mlsanityData to 7 Oct 2026

SoftGroup for 3D Instance Segmentation on Point Clouds

arXiv 2203.01509 ↗First posted 3 Mar 2022cs.CVcited by 197

Not on the map. Why?

Abstract (arXiv)

Existing state-of-the-art 3D instance segmentation methods perform semantic segmentation followed by grouping. The hard predictions are made when performing semantic segmentation such that each point is associated with a single class. However, the errors stemming from hard decision propagate into grouping that results in (1) low overlaps between the predicted instance with the ground truth and (2) substantial false positives. To address the aforementioned problems, this paper proposes a 3D instance segmentation method referred to as SoftGroup by performing bottom-up soft grouping followed by top-down refinement. SoftGroup allows each point to be associated with multiple classes to mitigate the problems stemming from semantic prediction errors and suppresses false positive instances by learning to categorize them as background. Experimental results on different datasets and multiple evaluation metrics demonstrate the efficacy of SoftGroup. Its performance surpasses the strongest prior method by a significant margin of +6.2% on the ScanNet v2 hidden test set and +6.8% on S3DIS Area 5 in terms of AP50. SoftGroup is also fast, running at 345ms per scan with a single Titan X on ScanNet v2 dataset. The source code and trained models for both datasets are available at urlhttps://github.com/thangvubk/SoftGroup.git.

Key facts. Cited by 197 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 3 Mar 2022.

Builds on

39 of 49 references.

References · 39

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2017Mask R-CNNcited by 7,259
  4. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  5. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 4,993
  6. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  7. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  8. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,220
  9. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,202
  10. 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,201
  11. 2018PointCNN: Convolution On X-Transformed Pointscited by 764
  12. 20173D Semantic Segmentation with Submanifold Sparse Convolutional Networkscited by 746
27 more
  1. 2018PointConv: Deep Convolutional Networks on 3D Point Cloudscited by 656
  2. 2019Deep Hough Voting for 3D Object Detection in Point Cloudscited by 648
  3. 2017Frustum PointNets for 3D Object Detection from RGB-D Datacited by 640
  4. 2018PartNet: A Large-scale Benchmark for Fine-grained and Hierarchical Part-level 3D Object Understandingcited by 521
  5. 2019Relation-Shape Convolutional Neural Network for Point Cloud Analysiscited by 346
  6. 2016OctNet: Learning Deep 3D Representations at High Resolutionscited by 340
  7. 2020PointGroup: Dual-Set Point Grouping for 3D Instance Segmentationcited by 318
  8. 2017Dynamic Edge-Conditioned Filters in Convolutional Neural Networks on Graphscited by 307
  9. 2019Mask Scoring R-CNNcited by 265
  10. 2018SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filterscited by 261
  11. 20183D-SIS: 3D Semantic Instance Segmentation of RGB-D Scanscited by 216
  12. 2020OccuSeg: Occupancy-aware 3D Instance Segmentationcited by 177
  13. 2017SGPN: Similarity Group Proposal Network for 3D Point Cloud Instance Segmentationcited by 168
  14. 2019Learning Object Bounding Boxes for 3D Instance Segmentation on Point Cloudscited by 156
  15. 2018GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in Point Cloudcited by 145
  16. 20203D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentationcited by 140
  17. 2017Pointwise Convolutional Neural Networkscited by 126
  18. 2021Hierarchical Aggregation for 3D Instance Segmentationcited by 126
  19. 2021Instance Segmentation in 3D Scenes using Semantic Superpoint Tree Networkscited by 98
  20. 2019PanopticFusion: Online Volumetric Semantic Mapping at the Level of Stuff and Thingscited by 97
  21. 2019Associatively Segmenting Instances and Semantics in Point Cloudscited by 86
  22. 20193D Instance Segmentation via Multi-Task Metric Learningcited by 79
  23. 2020DyCo3D: Robust Instance Segmentation of 3D Point Clouds through Dynamic Convolutioncited by 74
  24. 2019JSIS3D: Joint Semantic-Instance Segmentation of 3D Point Clouds with Multi-Task Pointwise Networks and Multi-Value Conditional Random Fieldscited by 62
  25. 2019Point Cloud Instance Segmentation using Probabilistic Embeddingscited by 52
  26. 2020Learning Gaussian Instance Segmentation in Point Cloudscited by 33
  27. 2019MASC: Multi-scale Affinity with Sparse Convolution for 3D Instance Segmentationcited by 25

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 15; 2023: 63; 2024: 52; 2025: 42; 2026: 25.

197 papers.

Sort:

By area:

  1. 2023ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenescited by 747
  2. 2023OpenMask3D: Open-Vocabulary 3D Instance Segmentationcited by 288
  3. 2022ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understandingcited by 262
  4. 2023ULIP-2: Towards Scalable Multimodal Pre-training for 3D Understandingcited by 199
  5. 2023Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D Understanding, Generation, and Instruction Followingcited by 192
  6. 2022PLA: Language-Driven Open-Vocabulary 3D Scene Understandingcited by 176
  7. 2023Chat-Scene: Bridging 3D Scene and Large Language Models with Object Identifierscited by 170
  8. 2022GAPartNet: Cross-Category Domain-Generalizable Object Perception and Manipulation via Generalizable and Actionable Partscited by 135
  9. 2024SceneVerse: Scaling 3D Vision-Language Learning for Grounded Scene Understandingcited by 130
  10. 2023Open3DIS: Open-Vocabulary 3D Instance Segmentation with 2D Mask Guidancecited by 123
  11. 2022PartSLIP: Low-Shot Part Segmentation for 3D Point Clouds via Pretrained Image-Language Modelscited by 117
  12. 2022Superpoint Transformer for 3D Scene Instance Segmentationcited by 115
38 more
  1. 2023Swin3D: A Pretrained Transformer Backbone for 3D Indoor Scene Understandingcited by 111
  2. 2024Grounded 3D-LLM with Referent Tokenscited by 111
  3. 2025GPT4Scene: Understand 3D Scenes from Videos with Vision-Language Modelscited by 106
  4. 2022EDA: Explicit Text-Decoupling and Dense Alignment for 3D Visual Groundingcited by 98
  5. 2023OneFormer3D: One Transformer for Unified Point Cloud Segmentationcited by 95
  6. 2024SceneScript: Reconstructing Scenes With An Autoregressive Structured Language Modelcited by 95
  7. 2023RegionPLC: Regional Point-Language Contrastive Learning for Open-World 3D Scene Understandingcited by 93
  8. 2025PARTFIELD: Learning 3D Feature Fields for Part Segmentation and Beyondcited by 91
  9. 2023OpenIns3D: Snap and Lookup for 3D Open-vocabulary Instance Segmentationcited by 82
  10. 2023Visual Programming for Zero-shot Open-Vocabulary 3D Visual Groundingcited by 64
  11. 2023Segment3D: Learning Fine-Grained Class-Agnostic 3D Segmentation without Manual Labelscited by 64
  12. 2022CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Cloudscited by 58
  13. 2024MaskClustering: View Consensus based Mask Graph Clustering for Open-Vocabulary 3D Instance Segmentationcited by 56
  14. 2025Open-Vocabulary Functional 3D Scene Graphs for Real-World Indoor Spacescited by 53
  15. 2023ISBNet: a 3D Point Cloud Instance Segmentation Network with Instance-aware Sampling and Box-aware Dynamic Convolutioncited by 52
  16. 20223D-Aware Object Goal Navigation via Simultaneous Exploration and Identificationcited by 46
  17. 2023Mask-Attention-Free Transformer for 3D Instance Segmentationcited by 45
  18. 2024EmbodiedOcc: Embodied 3D Occupancy Prediction for Vision-based Online Scene Understandingcited by 43
  19. 2023PartManip: Learning Cross-Category Generalizable Part Manipulation Policy from Point Cloud Observationscited by 40
  20. 2023Instance Neural Radiance Fieldcited by 39
  21. 2023Lowis3D: Language-Driven Open-World Instance-Level 3D Scene Understandingcited by 38
  22. 2023PartSLIP++: Enhancing Low-Shot 3D Part Segmentation via Multi-View Instance Segmentation and Maximum Likelihood Estimationcited by 37
  23. 2023SAM-guided Graph Cut for 3D Instance Segmentationcited by 35
  24. 2023AGILE3D: Attention Guided Interactive Multi-object 3D Segmentationcited by 33
  25. 2023Four Ways to Improve Verbo-visual Fusion for Dense 3D Visual Groundingcited by 32
  26. 2023Top-Down Beats Bottom-Up in 3D Instance Segmentationcited by 31
  27. 2023UnScene3D: Unsupervised 3D Instance Segmentation for Indoor Scenescited by 31
  28. 2024PO3AD: Predicting Point Offsets toward Better 3D Point Cloud Anomaly Detectioncited by 30
  29. 2025Spatial Understanding from Videos: Structured Prompts Meet Simulation Datacited by 28
  30. 2023SGFormer: Semantic Graph Transformer for Point Cloud-based 3D Scene Graph Generationcited by 27
  31. 2024TAMM: TriAdapter Multi-Modal Learning for 3D Shape Understandingcited by 26
  32. 2024ODIN: A Single Model for 2D and 3D Segmentationcited by 25
  33. 2022OGC: Unsupervised 3D Object Segmentation from Rigid Dynamics of Point Cloudscited by 24
  34. 20223D Segmentation of Humans in Point Clouds with Synthetic Datacited by 24
  35. 2023UrbanBIS: a Large-scale Benchmark for Fine-grained Urban Building Instance Segmentationcited by 24
  36. 2023Interactive Segment Anything NeRF with Feature Imitationcited by 24
  37. 2023Mask4Former: Mask Transformer for 4D Panoptic Segmentationcited by 24
  38. 2023Spherical Mask: Coarse-to-Fine 3D Point Cloud Instance Segmentation with Spherical Representationcited by 23

50 most cited of 197 All 197 (JSON)