Pith. sign in

REVIEW 4 major objections 5 minor 2 cited by

MAGiC-SLAM: Multi-Agent Gaussian Globally Consistent SLAM

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 MAGiC-SLAM claims that a rigidly deformable 3D Gaussian scene representation lets multiple agents track, map, and render a scene together with centimeter-level accuracy and far less compute than prior multi-agent SLAM.

desk verdict First multi-agent Gaussian SLAM with loop closure, worth a real review once the covariance update in Eq. (13) is fixed and code is released. read the letter →

arxiv 2411.16785 v1 pith:K7JD5P4T submitted 2024-11-25 cs.CV cs.AIcs.RO

Vladimir Yugay , Theo Gevers , Martin R. Oswald This is my paper · ORCID
classification cs.CVcs.AIcs.RO
keywords multi-agentSLAM3DGaussiansplattingnovelviewsynthesisloopclosureposegraphoptimizationRGB-DDINOv2globalmapmerging
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 introduces MAGiC-SLAM, a simultaneous localization and mapping system that lets several cameras or robots explore the same scene at once, sharing their work. The paper aims to show that a scene representation built from rigidly deformable 3D Gaussians makes multi-agent SLAM both accurate enough and fast enough for real-world use, solving problems in earlier neural multi-agent systems: slow tracking, poor rendering of real scenes, and a hard limit of two agents. By adding a loop-closure stage that detects revisited places with features from a large pretrained vision model, the method corrects drift across agents and merges their submaps into one globally consistent map. On synthetic and real-world benchmarks the authors report centimeter-level trajectory error, substantially higher novel-view rendering quality than the prior multi-agent system, and a large reduction in runtime and memory use. If the claims hold, multi-agent SLAM with photorealistic maps becomes practical for robotics, augmented reality, and large-scene reconstruction.

What carries the argument

The load-bearing object is the rigidly deformable 3D Gaussian submap: each agent's local map is a set of 3D Gaussians that can be moved and rotated as a whole when loop-closure corrections arrive, unlike neural fields that cannot be rigidly transformed. The other mechanisms that carry the argument are: a hybrid tracking module that first registers consecutive colored point clouds with coarse-to-fine ICP for a pose seed, then refines the pose by minimizing color and depth re-rendering errors against the frozen Gaussian map; a centralized loop-closure stage that uses DINOv2 image features stored in a FAISS database to find revisited submaps, registers anchor point clouds with FPFH plus RANSAC and ICP to form loop edges, and optimizes all submap poses with g2o; and a two-stage map merger that appends cached Gaussians coarsely and then optimizes the global Gaussian set against color and depth losses to remove artifacts. The key property that makes all of this work is that Gaussian submaps support rigid body transformations, so corrected poses can be applied directly to the map.

What would settle it

Run MAGiC-SLAM on an Aria room sequence that intentionally includes a moving person or another dynamic object for a substantial portion of the trajectory, and measure ATE RMSE and rendered PSNR against the provided ground truth; if the error rises far above the reported 0.90 cm average or rendering quality collapses, the claim that the method handles real-world data and generalizes beyond the selected sequences is falsified.

Watch

Extended reading notes

Core claim

The central claim is that replacing the distributed neural scene representation used by the previous state of the art, CP-SLAM, with a rigidly deformable 3D Gaussian representation yields a multi-agent SLAM system that is more accurate, faster, and not limited to two agents. Each agent builds small Gaussian submaps, tracks its camera with a two-stage implicit method (dense frame-to-frame initialization followed by frame-to-model refinement), and sends compact submaps plus image features to a central server. The server detects loop closures with DINOv2-based descriptors, optimizes a pose graph over submaps, pushes corrected poses back to the agents, and finally merges all submaps into a coherent global map that can render novel views. The authors report average ATE RMSE of 0.25 cm on ReplicaMultiagent and 0.90 cm on AriaMultiagent, rendering quality of 34.26 dB PSNR on ReplicaMultiagent training views and 22.61 dB PSNR on AriaMultiagent novel views, and per-frame mapping and tracking around 0.7 s, versus CP-SLAM's 17 s mapping and 3.4 s tracking, with peak GPU use dropping from 7.70 GiB to 1.12 GiB.

Load-bearing premise

The real-world evaluation depends on the assumption that hand-picked Aria sequences, chosen for having enough consecutive frames without dynamic objects, represent realistic multi-agent operation; if dynamic objects or less cooperative trajectories are present, the claimed real-world accuracy and rendering quality could drop substantially.

Editorial extensions

If this is right

  • Multi-agent teams can map larger spaces faster: each agent handles its own stream, so adding agents adds coverage without changing the core algorithm, limited only by server capacity.
  • Loop closure no longer requires revisiting with the same agent; any agent's revisit of another agent's mapped area can close a loop, which is what lets multi-agent trajectories be globally aligned.
  • The same pipeline produces a renderable 3D map usable for novel view synthesis, not just a sparse or dense geometric map, so downstream augmented-reality and inspection tasks get a photorealistic model from the same run.
  • Runtime and memory are low enough that per-agent mapping and tracking run at about 0.7 seconds per frame on an RTX A6000, much closer to interactive use than the previous neural system's 17 seconds per frame mapping.
  • The number of agents is not hard-coded; the architecture handles two agents in the synthetic dataset and three in the real-world evaluation, with no architectural limit beyond server resources.

Reading between the lines

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

  • If the reported speedups carry over to online operation, the main remaining bottleneck is the implicit tracker's iteration count, which the paper itself states; reducing those iterations could push the system toward real-time multi-agent use without changing the architecture.
  • Because loop detection relies on DINOv2 features rather than geometric place recognition, the method may generalize to outdoor and texture-poor scenes, but it also inherits DINOv2's failure modes, a tradeoff the paper does not test.
  • The coarse-to-fine map-merging stage suggests a standalone recipe: cache only visible Gaussians, append, then globally refine; this could be applied to single-agent lifelong SLAM systems that accumulate many submaps over time.
  • The evaluation reports novel-view rendering only on the real-world dataset, while synthetic rendering is on training views; a direct novel-view benchmark on synthetic scenes would separate tracking quality from rendering generalization, a test the paper leaves implicit.
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 MAGiC-SLAM, a centralized multi-agent RGB-D SLAM system built on 3D Gaussian splatting. Each agent maintains local sub-maps, performs two-stage tracking (coarse-to-fine colored ICP initialization followed by re-rendering-based pose refinement), and sends sub-map data and DINOv2 image features to a centralized server. The server detects loop closures, registers anchor point clouds from sub-maps, optimizes a pose graph, corrects camera poses and Gaussian parameters, and finally merges the refined sub-maps into a single global Gaussian map. Experiments on the MultiagentReplica dataset and a hand-picked three-agent subset of the Aria dataset report lower ATE RMSE, better rendering metrics, and lower runtime and memory usage than CP-SLAM and several single- and multi-agent baselines.

Significance. If the reported results reflect the actual implementation, the contribution is substantial: it is the first Gaussian-splatting SLAM system for more than two collaborative agents that supports loop closure and globally consistent map merging while retaining novel-view synthesis quality. The experimental numbers are strong (e.g., average ATE 0.25 cm on ReplicaMultiagent, PSNR 34.26 dB, and a large runtime reduction relative to CP-SLAM), and the paper includes useful ablations of pose initialization, loop detection, and map-merging strategies. The choice of DINOv2 for loop detection and the coarse-to-fine merging strategy are concrete and reasonable. However, the central pose-update equation is mathematically incorrect as written, and the absence of released code or data prevents the reader from resolving whether the implementation follows the correct transformation. The real-world evaluation also relies on a curated subset of the Aria data with no quantitative selection statistics.

major comments (4)
  1. [3.3, Eq. (13)] Equation (13) states that the covariance of each Gaussian is updated as Sigma_j <- T^c_{i,R} Sigma_j, where T^c_{i,R} is the rotation part of the SE(3) correction. Under a rigid rotation, a covariance matrix must transform congruently as Sigma' = R Sigma R^T. The published update is not a congruence transformation: for a general anisotropic Gaussian it does not preserve symmetry or positive semi-definiteness, and applying the rotation on only one side would scatter anisotropic splats when loop-closure corrections are integrated. If the intended operation is to left-multiply the per-Gaussian rotation quaternion while keeping scales unchanged, then the covariance representation is updated correctly but Eq. (13) is a misleading typo and must be rewritten, e.g., as Sigma' = R Sigma R^T or by explicitly stating that only the stored orientation is left-multiplied. Since the text says 'Sigma' and no code is released, the reader cannot tell whether the implementation follows the incorrect formula. This must be clarified because the claimed global consistency and rendering improvements depend on correct Gaussian transformation under loop closure.
  2. [4, Datasets] The real-world evaluation uses AriaMultiagent, described as three sequences per room selected for having 'sufficient consecutive frames without dynamic objects'. The paper gives no statistics on how much of the original Aria data was discarded, how the three sequences were chosen, or whether the 100 novel-view test frames are temporally separated from the 500 training frames. Without this information, it is difficult to assess whether Tables 2 and 5 support the claimed generalization to real multi-agent operation. Please report the selection procedure quantitatively, including the fraction of frames retained and the temporal gap between training and test frames.
  3. [4.1, Tables 1-3] All tracking evaluations report per-agent ATE RMSE, which measures each agent's trajectory against its own ground truth but does not directly measure cross-agent consistency of the merged global map. Since the central claim is 'globally consistent' multi-agent reconstruction, the evaluation should include an inter-agent consistency metric, such as relative pose error between agents in overlapping regions, map-to-map Chamfer distance, or registration residuals between corrected sub-maps. The current rendering metrics indirectly reflect map quality but do not isolate cross-agent alignment.
  4. [3.3 / Reproducibility] The ambiguity in Eq. (13) is compounded by the fact that no code or data is released. A precise pseudocode block for the Gaussian update (or a code release) is needed to show that the loop-closure correction is applied as a congruence transformation and not as the one-sided multiplication written in the paper. This is a load-bearing point for the reproducibility of the reported global-consistency results.
minor comments (5)
  1. [Table 3] The MonoGS row for ReplicaMultiagent Apt-0 appears garbled as '0.380.213.33'; the intended values should be separated or the table regenerated.
  2. [Table 5 caption] The caption says the global map is built by merging the maps from two agents, but the MAGiC-SLAM experiments in Table 2 process three agents; please clarify whether the rendering evaluation uses two or three agents for MAGiC-SLAM.
  3. [3.1, Eq. (2)] The SSIM weighting parameter lambda in the color loss is not listed in Supplementary Table B.1, which only reports lambda_c, learning rates, and iteration counts; please include the value used for lambda.
  4. [4, Evaluation protocol] Each table reports a single number per cell without error bars or repeated trials. If any component of the pipeline is stochastic (e.g., sampling of new Gaussians or optimization), please state determinism or provide repeated-run statistics; otherwise the 'more accurate' claim rests on single runs.
  5. [5, Limitations] The paper states the system operates 'slightly faster than 1 FPS', but Table 8 reports 0.69 s per tracking frame and 0.71 s per mapping frame, which is about 1.4 FPS for the combined pipeline; please make the throughput statement consistent with the reported per-frame times.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: MAGiC-SLAM's tracking, loop closure, and map merging are built from independently specified losses, external ICP/FPFH registration, and a pretrained DINOv2 detector; same-author citations are not load-bearing.

full rationale

The paper's derivation chain is self-contained and does not reduce any claimed result to its own inputs. Tracking (§3.2) predicts camera poses by initializing them with an external colored-ICP registration and refining against rendering losses (Eqs. 5–9) that compare the Gaussian splat rendering with the input RGB-D frames; no target pose value is fitted into the loss. Loop closure (§3.3) uses a pretrained DINOv2 model for detection, FPFH+RANSAC followed by ICP on raw point clouds for constraint estimation, and standard g2o pose-graph optimization (Eqs. 10–11), all independent of the Gaussian map parameters. The pose-update integration (Eqs. 12–13) is a direct geometric application of the optimized corrections, not an equivalence with the input. Map merging (§3.4) optimizes Gaussian parameters against color/depth rendering losses and evaluates rendering on held-out novel views; reporting PSNR/SSIM/LPIPS on these independent test frames is a standard evaluation, not a circular prediction. Same-author citations (Gaussian-SLAM [43], Point-SLAM [32], Loopy-SLAM [20], NICE-SLAM [47]) appear as baselines, related work, or as sources of generic sub-map and loop-closure techniques that are re-implemented with external components; none is invoked as an unverified authority for the central multi-agent claim. The potential issue in Eq. (13) (missing right-side transpose in the covariance update) is a correctness/implementation concern, not a circularity, and therefore does not affect this score.

Assumptions & free parameters 14 free parameters · 7 assumptions · 0 invented entities

The system's central claim rests on a large set of per-dataset hyperparameters (Table B.1), standard pose-graph and registration machinery, and the representativeness of the evaluation data. No new physical entities are introduced. The ledger shows that the paper's contribution is an engineered integration rather than a derivation, and the correctness of the covariance update in Eq. (13) is an unverified standard-math axiom.

free parameters (14)
  • tracking loss color weight λ_c = 0.95 (ReplicaMultiagent), 0.6 (AriaMultiagent)
    Tuned per dataset in Table B.1; directly controls pose refinement in Eq. (9).
  • rotation learning rate lr = 0.0002 (Replica), 0.002 (Aria)
    Per-dataset; affects tracking convergence.
  • translation learning rate lt = 0.002 (Replica), 0.01 (Aria)
    Per-dataset; affects tracking convergence.
  • tracking optimization iterations itert = 60 (Replica), 200 (Aria)
    Per-dataset; more iterations improve ATE at computational cost.
  • mapping optimization iterations iterm = 100 (both)
    Fixed per dataset; affects map quality.
  • sub-map creation interval = 50 frames (Replica), 20 frames (Aria)
    Chosen to synchronize agents and server; affects loop closure node density.
  • new Gaussian sample count Mk = 60K (Replica), 100K (Aria)
    Number of candidate Gaussians added per keyframe; affects map density.
  • alpha threshold α_thre = 0.98
    Condition for adding new Gaussians; manual threshold.
  • opacity pruning threshold o_thre = 0.1
    Prunes Gaussians after mapping; affects map size and quality.
  • loop detection feature threshold θ_feature = 0.35
    DINOv2 distance threshold for declaring loop; critical for loop closure behavior.
  • loop time threshold θ_time = not reported
    Filters same-agent loops; value absent from main text and supplementary.
  • SSIM weight λ in color loss = not reported
    Used in Eq. (2); exact value not given.
  • ICP color-geometric weight σ = not reported
    Weight in Eq. (5); value not specified.
  • global map refinement iterations = 3000
    Fine merging optimization budget; affects final rendering quality.
assumptions (7)
  • standard math SE(3) pose graph optimization with g2o correctly represents and minimizes the residuals in Eqs. (10)-(11).
    Pose graph optimization is taken as a reliable black box from [17].
  • standard math Rigidly transforming a 3D Gaussian requires updating the covariance as R Σ Rᵀ; the paper's update in Eq. (13) is inconsistent with this standard rule.
    If the covariance update is not corrected, rendering after pose updates is not mathematically faithful to the transformed map.
  • domain assumption DINOv2 features provide reliable loop detection across unseen environments at a fixed threshold θ_feature=0.35.
    The paper relies on this for loop closure generalization but provides no calibration or failure analysis.
  • domain assumption First-frame point clouds of different agents' sub-maps have sufficient overlap and density for coarse-to-fine FPFH+RANSAC+ICP registration.
    Loop constraints depend entirely on this registration succeeding across agents.
  • domain assumption The selected AriaMultiagent sequences, with dynamic objects excluded, are representative of real-world multi-agent SLAM.
    Section 4 states sequences were chosen for 'sufficient consecutive frames without dynamic objects'; the amount and criteria of exclusion are not quantified.
  • domain assumption The centralized server has sufficient bandwidth and compute to receive sub-maps and features and return optimized poses without violating the online multi-agent assumption.
    No communication latency or bandwidth model is provided; the system is described as centralized.
  • domain assumption Multi-scale colored ICP converges to an accurate relative pose from the proposed initialization.
    Tracking accuracy relies on this; details deferred to [28].

how reviews work

0 comments
Cite this review

Pith. "Pith review of MAGiC-SLAM: Multi-Agent Gaussian Globally Consistent SLAM." pith.science (2026). https://pith.science/paper/K7JD5P4T

@misc{pith2026241116785,
  author       = {Pith},
  title        = {Pith review of: MAGiC-SLAM: Multi-Agent Gaussian Globally Consistent SLAM},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K7JD5P4T}},
  note         = {Machine review of arXiv:2411.16785}
}
read the original abstract

Simultaneous localization and mapping (SLAM) systems with novel view synthesis capabilities are widely used in computer vision, with applications in augmented reality, robotics, and autonomous driving. However, existing approaches are limited to single-agent operation. Recent work has addressed this problem using a distributed neural scene representation. Unfortunately, existing methods are slow, cannot accurately render real-world data, are restricted to two agents, and have limited tracking accuracy. In contrast, we propose a rigidly deformable 3D Gaussian-based scene representation that dramatically speeds up the system. However, improving tracking accuracy and reconstructing a globally consistent map from multiple agents remains challenging due to trajectory drift and discrepancies across agents' observations. Therefore, we propose new tracking and map-merging mechanisms and integrate loop closure in the Gaussian-based SLAM pipeline. We evaluate MAGiC-SLAM on synthetic and real-world datasets and find it more accurate and faster than the state of the art.

Figures

Figures reproduced from arXiv: 2411.16785 by the authors.

Figure 1
Figure 1. MAGiC-SLAM is a multi-agent SLAM method capable of novel view synthesis. Given single-camera RGBD input streams from multiple simultaneously operating agents MAGiC-SLAM estimates their trajectories and reconstructs a 3D Gaussian map that can be rendered from previously unseen viewpoints. We showcase the high-fidelity 3D Gaussian map of a real-world environment alongside multiple agent trajectories (depicted in green… view at source ↗
Figure 2
Figure 2. MAGiC-SLAM Architecture. Agent Side: Each agent processes a separate RGBD stream, maintaining a local sub-map and estimating its trajectory. When an agent starts a new sub-map, it sends the previous sub-map and image features to the centralized server. Server Side: The server stores the image features and sub-maps from all agents and performs loop closure detection, loop constraint estimation, and pose graph optimiz… view at source ↗
Figure 3
Figure 3. Map Merging. Our coarse-to-fine strategy effectively removes (a) visual artifacts caused by the GS mechanism and (c) geometric artifacts resulting from Gaussian sub-map intersections. systems like SWARM-SLAM [18], CCM-SLAM [33], and CP-SLAM [11]. To make the comparison more comprehen￾sive, we include several single-agent systems like Gaussian￾SLAM [43], MonoGS [23], and Orb-SLAM3 [4]. We include Orb-SLAM3 since it i… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Rendering performance on ReplicaMultiagent [11]. Thanks to GS scene representation and effective merging strategy, MAGiC-SLAM encodes more high-frequency details and substantially increases the quality of the renderings. Methods Datasets Avg. MAGiC-SLAM (w. NetVLAD [1]…

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. Hierarchical Scoring with 3D Gaussian Splatting for Instance Image-Goal Navigation

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A two-stage scorer, CLIP semantic prefiltering plus DINOv2 geometric matching over a 3D Gaussian map, achieves a 0.784 success rate on HM3D instance image-goal navigation.

  2. HAMMER: Heterogeneous, Multi-Robot Semantic Gaussian Splatting

    cs.RO 2025-01 conditional novelty 5.0 of 10

    A multi-robot system aligns heterogeneous camera streams into a common frame and continually trains a semantic 3D Gaussian Splatting map for language-guided navigation.

Reference graph

Works this paper leans on

47 extracted references · 38 canonical work pages · cited by 2 Pith papers

  1. [1]

    Netvlad: Cnn architecture for weakly supervised place recognition, 2016

    Relja Arandjelovi ´c, Petr Gronat, Akihiko Torii, Tomas Pa- jdla, and Josef Sivic. Netvlad: Cnn architecture for weakly supervised place recognition, 2016. 5, 8, 2

    work page 2016

  2. [2]

    Barron, Ben Mildenhall, Matthew Tancik, Peter Hedman, Ricardo Martin-Brualla, and Pratul P

    Jonathan T. Barron, Ben Mildenhall, Matthew Tancik, Peter Hedman, Ricardo Martin-Brualla, and Pratul P. Srinivasan. Mip-nerf: A multiscale representation for anti-aliasing neu- ral radiance fields, 2021. 2

    2021

  3. [3]

    Besl and Neil D

    P.J. Besl and Neil D. McKay. A method for registration of 3-d shapes. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 14(2):239–256, 1992. 4, 5

    work page 1992

  4. [4]

    Gomez Rodriguez, Jose M

    Carlos Campos, Richard Elvira, Juan J. Gomez Rodriguez, Jose M. M. Montiel, and Juan D. Tardos. Orb-slam3: An accurate open-source library for visual, visual–inertial, and multimap slam. IEEE Transactions on Robotics , 37(6): 1874–1890, 2021. 6, 7

    work page 2021

  5. [5]

    How, and Luca Carlone

    Yun Chang, Yulun Tian, Jonathan P. How, and Luca Carlone. Kimera-multi: a system for distributed multi-robot metric- semantic simultaneous localization and mapping. In 2021 IEEE International Conference on Robotics and Automation (ICRA), pages 11210–11218, 2021. 2, 3

    work page 2021

  6. [6]

    Orbeez-slam: A real- time monocular visual slam with orb features and nerf- realized mapping

    Chi-Ming Chung, Yang-Che Tseng, Ya-Ching Hsu, Xiang- Qian Shi, Yun-Hung Hua, Jia-Fong Yeh, Wen-Chin Chen, Yi-Ting Chen, and Winston H Hsu. Orbeez-slam: A real- time monocular visual slam with orb features and nerf- realized mapping. arXiv preprint arXiv:2209.13274, 2022. 2

    arXiv 2022

  7. [7]

    V olumetric method for build- ing complex models from range images

    Brian Curless and Marc Levoy. V olumetric method for build- ing complex models from range images. InSIGGRAPH Con- ference on Computer Graphics. ACM, 1996. 1

    work page 1996

  8. [8]

    The faiss library,

    Matthijs Douze, Alexandr Guzhva, Chengqi Deng, Jeff Johnson, Gergely Szilvasy, Pierre-Emmanuel Mazar´e, Maria Lomeli, Lucas Hosseini, and Herv´e J´egou. The faiss library,

Show all 47 references
  1. [9]

    Visual simultaneous localization and mapping: a survey

    Jorge Fuentes-Pacheco, Jos ´e Ruiz-Ascencio, and Juan Manuel Rend ´on-Mancha. Visual simultaneous localization and mapping: a survey. Artificial intelligence review, 43:55–81, 2015. 1, 3

    2015

  2. [10]

    Dorian G ´alvez-L´opez and J. D. Tard´os. Bags of binary words for fast place recognition in image sequences. IEEE Trans- actions on Robotics, 28(5):1188–1197, 2012. 5

    2012

  3. [11]

    Cp-slam: Collaborative neural point-based slam system, 2023

    Jiarui Hu, Mao Mao, Hujun Bao, Guofeng Zhang, and Zhaopeng Cui. Cp-slam: Collaborative neural point-based slam system, 2023. 2, 3, 5, 6, 7, 8, 1

    2023

  4. [12]

    Cg-slam: Efficient dense rgb-d slam in a consistent uncertainty-aware 3d gaussian field, 2024

    Jiarui Hu, Xianhao Chen, Boyin Feng, Guanglin Li, Liangjing Yang, Hujun Bao, Guofeng Zhang, and Zhaopeng Cui. Cg-slam: Efficient dense rgb-d slam in a consistent uncertainty-aware 3d gaussian field, 2024. 2, 3

    2024

  5. [13]

    Photo-slam: Real-time simultaneous localization and photo- realistic mapping for monocular, stereo, and rgb-d cameras

    Huajian Huang, Longwei Li, Cheng Hui, and Sai-Kit Yeung. Photo-slam: Real-time simultaneous localization and photo- realistic mapping for monocular, stereo, and rgb-d cameras. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024. 2

    2024

  6. [14]

    A survey of state-of-the-art on visual slam

    Iman Abaspur Kazerouni, Luke Fitzgerald, Gerard Dooly, and Daniel Toal. A survey of state-of-the-art on visual slam. Expert Systems with Applications, 205:117734, 2022. 1, 3

    2022

  7. [15]

    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. arXiv preprint, 2023. 1, 2, 3, 4

    2023

  8. [16]

    3d gaussian splatting for real-time radiance field rendering

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

    2023

  9. [17]

    G2o: A general framework for graph optimization

    Rainer K ¨ummerle, Giorgio Grisetti, Hauke Strasdat, Kurt Konolige, and Wolfram Burgard. G2o: A general framework for graph optimization. In 2011 IEEE International Confer- ence on Robotics and Automation , pages 3607–3613, 2011. 5, 1

    2011

  10. [18]

    Swarm-slam: Sparse decentralized collaborative simultaneous localization and mapping framework for multi-robot systems

    Pierre-Yves Lajoie and Giovanni Beltrame. Swarm-slam: Sparse decentralized collaborative simultaneous localization and mapping framework for multi-robot systems. IEEE Robotics and Automation Letters , 9(1):475–482, 2024. 2, 3, 6, 7

    2024

  11. [19]

    Door-slam: Distributed, online, and outlier resilient slam for robotic teams

    Pierre-Yves Lajoie, Benjamin Ramtoula, Yun Chang, Luca Carlone, and Giovanni Beltrame. Door-slam: Distributed, online, and outlier resilient slam for robotic teams. IEEE Robotics and Automation Letters, 5(2):1656–1663, 2020. 3

    2020

  12. [20]

    Loopy-slam: Dense neural slam with loop closures

    Lorenzo Liso, Erik Sandstr ¨om, Vladimir Yugay, Luc Van Gool, and Martin R Oswald. Loopy-slam: Dense neural slam with loop closures. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 20363–20373, 2024. 2, 5

    2024

  13. [21]

    Eslam: Efficient dense slam system based on hy- brid representation of signed distance fields

    Mohammad Mahdi Johari, Camilla Carta, and Franc ¸ois Fleuret. Eslam: Efficient dense slam system based on hy- brid representation of signed distance fields. arXiv e-prints, pages arXiv–2211, 2022. 2

    2022

  14. [22]

    Newton: Neural view-centric mapping for on-the-fly large-scale slam

    Hidenobu Matsuki, Keisuke Tateno, Michael Niemeyer, and Federic Tombari. Newton: Neural view-centric mapping for on-the-fly large-scale slam. arXiv preprint arXiv:2303.13654, 2023. 3

    2023 arXiv

  15. [23]

    Hidenobu Matsuki, Riku Murai, Paul H. J. Kelly, and An- drew J. Davison. Gaussian Splatting SLAM. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, 2024. 1, 2, 3, 6, 7

    2024

  16. [24]

    Srinivasan, Matthew Tancik, Jonathan T

    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. In European Conference on Computer Vision (ECCV). CVF, 2020. 2

    2020

  17. [25]

    Instant neural graphics primitives with a mul- tiresolution hash encoding

    Thomas M ¨uller, Alex Evans, Christoph Schied, and Alexan- der Keller. Instant neural graphics primitives with a mul- tiresolution hash encoding. ACM Transactions on Graphics (ToG), 41(4):1–15, 2022. 2

    2022

  18. [26]

    Dinov2: Learning robust visual features with- out supervision, 2024

    Maxime Oquab, Timoth ´ee Darcet, Th ´eo Moutakanni, Huy V o, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, Mah- moud Assran, Nicolas Ballas, Wojciech Galuba, Russell Howes, Po-Yao Huang, Shang-Wen Li, Ishan Misra, Michae...

    2024

  19. [27]

    Aria digital twin: A new benchmark dataset for egocentric 3d machine percep- tion, 2023

    Xiaqing Pan, Nicholas Charron, Yongqian Yang, Scott Pe- ters, Thomas Whelan, Chen Kong, Omkar Parkhi, Richard Newcombe, and Carl Yuheng Ren. Aria digital twin: A new benchmark dataset for egocentric 3d machine percep- tion, 2023. 6

    2023

  20. [28]

    Colored point cloud registration revisited

    Jaesik Park, Qian-Yi Zhou, and Vladlen Koltun. Colored point cloud registration revisited. In 2017 IEEE Interna- tional Conference on Computer Vision (ICCV) , pages 143– 152, 2017. 4

    2017

  21. [29]

    Rtg-slam: Real-time 3d reconstruction at scale using gaussian splatting

    Zhexi Peng, Tianjia Shao, Liu Yong, Jingke Zhou, Yin Yang, Jingdong Wang, and Kun Zhou. Rtg-slam: Real-time 3d reconstruction at scale using gaussian splatting. In ACM SIGGRAPH Conference Proceedings, Denver, CO, United States, July 28 - August 1, 2024, 2024. 2

    2024

  22. [30]

    Leonard, and Luca Carlone

    Antoni Rosinol, John J. Leonard, and Luca Carlone. NeRF- SLAM: Real-Time Dense Monocular SLAM with Neural Radiance Fields. arXiv, 2022. 2

    2022

  23. [31]

    Fast point feature histograms (fpfh) for 3d registration

    Radu Bogdan Rusu, Nico Blodow, and Michael Beetz. Fast point feature histograms (fpfh) for 3d registration. In 2009 IEEE International Conference on Robotics and Automation, pages 3212–3217, 2009. 5

    2009

  24. [32]

    Point-slam: Dense neural point cloud-based slam

    Erik Sandstr ¨om, Yue Li, Luc Van Gool, and Martin R Os- wald. Point-slam: Dense neural point cloud-based slam. In International Conference on Computer Vision (ICCV) . IEEE/CVF, 2023. 2, 6

    2023

  25. [33]

    CCM-SLAM: Robust and efficient centralized collaborative monocular simultane- ous localization and mapping for robotic teams

    Patrik Schmuck and Margarita Chli. CCM-SLAM: Robust and efficient centralized collaborative monocular simultane- ous localization and mapping for robotic teams. In Journal of Field Robotics (JFR), 2018. 2, 3, 6, 7

    2018

  26. [34]

    Smith and Peter Cheeseman

    Randall C. Smith and Peter Cheeseman. On the representa- tion and estimation of spatial uncertainty. The International Journal of Robotics Research, 5(4):56–68, 1986. 2

    1986

  27. [35]

    A benchmark for the evalua- tion of RGB-D SLAM systems

    J ¨urgen Sturm, Nikolas Engelhard, Felix Endres, Wolfram Burgard, and Daniel Cremers. A benchmark for the evalua- tion of RGB-D SLAM systems. In International Conference on Intelligent Robots and Systems (IROS). IEEE/RSJ, 2012. 6

    2012

  28. [36]

    Davi- son

    Edgar Sucar, Shikun Liu, Joseph Ortiz, and Andrew J. Davi- son. iMAP: Implicit Mapping and Positioning in Real-Time. In International Conference on Computer Vision (ICCV) . IEEE/CVF, 2021. 1, 2

    2021

  29. [37]

    Oswald, and Matteo Poggi

    Fabio Tosi, Youmin Zhang, Ziren Gong, Erik Sandstr ¨om, Stefano Mattoccia, Martin R. Oswald, and Matteo Poggi. How nerfs and 3d gaussian splatting are reshaping slam: a survey, 2024. 1, 3, 6, 2

    2024

  30. [38]

    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 Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 13293–13302, 2023. 2

    2023

  31. [39]

    Image quality assessment: from error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Si- moncelli. Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing, 13(4):600–612, 2004. 6

    2004

  32. [40]

    Gs-slam: Dense visual slam with 3d gaussian splatting, 2024

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

    2024

  33. [41]

    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 IEEE International Symposium on Mixed and Augmented Reality (ISMAR), pages 499–507. IEEE, 2022. 2

    2022

  34. [42]

    Plenoxels: Radiance fields without neural networks, 2021

    Alex Yu, Sara Fridovich-Keil, Matthew Tancik, Qinhong Chen, Benjamin Recht, and Angjoo Kanazawa. Plenoxels: Radiance fields without neural networks, 2021. 2

    2021

  35. [43]

    Os- wald

    Vladimir Yugay, Yue Li, Theo Gevers, and Martin R. Os- wald. Gaussian-slam: Photo-realistic dense slam with gaus- sian splatting, 2024. 2, 3, 4, 5, 6, 7, 1

    2024

  36. [44]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In IEEE conference on com- puter vision and pattern recognition , pages 586–595, 2018. 6

    2018

  37. [45]

    Go-slam: Global optimization for consistent 3d instant reconstruction

    Youmin Zhang, Fabio Tosi, Stefano Mattoccia, and Mat- teo Poggi. Go-slam: Global optimization for consistent 3d instant reconstruction. arXiv preprint arXiv:2309.02436,

    arXiv

  38. [46]

    Loopsplat: Loop closure by registering 3d gaussian splats, 2024

    Liyuan Zhu, Yue Li, Erik Sandstr ¨om, Shengyu Huang, Kon- rad Schindler, and Iro Armeni. Loopsplat: Loop closure by registering 3d gaussian splats, 2024. 2

    2024

  39. [47]

    Nice-slam: Neural implicit scalable encoding for slam

    Zihan Zhu, Songyou Peng, Viktor Larsson, Weiwei Xu, Hu- jun Bao, Zhaopeng Cui, Martin R Oswald, and Marc Polle- feys. Nice-slam: Neural implicit scalable encoding for slam. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12786–12796, 2022. 2, 8 10 MAGi...

    2022

Pith tools

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