Pith. sign in

REVIEW 3 major objections 5 minor 48 references

GSFF-SLAM: 3D Semantic Gaussian Splatting SLAM via Feature Field

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read GSFF-SLAM claims online dense semantic mapping by attaching trainable feature vectors to 3D Gaussians, reaching 95.03% mIoU with ground-truth priors and up to 2.9x speedup.

desk verdict A solid GT-supervised semantic SLAM system whose headline noisy-prior claim needs numbers before it stands. read the letter →

arxiv 2504.19409 v2 pith:A3OTHRNP submitted 2025-04-28 cs.RO cs.AI

Zuxing Lu , Xin Yuan , Shaowen Yang , Jingyu Liu , Changyin Sun This is my paper · ORCID
classification cs.ROcs.AI
keywords 3DGaussianSplattingSemanticSLAMFeaturefieldsRGB-DOpen-vocabularysegmentationDensemappingNeuralrenderingFoundationmodels
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 tries to establish that an online RGB-D SLAM system can build dense, semantically meaningful 3D maps without relying on dense ground-truth labels. GSFF-SLAM stores semantics as trainable feature vectors on each 3D Gaussian and renders them with the same splatting pipeline used for color and depth, while optimizing the feature field separately from geometry. The payoff would be a robot that tracks its pose and reconstructs geometry while also learning what objects are, from sparse text or detection signals, and doing so fast enough for interactive use; with ground-truth labels it reports 95.03% mIoU and up to 2.9x faster semantic mapping than a leading baseline.

What carries the argument

The load-bearing object is the semantic feature field: each Gaussian point carries an optimizable vector $f\in\mathbb{R}^N$ that is rendered into a dense feature map $\hat{F}\in\mathbb{R}^{H\times W\times N}$ by the same front-to-back $\alpha$-blending used for RGB and depth. The design that makes the claim work is decoupling—the gradient of the feature embedding never propagates into the mean, covariance, opacity, or color parameters, and feature-field optimization runs only after geometric mapping has converged—combined with a co-visibility keyframe selector and, for text supervision, the L1 loss of Eq. 11 aligning rendered features with CLIP text embeddings of detections.

What would settle it

Run the sparse-text supervision variant on a sequence where a detector deliberately misses a common class in most frames, then measure the class's mIoU against a dense ground-truth evaluation: if the feature field cannot recover the missed class above chance, the claimed advantage of sparse-prior robustness is falsified. The authors' own observation that ceiling lights and vents are contaminated by low detection rates provides a concrete starting point for this test.

Watch

Extended reading notes

Core claim

GSFF-SLAM's central claim is that semantics can be decoupled from 3D reconstruction in a Gaussian Splatting SLAM system by giving every Gaussian its own feature vector $f\in\mathbb{R}^N$ ($N=128$) and rendering these vectors into dense feature maps through the same $\alpha$-blending as color and depth. Feature gradients are stopped before the geometric parameters, so the semantic field is optimized after—and independently of—the map's shape and appearance. With ground-truth 2D labels on the Replica dataset, this reaches a reported 95.03% mIoU (99.41% pixel accuracy), exceeding prior NeRF- and Gaussian-based semantic SLAM baselines; with noisy, sparse priors produced by Grounding-DINO detections, SAM masks, and CLIP text embeddings, it still yields coherent open-vocabulary segmentations and a runtime of 45 minutes versus 132 minutes for SNI-SLAM at only a small mIoU loss.

Load-bearing premise

The sparse-prior contribution rests on the assumption that Grounding-DINO boxes, SAM masks, and CLIP text embeddings line up spatially with the RGB-D frames and are accurate enough to supervise the feature field through Eq. 11; if that spatial alignment or label accuracy fails, the claimed robustness to noisy and sparse priors does not follow.

Editorial extensions

If this is right

  • With 2D ground-truth priors, semantic segmentation reaches 95.03% mIoU on Replica, improving on SNI-SLAM's 84.62% by 10.41 points.
  • Semantic feature rendering runs at 15.8 fps normally and 19.2 fps in the speedup configuration, compared with 0.87 fps for SNI-SLAM.
  • The decoupled design lets semantic labels be added after geometry has been built, so the same map can absorb new semantic supervision without re-optimizing shape and appearance.
  • On Replica, tracking error (ATE RMSE 0.311 cm) and rendering quality (PSNR 38.67 dB) remain competitive or better while the semantic field is being learned.
  • Sparse and noisy textual priors from foundation models can supervise semantic reconstruction, with qualitative results showing coherent segmentation even for objects that were never labeled.

Reading between the lines

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

  • Beyond the paper, decoupling suggests the same stored feature vectors could be queried with new text phrases after mapping, turning the map into a re-labelable representation without retraining the feature field.
  • Because sparse signals are densified through the 3D field, a natural extension is interactive correction: a user points at one instance and names it, and the field propagates that label to geometrically similar regions.
  • The reported drift on long sequences and motion-blurred frames points to loop closure and blur-robust tracking as the next bottlenecks; adding them would likely carry the semantic gains to longer real-world deployments.
  • A quantitative test of robustness to sparsity would measure per-class detection rate against final mIoU; the authors' own ceiling-light and vent contamination suggests errors worsen as detection rate falls.
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

3 major / 5 minor

Summary. The manuscript proposes GSFF-SLAM, an RGB-D dense semantic SLAM system based on 3D Gaussian Splatting. The core idea is to attach an N-dimensional semantic feature vector to each Gaussian and to optimize these semantic embeddings separately from geometric and color optimization. The paper claims improved tracking and rendering accuracy on Replica, ScanNet, and TUM-RGBD, state-of-the-art semantic segmentation (95.03 percent mIoU) on Replica under ground-truth supervision, a 2.9x runtime speedup at a small accuracy cost, and the ability to reconstruct semantics from sparse and noisy priors produced by Grounding-DINO, SAM, and CLIP. The ground-truth-supervised claims are backed by quantitative tables; the sparse and noisy prior claim is supported only by a qualitative figure.

Significance. If the results hold, the decoupled feature-field design is a useful contribution: it separates semantic optimization from geometry optimization, allows semantic supervision to be deferred, and provides a route for integrating foundation-model outputs into an online 3DGS SLAM pipeline. The paper includes extensive comparisons against several NeRF- and 3DGS-based SLAM baselines, an ablation of two key hyperparameters, and a runtime comparison. However, the most distinctive claim, robustness to sparse and noisy 2D priors, has no quantitative validation, and the equations describing noisy-textual supervision contain an undefined dimension mismatch. The ground-truth-supervised semantic results are solid, but they alone do not distinguish the method from prior work that already uses dense ground-truth priors.

major comments (3)
  1. [§3.3, Eqs. (11)-(12)] The noisy-textual supervision mechanism is incomplete as written. The rendered feature map is defined as F_hat in R^{HxWxN} with N=128 in §4.1, while the CLIP text feature q(l) is M-dimensional with M=512 and I_f is in R^{HxWxM}. Eq. (11) compares I_f with o(F_hat), so o(·) must perform a channel projection or expansion from 128 to 512, but this operation is never defined. Eq. (12) then computes a dot product f(x)q(l)^T with f(x) in R^128 and q(l) in R^512, which is undefined unless an additional learned projection is introduced and trained. This needs to be specified explicitly and consistently.
  2. [§4.3, Noisy Textual Label Supervision and Figure 5] The central robustness claim that the decoupled feature-field optimization supports semantic reconstruction from sparse and noisy 2D priors is supported only by a qualitative comparison with Feature-3DGS. No numeric metrics such as mIoU, pixel accuracy, or detection rates are reported for the Grounding-DINO/SAM/CLIP pipeline, and the authors concede that low detection rates contaminate objects such as ceiling lights and vents. The experimental protocol is also underspecified: the text does not state the text queries used, the detection confidence threshold, the source or amount of label noise, or the object-merge procedure. Without a quantitative evaluation, the abstract's claim about sparse and noisy signals is not established.
  3. [Tables 6 and 7] The headline 95.03 percent mIoU and the associated 114-minute runtime correspond to the configuration rho_pc=1/16 and tau_thresh=0.95 in Table 7, which is identified through an ablation on the same Replica benchmark used for the final evaluation. No error bars or standard deviations are reported, and the resource-constrained configuration in the first row of Table 7 achieves 90.54 percent mIoU. The 'up to 2.9x speedup with only marginal performance degradation' claim refers to a different configuration and should be presented as a trade-off curve rather than as a single state-of-the-art result.
minor comments (5)
  1. [§3.3, Eq. (9)] The symbol N is used for the number of Gaussians, the feature dimension, and the number of semantic classes. With N=128 in §4.1, it is unclear how the cross-entropy over N classes is aligned with the Replica label set and how the argmax in Eq. (10) avoids selecting an invalid empty feature channel. Please clarify the class mapping.
  2. [§3.3] The sentence 'Using the pipeline described in ??' contains an unresolved cross-reference that should be fixed.
  3. [Table 4] The text states that the best results are highlighted as first, second, and third, but no highlighting is visible in the submitted reading copy; please ensure the table formatting is preserved in the final version.
  4. [§4.3, Noisy Textual Label Supervision] The ablation in Table 7 varies only rho_pc and tau_thresh. Adding an ablation of the semantic feature dimension N and of the decoupling itself, for example joint versus separate optimization of semantic and geometric parameters, would more directly support the paper's main architectural claim.
  5. [§4.3, Ground Truth Supervision] Since ScanNet provides semantic annotations, reporting semantic segmentation results on ScanNet would strengthen the generalization claim beyond the synthetic Replica scenes.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found: GT and CLIP supervision are external inputs; metrics are standard benchmark evaluations, not self-defined predictions.

full rationale

GSFF-SLAM's derivation chain is self-contained. The semantic feature embedding f is an optimizable per-Gaussian parameter rendered by Eq. 3 and supervised either by Eq. 9 (cross-entropy against 2D ground-truth labels) or by Eq. 11 (L1 loss against CLIP feature maps from a foundation-model pipeline). Evaluation of the GT-supervised variant uses Eq. 10's argmax and reports mIoU on the same Replica scenes that provided supervision; this is standard semantic-SLAM mapping-fidelity evaluation, not a parameter-free prediction, and the 95.03% mIoU number is not a fitted constant renamed as a result. The sparse/noisy-prior claim rests on qualitative Figure 5 and is explicitly hedged by the authors' admission that low detection rates contaminate objects such as ceiling lights and vents, so it is an evidentiary limitation rather than a circular construction. There are no load-bearing self-citations, no imported uniqueness theorem, and no ansatz smuggled in via citation. The Eq. 11/Eq. 12 dimensionality inconsistency (rendered feature f(x) in R^128 vs CLIP text query q(l) in R^512) is a correctness/specification concern, not a circularity. Overall, no derivation step reduces by definition to its own inputs.

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

The core contributions rest on one hand-picked feature dimension, several unreported loss weights, two ablated hyperparameters selected on the same benchmark, and the modeling assumption that per-Gaussian semantic features can be decoupled from geometry. No new physical entities or forces are introduced.

free parameters (5)
  • semantic_feature_dimension_N = 128
    The dimension of the per-Gaussian semantic embedding; chosen by hand and not ablated, so its effect on capacity and accuracy is unmeasured.
  • tracking_loss_weight_lambda_t = not reported
    Balances color and depth terms in Eq. 6; the value is not given, making exact reproduction impossible.
  • mapping_loss_weights_lambda_m_lambda_r = not reported
    Controls color, depth, and regularization balance in Eq. 8; values are not specified.
  • keyframe_visibility_threshold_tau_thresh = 0.8 or 0.95
    Co-visibility IoU threshold for keyframe selection; ablated in Table 7, with 0.95 used for the final SOTA configuration.
  • pointcloud_downsample_rate_rho_pc = 1/64 or 1/16
    Controls Gaussian density; ablated in Table 7, with 1/16 used for the final SOTA configuration.
assumptions (5)
  • standard math The differentiable 3D Gaussian splatting rasterizer of Kerbl et al. correctly models color, depth, and feature rendering and supports the required gradients.
    Used as a black box throughout Sections 3.1 and 3.2; any approximation error in the rasterizer propagates into tracking and mapping losses.
  • domain assumption RGB-D input frames provide depth aligned with color, and the benchmark camera trajectories are reliable enough for ATE evaluation.
    Eq. 6 uses depth loss and Eq. 2 renders depth; incorrect depth or pose alignment would invalidate the reported tracking and mapping numbers.
  • domain assumption A single N-dimensional semantic vector per Gaussian, composited with the same alpha weights as color, is expressive enough to represent semantic fields across the scene.
    This is the representation choice behind Eq. 3; no capacity or expressivity analysis is provided.
  • ad hoc to paper Semantic feature optimization can be deferred until after geometric and color optimization without sacrificing final semantic accuracy.
    Section 3.3 states that this decoupling is motivated by computational burden and empirical efficiency; it is a design premise, not an established result.
  • domain assumption Foundation model outputs (Grounding-DINO, SAM, CLIP) are sufficiently accurate and spatially aligned to supervise feature fields in the text-guided setting.
    Invoked in Eqs. 11 and 12; only qualitative evidence in Figure 5 supports this, and the paper admits contamination from low detection rates.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GSFF-SLAM: 3D Semantic Gaussian Splatting SLAM via Feature Field." pith.science (2026). https://pith.science/paper/A3OTHRNP

@misc{pith2026250419409,
  author       = {Pith},
  title        = {Pith review of: GSFF-SLAM: 3D Semantic Gaussian Splatting SLAM via Feature Field},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/A3OTHRNP}},
  note         = {Machine review of arXiv:2504.19409}
}
abstract

Semantic-aware 3D scene reconstruction is essential for autonomous robots to perform complex interactions. Semantic SLAM, an online approach, integrates pose tracking, geometric reconstruction, and semantic mapping into a unified framework, shows significant potential. However, existing systems, which rely on 2D ground truth priors for supervision, are often limited by the sparsity and noise of these signals in real-world environments. To address this challenge, we propose GSFF-SLAM, a novel dense semantic SLAM system based on 3D Gaussian Splatting that leverages feature fields to achieve joint rendering of appearance, geometry, and N-dimensional semantic features. By independently optimizing feature gradients, our method supports semantic reconstruction using various forms of 2D priors, particularly sparse and noisy signals. Experimental results demonstrate that our approach outperforms previous methods in both tracking accuracy and photorealistic rendering quality. When utilizing 2D ground truth priors, GSFF-SLAM achieves state-of-the-art semantic segmentation performance with 95.03\% mIoU, while achieving up to 2.9$\times$ speedup with only marginal performance degradation.

Figures

Figures reproduced from arXiv: 2504.19409 by the authors.

Figure 1
Figure 1. Our GSFF-SLAM leverages different forms of signals to enhance various downstream online tasks. Our [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Overview of GSFF-SLAM. Our method takes an RGB-D stream as input, leveraging 3D Gaussian Splatting with semantic feature embedding f to generate RGB images, depth images, and dense feature maps. Semantic signals, derived from foundation models or ground truth, supervise the learning process, while the feature embedding f is optimized independently. Dense Visual SLAM. SLAM typically divided into two main tasks: mappi… view at source ↗
Figure 3
Figure 3. Qualitative comparison on rendering quality of baseline and our method. We select 4 scenes of Replica dataset [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison of semantic reconstruction performance using ground truth labels on the Replica [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparison of semantic reconstruction performance using noisy textual labels on the Replica [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 24 canonical work pages

  1. [1]

    Monoslam: Real-time single camera slam

    Andrew J Davison, Ian D Reid, Nicholas D Molton, and Olivier Stasse. Monoslam: Real-time single camera slam. IEEE transactions on pattern analysis and machine intelligence, 29(6):1052–1067, 2007

    work page 2007

  2. [2]

    Visual simultaneous localization and mapping: a survey

    Jorge Fuentes-Pacheco, José Ruiz-Ascencio, and Juan Manuel Rendón-Mancha. Visual simultaneous localization and mapping: a survey. Artificial intelligence review, 43:55–81, 2015

    work page 2015

  3. [3]

    Parallel tracking and mapping for small ar workspaces

    Georg Klein and David Murray. Parallel tracking and mapping for small ar workspaces. In 2007 6th IEEE and ACM international symposium on mixed and augmented reality, pages 225–234. IEEE, 2007

    work page 2007

  4. [4]

    Orb-slam: a versatile and accurate monocular slam system

    Raul Mur-Artal, Jose Maria Martinez Montiel, and Juan D Tardos. Orb-slam: a versatile and accurate monocular slam system. IEEE transactions on robotics, 31(5):1147–1163, 2015

    2015

  5. [5]

    Orb-slam2: An open-source slam system for monocular, stereo, and rgb-d cameras

    Raul Mur-Artal and Juan D Tardós. Orb-slam2: An open-source slam system for monocular, stereo, and rgb-d cameras. IEEE transactions on robotics, 33(5):1255–1262, 2017

    work page 2017

  6. [6]

    Dtam: Dense tracking and mapping in real-time

    Richard A Newcombe, Steven J Lovegrove, and Andrew J Davison. Dtam: Dense tracking and mapping in real-time. In 2011 international conference on computer vision, pages 2320–2327. IEEE, 2011

    2011

  7. [7]

    Nerf: Representing scenes as neural radiance fields for view synthesis

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

    2021

  8. [8]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimkühler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM Trans. Graph., 42(4):139–1, 2023

    2023

Show all 48 references
  1. [9]

    Nice-slam: Neural implicit scalable encoding for slam

    Zihan Zhu, Songyou Peng, Viktor Larsson, Weiwei Xu, Hujun Bao, Zhaopeng Cui, Martin R Oswald, and Marc Pollefeys. Nice-slam: Neural implicit scalable encoding for slam. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12786–12796, 2022

    2022

  2. [10]

    V ox-fusion: Dense tracking and mapping with voxel-based neural implicit representation

    Xingrui Yang, Hai Li, Hongjia Zhai, Yuhang Ming, Yuqian Liu, and Guofeng Zhang. V ox-fusion: Dense tracking and mapping with voxel-based neural implicit representation. In 2022 IEEE International Symposium on Mixed and Augmented Reality (ISMAR), pages 499–507. IEEE, 2022

    2022

  3. [11]

    Point-slam: Dense neural point cloud-based slam

    Erik Sandström, Yue Li, Luc Van Gool, and Martin R Oswald. Point-slam: Dense neural point cloud-based slam. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 18433–18444, 2023

    2023

  4. [12]

    Splatam: Splat track & map 3d gaussians for dense rgb-d slam

    Nikhil Keetha, Jay Karhade, Krishna Murthy Jatavallabhula, Gengshan Yang, Sebastian Scherer, Deva Ramanan, and Jonathon Luiten. Splatam: Splat track & map 3d gaussians for dense rgb-d slam. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, p...

    2024

  5. [13]

    Gaussian splatting slam

    Hidenobu Matsuki, Riku Murai, Paul HJ Kelly, and Andrew J Davison. Gaussian splatting slam. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 18039–18048, 2024

    2024

  6. [14]

    Structure-from-motion revisited

    Johannes Lutz Schönberger and Jan-Michael Frahm. Structure-from-motion revisited. In Conference on Computer Vision and Pattern Recognition (CVPR), 2016

    2016

  7. [15]

    Semanticfusion: Dense 3d semantic mapping with convolutional neural networks

    John McCormac, Ankur Handa, Andrew Davison, and Stefan Leutenegger. Semanticfusion: Dense 3d semantic mapping with convolutional neural networks. In 2017 IEEE International Conference on Robotics and automation (ICRA), pages 4628–4635. IEEE, 2017

    2017

  8. [16]

    In-place scene labelling and understanding with implicit scene representation

    Shuaifeng Zhi, Tristan Laidlow, Stefan Leutenegger, and Andrew J Davison. In-place scene labelling and understanding with implicit scene representation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 15838–15847, 2021

    2021

  9. [17]

    Decomposing nerf for editing via feature field distillation

    Sosuke Kobayashi, Eiichi Matsumoto, and Vincent Sitzmann. Decomposing nerf for editing via feature field distillation. Advances in Neural Information Processing Systems, 35:23311–23330, 2022

    2022

  10. [18]

    Neural implicit dense semantic slam

    Yasaman Haghighi, Suryansh Kumar, Jean-Philippe Thiran, and Luc Van Gool. Neural implicit dense semantic slam. arXiv preprint arXiv:2304.14560, 2023

    2023 arXiv

  11. [19]

    Dns-slam: Dense neural semantic-informed slam

    Kunyi Li, Michael Niemeyer, Nassir Navab, and Federico Tombari. Dns-slam: Dense neural semantic-informed slam. In 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 7839–7846. IEEE, 2024

    2024

  12. [20]

    Sni-slam: Semantic neural implicit slam

    Siting Zhu, Guangming Wang, Hermann Blum, Jiuming Liu, Liang Song, Marc Pollefeys, and Hesheng Wang. Sni-slam: Semantic neural implicit slam. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 21167–21177, 2024

    2024

  13. [21]

    Sgs-slam: Semantic gaussian splatting for neural dense slam

    Mingrui Li, Shuhong Liu, Heng Zhou, Guohao Zhu, Na Cheng, Tianchen Deng, and Hongyu Wang. Sgs-slam: Semantic gaussian splatting for neural dense slam. In European Conference on Computer Vision, pages 163–179. Springer, 2024

    2024

  14. [22]

    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, 2019

    1906 arXiv

  15. [23]

    Gaussian-slam: Photo-realistic dense slam with gaussian splatting

    Vladimir Yugay, Yue Li, Theo Gevers, and Martin R Oswald. Gaussian-slam: Photo-realistic dense slam with gaussian splatting. arXiv preprint arXiv:2312.10070, 2023

    2023 arXiv

  16. [24]

    Gaussianeditor: Swift and controllable 3d editing with gaussian splatting

    Yiwen Chen, Zilong Chen, Chi Zhang, Feng Wang, Xiaofeng Yang, Yikai Wang, Zhongang Cai, Lei Yang, Huaping Liu, and Guosheng Lin. Gaussianeditor: Swift and controllable 3d editing with gaussian splatting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...

    2024

  17. [25]

    Language embedded 3d gaussians for open- vocabulary scene understanding

    Jin-Chuan Shi, Miao Wang, Hao-Bin Duan, and Shao-Hua Guan. Language embedded 3d gaussians for open- vocabulary scene understanding. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5333–5343, 2024

    2024

  18. [26]

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

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

    2024

  19. [27]

    Berg, Wan-Yen Lo, Piotr Dollár, and Ross Girshick

    Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alexander C. Berg, Wan-Yen Lo, Piotr Dollár, and Ross Girshick. Segment anything. arXiv:2304.02643, 2023

    2023 arXiv

  20. [28]

    Parallel tracking and mapping on a camera phone

    Georg Klein and David Murray. Parallel tracking and mapping on a camera phone. In2009 8th IEEE International Symposium on Mixed and Augmented Reality, pages 83–86. IEEE, 2009

    2009

  21. [29]

    Orb-slam3: An accurate open-source library for visual, visual–inertial, and multimap slam

    Carlos Campos, Richard Elvira, Juan J Gómez Rodríguez, José MM Montiel, and Juan D Tardós. Orb-slam3: An accurate open-source library for visual, visual–inertial, and multimap slam. IEEE Transactions on Robotics, 37(6):1874–1890, 2021

    2021

  22. [30]

    imap: Implicit mapping and positioning in real-time

    Edgar Sucar, Shikun Liu, Joseph Ortiz, and Andrew J Davison. imap: Implicit mapping and positioning in real-time. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6229–6238, 2021

    2021

  23. [31]

    Deepfactors: Real-time probabilistic dense monocular slam

    Jan Czarnowski, Tristan Laidlow, Ronald Clark, and Andrew J Davison. Deepfactors: Real-time probabilistic dense monocular slam. IEEE Robotics and Automation Letters, 5(2):721–728, 2020

    2020

  24. [32]

    Real-time high-accuracy three-dimensional reconstruction with consumer rgb-d cameras

    Yan-Pei Cao, Leif Kobbelt, and Shi-Min Hu. Real-time high-accuracy three-dimensional reconstruction with consumer rgb-d cameras. ACM Transactions on Graphics (TOG), 37(5):1–16, 2018. 12 PrePrint Paper

    2018

  25. [33]

    Droid-slam: Deep visual slam for monocular, stereo, and rgb-d cameras

    Zachary Teed and Jia Deng. Droid-slam: Deep visual slam for monocular, stereo, and rgb-d cameras. Advances in neural information processing systems, 34:16558–16569, 2021

    2021

  26. [34]

    Elastic fragments for dense scene reconstruction

    Qian-Yi Zhou, Stephen Miller, and Vladlen Koltun. Elastic fragments for dense scene reconstruction. In Proceedings of the IEEE International Conference on Computer Vision, pages 473–480, 2013

    2013

  27. [35]

    Bundlefusion: Real- time globally consistent 3d reconstruction using on-the-fly surface reintegration

    Angela Dai, Matthias Nießner, Michael Zollhöfer, Shahram Izadi, and Christian Theobalt. Bundlefusion: Real- time globally consistent 3d reconstruction using on-the-fly surface reintegration. ACM Transactions on Graphics (ToG), 36(4):1, 2017

    2017

  28. [36]

    V oxel map for visual slam

    Manasi Muglikar, Zichao Zhang, and Davide Scaramuzza. V oxel map for visual slam. In2020 IEEE International Conference on Robotics and Automation (ICRA), pages 4181–4187. IEEE, 2020

    2020

  29. [37]

    Visual language maps for robot navigation

    Chenguang Huang, Oier Mees, Andy Zeng, and Wolfram Burgard. Visual language maps for robot navigation. In 2023 IEEE International Conference on Robotics and Automation (ICRA), pages 10608–10615. IEEE, 2023

    2023

  30. [38]

    Distilled feature fields enable few-shot language-guided manipulation

    William Shen, Ge Yang, Alan Yu, Jansen Wong, Leslie Pack Kaelbling, and Phillip Isola. Distilled feature fields enable few-shot language-guided manipulation. arXiv preprint arXiv:2308.07931, 2023

    2023 arXiv

  31. [39]

    Ovexp: Open vocabulary exploration for object-oriented navigation

    Meng Wei, Tai Wang, Yilun Chen, Hanqing Wang, Jiangmiao Pang, and Xihui Liu. Ovexp: Open vocabulary exploration for object-oriented navigation. arXiv preprint arXiv:2407.09016, 2024

    2024 arXiv

  32. [40]

    Instant neural graphics primitives with a multiresolution hash encoding

    Thomas Müller, Alex Evans, Christoph Schied, and Alexander Keller. Instant neural graphics primitives with a multiresolution hash encoding. ACM transactions on graphics (TOG), 41(4):1–15, 2022

    2022

  33. [41]

    Language-driven semantic segmentation

    Boyi Li, Kilian Q Weinberger, Serge Belongie, Vladlen Koltun, and René Ranftl. Language-driven semantic segmentation. arXiv preprint arXiv:2201.03546, 2022

    2022 arXiv

  34. [42]

    Open3d: A modern library for 3d data processing

    Qian-Yi Zhou, Jaesik Park, and Vladlen Koltun. Open3d: A modern library for 3d data processing. arXiv preprint arXiv:1801.09847, 2018

    2018 arXiv

  35. [43]

    Grounding dino: Marrying dino with grounded pre-training for open-set object detection

    Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, Jie Yang, Chunyuan Li, Jianwei Yang, Hang Su, Jun Zhu, et al. Grounding dino: Marrying dino with grounded pre-training for open-set object detection. arXiv preprint arXiv:2303.05499, 2023

    2023 arXiv

  36. [44]

    Sam 2: Segment anything in images and videos

    Nikhila Ravi, Valentin Gabeur, Yuan-Ting Hu, Ronghang Hu, Chaitanya Ryali, Tengyu Ma, Haitham Khedr, Roman Rädle, Chloe Rolland, Laura Gustafson, Eric Mintun, Junting Pan, Kalyan Vasudev Alwala, Nicolas Carion, Chao-Yuan Wu, Ross Girshick, Piotr Dollár, and Christoph Feichtenh...

    2024 arXiv

  37. [45]

    Learning transferable visual models from natural language supervision

    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 supervision. In International conference on machine learning, pa...

    2021

  38. [46]

    Scannet: Richly-annotated 3d reconstructions of indoor scenes

    Angela Dai, Angel X Chang, Manolis Savva, Maciej Halber, 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

    2017

  39. [47]

    A benchmark for the evaluation of rgb-d slam systems

    Jürgen Sturm, Nikolas Engelhard, Felix Endres, Wolfram Burgard, and Daniel Cremers. A benchmark for the evaluation of rgb-d slam systems. In 2012 IEEE/RSJ international conference on intelligent robots and systems, pages 573–580. IEEE, 2012

    2012

  40. [48]

    Loopsplat: Loop closure by registering 3d gaussian splats

    Liyuan Zhu, Yue Li, Erik Sandström, Shengyu Huang, Konrad Schindler, and Iro Armeni. Loopsplat: Loop closure by registering 3d gaussian splats. In International Conference on 3D Vision (3DV), 2025. 13

    2025

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.