CVPR 2021 accepted papers on Point clouds
85 of the 1,661 papers accepted at CVPR 2021 (5.1%) are on Point clouds, compared with 4.8% at CVPR 2020. They fall into 4 topics.
See CVPR 2021 in the interactive map
Submitting to CVPR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Point cloud classification · 31 papers
- PAConv: Position Adaptive Convolution With Dynamic Kernel Assembling on Point Clouds (517 citations)
- PU-GCN: Point Cloud Upsampling Using Graph Convolutional Networks (261 citations)
- Point 4D Transformer Networks for Spatio-Temporal Modeling in Point Cloud Videos (210 citations)
- VoxelContext-Net: An Octree Based Framework for Point Cloud Compression (192 citations)
- SOE-Net: A Self-Attention and Orientation Encoding Network for Point Cloud Based Place Recognition (188 citations)
- TPCN: Temporal Point Cloud Networks for Motion Forecasting (187 citations)
- Scan2Cap: Context-Aware Dense Captioning in RGB-D Scans (136 citations)
- 3D AffordanceNet: A Benchmark for Visual Object Affordance Understanding (120 citations)
- Bidirectional Projection Network for Cross Dimension Scene Understanding (116 citations)
- UV-Net: Learning From Boundary Representations (115 citations)
- BRepNet: A Topological Message Passing System for Solid Models (107 citations)
- PV-RAFT: Point-Voxel Correlation Fields for Scene Flow Estimation of Point Clouds (95 citations)
and 19 more in the interactive map.
Segmentation and detection · 23 papers
- Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation (599 citations)
- PointAugmenting: Cross-Modal Augmentation for 3D Object Detection (472 citations)
- 3D Object Detection With Pointformer (460 citations)
- SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud (426 citations)
- SCF-Net: Learning Spatial Contextual Features for Large-Scale Point Cloud Segmentation (307 citations)
- Exploring Data-Efficient 3D Scene Understanding With Contrastive Scene Contexts (277 citations)
- Semantic Segmentation for Real Point Cloud Scenes via Bilateral Augmentation and Adaptive Fusion (269 citations)
- LiDAR R-CNN: An Efficient and Universal 3D Object Detector (245 citations)
- Towards Semantic Segmentation of Urban-Scale 3D Point Clouds: A Dataset, Benchmarks and Challenges (217 citations)
- Offboard 3D Object Detection From Point Cloud Sequences (176 citations)
- HVPR: Hybrid Voxel-Point Representation for Single-Stage 3D Object Detection (167 citations)
- One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation (156 citations)
and 11 more in the interactive map.
Shape and surface completion · 18 papers
- Diffusion Probabilistic Models for 3D Point Cloud Generation (620 citations)
- Variational Relational Point Completion Network (237 citations)
- PMP-Net: Point Cloud Completion by Learning Multi-Step Point Moving Paths (206 citations)
- Point Cloud Upsampling via Disentangled Refinement (178 citations)
- Cycle4Completion: Unpaired Point Cloud Completion Using Cycle Transformation With Missing Region Coding (118 citations)
- View-Guided Point Cloud Completion (93 citations)
- CorrNet3D: Unsupervised End-to-End Learning of Dense Correspondence for 3D Point Clouds (85 citations)
- Style-Based Point Generator With Adversarial Rendering for Point Cloud Completion (81 citations)
- RfD-Net: Point Scene Understanding by Semantic Instance Reconstruction (76 citations)
- Point2Skeleton: Learning Skeletal Representations from Point Clouds (74 citations)
- Learning Delaunay Surface Elements for Mesh Reconstruction (55 citations)
- Skeleton Merger: An Unsupervised Aligned Keypoint Detector (36 citations)
and 6 more in the interactive map.
Registration and matching · 13 papers
- Predator: Registration of 3D Point Clouds With Low Overlap (715 citations)
- SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registration (392 citations)
- PointDSC: Robust Point Cloud Registration Using Deep Spatial Consistency (359 citations)
- Robust Point Cloud Registration Framework Based on Deep Graph Matching (247 citations)
- DeepI2P: Image-to-Point Cloud Registration via Deep Classification (113 citations)
- PointNetLK Revisited (74 citations)
- StickyPillars: Robust and Efficient Feature Matching on Point Clouds Using Graph Neural Networks (69 citations)
- Locally Aware Piecewise Transformation Fields for 3D Human Mesh Registration (66 citations)
- ReAgent: Point Cloud Registration Using Imitation and Reinforcement Learning (51 citations)
- UnsupervisedR&R: Unsupervised Point Cloud Registration via Differentiable Rendering (43 citations)
- Self-Supervised Geometric Perception (30 citations)
- RPSRNet: End-to-End Trainable Rigid Point Set Registration Network Using Barnes-Hut 2D-Tree Representation (13 citations)
and 1 more in the interactive map.
