mlsanityData to 7 Oct 2026

BSNet: Box-Supervised Simulation-assisted Mean Teacher for 3D Instance Segmentation

arXiv 2403.15019 ↗First posted 22 Mar 2024cs.CVcited by 9

Not on the map. Why?

Abstract (arXiv)

3D instance segmentation (3DIS) is a crucial task, but point-level annotations are tedious in fully supervised settings. Thus, using bounding boxes (bboxes) as annotations has shown great potential. The current mainstream approach is a two-step process, involving the generation of pseudo-labels from box annotations and the training of a 3DIS network with the pseudo-labels. However, due to the presence of intersections among bboxes, not every point has a determined instance label, especially in overlapping areas. To generate higher quality pseudo-labels and achieve more precise weakly supervised 3DIS results, we propose the Box-Supervised Simulation-assisted Mean Teacher for 3D Instance Segmentation (BSNet), which devises a novel pseudo-labeler called Simulation-assisted Transformer. The labeler consists of two main components. The first is Simulation-assisted Mean Teacher, which introduces Mean Teacher for the first time in this task and constructs simulated samples to assist the labeler in acquiring prior knowledge about overlapping areas. To better model local-global structure, we also propose Local-Global Aware Attention as the decoder for teacher and student labelers. Extensive experiments conducted on the ScanNetV2 and S3DIS datasets verify the superiority of our designs. Code is available at hrefhttps://github.com/peoplelu/BSNethttps://github.com/peoplelu/BSNet.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 42 of its 53 references that are in the map. First posted on arXiv 22 Mar 2024.

Builds on

42 of 53 references.

References · 42

  1. 2020End-to-End Object Detection with Transformerscited by 7,384
  2. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  3. 2021Masked-attention Mask Transformer for Universal Image Segmentationcited by 2,338
  4. 2016V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentationcited by 2,284
  5. 2021Per-Pixel Classification is Not All You Need for Semantic Segmentationcited by 1,148
  6. 20173D Semantic Segmentation with Submanifold Sparse Convolutional Networkscited by 746
  7. 2019Deep Hough Voting for 3D Object Detection in Point Cloudscited by 648
  8. 2020PointContrast: Unsupervised Pre-training for 3D Point Cloud Understandingcited by 502
  9. 2021An End-to-End Transformer Model for 3D Object Detectioncited by 329
  10. 2020PointGroup: Dual-Set Point Grouping for 3D Instance Segmentationcited by 318
  11. 2021End-to-End Semi-Supervised Object Detection with Soft Teachercited by 315
  12. 2021DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentationcited by 299
30 more
  1. 2022Mask3D: Mask Transformer for 3D Semantic Instance Segmentationcited by 299
  2. 2021Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentationcited by 268
  3. 2021Group-Free 3D Object Detection via Transformerscited by 215
  4. 2020Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identificationcited by 214
  5. 2020Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contextscited by 210
  6. 2022SoftGroup for 3D Instance Segmentation on Point Cloudscited by 197
  7. 2019Learning Object Bounding Boxes for 3D Instance Segmentation on Point Cloudscited by 156
  8. 2021Self-supervised Augmentation Consistency for Adapting Semantic Segmentationcited by 149
  9. 2018GSPN: Generative Shape Proposal Network for 3D Instance Segmentation in Point Cloudcited by 145
  10. 20203D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentationcited by 140
  11. 2021Hierarchical Aggregation for 3D Instance Segmentationcited by 126
  12. 2021Cross-Domain Adaptive Teacher for Object Detectioncited by 125
  13. 2022Superpoint Transformer for 3D Scene Instance Segmentationcited by 115
  14. 2021Instance Segmentation in 3D Scenes using Semantic Superpoint Tree Networkscited by 98
  15. 2021FCAF3D: Fully Convolutional Anchor-Free 3D Object Detectioncited by 97
  16. 2022CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Cloudscited by 58
  17. 2023ISBNet: a 3D Point Cloud Instance Segmentation Network with Instance-aware Sampling and Box-aware Dynamic Convolutioncited by 52
  18. 2023Contrastive Mean Teacher for Domain Adaptive Object Detectorscited by 52
  19. 2020Exploiting Sample Uncertainty for Domain Adaptive Person Re-Identificationcited by 50
  20. 2022Consistent-Teacher: Towards Reducing Inconsistent Pseudo-targets in Semi-supervised Object Detectioncited by 43
  21. 2023Active Teacher for Semi-Supervised Object Detectioncited by 43
  22. 2021RandomRooms: Unsupervised Pre-training from Synthetic Shapes and Randomized Layouts for 3D Object Detectioncited by 37
  23. 20223D Instances as 1D Kernelscited by 36
  24. 2020Learning Gaussian Instance Segmentation in Point Cloudscited by 33
  25. 2022MaskGroup: Hierarchical Point Grouping and Masking for 3D Instance Segmentationcited by 29
  26. 2022Omni-DETR: Omni-Supervised Object Detection with Transformerscited by 28
  27. 2023SE-ORNet: Self-Ensembling Orientation-aware Network for Unsupervised Point Cloud Shape Correspondencecited by 19
  28. 2023ALWOD: Active Learning for Weakly-Supervised Object Detectioncited by 8
  29. 2021Delving into Probabilistic Uncertainty for Unsupervised Domain Adaptive Person Re-Identificationcited by 7
  30. 2023GaPro: Box-Supervised 3D Point Cloud Instance Segmentation Using Gaussian Processes as Pseudo Labelerscited by 6

Built on by

Cited by year (2026 to 7 Oct)

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

9 papers.

  1. 2024Instance-Adaptive and Geometric-Aware Keypoint Learning for Category-Level 6D Object Pose Estimationcited by 41
  2. 2024DN-4DGS: Denoised Deformable Network with Temporal-Spatial Aggregation for Dynamic Scene Renderingcited by 28
  3. 2025Relation3D: Enhancing Relation Modeling for Point Cloud Instance Segmentationcited by 10
  4. 2025SAS: Segment Any 3D Scene with Integrated 2D Priorscited by 5
  5. 2025LiDAR Remote Sensing Meets Weak Supervision: Concepts, Methods, and Perspectivescited by 4
  6. 2025Sketchy Bounding-box Supervision for 3D Instance Segmentationcited by 4
  7. 2025Beyond the Final Layer: Hierarchical Query Fusion Transformer with Agent-Interpolation Initialization for 3D Instance Segmentationcited by 3
  8. 2025BEEP3D: Box-Supervised End-to-End Pseudo-Mask Generation for 3D Instance Segmentationcited by 1
  9. 2026GeoGuide: Hierarchical Geometric Guidance for Open-Vocabulary 3D Semantic Segmentationcited by 1