ECCV 2020 3D点云 Point Cloud 文章汇总

一、点云文章资源

近年来，对于点云处理的研究越来越火热。Github上面有一个工程，汇总了从2017年以来各大会议上点云论文，awesome-point-cloud-analysis ,本文作者之前整理了CVPR 2020 中的点云论文。

本文主要整理ECCV2020中的点云相关论文，总共70多篇，供大家查阅。

二、ECCV 2020 点云文章汇总

点云分析

1. A Closer Look at Local Aggregation Operators in Point Cloud Analysis 用一个统一的deep resnet框架来对比了几种不同的聚合操作算子，发现几种算子对最后效果的差别不大，因此可用作者设计的最简单的没有学习参数的聚合操作代替。
2. PointContrast: Unsupervised Pre-training for 3D Point Cloud Understanding
3. Deep Positional and Relational Feature Learning for Rotation-Invariant Point Cloud Analysis
4. Label-Efficient Learning on Point Clouds using Approximate Convex Decompositions 使用了一种ACD分解方法对物体进行分割再学习，提升label的效率。
5. Orderly Disorder in Point Cloud Domain

点云分割

1. Deep FusionNet for Point Cloud Semantic Segmentation
2. Efficient Outdoor 3D Point Cloud Semantic Segmentation for Critical Road Objects and Distributed Contexts
3. Instance-Aware Embedding for Point Cloud Instance Segmentation
4. Joint Semantic Segmentation and Edge Detection Network for 3D Point Clouds
5. Learning and Memorizing Representative Prototypes for 3D Point Cloud Semantic and Instance Segmentation
6. Self-Prediction for Joint Instance and Semantic Segmentation of Point Clouds
7. SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentation

点云补全

1. （Oral）SoftPoolNet: Shape Descriptor for Point Cloud Completion and Classification
2. Multimodal Shape Completion via Conditional Generative Adversarial Networks
3. Weakly-supervised 3D Shape Completion in the Wild
4. Detail Preserved Point Cloud Completion via Separated Feature Aggregation
5. GRNet: Gridding Residual Network for Dense Point Cloud Completion
6. Towards Precise Completion of Deformable Shapes
7. Mapping in a Cycle: Sinkhorn Regularized Unsupervised Learning for Point Cloud Shapes

目标检测

1. （*）Active Perception using Light Curtains for Autonomous Driving
2. Generative Sparse Detection Networks for 3D Single-shot Object Detection
3. 3D-CVF: Generating Joint Camera and LiDAR Features Using Cross-View Spatial Feature Fusion for 3D Object Detection
4. An LSTM Approach to Temporal 3D Object Detection in LiDAR Point Clouds
5. Improving 3D Object Detection through Progressive Population Based Augmentation
6. EPNet: Enhancing Point Features with Image Semantics for 3D Object Detection
7. Finding Your (3D) Center: 3D Object Detection Using a Learned Loss
8. H3DNet: 3D Object Detection Using Hybrid Geometric Primitives
9. InfoFocus: 3D Object Detection for Autonomous Driving with Dynamic Information Modeling
10. Monocular 3D Object Detection via Feature Domain Adaptation
11. Object as Hotspots: An Anchor-Free 3D Object Detection Approach via Firing of Hotspots
12. Pillar-based Object Detection for Autonomous Driving
13. SPOT: Selective Point Cloud Voting for Better Proposal in Point Cloud Object Detection
14. SSN: Shape Signature Networks for Multi-class Object Detection from Point Clouds
15. Monocular Differentiable Rendering for Self-Supervised 3D Object Detection
16. Streaming Object Detection for 3-D Point Clouds
17. Weakly Supervised 3D Object Detection from Lidar Point Cloud

Re-ID

1. （Oral）ReferIt3D: Neural Listeners for Fine-Grained 3D Object Identification in Real-World Scenes

差值

1. （Oral）Intrinsic Point Cloud Interpolation via Dual Latent Space Navigation

Correspondence

1. Cyclic Functional Mapping: Self-supervised Correspondence between Non-isometric Deformable Shapes
2. Human Correspondence Consensus for 3D Object Semantic Understanding

surface fitting

1. （Oral）DeepFit: 3D Surface Fitting via Neural Network Weighted Least Squares 拟合曲面时，给每个点赋予一个权重，用一个网络（Pointnet）去学习这个权重。
2. A Parametric Surface Fitting Network for 3D Point Clouds

Registration

1. DH3D: Deep Hierarchical 3D Descriptors for Robust Large-Scale 6DoF Relocalization
2. DeepGMR: Learning Latent Gaussian Mixture Models for Registration
3. A Closest Point Proposal for MCMC-based Probabilistic Surface Registration
4. CN: Channel Normalization For Point Cloud Recognition
5. FreeCam3D: Snapshot Structured Light 3D with Freely-Moving Cameras
6. Iterative Distance-Aware Similarity Matrix Convolution with Mutual-Supervised Point Elimination for Efficient Point Cloud Registration

场景流

1. PointPWC-Net: Cost Volume on Point Clouds for (Self-)Supervised Scene Flow Estimation
2. FLOT: Scene Flow on Point Clouds guided by Optimal Transport

数据集

1. A Large-scale Annotated Mechanical Components Benchmark for Classification and Retrieval Tasks with Deep Neural Networks
2. CLOTH3D: Clothed 3D Humans

数据增强

1. PointMixup: Augmentation for Point Clouds

重建

1. Accurate Reconstruction of Oriented 3D Points using Affine Correspondences
2. （Oral）Combining Implicit Function Learning and Parametric Models for 3D Human Reconstruction 稀疏点经过一个网络后按照身体部位进行分类，之后通过一个人体参数模型来拟合。

安全

1. AdvPC: Transferable Adversarial Perturbations on 3D Point Clouds
2. （Oral）Privacy Preserving Structure-from-Motion

姿态

1. Hand-Transformer: Non-Autoregressive Structured Modeling for 3D Hand Pose Estimation

去噪

1. Learning Graph-Convolutional Representations for Point Cloud Denoising

其他

1. Learning Gradient Fields for Shape Generation
2. PT2PC: Learning to Generate 3D Point Cloud Shapes from Part Tree Conditions
3. Progressive Point Cloud Deconvolution Generation Network
4. Points2Surf Learning Implicit Surfaces from Point Clouds
5. Coupling Explicit and Implicit Surface Representations for Generative 3D Modeling
6. Discrete Point Flow Networks for Efficient Point Cloud Generation
7. （Oral）Quaternion Equivariant Capsule Networks for 3D Point Clouds
8. 3D-Rotation-Equivariant Quaternion Neural Networks
9. DPDist: Comparing Point Clouds Using Deep Point Cloud Distance
10. Deformation-Aware 3D Model Embedding and Retrieval
11. DELTAS: Depth Estimation by Learning Triangulation And densification of Sparse points
12. Implicit Latent Variable Model for Scene-Consistent Motion Forecasting
13. Interactive Annotation of 3D Object Geometry using 2D Scribbles
14. Neural Point-Based Graphics
15. Meshing Point Clouds with Predicted Intrinsic-Extrinsic Ratio Guidance
16. PUGeo-Net: A Geometry-centric Network for 3D Point Cloud Upsampling
17. Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution
18. Unsupervised Learning of Category-Specific Symmetric 3D Keypoints from Point Sets
19. PointTriNet: Learned Triangulation of 3D Point Sets
20. Info3D: Representation Learning on 3D Objects using Mutual Information Maximization and Contrastive Learning