstruction than the distance relation.

Learning to Train a Point Cloud Reconstruction Network Without Matching

A novel framework named PCLossNet is proposed which learns to train a point cloud reconstruction network without any matching, by training through an adversarial process together with the reconstruction network, to better explore the differences between point clouds and create more precise reconstruction results.

D-LC-Nets: Robust Denoising and Loop Closing Networks for LiDAR SLAM in Complicated Circumstances with Noisy Point Clouds

It is demonstrated by extensive experiments and benchmark studies that this method can have a significant boost on the localization accuracy of the LiDAR SLAM system when faced with noisy point clouds, with a marginal increase in computational cost.

An Integrated LiDAR-SLAM System for Complex Environment with Noisy Point Clouds

This work proposed a lightweight network for large-scale point clouds denoising and designed an efficient loop closure network for place recognition in global optimization to improve the localization accuracy of the whole system.

Comprehensive Review of Deep Learning-Based 3D Point Cloud Completion Processing and Analysis

This work aims to conduct a comprehensive survey on various methods of point cloud completion, including point-based, view- based, convolution-based), convolution, graph, graph based, generative model based, transformer-based approaches, etc, and summarizes the comparisons among these methods to provoke further research insights.

An Interpretable Joint Nonnegative Matrix Factorization-Based Point Cloud Distance Measure

A new method for deter- mining shared features of and measuring the distance between data sets or point clouds using the joint factorization of two data matrices, which reveals structural differences in both image and text data.



Morphing and Sampling Network for Dense Point Cloud Completion

This work proposes a novel approach to complete the partial point cloud in two stages, which outperforms the existing methods in both the Earth Mover's Distance (EMD) and the Chamfer Distance (CD).

Point-set Distances for Learning Representations of 3D Point Clouds

Experiments show that the sliced Wasserstein distance allows the neural network to learn a more efficient representation compared to the Chamfer discrepancy, which is demonstrated on several tasks in 3D computer vision including training a point cloud autoencoder, generative modeling, transfer learning, and point cloud registration.

DPDist : Comparing Point Clouds Using Deep Point Cloud Distance

This work introduces a new deep learning method for point cloud comparison that measures the distance between the points in one cloud and the estimated surface from which the other point cloud is sampled, using the 3D modified Fisher vector representation.

Variational Relational Point Completion Network

A variational framework, Variational Relational point Completion network (VRC-Net), with two appealing properties: Probabilistic Modeling and Relational Enhancement, which shows great generalizability and robustness on real-world point cloud scans.

Detail Preserved Point Cloud Completion via Separated Feature Aggregation

Qualitative and quantitative evaluations demonstrate that the proposed network outperforms current state-of-the art methods especially on detail preservation.

Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation

  • Xinge ZhuHui Zhou Dahua Lin
  • Computer Science
    2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2021
A new framework for the outdoor LiDAR segmentation is proposed, where cylindrical partition and asymmetrical 3D convolution networks are designed to explore the 3D geometric pattern while maintaining these inherent properties.

PF-Net: Point Fractal Network for 3D Point Cloud Completion

A Point Fractal Network (PF-Net), a novel learning-based approach for precise and high-fidelity point cloud completion that preserves the spatial arrangements of the incomplete point cloud and can figure out the detailed geometrical structure of the missing region(s) in the prediction.

TopNet: Structural Point Cloud Decoder

This work proposes a novel decoder that generates a structured point cloud without assuming any specific structure or topology on the underlying point set, and significantly outperforms state-of-the-art 3D point cloud completion methods on the Shapenet dataset.

GRNet: Gridding Residual Network for Dense Point Cloud Completion

This work devise two novel differentiable layers, named Gridding and Gridding Reverse, to convert between point clouds and 3D grids without losing structural information, and presents the differentiable Cubic Feature Sampling layer to extract features of neighboring points, which preserves context information.

Cascaded Refinement Network for Point Cloud Completion

This work proposes a cascaded refinement network together with a coarse-to-fine strategy to synthesize the detailed object shapes and designs a patch discriminator that guarantees every local area has the same pattern with the ground truth to learn the complicated point distribution.