Pith. sign in

REVIEW 4 major objections 5 minor 44 references

A transient, local blur of rendered keyframes makes 3D Gaussian Splatting SLAM track more accurately, without touching the underlying map.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 19:33 UTC pith:DE7EQAKG

load-bearing objection Plausible idea, uncontrolled evidence: CB-KNN smoothing could help 3DGS-SLAM tracking, but no ablation isolates it, and Eq. (10) contradicts its own adaptive story. the 4 major comments →

arxiv 2511.23221 v3 pith:DE7EQAKG submitted 2025-11-28 cs.CV

Robust 3DGS-based SLAM via Adaptive Kernel Smoothing

Shouhe Zhang , Dayong Ren , Wen Jie Li , Piaopiao Yu , Sensen Song , Kaikai Shao , Yurong Qian This is my paper
classification cs.CV
keywords 3D Gaussian SplattingSLAMcamera pose trackingadaptive kernel smoothingK-nearest neighborrendering robustnessRGB-D SLAMkeyframe optimization
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper challenges the assumption that higher rendering fidelity automatically yields better pose tracking in 3DGS-based SLAM. It argues that making the rasterization process robust to Gaussian parameter errors is more critical than perfect scene representation, and that a controlled, local smoothing of rendered keyframes can stabilize pose optimization. The proposed CB-KNN method adaptively shifts the 2D projections of the K-nearest Gaussians toward their centroid and blends their colors, applied only during keyframe rendering. On Replica, TUM-RGBD, and ScanNet, this reduces trajectory error while preserving or slightly improving rendering quality, supporting the claim that temporary rendering regularization is a practical alternative to redesigning the rasterizer.

Core claim

The paper establishes that a 'Corrective Blurry KNN' (CB-KNN) pre-rendering step—which, for each pixel, nudges the K nearest Gaussians toward their common centroid in the image plane and replaces their colors with a contribution-weighted average—acts as a regularizer that dampens parameter noise from outlier Gaussians. This yields smoother depth/color/contour images for pose optimization, improving ATE RMSE from 0.39 to 0.32 cm on Replica, 3.31 to 3.18 cm on TUM-RGBD, and 13.43 to 12.48 cm on ScanNet, with comparable or better rendering metrics. The corrections are transient and leave the map parameters unchanged.

What carries the argument

CB-KNN (Corrective Blurry K-Nearest Neighbors): a per-pixel correction applied at render time. Select the K Gaussians with highest contribution, compute the centroid of their 2D projections, shift each projection by a fraction α of its distance to the centroid (Eq. 6), and set each color to the normalized attenuation-weighted average of the neighbors (Eq. 7). This produces a smoothed temporary Gaussian set used only for that frame's rendering, preserving differentiability for pose and map optimization while not altering the stored map.

Load-bearing premise

The corrected Gaussian position is specified only in terms of its 2D projection, so the paper never defines how that offset is turned back into a 3D location for the attenuation function, nor how gradients flow through that corrected projection during pose and map optimization. If that step is not implemented with a valid differentiable 3D correction, the reported improvements could stem from the implementation choice rather than from the smoothing principle.

What would settle it

Render a keyframe with CB-KNN enabled but detach the gradients of the corrected positions and colors (stop-gradient), so the optimizer sees the smoothed image but receives no gradient from the correction. If pose tracking accuracy remains the same, the smoothing effect alone—not its differentiability—is what helps; if accuracy degrades, the gradient through the correction is load-bearing. Alternatively, compute the Jacobian of the rendered color with respect to the original Gaussian center through Eq. (6): if that Jacobian is inconsistent with the actual 3D position used in the attenuation fun

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • If the central claim is right, 3DGS-SLAM systems can be made more robust without modifying the map or the rasterizer core, just by inserting a lightweight correction at keyframe rendering.
  • The separation of mapping quality and tracking quality suggests that future SLAM systems may explicitly optimize for trackability, e.g., rendering with regularization rather than only for photometric accuracy.
  • CB-KNN's adaptive K selection based on local density and motion amplitude indicates the smoothing strength can be tuned per-frame, potentially generalizing to other 3DGS applications where robustness matters.
  • Improved LPIPS on Replica and PSNR on TUM suggest that smoothing during keyframe optimization can also improve the learned map's rendering quality, implying a beneficial side effect.
  • Because the correction is transient and keyframe-only, the added computational cost is small; the paper reports higher FPS and lower per-frame time than SplaTAM on Replica/R0.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The paper leaves unspecified how the corrected 2D projection is lifted back to a 3D position for the attenuation function; a careful reader should check whether the implementation uses an approximate depth or a projective lifting, and whether the gradient path is consistent with the stated formulas.
  • A natural extension is to test CB-KNN in stereo or monocular 3DGS-SLAM, where depth is estimated and parameter noise is higher; the smoothing may yield larger gains there.
  • The robustness mechanism suggests an analogy to robust estimation: CB-KNN acts as a nonparametric outlier filter on Gaussian parameters; one could try other robust statistics (median, trimmed mean) in the same slot.
  • The adaptive K formula depends on two hand-set constants (K0, β); ablating them could reveal whether the adaptive selection is the active ingredient or simply the fixed blur level.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper challenges the assumption that rendering fidelity is the main driver of tracking accuracy in 3DGS-based SLAM. It proposes Corrective Blurry KNN (CB-KNN), a temporary correction applied during keyframe rendering that shifts the 2D projections of the K-nearest Gaussians toward their centroid and replaces their colors with a contribution-weighted average. The authors claim this controlled blur/densification makes pose optimization more robust to Gaussian parameter noise without modifying the underlying map. They report ATE RMSE reductions versus SplaTAM on Replica (0.39→0.32 cm), TUM-RGBD (3.31→3.18 cm), and ScanNet (13.43→12.48 cm), comparable or slightly improved rendering metrics, and lower runtime on Replica/R0. The paper includes an adaptive K formula depending on local Gaussian density and inter-frame motion.

Significance. If the central claim is upheld, the paper offers a simple, modular mechanism that could be plugged into existing 3DGS-SLAM systems to improve tracking robustness at negligible mapping cost. The empirical direction is consistent across three benchmarks, which is encouraging. The paper also provides constructive specificity by naming its hyperparameters (K0, β, α, color weight) and by restricting the correction to keyframes, which makes the proposal concrete and testable. However, the current evidence does not isolate the proposed mechanism: the comparison against SplaTAM involves several simultaneous changes, key geometric/differentiability details of the correction are undefined, and no error bars or ablations are provided. The significance of the contribution is therefore real but conditional on resolving these load-bearing issues.

major comments (4)
  1. [§III-B, Eq. (6)] Equation (6) defines the corrected position only through its 2D projection: π(µ'_pk)=π(µ_pk)+α(...). No 3D vector µ'_pk is ever specified, yet Eqs. (2)–(4) use f_pk(p) for the "corrected Gaussian" and the text says f_pk uses the corrected 2D projection position as input. Since Eq. (1) defines the Gaussian decay in 3D, and the rasterizer in 3DGS requires a 3D mean (or a well-defined 2D Gaussian with covariance), the rendering operation is underspecified. More critically, the paper claims pose and map optimization remain end-to-end differentiable through the corrected rendering. No Jacobian of the correction with respect to µ_pk or E_t is given. Please define the 3D update (e.g., depth-based unprojection of the 2D offset), the covariance used after correction, and the gradient derivation. Without this, the reported ATE gains may be artifacts of an unspecified implementation.
  2. [§III-C, Eq. (10)] The adaptive-K description and formula are contradictory. The text states that K should increase when Gaussian density is low or scene complexity is high ("increasing the K value enhances the smoothing effect when Gaussian density is low or scene complexity is high"); however K=K0·max(0.5, 1−βγ/(ρ+ε)) decreases as γ/(ρ+ε) grows, i.e., in exactly those regimes. Either the formula or the stated intuition is wrong, and this matters because the claimed adaptivity is a named contribution. Please align the two and report which behavior was actually implemented in the experiments.
  3. [§IV, Table I; §III-C] The central claim is that CB-KNN smoothing is responsible for the ATE reductions, but the comparison to SplaTAM is uncontrolled. The pipeline also changes keyframe selection (every nth frame vs. SplaTAM's overlap-based), the loss (color weight 0.6 in Eq. (8), tuned by repeated experiments), keyframe count in map updates, and K0 per dataset. No ablation disables CB-KNN (e.g., α=0 or no position/color correction) while keeping the other modifications. Without that ablation, the improvement cannot be attributed to the smoothing step; it could come from any of the other differences. This experiment is essential for the paper's hypothesis.
  4. [§IV-A, Tables I–III] Experimental support is statistically thin. All metrics are single runs without error bars, and several hyperparameters (K0=5 vs. 8, color weight 0.6, β=0.3) are tuned on the same benchmark suites. The reported gains are small (e.g., Replica ATE 0.39→0.32 cm, TUM 3.31→3.18 cm), so without multiple runs or a validation protocol one cannot exclude that the differences are due to noise or to tuning. Please report mean±std over at least 3 runs and separate validation/test selection of hyperparameters.
minor comments (5)
  1. [§II-B] Duplicate phrase in "utilizing rapidly rendered rapidly rendered color maps and depth maps."
  2. [§IV-B] The acronym "CK-KNN" is used in the Relative Pose Error Analysis section; it should be CB-KNN.
  3. [Eq. (8)] The loss expression appears to be missing a plus sign and a parenthesis: it should read L1(D(p)−D_GT(p)) + 0.6·L1(C(p)−C_GT(p)), not L1(D(p)−D_GT(p) + 0.6·L1(C(p)−C_GT(p))).
  4. [§III-C and §IV-A] K0 is inconsistently specified: §III-C says the baseline value K0=8, while §IV-A states K0=5 for Replica and K0=8 for TUM-RGBD/ScanNet. Please clarify which value is used and why.
  5. [References and captions] Reference [15] and [25] are the same paper (Woodford & Rosten); Fig. 3's caption cites SplaTAM as [9] instead of [10]; some table entries are typeset inconsistently (e.g., lowercase scene names in Table III).

Circularity Check

0 steps flagged

No significant circularity: CB-KNN is an empirically evaluated heuristic and its equations do not reduce to fitted outputs or self-citations.

full rationale

The paper's derivation chain is self-contained in the sense required by the circularity pass. Equations (1)-(10) define the proposed rendering corrections (position offset, color averaging, adaptive K) from Gaussian parameters and camera inputs; none of these quantities is defined in terms of the evaluation metrics (ATE RMSE, RPE, PSNR) that are reported in Tables I-IV. The comparison to SplaTAM/MonoGS is an external benchmark evaluation, not a prediction manufactured from fitted values. Hyperparameters such as K0, alpha in Eq. (6), beta in Eq. (10), and the 0.6 color weight in Eq. (8) are stated as tuned choices ('The color weight of 0.6 is determined as the optimal value through repeated experiments'), which is ordinary hyperparameter tuning rather than fitting a parameter to the target result. The method borrows standard components (radial Gaussian decay, SplaTAM densification and contour rendering) with explicit external citations, and the many Ren et al. self-citations in the bibliography support background claims about point-cloud/segmentation processing and are not load-bearing for the CB-KNN mechanism or the tracking result. The formula-text mismatch around Eq. (10) (text says K should increase under low density/high complexity, while the formula decreases K in exactly those regimes) is an internal consistency defect and a legitimate reproducibility concern, but it is not circularity because it does not make any result true by construction. The absence of an ablation that disables CB-KNN while keeping the other pipeline changes is an experimental attribution gap, not a circular-derivation gap. Therefore no circular step meeting the required quote-and-reduction standard was found.

Axiom & Free-Parameter Ledger

5 free parameters · 4 axioms · 0 invented entities

The central results are empirical. The method rests on a stated but unproven hypothesis that temporary KNN smoothing regularizes pose optimization, and several coefficients are hand-set or tuned on the same benchmarks. No new physical or algorithmic entities beyond the CB-KNN correction are introduced. The unspecified 2D-to-3D lift is the largest hidden assumption in the method.

free parameters (5)
  • K0 = 5 on Replica, 8 on TUM-RGBD and ScanNet
    Baseline neighborhood size in Eq. (10); set per dataset without an ablation or held-out tuning protocol.
  • beta = 0.3
    Adjustment coefficient in Eq. (10); hand-selected, no sensitivity analysis.
  • alpha = range [0.1, 0.3]
    Position offset coefficient in Eq. (6); the paper does not report the value used per dataset or per scene.
  • color weight in loss = 0.6
    Weight on color L1 term in Eq. (8); the paper states it was 'determined as the optimal value through repeated experiments.'
  • keyframe interval n = unspecified
    Keyframes are selected as 'every nth frame' but n is never reported, affecting which frames receive smoothing.
axioms (4)
  • domain assumption 3DGS rendering model: pixel color/depth are obtained by alpha-compositing radially decaying Gaussians (Eqs. 1-4).
    Borrowed from Kerbl et al. and SplaTAM; no verification or derivation in this paper.
  • ad hoc to paper Smoothing positions and colors of K-nearest Gaussians toward their centroid/weighted average reduces pose-tracking error caused by Gaussian parameter noise.
    Core hypothesis of the paper; asserted in the introduction and tested only indirectly through final metrics, with no direct evidence that the smoothing suppresses noise rather than biasing the signal.
  • ad hoc to paper The adaptive K formula in Eq. (10), with K0, beta, density rho, and motion gamma, improves tracking in the assumed direction.
    No derivation or ablation supports the specific functional form or the chosen constants.
  • ad hoc to paper Corrected rendering remains differentiable with respect to original Gaussian parameters and camera pose.
    The paper claims gradients are returned to the original map, but Eq. (6) defines a corrected position only via its 2D projection; the inverse lift to 3D and the associated Jacobian are not specified.

pith-pipeline@v1.3.0-alltime-deepseek · 13576 in / 11374 out tokens · 107852 ms · 2026-08-03T19:33:22.379842+00:00 · methodology

0 comments
read the original abstract

In this paper, we challenge the conventional notion in 3DGS-SLAM that rendering quality is the primary determinant of tracking accuracy. We argue that, compared to solely pursuing a perfect scene representation, it is more critical to enhance the robustness of the rasterization process against parameter errors to ensure stable camera pose tracking. To address this challenge, we propose a novel approach that leverages a smooth kernel strategy to enhance the robustness of 3DGS-based SLAM. Unlike conventional methods that focus solely on minimizing rendering error, our core insight is to make the rasterization process more resilient to imperfections in the 3DGS parameters. We hypothesize that by allowing each Gaussian to influence a smoother, wider distribution of pixels during rendering, we can mitigate the detrimental effects of parameter noise from outlier Gaussians. This approach intentionally introduces a controlled blur to the rendered image, which acts as a regularization term, stabilizing the subsequent pose optimization. While a complete redesign of the rasterization pipeline is an ideal solution, we propose a practical and effective alternative that is readily integrated into existing 3DGS frameworks. Our method, termed Corrective Blurry KNN (CB-KNN), adaptively modifies the RGB values and locations of the K-nearest neighboring Gaussians within a local region. This dynamic adjustment generates a smoother local rendering, reducing the impact of erroneous GS parameters on the overall image. Experimental results demonstrate that our approach, while maintaining the overall quality of the scene reconstruction (mapping), significantly improves the robustness and accuracy of camera pose tracking.

Figures

Figures reproduced from arXiv: 2511.23221 by Dayong Ren, Kaikai Shao, Piaopiao Yu, Sensen Song, Shouhe Zhang, Wen Jie Li, Yurong Qian.

Figure 1
Figure 1. Figure 1: In the fr1/desk scene of the TUM-RGBD[9] dataset, [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The flowchart illustrates the workflow of a 3DGS - SLAM method based on CB - KNN. It starts with initial Gaussian [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Rendering results on the Room0 from Replica[16]. It can be observed from the rendering results that our CB-KNN [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Trajectory maps on Replica[16], TUM-RGBD[9], and ScanNet[45]. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

44 extracted references · 4 linked inside Pith

  1. [1]

    3d gaussian splatting for real-time radiance field render- ing.,

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis, “3d gaussian splatting for real-time radiance field render- ing.,”ACM Trans. Graph., vol. 42, no. 4, pp. 139–1, 2023

    2023

  2. [2]

    Gaussian splatting slam,

    Hidenobu Matsuki, Riku Murai, Paul HJ Kelly, and Andrew J Davi- son, “Gaussian splatting slam,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 18039–18048

    2024

  3. [3]

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

    Chi Yan, Delin Qu, Dan Xu, Bin Zhao, Zhigang Wang, Dong Wang, and Xuelong Li, “Gs-slam: Dense visual slam with 3d gaussian splatting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 19595–19604

    2024

  4. [4]

    Nerf: Neural radiance field in 3d vision, a comprehensive review,

    Kyle Gao, Yina Gao, Hongjie He, Dening Lu, Linlin Xu, and Jonathan Li, “Nerf: Neural radiance field in 3d vision, a comprehensive review,” arXiv preprint arXiv:2210.00379, 2022

    Pith/arXiv arXiv 2022

  5. [5]

    A fuzzy c-means clustering algorithm for real medical image segmentation,

    Feifei Zhang, Fei Shi, Dayong Ren, and Yue Li, “A fuzzy c-means clustering algorithm for real medical image segmentation,” inICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2025, pp. 1–5

    2025

  6. [6]

    Plenoctrees for real-time rendering of neural radiance fields,

    Alex Yu, Ruilong Li, Matthew Tancik, Hao Li, Ren Ng, and Angjoo Kanazawa, “Plenoctrees for real-time rendering of neural radiance fields,” inProceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 5752–5761

    2021

  7. [7]

    Neural point-based graphics,

    Kara-Ali Aliev, Artem Sevastopolsky, Maria Kolos, Dmitry Ulyanov, and Victor Lempitsky, “Neural point-based graphics,” inEuropean conference on computer vision. Springer, 2020, pp. 696–712

    2020

  8. [8]

    Rpbg: Towards robust neural point-based graphics in the wild,

    Qingtian Zhu, Zizhuang Wei, Zhongtian Zheng, Yifan Zhan, Zhuyu Yao, Jiawang Zhang, Kejian Wu, and Yinqiang Zheng, “Rpbg: Towards robust neural point-based graphics in the wild,” inEuropean Conference on Computer Vision. Springer, 2024, pp. 389–406. TABLE I: The tracking performance on Replica[16], TUM-RGBD[9], and ScanNet[45] (ATE RMSE↓[cm]) Dataset Repli...

    arXiv 2024

  9. [9]

    A benchmark for the evaluation of rgb-d slam systems,

    J ¨urgen Sturm, Nikolas Engelhard, Felix Endres, Wolfram Burgard, and Daniel Cremers, “A benchmark for the evaluation of rgb-d slam systems,” in2012 IEEE/RSJ international conference on intelligent robots and systems. IEEE, 2012, pp. 573–580

    2012

  10. [10]

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

    Nikhil Keetha, Jay Karhade, Krishna Murthy Jatavallabhula, Geng- shan Yang, Sebastian Scherer, Deva Ramanan, and Jonathon Luiten, “Splatam: Splat track & map 3d gaussians for dense rgb-d slam,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 21357–21366

    2024

  11. [11]

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

    Carlos Campos, Richard Elvira, Juan J G ´omez Rodr´ıguez, Jos ´e MM Montiel, and Juan D Tard ´os, “Orb-slam3: An accurate open-source library for visual, visual–inertial, and multimap slam,”IEEE transac- Fig. 4: Trajectory maps on Replica[16], TUM-RGBD[9], and ScanNet[45]. TABLE IV: Novel And Train View Rendering Performance on Replica/R0 Methods Metrics ...

    2021

  12. [12]

    Direct sparse odometry,

    Jakob Engel, Vladlen Koltun, and Daniel Cremers, “Direct sparse odometry,”IEEE transactions on pattern analysis and machine intelligence, vol. 40, no. 3, pp. 611–625, 2017

    2017

  13. [13]

    Lsd-slam: Large- scale direct monocular slam,

    Jakob Engel, Thomas Sch ¨ops, and Daniel Cremers, “Lsd-slam: Large- scale direct monocular slam,” inEuropean conference on computer vision. Springer, 2014, pp. 834–849

    2014

  14. [14]

    Elasticfusion: Dense slam without a pose graph.,

    Thomas Whelan, Stefan Leutenegger, Renato F Salas-Moreno, Ben Glocker, and Andrew J Davison, “Elasticfusion: Dense slam without a pose graph.,” inRobotics: science and systems. Rome, 2015, vol. 11

    2015

  15. [16]

    The replica dataset: A digital replica of indoor spaces,

    Julian Straub, Thomas Whelan, Lingni Ma, Yufan Chen, Erik Wij- mans, Simon Green, Jakob J Engel, Raul Mur-Artal, Carl Ren, Shobhit Verma, et al., “The replica dataset: A digital replica of indoor spaces,” arXiv preprint arXiv:1906.05797, 2019

    Pith/arXiv arXiv 1906

  16. [17]

    The euroc micro aerial vehicle datasets,

    Michael Burri, Janosch Nikolic, Pascal Gohl, Thomas Schneider, Joern Rehder, Sammy Omari, Markus W Achtelik, and Roland Siegwart, “The euroc micro aerial vehicle datasets,”The International Journal of Robotics Research, vol. 35, no. 10, pp. 1157–1163, 2016

    2016

  17. [18]

    A non-local algo- rithm for image denoising,

    Antoni Buades, Bartomeu Coll, and J-M Morel, “A non-local algo- rithm for image denoising,” in2005 IEEE computer society conference on computer vision and pattern recognition (CVPR’05). Ieee, 2005, vol. 2, pp. 60–65

    2005

  18. [19]

    Dl-posenet: A differential lightweight network for pose regression over se (3),

    Wenjie Li, Jia Liu, Yanyan Wang, Wei Hao, Dayong Ren, and Lijun Chen, “Dl-posenet: A differential lightweight network for pose regression over se (3),” in2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024, pp. 16834–16840

    2024

  19. [20]

    Online deep bingham network for probabilistic orientation estimation,

    Wenjie Li, Jia Liu, Wei Hao, Haisong Liu, Dayong Ren, Yanyan Wang, and Lijun Chen, “Online deep bingham network for probabilistic orientation estimation,”IET Computer Vision, vol. 17, no. 6, pp. 663– 675, 2023

    2023

  20. [21]

    Robust moving least-squares fitting with sharp features,

    Shachar Fleishman, Daniel Cohen-Or, and Cl ´audio T Silva, “Robust moving least-squares fitting with sharp features,”ACM transactions on graphics (TOG), vol. 24, no. 3, pp. 544–552, 2005

    2005

  21. [22]

    Denoising point sets via l0 minimization,

    Yujing Sun, Scott Schaefer, and Wenping Wang, “Denoising point sets via l0 minimization,”Computer Aided Geometric Design, vol. 35, pp. 2–15, 2015

    2015

  22. [23]

    Applications of robust statistics in autonomous driving,

    Tino Werner, “Applications of robust statistics in autonomous driving,” 2025

    2025

  23. [24]

    Online photometric calibration of auto exposure video for realtime visual odometry and slam,

    Paul Bergmann, Rui Wang, and Daniel Cremers, “Online photometric calibration of auto exposure video for realtime visual odometry and slam,”IEEE Robotics and Automation Letters, vol. 3, no. 2, pp. 627– 634, 2017

    2017

  24. [25]

    Large scale photometric bundle adjustment,

    Oliver J Woodford and Edward Rosten, “Large scale photometric bundle adjustment,”arXiv preprint arXiv:2008.11762, 2020

    Pith/arXiv arXiv 2008

  25. [26]

    Physically- based photometric bundle adjustment in non-lambertian environ- ments,

    Lei Cheng, Junpeng Hu, Haodong Yan, Mariia Gladkova, Tianyu Huang, Yun-Hui Liu, Daniel Cremers, and Haoang Li, “Physically- based photometric bundle adjustment in non-lambertian environ- ments,” in2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2024, pp. 10461–10468

    2024

  26. [27]

    Sae: Estimation for transition matrix in annotation algorithms,

    Dayong Ren, Shuangyu Yang, Wenjie Li, Jie Guo, and Yanwen Guo, “Sae: Estimation for transition matrix in annotation algorithms,”arXiv preprint, 2022

    2022

  27. [28]

    Point attention network for point cloud semantic segmentation,

    Dayong Ren, Zhengyi Wu, Jiawei Li, Piaopiao Yu, Jie Guo, Mingqiang Wei, and Yanwen Guo, “Point attention network for point cloud semantic segmentation,”Science China Information Sciences, vol. 65, no. 9, pp. 192104, 2022

    2022

  28. [29]

    Spiking pointnet: Spiking neural net- works for point clouds,

    Dayong Ren, Zhe Ma, Yuanpei Chen, Weihang Peng, Xiaode Liu, Yuhan Zhang, and Yufei Guo, “Spiking pointnet: Spiking neural net- works for point clouds,”Advances in Neural Information Processing Systems, vol. 36, 2024

    2024

  29. [30]

    Mffnet: multimodal feature fusion network for point cloud semantic segmentation,

    Dayong Ren, Jiawei Li, Zhengyi Wu, Jie Guo, Mingqiang Wei, and Yanwen Guo, “Mffnet: multimodal feature fusion network for point cloud semantic segmentation,”The Visual Computer, vol. 40, no. 8, pp. 5155–5167, 2024

    2024

  30. [31]

    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,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 13293–13302

    2023

  31. [32]

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

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

    Pith/arXiv arXiv 2023

  32. [33]

    Es- lam: Efficient dense slam system based on hybrid representation of signed distance fields,

    Mohammad Mahdi Johari, Camilla Carta, and Franc ¸ois Fleuret, “Es- lam: Efficient dense slam system based on hybrid representation of signed distance fields,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 17408–17419

    2023

  33. [34]

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

    Erik Sandstr ¨om, Yue Li, Luc Van Gool, and Martin R Oswald, “Point- slam: Dense neural point cloud-based slam,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 18433–18444

    2023

  34. [35]

    Nice-slam: Neural implicit scalable encoding for slam,

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

    2022

  35. [36]

    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,” in2022 IEEE International Symposium on Mixed and Augmented Reality (ISMAR). IEEE, 2022, pp. 499–507

    2022

  36. [37]

    π-lsam: Lidar smoothing and mapping with planes,

    Lipu Zhou, Shengze Wang, and Michael Kaess, “π-lsam: Lidar smoothing and mapping with planes,” in2021 IEEE international conference on robotics and automation (ICRA). IEEE, 2021, pp. 5751– 5757

    2021

  37. [38]

    Lidar-net: A real-scanned 3d point cloud dataset for indoor scenes,

    Yanwen Guo, Yuanqi Li, Dayong Ren, Xiaohong Zhang, Jiawei Li, Liang Pu, Changfeng Ma, Xiaoyu Zhan, Jie Guo, Mingqiang Wei, et al., “Lidar-net: A real-scanned 3d point cloud dataset for indoor scenes,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 21989–21999

    2024

  38. [39]

    Adaptive gaussian regularization constrained sparse subspace clustering for image segmentation,

    Sensen Song, Dayong Ren, Zhenhong Jia, and Fei Shi, “Adaptive gaussian regularization constrained sparse subspace clustering for image segmentation,” inICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2024, pp. 4400–4404

    2024

  39. [40]

    Akdt: Adaptive kernel dilation transformer for effective image de- noising,

    Alexandru Brateanu, Raul Balmez, Adrian Avram, and Ciprian Orhei, “Akdt: Adaptive kernel dilation transformer for effective image de- noising,”Proceedings Copyright, vol. 418, no. 425, pp. 19

  40. [41]

    Point cloud denoising and feature preservation: an adaptive kernel approach based on local density and global statistics,

    Lianchao Wang, Yijin Chen, Wenhui Song, and Hanghang Xu, “Point cloud denoising and feature preservation: an adaptive kernel approach based on local density and global statistics,”Sensors, vol. 24, no. 6, pp. 1718, 2024

    2024

  41. [42]

    Geosegnet: point cloud semantic segmentation via geometric encoder–decoder modeling,

    Chen Chen, Yisen Wang, Honghua Chen, Xuefeng Yan, Dayong Ren, Yanwen Guo, Haoran Xie, Fu Lee Wang, and Mingqiang Wei, “Geosegnet: point cloud semantic segmentation via geometric encoder–decoder modeling,”The Visual Computer, vol. 40, no. 8, pp. 5107–5121, 2024

    2024

  42. [43]

    Zigza- gpointmamba: Spatial-semantic mamba for point cloud understand- ing,

    Linshuang Diao, Dayong Ren, Sensen Song, and Yurong Qian, “Zigza- gpointmamba: Spatial-semantic mamba for point cloud understand- ing,”arXiv preprint arXiv:2505.21381, 2025

    arXiv 2025

  43. [44]

    A practical grabcut color image segmentation based on bayes classifica- tion and simple linear iterative clustering,

    Dayong Ren, Zhenhong Jia, Jie Yang, and Nikola K Kasabov, “A practical grabcut color image segmentation based on bayes classifica- tion and simple linear iterative clustering,”IEEE Access, vol. 5, pp. 18480–18487, 2017

    2017

  44. [45]

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

    Angela Dai, Angel X Chang, Manolis Savva, Maciej Halber, Thomas Funkhouser, and Matthias Nießner, “Scannet: Richly-annotated 3d reconstructions of indoor scenes,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 5828–5839

    2017