Pith. sign in

REVIEW 3 major objections 5 minor 66 references

Scalable Benchmarking and Robust Learning for Noise-Free Ego-Motion and 3D Reconstruction from Noisy Video

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read CorrGS, a test-time adaptation method built on Gaussian Splatting, uses visual correspondences between noisy observations and clean rendered frames to keep ego-motion accurate and 3D reconstruction noise-free under fast motion and…

desk verdict Solid benchmark with an overclaimed method; Robust-Ego3D deserves publication, but CorrGS is demonstrated only for brightness change plus fast motion. read the letter →

arxiv 2501.14319 v1 pith:HLYCBGRR submitted 2025-01-24 cs.CV cs.RO

Xiaohao Xu , Tianyi Zhang , Shibo Zhao , Xiang Li , Sibo Wang , Yongqi Chen , Ye Li , Bhiksha Raj
show 3 more authors
Matthew Johnson-Roberson Sebastian Scherer Xiaonan Huang
This is my paper · ORCID
classification cs.CVcs.RO
keywords noisyvideoego-motionestimation3DreconstructionGaussianSplattingvisualcorrespondencetest-timeadaptationSLAMrobustnesssyntheticbenchmark
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper argues that the usual assumption of noise-free sensor data has held back ego-motion estimation and photorealistic 3D reconstruction, and that robustness should be tested and engineered for. To support this, it builds a scalable pipeline that corrupts clean 3D scenes into noisy RGB-D videos, instantiates a 124-setting benchmark called Robust-Ego3D, and shows that existing dense neural SLAM models degrade sharply under fast motion, depth noise, and dynamic lighting. It then proposes CorrGS, which uses a clean Gaussian-splat map as reference: correspondence-guided pose learning initializes each camera from 2D matches lifted to 3D, and correspondence-guided appearance restoration learns to map noisy colors back to clean ones. On synthetic sparse-view video with partial brightness reduction, CorrGS lifts tracking success from 1/8 to 8/8 sequences and reaches 0.54 cm ATE, close to the noise-free 0.45 cm, while restoring RGB to 35.38 dB PSNR.

What carries the argument

The load-bearing mechanism is the pair of correspondence-guided modules built on a Gaussian Splatting map. Correspondence-guided Pose Learning (CPL) computes 2D matches between the rendered and observed RGB frames, lifts them to 3D via rendered and observed depth, solves a soft-L1 relative pose problem, and feeds the result into differentiable pose optimization after a Pose Quality Verification step rejects bad initializations. Correspondence-guided Appearance Restoration Learning (CARL) fits a restoration model that maps noisy observed colors to clean rendered colors at the matched points, applies it to the whole observed frame, and uses the restored frame for a second round of CPL and for map updates. The Gaussian-splat representation provides the fast, differentiable rendering that makes this loop practical.

What would settle it

A sequence where the map itself is progressively corrupted (e.g., by adding noise to the Gaussian splat colors or depths during tracking) and where matching under strong motion blur or severe brightness reduction breaks down; if CorrGS then diverges or produces noisy reconstructions, the claim that it recovers noise-free ego-motion and 3D from noisy video would be falsified for those conditions.

Watch

Extended reading notes

Core claim

The central claim is that a dense neural SLAM system can stay robust under realistic noise if it treats its own clean 3D map as a reference and aligns each noisy observation to it by visual correspondence. CorrGS renders RGB-D frames from a maintained clean Gaussian-splat map, establishes 2D matches between rendered and observed frames, lifts them to 3D to initialize the camera pose, and then uses the aligned correspondences to learn a restoration function that maps observed colors to clean rendered colors. The restored image is used both for tracking and for map update, so pose accuracy and appearance fidelity reinforce each other. The paper reports that this design outperforms prior state-of-the-art neural SLAM methods in rapid-motion and dynamic-illumination scenarios, converting tracking failures into successful trajectories and producing photorealistic reconstructions from noisy sparse-view video.

Load-bearing premise

The method assumes that a clean, trustworthy 3D map and reliable visual correspondences between noisy observations and its renderings are available; if the map is already corrupted or matching fails, there is no clean reference to recover from.

Editorial extensions

If this is right

  • If CorrGS is correct, test-time adaptation to a clean internal reference is a viable route to robustness, so SLAM systems need not be retrained for each perturbation type.
  • The Robust-Ego3D benchmark exposes that no single existing model is robust across all perturbation categories, implying that robustness evaluation should include motion, sensor, and synchronization noise, not just image-level corruptions.
  • Highly correlated perturbations (e.g., Gaussian, shot, impulse noise, and spatter) can act as proxies in benchmarking, cutting evaluation cost while retaining a rough performance estimate.
  • Correspondence initialization plus a quality gate can rescue differentiable pose optimization under fast motion, where large pose changes otherwise amplify gradients and cause divergence.

Reading between the lines

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

  • The same correspondence-guided restoration loop could be extended to depth noise and desynchronization, since the paper only tests brightness reduction; restoring depth against rendered depth is a natural next step the authors leave implicit.
  • Because CARL learns the mapping from noisy to clean colors at correspondence points, it could also serve as a self-supervised denoiser for other degradation types (blur, snow) as long as the clean historical map remains trustworthy.
  • The reliance on a clean historical map suggests an online monitoring scheme: when rendering loss rises, the system could flag map corruption or matching failure, effectively using the method's own internal signal as a failure detector.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces Robust-Ego3D, a synthetic benchmark for dense neural SLAM robustness under 124 RGB-D perturbation settings spanning motion deviations, faster motion, RGB imaging corruptions, depth imaging perturbations, and RGB-D desynchronization, and it proposes CorrGS, a Gaussian-splatting SLAM method that uses learned correspondences between noisy observations and rendered RGB-D frames from a historical map to initialize poses and to learn a linear appearance restorer. The benchmark compares five neural SLAM baselines plus ORB-SLAM3 across these settings, with each setting averaged over 24 trials. CorrGS is evaluated on sparse-view synthetic video with 10× speed and partial brightness reduction, reporting ATE 0.54 cm versus baseline tracking failures and restored PSNR 35.38 dB, together with one qualitative real-world sequence with dynamic illumination and fast motion.

Significance. If the benchmark and pipeline are released as described, they form a useful community resource: the perturbation taxonomy is physically motivated and spans a broader range of degradations than existing RGB-D SLAM benchmarks, the pipeline is customizable, and the 24-trial averaging supports more reliable baseline comparisons than typical single-run evaluations. CorrGS also demonstrates a sensible and lightweight combination of correspondence-based pose initialization, pose quality verification, and online appearance restoration, with a clean ablation of the restoration and tracking components. However, the paper's central claim of robust ego-motion and noise-free 3D reconstruction from generalized noisy video is validated only under partial brightness reduction in the latter half of each sequence, so the durable contribution is currently the benchmark rather than the demonstrated scope of CorrGS.

major comments (3)
  1. [§5.2, 'Noisy data synthesis setup'] The abstract and title claim robust ego-motion and noise-free 3D reconstruction from noisy video, and Section 4 frames CorrGS as addressing "generalized noisy conditions," but the only synthetic test of CorrGS uses partial brightness reduction in the latter half of each video, with the paper stating "We focus on illumination changes, leaving other perturbations for future work." Motion perturbations are tested only as fast motion without simultaneous imaging noise, and depth noise, desynchronization, and mixed perturbations from the Robust-Ego3D benchmark are not applied to CorrGS. This mismatch between the claim and the tested setting is load-bearing; please either add CorrGS experiments on the other perturbation families (especially depth noise and RGB-D desynchronization) or narrow the claims to illumination degradation and fast motion.
  2. [§5.1, Algorithm 1 and CARL] CorrGS's restoration target in Eq. (5) is the color rendered from the historical map, and CPL matches the noisy observation against rendered RGB-D from that same map, so the method assumes a trustworthy clean historical map. In the synthetic protocol the first half of the sequence is clean, and the map is therefore uncontaminated when the brightness reduction begins, so CARL only has to invert a global brightness change and CPL is matching against an essentially clean render. This does not test the regime in which noise is present from the first frame, where the historical map and rendered targets are themselves degraded and CARL has no clean reference to fall back on. The real-world pilot in Section 5.3 is qualitative and does not report ATE or reconstruction metrics, so the clean-map assumption is not stress-tested under the stated scope. Please add a controlled experiment with perturbations present from frame 1, or explicitly state this as a limitation of the method.
  3. [Tables 5 and 6] The benchmark section averages 24 trials per setting and reports standard deviations for ORB-SLAM3, but the CorrGS comparison tables report single numbers without standard deviations or the number of seeds. Because the central claim is that CorrGS "consistently outperforms" prior methods, and the baselines exhibit tracking failures that may be high-variance, please report repeated-run statistics for CorrGS and the baselines in these tables to support the consistency claim.
minor comments (5)
  1. [§5.2, text and Table 6] The text reports an average restored PSNR of 35.48 dB, while Table 6 lists the average as 35.38 dB; please reconcile the discrepancy.
  2. [Appendix C, Q6] The datasheet says the benchmark "includes 1,000 perturbed settings, each with 2,000 RGB-D video sequences," which conflicts with the earlier description of 1,000 videos and 2,000 frames per sequence; please correct the wording.
  3. [Appendix E] Theorems A–E are informal gradient-magnitude arguments rather than formal theorems with explicit assumptions and proofs; consider relabeling them as observations or propositions and adding the missing hypotheses.
  4. [Table 5] The column headers mix loop-closure indicators and sequence identifiers in a way that is hard to parse; please clarify which runs use loop closure and what O-0 through R-2 denote.
  5. [Title and abstract] The phrase "noise-free ego-motion and 3D reconstruction from noisy video" could be misread as claiming noise-free input; consider rewording to clarify that the reconstructed map and trajectory estimates, not the input observations, are noise-free.

Circularity Check

1 steps flagged · score 2.0 of 10

Core benchmarking and ATE comparisons are externally anchored; the only self-referential element is CARL's restoration PSNR, which is measured on the same frames used to fit the linear restorer, making that metric a calibration fit rather than an independent prediction.

  1. fitted input called prediction [Sec. 5.1 (Eq. 5, CARL) and Sec. 5.2 (Table 6, RGB Restoration Quality)]
    "CARL mitigates color degradation by learning a restoration model f(·; θ) that maps noisy colors Co,i of observed points to their clean counterparts Cr,i, rendered from historical map which is maintained to be noise-free: θ∗ = arg min ... (5). ... CARL uses a linear model for restoration, optimized using the Adam optimizer over 100 iterations with a learning rate of 0.2. ... The learnt model is then applied to the observed image."

    The linear restorer is fit on correspondences between the observed noisy frame and the render of the historical clean map (Eq. 5), then applied to that same observed frame. Table 6's 'RGB Restoration Quality (PSNR)' therefore reports the fit quality on the training frame, not an out-of-sample prediction. Under the paper's protocol—global brightness reduction in the latter half while the map is built from the clean first half—a linear least-squares fit on matched pixels essentially determines the full-frame restoration, so the reported 35.38 dB PSNR is partly a consequence of calibrating to the model's own clean render. The ATE and depth results are not affected by this and are measured against external Replica ground truth.

full rationale

No circularity appears in the benchmark construction or in the central pose-tracking comparison. Robust-Ego3D is generated from external Replica meshes and trajectories, and CorrGS is compared against published baselines (SplaTAM, GO-SLAM, CO-SLAM, iMAP, Nice-SLAM, ORB-SLAM3) using ATE against ground-truth trajectories. CPL relies on externally trained LoFTR correspondences, and the pose optimization is not fitted to the reported ATE. The only self-referential component is CARL, whose clean target is the model's own historical map; because the synthetic protocol makes the first half of each video clean and evaluates restoration PSNR on the same frames used to fit the linear model, that metric is a calibration fit rather than a forced independent prediction. This is a minor evaluation limitation, not a derivation that reduces the paper's central claims to its inputs. No load-bearing self-citation chain or imported uniqueness theorem appears; the cited prior work (SplaTAM, LoFTR, Replica) is external and reproducible.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

No new physical entities, forces, or mediators are introduced. The benchmark uses hand-selected perturbation severities as free inputs, and CorrGS relies on standard domain assumptions about static scenes, dense depth, reliable correspondences, and a clean historical map.

free parameters (5)
  • RGB imaging perturbation severity levels = Levels 1-5 per type (Table A1)
    Hand-selected following Hendrycks and Dietterich (2019); they set benchmark difficulty but are not fitted to model outputs.
  • Depth perturbation severity parameters = Gaussian sigma 0.1-0.5, missing rate 10-30%, clipping ranges in Table A2
    Chosen to approximate TUM-RGBD depth statistics; not fitted to benchmark results.
  • Motion deviation magnitudes = Rotation sigma 1/3/5 degrees; translation sigma 1.25/2.5/5 cm
    Selected to span a progressive difficulty range and to emulate robot vibration and jumping; not fitted.
  • Desynchronization frame intervals = Delta = 5, 10, 20 frames (static and dynamic)
    Hand-chosen temporal misalignments to test multi-sensor timing; not fitted.
  • CARL online restoration parameters = Linear map optimized over 100 iterations with learning rate 0.2
    A per-sequence fitted color transformation from noisy colors to rendered clean colors; it is a learned adaptation, not a global constant.
assumptions (6)
  • standard math SO(3) rotation perturbations are generated via the exponential map of Gaussian angular velocities (Appendix A.1, Eq. A2).
    Used to synthesize rotation and translation deviations; standard rigid-body kinematics.
  • domain assumption Static, bounded indoor scenes with known camera intrinsics, dense depth, and known motion and observation models (Appendix B.1).
    Defines the benchmark scope and excludes dynamic objects, outdoor scenes, sparse LiDAR depth, and active decision-making.
  • domain assumption Simplified linear perturbation models, such as additive Gaussian noise, approximate real-world sensing degradation (Appendix B.1).
    Needed to justify transferring benchmark conclusions to real deployment; the authors acknowledge this simplification.
  • domain assumption LoFTR correspondences between rendered and observed RGB frames can be lifted to reliable 3D correspondences via depth (Sec. 5.1, Eq. 4).
    CPL's pose initialization fails if matching or depth lifting is unreliable under heavy noise or large viewpoint change.
  • domain assumption A clean historical map renders trustworthy RGB-D targets for restoration learning (Sec. 5.1, CARL).
    CARL fits a color mapping from noisy observations to the rendered historical map; if the historical map contains noise or drift, the restoration target is corrupted.
  • domain assumption Theoretical analyses assume L2 photometric and depth losses and additive Gaussian noise (Appendix E, Theorems A-E).
    The informal theorems explain gradient behavior under these models; real perturbations may deviate from these assumptions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Scalable Benchmarking and Robust Learning for Noise-Free Ego-Motion and 3D Reconstruction from Noisy Video." pith.science (2026). https://pith.science/paper/HLYCBGRR

@misc{pith2026250114319,
  author       = {Pith},
  title        = {Pith review of: Scalable Benchmarking and Robust Learning for Noise-Free Ego-Motion and 3D Reconstruction from Noisy Video},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HLYCBGRR}},
  note         = {Machine review of arXiv:2501.14319}
}
read the original abstract

We aim to redefine robust ego-motion estimation and photorealistic 3D reconstruction by addressing a critical limitation: the reliance on noise-free data in existing models. While such sanitized conditions simplify evaluation, they fail to capture the unpredictable, noisy complexities of real-world environments. Dynamic motion, sensor imperfections, and synchronization perturbations lead to sharp performance declines when these models are deployed in practice, revealing an urgent need for frameworks that embrace and excel under real-world noise. To bridge this gap, we tackle three core challenges: scalable data generation, comprehensive benchmarking, and model robustness enhancement. First, we introduce a scalable noisy data synthesis pipeline that generates diverse datasets simulating complex motion, sensor imperfections, and synchronization errors. Second, we leverage this pipeline to create Robust-Ego3D, a benchmark rigorously designed to expose noise-induced performance degradation, highlighting the limitations of current learning-based methods in ego-motion accuracy and 3D reconstruction quality. Third, we propose Correspondence-guided Gaussian Splatting (CorrGS), a novel test-time adaptation method that progressively refines an internal clean 3D representation by aligning noisy observations with rendered RGB-D frames from clean 3D map, enhancing geometric alignment and appearance restoration through visual correspondence. Extensive experiments on synthetic and real-world data demonstrate that CorrGS consistently outperforms prior state-of-the-art methods, particularly in scenarios involving rapid motion and dynamic illumination.

Figures

Figures reproduced from arXiv: 2501.14319 by the authors.

Figure 1
Figure 1. Towards robust ego-motion and photorealistic 3D reconstruction. (a) Previous ap￾proaches rely on synthetic datasets with perfect conditions (noise-free and smooth motion). (b) Real-world data introduce inherent noise and complexities. We present a customizable noisy data synthesis pipeline to evaluate methods under realistic noisy conditions. (c) Our Robust-Ego3D benchmark reveals that existing methods produce noisy… view at source ↗
Figure 2
Figure 2. Framework for noisy data syn￾thesis, model estimation, and evaluation. p(z1:t, x1:t | m, x 0 1:t , ξ1:t). (1) Model estimation (from noisy video to noise-free ego-motion and 3D reconstruction). From noisy ob￾servations z1:t, the model estimates the 3D scene ˆm and the trajectory ˆx1:t: p( ˆm, ˆx1:t | z1:t). (2) Evaluation and feedback. The estimates ˆm and ˆx1:t are compared with ground truth m and x1:t, revealing p… view at source ↗
Figure 3
Figure 3. Perturbation taxonomy for RGB-D sensing. 3.2 PERTURBATION TAXONOMY FOR RGB-D SENSING SYSTEMS Perturbation sources. As shown in [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Noisy data synthesis pipeline. from the 3D scene mesh, perturbed tra￾jectory, and sensor configurations. Sub￾sequently, sensor-related perturbations, in￾cluding imaging corruptions and multi￾sensor desynchronization, are composed, resulting in noisy and unstable sensor…
Figure 5
Figure 5. Figure 5: Effect of Shot Noise on 3D reconstruction of Nice-SLAM. 3) Comparison across perturbations categories. Different perturbations impact performance to varying degrees, with environmental effects posing the most significant challenge, followed by sensor noise, while image…
Figure 7
Figure 7. Figure 7: Effect of fast motion [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 9
Figure 9. Figure 9: Effect of RGB-D desynchronization. improve robustness, combining the learnable components of Neural methods with ORB-SLAM3’s resilience to fast motion could offer a more balanced approach [PITH_FULL_IMAGE:figures/full_fig_p007_9.png]
Figure 10
Figure 10. Figure 10: Rendering under mixed perturbations (same as [PITH_FULL_IMAGE:figures/full_fig_p008_10.png]
Figure 11
Figure 11. Figure 11: Correlation of performance (ATE) across static RGB perturbations. under large-scale perturbations can be resource-intensive, requiring numerous runs across varying perturbation types, severity levels, scenes, and models. However, by under￾standing correlations between…
Figure 12
Figure 12. Figure 12: Ablation of CPL on 3D re￾construction from sparse-view video. 4) However, the initialization from Eq. 4 may fail when the two viewpoints are very close together. In such scenarios, it is challenging to accurately estimate the relative pose by optimizing the transforma…
Figure 13
Figure 13. Figure 13: Effect of CARL on 3D recon￾struction from sparse-view noisy video. Effect of CARL: Online restoration learning enables consistent noise-free 3D reconstruction [PITH_FULL_IMAGE:figures/full_fig_p010_13.png]
Figure 14
Figure 14. Figure 14: RGB-D rendering and 3D reconstruction by [PITH_FULL_IMAGE:figures/full_fig_p010_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

66 extracted references · 49 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    IEEEtran_HOWTO

    11em plus .33em minus .07em @technote 4000 4000 100 4000 4000 500 `\.=1000 = #1 #1 #1 0pt [0pt][0pt] #1 * \| ** #1 \@IEEEauthorblockNstyle \@IEEEauthorblockAstyle \@IEEEauthordefaulttextstyle \@IEEEauthorblockconfadjspace -0.25em \@IEEEauthorblockNtopspace 0.0ex \@IEEEauthorblockAtopspace 0.0ex \@IEEEauthorblockNinterlinespace 2.6ex \@IEEEauthorblockAinte...

    work page 2007

  3. [3]

    11em plus .33em minus .07em 4000 4000 100 4000 4000 500 `\.=1000 = #1 \@IEEEnotcompsoconly \@IEEEcompsoconly #1 * [1] 0pt [0pt][0pt] #1 * \| ** #1 \@IEEEauthorblockNstyle \@IEEEcompsocnotconfonly \@IEEEcompsocconfonly \@IEEEauthorblockAstyle \@IEEEcompsocnotconfonly \@IEEEcompsocconfonly \@IEEEcompsocconfonly \@IEEEauthordefaulttextstyle \@IEEEcompsocnotc...

    work page

  4. [4]

    Powerful, soft combustion actuators for insect-scale robots

    Cameron A Aubin, Ronald H Heisser, Ofek Peretz, Julia Timko, Jacqueline Lo, E Farrell Helbling, Sadaf Sobhani, Amir D Gat, and Robert F Shepherd. Powerful, soft combustion actuators for insect-scale robots. Science, 381 0 (6663): 0 1212--1217, 2023

    work page 2023

  5. [5]

    Robust slam systems: Are we there yet? In 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp.\ 5320--5327

    Mihai Bujanca, Xuesong Shi, Matthew Spear, Pengpeng Zhao, Barry Lennox, and Mikel Luj \'a n. Robust slam systems: Are we there yet? In 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp.\ 5320--5327. IEEE, 2021

    work page 2021

  6. [6]

    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, 2016

    work page 2016

  7. [7]

    Cesar Cadena, Luca Carlone, Henry Carrillo, Yasir Latif, Davide Scaramuzza, José Neira, Ian Reid, and John J. Leonard. Past, present, and future of simultaneous localization and mapping: Toward the robust-perception age. IEEE Transactions on Robotics, 32 0 (6): 0 1309--1332, 2016

    work page 2016

  8. [8]

    Gómez Rodríguez, José M

    Carlos Campos, Richard Elvira, Juan J. Gómez Rodríguez, José M. M. Montiel, and Juan D. Tardós. Orb-slam3: An accurate open-source library for visual, visual–inertial, and multimap slam. IEEE Transactions on Robotics, 37 0 (6): 0 1874--1890, 2021

    work page 2021

Show all 66 references
  1. [9]

    Shapenet: An information-rich 3d model repository

    Angel X Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese, Manolis Savva, Shuran Song, Hao Su, et al. Shapenet: An information-rich 3d model repository. arXiv preprint arXiv:1512.03012, 2015

    2015 arXiv

  2. [10]

    Robustnav: Towards benchmarking robustness in embodied navigation

    Prithvijit Chattopadhyay, Judy Hoffman, Roozbeh Mottaghi, and Aniruddha Kembhavi. Robustnav: Towards benchmarking robustness in embodied navigation. In IEEE/CVF International Conference on Computer Vision, pp.\ 15691--15700, 2021

    2021

  3. [11]

    Pearson correlation coefficient

    Israel Cohen, Yiteng Huang, Jingdong Chen, Jacob Benesty, Jacob Benesty, Jingdong Chen, Yiteng Huang, and Israel Cohen. Pearson correlation coefficient. Noise reduction in speech processing, pp.\ 1--4, 2009

    2009

  4. [12]

    Scannet: Richly-annotated 3d reconstructions of indoor scenes

    Angela Dai, Angel X Chang, Manolis Savva, Maciej Halber, Thomas Funkhouser, and Matthias Nie ner. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 5828--5839, 2017 a

    2017

  5. [13]

    Bundlefusion: Real-time globally consistent 3d reconstruction using on-the-fly surface reintegration

    Angela Dai, Matthias Nie ner, Michael Zollh\" o fer, Shahram Izadi, and Christian Theobalt. Bundlefusion: Real-time globally consistent 3d reconstruction using on-the-fly surface reintegration. ACM Trans. Graph., 36 0 (4), jul 2017 b

    2017

  6. [14]

    Carla: An open urban driving simulator

    Alexey Dosovitskiy, German Ros, Felipe Codevilla, Antonio Lopez, and Vladlen Koltun. Carla: An open urban driving simulator. In Conference on robot learning, pp.\ 1--16. PMLR, 2017

    2017

  7. [15]

    Kamak Ebadi, Lukas Bernreiter, Harel Biggie, Gavin Catt, Yun Chang, Arghya Chatterjee, Christopher E. Denniston, Simon-Pierre Deschênes, Kyle Harlow, Shehryar Khattak, Lucas Nogueira, Matteo Palieri, Pavel Petráček, Matěj Petrlík, Andrzej Reinke, Vít Krátký, Shibo Zhao, Ali-ak...

    2023

  8. [16]

    Optimization based full body control for the atlas robot

    Siyuan Feng, Eric Whitman, X Xinjilefu, and Christopher G Atkeson. Optimization based full body control for the atlas robot. In 2014 IEEE-RAS International Conference on Humanoid Robots, pp.\ 120--127. IEEE, 2014

    2014

  9. [17]

    Datasheets for datasets

    Timnit Gebru, Jamie Morgenstern, Briana Vecchione, Jennifer Wortman Vaughan, Hanna Wallach, Hal Daum\' e III, and Kate Crawford. Datasheets for datasets. Commun. ACM, 64 0 (12): 0 86–92, nov 2021

    2021

  10. [18]

    Hopkins, Georg Wiedebach, Jared Bishop, Steven Pickles, David Muller, and Moritz Bacher

    Ruben Grandia, Espen Knoop, Michael A. Hopkins, Georg Wiedebach, Jared Bishop, Steven Pickles, David Muller, and Moritz Bacher. Design and control of a bipedal robotic character. In Robotics: Science and Systems 2024, Delft, Netherlands, July 15-July 19 2024. Disney Research, ...

    2024

  11. [19]

    Photon, poisson noise

    Samuel W Hasinoff. Photon, poisson noise. Computer Vision, A Reference Guide, 4 0 (16): 0 1, 2014

    2014

  12. [20]

    The hilti slam challenge dataset

    Michael Helmberger, Kristian Morin, Beda Berner, Nitish Kumar, Giovanni Cioffi, and Davide Scaramuzza. The hilti slam challenge dataset. IEEE Robotics and Automation Letters, 7 0 (3): 0 7518--7525, 2022

    2022

  13. [21]

    Benchmarking neural network robustness to common corruptions and perturbations

    Dan Hendrycks and Thomas Dietterich. Benchmarking neural network robustness to common corruptions and perturbations. Proceedings of the International Conference on Learning Representations, 2019

    2019

  14. [22]

    Live demonstration: Tensegrity state estimation

    Xiaonan Huang, William R Johnson, Joran Booth, and Rebecca Kramer-Bottiglio. Live demonstration: Tensegrity state estimation. In 2022 IEEE Sensors, pp.\ 1--1. IEEE, 2022

    2022

  15. [23]

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

    Mohammad Mahdi Johari, Camilla Carta, and Fran c ois 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, pp.\ 17408--17419, 2023

    2023

  16. [24]

    Benchmarking the robustness of semantic segmentation models

    Christoph Kamann and Carsten Rother. Benchmarking the robustness of semantic segmentation models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 8828--8838, 2020

    2020

  17. [25]

    Champion-level drone racing using deep reinforcement learning

    Elia Kaufmann, Leonard Bauersfeld, Antonio Loquercio, Matthias M \"u ller, Vladlen Koltun, and Davide Scaramuzza. Champion-level drone racing using deep reinforcement learning. Nature, 620 0 (7976): 0 982--987, 2023

    2023

  18. [26]

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

    Nikhil Keetha, Jay Karhade, Krishna Murthy Jatavallabhula, Gengshan Yang, Sebastian Scherer, Deva Ramanan, and Jonathon Luiten. Splatam: Splat, track map 3d gaussians for dense rgb-d slam. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024

    2024

  19. [27]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk \"u hler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics, 42 0 (4), 2023

    2023

  20. [28]

    T test as a parametric statistic

    Tae Kyun Kim. T test as a parametric statistic. Korean journal of anesthesiology, 68 0 (6): 0 540--546, 2015

    2015

  21. [29]

    Ascento: A two-wheeled jumping robot

    Victor Klemm, Alessandro Morra, Ciro Salzmann, Florian Tschopp, Karen Bodie, Lionel Gulich, Nicola K \"u ng, Dominik Mannhart, Corentin Pfister, Marcus Vierneisel, et al. Ascento: A two-wheeled jumping robot. In 2019 International Conference on Robotics and Automation (ICRA), ...

    2019

  22. [30]

    Robo3d: Towards robust and reliable 3d perception against corruptions

    Lingdong Kong, Youquan Liu, Xin Li, Runnan Chen, Wenwei Zhang, Jiawei Ren, Liang Pan, Kai Chen, and Ziwei Liu. Robo3d: Towards robust and reliable 3d perception against corruptions. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pp.\ 19994--20006, 2023

    2023

  23. [31]

    Robust referring video object segmentation with cyclic structural consensus

    Xiang Li, Jinglu Wang, Xiaohao Xu, Xiao Li, Bhiksha Raj, and Yan Lu. Robust referring video object segmentation with cyclic structural consensus. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pp.\ 22236--22245, 2023

    2023

  24. [32]

    Gaussian splatting slam

    Hidenobu Matsuki, Riku Murai, Paul HJ Kelly, and Andrew J Davison. Gaussian splatting slam. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 18039--18048, 2024

    2024

  25. [33]

    Benchmarking robustness in object detection: Autonomous driving when winter is coming

    Claudio Michaelis, Benjamin Mitzkus, Robert Geirhos, Evgenia Rusak, Oliver Bringmann, Alexander S Ecker, Matthias Bethge, and Wieland Brendel. Benchmarking robustness in object detection: Autonomous driving when winter is coming. arXiv preprint arXiv:1907.07484, 2019

    1907 arXiv

  26. [34]

    Raúl Mur-Artal and Juan D. Tardós. Orb-slam2: An open-source slam system for monocular, stereo, and rgb-d cameras. IEEE Transactions on Robotics, 33 0 (5): 0 1255--1262, 2017

    2017

  27. [35]

    Newcombe, Shahram Izadi, Otmar Hilliges, David Molyneaux, David Kim, Andrew J

    Richard A. Newcombe, Shahram Izadi, Otmar Hilliges, David Molyneaux, David Kim, Andrew J. Davison, Pushmeet Kohli, Jamie Shotton, Steve Hodges, and Andrew W. Fitzgibbon. Kinectfusion: Real-time dense surface mapping and tracking. In 10th IEEE International Symposium on Mixed a...

    2011

  28. [36]

    Highly dynamic bistable soft actuator for reconfigurable multimodal soft robots (adv

    Dinesh K Patel, Xiaonan Huang, Yichi Luo, Mrunmayi Mungekar, M Khalid Jawed, Lining Yao, and Carmel Majidi. Highly dynamic bistable soft actuator for reconfigurable multimodal soft robots (adv. mater. technol. 2/2023). Advanced Materials Technologies, 8 0 (2): 0 2370006, 2023 a

    2023

  29. [37]

    Covins-g: A generic back-end for collaborative visual-inertial slam

    Manthan Patel, Marco Karrer, Philipp Bänninger, and Margarita Chli. Covins-g: A generic back-end for collaborative visual-inertial slam. In 2023 IEEE International Conference on Robotics and Automation (ICRA), pp.\ 2076--2082, 2023 b

    2023

  30. [38]

    Measuring robustness of visual slam

    David Prokhorov, Dmitry Zhukov, Olga Barinova, Konushin Anton, and Anna Vorontsova. Measuring robustness of visual slam. In 2019 16th International conference on machine vision applications (MVA), pp.\ 1--6. IEEE, 2019

    2019

  31. [39]

    Nerf-slam: Real-time dense monocular slam with neural radiance fields

    Antoni Rosinol, John J Leonard, and Luca Carlone. Nerf-slam: Real-time dense monocular slam with neural radiance fields. arXiv preprint arXiv:2210.13641, 2022

    2022 arXiv

  32. [40]

    System design and locomotion of superball, an untethered tensegrity robot

    Andrew P Sabelhaus, Jonathan Bruce, Ken Caluwaerts, Pavlo Manovi, Roya Fallah Firoozi, Sarah Dobi, Alice M Agogino, and Vytas SunSpiral. System design and locomotion of superball, an untethered tensegrity robot. In 2015 IEEE international conference on robotics and automation ...

    2015

  33. [41]

    Erik Sandström, Yue Li, Luc Van Gool, and Martin R. Oswald. Point-slam: Dense neural point cloud-based slam. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023

    2023

  34. [42]

    BAD SLAM: bundle adjusted direct RGB-D SLAM

    Thomas Sch \" o ps, Torsten Sattler, and Marc Pollefeys. BAD SLAM: bundle adjusted direct RGB-D SLAM . In IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2019, Long Beach, CA, USA, June 16-20, 2019 , pp.\ 134--144. Computer Vision Foundation / IEEE , 2019

    2019

  35. [43]

    o rg St \

    David Schubert, Thore Goll, Nikolaus Demmel, Vladyslav Usenko, J \"o rg St \"u ckler, and Daniel Cremers. The tum vi benchmark for evaluating visual-inertial odometry. In 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp.\ 1680--1687, 2018

    2018

  36. [44]

    The proof and measurement of association between two things

    Charles Spearman. The proof and measurement of association between two things. 1961

    1961

  37. [45]

    The replica dataset: A digital replica of indoor spaces

    Julian Straub, Thomas Whelan, Lingni Ma, Yufan Chen, Erik Wijmans, 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

    1906 arXiv

  38. [46]

    A benchmark for the evaluation of rgb-d slam systems

    Jürgen Sturm, Nikolas Engelhard, Felix Endres, Wolfram Burgard, and Daniel Cremers. A benchmark for the evaluation of rgb-d slam systems. In 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp.\ 573--580, 2012

    2012

  39. [47]

    imap: Implicit mapping and positioning in real-time

    Edgar Sucar, Shikun Liu, Joseph Ortiz, and Andrew J Davison. imap: Implicit mapping and positioning in real-time. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pp.\ 6229--6238, 2021

    2021

  40. [48]

    Loftr: Detector-free local feature matching with transformers

    Jiaming Sun, Zehong Shen, Yuang Wang, Hujun Bao, and Xiaowei Zhou. Loftr: Detector-free local feature matching with transformers. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp.\ 8922--8931, 2021

    2021

  41. [49]

    Droid-slam: Deep visual slam for monocular, stereo, and rgb-d cameras

    Zachary Teed and Jia Deng. Droid-slam: Deep visual slam for monocular, stereo, and rgb-d cameras. Advances in neural information processing systems, 34: 0 16558--16569, 2021

    2021

  42. [50]

    Resilient and distributed multi-robot visual slam: Datasets, experiments, and lessons learned

    Yulun Tian, Yun Chang, Long Quang, Arthur Schang, Carlos Nieto-Granda, Jonathan P How, and Luca Carlone. Resilient and distributed multi-robot visual slam: Datasets, experiments, and lessons learned. In 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (...

    2023

  43. [51]

    How nerfs and 3d gaussian splatting are reshaping slam: a survey

    Fabio Tosi, Youmin Zhang, Ziren Gong, Erik Sandstr \"o m, Stefano Mattoccia, Martin R Oswald, and Matteo Poggi. How nerfs and 3d gaussian splatting are reshaping slam: a survey. arXiv preprint arXiv:2402.13255, 4, 2024

    2024 arXiv

  44. [52]

    Scipy 1.0: fundamental algorithms for scientific computing in python

    Pauli Virtanen, Ralf Gommers, Travis E Oliphant, Matt Haberland, Tyler Reddy, David Cournapeau, Evgeni Burovski, Pearu Peterson, Warren Weckesser, Jonathan Bright, et al. Scipy 1.0: fundamental algorithms for scientific computing in python. Nature methods, 17 0 (3): 0 261--272, 2020

    2020

  45. [53]

    Simulating photo-realistic snow and fog on existing images for enhanced cnn training and evaluation

    Alexander Von Bernuth, Georg Volk, and Oliver Bringmann. Simulating photo-realistic snow and fog on existing images for enhanced cnn training and evaluation. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC), pp.\ 41--46. IEEE, 2019

    2019

  46. [54]

    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 Conference on Computer Vision and Pattern Recognition, pp.\ 13293--13302, 2023

    2023

  47. [55]

    Tartanair: A dataset to push the limits of visual slam

    Wenshan Wang, Delong Zhu, Xiangwei Wang, Yaoyu Hu, Yuheng Qiu, Chen Wang, Yafei Hu, Ashish Kapoor, and Sebastian Scherer. Tartanair: A dataset to push the limits of visual slam. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp.\ 4909--4916...

    2020

  48. [56]

    Hero-slam: Hybrid enhanced robust optimization of neural slam

    Zhe Xin, Yufeng Yue, Liangjun Zhang, and Chenming Wu. Hero-slam: Hybrid enhanced robust optimization of neural slam. In 2024 IEEE International Conference on Robotics and Automation (ICRA), pp.\ 8610--8616. IEEE, 2024

    2024

  49. [57]

    Towards robust video object segmentation with adaptive object calibration

    Xiaohao Xu, Jinglu Wang, Xiang Ming, and Yan Lu. Towards robust video object segmentation with adaptive object calibration. In Proceedings of the 30th ACM International Conference on Multimedia, pp.\ 2709--2718, 2022

    2022

  50. [58]

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

    Xingrui Yang, Hai Li, Hongjia Zhai, Yuhang Ming, Yuqian Liu, and Guofeng Zhang. Vox-fusion: Dense tracking and mapping with voxel-based neural implicit representation. In Henry B. L. Duh, Ian Williams, Jens Grubert, J. Adam Jones, and Jianmin Zheng (eds.), IEEE International S...

    2022

  51. [59]

    Precision jumping limits from flight-phase control in salto-1p

    Justin K Yim and Ronald S Fearing. Precision jumping limits from flight-phase control in salto-1p. In 2018 IEEE/RSJ international conference on intelligent robots and systems (IROS), pp.\ 2229--2236. IEEE, 2018

    2018

  52. [60]

    Benchmarking augmentation methods for learning robust navigation agents: the winning entry of the 2021 igibson challenge

    Naoki Yokoyama, Qian Luo, Dhruv Batra, and Sehoon Ha. Benchmarking augmentation methods for learning robust navigation agents: the winning entry of the 2021 igibson challenge. In 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp.\ 1748--1755. ...

    2021

  53. [61]

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

    Youmin Zhang, Fabio Tosi, Stefano Mattoccia, and Matteo Poggi. Go-slam: Global optimization for consistent 3d instant reconstruction. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2023

    2023

  54. [62]

    SubT-MRS dataset: Pushing slam towards all-weather environments

    Shibo Zhao, Yuanjun Gao, Tianhao Wu, Damanpreet Singh, Rushan Jiang, Haoxiang Sun, Mansi Sarawata, Warren C Whittaker, Ian Higgins, Shaoshu Su, Yi Du, Can Xu, John Keller, Jay Karhade, Lucas Nogueira, Sourojit Saha, Yuheng Qiu, Ji Zhang, Wenshan Wang, Chen Wang, and Sebastian ...

    2024

  55. [63]

    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. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 12786--12796, 2022

    2022

  56. [64]

    @esa (Ref

    \@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should ...

  57. [65]

    \@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@firs...

  58. [66]

    New" or

    @open @close @open @close and [1] URL: #1 \@ifundefined chapter * \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibset...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.