mlsanityData to 7 Oct 2026

Benchmarking and Analyzing Point Cloud Classification under Corruptions

arXiv 2202.03377 ↗First posted 7 Feb 2022cs.CVcited by 70

Not on the map. Why?

Abstract (arXiv)

3D perception, especially point cloud classification, has achieved substantial progress. However, in real-world deployment, point cloud corruptions are inevitable due to the scene complexity, sensor inaccuracy, and processing imprecision. In this work, we aim to rigorously benchmark and analyze point cloud classification under corruptions. To conduct a systematic investigation, we first provide a taxonomy of common 3D corruptions and identify the atomic corruptions. Then, we perform a comprehensive evaluation on a wide range of representative point cloud models to understand their robustness and generalizability. Our benchmark results show that although point cloud classification performance improves over time, the state-of-the-art methods are on the verge of being less robust. Based on the obtained observations, we propose several effective techniques to enhance point cloud classifier robustness. We hope our comprehensive benchmark, in-depth analysis, and proposed techniques could spark future research in robust 3D perception.

Key facts. Cited by 70 papers in the map (data through 7 Oct 2026). Builds on 39 of its 49 references that are in the map. First posted on arXiv 7 Feb 2022.

Builds on

39 of 49 references.

References · 39

  1. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  2. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  3. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  4. 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,875
  5. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  6. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,220
  7. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,842
  8. 2019Do ImageNet Classifiers Generalize to ImageNet?cited by 1,507
  9. 2020Point Transformercited by 1,462
  10. 2019Natural Adversarial Examplescited by 1,401
  11. 2020PCT: Point cloud transformercited by 730
  12. 2021Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modelingcited by 647
27 more
  1. 2021Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstructioncited by 646
  2. 2019Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Datacited by 532
  3. 2017Graph Signal Processing: Overview, Challenges and Applicationscited by 409
  4. 2019Relation-Shape Convolutional Neural Network for Point Cloud Analysiscited by 346
  5. 2017Dynamic Edge-Conditioned Filters in Convolutional Neural Networks on Graphscited by 307
  6. 2021Vector Neurons: A General Framework for SO(3)-Equivariant Networkscited by 304
  7. 2019A Fourier Perspective on Model Robustness in Computer Visioncited by 288
  8. 2016Volumetric and Multi-View CNNs for Object Classification on 3D Datacited by 240
  9. 2020PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Samplingcited by 238
  10. 2018SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloudcited by 228
  11. 2021PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Cloudscited by 211
  12. 2020Unsupervised Point Cloud Pre-Training via Occlusion Completioncited by 186
  13. 2021Walk in the Cloud: Learning Curves for Point Clouds Shape Analysiscited by 172
  14. 2020Objectron: A Large Scale Dataset of Object-Centric Videos in the Wild with Pose Annotationscited by 155
  15. 2021Revisiting Point Cloud Shape Classification with a Simple and Effective Baselinecited by 145
  16. 2020PointMixup: Augmentation for Point Cloudscited by 104
  17. 2019Rotation Invariant Convolutions for 3D Point Clouds Deep Learningcited by 80
  18. 2013Discrete Signal Processing on Graphs: Frequency Analysiscited by 72
  19. 2020Learning Geometry-Disentangled Representation for Complementary Understanding of 3D Object Point Cloudcited by 71
  20. 2020PointAugment: an Auto-Augmentation Framework for Point Cloud Classificationcited by 65
  21. 2021Regularization Strategy for Point Cloud via Rigidly Mixed Samplecited by 59
  22. 2021PointGuard: Provably Robust 3D Point Cloud Classificationcited by 49
  23. 2021Point Cloud Augmentation with Weighted Local Transformationscited by 39
  24. 2020Cloud Transformers: A Universal Approach To Point Cloud Processing Taskscited by 24
  25. 2020Triangle-Net: Towards Robustness in Point Cloud Learningcited by 13
  26. 2020RobustPointSet: A Dataset for Benchmarking Robustness of Point Cloud Classifierscited by 10
  27. 2020Orderly Disorder in Point Cloud Domaincited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 9; 2023: 22; 2024: 20; 2025: 10; 2026: 9.

70 papers.

Sort:

By area:

  1. 2023OmniObject3D: Large-Vocabulary 3D Object Dataset for Realistic Perception, Reconstruction and Generationcited by 324
  2. 2024ShapeLLM: Universal 3D Object Understanding for Embodied Interactioncited by 133
  3. 2023Robo3D: Towards Robust and Reliable 3D Perception against Corruptionscited by 128
  4. 2023Rethinking Range View Representation for LiDAR Segmentationcited by 126
  5. 2023Benchmarking Robustness of 3D Object Detection to Common Corruptions in Autonomous Drivingcited by 112
  6. 2024Benchmarking and Improving Bird's Eye View Perception Robustness in Autonomous Drivingcited by 39
  7. 2024Is Your HD Map Constructor Reliable under Sensor Corruptions?cited by 28
  8. 2022MATE: Masked Autoencoders are Online 3D Test-Time Learnerscited by 23
  9. 2023BiBench: Benchmarking and Analyzing Network Binarizationcited by 22
  10. 2022Exploring Point-BEV Fusion for 3D Point Cloud Object Tracking with Transformercited by 18
  11. 2023Benchmarking the Robustness of LiDAR Semantic Segmentation Modelscited by 18
  12. 2024GEAL: Generalizable 3D Affordance Learning with Cross-Modal Consistencycited by 17
38 more
  1. 2023Benchmarking Robustness in Neural Radiance Fieldscited by 13
  2. 2023The RoboDepth Challenge: Methods and Advancements Towards Robust Depth Estimationcited by 13
  3. 2024DSRC: Learning Density-insensitive and Semantic-aware Collaborative Representation against Corruptionscited by 13
  4. 2023Annotator: A Generic Active Learning Baseline for LiDAR Semantic Segmentationcited by 11
  5. 2024Calib3D: Calibrating Model Preferences for Reliable 3D Scene Understandingcited by 11
  6. 2022Common Corruption Robustness of Point Cloud Detectors: Benchmark and Enhancementcited by 9
  7. 2022Point-DAE: Denoising Autoencoders for Self-supervised Point Cloud Learningcited by 9
  8. 2022PointCA: Evaluating the Robustness of 3D Point Cloud Completion Models Against Adversarial Examplescited by 9
  9. 2023Risk-optimized Outlier Removal for Robust 3D Point Cloud Classificationcited by 6
  10. 2023Sample-adaptive Augmentation for Point Cloud Recognition Against Real-world Corruptionscited by 6
  11. 2024FSC: Few-point Shape Completioncited by 6
  12. 2024Fast and Simple Explainability for Point Cloud Networkscited by 6
  13. 2024UPose3D: Uncertainty-Aware 3D Human Pose Estimation with Cross-View and Temporal Cuescited by 6
  14. 2024Point-PRC: A Prompt Learning Based Regulation Framework for Generalizable Point Cloud Analysiscited by 6
  15. 2022On the Robustness of 3D Object Detectorscited by 5
  16. 2023EPiC: Ensemble of Partial Point Clouds for Robust Classificationcited by 5
  17. 2023Persistent Homology Meets Object Unity: Object Recognition in Cluttercited by 5
  18. 2024PointGL: A Simple Global-Local Framework for Efficient Point Cloud Analysiscited by 5
  19. 2023PointPatchMix: Point Cloud Mixing with Patch Scoringcited by 4
  20. 2023Benchmarking and Analyzing Robust Point Cloud Recognition: Bag of Tricks for Defending Adversarial Examplescited by 4
  21. 2024Explainability of Point Cloud Neural Networks Using SMILE: Statistical Model-Agnostic Interpretability with Local Explanationscited by 4
  22. 2025Point-Cache: Test-time Dynamic and Hierarchical Cache for Robust and Generalizable Point Cloud Analysiscited by 4
  23. 2025SMART-PC: Skeletal Model Adaptation for Robust Test-Time Training in Point Cloudscited by 4
  24. 2023Variational Relational Point Completion Network for Robust 3D Classificationcited by 3
  25. 2024ModelNet-O: A Large-Scale Synthetic Dataset for Occlusion-Aware Point Cloud Classificationcited by 3
  26. 2024Team Samsung-RAL: Technical Report for 2024 RoboDrive Challenge-Robust Map Segmentation Trackcited by 3
  27. 2023Robustifying Point Cloud Networks by Refocusingcited by 2
  28. 2024Target-Guided Adversarial Point Cloud Transformer Towards Recognition Against Real-world Corruptionscited by 2
  29. 2024OODFace: Benchmarking Robustness of Face Recognition under Common Corruptions and Appearance Variationscited by 2
  30. 2025Occlusion-aware Text-Image-Point Cloud Pretraining for Open-World 3D Object Recognitioncited by 2
  31. 2025Content-Distortion High-Order Interaction for Blind Image Quality Assessmentcited by 2
  32. 2025APCoTTA: Continual Test-Time Adaptation for Semantic Segmentation of Airborne LiDAR Point Cloudscited by 2
  33. 2025TACO-Net: Topological Signatures Triumph in 3D Object Classificationcited by 2
  34. 2022Point-Voxel Adaptive Feature Abstraction for Robust Point Cloud Classificationcited by 1
  35. 2022Semantic keypoint extraction for scanned animals using multi-depth-camera systemscited by 1
  36. 2023Enhancing Infrared Small Target Detection Robustness with Bi-Level Adversarial Frameworkcited by 1
  37. 2023CSI: Enhancing the Robustness of 3D Point Cloud Recognition against Corruptioncited by 1
  38. 2024MM-Mixing: Multi-Modal Mixing Alignment for 3D Understandingcited by 1

50 most cited of 70 All 70 (JSON)