Pith. sign in

REVIEW 3 major objections 6 minor 55 references

MBA-SLAM: Motion Blur Aware Gaussian Splatting SLAM

T0 review · 3 major / 6 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 MBA-SLAM claims that dense RGB-D SLAM can track and map from severely blurred video by modeling each blurred frame as an average of sharp views along the exposure trajectory, jointly optimizing trajectory and scene.

desk verdict Solid system paper on blur-robust dense SLAM, but the flagship 'physical formation' claim is undercut by an inconsistency between the SE(3) interpolation and its translation/rotation decomposition. read the letter →

arxiv 2411.08279 v2 pith:HBJOP37P submitted 2024-11-13 cs.CV cs.RO

Peng Wang , Lingzhe Zhao , Yin Zhang , Shiyu Zhao , Peidong Liu This is my paper · ORCID
classification cs.CVcs.RO
keywords motionblurdenseRGB-DSLAM3DGaussianSplattingNeuralRadianceFieldsbundleadjustmentcameratrajectoryestimationimagedeblurringvisualodometry
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 sets out to establish that dense visual SLAM does not require sharp input frames: by inserting the physical motion blur formation process into both tracking and mapping, a stream of blurry RGB-D images can yield accurate camera trajectories and a sharp, high-fidelity 3D map. The method represents the camera motion during each exposure by two poses, one at the start and one at the end, with intermediate poses interpolated in $\mathrm{SE}(3)$; blurred images are synthesized by averaging the sharp views rendered along that path and are compared against the captured frames. Because the synthesis is differentiable, the trajectory and the scene representation (a tri-plane radiance field or a set of 3D Gaussians) are optimized jointly, which turns blur into a source of constraint rather than a failure mode. On synthetic and real blurry datasets the reported localization and reconstruction results surpass earlier NeRF- and Gaussian-based SLAM systems, and the same pipeline remains competitive on standard sharp datasets when the blur model is switched off.

What carries the argument

The load-bearing mechanism is differentiable re-blurring. For each sampled image patch the tracker transfers pixels to virtual poses interpolated along the trajectory, reads intensities from the sharp reference by bilinear interpolation, averages them as in Eq. (24), and minimizes the difference from the captured blurry intensities. The trajectory itself is compactly parameterized by two $\mathrm{SE}(3)$ poses, $T_{\mathrm{start}}$ and $T_{\mathrm{end}}$, connected by Lie-algebra interpolation (Eq. (5)); this small unknown set is what lets the tracker run at real-time speed and gives the mapper a well-posed bundle adjustment over the exposure intervals. The end-to-end differentiability of the averaging model lets gradients flow from the blur residual into both the scene representation and the per-frame trajectory, which is why the same formulation serves tracking, mapping, and deblurring.

What would settle it

Take a camera whose ground-truth motion inside each exposure contains an inflection point, for example a jerk-and-reverse trajectory with the same total displacement as a straight segment, and run the method on the resulting RGB-D sequence. If the per-frame trajectory estimates and deblurred renderings degrade substantially relative to a straight-line exposure of equal displacement, that is direct evidence that the linear-in-SE(3) model is the limiting factor; the expected signature is residual streak directions in the deblurred image that the average of linearly interpolated views cannot produce.

Watch

Extended reading notes

Core claim

On its own terms, the central discovery is that a blurred frame can be treated as a physically differentiable average of sharp views: $B(\mathbf{x}) \approx \frac{1}{n}\sum_{i=0}^{n-1} I_i(\mathbf{x})$, where each virtual view $I_i$ is rendered from a pose $T_t = T_{\mathrm{start}} \exp(\frac{t}{\tau}\log(T_{\mathrm{start}}^{-1}T_{\mathrm{end}}))$ inside the exposure. Equipped with this model, the tracker aligns a rendered sharp keyframe to the current blurry frame by re-blurring, and the mapper jointly optimizes keyframe trajectories and scene parameters by minimizing photometric and depth residuals on the synthesized blur. The paper further claims that the resulting system outperforms previous state-of-the-art dense visual SLAM methods on motion-blurred data while also holding its own on sharp benchmarks, and that either a tri-plane NeRF or a 3D Gaussian Splatting backend can carry the mapping role.

Load-bearing premise

The load-bearing premise is that during one exposure the camera follows a straight path in six-degree-of-freedom motion between its start and end poses, so every intermediate view lies on a linear interpolation; if the camera changes direction inside the exposure, the synthesized blur no longer matches the captured blur and the joint optimization cannot fully repair the mismatch.

Editorial extensions

If this is right

  • Dense SLAM systems could operate directly on video captured in low light or with long exposure, where current NeRF- and Gaussian-based pipelines typically lose tracking or produce corrupted maps.
  • The learned map is a sharp representation of the scene, so the same pipeline outputs deblurred renderings of the environment without a separate deblurring network.
  • Per-frame start and end poses provide temporal information about camera motion during each exposure, not just a single instantaneous pose.
  • Because the blur model is representation-agnostic, the approach can be ported to future differentiable scene representations by keeping the same re-blurring loss.
  • On sharp inputs, keeping the blur model off restores full speed while retaining competitive accuracy, so one system covers both sharp and blurry operation.

Reading between the lines

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

  • Editorial inference: the linear trajectory model is the natural ceiling of the method; a sequence with a sharp direction reversal inside one exposure should expose residual blur that the two-pose model cannot explain, and replacing the interpolation with a spline or fusing inertial readings is the obvious extension.
  • Editorial inference: the re-blurring loss does not fundamentally require measured depth once a coarse scene exists, so a monocular version that renders depth from the map and then re-blurs could inherit some of the same tolerance to blur.
  • Editorial inference: if blurred frames become constraints instead of outliers, motion blur handling in SLAM may shift from preprocessing the frames to estimating a per-frame trajectory, with implications for rolling-shutter and hand-held capture where blur is frequent.
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

3 major / 6 minor

Summary. MBA-SLAM proposes a dense RGB-D SLAM system for motion-blurred video. The key idea is to represent the camera motion during each frame's exposure by two poses, T_start and T_end, and to synthesize the observed blur by averaging virtual sharp views rendered at interpolated poses along the trajectory. The system has a CUDA-implemented blur-aware tracker that re-blurs a rendered keyframe and aligns it to the current blurry frame, and a mapper that jointly optimizes the scene representation (either tri-plane NeRF or 3DGS) and the intra-exposure trajectories. The paper evaluates on the synthetic ArchViz blur dataset, selected blurry ScanNet/TUM sequences, a newly captured RealSense+Vicon blur dataset, and sharp Replica/ScanNet/TUM benchmarks. The reported ATE, PSNR/SSIM/LPIPS, and mesh reconstruction results are strong on the blur datasets, and ablations show that the blur-aware mapper and the frame-to-frame tracker each improve performance. The method explicitly extends the authors' earlier MBA-VO, BAD-NeRF, and BAD-Gaussians work into a complete SLAM pipeline with a faster tri-plane NeRF backend.

Significance. If correct, the paper would show that dense RGB-D SLAM can operate directly on severely blurred video by jointly recovering the scene and the intra-exposure trajectory, avoiding the usual front-end deblurring or pose-failure cascade. The empirical support is substantial: on ArchViz the proposed GS version reaches average ATE 0.84 cm and PSNR 28.82 dB, far exceeding the compared baselines; the ablations isolate the contribution of the blur model and of the new tracker; and the authors commit to releasing code and introduce a real blur dataset with motion-capture ground truth. The central caveat is that the motion-model derivation contains an internal inconsistency (Eq. (7) is not the translation component of the SE(3) interpolation in Eq. (5)), so the 'physical image formation' claim needs to be repaired or softened. In addition, the abstract's claim to surpass all prior methods is not supported on the sharp Replica and TUM datasets.

major comments (3)
  1. [Sec. 3.2.2, Eqs. (5)–(7)] The motion model is not what it is claimed to be. Eq. (7) states t_t = t_start + (t/τ)(t_end − t_start), and the text says it is a decomposition of Eq. (5), but the translation component of exp((t/τ) log(T_start^{-1} T_end)) is only equal to that linear expression for pure translation. For a general relative transform (R_rel, t_rel), the exponential-map translation contains rotation–translation coupling terms (e.g., terms proportional to [ω]v via the left Jacobian of SO(3)); these are absent from Eq. (7). Because the tracker's re-blurring in Eq. (24) and the Jacobians in Eqs. (10)–(12) and (22)–(23) are built from Eq. (7), the implemented trajectory is a decoupled rotation-plus-linear-translation interpolation, not constant-velocity SE(3) motion. The abstract's phrase 'accurately modeling the physical image formation process' is therefore an overstatement. Please either implement the true SE(3) interpolation and update the Jacobians, or explicitly present Eq. (7) as an approximation and reword the physical-model claims.
  2. [Sec. 3.2.4, Eqs. (24) and (26)] Eq. (26) restates the virtual pose as the full SE(3) exponential T_start exp((i/(n−1))τ log(T_start^{-1} T_end)), while the Jacobian derivations in Sec. 3.2.2, particularly Eqs. (10)–(12) and (22)–(23), are valid only for the decoupled model of Eq. (7). These two models are not equivalent. The manuscript should specify unambiguously whether the tracker and the mapper use Eq. (26) or Eqs. (6)–(7); if the implementation uses the decoupled model, Eq. (26) and any downstream uses must be changed, and the effect of the mismatch on blur synthesis should be discussed.
  3. [Abstract, Sec. 4.3, Tables 4 and 7] The abstract and introduction claim that MBA-SLAM 'surpasses previous state-of-the-art methods in both camera localization and map reconstruction' without qualification. This is contradicted by the paper's own sharp-dataset results: in Table 4, RTG-SLAM has average ATE 0.18 cm on Replica, substantially lower than Ours-GS 0.35 cm and Ours-NeRF 0.41 cm; in Table 7, on TUM RGB-D, RTG-SLAM (1.06 cm), Photo-SLAM (1.28 cm), and MonoGS (1.47 cm) all outperform Ours-GS (1.98 cm). The text in Sec. 4.3 acknowledges 'except RTG-SLAM', but the abstract and contribution list do not. Please restrict the SOTA claim to the motion-blur datasets, or rephrase it as 'among NeRF-based and 3DGS-based methods' with the specific exceptions stated.
minor comments (6)
  1. [Sec. 4.1] The self-captured Realsense dataset is not described with exposure times, blur magnitudes, motion speeds, or synchronization details; please add these to support the dataset as a community resource.
  2. [Tables 8 and 9] The row 'Photo-SLAM [16]' in Table 8 uses the reference number of MonoGS, and Table 9's 'PointSLAM [12]' uses ESLAM's number; please correct the table citations.
  3. [Tables 2 and 3] Many baseline cells are marked with ✗ or ✖; the captions should state the exact failure criterion (e.g., tracking divergence threshold or code error) and whether these runs were excluded from the averages.
  4. [Tables 1 and 4] The text says all methods were run five times and averaged, but no standard deviations are reported; given the large run-to-run spread visible in some baselines (e.g., Point-SLAM on ArchViz), adding variance or per-run values would improve the comparison.
  5. [Sec. 4.5, Table 12] The rationale for n=13 is saturation, but ATE on ArchViz-3 keeps improving from 1.413 cm at n=13 to 1.223 cm at n=17; please state the trade-off criterion more explicitly.
  6. [Sec. 3.3.3] The depth loss D(x) uses the middle pose while the color loss B(x) integrates over the whole trajectory; one sentence explaining why depth is not blurred in the same way would avoid confusion.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular step found; the blur-aware tracker and mapper are derived from the physical image formation model and validated on external benchmarks, with only minor reliance on the authors' prior components.

full rationale

The claimed derivation chain is self-contained. Equations (1)-(2) define the motion blur image as the temporal average of virtual sharp images; Equation (5) parameterizes the intra-exposure camera pose by SE(3) interpolation between T_start and T_end; Equations (24)-(25) re-blur the reference keyframe and minimize photometric error against the captured blurred frame; Section 3.3 jointly optimizes T_start, T_end, and the scene representation with the losses in Equations (58)-(62). In each stage the optimization variable is the quantity being estimated (the trajectory), and the supervision is the captured image, so no fitted parameter is later relabeled as a prediction. Performance is measured on Replica, ScanNet, TUM RGB-D, and real Realsense sequences against independent published systems, so the headline results do not reduce to a constructed identity. The paper explicitly builds on MBA-VO, BAD-NeRF, and BAD-Gaussians, all peer-reviewed prior work; citing these components as building blocks is legitimate reuse, not circular justification. The only caveat is a modeling approximation: the paper writes Equation (7) as linear translation interpolation, which does not exactly equal the translation component of the SE(3) exponential in Equation (5) for combined rotation-plus-translation motion. This is a correctness and approximation risk that could bias estimated trajectories under such motion, but it is not a circular reduction of the output to the input. Hence the low score.

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

No new physical entities are introduced. The central claim rests on the physical blur model, the linear SE(3) trajectory model, the fronto-parallel plane transfer, and hand-set hyperparameters such as the number of virtual frames and loss weights. These are reasonable domain assumptions for short exposures, but they define the regime of validity: the method is designed for blur dominated by camera motion during exposure, not for defocus or rolling-shutter blur.

free parameters (3)
  • number of virtual frames n = 13 for Gaussian Splatting, 7 for NeRF
    Number of samples in the discrete blur integral in Eq. (2); chosen by ablation in Table 12 as a speed and quality trade-off, not derived from physics.
  • scale regularization ratio r = 1.0
    Hyperparameter in Eq. (61) that prevents Gaussians from becoming too thin; hand-set value reported in Sec. 4.1.
  • loss weights lambda_c, lambda_d, lambda_ssim, lambda_fs, lambda_sdf = not fully specified
    Hand-chosen balancing weights in Eq. (60) and Eq. (62); values are not tabulated, so the exact objective is not reproducible from the paper alone.
assumptions (5)
  • domain assumption Motion-blurred image equals the time-average of virtual sharp images during exposure (Eq. 1)
    Assumes blur is purely due to global camera motion during exposure, ignoring rolling shutter, defocus, and sensor noise. Invoked in Sec. 3.1 and used throughout the tracker and mapper.
  • domain assumption Camera trajectory during exposure is a straight line in SE(3) between Tstart and Tend (Eq. 5)
    Linear interpolation in the Lie algebra is used to generate virtual poses for re-blurring. The authors admit higher-order splines could capture more complex motion but state the linear model worked in experiments (Sec. 3.2.2).
  • domain assumption Re-blurring pixel transfer assumes 3D points lie on a fronto-parallel plane relative to the reference keyframe (Sec. 3.2.3, Fig. 2)
    The transfer equations in Eqs. (31)-(34) use a fronto-parallel plane approximation to map pixels between virtual views and the reference image, which only holds locally for planar patches.
  • domain assumption Measured depth corresponds to the middle pose of the RGB exposure trajectory (Eq. 59)
    The depth loss compares rendered depth at the middle pose with the measured depth Dgt, assuming the depth sensor reading aligns with the middle of the exposure, which is not physically modeled.
  • standard math Differentiable volume rendering and Gaussian rasterization from prior works are correct (Eqs. 53-57)
    The paper relies on the standard NeRF volume rendering equations and the 3DGS rasterization formulation from [7], [8], and [12] without re-deriving them.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MBA-SLAM: Motion Blur Aware Gaussian Splatting SLAM." pith.science (2026). https://pith.science/paper/HBJOP37P

@misc{pith2026241108279,
  author       = {Pith},
  title        = {Pith review of: MBA-SLAM: Motion Blur Aware Gaussian Splatting SLAM},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HBJOP37P}},
  note         = {Machine review of arXiv:2411.08279}
}
read the original abstract

Emerging 3D scene representations, such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS), have demonstrated their effectiveness in Simultaneous Localization and Mapping (SLAM) for photo-realistic rendering, particularly when using high-quality video sequences as input. However, existing methods struggle with motion-blurred frames, which are common in real-world scenarios like low-light or long-exposure conditions. This often results in a significant reduction in both camera localization accuracy and map reconstruction quality. To address this challenge, we propose a dense visual deblur SLAM pipeline (i.e. MBA-SLAM) to handle severe motion-blurred inputs and enhance image deblurring. Our approach integrates an efficient motion blur-aware tracker with either neural radiance fields or Gaussian Splatting based mapper. By accurately modeling the physical image formation process of motion-blurred images, our method simultaneously learns 3D scene representation and estimates the cameras' local trajectory during exposure time, enabling proactive compensation for motion blur caused by camera movement. In our experiments, we demonstrate that MBA-SLAM surpasses previous state-of-the-art methods in both camera localization and map reconstruction, showcasing superior performance across a range of datasets, including synthetic and real datasets featuring sharp images as well as those affected by motion blur, highlighting the versatility and robustness of our approach. Code is available at https://github.com/WU-CVGL/MBA-SLAM.

Figures

Figures reproduced from arXiv: 2411.08279 by the authors.

Figure 1
Figure 1. The pipeline of MBA-SLAM. Our framework consists of blur aware tracking process and bundle adjustment deblurring mapping process. Tracking: Given the current blurry frame, the mapper first renders a virtual sharp image of the lastest blurry keyframe from the 3D scene. Our motion blur-aware tracker directly estimates the camera motion trajectory during the exposure time, represented by the camera positions at the sta… view at source ↗
Figure 2
Figure 2. Pixel point transfer strategies. Note that we assume the pixel center lies at the grid intersection, e.g.the green grid is considered as a 3×3 patch. 3.2.3 Direct Image Alignment with Blurry Images Our motion blur-aware tracker operates by directly aligning the keyframe, assumed to be sharp, with the current frame, which may suffer from motion blur. To exploit photometric consistency during alignment, we must either… view at source ↗
Figure 3
Figure 3. Geometric relationship between x ∈ R2 (i.e.the red pixel) of the virtual sharp image Ii and x iτ n−1 ∈ R2 (i.e.the black pixel) of the reference image Iref. The right figure is its simplified 2D top-down view of the left figure. In practice, many direct image alignment methods use local patches to facilitate convergence. In contrast to direct image alignment algorithm for sharp images, which typi￾cally selects the l… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Estimated trajectories of MBA-SLAM from the motion blurred image sequences of the ArchViz dataset. It demonstrates that MBA￾SLAM can estimate accurate trajectories, although the camera motions are very challenging. 3.3.3 Loss Functions Given a batch of pixels R, the ph…
Figure 5
Figure 5. Figure 5: Qualitative rendering results of different methods with synthetic ArchViz datasets. It demonstrates that MBA-SLAM can restore and render sharp images from blurry input and outperform other dense visual SLAMs. Best viewed in high resolution. particularly in areas with r…
Figure 6
Figure 6. Figure 6: Qualitative mesh visualization of different methods with ArchViz-1 datasets. The result reveals that implicit Radiance Fields (e.g.CoSLAM, ESLAM) deliver better reconstruction mesh performance than explicit point based methods (i.e.Point-SLAM, SplaTAM). MBA-SLAM always…
Figure 7
Figure 7. Figure 7: Qualitative rendering results of different methods with the real public ScanNet and TUM RGB-D datasets. The experimental results demonstrate that our method achieves superior performance over prior methods on the real public dataset. end mapper. The tracking results fo…
Figure 8
Figure 8. Figure 8: Qualitative rendering results of different methods with our real captured Realsense datasets. The experimental results demonstrate that our method achieves superior performance over prior methods on the real captured dataset as well. Note that the first column shows an…
Figure 9
Figure 9. Figure 9: Qualitative mesh visualization results of different methods with Replica datasets. It demonstrates MBA-SLAM surpasses other state￾of-the-art dense visual SLAMs even on standard sharp datasets. Note that RTG-SLAM produces incomplete meshes due to the presence of numerou…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 50 canonical work pages

  1. [1]

    ORB-SLAM2: An open-source slam system for monocular, stereo, and RGB-D cameras,

    R. Mur-Artal and J. D. Tard ´os, “ORB-SLAM2: An open-source slam system for monocular, stereo, and RGB-D cameras,” IEEE transactions on robotics , vol. 33, no. 5, pp. 1255–1262, 2017. 1, 2, 3

    work page 2017

  2. [2]

    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, 2017. 1, 3

    work page 2017

  3. [3]

    CodeSLAM—learning a compact, optimisable represen- tation for dense visual SLAM,

    M. Bloesch, J. Czarnowski, R. Clark, S. Leutenegger, and A. J. Davison, “CodeSLAM—learning a compact, optimisable represen- tation for dense visual SLAM,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 2560–2568. 1

    work page 2018

  4. [4]

    Deep- Factors: Real-time probabilistic dense monocular SLAM,

    J. Czarnowski, T. Laidlow, R. Clark, and A. J. Davison, “Deep- Factors: Real-time probabilistic dense monocular SLAM,” IEEE Robotics and Automation Letters , vol. 5, no. 2, pp. 721–728, 2020. 1

    work page 2020

  5. [5]

    SceneCode: Monocular dense semantic reconstruction using learned encoded scene representations,

    S. Zhi, M. Bloesch, S. Leutenegger, and A. J. Davison, “SceneCode: Monocular dense semantic reconstruction using learned encoded scene representations,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2019, pp. 11 776–11 785. 1

    work page 2019

  6. [6]

    Seman- ticFusion: Dense 3d semantic mapping with convolutional neural networks,

    J. McCormac, A. Handa, A. Davison, and S. Leutenegger, “Seman- ticFusion: Dense 3d semantic mapping with convolutional neural networks,” in 2017 IEEE International Conference on Robotics and automation (ICRA). IEEE, 2017, pp. 4628–4635. 1

    work page 2017

  7. [7]

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

    B. Mildenhall, P . P . Srinivasan, M. Tancik, J. T. Barron, R. Ra- mamoorthi, and R. Ng, “NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis,” in ECCV, 2020. 1, 2, 3, 7

    work page 2020

  8. [8]

    3D Gaus- sian Splatting for Real-Time Radiance Field Rendering,

    B. Kerbl, G. Kopanas, T. Leimk ¨uhler, and G. Drettakis, “3D Gaus- sian Splatting for Real-Time Radiance Field Rendering,” ACM T ransactions on Graphics, vol. 42, no. 4, July 2023. [Online]. Avail- able: https://repo-sam.inria.fr/fungraph/3d-gaussian-splatting/ 1, 2, 3, 7, 8

    work page 2023

Show all 55 references
  1. [9]

    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. 1, 2, 8, 10

    2021

  2. [10]

    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 Conference on Computer Vision and Pattern Recognition, 2022, pp. 12 786–12 796. 1, 2, 8, 9, 10, 14, 15, 16

    2022

  3. [11]

    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. 1, 2, 8, 9, 10, 11, 12, 13, 14, 15, 16

    2023

  4. [12]

    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. 1, 2, 7, 8, 9, 10, 11, 12, 13,...

    2023

  5. [13]

    NeRF-SLAM: Real- time dense monocular SLAM with neural radiance fields,

    A. Rosinol, J. J. Leonard, and L. Carlone, “NeRF-SLAM: Real- time dense monocular SLAM with neural radiance fields,” in 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2023, pp. 3437–3444. 1, 2

    2023

  6. [14]

    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 CVPR,

  7. [15]

    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,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024. 1, 2, 8, 9, 10, 11, 12, 13,...

    2024

  8. [16]

    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. 1, 2, 8, 9, 10, 11, 12, 13, 14, 15, 16

    2024

  9. [17]

    Photo-SLAM: Real-time simultaneous localization and photorealistic mapping for monoc- ular, stereo, and rgb-d cameras,

    H. Huang, L. Li, C. Hui, and S.-K. Yeung, “Photo-SLAM: Real-time simultaneous localization and photorealistic mapping for monoc- ular, stereo, and rgb-d cameras,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024. 1, 2, 8, 9, 10, 11, 1...

    2024

  10. [18]

    MBA-VO: Motion Blur Aware Visual Odometry,

    P . Liu, X. Zuo, V . Larsson, and M. Pollefeys, “MBA-VO: Motion Blur Aware Visual Odometry,” in ICCV, 2021, pp. 5550–5559. 2, 3, 8, 9

    2021

  11. [19]

    The Replica dataset: A digital replica of indoor spaces,

    J. Straub, T. Whelan, L. Ma, Y. Chen, E. Wijmans, S. Green, J. J. Engel, R. Mur-Artal, C. Ren, S. Verma, A. Clarkson, M. Yan, B. Budge, Y. Yan, X. Pan, J. Yon, Y. Zou, K. Leon, N. Carter, J. Briales, T. Gillingham, E. Mueggler, L. Pesqueira, M. Savva, D. Batra, H. M. Strasdat,...

    1906 arXiv

  12. [20]

    ScanNet: Richly-annotated 3d reconstructions of indoor scenes,

    A. Dai, A. X. Chang, M. Savva, M. Halber, T. Funkhouser, and M. Nießner, “ScanNet: Richly-annotated 3d reconstructions of indoor scenes,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2017, pp. 5828–5839. 2, 8, 9, 10

    2017

  13. [21]

    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 2012 IEEE/RSJ international conference on intelligent robots and systems . IEEE, 2012, pp. 573–580. 2, 8, 9, 10

    2012

  14. [22]

    BAD-NeRF: Bundle Adjusted Deblur Neural Radiance Fields,

    P . Wang, L. Zhao, R. Ma, and P . Liu, “BAD-NeRF: Bundle Adjusted Deblur Neural Radiance Fields,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , June 2023, pp. 4170–4179. 2, 9

    2023

  15. [23]

    BAD-Gaussians: Bundle adjusted deblur gaussian splatting,

    L. Zhao, P . Wang, and P . Liu, “BAD-Gaussians: Bundle adjusted deblur gaussian splatting,” arXiv preprint arXiv:2403.11831, 2024. 2

    2024 arXiv

  16. [24]

    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 Proceed- ings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 3727–3737. 2

    2023

  17. [25]

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

    C.-M. Chung, Y.-C. Tseng, Y.-C. Hsu, X.-Q. Shi, Y.-H. Hua, J.-F. Yeh, W.-C. Chen, Y.-T. Chen, and W. H. Hsu, “Orbeez-SLAM: A real- time monocular visual slam with orb features and nerf-realized mapping,” in 2023 IEEE International Conference on Robotics and Automation (ICRA). ...

    2023

  18. [26]

    Point- SLAM: Dense neural point cloud-based SLAM,

    E. Sandstr ¨om, 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. 2, 8, 9, 10, 11, 12, 13, 14, 15, 16

    2023

  19. [27]

    Droid-SLAM: Deep visual slam for monoc- ular, stereo, and RGB-D cameras,

    Z. Teed and J. Deng, “Droid-SLAM: Deep visual slam for monoc- ular, stereo, and RGB-D cameras,” Advances in neural information processing systems, vol. 34, pp. 16 558–16 569, 2021. 2

    2021

  20. [28]

    Deblur-NeRF: Neural Radiance Fields from Blurry Images,

    L. Ma, X. Li, J. Liao, Q. Zhang, X. Wang, J. Wang, and P . V . Sander, “Deblur-NeRF: Neural Radiance Fields from Blurry Images,” in CVPR, 2022, pp. 12 861–12 870. 2

    2022

  21. [29]

    DP-NeRF: Deblurred neural radiance field with physical scene priors,

    D. Lee, M. Lee, C. Shin, and S. Lee, “DP-NeRF: Deblurred neural radiance field with physical scene priors,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2023, pp. 12 386–12 396. 2

    2023

  22. [30]

    Deblurring 3d gaussian splatting,

    B. Lee, H. Lee, X. Sun, U. Ali, and E. Park, “Deblurring 3d gaussian splatting,” arXiv preprint arXiv:2401.00834 , 2024. 2

    2024 arXiv

  23. [31]

    BAGS: Blur agnostic gaussian splatting through multi-scale ker- nel modeling,

    C. Peng, Y. Tang, Y. Zhou, N. Wang, X. Liu, D. Li, and R. Chellappa, “BAGS: Blur agnostic gaussian splatting through multi-scale ker- nel modeling,” arXiv preprint arXiv:2403.04926 , 2024. 2

    2024 arXiv

  24. [32]

    ExBluRF: Efficient radiance fields for extreme motion blurred images,

    D. Lee, J. Oh, J. Rim, S. Cho, and K. M. Lee, “ExBluRF: Efficient radiance fields for extreme motion blurred images,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 17 639–17 648. 2

    2023

  25. [33]

    Structure-from-Motion Revis- ited,

    J. L. Schonberger and J.-M. Frahm, “Structure-from-Motion Revis- ited,” in CVPR, 2016, pp. 4104–4113. 2

    2016

  26. [34]

    LSD-SLAM: Large-scale JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 18 direct monocular SLAM,

    J. Engel, T. Sch ¨ops, and D. Cremers, “LSD-SLAM: Large-scale JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 18 direct monocular SLAM,” in European conference on computer vision. Springer, 2014, pp. 834–849. 3

    2021

  27. [35]

    LDSO: Direct sparse odometry with loop closure,

    X. Gao, R. Wang, N. Demmel, and D. Cremers, “LDSO: Direct sparse odometry with loop closure,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2018, pp. 2198–2204. 3

    2018

  28. [36]

    Towards robust visual odometry with a multi-camera system,

    P . Liu, M. Geppert, L. Heng, T. Sattler, A. Geiger, and M. Pollefeys, “Towards robust visual odometry with a multi-camera system,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2018, pp. 1154–1161. 3

    2018

  29. [37]

    Direct visual odometry for a fisheye-stereo camera,

    P . Liu, L. Heng, T. Sattler, A. Geiger, and M. Pollefeys, “Direct visual odometry for a fisheye-stereo camera,” in 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2017, pp. 1746–1752. 3

    2017

  30. [38]

    Rolling-shutter modelling for direct visual-inertial odometry,

    D. Schubert, N. Demmel, L. Von Stumberg, V . Usenko, and D. Cremers, “Rolling-shutter modelling for direct visual-inertial odometry,” in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2019, pp. 2462–2469. 3

    2019

  31. [39]

    Real-time simultaneous localisation and mapping with a single camera,

    Davison, “Real-time simultaneous localisation and mapping with a single camera,” in Proceedings Ninth IEEE International Conference on Computer Vision. IEEE, 2003, pp. 1403–1410. 3

    2003

  32. [40]

    Visual odometry,

    D. Nist ´er, O. Naroditsky, and J. Bergen, “Visual odometry,” inPro- ceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004. , vol. 1. Ieee, 2004, pp. I–I. 3

    2004

  33. [41]

    DeMoN: Depth and Motion Network for Learning Monocular Stereo,

    B. Ummenhofer, H. Zhou, J. Uhrig, N. Mayer, E. Ilg, A. Doso- vitskiy, and T. Brox, “DeMoN: Depth and Motion Network for Learning Monocular Stereo,” in CVPR, 2017. 3

    2017

  34. [42]

    DeepTAM: Deep tracking and mapping,

    H. Zhou, B. Ummenhofer, and T. Brox, “DeepTAM: Deep tracking and mapping,” in ECCV, 2018. 3

    2018

  35. [43]

    Unsupervised learning of depth and ego-motion from video,

    T. Zhou, M. Brown, N. Snavely, and D. G. Lowe, “Unsupervised learning of depth and ego-motion from video,” in CVPR, 2017. 3

    2017

  36. [44]

    Simultaneous localization, mapping and deblurring,

    H. S. Lee, J. Kwon, and K. M. Lee, “Simultaneous localization, mapping and deblurring,” in ICCV, 2011. 3

    2011

  37. [45]

    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 T ransactions on Robotics, 2016. 3

    2016

  38. [46]

    Efficient geometry-aware 3d generative adversarial networks,

    E. R. Chan, C. Z. Lin, M. A. Chan, K. Nagano, B. Pan, S. De Mello, O. Gallo, L. J. Guibas, J. Tremblay, S. Khamis et al. , “Efficient geometry-aware 3d generative adversarial networks,” in Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, 202...

    2022

  39. [47]

    Optical models for direct volume rendering,

    N. Max, “Optical models for direct volume rendering,” IEEE T ransactions on Visualization and Computer Graphics , vol. 1, no. 2, pp. 99–108, 1995. 7

    1995

  40. [48]

    StyleSDF: High-resolution 3d- consistent image and geometry generation,

    R. Or-El, X. Luo, M. Shan, E. Shechtman, J. J. Park, and I. Kemelmacher-Shlizerman, “StyleSDF: High-resolution 3d- consistent image and geometry generation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 13 503–13 513. 7

    2022

  41. [49]

    PhysGaussian: Physics-integrated 3d gaussians for generative dynamics,

    T. Xie, Z. Zong, Y. Qiu, X. Li, Y. Feng, Y. Yang, and C. Jiang, “PhysGaussian: Physics-integrated 3d gaussians for generative dynamics,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 4389–4398. 8

    2024

  42. [50]

    Vox- Fusion: Dense tracking and mapping with voxel-based neural implicit representation,

    X. Yang, H. Li, H. Zhai, Y. Ming, Y. Liu, and G. Zhang, “Vox- Fusion: Dense tracking and mapping with voxel-based neural implicit representation,” in 2022 IEEE International Symposium on Mixed and Augmented Reality (ISMAR) . IEEE, 2022, pp. 499–507. 8, 9, 10, 11, 15

    2022

  43. [51]

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

    Z. Peng, T. Shao, L. Yong, J. Zhou, Y. Yang, J. Wang, and K. 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. 8, 9, 10, 12, 13, 14, 15, 16

    2024

  44. [52]

    Marching cubes: A high reso- lution 3d surface construction algorithm,

    W. E. Lorensen and H. E. Cline, “Marching cubes: A high reso- lution 3d surface construction algorithm,” ACM siggraph computer graphics, vol. 21, no. 4, pp. 163–169, 1987. 9

    1987

  45. [53]

    The Unreasonable Effectiveness of Deep Features as a Perceptual Metric,

    R. Zhang, P . Isola, A. A. Efros, E. Shechtman, and O. Wang, “The Unreasonable Effectiveness of Deep Features as a Perceptual Metric,” in CVPR, 2018, pp. 586–595. 9

    2018

  46. [55]

    A volumetric method for building complex models from range images,

    B. Curless and M. Levoy, “A volumetric method for building complex models from range images,” in Proceedings of the 23rd annual conference on Computer graphics and interactive techniques , 1996, pp. 303–312. 10

    1996

  47. [2024]

    Available: https://arxiv.org/abs/2409.06765 9

    [Online]. Available: https://arxiv.org/abs/2409.06765 9

    arXiv

Pith tools

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