Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

PG-SLAM: Photo-realistic and Geometry-aware RGB-D SLAM in Dynamic Environments

T0 review · 4 major / 5 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 The paper presents PG-SLAM, an RGB-D SLAM system that simultaneously reconstructs dynamic humans and rigid items, maps the static background, and estimates the camera trajectory using both foreground and background observations.

desk verdict Genuine first in dynamic GS SLAM, with consistent experiments, but the unguarded dependence on pretrained perception and missing code/baselines keep the empirical claim conditional. read the letter →

arxiv 2411.15800 v1 pith:PZCPJOYQ submitted 2024-11-24 cs.RO cs.CV

Haoang Li , Xiangqi Meng , Xingxing Zuo , Zhe Liu , Hesheng Wang , Daniel Cremers This is my paper · ORCID
classification cs.ROcs.CV
keywords RGB-DSLAMdynamicenvironments3DGaussiansplattingnon-rigidhumanreconstructionrigiditemtrackingopticalflowSMPLbodymodelcameralocalization
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

PG-SLAM is an RGB-D SLAM system that treats dynamic foreground as something to reconstruct rather than delete. It attaches 3D Gaussians to an SMPL body model to represent non-rigid humans, uses optical flow to track rigid items, and builds the static background in local maps. The same Gaussians are rendered into images and depths to provide appearance constraints, and their projected centers generate optical flows that constrain camera pose. The paper reports that this single pipeline maps moving humans and objects photorealistically while achieving lower camera trajectory error than prior methods on Bonn, TUM, and NeuMan benchmarks. If true, dynamic scenes no longer require choosing between complete mapping and accurate localization.

What carries the argument

The unifying object is a set of differentiable 3D Gaussians, each with a center, covariance, opacity, and color, rendered by alpha-blending along rays. The paper makes this representation dynamic: human Gaussians are bound to SMPL mesh facets and deformed by a per-frame MLP, item Gaussians are moved by rigid transformations estimated from optical-flow correspondences, and background Gaussians are organized into local maps. The same renderer supplies photometric and depth losses, and projected Gaussian centers produce synthesized optical flows that act as a geometric pose constraint. All three modules share one representation, which is what lets foreground mapping and camera localization reinforce each other.

What would settle it

Run the system on a synthetic dynamic scene with ground-truth masks, SMPL parameters, and optical flow, then add calibrated noise to each perception module separately and measure camera trajectory error and human trajectory error. The claim that foreground information improves localization would be falsified if noiseless perception does not improve over background-only pose estimation, or if perturbing one module collapses accuracy despite the multi-constraint design.

Watch

Extended reading notes

Core claim

The paper proposes that 3D Gaussian splatting can be extended from static scenes to dynamic ones by constraining Gaussians with two motion models: human Gaussians are attached to the SMPL model and deformed by a small MLP driven by pose changes, while item Gaussians are associated across frames by optical flow and aligned rigidly. A two-stage localizer first estimates pose from the static background, then refines it using both background and foreground, with a geometric constraint that compares observed RAFT optical flows to flows synthesized by projecting weighted Gaussian centers. The authors claim this is the first Gaussian splatting-based SLAM method that maps dynamic humans and items, and they report camera localization and human-trajectory accuracy better than existing methods on Bonn, TUM, and NeuMan.

Load-bearing premise

The load-bearing premise is that the pretrained perception modules — Mask R-CNN for segmentation, ReFit for SMPL pose and shape, and RAFT for optical flow — work reliably in the scenes being mapped, since errors in any one feed directly into Gaussian initialization, motion estimation, and the projected-flow pose constraint.

Editorial extensions

If this is right

  • Dynamic objects no longer have to be discarded: the same map that reconstructs a human or box can be reused to refine the camera pose, so scenes with large foregrounds become usable for tracking rather than being obstacles.
  • A single run produces a photorealistic, temporally consistent model of the static scene plus articulated humans and rigid objects, enabling novel-view rendering and human trajectory estimation as by-products of SLAM.
  • Local-map alignment that combines ICP-style geometry with appearance losses can reduce accumulated drift in loop-closure situations, not only between temporally adjacent maps.
  • On depth-noisy data, the scale-regularization loss keeps human sizes consistent over time, which directly improves human trajectory estimates.

Reading between the lines

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

  • The same Gaussian-plus-deformation machinery could extend to other deformable object classes: replace the SMPL prior with a category-level shape model and the pipeline could track hands, animals, or soft packages.
  • Because the projected-flow constraint derives from Gaussian centers and rendering weights, it could in principle replace RAFT with flows computed from optimized depths and poses, removing one pretrained dependency.
  • The reported dependence on ReFit and RAFT implies a performance ceiling: scenes with fast motion, occlusions, or depth noise that break those pretrained modules would violate the assumptions, so stress-testing on crowded multi-person scenes would reveal where the pipeline degrades.
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.

Referee Report

4 major / 5 minor

Summary. The paper proposes PG-SLAM, an RGB-D SLAM system that extends 3D Gaussian splatting to dynamic scenes. It represents non-rigid humans by Gaussians attached to an SMPL mesh and deformed over time by an MLP, represents rigid items by Gaussians moved by estimated rigid transformations, and reconstructs the static background through local maps that are aligned by a combined geometric-and-appearance optimization. Camera localization is formulated in two stages: a coarse pose from static background, followed by a refinement stage that jointly optimizes the pose and the foreground Gaussians using photometric, depth, and projected-optical-flow geometric constraints. Experiments on Bonn, TUM, and NeuMan report absolute trajectory error for camera and human trajectories, with ablations supporting the proposed geometric constraint, foreground use, human scale regularization, and local-map correction.

Significance. If the results hold, PG-SLAM would be a useful contribution: a single pipeline that maintains a photorealistic map of dynamic humans and items while also tracking the camera, rather than discarding dynamic content like filtering-based systems. The formulation is coherent and the ablations are consistently ordered in the expected direction: the geometric flow constraint improves camera ATE (Table II), using foreground improves it further (Table III), scale regularization substantially improves human-trajectory ATE on NeuMan (Table IV), and local-map correction helps on TUM (Table V). However, the empirical support has important gaps: the declared development baseline Gaussian-SLAM [33] is missing from the main comparison, no run-to-run variance is reported, scene-representation quality is only qualitative, and the system's dependence on pretrained Mask R-CNN, ReFit, and RAFT is not stress-tested. These gaps leave the headline 'outperforms state-of-the-art' and 'photo-realistic mapping' claims only partially verified.

major comments (4)
  1. [§VII-A3, §VII-B, Fig. 13] The paper states in §VII-A3, 'We treat Gaussian-SLAM [33] as the baseline to develop our SLAM method,' yet Table I compares only ESLAM, MonoGS, and Rodyn-SLAM. The one direct visual comparison to Gaussian-SLAM appears in Fig. 13, where the caption labels it as '(a) Gaussian-SLAM [33]' while the text calls the same panel 'Before optimization,' so it is unclear whether that panel is the ablation baseline or the development baseline. Because the closest baseline is precisely the one omitted, the claim in §VII-B3 that PG-SLAM 'outperforms state-of-the-art approaches' cannot be separated from the improvements inherited from Gaussian-SLAM. Please add Gaussian-SLAM (and, if available, the concurrent DG-SLAM [34]) to the quantitative comparison, and clarify the role of Fig. 13(a).
  2. [§VI-A, Eq. (14)] In the second-stage localization, the camera pose T_{t+1} and foreground Gaussians F^W_{t+1} are jointly optimized, but F^W_{t+1} is initialized from the first-stage pose and from the pretrained perception stack (Mask R-CNN masks, ReFit SMPL parameters, and RAFT optical flows described in Sections IV-A, IV-B2, and VI-B). This is not a formal circularity, but it is an unguarded feedback path: errors in any perception component enter both the foreground model and the pose-refinement objective, and no mechanism down-weights foreground constraints when those components are unreliable. The paper reports no sensitivity analysis or failure cases for the perception dependencies, and the 'SD' columns in Table I are per-trajectory standard deviations of ATE, not run-to-run variances. On NeuMan, where §VII-A1 notes limited depth quality, the root joint is obtained by back-projecting ReFit's 2D result with noisy depth, and the scale regularizer in §IV-A2 constrains Gaussian scale but not root translation; a small ReFit or depth error can therefore shift the human model and bias the refined pose. Please add perturbation or ablation experiments for mask, SMPL, and flow errors, and report run-to-run variance if any stochastic initialization exists.
  3. [§VII-B2, §VII-A2] The paper claims 'photo-realistic scene representation' as a central contribution, but the mapping quality is supported only by qualitative figures (Figs. 7, 8, 9) and human-trajectory ATE (Table IV). Table I quantifies camera localization only. Since the contribution includes mapping dynamic humans and items, quantitative reconstruction metrics such as rendered-frame PSNR/SSIM/LPIPS against observed RGB-D frames, or held-out-view synthesis quality, should be reported for PG-SLAM and for the compared methods. Without such metrics, the claim that PG-SLAM outperforms state-of-the-art approaches in scene representation is not quantitatively grounded.
  4. [Abstract, §I, §II-C, §VIII] The headline novelty claim—'the first Gaussian splatting-based SLAM method that can not only localize the camera and reconstruct the static background, but also map the dynamic humans and items'—is asserted without comparison to the concurrent DG-SLAM [34] discussed in §II-C, or to any other concurrent dynamic-GS SLAM system. The current text describes [34] only as a filtering-based method, but does not demonstrate that no other concurrent system maps dynamic foreground. Please add the relevant comparison or soften the claim to 'one of the first.' This is a documentation issue, but it directly concerns the paper's main claimed contribution.
minor comments (5)
  1. [§IV-A2, §IV-B2, §VII-B1] There are typos: 'Gaussisans' should be 'Gaussians' (two occurrences), and 'isotropoic' should be 'isotropic.'
  2. [Fig. 10, Table I] Sequence names are inconsistent (e.g., 'Walking_xyz' vs 'walking_xyz' vs 'wk_xyz'; 'Citron' vs 'citron'; 'parkinglot' vs 'parking_lot'). Please normalize the naming across the text, tables, and figure captions.
  3. [Reference [34]] The arXiv identifier for DG-SLAM appears as 'arXiv:411.08373'; this should likely be 'arXiv:2411.08373'.
  4. [§V-B] The text says the local-map optimization 'can be applied to loop closure,' but no loop-closure experiment or implementation detail is provided. Please state explicitly whether loop closure is evaluated, or mark it as future work.
  5. [§VII-A3] No runtime or per-frame processing time is reported. The paper does not claim real-time performance, but reporting the runtime on the RTX 4090 setup would help readers assess practical applicability.

Circularity Check

2 steps flagged · score 6.0 of 10

Dynamic-foreground pose refinement is gauge-equivalent to foreground re-fitting: Eq. (14) and Eq. (17) do not constrain the camera pose by construction.

  1. self definitional [Section IV-A2 (world-frame transform) and Section VI-A, Eq. (14)]
    "After optimizing Gaussians Gk+1 in the (k + 1)-th camera frame, we use the camera pose Tt+1 to transform these Gaussians to the world frame. ... Recall that given RGB-D images (It+1, Dt+1) and coarse camera pose Tt+1 obtained in the first stage, we can roughly estimate a set of foreground Gaussians F W t+1 in the world frame (see Section IV). ... foreground Gaussians may not be accurate enough due to fewer views, and thus are treated as variables to optimize."

    The foreground map F^W_{t+1} is constructed by applying the very pose T_{t+1} that Eq. (14) claims to refine: Section IV transforms the camera-frame foreground Gaussians into the world frame using T_{t+1}, and Eq. (14) then minimizes over both T_{t+1} and F^W_{t+1} against the same single-view observation. Since F^W can be redefined as T' T^{-1}_coarse F^W_coarse for any candidate pose T', the rendered foreground patches are unchanged and the foreground photometric/depth residual has no dependence on T.

  2. self definitional [Section VI-B2, Eq. (17)]
    "The projected optical flow ˜f is additionally with respect to the Gaussian center µW_k, compared with Eq. (15). Then by analogy with the static foreground, we minimize the difference between the projected flow ˜f in Eq. (17) and the observed flow f extracted by RAFT algorithm. This achieves a joint optimization of the transformation Tt+1 and Gaussian centers {µW_k}."

    The dynamic-foreground optical-flow constraint has the same gauge freedom. The centers µ^W_k are world-frame dynamic Gaussian centers obtained in Section IV using the pose T_{t+1} being estimated, and Eq. (17) makes the projected flow depend on both T_{t+1} and µ^W_k before jointly optimizing both. A simultaneous change T→T' and µ^W→T' T^{-1} µ^W leaves each projected point q_k, and hence the projected flow ˜f, unchanged; the residual ∥˜f−f∥ therefore does not constrain T. By construction, the dynamic-foreground geometric constraint is vacuous for camera localization, so the paper's stated mechanism for using foreground to improve pose accuracy is not supported by its own equations.

full rationale

The derivation is largely self-contained and externally grounded: foreground initialization relies on external pretrained components (Mask R-CNN, ReFit, RAFT, SMPL), and the main empirical claims are evaluated against ground-truth trajectories on Bonn, TUM, and NeuMan. Self-references such as the TUM benchmark and DSO are standard external evidence and are not load-bearing. The genuine circularity is confined to the foreground pose-refinement loop. Section IV builds the world-frame foreground Gaussians F^W_{t+1} using the very camera pose T_{t+1} that Section VI-A then tries to refine (for humans: 'we use the camera pose Tt+1 to transform these Gaussians to the world frame'; for items, the rigid transformation ∆M also uses T_{t+1}). Eq. (14) then optimizes T_{t+1} and F^W_{t+1} jointly against a single view, so any pose change can be absorbed by re-defining F^W; the foreground appearance term is invariant under (T,F^W)→(T',T'T^{-1}F^W). The same gauge freedom invalidates the dynamic-foreground optical-flow constraint in Eq. (17), where centers µ^W_k and T are jointly optimized. Thus the paper's central explanation of why foreground improves localization—'leverage information of both static background and dynamic foreground to compensate for noise'—does not follow from the stated equations; the foreground term does not constrain the pose by construction. The Table III improvements may reflect the extra optimization of F rather than foreground pose constraints; this is a partial circularity rather than a fully vacuous paper, because the background-only first stage and the background constraints are independent. Section VII-A1's caveat about NeuMan depth quality is also relevant: the root joint is back-projected from noisy depth, and the scale regularizer in Table IV constrains Gaussian scale rather than root translation, so human-tracking numbers are sensitive to perception noise; this is a robustness risk rather than a circularity. Novelty and coverage issues (the uncompared Gaussian-SLAM baseline and the characterization of concurrent DG-SLAM) are correctness concerns, not circularity.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The system introduces no new physical entities; it is an empirical pipeline. Its central claim rests on hand-set loss weights, a data-derived scale reference, and several pretrained perception models whose outputs are trusted without sensitivity analysis.

free parameters (3)
  • Loss weights lambda_P and lambda_D = 0.6 and 0.4
    Hand-selected weights for the photometric and depth losses; all mapping and localization losses use these values, and no sensitivity study is reported.
  • Human scale reference s_hat = average scale of Gaussians initialized from the first frame
    Used as the constant standard scale in the human scale regularization loss in Section IV-A2; it is computed from the first observation and constrains later human sizes. The ablation in Table IV shows it strongly affects human trajectory error.
  • Deformation network D architecture = 8 layers, 256-dimensional hidden features, two readout heads
    Architecture chosen by the authors for the human deformation network; no ablation of capacity or training details is reported.
assumptions (5)
  • domain assumption SMPL and ReFit provide accurate human pose, shape, and root-joint estimates in every frame.
    Used to initialize and deform human Gaussians in Section IV-A; errors in pose or root depth propagate to the human map.
  • domain assumption Mask R-CNN produces accurate masks separating humans, items, and background.
    All sets of Gaussians are initialized and rendered from these masks; no failure analysis for segmentation errors is provided.
  • domain assumption RAFT optical flow provides reliable 2D correspondences on both static and dynamic regions.
    Used for rigid item transformations in Section IV-B2 and for projected-flow geometric constraints in Section VI-B.
  • domain assumption Depth images are approximately metric and aligned with RGB images.
    Back-projection, scale recovery, and depth losses assume this; the paper notes depth quality is partly limited on NeuMan in Section VII-A1.
  • domain assumption The constant-velocity motion model gives a good initialization for the next camera pose.
    Used in the first-stage localization in Section VI-A, inherited from ORB-SLAM2 [3].

how reviews work

0 comments
Cite this review

Pith. "Pith review of PG-SLAM: Photo-realistic and Geometry-aware RGB-D SLAM in Dynamic Environments." pith.science (2026). https://pith.science/paper/PZCPJOYQ

@misc{pith2026241115800,
  author       = {Pith},
  title        = {Pith review of: PG-SLAM: Photo-realistic and Geometry-aware RGB-D SLAM in Dynamic Environments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PZCPJOYQ}},
  note         = {Machine review of arXiv:2411.15800}
}
read the original abstract

Simultaneous localization and mapping (SLAM) has achieved impressive performance in static environments. However, SLAM in dynamic environments remains an open question. Many methods directly filter out dynamic objects, resulting in incomplete scene reconstruction and limited accuracy of camera localization. The other works express dynamic objects by point clouds, sparse joints, or coarse meshes, which fails to provide a photo-realistic representation. To overcome the above limitations, we propose a photo-realistic and geometry-aware RGB-D SLAM method by extending Gaussian splatting. Our method is composed of three main modules to 1) map the dynamic foreground including non-rigid humans and rigid items, 2) reconstruct the static background, and 3) localize the camera. To map the foreground, we focus on modeling the deformations and/or motions. We consider the shape priors of humans and exploit geometric and appearance constraints of humans and items. For background mapping, we design an optimization strategy between neighboring local maps by integrating appearance constraint into geometric alignment. As to camera localization, we leverage both static background and dynamic foreground to increase the observations for noise compensation. We explore the geometric and appearance constraints by associating 3D Gaussians with 2D optical flows and pixel patches. Experiments on various real-world datasets demonstrate that our method outperforms state-of-the-art approaches in terms of camera localization and scene representation. Source codes will be publicly available upon paper acceptance.

Figures

Figures reproduced from arXiv: 2411.15800 by the authors.

Figure 1
Figure 1. Overview of our SLAM method. Given sequential RGB-D images obtained in a dynamic environment, our method can not only reconstruct the static [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 3
Figure 3. Update of human Gaussians. In the SMPL frame, we use a neural [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 2
Figure 2. Initialization of human Gaussians. Given the first RGB-D images, [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Rigid transformation and addition of item Gaussians. We first estimate [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Optimization between n-th and (n + 1)-th local maps. Here, we show the centers of Gaussians. We iteratively align Gaussians Gn+1 to Gaussians Gn based on geometric constraint. To improve the robustness of optimization, we integrate the appearance constraint into each i…
Figure 6
Figure 6. Figure 6: Geometric constraint of optical flows. We mainly consider the static [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 8
Figure 8. Figure 8: Representative environment mapping results of our [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: Representative human trajectories estimated by our [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Representative camera trajectories estimated by various SLAM methods on the Sequence [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 11
Figure 11. Figure 11: Ablation study of camera localization using geometric constraint and dynamic foreground on Sequence [PITH_FULL_IMAGE:figures/full_fig_p013_11.png]
Figure 13
Figure 13. Figure 13: Ablation study of optimization between neighboring local maps [PITH_FULL_IMAGE:figures/full_fig_p014_13.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. EndoFlow-SLAM: Real-Time Endoscopic SLAM with Flow-Constrained Gaussian Splatting

    cs.CV 2025-06 conditional novelty 5.0 of 10

    EndoFlow-SLAM couples 3D Gaussian Splatting with optical-flow and depth-gradient supervision, and reports improved pose and rendering accuracy on static and dynamic endoscopic datasets.

Reference graph

Works this paper leans on

44 extracted references · 41 canonical work pages · cited by 1 Pith paper

  1. [33]

    Gaussian-SLAM: Photo- realistic Dense SLAM with Gaussian Splatting,

    V . Yugay, Y . Li, T. Gevers, and M. R. Oswald, “Gaussian-SLAM: Photo- realistic Dense SLAM with Gaussian Splatting,” 2024

    work page 2024

  2. [34]

    DG-SLAM: Robust Dynamic Gaussian Splatting SLAM with Hybrid Pose Optimization,

    Y . Xu, H. Jiang, Z. Xiao, J. Feng, and L. Zhang, “DG-SLAM: Robust Dynamic Gaussian Splatting SLAM with Hybrid Pose Optimization,” arXiv preprint arXiv:411.08373 , 2024

    work page 2024

  3. [1]

    Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age,

    C. Cadena, L. Carlone, H. Carrillo, Y . Latif, D. Scaramuzza, J. Neira, I. Reid, and J. J. Leonard, “Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age,” IEEE Transactions on Robotics , vol. 32, no. 6, pp. 1309–1332, Dec. 2016

    work page 2016

  4. [2]

    MonoSLAM: Real-Time Single Camera SLAM,

    A. J. Davison, I. D. Reid, N. D. Molton, and O. Stasse, “MonoSLAM: Real-Time Single Camera SLAM,” IEEE Transactions on Pattern Anal- ysis and Machine Intelligence , vol. 29, no. 6, pp. 1052–1067, Jul. 2007

    work page 2007

  5. [3]

    ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras,

    R. Mur-Artal and J. D. Tardós, “ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras,” IEEE Transac- tions on Robotics , vol. 33, no. 5, pp. 1255–1262, Oct. 2017

    work page 2017

  6. [4]

    Direct Sparse Odometry,

    J. Engel, V . Koltun, and D. Cremers, “Direct Sparse Odometry,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 40, no. 3, pp. 611–625, Mar. 2017

    work page 2017

  7. [5]

    iMAP: Implicit Map- ping and Positioning in Real-Time,

    E. Sucar, S. Liu, J. Ortiz, and A. J. Davison, “iMAP: Implicit Map- ping and Positioning in Real-Time,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 6229–6238

    work page 2021

  8. [6]

    NICE-SLAM: Neural Implicit Scalable Encoding for SLAM,

    Z. Zhu, S. Peng, V . Larsson, W. Xu, H. Bao, Z. Cui, M. R. Oswald, and M. Pollefeys, “NICE-SLAM: Neural Implicit Scalable Encoding for SLAM,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2022, pp. 12 786–12 796

    work page 2022

Show all 44 references
  1. [7]

    GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting,

    C. Yan, D. Qu, D. Xu, B. Zhao, Z. Wang, D. Wang, and X. Li, “GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 19 595–19 604

    2024

  2. [8]

    R. I. Hartley and A. Zisserman, Multiple View Geometry in Computer Vision, 2nd ed. Cambridge University Press, 2004

    2004

  3. [9]

    Motion removal for reliable RGB- D SLAM in dynamic environments,

    Y . Sun, M. Liu, and M. Q.-H. Meng, “Motion removal for reliable RGB- D SLAM in dynamic environments,” Robotics and Autonomous Systems , vol. 108, pp. 115–128, Oct. 2018

    2018

  4. [10]

    DS- SLAM: A Semantic Visual SLAM towards Dynamic Environments,

    C. Yu, Z. Liu, X.-J. Liu, F. Xie, Y . Yang, Q. Wei, and Q. Fei, “DS- SLAM: A Semantic Visual SLAM towards Dynamic Environments,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems , 2018, pp. 1168–1174

    2018

  5. [11]

    RoDyn-SLAM: Robust Dynamic Dense RGB-D SLAM with Neural Radiance Fields,

    H. Jiang, Y . Xu, K. Li, J. Feng, and L. Zhang, “RoDyn-SLAM: Robust Dynamic Dense RGB-D SLAM with Neural Radiance Fields,” IEEE Robotics and Automation Letters , vol. 9, no. 9, pp. 7509–7516, Sept. 2024

    2024

  6. [12]

    VDO-SLAM: A visual dynamic object-aware SLAM system,

    J. Zhang, M. Henein, R. Mahony, and V . Ila, “VDO-SLAM: A visual dynamic object-aware SLAM system,” arXiv preprint arXiv:2005.11052, 2020

    2005 arXiv

  7. [13]

    DynaSLAM II: Tightly-Coupled Multi-Object Tracking and SLAM,

    B. Bescos, C. Campos, J. D. Tardós, and J. Neira, “DynaSLAM II: Tightly-Coupled Multi-Object Tracking and SLAM,” IEEE Robotics and Automation Letters, vol. 6, no. 3, pp. 5191–5198, Jul. 2021

    2021

  8. [14]

    AirDOS: Dynamic SLAM benefits from Articulated Objects,

    Y . Qiu, C. Wang, W. Wang, M. Henein, and S. Scherer, “AirDOS: Dynamic SLAM benefits from Articulated Objects,” in Proceedings of the IEEE International Conference on Robotics and Automation , 2022, pp. 8047–8053

    2022

  9. [15]

    BodySLAM: Joint Camera Localisation, Mapping, and Human Motion Tracking,

    D. F. Henning, T. Laidlow, and S. Leutenegger, “BodySLAM: Joint Camera Localisation, Mapping, and Human Motion Tracking,” in Pro- ceedings of the European Conference on Computer Vision , 2022, pp. 656–673

    2022

  10. [16]

    3D Gaussian Splatting for Real-Time Radiance Field Rendering,

    B. Kerbl, G. Kopanas, T. Leimkühler, and G. Drettakis, “3D Gaussian Splatting for Real-Time Radiance Field Rendering,” ACM Transactions on Graphics , vol. 42, no. 4, Jul. 2023

    2023

  11. [17]

    SMPL: A Skinned Multi-Person Linear Model,

    M. Loper, N. Mahmood, J. Romero, G. Pons-Moll, and M. J. Black, “SMPL: A Skinned Multi-Person Linear Model,” ACM Transactions on Graphics, vol. 34, no. 6, pp. 248:1–248:16, Oct. 2015

    2015

  12. [18]

    ElasticFusion: Real-time dense SLAM and light source estimation,

    T. Whelan, R. F. Salas-Moreno, B. Glocker, A. J. Davison, and S. Leutenegger, “ElasticFusion: Real-time dense SLAM and light source estimation,” The International Journal of Robotics Research , vol. 35, no. 14, pp. 1697–1716, Sept. 2016

    2016

  13. [19]

    3-D Mapping With an RGB-D Camera,

    F. Endres, J. Hess, J. Sturm, D. Cremers, and W. Burgard, “3-D Mapping With an RGB-D Camera,”IEEE Transactions on Robotics, vol. 30, no. 1, pp. 177–187, Feb. 2014

    2014

  14. [20]

    Learning Meshes for Dense Visual SLAM,

    M. Bloesch, T. Laidlow, R. Clark, S. Leutenegger, and A. Davison, “Learning Meshes for Dense Visual SLAM,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2019, pp. 5854–5863

    2019

  15. [21]

    V oxel Map for Visual SLAM,

    M. Muglikar, Z. Zhang, and D. Scaramuzza, “V oxel Map for Visual SLAM,” in Proceedings of the IEEE International Conference on Robotics and Automation , 2020, pp. 4181–4187

    2020

  16. [22]

    Flowfusion: Dynamic dense rgb-d slam based on optical flow,

    T. Zhang, H. Zhang, Y . Li, Y . Nakamura, and L. Zhang, “Flowfusion: Dynamic dense rgb-d slam based on optical flow,” in Proceedings of the IEEE International Conference on Robotics and Automation , 2020, pp. 7322–7328

    2020

  17. [23]

    RGB-D SLAM in Dynamic Environments Using Point Correlations,

    W. Dai, Y . Zhang, P. Li, Z. Fang, and S. Scherer, “RGB-D SLAM in Dynamic Environments Using Point Correlations,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 44, no. 1, pp. 373– 389, Jan. 2022

    2022

  18. [24]

    DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation,

    J. J. Park, P. Florence, J. Straub, R. Newcombe, and S. Lovegrove, “DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2019, pp. 165–174

    2019

  19. [25]

    NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis,

    B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoorthi, and R. Ng, “NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis,” in Proceedings of the European Conference on Computer Vision, 2020, pp. 405–421

    2020

  20. [26]

    V ox- Fusion: Dense Tracking and Mapping with V oxel-based Neural Implicit Representation,

    X. Yang, H. Li, H. Zhai, Y . Ming, Y . Liu, and G. Zhang, “V ox- Fusion: Dense Tracking and Mapping with V oxel-based Neural Implicit Representation,” in Proceedings of the IEEE International Symposium on Mixed and Augmented Reality , 2022, pp. 499–507

    2022

  21. [27]

    ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of Signed Distance Fields,

    M. M. Johari, C. Carta, and F. Fleuret, “ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of Signed Distance Fields,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 17 408–17 419

    2023

  22. [28]

    Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural Real-Time SLAM,

    H. Wang, J. Wang, and L. Agapito, “Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural Real-Time SLAM,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 13 293–13 302

    2023

  23. [29]

    GO-SLAM: Global Optimization for Consistent 3D Instant Reconstruction,

    Y . Zhang, F. Tosi, S. Mattoccia, and M. Poggi, “GO-SLAM: Global Optimization for Consistent 3D Instant Reconstruction,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 3727–3737

    2023

  24. [30]

    Point- SLAM: Dense Neural Point Cloud-based SLAM,

    E. Sandström, Y . Li, L. Van Gool, and M. R. Oswald, “Point- SLAM: Dense Neural Point Cloud-based SLAM,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 18 433–18 444

    2023

  25. [31]

    Gaussian Splatting SLAM,

    H. Matsuki, R. Murai, P. H. J. Kelly, and A. J. Davison, “Gaussian Splatting SLAM,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 18 039–18 048

    2024

  26. [32]

    SplaTAM: Splat, Track & Map 3D Gaussians for Dense RGB-D SLAM,

    N. Keetha, J. Karhade, K. M. Jatavallabhula, G. Yang, S. Scherer, D. Ramanan, and J. 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 , 2024, pp. 21 357–21 366

    2024

  27. [35]

    Image quality assess- ment: from error visibility to structural similarity,

    Z. Wang, A. Bovik, H. Sheikh, and E. Simoncelli, “Image quality assess- ment: from error visibility to structural similarity,” IEEE Transactions on Image Processing , vol. 13, no. 4, pp. 600–612, Apr. 2004

    2004

  28. [36]

    Mask R-CNN,

    K. He, G. Gkioxari, P. Dollár, and R. Girshick, “Mask R-CNN,” in Proceedings of the IEEE International Conference on Computer Vision , 2017, pp. 2980–2988

    2017

  29. [37]

    ReFit: Recurrent Fitting Network for 3D Human Recovery,

    Y . Wang and K. Daniilidis, “ReFit: Recurrent Fitting Network for 3D Human Recovery,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 14 598–14 608

    2023

  30. [38]

    RAFT: Recurrent All-Pairs Field Transforms for Optical Flow ,

    Z. Teed and J. Deng, “RAFT: Recurrent All-Pairs Field Transforms for Optical Flow ,” in Proceedings of the European Conference on Computer Vision, 2021, pp. 4839–4843

    2021

  31. [39]

    Least-Squares Fitting of Two 3-D Point Sets,

    K. S. Arun, T. S. Huang, and S. D. Blostein, “Least-Squares Fitting of Two 3-D Point Sets,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. PAMI-9, no. 5, pp. 698–700, Sept. 1987

    1987

  32. [40]

    LoopSplat: Loop Closure by Registering 3D Gaussian Splats,

    L. Zhu, Y . Li, E. Sandström, S. Huang, K. Schindler, and I. Armeni, “LoopSplat: Loop Closure by Registering 3D Gaussian Splats,” arXiv preprint arXiv:2408.10154, 2024

    2024 arXiv

  33. [41]

    A benchmark for the evaluation of RGB-D SLAM systems,

    J. Sturm, N. Engelhard, F. Endres, W. Burgard, and D. Cremers, “A benchmark for the evaluation of RGB-D SLAM systems,” in Proceed- ings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, 2012, pp. 573–580

    2012

  34. [42]

    NeuMan: Neural Human Radiance Field from a Single Video,

    W. Jiang, K. M. Yi, G. Samei, O. Tuzel, and A. Ranjan, “NeuMan: Neural Human Radiance Field from a Single Video,” in Proceedings of the European Conference on Computer Vision , 2022, pp. 402–418

    2022

  35. [43]

    ReFu- sion: 3D Reconstruction in Dynamic Environments for RGB-D Cameras Exploiting Residuals,

    E. Palazzolo, J. Behley, P. Lottes, P. Giguère, and C. Stachniss, “ReFu- sion: 3D Reconstruction in Dynamic Environments for RGB-D Cameras Exploiting Residuals,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems , 2019, pp. 7855–7862

    2019

  36. [44]

    A Tutorial on Quantitative Trajec- tory Evaluation for Visual(-Inertial) Odometry,

    Z. Zhang and D. Scaramuzza, “A Tutorial on Quantitative Trajec- tory Evaluation for Visual(-Inertial) Odometry,” in Proceedings of the 16 IEEE/RSJ International Conference on Intelligent Robots and Systems , 2018, pp. 7244–7251

    2018

Pith tools

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