Bootstraping Clustering of Gaussians for View-consistent 3D Scene Understanding

Huchuan Lu; Liqian Ma; Lu Zhang; Ping Hu; Wenbo Zhang; Yunzhi Zhuge

REVIEW 4 major objections 6 minor 38 references

Bootstraping Clustering of Gaussians for View-consistent 3D Scene Understanding

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read FreeGS learns view-consistent 3D scene semantics from raw multi-view images alone, with no 2D labels, no SAM masks, and no preprocessing.

desk verdict Clever bootstrapping design that delivers competitive label-free 3D segmentation, but the clustering engine is unquantified and the view-consistency claim outruns the evidence. read the letter →

arxiv 2411.19551 v2 pith:YT2RHEYZ submitted 2024-11-29 cs.CV cs.LG

Wenbo Zhang , Lu Zhang , Ping Hu , Liqian Ma , Yunzhi Zhuge , Huchuan Lu This is my paper · ORCID

classification cs.CVcs.LG

keywords 3DGaussianSplattingunsupervisedsceneunderstandingview-consistentsemanticsinstanceclusteringopen-vocabularysegmentationobjectdetectionbootstrappingsemanticdistillation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

FreeGS claims that a 3D Gaussian scene representation can learn view-consistent semantics and instance identities from raw multi-view images alone, with no 2D labels, no SAM masks, and no tracking-based preprocessing. The paper argues that alternating between clustering Gaussians in a joint geometry–appearance–semantics space and distilling CLIP features into the resulting instance groups lets the two tasks improve each other. If the claim holds, open-vocabulary novel-view segmentation, text- and click-driven object selection, and 3D object detection all become possible using only the images and camera poses already needed for reconstruction. The reported experiments show performance comparable to supervised counterparts on public benchmarks while removing the preprocessing step entirely.

What carries the argument

The identity-coupled semantic field (IDSF) is the central object: each Gaussian stores a semantic vector $f_i \in \mathbb{R}^{128}$ and a scalar instance index $d_i$, so identity and semantics live on the same primitive. The engine is the alternating bootstrap. Union-space clustering runs HDBSCAN on the concatenation of normalized position, color, and PCA-reduced semantics ($S_{\text{union}} = S_{\text{pos}} \cup S_{\text{app}} \cup S_{\text{sem}}$) to produce instance indices; the indices are $\alpha$-blended and argmaxed into a 2D ID map, refined by DenseCRF, and used to form instance-level CLIP supervision. The multi-level distillation loss $L_F = \|F_{\text{pix}} - \hat{F}\|_1 + \gamma \|F_{\text{ins}} - F\|_1$ injects pixel- and instance-level semantics, and the 2D–3D joint contrastive loss $L_C$ re-merges fragments by treating similar groups as positive samples. Feature-field smoothing $L_S$ over $K$-nearest neighbors keeps the field continuous, and together these losses close the loop that lets clustering and semantics refine each other.

What would settle it

Take a scene with two visually near-identical, touching objects of the same category, and run FreeGS from unlabeled multi-view images. If the union-space clustering initially merges them into one instance and the bootstrapped distillation plus contrastive loss never separates them over the full 7k-iteration second phase, the central claim that clustering and semantics bootstrap toward view-consistent instances fails on that scene. A quantitative version: track cluster purity against ground-truth instance masks during training; the paper's claim predicts a monotone improvement, so a non-increasing or oscillating purity trajectory would contradict it.

Watch

Extended reading notes

Core claim

The central discovery is that cross-view instance consistency can be bootstrapped rather than imported from 2D models. FreeGS attaches to each Gaussian an identity-coupled semantic field: a view-independent semantic vector plus a scalar instance index. A clustering pass groups Gaussians in a union space of position, color, and PCA-reduced semantics, assigning each Gaussian an instance index; a distillation pass renders those indices to 2D, uses the resulting masks to extract CLIP features per instance, and aligns them with the rendered semantic field. A 2D–3D joint contrastive loss pulls features of the same clustered group together and pushes different groups apart. The paper argues that this alternating schedule yields view-consistent semantics without any label, because the 3D grouping regularizes the 2D semantic injection and the semantic field in turn improves the grouping.

Load-bearing premise

The load-bearing premise is that HDBSCAN clustering of position, color, and partially trained semantics produces correct object groupings, and that masks derived from those groupings give supervision that improves the semantics rather than locking in initial clustering errors.

Editorial extensions

If this is right

On LERF-Mask and 3D-OVS, FreeGS matches the segmentation mIoU of SAM-supervised methods like Gaussian Grouping and LangSplat, without generating any SAM masks.
The same trained field answers text and click prompts by matching in 3D, so novel-view segmentation, object selection, and detection share one inference pipeline.
Removing preprocessing cuts the total pipeline from 92–246 minutes to about 62 minutes on the ramen scene, making the semantic field practical for quick 3D capture.
Because instance indices are stored per Gaussian, 3D editing and prompt-based selection become direct manipulations of the cluster rather than pixel-space postprocessing.
The contrastive loss stabilizes cluster assignments across training iterations, which is what the paper credits for avoiding fragmented objects in cluttered scenes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural stress test is a scene with two visually near-identical adjacent objects: the union-space clustering has no appearance or semantic signal to separate them initially, so whether the bootstrapping loop can ever split them is an open question not addressed by the chosen benchmarks.
The union-space bootstrapping principle is not tied to Gaussian splatting; any explicit 3D representation with a differentiable renderer and a 2D feature extractor could in principle run the same alternating clustering–distillation loop.
The paper fixes PCA dimension and HDBSCAN min_samples without a sensitivity study; measuring how cluster purity and downstream mIoU vary with those hyperparameters would directly test how much of the result rests on the clustering configuration.
The authors acknowledge 3D detection accuracy has significant room for improvement, so the framework's most immediate practical value is likely segmentation and object selection, with precise 3D localization as the next target for the same bootstrapping idea.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Hi [name],

Quick take on arXiv:2411.19551: the paper has a genuinely new mechanism—the union-space clustering bootstrapped with alternating 2D distillation—and it delivers results that are competitive with SAM-supervised baselines while skipping preprocessing. That is a real practical contribution. The code is out, the ablations show each component pulls weight, and the writing is honest about limitations. I'd send it to referees.

What's new: instead of requiring SAM masks or view-consistent labels, they add a per-Gaussian semantic vector and instance index (IDSF) and optimize them alternately. HDBSCAN clusters Gaussians in position/color/PCA-reduced semantic space to get instance indices; those indices are rendered to 2D to produce pseudo-masks that constrain the CLIP distillation, and a joint 2D-3D contrastive loss stabilizes the loop. It's a clean bootstrapping design, and the efficiency gain over LangSplat is real (141 min preprocessing to zero).

The soft spots are in the empirical support. The central claim is view consistency, but they never measure it directly—no cross-view segmentation stability metric, no cluster accuracy over training. The clustering quality is the engine of the bootstrap, and it's completely unquantified. The stress-test note about noise points is fair: HDBSCAN assigns noise labels, but the paper never says how those Gaussians are treated in the one-hot index or the ID map. That could pollute masks with a spurious "noise instance." Also, there are only 3-5 scenes per benchmark, no error bars or multiple seeds, and the HDBSCAN min_samples is tuned per dataset (20 vs 60), so the method's robustness to its most important hyperparameter is unknown. Those are real gaps, not fatal flaws: the ablations do show the designed components matter, and the qualitative results look consistent.

Overall: this is a solid method paper with a clever core and a fair shot at being useful. A serious referee should require an added cluster-quality analysis, noise-point handling details, and at least a sensitivity study on min_samples. I'd accept it for review rather than desk-reject, and I'd probably cite it if I were working on label-free 3D scene understanding.

Referee Report

4 major / 6 minor

Summary. FreeGS proposes an unsupervised 3D Gaussian Splatting framework for view-consistent scene understanding. The core idea is an IDentity-coupled Semantic Field (IDSF) that augments each Gaussian with a semantic vector and a scalar instance index. Training alternates between HDBSCAN clustering of Gaussians in a position-color-semantics union space to obtain pseudo-instance indices, and multi-level 2D semantic distillation from frozen CLIP/MaskCLIP features guided by these indices; a 2D-3D contrastive loss is added to stabilize the loop. The paper evaluates on LERF-Mask, 3D-OVS, and ScanNet for novel-view segmentation, 3D detection, and interactive selection, reporting performance comparable to SAM-supervised baselines while eliminating SAM preprocessing.

Significance. If the bootstrapping mechanism is sound, FreeGS is a meaningful advance: it removes the SAM-mask preprocessing bottleneck, introduces an explicit view-consistent instance index in 3D, and unifies several tasks in one field. The paper ships code and includes ablations showing that each component contributes on the tested scenes. The significance is currently tempered by the absence of a direct view-consistency metric, the lack of clustering-quality analysis, and the small evaluation scale; these gaps leave the central claim only partially substantiated.

major comments (4)

[Section 3.2, Eq. (5)] The handling of HDBSCAN noise labels is unspecified. HDBSCAN assigns a noise label to low-density points, but the paper never states how such Gaussians are treated when constructing the one-hot vector d'_i or when rendering the ID map in Eq. (5). If noise points are assigned a common index, they will appear as a spurious instance in the argmax map; if they are excluded, the definition of the rendered index and of the masks M_i in Eq. (6) is incomplete. This is not a cosmetic detail: Eq. (6) and Eq. (8) use these masks and indices directly, so the ambiguity affects the core bootstrapping loop.
[Section 3.2 and Section 4 (Training)] The paper provides no evaluation of clustering quality or of bootstrap error propagation, despite the method's reliance on union-space clustering as the engine of the alternating optimization. At the start of the second training phase, the semantic features f_i are not yet trained, so the first clustering is driven almost entirely by position and color; the paper does not show that this initialization is sufficient, nor that the smoothing loss (Eq. 3) and contrastive loss (Eq. 8) improve rather than reinforce initial grouping errors. A quantitative analysis of cluster purity/accuracy and of how the pseudo-masks evolve over training iterations is needed to substantiate the claim that the two steps 'bootstrap view-consistent semantics.'
[Section 4, Tables 1-3] The view-consistency claim is not directly measured. The reported metrics (per-frame mIoU for segmentation, 3D box mIoU and recall for detection) evaluate average accuracy, not the stability of instance identities across views. To support the title and abstract, the paper should report a direct consistency measure, for example the mean pairwise IoU of rendered ID maps across different viewpoints, or the fraction of Gaussians assigned to the same cluster when the process is repeated with different training seeds. Without such a measure, the qualitative claim of consistency (e.g., Fig. 3) is not quantified.
[Section 4, Implementation Details and Tables 1-3] The evaluation is too small to support the strength of the comparisons, and no variance is reported. The benchmarks comprise 3, 5, and 3 scenes, and differences such as Table 1's 69.5 versus 72.8 and Table 2's 77.0 versus 77.0 are within the noise expected for this sample size. Moreover, the HDBSCAN min_samples hyperparameter is tuned per dataset (20 for LERF-Mask, 60 for 3D-OVS and ScanNet) with no sensitivity analysis, even though clustering is the load-bearing component of the method. At minimum, multiple runs with standard deviations and a sensitivity study for min_samples and the loss weights (lambda_C, lambda_S) should be reported.

minor comments (6)

[Title and Section 3.3] 'Bootstraping' should be 'Bootstrapping' in the title and in the heading of Section 3.3.
[Section 3.2] 'Sapp is the gather of view-independent color part c′' should read 'Sapp is the set of view-independent color components c′'.
[Eq. (8)] The notation for the positive and negative sets (F Gi+ and F Gi-) is introduced as sets of mean features, but the overbar on F is not defined; please clarify the notation.
[Table 2] The caption says the LERF results are taken from LangSplat, which conflicts with the text stating that methods were retrained on training views; please clarify the protocol.
[Fig. 6] The interactive object selection task is presented qualitatively; a short protocol description or a small user study would strengthen the claim that the method supports this task.
[Section 4] 'Gasussians' should be 'Gaussians'; also, 'min samples' should be 'min_samples' for consistency with the HDBSCAN library naming.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the semantic supervision comes from external CLIP/MaskCLIP features, and the alternating clustering/distillation loop is iterative bootstrapping, not a definitional reduction.

full rationale

The paper's derivation chain is self-contained with respect to external evidence. Union-space clustering (Sec. 3.2) groups Gaussians using position, color, and PCA-reduced semantic features, and the resulting instance indices are rendered into 2D masks via Eq. (5). These masks are then used only to select image patches for the frozen CLIP encoder in the instance-level distillation term (Eq. 6), while pixel-level distillation uses external MaskCLIP and FeatUP features. None of the losses is minimized toward a target that is, by construction, the same function of the predicted field; the mutual update of semantic features and instance indices is an explicit iterative bootstrap, not a fitted parameter renamed as a prediction. The paper also does not rely on self-citation chains or imported uniqueness theorems. The concern that HDBSCAN cluster quality is unevaluated and that errors could self-reinforce is an empirical robustness and correctness risk, not a circularity, because no equation reduces the claimed output to its own input by construction.

Assumptions & free parameters 4 free parameters · 4 assumptions · 1 invented entities

The central claim relies on external CLIP features and on the empirical stability of the alternating clustering-distillation loop. The per-dataset HDBSCAN min_samples is a free parameter that directly affects the reported metrics. No new physical entities are introduced; IDSF is a software construct.

free parameters (4)

HDBSCAN min_samples = 20 (LERF-Mask); 60 (3D-OVS and ScanNet)
Tuned per dataset; directly controls the granularity of object grouping and thus segmentation quality.
loss weights = lambda_C=0.05, lambda_S=0.1, gamma=0.3
Chosen by hand to balance distillation, contrastive, and smoothing losses.
contrastive temperature tau = 0.1
Set in all experiments; controls the softness of the contrastive distribution.
feature dimension D = 128
Dimensionality of the per-Gaussian semantic vector.

assumptions (4)

domain assumption CLIP and MaskCLIP features carry semantically meaningful and view-independent visual representations.
The method distills these features into 3D Gaussians; if CLIP features are unreliable in novel views, the bootstrapping inherits errors. Invoked throughout Section 3.3.
domain assumption Clustering in the union space of position, color, and PCA-reduced semantics yields object-level groups.
The correctness of instance indices depends on HDBSCAN's ability to find clusters in this space; Section 3.2.
domain assumption FeatUP super-resolution preserves feature semantics at higher resolution.
Pixel-level distillation uses FeatUP to upsample MaskCLIP features; grid artifacts and semantic mixups are assumed mitigated.
ad hoc to paper The alternating optimization converges to a non-trivial fixed point rather than a degenerate grouping.
No convergence proof is given; the success relies on empirical stability, e.g., applying losses every 10 iterations. Section 3.4.

invented entities (1)

IDentity-coupled Semantic Field (IDSF)
purpose: Attaches a semantic vector and a view-consistent instance index to each Gaussian.
A new latent representation proposed in the paper; no external falsifiable handle.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Bootstraping Clustering of Gaussians for View-consistent 3D Scene Understanding." pith.science (2026). https://pith.science/paper/YT2RHEYZ

@misc{pith2026241119551,
  author       = {Pith},
  title        = {Pith review of: Bootstraping Clustering of Gaussians for View-consistent 3D Scene Understanding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YT2RHEYZ}},
  note         = {Machine review of arXiv:2411.19551}
}

read the original abstract

Injecting semantics into 3D Gaussian Splatting (3DGS) has recently garnered significant attention. While current approaches typically distill 3D semantic features from 2D foundational models (e.g., CLIP and SAM) to facilitate novel view segmentation and semantic understanding, their heavy reliance on 2D supervision can undermine cross-view semantic consistency and necessitate complex data preparation processes, therefore hindering view-consistent scene understanding. In this work, we present FreeGS, an unsupervised semantic-embedded 3DGS framework that achieves view-consistent 3D scene understanding without the need for 2D labels. Instead of directly learning semantic features, we introduce the IDentity-coupled Semantic Field (IDSF) into 3DGS, which captures both semantic representations and view-consistent instance indices for each Gaussian. We optimize IDSF with a two-step alternating strategy: semantics help to extract coherent instances in 3D space, while the resulting instances regularize the injection of stable semantics from 2D space. Additionally, we adopt a 2D-3D joint contrastive loss to enhance the complementarity between view-consistent 3D geometry and rich semantics during the bootstrapping process, enabling FreeGS to uniformly perform tasks such as novel-view semantic segmentation, object selection, and 3D object detection. Extensive experiments on LERF-Mask, 3D-OVS, and ScanNet datasets demonstrate that FreeGS performs comparably to state-of-the-art methods while avoiding the complex data preprocessing workload. Our code is publicly available at https://github.com/wb014/FreeGS.

Figures

Figures reproduced from arXiv: 2411.19551 by the authors.

**Figure 1.** We present a novel framework, FreeGS, to inject 2D semantics ( ) into 3DGS ( ), without the need of any 2D labels. After a 2D-3D collaborative bootstrapping learning strategy, the model can support versatile applications, such as novel-view 2D segmentation, open-vocabulary 3D detection, and interactive object selection. 2023) have integrated learnable semantic fields into NeRF, performing 3D feature distillation fr… view at source ↗

**Figure 2.** Framework overview of FreeGS. The framework consists of three key components: Union-space 3D Gaussian Clustering, Multi-level 2D Semantic Distillation, and 2D-3D Joint Contrastive Learning. In the 3D space, Gaussians equipped with the IDentity-coupled Semantic Fields (IDSF) are input to the union-space clustering module to extract view-consistent instance indices. Subsequently, the IDSF is rendered onto 2D space an… view at source ↗

**Figure 3.** Qualitative comparisons of different methods on LERF-Mask and 3D-OVS dataset. Our method successfully segment [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Qualitative comparisons of different methods on “Scene0462” of the ScanNet dataset. We use yellow lines to draw [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 6.** Figure 6: Visualization of interactive object selection results [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 11 canonical work pages

[1]

, " * write output.state after.block = add.period write newline

ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...
[2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...
[3]

Caron, M.; Touvron, H.; Misra, I.; J \'e gou, H.; Mairal, J.; Bojanowski, P.; and Joulin, A. 2021. Emerging properties in self-supervised vision transformers. In Proceedings of the IEEE/CVF international conference on computer vision, 9650--9660

2021
[4]

K.; Oh, S

Cheng, H. K.; Oh, S. W.; Price, B.; Schwing, A.; and Lee, J.-Y. 2023. Tracking anything with decoupled video segmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 1316--1326

2023
[5]

X.; Savva, M.; Halber, M.; Funkhouser, T.; and Nie ner, M

Dai, A.; Chang, A. X.; Savva, M.; Halber, M.; Funkhouser, T.; and Nie ner, M. 2017. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In Proceedings of the IEEE conference on computer vision and pattern recognition, 5828--5839

2017
[6]

Dong, X.; Bao, J.; Zheng, Y.; Zhang, T.; Chen, D.; Yang, H.; Zeng, M.; Zhang, W.; Yuan, L.; Chen, D.; et al. 2023. Maskclip: Masked self-distillation advances contrastive language-image pretraining. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 10995--11005

work page 2023
[7]

Fan, Z.; Wang, P.; Jiang, Y.; Gong, X.; Xu, D.; and Wang, Z. 2022. Nerf-sos: Any-view self-supervised object segmentation on complex scenes. arXiv preprint arXiv:2209.08776

arXiv 2022
[8]

Fang, G.; and Wang, B. 2024. Mini-Splatting: Representing Scenes with a Constrained Number of Gaussians. arXiv preprint arXiv:2403.14166

arXiv 2024

Show all 38 references

[9]

Fang, J.; Wang, J.; Zhang, X.; Xie, L.; and Tian, Q. 2024. GaussianEditor: Editing 3D Gaussians Delicately with Text Instructions. In CVPR

2024
[10]

Fu, S.; Hamilton, M.; Brandt, L.; Feldman, A.; Zhang, Z.; and Freeman, W. T. 2024. Featup: A model-agnostic framework for features at any resolution. arXiv preprint arXiv:2403.10516

2024 arXiv
[11]

Hu, X.; Wang, Y.; Fan, L.; Fan, J.; Peng, J.; Lei, Z.; Li, Q.; and Zhang, Z. 2024. Semantic Anything in 3D Gaussians. arXiv preprint arXiv:2401.17857

2024 arXiv
[12]

Huang, J.; and Yu, H. 2023. Point'n Move: Interactive Scene Object Manipulation on Gaussian Splatting Radiance Fields. arXiv preprint arXiv:2311.16737

2023 arXiv
[13]

Kerbl, B.; Kopanas, G.; Leimk \"u hler, T.; and Drettakis, G. 2023. 3d gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics, 42(4): 1--14

2023
[14]

M.; Goldberg, K.; Kanazawa, A.; and Tancik, M

Kerr, J.; Kim, C. M.; Goldberg, K.; Kanazawa, A.; and Tancik, M. 2023. Lerf: Language embedded radiance fields. In ICCV, 19729--19739

2023
[15]

C.; Lo, W.-Y.; et al

Kirillov, A.; Mintun, E.; Ravi, N.; Mao, H.; Rolland, C.; Gustafson, L.; Xiao, T.; Whitehead, S.; Berg, A. C.; Lo, W.-Y.; et al. 2023. Segment anything. In ICCV, 4015--4026

2023
[16]

Kobayashi, S.; Matsumoto, E.; and Sitzmann, V. 2022. Decomposing nerf for editing via feature field distillation. Advances in Neural Information Processing Systems, 35: 23311--23330

2022
[17]

Kopanas, G.; Philip, J.; Leimk \"u hler, T.; and Drettakis, G. 2021. Point-Based Neural Rendering with Per-View Optimization. In Computer Graphics Forum, volume 40, 29--43. Wiley Online Library

2021
[18]

Lan, K.; Li, H.; Shi, H.; Wu, W.; Liao, Y.; Wang, L.; and Zhou, P. 2023. 2d-guided 3d gaussian segmentation. arXiv preprint arXiv:2312.16047

2023 arXiv
[19]

Q.; Belongie, S.; Koltun, V.; and Ranftl, R

Li, B.; Weinberger, K. Q.; Belongie, S.; Koltun, V.; and Ranftl, R. 2022. Language-driven semantic segmentation. arXiv preprint arXiv:2201.03546

2022 arXiv
[20]

Liao, G.; Li, J.; Bao, Z.; Ye, X.; Wang, J.; Li, Q.; and Liu, K. 2024. CLIP-GS: CLIP-Informed Gaussian Splatting for Real-time and View-consistent 3D Semantic Understanding. arXiv preprint arXiv:2404.14249

2024 arXiv
[21]

Liu, K.; Zhan, F.; Zhang, J.; Xu, M.; Yu, Y.; El Saddik, A.; Theobalt, C.; Xing, E.; and Lu, S. 2023. Weakly supervised 3d open-vocabulary segmentation. NeurIPS, 36: 53433--53456

2023
[22]

McInnes, L.; Healy, J.; Astels, S.; et al. 2017. hdbscan: Hierarchical density based clustering. J. Open Source Softw., 2(11): 205

2017
[23]

P.; Tancik, M.; Barron, J

Mildenhall, B.; Srinivasan, P. P.; Tancik, M.; Barron, J. T.; Ramamoorthi, R.; and Ng, R. 2021. Nerf: Representing scenes as neural radiance fields for view synthesis. Communications of the ACM, 65(1): 99--106

2021
[24]

Qin, M.; Li, W.; Zhou, J.; Wang, H.; and Pfister, H. 2023. LangSplat: 3D Language Gaussian Splatting. arXiv preprint arXiv:2312.16084

2023 arXiv
[25]

W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; et al

Radford, A.; Kim, J. W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; et al. 2021. Learning transferable visual models from natural language supervision. In International conference on machine learning, 8748--8763. PMLR

2021
[26]

Ren, J.; Pan, L.; Tang, J.; Zhang, C.; Cao, A.; Zeng, G.; and Liu, Z. 2023. Dreamgaussian4d: Generative 4d gaussian splatting. arXiv preprint arXiv:2312.17142

2023 arXiv
[27]

C.; Dahaghin, M.; Toso, M.; and Del Bue, A

Silva, M. C.; Dahaghin, M.; Toso, M.; and Del Bue, A. 2024. Contrastive Gaussian Clustering: Weakly Supervised 3D Scene Segmentation. arXiv preprint arXiv:2404.12784

2024 arXiv
[28]

M.; and Szeliski, R

Snavely, N.; Seitz, S. M.; and Szeliski, R. 2006. Photo tourism: exploring photo collections in 3D. In ACM siggraph 2006 papers, 835--846

2006
[29]

Tang, J.; Chen, Z.; Chen, X.; Wang, T.; Zeng, G.; and Liu, Z. 2024 a . LGM: Large Multi-View Gaussian Model for High-Resolution 3D Content Creation. arXiv preprint arXiv:2402.05054

2024 arXiv
[30]

Tang, J.; Ren, J.; Zhou, H.; Liu, Z.; and Zeng, G. 2024 b . Dreamgaussian: Generative gaussian splatting for efficient 3d content creation. ICLR

2024
[31]

Tschernezki, V.; Laina, I.; Larlus, D.; and Vedaldi, A. 2022. Neural feature fusion fields: 3d distillation of self-supervised 2d image representations. In 2022 International Conference on 3D Vision (3DV), 443--453. IEEE

2022
[32]

Wu, Z.; Yu, C.; Jiang, Y.; Cao, C.; Wang, F.; and Bai, X. 2024. SC4D: Sparse-Controlled Video-to-4D Generation and Motion Transfer. arXiv preprint arXiv:2404.03736

2024 arXiv
[33]

Yang, C.; Li, S.; Fang, J.; Liang, R.; Xie, L.; Zhang, X.; Shen, W.; and Tian, Q. 2024. GaussianObject: Just Taking Four Images to Get A High-Quality 3D Object with Gaussian Splatting. arXiv preprint arXiv:2402.10259

2024 arXiv
[34]

Ye, M.; Danelljan, M.; Yu, F.; and Ke, L. 2023. Gaussian grouping: Segment and edit anything in 3d scenes. arXiv preprint arXiv:2312.00732

2023 arXiv
[35]

Yi, T.; Fang, J.; Wang, J.; Wu, G.; Xie, L.; Zhang, X.; Liu, W.; Tian, Q.; and Wang, X. 2024. GaussianDreamer: Fast Generation from Text to 3D Gaussians by Bridging 2D and 3D Diffusion Models. In CVPR

2024
[36]

Zeng, Y.; Jiang, Y.; Zhu, S.; Lu, Y.; Lin, Y.; Zhu, H.; Hu, W.; Cao, X.; and Yao, Y. 2024. Stag4d: Spatial-temporal anchored generative 4d gaussians. arXiv preprint arXiv:2403.14939

2024 arXiv
[37]

Zhi, S.; Laidlow, T.; Leutenegger, S.; and Davison, A. J. 2021. In-Place Scene Labelling and Understanding with Implicit Scene Representation. In 2021 IEEE/CVF International Conference on Computer Vision (ICCV)

2021
[38]

Zhou, S.; Chang, H.; Jiang, S.; Fan, Z.; Zhu, Z.; Xu, D.; Chari, P.; You, S.; Wang, Z.; and Kadambi, A. 2024. Feature 3dgs: Supercharging 3d gaussian splatting to enable distilled feature fields. CVPR

2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

, " * write output.state after.block = add.period write newline

ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

[2] [2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

[3] [3]

Caron, M.; Touvron, H.; Misra, I.; J \'e gou, H.; Mairal, J.; Bojanowski, P.; and Joulin, A. 2021. Emerging properties in self-supervised vision transformers. In Proceedings of the IEEE/CVF international conference on computer vision, 9650--9660

2021

[4] [4]

K.; Oh, S

Cheng, H. K.; Oh, S. W.; Price, B.; Schwing, A.; and Lee, J.-Y. 2023. Tracking anything with decoupled video segmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 1316--1326

2023

[5] [5]

X.; Savva, M.; Halber, M.; Funkhouser, T.; and Nie ner, M

Dai, A.; Chang, A. X.; Savva, M.; Halber, M.; Funkhouser, T.; and Nie ner, M. 2017. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In Proceedings of the IEEE conference on computer vision and pattern recognition, 5828--5839

2017

[6] [6]

Dong, X.; Bao, J.; Zheng, Y.; Zhang, T.; Chen, D.; Yang, H.; Zeng, M.; Zhang, W.; Yuan, L.; Chen, D.; et al. 2023. Maskclip: Masked self-distillation advances contrastive language-image pretraining. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 10995--11005

work page 2023

[7] [7]

Fan, Z.; Wang, P.; Jiang, Y.; Gong, X.; Xu, D.; and Wang, Z. 2022. Nerf-sos: Any-view self-supervised object segmentation on complex scenes. arXiv preprint arXiv:2209.08776

arXiv 2022

[8] [8]

Fang, G.; and Wang, B. 2024. Mini-Splatting: Representing Scenes with a Constrained Number of Gaussians. arXiv preprint arXiv:2403.14166

arXiv 2024