Pith. sign in

REVIEW 5 major objections 6 minor 34 references

DCSEG: Decoupled 3D Open-Set Segmentation using Gaussian Splatting

T0 review · 5 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read DCSEG decouples 3D mask proposal from 2D labeling to beat NeRF baselines in open-set segmentation.

desk verdict DCSEG is a clean, modular 3DGS open-vocabulary segmentation pipeline that recombines known parts in a sensible way, but the empirical case against OpenNeRF is thinner than the headline suggests. read the letter →

arxiv 2412.10972 v2 pith:5MGVME6D submitted 2024-12-14 cs.CV

Luis Wiedmann , Luca Wiehe , David Rozenberszki This is my paper · ORCID
classification cs.CV
keywords 3DGaussianSplattingopen-vocabularysegmentationdecoupledclass-agnosticmaskproposalcontrastivelearninginstancebipartiteassignmentScanNet
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

This paper claims that 3D open-vocabulary segmentation can be built by decoupling the two core tasks: proposing class-agnostic object or part masks in 3D, and assigning semantic labels using a 2D open-vocabulary segmentation model. The authors construct this pipeline on 3D Gaussian Splatting, learning per-Gaussian features with contrastive supervision from SAM masks, clustering those features with HDBSCAN, and then matching the resulting 3D masks to class-aware 2D masks from OVSeg or OpenSeg. If this works, it would matter because it shows a modular alternative to NeRF-based segmentation that is faster to render, gives instance and part segmentation without extra training, and can adapt to new 2D foundation models without retraining. The paper reports improved mIoU and mAcc over OpenNeRF on Replica and ScanNet, while explicitly acknowledging that tail-class accuracy is limited by the 2D foundation model.

What carries the argument

The pipeline is carried by three coupled components: (1) per-Gaussian affinity features $f_{g_i}$ in $\mathbb{R}^n$, learned by a scale-aware contrastive loss that pulls together features of pixels inside the same SAM mask and pushes apart those in different masks; (2) density-based hierarchical clustering (HDBSCAN) on those Gaussian features to produce class-agnostic clusters, which are rasterized into binary 2D masks $M_a$; and (3) a bipartite assignment between $M_a$ and the 2D class-aware masks $M_b$ from a foundation segmentation model, using the inverse Jaccard index as the cost and duplicating class vertices so multiple instances of the same class can each receive a label. The duplicated-vertex assignment, solvable by the Jonker-Volgenant variant of the Hungarian algorithm, is the workaround the authors adopt to avoid an NP-hard generalized assignment problem.

What would settle it

Take a Replica scene and a tail class that OVSeg frequently misses, then rerun only the classification stage with that class's 2D masks artificially removed; if the 3D clusters for that class still exist but receive wrong or no labels despite unchanged geometry, the claimed dependence on the 2D foundation model is directly demonstrated, confirming the paper's own limitation statement.

Watch

Extended reading notes

Core claim

The central claim is that decoupling class-agnostic 3D mask proposal from 2D class-aware mask classification yields a 3D open-set segmentation system that outperforms the NeRF-based OpenNeRF baseline in mean IoU and mean accuracy on both Replica and ScanNet scenes, while also producing instance-level and part-level masks without any retraining of an instance segmentation network. The 3D reconstruction uses explicit Gaussian primitives, and the class-agnostic masks come from clustering scale-aware contrastive features derived from SAM masks, while semantic labels come from matching these clusters to 2D open-vocabulary masks via a relaxed bipartite assignment. The authors further claim that this design is modular: swapping the 2D segmentation backbone (OpenSeg vs OVSeg) changes results but requires no retraining, and the same logic would extend to replacing the 3D representation.

Load-bearing premise

The assignment stage assumes that class-agnostic 3D clusters, when projected to 2D, overlap with class-aware masks from OVSeg or OpenSeg well enough for the weighted bipartite match with duplicated vertices to assign correct labels to every cluster, including instances the 2D model might misrecognize.

Editorial extensions

If this is right

  • If the central claim is correct, 3D open-set segmentation no longer requires training a dedicated 3D semantic head; it can be assembled from existing 2D open-vocabulary models plus a 3D clustering stage, meaning future improvements in 2D segmentation directly transfer to 3D.
  • Instance and part masks emerge as a byproduct of the class-agnostic clustering, so downstream tasks like robotic grasping or AR object manipulation get fine-grained 3D structure without an extra instance-segmentation network.
  • The explicit geometry of Gaussian splatting reduces noise on large uniform surfaces compared to NeRF-based pipelines, as the paper demonstrates with qualitative comparisons on walls and floors.
  • The plug-and-play modularity suggests that users can trade off speed versus tail-class accuracy simply by swapping the 2D backbone, as shown by the OVSeg versus OpenSeg comparison.
  • Because the method is zero-shot with respect to 3D labels, it can segment classes not present in the ground-truth annotations, such as posters on walls in ScanNet, potentially providing a cleaner signal for scene understanding than the provided labels.

Reading between the lines

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

  • The decoupling principle is a broader recipe: geometry determines what constitutes a coherent object or part, while a 2D language-conditioned model determines what that object is called; the bottleneck is the 2D classifier, so 3D segmentation quality will track 2D open-vocabulary progress.
  • The bipartite assignment with duplicated vertices is a heuristic that could fail when multiple instances of one class partially overlap in a given view; a testable extension would replace it with a globally consistent assignment across all views, or add a 3D language-feature disambiguation step, which the paper does not pursue.
  • Since the Gaussian representation is explicit and renderable, the same decoupled pipeline could be run iteratively on a live scene, enabling open-set segmentation in interactive or robotic settings without retraining whenever a novel object class is queried.
  • The paper's limitation on tail classes is not a flaw of the 3D clustering but a property of the 2D model; a straightforward stress test would feed a class that OVSeg cannot recognize and check that the 3D mask is still proposed, isolating the failure point.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper proposes DCSEG, a decoupled open-vocabulary 3D semantic segmentation pipeline built on 3D Gaussian Splatting. The method first reconstructs a scene with 3DGS, attaches class-agnostic affinity features to each Gaussian using a SAGA-style contrastive loss supervised by 2D SAM masks, and clusters these features with HDBSCAN to produce coarse instance- or part-level masks. A second stage projects these 3D clusters into 2D and assigns semantic labels by matching them to class-aware masks from a 2D open-vocabulary model (OVSeg or OpenSeg), using a bipartite assignment with duplicated vertices to handle multiple instances per class. The authors evaluate on Replica and ScanNet, reporting mIoU/mAcc against OpenNeRF, OpenScene, LERF, and MinkowskiNet, and include an ablation of the 2D backbone and matching versus assignment. The central claim is that the decoupled design outperforms NeRF-based OpenNeRF on total mIoU and mAcc while providing instance and part segmentation without retraining.

Significance. If the empirical claims are supported, the paper makes a useful contribution: it demonstrates a modular, representation-agnostic decomposition of 3D segmentation into mask proposal and mask classification, with the practical advantages of explicit geometry, fast rendering, and no retraining when swapping 2D foundation models. The ability to obtain instance- and part-level masks without a dedicated instance segmentation network is a genuine strength, and the ablation of OpenSeg versus OVSeg gives concrete evidence for the modularity claim. The paper is honest about the limitation that tail-class performance is bounded by the 2D classifier. However, the central comparison to OpenNeRF rests on margins that are small and unquantified, and the absence of the closest 3DGS baseline (SAGA) leaves the broader claim about 3DGS-based segmentation untested. The methodological skeleton is sound and the paper is clearly written, but the experimental evidence needs strengthening before the headline claim can be accepted.

major comments (5)
  1. [§4.3, Table 1] The claimed superiority over OpenNeRF on Replica is not statistically established. The total mIoU margin is 19.9 versus 19.1 (+0.8), and the tail mIoU margin is 6.7 versus 6.6 (+0.1), averaged over only 8 scenes with no error bars or per-scene results. These differences are within run-to-run noise for stochastic training and evaluation. Please provide per-scene numbers, multiple seeds or confidence intervals, or a statistical test to show the improvement is reliable.
  2. [§4.4, Table 3] The ScanNet comparison uses only 200 images per scene for both DCSEG and OpenNeRF, but NeRF-based methods are known to degrade sharply with sparse views, so this protocol may systematically disadvantage the baseline. Since OpenNeRF does not report ScanNet numbers in its original paper, the authors run it themselves; please justify the 200-image protocol, report the original full-data numbers if available, or include a dense-view comparison to ensure the baseline is not handicapped.
  3. [§3, Eq. (1)] Equation (1) is written with the inverse Jaccard formula: w(ma, mb) = sum_i sum_j |ma ∪ mb| / |ma ∩ mb|. As written, this is a pixel-summed inverse Jaccard index, and the standard Jaccard index should appear as the denominator of the cost or the cost should be 1 − Jaccard. This formula is load-bearing because it defines the assignment cost in the bipartite matching used for classification. Please correct the equation and define the notation clearly.
  4. [§4, overall] The paper reuses SAGA's contrastive feature learning stage but never compares against SAGA, which is the closest 3DGS-based baseline. Without this comparison, the claim that the decoupled 3DGS pipeline improves over existing 3DGS segmentation approaches is untested. Please add SAGA results on the same scenes and protocol, or explicitly scope the claim to NeRF-based baselines only.
  5. [§5.2, Limitations] The acknowledged ceiling that tail-class performance is limited by the 2D foundation model is a structural property of the assignment stage: the 3D clustering has no independent semantic signal, so any object the 2D model fails to recognize cannot be labeled even if the 3D mask is perfect. This is not circular, but it means the reported gains over OpenNeRF depend on the specific 2D model used. The paper should state this dependency more prominently when interpreting the comparisons, and ideally include a per-class breakdown showing which classes drive the differences.
minor comments (6)
  1. [§3.1 / §4.1] There is a typo in Section 4.1: 'is is not bound to the evaluation classes' should read 'is not bound'.
  2. [§4.4 / Table 3] The caption of Table 3 says 'choosen' — should be 'chosen'.
  3. [§1, Contributions] The contribution bullet 'event parts' appears to be a typo for 'even parts'.
  4. [References] Reference [31] cites SAM3D, but the text refers to SAM (Segment Anything Model). The attribution is likely incorrect; please cite the original SAM paper (Kirillov et al.) or clarify which model is actually used.
  5. [Figure 2] The caption 'e.g. the blanket/pillows or the wall behind the bed-lamps' is unclear; please specify which images show the claimed sharper boundaries.
  6. [Equation (1)] The notation |ma,ij ∪ mb,ij| and |ma,ij ∩ mb,ij| is ambiguous because ma,ij and mb,ij are binary values at a pixel, not sets; please define the union and intersection in terms of binary masks.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: DCSEG composes external 2D foundation models and 3D clustering without fitting parameters to the evaluation metrics or relying on load-bearing self-citations.

full rationale

The derivation chain is self-contained. Stage 1 learns class-agnostic Gaussian features via a contrastive loss inspired by and cited to SAGA [3], an external method, using masks from SAM; stage 2 attaches semantic labels by matching projected 3D clusters to class-aware masks from external models OVSeg/OpenSeg. No equation or fitted parameter reduces to the reported mIoU/mAcc values, and the labels are inherited from external 2D models rather than derived from the 3D representation itself. The only self-citation is UnScene3D [28] by co-author Rozenberszki, mentioned in Related Work as an example of unsupervised instance segmentation; it is not load-bearing for any central claim. The limitations in Sec. 5.2 (tail-class ceiling set by the 2D foundation model, suboptimal bipartite assignment, Gaussian edge artifacts) are honest dependencies, not circularities. The comparison against OpenNeRF and other baselines is an empirical evaluation with external ground truth, not a prediction forced by construction. Therefore no circular step is present.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on hand-chosen hyperparameters inherited from SAGA and on the reliability of external 2D models. No new physical or computational entities are postulated; the pipeline combines existing modules.

free parameters (5)
  • Scale hyperparameter s = not reported
    Controls granularity of SAM masks in the contrastive loss, inherited from SAGA [3]; changes object versus part level clustering without retraining, and is hand-tuned.
  • HDBSCAN clustering parameters = not reported
    Min cluster size and other density parameters for HDBSCAN are not specified; they directly determine the number and shape of class-agnostic masks.
  • Low-connectivity cluster threshold = not reported
    Post-processing step removes outlier clusters with low pixel connectivity; threshold is chosen by hand and affects mask quality.
  • Number of input images = 200 per scene
    All experiments use 200 posed RGB-D images per Replica or ScanNet scene; this subsampling affects reconstruction quality and mask coverage, and is chosen by the authors.
  • Number of duplicated vertices per class = not reported
    Bipartite assignment duplicates 2D class masks to match instance counts, but the paper does not specify how the number of instances per class is estimated.
assumptions (4)
  • domain assumption 3D Gaussian Splatting reconstruction yields accurate enough geometry to project clusters to 2D
    The pipeline rasterizes 3D clusters back to 2D for matching; if the reconstruction drifts or Gaussians blur sharp edges, mask alignment fails (Limitations, Sec. 5.2).
  • domain assumption SAM masks provide coherent object or part level supervision
    Contrastive features are optimized so pixels in the same SAM mask have similar 3D features; incoherent SAM masks would produce fragmented or merged clusters.
  • domain assumption OVSeg and OpenSeg class-aware masks are accurate enough for label assignment
    The authors state tail-class performance is limited by the 2D foundation model (Sec. 5.2); the whole classification stage depends on these models.
  • domain assumption Bipartite assignment with duplicated vertices approximates the generalized assignment
    The paper acknowledges bipartite assignment is not optimal (Sec. 5.2) and relies on the approximation for correct labels.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DCSEG: Decoupled 3D Open-Set Segmentation using Gaussian Splatting." pith.science (2026). https://pith.science/paper/5MGVME6D

@misc{pith2026241210972,
  author       = {Pith},
  title        = {Pith review of: DCSEG: Decoupled 3D Open-Set Segmentation using Gaussian Splatting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5MGVME6D}},
  note         = {Machine review of arXiv:2412.10972}
}
read the original abstract

Open-set 3D segmentation represents a major point of interest for multiple downstream robotics and augmented/virtual reality applications. We present a decoupled 3D segmentation pipeline to ensure modularity and adaptability to novel 3D representations as well as semantic segmentation foundation models. We first reconstruct a scene with 3D Gaussians and learn class-agnostic features through contrastive supervision from a 2D instance proposal network. These 3D features are then clustered to form coarse object- or part-level masks. Finally, we match each 3D cluster to class-aware masks predicted by a 2D open-vocabulary segmentation model, assigning semantic labels without retraining the 3D representation. Our decoupled design (1) provides a plug-and-play interface for swapping different 2D or 3D modules, (2) ensures multi-object instance segmentation at no extra cost, and (3) leverages rich 3D geometry for robust scene understanding. We evaluate on synthetic and real-world indoor datasets, demonstrating improved performance over comparable NeRF-based pipelines on mIoU and mAcc, particularly for challenging or long-tail classes. We also show how varying the 2D backbone affects the final segmentation, highlighting the modularity of our framework. These results confirm that decoupling 3D mask proposal and semantic classification can deliver flexible, efficient, and open-vocabulary 3D segmentation.

Figures

Figures reproduced from arXiv: 2412.10972 by the authors.

Figure 1
Figure 1. Decoupling the semantic segmentation pipeline. We present DCSEG, a holistic 3D reconstruction and scene understanding method. At the core of our method, we leverage pre-trained 2D foundation models to recognize uniform semantic concepts in 2D images of 3D scenes and use these predicted masks as contrastive optimization targets from multi-view images to class-agnostic 3D instances and object parts. These features are… view at source ↗
Figure 2
Figure 2. Segmentation results of our method (DCSEG) compared to the ground truth and OpenNeRF. Our segmentation masks can detect boundaries more accurately e.g. the blanket/pillows or the wall behind the bed-lamps. Large uniform areas, such as the floor, can be detected with significantly less noise. Switching between Openseg and OVSeg can be done without retraining and demonstrates adaptability with respect to foundation mo… view at source ↗
Figure 3
Figure 3. Shortcomings of the ScanNet GT. Our Method accurately recognizes and segments the posters on the wall, but they are not represented in the provided ScanNet Ground Truth, therefore hurting our performance despite a more accurate segmentation of the scene. Total Head Common Tail mIoU mAcc mIoU mAcc mIoU mAcc mIoU mAcc OpenSeg + Matching 16.17 29.61 31.89 43.76 14.63 32.50 2.93 14.28 OVSeg + Matching 17.96 32.41 35.35 … view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Further Results on ScanNet (scene0030 01) and Replica (office2) 5. Discussion 5.1. Instance and Part Segmentation. As mentioned previously and demonstrated in [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 6
Figure 6. Figure 6: Part Segmentation. For objects with clearly separable parts, our approach tends to propose masks that correspond to part segmentation. In conclusion, our methodology offers distinct advan￾tages over our baselines by enabling instance or even part segmentation without n…
Figure 7
Figure 7. Figure 7: Limitations. The nature of 3D Gaussians prevents clear edges when segmenting (top). Additionally, we can observe low￾connectivity clusters with few pixels between the two walls (right). threshold remain in the data. Such clusters continue to pose a problem, indicating …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 22 canonical work pages

  1. [1]

    Ricardo J. G. B. Campello, Davoud Moulavi, and J ¨org Sander. Density-based clustering based on hierarchical den- sity estimates. In Pacific-Asia Conference on Knowledge Discovery and Data Mining, 2013. 3

    work page 2013

  2. [2]

    A survey of algorithms for the generalized assignment problem

    Dirk G Cattrysse and Luk N Van Wassenhove. A survey of algorithms for the generalized assignment problem. Euro- pean journal of operational research, 60(3):260–272, 1992. 4

    work page 1992

  3. [3]

    Segment any 3d gaussians

    Jiazhong Cen, Jiemin Fang, Chen Yang, Lingxi Xie, Xi- aopeng Zhang, Wei Shen, and Qi Tian. Segment any 3d gaussians. arXiv preprint arXiv:2312.00860, 2023. 2, 3

    arXiv 2023

  4. [4]

    Splat-nav: Safe real-time robot navigation in gaussian splatting maps, 2024

    Timothy Chen, Ola Shorinwa, Joseph Bruno, Aiden Swann, Javier Yu, Weijia Zeng, Keiko Nagami, Philip Dames, and Mac Schwager. Splat-nav: Safe real-time robot navigation in gaussian splatting maps, 2024. 2

    work page 2024

  5. [5]

    Masked-attention mask transformer for universal image segmentation

    Bowen Cheng, Ishan Misra, Alexander G Schwing, Alexan- der Kirillov, and Rohit Girdhar. Masked-attention mask transformer for universal image segmentation. In Proceed- 8 ings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1290–1299, 2022. 8

    work page 2022

  6. [6]

    4d spatio-temporal convnets: Minkowski convolutional neural networks

    Christopher Choy, JunYoung Gwak, and Silvio Savarese. 4d spatio-temporal convnets: Minkowski convolutional neural networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 3075–3084,

  7. [7]

    On implementing 2d rectangular assign- ment algorithms

    David F Crouse. On implementing 2d rectangular assign- ment algorithms. IEEE Transactions on Aerospace and Elec- tronic Systems, 52(4):1679–1696, 2016. 4

    work page 2016

  8. [8]

    Scannet: Richly-annotated 3d reconstructions of indoor scenes

    Angela Dai, Angel X Chang, Manolis Savva, Maciej Hal- ber, Thomas Funkhouser, and Matthias Nießner. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 5828–5839, 2017. 2, 5

    2017

Show all 34 references
  1. [9]

    De- coupling zero-shot semantic segmentation

    Jian Ding, Nan Xue, Gui-Song Xia, and Dengxin Dai. De- coupling zero-shot semantic segmentation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11583–11592, 2022. 3

    2022

  2. [10]

    OpenNeRF: Open Set 3D Neural Scene Segmentation with Pixel-Wise Features and Rendered Novel Views

    Francis Engelmann, Fabian Manhardt, Michael Niemeyer, Keisuke Tateno, Marc Pollefeys, and Federico Tombari. OpenNeRF: Open Set 3D Neural Scene Segmentation with Pixel-Wise Features and Rendered Novel Views. In Interna- tional Conference on Learning Representations, 2024. 2, 4, 6

    2024

  3. [11]

    Comparing sets of patterns with the jaccard index

    Sam Fletcher, Md Zahidul Islam, et al. Comparing sets of patterns with the jaccard index. Australasian Journal of In- formation Systems, 22, 2018. 4

    2018

  4. [12]

    Panoptic nerf: 3d-to-2d label transfer for panoptic urban scene segmentation

    Xiao Fu, Shangzhan Zhang, Tianrun Chen, Yichong Lu, Lanyun Zhu, Xiaowei Zhou, Andreas Geiger, and Yiyi Liao. Panoptic nerf: 3d-to-2d label transfer for panoptic urban scene segmentation. In 2022 International Conference on 3D Vision (3DV), pages 1–11. IEEE, 2022. 2

    2022

  5. [13]

    Scal- ing open-vocabulary image segmentation with image-level labels

    Golnaz Ghiasi, Xiuye Gu, Yin Cui, and Tsung-Yi Lin. Scal- ing open-vocabulary image segmentation with image-level labels. In European Conference on Computer Vision, pages 540–557. Springer, 2022. 2, 3

    2022

  6. [14]

    Semantic gaussians: Open-vocabulary scene understanding with 3d gaussian splatting.arXiv preprint arXiv:2403.15624,

    Jun Guo, Xiaojian Ma, Yue Fan, Huaping Liu, and Qing Li. Semantic gaussians: Open-vocabulary scene understanding with 3d gaussian splatting.arXiv preprint arXiv:2403.15624,

    arXiv

  7. [15]

    Open-vocabulary semantic segmentation with decou- pled one-pass network

    Cong Han, Yujie Zhong, Dengjie Li, Kai Han, and Lin Ma. Open-vocabulary semantic segmentation with decou- pled one-pass network. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 1086– 1096, 2023. 3

    2023

  8. [16]

    Sagd: Boundary- enhanced segment anything in 3d gaussian via gaussian de- composition, 2024

    Xu Hu, Yuxi Wang, Lue Fan, Junsong Fan, Junran Peng, Zhen Lei, Qing Li, and Zhaoxiang Zhang. Sagd: Boundary- enhanced segment anything in 3d gaussian via gaussian de- composition, 2024. 8

    2024

  9. [17]

    Segment3d: Learning fine-grained class-agnostic 3d segmentation without manual labels

    Rui Huang, Songyou Peng, Ayca Takmaz, Federico Tombari, Marc Pollefeys, Shiji Song, Gao Huang, and Francis Engel- mann. Segment3d: Learning fine-grained class-agnostic 3d segmentation without manual labels. In European Confer- ence on Computer Vision , pages 278–295. Springer, 2024. 2

    2024

  10. [18]

    A shortest augmenting path algorithm for dense and sparse linear assignment problems

    Roy Jonker and Ton V olgenant. A shortest augmenting path algorithm for dense and sparse linear assignment problems. Computing, 38:325–340, 1987. 4

    1987

  11. [19]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics, 42 (4):1–14, 2023. 2

    2023

  12. [20]

    Lerf: Language embedded radiance fields

    Justin Kerr, Chung Min Kim, Ken Goldberg, Angjoo Kanazawa, and Matthew Tancik. Lerf: Language embedded radiance fields. In Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision, pages 19729–19739,

  13. [21]

    The hungarian method for the assignment problem

    Harold W Kuhn. The hungarian method for the assignment problem. Naval research logistics quarterly , 2(1-2):83–97,

  14. [22]

    Gaussnav: Gaussian splatting for visual navigation, 2024

    Xiaohan Lei, Min Wang, Wengang Zhou, and Houqiang Li. Gaussnav: Gaussian splatting for visual navigation, 2024. 2

    2024

  15. [23]

    Open-vocabulary semantic segmentation with mask-adapted clip

    Feng Liang, Bichen Wu, Xiaoliang Dai, Kunpeng Li, Yinan Zhao, Hang Zhang, Peizhao Zhang, Peter Vajda, and Diana Marculescu. Open-vocabulary semantic segmentation with mask-adapted clip. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pag...

    2023

  16. [24]

    Nerf: Representing scenes as neural radiance fields for view syn- thesis

    Ben Mildenhall, Pratul P Srinivasan, Matthew Tancik, Jonathan T Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view syn- thesis. Communications of the ACM , 65(1):99–106, 2021. 2

    2021

  17. [25]

    Openscene: 3d scene understanding with open vocabularies

    Songyou Peng, Kyle Genova, Chiyu Jiang, Andrea Tagliasacchi, Marc Pollefeys, Thomas Funkhouser, et al. Openscene: 3d scene understanding with open vocabularies. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition , pages 815–824, 2023. 2, 4, 6

    2023

  18. [26]

    Langsplat: 3d language gaussian splatting

    Minghan Qin, Wanhua Li, Jiawei Zhou, Haoqian Wang, and Hanspeter Pfister. Langsplat: 3d language gaussian splatting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 20051–20060, 2024. 3

    2024

  19. [27]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. In International conference on machine learning, ...

    2021

  20. [28]

    Unscene3d: Unsupervised 3d instance segmentation for indoor scenes

    David Rozenberszki, Or Litany, and Angela Dai. Unscene3d: Unsupervised 3d instance segmentation for indoor scenes. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19957–19967, 2024. 2

    2024

  21. [29]

    The replica dataset: A digital replica of indoor spaces

    Julian Straub, Thomas Whelan, Lingni Ma, Yufan Chen, Erik Wijmans, Simon Green, Jakob J Engel, Raul Mur-Artal, Carl Ren, Shobhit Verma, et al. The replica dataset: A digital replica of indoor spaces. arXiv preprint arXiv:1906.05797,

    1906 arXiv

  22. [30]

    Sumner, Marc Pollefeys, Federico Tombari, and Francis Engelmann

    Ayc ¸a Takmaz, Elisabetta Fedele, Robert W. Sumner, Marc Pollefeys, Federico Tombari, and Francis Engelmann. 9 OpenMask3D: Open-V ocabulary 3D Instance Segmentation. In Advances in Neural Information Processing Systems (NeurIPS), 2023. 2, 3, 5, 8

    2023

  23. [31]

    Sam3d: Segment anything in 3d scenes.arXiv preprint arXiv:2306.03908, 2023

    Yunhan Yang, Xiaoyang Wu, Tong He, Hengshuang Zhao, and Xihui Liu. Sam3d: Segment anything in 3d scenes.arXiv preprint arXiv:2306.03908, 2023. 2

    2023 arXiv

  24. [32]

    Sai3d: Segment any instance in 3d scenes

    Yingda Yin, Yuzheng Liu, Yang Xiao, Daniel Cohen-Or, Jingwei Huang, and Baoquan Chen. Sai3d: Segment any instance in 3d scenes. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 3292–3302, 2024. 2, 3, 8

    2024

  25. [33]

    Feature 3dgs: Supercharging 3d gaussian splatting to enable distilled feature fields

    Shijie Zhou, Haoran Chang, Sicheng Jiang, Zhiwen Fan, Ze- hao Zhu, Dejia Xu, Pradyumna Chari, Suya You, Zhangyang Wang, and Achuta Kadambi. Feature 3dgs: Supercharging 3d gaussian splatting to enable distilled feature fields. InPro- ceedings of the IEEE/CVF Conference on Compu...

    2024

  26. [34]

    V oxelnet: End-to-end learning for point cloud based 3d object detection

    Yin Zhou and Oncel Tuzel. V oxelnet: End-to-end learning for point cloud based 3d object detection. In Proceedings of the IEEE conference on computer vision and pattern recog- nition, pages 4490–4499, 2018. 2 10

    2018

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.