mlsanityData to 7 Oct 2026

Segment Any 3D Gaussians

arXiv 2312.00860 ↗First posted 1 Dec 2023cs.CVcited by 158

Not on the map. Why?

Abstract (arXiv)

This paper presents SAGA (Segment Any 3D GAussians), a highly efficient 3D promptable segmentation method based on 3D Gaussian Splatting (3D-GS). Given 2D visual prompts as input, SAGA can segment the corresponding 3D target represented by 3D Gaussians within 4 ms. This is achieved by attaching an scale-gated affinity feature to each 3D Gaussian to endow it a new property towards multi-granularity segmentation. Specifically, a scale-aware contrastive training strategy is proposed for the scale-gated affinity feature learning. It 1) distills the segmentation capability of the Segment Anything Model (SAM) from 2D masks into the affinity features and 2) employs a soft scale gate mechanism to deal with multi-granularity ambiguity in 3D segmentation through adjusting the magnitude of each feature channel according to a specified 3D physical scale. Evaluations demonstrate that SAGA achieves real-time multi-granularity segmentation with quality comparable to state-of-the-art methods. As one of the first methods addressing promptable segmentation in 3D-GS, the simplicity and effectiveness of SAGA pave the way for future advancements in this field. Our code will be released.

Key facts. Cited by 158 papers in the map (data through 7 Oct 2026). Builds on 51 of its 65 references that are in the map. First posted on arXiv 1 Dec 2023.

Builds on

51 of 65 references.

References · 51

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2023Segment Anythingcited by 8,158
  3. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,893
  4. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,141
  5. 2022Instant Neural Graphics Primitives with a Multiresolution Hash Encodingcited by 3,698
  6. 2023Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detectioncited by 3,274
  7. 2021Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fieldscited by 2,013
  8. 2021Plenoxels: Radiance Fields without Neural Networkscited by 1,607
  9. 2022TensoRF: Tensorial Radiance Fieldscited by 1,275
  10. 20234D Gaussian Splatting for Real-Time Dynamic Scene Renderingcited by 1,144
  11. 2021Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstructioncited by 961
  12. 2019The Replica Dataset: A Digital Replica of Indoor Spacescited by 885
38 more
  1. 2020GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fieldscited by 784
  2. 2019Local Light Field Fusion: Practical View Synthesis with Prescriptive Sampling Guidelinescited by 672
  3. 2022OpenScene: 3D Scene Understanding with Open Vocabulariescited by 556
  4. 2023Segment Everything Everywhere All at Oncecited by 546
  5. 2023LERF: Language Embedded Radiance Fieldscited by 538
  6. 2021Baking Neural Radiance Fields for Real-Time View Synthesiscited by 465
  7. 2021In-Place Scene Labelling and Understanding with Implicit Scene Representationcited by 411
  8. 2022Decomposing NeRF for Editing via Feature Field Distillationcited by 370
  9. 2022Fast Dynamic Radiance Fields with Time-Aware Neural Voxelscited by 344
  10. 2023ConceptFusion: Open-set Multimodal 3D Mappingcited by 302
  11. 2022Unsupervised Semantic Segmentation by Distilling Feature Correspondencescited by 241
  12. 2021NeX: Real-time View Synthesis with Neural Basis Expansioncited by 215
  13. 2020AutoInt: Automatic Integration for Fast Neural Volume Renderingcited by 204
  14. 2022Neural Feature Fusion Fields: 3D Distillation of Self-Supervised 2D Image Representationscited by 195
  15. 2022PLA: Language-Driven Open-Vocabulary 3D Scene Understandingcited by 176
  16. 2021Reviving Iterative Training with Mask Guidance for Interactive Segmentationcited by 162
  17. 2022Panoptic NeRF: 3D-to-2D Label Transfer for Panoptic Urban Scene Segmentationcited by 152
  18. 2022SPIn-NeRF: Multiview Segmentation and Perceptual Inpainting with Neural Radiance Fieldscited by 144
  19. 2023Weakly Supervised 3D Open-vocabulary Segmentationcited by 133
  20. 2021NeSF: Neural Semantic Fields for Generalizable Semantic Segmentation of 3D Scenescited by 119
  21. 2022PartSLIP: Low-Shot Part Segmentation for 3D Point Clouds via Pretrained Image-Language Modelscited by 117
  22. 2022SimpleClick: Interactive Image Segmentation with Simple Vision Transformerscited by 109
  23. 2022FocalClick: Towards Practical Interactive Image Segmentationcited by 101
  24. 2021Unsupervised Discovery of Object Radiance Fieldscited by 100
  25. 2022Semantic Abstraction: Open-World 3D Scene Understanding from 2D Vision-Language Modelscited by 95
  26. 2023RegionPLC: Regional Point-Language Contrastive Learning for Open-World 3D Scene Understandingcited by 93
  27. 2023CLIP-FO3D: Learning Free Open-world 3D Scene Representations from 2D Dense CLIPcited by 89
  28. 2021Decomposing 3D Scenes into Objects via Unsupervised Volume Segmentationcited by 88
  29. 2022NeRF-Supervision: Learning Dense Object Descriptors from Neural Radiance Fieldscited by 86
  30. 2023OmniSeg3D: Omniversal 3D Segmentation via Hierarchical Contrastive Learningcited by 75
  31. 2022NeRF-SOS: Any-View Self-supervised Object Segmentation on Complex Scenescited by 71
  32. 2022DM-NeRF: 3D Scene Geometry Decomposition and Manipulation from 2D Imagescited by 69
  33. 2022Neural Volumetric Object Selectioncited by 61
  34. 2023Segment Anything in 3D with Radiance Fieldscited by 61
  35. 2022Interactive Segmentation of Radiance Fieldscited by 60
  36. 2023Instance Neural Radiance Fieldcited by 39
  37. 2023Interactive Segment Anything NeRF with Feature Imitationcited by 24
  38. 2022Unsupervised Multi-View Object Segmentation Using Radiance Field Propagationcited by 22

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 1; 2024: 47; 2025: 58; 2026: 52.

158 papers.

Sort:

By area:

  1. 2024A Survey on 3D Gaussian Splattingcited by 221
  2. 2024OpenGaussian: Towards Point-Level 3D Gaussian-based Open Vocabulary Understandingcited by 141
  3. 2024Recent Advances in 3D Gaussian Splattingcited by 130
  4. 20243D Gaussian as a New Era: A Surveycited by 102
  5. 2024DGE: Direct Gaussian 3D Editing by Consistent Multi-view Editingcited by 92
  6. 2024MVSGaussian: Fast Generalizable Gaussian Splatting Reconstruction from Multi-View Stereocited by 86
  7. 2024Semantic Gaussians: Open-Vocabulary Scene Understanding with 3D Gaussian Splattingcited by 81
  8. 2024GS2Mesh: Surface Reconstruction from Gaussian Splatting via Novel Stereo Viewscited by 65
  9. 2024FlashSplat: 2D to 3D Gaussian Splatting Segmentation Solved Optimallycited by 57
  10. 2024InFusion: Inpainting 3D Gaussians via Learning Depth Completion from Diffusion Priorcited by 55
  11. 2024Point-SAM: Promptable 3D Segmentation Model for Point Cloudscited by 49
  12. 2024SemGauss-SLAM: Dense Semantic Gaussian Splatting SLAMcited by 41
38 more
  1. 2024Gaga: Group Any Gaussians via 3D-aware Memory Bankcited by 40
  2. 2024Gaussian Splatting: 3D Reconstruction and Novel View Synthesis, a Reviewcited by 40
  3. 2024On the Error Analysis of 3D Gaussian Splatting and an Optimal Projection Strategycited by 39
  4. 2024LUDVIG: Learning-Free Uplifting of 2D Visual Features to Gaussian Splatting Scenescited by 36
  5. 2024InstanceGaussian: Appearance-Semantic Joint Gaussian Representation for 3D Instance-Level Perceptioncited by 35
  6. 2024N2F2: Hierarchical Scene Understanding with Nested Neural Feature Fieldscited by 34
  7. 2024Moving Object Segmentation: All You Need Is SAM (and Flow)cited by 30
  8. 2024SAGD: Boundary-Enhanced Segment Anything in 3D Gaussian via Gaussian Decompositioncited by 29
  9. 2024GaussianCut: Interactive segmentation via graph cut for 3D Gaussian Splattingcited by 25
  10. 2025ObjectGS: Object-aware Scene Reconstruction and Scene Understanding via Gaussian Splattingcited by 24
  11. 2024StylizedGS: Controllable Stylization for 3D Gaussian Splattingcited by 23
  12. 2025Lifting by Gaussians: A Simple, Fast and Flexible Method for 3D Instance Segmentationcited by 23
  13. 2024Contrastive Gaussian Clustering: Weakly Supervised 3D Scene Segmentationcited by 22
  14. 2024MirrorGaussian: Reflecting 3D Gaussians for Reconstructing Mirror Reflectionscited by 21
  15. 20243DSceneEditor: Controllable 3D Scene Editing with Gaussian Splattingcited by 20
  16. 2025PUGS: Zero-shot Physical Understanding with Gaussian Splattingcited by 20
  17. 2025ReferSplat: Referring Segmentation in 3D Gaussian Splattingcited by 20
  18. 2024Segment Any 4D Gaussianscited by 19
  19. 2025COB-GS: Clear Object Boundaries in 3DGS Segmentation Based on Boundary-Adaptive Gaussian Splittingcited by 19
  20. 2025A Survey on 3D Gaussian Splatting Applications: Segmentation, Editing, and Generationcited by 19
  21. 2024PLGS: Robust Panoptic Lifting with 3D Gaussian Splattingcited by 17
  22. 2024SLGaussian: Fast Language Gaussian Splatting in Sparse Viewscited by 17
  23. 2025GaussianArt: Unified Modeling of Geometry and Motion for Articulated Objectscited by 17
  24. 2025DecoupledGaussian: Object-Scene Decoupling for Physics-Based Interactioncited by 16
  25. 20244D Gaussian Splatting: Modeling Dynamic Scenes with Native 4D Primitivescited by 15
  26. 2025Tackling View-Dependent Semantics in 3D Language Gaussian Splattingcited by 15
  27. 2025SemanticSplat: Feed-Forward 3D Scene Understanding with Language-Aware Gaussian Fieldscited by 13
  28. 20232D-Guided 3D Gaussian Segmentationcited by 12
  29. 2024Unsupervised Discovery of Object-Centric Neural Fieldscited by 12
  30. 2024Reference-based Controllable Scene Stylization with Gaussian Splattingcited by 12
  31. 20244-LEGS: 4D Language Embedded Gaussian Splattingcited by 12
  32. 2024Open-Vocabulary Online Semantic Mapping for SLAMcited by 11
  33. 2025iSegMan: Interactive Segment-and-Manipulate 3D Gaussianscited by 11
  34. 2024GradiSeg: Gradient-Guided Gaussian Segmentation with Enhanced 3D Boundary Precisioncited by 10
  35. 2024SAMa: Material-aware 3D Selection and Segmentationcited by 9
  36. 2025PE3R: Perception-Efficient 3D Reconstructioncited by 9
  37. 2025TexGS-VolVis: Expressive Scene Editing for Volume Visualization via Textured Gaussian Splattingcited by 9
  38. 2024Self-supervised Learning of Hybrid Part-aware 3D Representations of 2D Gaussians and Superquadricscited by 8

50 most cited of 158 All 158 (JSON)