mlsanityData to 7 Oct 2026

COARSE3D: Class-Prototypes for Contrastive Learning in Weakly-Supervised 3D Point Cloud Segmentation

arXiv 2210.01784 ↗First posted 4 Oct 2022cs.CVcited by 15

Not on the map. Why?

Abstract (arXiv)

Annotation of large-scale 3D data is notoriously cumbersome and costly. As an alternative, weakly-supervised learning alleviates such a need by reducing the annotation by several order of magnitudes. We propose COARSE3D, a novel architecture-agnostic contrastive learning strategy for 3D segmentation. Since contrastive learning requires rich and diverse examples as keys and anchors, we leverage a prototype memory bank capturing class-wise global dataset information efficiently into a small number of prototypes acting as keys. An entropy-driven sampling technique then allows us to select good pixels from predictions as anchors. Experiments on three projection-based backbones show we outperform baselines on three challenging real-world outdoor datasets, working with as low as 0.001% annotations.

Key facts. Cited by 15 papers in the map (data through 7 Oct 2026). Builds on 69 of its 93 references that are in the map. First posted on arXiv 4 Oct 2022.

Builds on

69 of 93 references.

References · 69

  1. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  2. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  3. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  4. 2019nuScenes: A multimodal dataset for autonomous drivingcited by 4,485
  5. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  6. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  7. 2015FaceNet: A Unified Embedding for Face Recognition and Clusteringcited by 2,833
  8. 2020Supervised Contrastive Learningcited by 2,803
  9. 2020Improved Baselines with Momentum Contrastive Learningcited by 2,282
  10. 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,277
  11. 2016Matching Networks for One Shot Learningcited by 2,246
  12. 2018Unsupervised Feature Learning via Non-Parametric Instance-level Discriminationcited by 1,838
38 more
  1. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,202
  2. 2020Contrastive Learning with Hard Negative Samplescited by 619
  3. 2019RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Cloudscited by 584
  4. 2021KITTI-360: A Novel Dataset and Benchmarks for Urban Scene Understanding in 2D and 3Dcited by 570
  5. 2020Dense Contrastive Learning for Self-Supervised Visual Pre-Trainingcited by 512
  6. 2020PointContrast: Unsupervised Pre-training for 3D Point Cloud Understandingcited by 502
  7. 2019PANet: Few-Shot Image Semantic Segmentation with Prototype Alignmentcited by 449
  8. 2020Searching Efficient 3D Architectures with Sparse Point-Voxel Convolutioncited by 422
  9. 2020Hard Negative Mixing for Contrastive Learningcited by 399
  10. 2020Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentationcited by 388
  11. 2017Large-scale Point Cloud Semantic Segmentation with Superpoint Graphscited by 384
  12. 2020Contrastive learning of global and local features for medical image segmentation with limited annotationscited by 309
  13. 2021In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learningcited by 301
  14. 2020Propagate Yourself: Exploring Pixel-Level Consistency for Unsupervised Visual Representation Learningcited by 297
  15. 2017Sampling Matters in Deep Embedding Learningcited by 285
  16. 2020PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentationcited by 274
  17. 2022Language-Grounded Indoor 3D Semantic Segmentation in the Wildcited by 243
  18. 2021Exploring Cross-Image Pixel Contrast for Semantic Segmentationcited by 242
  19. 2017Deep Learning for Case-Based Reasoning through Prototypes: A Neural Network that Explains Its Predictionscited by 234
  20. 2017SqueezeSeg: Convolutional Neural Nets with Recurrent CRF for Real-Time Road-Object Segmentation from 3D LiDAR Point Cloudcited by 231
  21. 2018SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloudcited by 228
  22. 2020SalsaNext: Fast, Uncertainty-aware Semantic Segmentation of LiDAR Point Clouds for Autonomous Drivingcited by 226
  23. 2017No Fuss Distance Metric Learning using Proxiescited by 225
  24. 2020Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning Contextual Shape Priors from Scene Completioncited by 222
  25. 2021Self-Supervised Pretraining of 3D Features on any Point-Cloudcited by 217
  26. 2020Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contextscited by 210
  27. 2019Local Aggregation for Unsupervised Learning of Visual Embeddingscited by 204
  28. 2021Spatio-temporal Self-Supervised Representation Learning for 3D Point Cloudscited by 180
  29. 2018Tangent Convolutions for Dense Prediction in 3Dcited by 173
  30. 2020SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentationcited by 170
  31. 2021(AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection for Sparse Semantic Segmentation Networkcited by 164
  32. 2020Attribute Prototype Network for Zero-Shot Learningcited by 160
  33. 2018SPLATNet: Sparse Lattice Networks for Point Cloud Processingcited by 158
  34. 2022Rethinking Semantic Segmentation: A Prototype Viewcited by 140
  35. 2019Cross-Batch Memory for Embedding Learningcited by 130
  36. 2022Image-to-Lidar Self-Supervised Distillation for Autonomous Driving Datacited by 126
  37. 2021Semi-Supervised Semantic Segmentation with Pixel-Level Contrastive Learning from a Class-wise Memory Bankcited by 125
  38. 2020SemanticPOSS: A Point Cloud Dataset with Large Quantity of Dynamic Instancescited by 123

50 most cited of 69 All 69 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 6; 2024: 6; 2025: 2; 2026: 1.

15 papers.

Sort:

By area:

  1. 2023Segment Any Point Cloud Sequences by Distilling Vision Foundation Modelscited by 123
  2. 2024SeeGround: See and Ground for Zero-Shot Open-Vocabulary 3D Visual Groundingcited by 66
  3. 2023UniSeg: A Unified Multi-Modal LiDAR Segmentation Network and the OpenPCSeg Codebasecited by 57
  4. 2023FRNet: Frustum-Range Networks for Scalable LiDAR Segmentationcited by 50
  5. 20244D Contrastive Superflows are Dense 3D Representation Learnerscited by 32
  6. 2024Is Your LiDAR Placement Optimized for 3D Scene Understanding?cited by 25
  7. 2024Multi-Space Alignments Towards Universal LiDAR Segmentationcited by 24
  8. 2023Synchronize Feature Extracting and Matching: A Single Branch Framework for 3D Object Trackingcited by 16
  9. 2024An Empirical Study of Training State-of-the-Art LiDAR Segmentation Modelscited by 12
  10. 2025LargeAD: Large-Scale Cross-Sensor Data Pretraining for Autonomous Drivingcited by 11
  11. 2023COLA: COarse-LAbel multi-source LiDAR semantic segmentation for autonomous drivingcited by 8
  12. 2025Zero-Shot 3D Visual Grounding from Vision-Language Modelscited by 2
3 more
  1. 2023Generalized Robot 3D Vision-Language Model with Fast Rendering and Pre-Training Vision-Language Alignmentcited by 1
  2. 2024MILAN: Milli-Annotations for Lidar Semantic Segmentationcited by 0
  3. 2026UniGround: Universal 3D Visual Grounding via Training-Free Scene Parsingcited by 0