mlsanityData to 7 Oct 2026

KPConv: Flexible and Deformable Convolution for Point Clouds

arXiv 1904.08889 ↗First posted 18 Apr 2019cs.CVcited by 1,202

Not on the map. Why?

Abstract (arXiv)

We present Kernel Point Convolution (KPConv), a new design of point convolution, i.e. that operates on point clouds without any intermediate representation. The convolution weights of KPConv are located in Euclidean space by kernel points, and applied to the input points close to them. Its capacity to use any number of kernel points gives KPConv more flexibility than fixed grid convolutions. Furthermore, these locations are continuous in space and can be learned by the network. Therefore, KPConv can be extended to deformable convolutions that learn to adapt kernel points to local geometry. Thanks to a regular subsampling strategy, KPConv is also efficient and robust to varying densities. Whether they use deformable KPConv for complex tasks, or rigid KPconv for simpler tasks, our networks outperform state-of-the-art classification and segmentation approaches on several datasets. We also offer ablation studies and visualizations to provide understanding of what has been learned by KPConv and to validate the descriptive power of deformable KPConv.

Key facts. Cited by 1,202 papers in the map (data through 7 Oct 2026). First posted on arXiv 18 Apr 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 214; 2022: 253; 2023: 256; 2024: 189; 2025: 181; 2026: 109.

1,202 papers.

Sort:

By area:

  1. 2023ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenescited by 747
  2. 2021Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modelingcited by 647
  3. 2023Point Transformer V3: Simpler, Faster, Strongercited by 579
  4. 2021PointCLIP: Point Cloud Understanding by CLIPcited by 461
  5. 2022Masked Autoencoders for Point Cloud Self-supervised Learningcited by 456
  6. 2022PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategiescited by 432
  7. 2021Dynamic Neural Networks: A Surveycited by 405
  8. 2022Point Transformer V2: Grouped Vector Attention and Partition-based Poolingcited by 384
  9. 2022Rethinking Network Design and Local Geometry in Point Cloud: A Simple Residual MLP Frameworkcited by 376
  10. 2021An End-to-End Transformer Model for 3D Object Detectioncited by 329
  11. 2022Mask3D: Mask Transformer for 3D Semantic Instance Segmentationcited by 299
  12. 2023OpenMask3D: Open-Vocabulary 3D Instance Segmentationcited by 288
38 more
  1. 2022ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understandingcited by 262
  2. 2022Language-Grounded Indoor 3D Semantic Segmentation in the Wildcited by 243
  3. 2021PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detectioncited by 232
  4. 2021Group-Free 3D Object Detection via Transformerscited by 215
  5. 2021PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Cloudscited by 211
  6. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 200
  7. 2022CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understandingcited by 198
  8. 2022SoftGroup for 3D Instance Segmentation on Point Cloudscited by 197
  9. 2021ManiSkill: Generalizable Manipulation Skill Benchmark with Large-Scale Demonstrationscited by 193
  10. 2021Spatio-temporal Self-Supervised Representation Learning for 3D Point Cloudscited by 180
  11. 2022PLA: Language-Driven Open-Vocabulary 3D Scene Understandingcited by 176
  12. 2021RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR Point Cloud Segmentationcited by 173
  13. 2022Stratified Transformer for 3D Point Cloud Segmentationcited by 166
  14. 2021Panoptic nuScenes: A Large-Scale Benchmark for LiDAR Panoptic Segmentation and Trackingcited by 165
  15. 2021(AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection for Sparse Semantic Segmentation Networkcited by 164
  16. 20222DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Cloudscited by 157
  17. 2022Masked Discrimination for Self-Supervised Learning on Point Cloudscited by 155
  18. 2021CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registrationcited by 151
  19. 2021Revisiting Point Cloud Shape Classification with a Simple and Effective Baselinecited by 145
  20. 2022Focal Sparse Convolutional Networks for 3D Object Detectioncited by 142
  21. 2021InstanceRefer: Cooperative Holistic Understanding for Visual Grounding on Point Clouds through Instance Multi-level Contextual Referringcited by 140
  22. 2021RangeDet:In Defense of Range View for LiDAR-based 3D Object Detectioncited by 140
  23. 2023Meta-Transformer: A Unified Framework for Multimodal Learningcited by 136
  24. 2023Robo3D: Towards Robust and Reliable 3D Perception against Corruptionscited by 128
  25. 2021Mix3D: Out-of-Context Data Augmentation for 3D Scenescited by 126
  26. 2023Rethinking Range View Representation for LiDAR Segmentationcited by 126
  27. 2023Spherical Transformer for LiDAR-based 3D Recognitioncited by 124
  28. 2023Segment Any Point Cloud Sequences by Distilling Vision Foundation Modelscited by 123
  29. 2022Delving into the Devils of Bird's-eye-view Perception: A Review, Evaluation and Recipecited by 122
  30. 2023DSVT: Dynamic Sparse Voxel Transformer with Rotated Setscited by 122
  31. 2022Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning?cited by 121
  32. 2021NeSF: Neural Semantic Fields for Generalizable Semantic Segmentation of 3D Scenescited by 119
  33. 2022Point-to-Voxel Knowledge Distillation for LiDAR Semantic Segmentationcited by 118
  34. 2022PartSLIP: Low-Shot Part Segmentation for 3D Point Clouds via Pretrained Image-Language Modelscited by 117
  35. 2023Swin3D: A Pretrained Transformer Backbone for 3D Indoor Scene Understandingcited by 111
  36. 2021Beyond Self-attention: External Attention using Two Linear Layers for Visual Taskscited by 109
  37. 2025Sonata: Self-Supervised Learning of Reliable Point Representationscited by 108
  38. 2021Panoptic-PolarNet: Proposal-free LiDAR Point Cloud Panoptic Segmentationcited by 107

50 most cited of 1,202 All 1,202 (JSON)