Pith. sign in

REVIEW 4 major objections 5 minor 2 cited by

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

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 DG-SLAM claims to be the first robust dynamic visual SLAM system grounded in 3D Gaussians, using fused depth-warp and semantic motion masks plus coarse-to-fine hybrid pose optimization to keep moving objects from corrupting the static map.

desk verdict A solid dynamic 3DGS SLAM system with real benchmark gains, but the motion mask's dependence on DROID-VO pose error is never stress-tested. read the letter →

arxiv 2411.08373 v1 pith:HJX2UDG6 submitted 2024-11-13 cs.RO

Yueming Xu , Haochen Jiang , Zhongyang Xiao , Jianfeng Feng , Li Zhang This is my paper · ORCID
classification cs.RO
keywords dynamicSLAM3DGaussiansplattingmotionmaskhybridposeoptimizationRGB-Dscenesnovel-viewsynthesiscameratracking
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

The paper sets out to build a visual SLAM system that keeps tracking accurately while people and objects move through the scene, and that rebuilds only the static background as a 3D Gaussian map. It claims that a fused motion mask—combining multi-frame depth-warp residuals with a semantic segmentation prior—can suppress dynamic pixels during both pose optimization and mapping, and that a hybrid coarse-to-fine tracker, starting from a learned dense optical-flow odometry [19] and refined by Gaussian-splatting alignment, gives precise poses. If correct, this would extend Gaussian Splatting SLAM from empty static rooms to ordinary indoor scenes, and the paper reports the lowest absolute trajectory error and best reconstruction quality among the compared systems on the TUM, BONN, and ScanNet datasets.

What carries the argument

The load-bearing mechanism is the motion mask plus a coarse-to-fine pose optimizer. The motion mask is built from depth-warp residuals (Eq. 6) and their fusion (Eq. 7): multiple observation masks within a sliding window are intersected for spatial consistency, then the semantic mask is unioned in, so the final mask is precise at edges and still catches non-semantic movers. The hybrid tracking uses [19] as the coarse pose through masked dense bundle adjustment (Eq. 8), then Gaussian-splatting fine alignment (Eq. 10) with a reliability mask derived from accumulated opacity, which prevents dynamic pixels and under-optimized Gaussians from steering the pose. Adaptive Gaussian point addition and pruning keeps the map geometry clean and prevents floaters in the rendered images.

What would settle it

Take a sequence with a large, slow-moving object covering most of the frame, or induce coarse-pose drift through fast camera motion, and compare the generated motion mask's precision and recall against ground-truth dynamic annotations; if the depth-warp mask mislabels large static regions as dynamic or misses the moving object when the coarse pose is off by a few centimeters, the claimed robustness fails in exactly the regime the paper targets.

Watch

Extended reading notes

Core claim

DG-SLAM's central claim is that a 3D-Gaussian SLAM system can be made robust in dynamic scenes by explicitly separating moving objects from the static map. The core object is a fused motion mask: per-pixel depth-warp residuals between keyframes in a sliding window are intersected over time to remove noise at edges, then unioned with a semantic dynamic-object mask, so that objects a semantic network misses (balloons, boxes) are still caught. This mask suppresses dynamic pixels in the masked dense bundle adjustment of the coarse stage and in the photometric and depth losses of the fine Gaussian-splatting stage. With this separation, the Gaussian map is optimized on static pixels only, and pose estimates are refined against a clean rendering. The paper reports that this yields lower ATE than optical-flow-plus-semantic dynamic SLAM baselines on TUM and BONN, and lower reconstruction error than neural-implicit or other Gaussian SLAM baselines on dynamic BONN sequences, while preserving real-time rendering ability.

Load-bearing premise

The whole pipeline depends on the coarse pose from the learned odometry and the depth values being accurate enough that the depth-warp residual cleanly separates moving from static pixels; if the initial pose drifts badly in scenes with large moving regions, the mask itself becomes wrong and the masked optimization will reinforce the error instead of fixing it.

Editorial extensions

If this is right

  • Dynamic objects no longer corrupt the Gaussian map: map optimization uses only pixels outside the motion mask, so the reconstructed static background is clean even when people carry balloons or boxes.
  • Pose estimation becomes robust in dynamic scenes: masking dynamic pixels in both the dense bundle adjustment and the fine Gaussian alignment removes a major source of photometric and geometric inconsistency.
  • The system retains real-time rendering ability while achieving the paper's reported ATE and reconstruction improvements over prior Gaussian and neural-implicit SLAM baselines on dynamic sequences.
  • Semantic segmentation errors can be compensated by the depth-warp masks, and depth-warp edge noise is reduced by spatio-temporal consistency, so the fused mask is more precise than either cue alone.
  • The masked Gaussian map supports novel-view synthesis of the static scene, extending the utility of the map beyond tracking to rendering and editing.

Reading between the lines

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

  • If this approach holds, the same motion-mask fusion could be dropped into other Gaussian or neural-implicit SLAM backends as a front-end filter, not only for SLAM but for static-map reconstruction from handheld RGB-D video in busy scenes.
  • The paper's reliance on a learned dense optical-flow odometry for the coarse pose suggests a failure mode when that odometry itself drifts; testing with degraded coarse poses would show how much of the claimed robustness comes from the mask versus the oracle pose.
  • The 163 ms semantic segmentation step could be swapped for a faster lightweight segmenter or self-supervised motion cues to push the system closer to true real-time operation without changing the fusion strategy.
  • Because the motion mask only removes pixels labeled dynamic through the sliding-window intersection, an object that is static at first but starts moving later (or vice versa) may need a forgetting mechanism for stale warp-mask labels.
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

4 major / 5 minor

Summary. The manuscript proposes DG-SLAM, a dynamic RGB-D SLAM system built on 3D Gaussian Splatting. The pipeline uses a motion mask computed from multi-frame depth-warp residuals and semantic segmentation to avoid optimizing over dynamic pixels, a coarse-to-fine tracking stage that starts from DROID-VO poses and then refines them against the Gaussian map, and adaptive Gaussian point addition/pruning for map quality. The system is evaluated on TUM RGB-D, BONN RGB-D Dynamic, and ScanNet, with tables reporting ATE, reconstruction accuracy/completion/completion ratio, and runtime. The paper claims state-of-the-art pose estimation, reconstruction, and novel-view synthesis in dynamic scenes while retaining real-time rendering.

Significance. If the results hold, DG-SLAM is a useful step toward making 3D Gaussian Splatting SLAM work in scenes with moving objects, an area where existing Gaussian SLAM systems are weak. The design is sensible: masking dynamic pixels before bundle adjustment and fine tracking, and using coarse-to-fine pose optimization, are technically reasonable. The paper is accompanied by a public code link, evaluates on standard benchmarks, and includes ablations that support the contribution of each component. However, the strength of the claim depends on sensitivity to the coarse pose source and on statistical reliability, neither of which is demonstrated beyond single-run point estimates.

major comments (4)
  1. [Sec. 3.2-3.3, Eqs. (6)-(10)] The central robustness claim rests on the depth-warp motion mask, but the paper does not quantify how the system behaves when the DROID-VO coarse pose is inaccurate. Eq. (6) warps pixels using the coarse pose; the resulting mask is applied in the DBA objective (Eq. 8) and in both terms of the fine tracking loss (Eq. 10). Section 3.3 itself acknowledges that an inaccurate initial pose degrades the depth-warp mask, yet no experiment perturbs the coarse pose or analyzes a sequence where DROID-VO drifts. Since DROID-VO alone achieves 3.3 cm average ATE on TUM (Table 2) and 15.4 cm on BONN (Table 3), the improvement to 2.2 cm and 5.5 cm could, in some regimes, be due to the mask-and-refine loop recovering from good coarse poses rather than to genuine robustness. A controlled pose-noise ablation (e.g., adding known translation/rotation perturbations to the DROID-VO output and measuring final ATE and mask accuracy) would settle whether the system degrades gracefully or fails exactly where it claims robustness.
  2. [Sec. 4, Tables 2-4 and 6] All quantitative results are single-run point estimates. Tables 2, 3, 4, and 6 report no error bars, no standard deviations across multiple runs, and no significance tests. The implementation section lists hand-set hyperparameters (loss weights lambda1=0.9, lambda2=0.2, lambda3=0.1; tracking/mapping iterations 20/40; pruning thresholds tau_alpha=0.005, tau_S1=0.4, tau_S2=36; depth threshold 0.6; window size 4), but no sensitivity analysis is given. This matters because several ablations in Table 6 differ by only about 1 cm (e.g., w/o Add 6.63 vs. 5.51, w/o Prune 6.89 vs. 5.51), and the 'state-of-the-art' claim is based on margins that may be within run-to-run noise. At minimum, the authors should provide repeated runs or a hyperparameter sweep to show that the reported margins are stable.
  3. [Eq. (6)] The depth-warp mask definition in Eq. (6) is ambiguous and is not fully consistent with the surrounding text. The equation states a condition D_j(p_{i->j}) - D_i(p) < e_th, while Section 3.2 then says 'we only mask the foreground pixels where the depth residual is positive.' A one-sided inequality with a positive threshold already excludes negative residuals, so the additional sentence is confusing; if a signed residual is intended, the equation should state it explicitly. The notation with the tensor product and I_{m x n} is nonstandard and should be replaced by an explicit elementwise predicate so that the mask can be reproduced unambiguously.
  4. [Sec. 3.4, Eq. (13)] The point-pruning criterion in Eq. (13) uses max(S) and min(S), but the scale vector S_i is never defined in the Gaussian representation of Eq. (1), which instead introduces the covariance matrix Sigma_i. This makes the pruning rule and the initialization description in Section 3.4 ('the scale vector S_i is initialized...') formally incomplete. The authors should either define the scale-rotation decomposition of Sigma_i explicitly or rewrite Eq. (13) in terms of Sigma_i's eigenvalues.
minor comments (5)
  1. [Various headings] There are typos and grammatical issues, e.g., 'Adapative Gaussian point adding strategy' and 'we reduce the gradient-based dynamic radius to half so that can add more Gaussian points.' These do not affect the technical content but should be corrected.
  2. [Eq. (2) vs. Eq. (5)] The transformation notation is inconsistent: Eq. (2) uses T^c_w while Eq. (5) uses T_ji for the relative pose. Please unify the notation.
  3. [References] References [14] and [29] are the same paper (Mildenhall et al., NeRF); one duplicate should be removed and the citation indices updated.
  4. [Table 5] Table 5 lists only iteration counts, but Section 4.3 says these results 'take TUM as an example' without specifying which sequences or whether the reported values are averaged; please clarify.
  5. [Section 4.5] The runtime comparison in Table 7 excludes semantic segmentation time, which is later reported as 163 ms per frame. The abstract's 'real-time rendering ability' claim should be reconciled with this extra cost, perhaps by reporting end-to-end runtime including segmentation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; pose and reconstruction claims are measured against external benchmarks, and the coarse-to-fine mask/pose loop is iterative refinement, not derivation-by-construction.

full rationale

The paper's central claims—camera pose accuracy, reconstruction quality, and novel-view synthesis—are evaluated on external datasets (TUM, BONN, ScanNet) against ground-truth trajectories and point clouds, so they are not manufactured from the method's own fitted parameters. The coarse-to-fine pipeline is a feedback loop rather than a circular definition: DROID-VO supplies a coarse pose, Eqs. (5)-(7) build a motion mask from depth warping and semantics, Eq. (8) performs masked bundle adjustment, and Eq. (10) refines the pose by minimizing photometric and depth residuals against rendered Gaussians. The final pose is the minimizer of an independent objective, not an algebraic consequence of the mask; the mask itself is an intermediate estimate that can be wrong, as the paper acknowledges. The self-citation of Rodyn-SLAM [33] is used as a comparison baseline and as related work, and it is not load-bearing for any of DG-SLAM's derivations. The paper's stated limitations—large-scale tracking and sensitivity to semantic segmentation precision—are robustness concerns, not evidence of circular reasoning. No fitted constant is relabeled as a prediction, and no result is forced by a self-citation chain.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The central claim is supported by a pipeline of tuned thresholds and pretrained black boxes (3DGS, DROID-VO, OneFormer) rather than by new physical entities. The listed free parameters are hand-set rather than fitted to the evaluation metrics in a disclosed way, and no sensitivity analysis is reported.

free parameters (7)
  • Loss weights lambda1, lambda2, lambda3 = 0.9, 0.2, 0.1
    Hand-chosen weights in Eqs. 10 and 12 balancing color, SSIM, and depth losses; no sensitivity analysis is provided.
  • Depth warp threshold e_th = 0.6
    Classifies dynamic pixels in Eq. 6; directly controls motion mask quality.
  • Depth warp window size N = 4
    Number of keyframes combined in Eq. 7 for spatio-temporal mask consistency.
  • Gaussian pruning thresholds tau_alpha, tau_S1, tau_S2 = 0.005, 0.4, 36
    Eq. 13 criteria for deleting abnormal Gaussians; affects map cleanliness and tracking quality.
  • Reliable mask opacity threshold tau_track = unspecified
    Used in Eq. 9 to decide which rendered pixels are trustworthy for fine tracking; value is not reported.
  • Opacity-based point addition threshold o_th = unspecified
    Used in adaptive point adding when accumulated opacity falls below threshold; value is not reported.
  • Tracking and mapping iteration counts = 20 and 40
    Fixed optimization budgets; performance and runtime depend on these values.
assumptions (5)
  • domain assumption The 3D Gaussian Splatting alpha-blending model (Eqs. 2-3) faithfully represents the scene for rendering and pose optimization.
    Inherited from 3DGS [18]; the entire fine tracking and mapping pipeline relies on this rendering model being differentiable and accurate.
  • ad hoc to paper Dynamic objects can be recognized by depth-warp residual thresholding combined with semantic segmentation (Eqs. 6-7).
    Core design assumption; threshold e_th and window N are hand-set, and failure modes at object edges or in mostly dynamic scenes are not analyzed.
  • ad hoc to paper Accumulated opacity O_hat (Eq. 4) is a reliable indicator of Gaussian map optimization quality.
    Used to build the reliable mask in Eq. 9; no evidence is given that opacity correlates with geometric correctness.
  • domain assumption DROID-VO provides a sufficiently accurate coarse pose in dynamic scenes for the warp mask and fine tracking.
    Section 3.3 uses DROID-VO as the coarse stage; if it fails, the mask and refinement both degrade.
  • domain assumption The semantic segmentation network OneFormer produces priors accurate enough to support the motion mask.
    Table 6 ablation shows removing the semantic mask increases average ATE from 5.51 cm to 15.27 cm, so the central result depends on this external pretrained model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DG-SLAM: Robust Dynamic Gaussian Splatting SLAM with Hybrid Pose Optimization." pith.science (2026). https://pith.science/paper/HJX2UDG6

@misc{pith2026241108373,
  author       = {Pith},
  title        = {Pith review of: DG-SLAM: Robust Dynamic Gaussian Splatting SLAM with Hybrid Pose Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HJX2UDG6}},
  note         = {Machine review of arXiv:2411.08373}
}
read the original abstract

Achieving robust and precise pose estimation in dynamic scenes is a significant research challenge in Visual Simultaneous Localization and Mapping (SLAM). Recent advancements integrating Gaussian Splatting into SLAM systems have proven effective in creating high-quality renderings using explicit 3D Gaussian models, significantly improving environmental reconstruction fidelity. However, these approaches depend on a static environment assumption and face challenges in dynamic environments due to inconsistent observations of geometry and photometry. To address this problem, we propose DG-SLAM, the first robust dynamic visual SLAM system grounded in 3D Gaussians, which provides precise camera pose estimation alongside high-fidelity reconstructions. Specifically, we propose effective strategies, including motion mask generation, adaptive Gaussian point management, and a hybrid camera tracking algorithm to improve the accuracy and robustness of pose estimation. Extensive experiments demonstrate that DG-SLAM delivers state-of-the-art performance in camera pose estimation, map reconstruction, and novel-view synthesis in dynamic scenes, outperforming existing methods meanwhile preserving real-time rendering ability.

Figures

Figures reproduced from arXiv: 2411.08373 by the authors.

Figure 1
Figure 1. Overview of DG-SLAM. Given a series of RGB-D frames, we reconstruct the static high-fidelity 3D Gaussian map and optimize the camera pose represented with lie algebra ξi . RGB-D SLAM with neural implicit representation. Neural implicit scene representations, also known as neural fields [29], have attracted considerable attention in the field of RGB-D SLAM for their impressive expressiveness and low memory footprint.… view at source ↗
Figure 2
Figure 2. Qualitative results of the motion mask generation. By fusing the semantic mask and depth warp mask, the final mask will be more precise. geometric depth loss and photometric color loss : Lmapping = λ1Lrgb + λ2Lssim + λ3Ldepth, (12) where Lssim denotes the structural similarity loss between two images. Moreover, λ1, λ2, λ3 are weight factors for balance in the optimization process. Adapative Gaussian point adding str… view at source ↗
Figure 3
Figure 3. Visual comparison of the rendering image on the TUM and BONN datasets. Our results are more complete and accurate without the dynamic object floaters. ball ball2 ps_trk ps_trk2 mv_box2 Avg. NICE-SLAM[7] Acc.[cm]↓ X 24.30 43.11 74.92 17.56 39.97 Comp.[cm]↓ X 16.65 117.95 172.20 18.19 81.25 Comp. Ratio[≤ 5cm%]↑ X 29.68 15.89 13.96 32.18 22.93 Co-SLAM[8] Acc.[cm]↓ 10.61 14.49 26.46 26.00 12.73 18.06 Comp.[cm]↓ 10.65 40… view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. ContraGS: Codebook-Condensed and Trainable Gaussian Splatting for Fast, Memory-Efficient Reconstruction

    cs.GR 2025-09 reject novelty 7.0 of 10

    ContraGS trains 3D Gaussian Splatting directly on codebook-compressed representations, cutting peak model memory ~3.5x with small quality loss.

  2. DyPho-SLAM : Real-time Photorealistic SLAM in Dynamic Environments

    cs.RO 2025-08 conditional novelty 6.0 of 10

    DyPho-SLAM uses prior-image masks and adaptive feature selection to keep camera tracking accurate while building a photorealistic static 3D map in real time.

Reference graph

Works this paper leans on

44 extracted references · 44 canonical work pages · cited by 2 Pith papers

  1. [1]

    Simultaneous map building and localization for an autonomous mobile robot

    John J Leonard and Hugh F Durrant-Whyte. Simultaneous map building and localization for an autonomous mobile robot. In IROS, 1991. 1

    work page 1991

  2. [2]

    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 TRO, 2017. 1, 2 10

    work page 2017

  3. [3]

    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 TRO, 2021. 1, 2, 8, 9

    work page 2021

  4. [4]

    Ul-slam: A universal monocular line-based slam via unifying structural and non-structural constraints

    Haochen Jiang, Rui Qian, Liang Du, Jian Pu, and Jianfeng Feng. Ul-slam: A universal monocular line-based slam via unifying structural and non-structural constraints. IEEE TASE,

    work page

  5. [5]

    Refusion: 3d reconstruction in dynamic environments for rgb-d cameras exploiting residuals

    Emanuele Palazzolo, Jens Behley, Philipp Lottes, Philippe Giguere, and Cyrill Stachniss. Refusion: 3d reconstruction in dynamic environments for rgb-d cameras exploiting residuals. In IROS, 2019. 1, 2, 4, 6, 8, 9

    work page 2019

  6. [6]

    iMAP: Implicit mapping and positioning in real-time

    Edgar Sucar, Shikun Liu, Joseph Ortiz, and Andrew Davison. iMAP: Implicit mapping and positioning in real-time. In ICCV, 2021. 1, 3, 8, 9

    work page 2021

  7. [7]

    Oswald, and Marc Pollefeys

    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 CVPR,

    work page

  8. [8]

    Co-slam: Joint coordinate and sparse parametric encodings for neural real-time slam

    Hengyi Wang, Jingwen Wang, and Lourdes Agapito. Co-slam: Joint coordinate and sparse parametric encodings for neural real-time slam. In CVPR, 2023. 1, 3, 7, 8, 9, 10

    work page 2023

Show all 44 references
  1. [9]

    Eslam: Efficient dense slam system based on hybrid representation of signed distance fields

    Mohammad Mahdi Johari, Camilla Carta, and François Fleuret. Eslam: Efficient dense slam system based on hybrid representation of signed distance fields. In CVPR, 2023. 1, 3, 7, 8, 9, 10

    2023

  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 ISMAR,

  3. [11]

    Kinectfusion: Real-time dense surface mapping and tracking

    Richard A Newcombe, Shahram Izadi, Otmar Hilliges, David Molyneaux, David Kim, Andrew J Davison, Pushmeet Kohi, Jamie Shotton, Steve Hodges, and Andrew Fitzgibbon. Kinectfusion: Real-time dense surface mapping and tracking. In ISMAR, 2011. 1

    2011

  4. [12]

    Surfelmeshing: Online surfel-based mesh reconstruction

    Thomas Schöps, Torsten Sattler, and Marc Pollefeys. Surfelmeshing: Online surfel-based mesh reconstruction. IEEE TPAMI, 2019. 1

    2019

  5. [13]

    Bad slam: Bundle adjusted direct rgb-d slam

    Thomas Schops, Torsten Sattler, and Marc Pollefeys. Bad slam: Bundle adjusted direct rgb-d slam. In CVPR, 2019. 1

    2019

  6. [14]

    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. Commun Acm, 2021. 1

    2021

  7. [15]

    Gs-slam: Dense visual slam with 3d gaussian splatting

    Chi Yan, Delin Qu, Dong Wang, Dan Xu, Zhigang Wang, Bin Zhao, and Xuelong Li. Gs-slam: Dense visual slam with 3d gaussian splatting. In CVPR, 2024. 2, 3

    2024

  8. [16]

    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 CVPR, 2024. 2, 3, 7, 8, 9

    2024

  9. [17]

    Hidenobu Matsuki, Riku Murai, Paul H. J. Kelly, and Andrew J. Davison. Gaussian Splatting SLAM. In CVPR, 2024. 2, 3, 7, 8, 9

    2024

  10. [18]

    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 TOG, 2023. 2, 3, 4

    2023

  11. [19]

    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. In NeurIPS, 2021. 2, 5, 7, 8

    2021

  12. [20]

    Ds-slam: A semantic visual slam towards dynamic environments

    Chao Yu, Zuxin Liu, Xin-Jun Liu, Fugui Xie, Yi Yang, Qi Wei, and Qiao Fei. Ds-slam: A semantic visual slam towards dynamic environments. In IROS, 2018. 2

    2018

  13. [21]

    Dynaslam: Tracking, mapping, and inpainting in dynamic scenes

    Berta Bescos, José M Fácil, Javier Civera, and José Neira. Dynaslam: Tracking, mapping, and inpainting in dynamic scenes. IEEE RAL, 2018. 2 11

    2018

  14. [22]

    Dynamic-slam: Semantic monocular visual localization and mapping based on deep learning in dynamic environment

    Linhui Xiao, Jinge Wang, Xiaosong Qiu, Zheng Rong, and Xudong Zou. Dynamic-slam: Semantic monocular visual localization and mapping based on deep learning in dynamic environment. IEEE RAS, 2019. 2

    2019

  15. [23]

    Dynaslam ii: Tightly-coupled multi-object tracking and slam

    Berta Bescos, Carlos Campos, Juan D Tardós, and José Neira. Dynaslam ii: Tightly-coupled multi-object tracking and slam. IEEE RAL, 2021. 2

    2021

  16. [24]

    Vdo-slam: a visual dynamic object-aware slam system

    Jun Zhang, Mina Henein, Robert Mahony, and Viorela Ila. Vdo-slam: a visual dynamic object-aware slam system. arXiv preprint, 2020. 2

    2020

  17. [25]

    Crowd-slam: visual slam towards crowded environments using object detection

    João Carlos Virgolino Soares, Marcelo Gattass, and Marco Antonio Meggiolaro. Crowd-slam: visual slam towards crowded environments using object detection. Journal of Intelligent & Robotic Systems, 2021. 2

    2021

  18. [26]

    Ddn-slam: Real-time dense dynamic neural implicit slam with joint semantic encoding

    Mingrui Li, Jiaming He, Guangan Jiang, and Hongyu Wang. Ddn-slam: Real-time dense dynamic neural implicit slam with joint semantic encoding. arXiv preprin, 2024. 2

    2024

  19. [27]

    Dytanvo: Joint refinement of visual odometry and motion segmentation in dynamic environments

    Shihao Shen, Yilin Cai, Wenshan Wang, and Sebastian Scherer. Dytanvo: Joint refinement of visual odometry and motion segmentation in dynamic environments. In ICRA, 2023. 2

    2023

  20. [28]

    Deflowslam: Self-supervised scene motion decomposition for dynamic dense slam

    Weicai Ye, Xingyuan Yu, Xinyue Lan, Yuhang Ming, Jinyu Li, Hujun Bao, Zhaopeng Cui, and Guofeng Zhang. Deflowslam: Self-supervised scene motion decomposition for dynamic dense slam. arXiv preprint, 2022. 2

    2022

  21. [29]

    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. Commun ACM, 2021. 3

    2021

  22. [30]

    Di-fusion: Online implicit 3d reconstruction with deep priors

    Jiahui Huang, Shi-Sheng Huang, Haoxuan Song, and Shi-Min Hu. Di-fusion: Online implicit 3d reconstruction with deep priors. In CVPR, 2021. 3

    2021

  23. [31]

    Erik Sandström, Yue Li, Luc Van Gool, and Martin R. Oswald. Point-slam: Dense neural point cloud-based slam. In ICCV, 2023. 3, 6, 9, 10

    2023

  24. [32]

    Point-nerf: Point-based neural radiance fields

    Qiangeng Xu, Zexiang Xu, Julien Philip, Sai Bi, Zhixin Shu, Kalyan Sunkavalli, and Ulrich Neumann. Point-nerf: Point-based neural radiance fields. In CVPR, 2022. 3

    2022

  25. [33]

    Rodyn-slam: Robust dynamic dense rgb-d slam with neural radiance fields

    Haochen Jiang, Yueming Xu, Kejie Li, Jianfeng Feng, and Li Zhang. Rodyn-slam: Robust dynamic dense rgb-d slam with neural radiance fields. RAL, 2024. 3, 8, 9

    2024

  26. [34]

    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, 2023. 3

    2023

  27. [35]

    Compact 3d gaussian splatting for dense visual slam

    Tianchen Deng, Yaohui Chen, Leyan Zhang, Jianfei Yang, Shenghai Yuan, Danwei Wang, and Weidong Chen. Compact 3d gaussian splatting for dense visual slam. arXiv preprint, 2024. 3

    2024

  28. [36]

    Parallel tracking and mapping for small ar workspaces

    Georg Klein and David Murray. Parallel tracking and mapping for small ar workspaces. In ISMAR, 2007. 3

    2007

  29. [37]

    Newcombe, Steven J

    Richard A. Newcombe, Steven J. Lovegrove, and Andrew J. Davison. Dtam: Dense tracking and mapping in real-time. In ICCV, 2011. 3

    2011

  30. [38]

    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 IROS, 2012. 6, 7

    2012

  31. [39]

    Chang, Manolis Savva, Maciej Halber, Thomas Funkhouser, and Matthias Nießner

    Angela Dai, Angel X. Chang, Manolis Savva, Maciej Halber, Thomas Funkhouser, and Matthias Nießner. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In CVPR, 2017. 6, 9

    2017

  32. [40]

    Closed-form solution of absolute orientation using unit quaternions

    Berthold KP Horn. Closed-form solution of absolute orientation using unit quaternions. Josa a,

  33. [41]

    One- Former: One Transformer to Rule Universal Image Segmentation

    Jitesh Jain, Jiachen Li, MangTik Chiu, Ali Hassani, Nikita Orlov, and Humphrey Shi. One- Former: One Transformer to Rule Universal Image Segmentation. In CVPR, 2023. 7 12

    2023

  34. [42]

    Co-fusion: Real-time segmentation, tracking and fusion of multiple objects

    Martin Rünz and Lourdes Agapito. Co-fusion: Real-time segmentation, tracking and fusion of multiple objects. In ICRA, 2017. 8

    2017

  35. [43]

    Mid-fusion: Octree-based object-level multi-instance dynamic slam

    Binbin Xu, Wenbin Li, Dimos Tzoumanikas, Michael Bloesch, Andrew Davison, and Stefan Leutenegger. Mid-fusion: Octree-based object-level multi-instance dynamic slam. In ICRA,

  36. [44]

    Em-fusion: Dynamic object-level slam with probabilistic data association

    Michael Strecke and Jorg Stuckler. Em-fusion: Dynamic object-level slam with probabilistic data association. In ICCV, 2019. 8, 9 13

    2019

Pith tools

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