mlsanityData to 7 Oct 2026

Unsupervised Point Cloud Pre-Training via Occlusion Completion

arXiv 2010.01089 ↗First posted 2 Oct 2020cs.CV cs.LGcited by 186

Not on the map. Why?

Abstract (arXiv)

We describe a simple pre-training approach for point clouds. It works in three steps: 1. Mask all points occluded in a camera view; 2. Learn an encoder-decoder model to reconstruct the occluded points; 3. Use the encoder weights as initialisation for downstream point cloud tasks. We find that even when we construct a single pre-training dataset (from ModelNet40), this pre-training method improves accuracy across different datasets and encoders, on a wide range of downstream tasks. Specifically, we show that our method outperforms previous pre-training methods in object classification, and both part-based and semantic segmentation tasks. We study the pre-trained features and find that they lead to wide downstream minima, have high transformation invariance, and have activations that are highly correlated with part labels. Code and data are available at: https://github.com/hansen7/OcCo

Key facts. Cited by 186 papers in the map (data through 7 Oct 2026). First posted on arXiv 2 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 3; 2022: 42; 2023: 61; 2024: 38; 2025: 31; 2026: 11.

186 papers.

Sort:

By area:

  1. 2021Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modelingcited by 647
  2. 2022Masked Autoencoders for Point Cloud Self-supervised Learningcited by 456
  3. 2022Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud Pre-trainingcited by 263
  4. 2023Uni3D: Exploring Unified 3D Representation at Scalecited by 231
  5. 2023MVImgNet: A Large-scale Dataset of Multi-view Imagescited by 219
  6. 2022CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understandingcited by 198
  7. 2023Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D Understanding, Generation, and Instruction Followingcited by 192
  8. 2023OpenShape: Scaling Up 3D Shape Representation Towards Open-World Understandingcited by 183
  9. 2023Chat-Scene: Bridging 3D Scene and Large Language Models with Object Identifierscited by 170
  10. 2023Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretrainingcited by 162
  11. 2022Masked Discrimination for Self-Supervised Learning on Point Cloudscited by 155
  12. 2022Learning 3D Representations from 2D Pre-trained Models via Image-to-Point Masked Autoencoderscited by 143
38 more
  1. 2022Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Datacited by 126
  2. 2023Chat-3D: Data-efficiently Tuning Large Language Model for Universal Dialogue of 3D Scenescited by 125
  3. 2023Segment Any Point Cloud Sequences by Distilling Vision Foundation Modelscited by 123
  4. 2022Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning?cited by 121
  5. 2023PointGPT: Auto-regressively Generative Pre-training from Point Cloudscited by 115
  6. 2022P2P: Tuning Pre-trained Image Models for Point Cloud Analysis with Point-to-Pixel Promptingcited by 74
  7. 2023UniPAD: A Universal Pre-training Paradigm for Autonomous Drivingcited by 73
  8. 2022Benchmarking and Analyzing Point Cloud Classification under Corruptionscited by 70
  9. 2023CLIP2: Contrastive Language-Image-Point Pretraining from Real-World Point Cloud Datacited by 69
  10. 2022ALSO: Automotive Lidar Self-supervision by Occupancy estimationcited by 68
  11. 2023Instance-aware Dynamic Prompt Tuning for Pre-trained Point Cloud Modelscited by 67
  12. 2023DexArt: Benchmarking Generalizable Dexterous Manipulation with Articulated Objectscited by 64
  13. 2022GD-MAE: Generative Decoder for MAE Pre-training on LiDAR Point Cloudscited by 57
  14. 2024Mamba3D: Enhancing Local Features for 3D Point Cloud Analysis via State Space Modelcited by 57
  15. 2023PonderV2: Pave the Way for 3D Foundation Model with A Universal Pre-training Paradigmcited by 56
  16. 20214DContrast: Contrastive Learning with Dynamic Correspondences for 3D Scene Understandingcited by 54
  17. 2022Unsupervised Point Cloud Representation Learning with Deep Neural Networks: A Surveycited by 54
  18. 2023Joint-MAE: 2D-3D Joint Masked Autoencoders for 3D Point Cloud Pre-trainingcited by 52
  19. 2024Dynamic Adapter Meets Prompt Tuning: Parameter-Efficient Transfer Learning for Point Cloud Analysiscited by 52
  20. 2022Ponder: Point Cloud Pre-training via Neural Renderingcited by 50
  21. 2022Implicit Autoencoder for Point-Cloud Self-Supervised Representation Learningcited by 46
  22. 2022Comprehensive Review of Deep Learning-Based 3D Point Cloud Completion Processing and Analysiscited by 44
  23. 2024Parameter-Efficient Fine-Tuning in Spectral Domain for Point Cloud Learningcited by 43
  24. 2023Towards Compact 3D Representations via Point Feature Enhancement Masked Autoencoderscited by 42
  25. 2023GeoMAE: Masked Geometric Target Prediction for Self-supervised Point Cloud Pre-Trainingcited by 38
  26. 2023GrowSP: Unsupervised Semantic Segmentation of 3D Point Cloudscited by 34
  27. 2022Complete-to-Partial 4D Distillation for Self-Supervised Point Cloud Sequence Representation Learningcited by 30
  28. 2023Point2Vec for Self-Supervised Representation Learning on Point Cloudscited by 28
  29. 2023Take-A-Photo: 3D-to-2D Generative Pre-training of Point Cloud Modelscited by 25
  30. 2023Building3D: An Urban-Scale Dataset and Benchmarks for Learning Roof Structures from Point Cloudscited by 24
  31. 2024PCP-MAE: Learning to Predict Centers for Point Masked Autoencoderscited by 23
  32. 2022Evaluating Self-Supervised Learning for Molecular Graph Embeddingscited by 22
  33. 2023Self-supervised Learning for Pre-Training 3D Point Clouds: A Surveycited by 21
  34. 2024PointRWKV: Efficient RWKV-Like Model for Hierarchical Point Cloud Learningcited by 21
  35. 2022MAELi: Masked Autoencoder for Large-Scale LiDAR Point Cloudscited by 20
  36. 20223D-OAE: Occlusion Auto-Encoders for Self-Supervised Learning on Point Cloudscited by 18
  37. 2022EPCL: Frozen CLIP Transformer is An Efficient Point Cloud Encodercited by 18
  38. 2023Lang3DSG: Language-based contrastive pre-training for 3D Scene Graph predictioncited by 18

50 most cited of 186 All 186 (JSON)