CVPR 2026 accepted papers on Point clouds
82 of the 4,042 papers accepted at CVPR 2026 (2.0%) are on Point clouds, compared with 3.6% at CVPR 2025. They fall into 4 topics.
See CVPR 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Point cloud classification · 35 papers
- LitePT: Lighter Yet Stronger Point Transformer (1 citation)
- 3D sans 3D Scans: Scalable Pre-training from Video-Generated Point Clouds
- Action-Geometry Prediction with 3D Geometric Prior for Bimanual Manipulation
- AffordMatcher: Affordance Learning in 3D Scenes from Visual Signifiers
- AnyPcc: Compressing Any Point Cloud with a Single Universal Model
- AvatarPointillist: AutoRegressive 4D Gaussian Avatarization
- Back to Point: Exploring Point-Language Models for Zero-Shot 3D Anomaly Detection
- CLIPoint3D: Language-Grounded Few-Shot Unsupervised 3D Point Cloud Domain Adaptation
- Deformation-based In-Context Learning for Point Cloud Understanding
- Direction-aware 3D Large Multimodal Models
- Efficient Encoder-Free Fourier-based 3D Large Multimodal Model
- FILTR: Extracting Topological Features from Pretrained 3D Models
and 23 more in the interactive map.
Segmentation and detection · 19 papers
- Rewis3d: Reconstruction Improves Weakly-Supervised Semantic Segmentation (1 citation)
- CCF: Complementary Collaborative Fusion for Domain Generalized Multi-Modal 3D Object Detection
- ESAM++: Efficient Online 3D Perception on the Edge
- Geometric-Aware Hypergraph Reasoning for Novel Class Discovery in Point Cloud Segmentation
- Ghost-FWL: A Large-Scale Full-Waveform LiDAR Dataset for Ghost Detection and Removal
- Learning to Identify Out-of-Distribution Objects for 3D LiDAR Anomaly Segmentation
- PointGS: Semantic-Consistent Unsupervised 3D Point Cloud Segmentation with 3D Gaussian Splatting
- QD-PCQA: Quality-Aware Domain Adaptation for Point Cloud Quality Assessment
- QueryOcc: Query-based Self-Supervision for 3D Semantic Occupancy
- TerraSeg: Self-Supervised Ground Segmentation for Any LiDAR
- Towards Foundation Models for 3D Scene Understanding: Instance-Aware Self-Supervised Learning for Point Clouds
- Adaptive 3D Perception for Small Aerial Targets Under Sparse Sampling via Reinforcement Learning
and 7 more in the interactive map.
Shape and surface completion · 18 papers
- BuildAnyPoint: 3D Building Structured Abstraction from Diverse Point Clouds
- FUSER: Feed-Forward Multiview 3D Registration Transformer and SE(3)$^N$ Diffusion Refinement
- Learning Eigenstructures of Unstructured Data Manifolds
- Learning to Infer Parameterized Representations of Plants from 3D Scans
- Neu-PiG: Neural Preconditioned Grids for Fast Dynamic Surface Reconstruction on Long Sequences
- PartDiffuser: Part-wise 3D Mesh Generation via Discrete Diffusion
- PointNSP: Autoregressive 3D Point Cloud Generation with Next-Scale Level-of-Detail Prediction
- Points-to-3D: Structure-Aware 3D Generation with Point Cloud Priors
- PQDT: Pseudo-Query Dual Transformer for Robust Point Cloud Restoration
- Velox: Learning Representations of 4D Geometry and Appearance
- Edges Compete for Trust: Group Relative Edge Optimization for Building Reconstruction from Point Clouds
- Hyper-PCN: Hypergraph-Based Point Cloud Completion via High-Order Correlation Modeling
and 6 more in the interactive map.
Registration and matching · 10 papers
- Rethinking 2D-3D Registration: A Novel Network for High-Value Zone Selection and Representation Consistency Alignment (2 citations)
- C-GenReg: Training-Free 3D Point Cloud Registration by Multi-View-Consistent Geometry-to-Image Generation with Probabilistic Modalities Fusion
- Generalized-CVO: Fast and Correspondence-Free Local Point Cloud Registration with Second Order Riemannian Optimization
- Hg-I2P: Bridging Modalities for Generalizable Image-to-Point-Cloud Registration via Heterogeneous Graphs
- UniCorrn: Unified Correspondence Transformer Across 2D and 3D
- DualReg: Dual-Space Filtering and Reinforcement for Rigid Registration
- MHopReg: Efficient Hierarchical Multi-Hop Graph Search for Point Cloud Registration
- mmWaveFlow: Unified Enhancement and Generation of mmWave Human Point Clouds
- SuP: Sub-cloud Driven Point Cloud Registration
- Topology-aware Feature Propagation for Unsupervised Non-rigid Point Cloud Correspondence