Pith. sign in

REVIEW 3 major objections 6 minor 45 references

Sparse-anchor metric depth methods collapse when range sensors return present-but-wrong values; a parameter-free foundation-consistency gate closes the multipath blind spot and cuts KITTI multipath error 3.2× with no retraining.

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 · grok-4.5

2026-07-11 21:40 UTC pith:TCBNTQP5

load-bearing objection Solid diagnosis of a real VI-Depth failure mode plus a cheap, well-controlled fix; immediately usable systems work, not a paradigm shift. the 3 major comments →

arxiv 2607.04101 v1 pith:TCBNTQP5 submitted 2026-07-05 cs.CV

The Multipath Blind Spot: K-Agnostic Robust Calibration for Sparse-Anchor Metric Depth from Frozen Foundations

Sohag Roy , Rajesh Misra , Swami Shastravidyananda , Tamal Maharaj This is my paper
classification cs.CV
keywords monocular depth estimationmetric depthsparse depthrobust estimationdepth foundation modelssensor outlierscalibration
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.

Monocular depth foundations give strong relative scene geometry but no absolute scale. A handful of sparse metric anchors from a range sensor can recover that scale, yet real sensors often return depths that are present with the wrong value—time-of-flight multipath, mixed pixels, max-range dropouts—not merely missing. The paper shows that the field’s residual-on-CFA calibration recipe collapses under these outliers, and that the strongest public method is prepared only for missing anchors, so it falls behind an unprotected baseline on three of four datasets when anchors are present but wrong. The authors introduce MRAC, an inference-time wrapper that tests each anchor against the foundation’s own relative-depth ordering with a Theil–Sen fit and a median-absolute-deviation gate, then feeds only the consistent anchors into a single calibration pass. On a 320-cell benchmark it wins 84% of same-backbone cells across four outlier families and all multipath-corrupted cells against the prior method, at roughly 50 microseconds of CPU cost, no new parameters, and one checkpoint for anchor budgets from 5 to 200.

Core claim

The strongest publicly deployed sparse-anchor calibrator has a structural multipath blind spot: its training prepares it for missing anchors but supplies no mechanism that can reject anchors present with the wrong value. Under multipath-style corruption its accuracy falls behind even an unprotected baseline on three of four datasets. Gating anchors by consistency with the frozen foundation’s relative-depth field—via a Theil–Sen affine fit and a MAD residual test—before a single residual-on-CFA head call restores accuracy across four sensor-grounded outlier families without retraining or added parameters.

What carries the argument

Multipath-Robust Anchor Calibration (MRAC): a parameter-free inference-time wrapper that estimates the affine link between the foundation’s relative depths and the sparse anchors with Theil–Sen, gates anchors by a median-absolute-deviation residual test, and passes only the survivors to one call of the residual-on-CFA calibration head.

Load-bearing premise

Legitimate anchors obey a single global affine relationship with the foundation’s relative-depth field, so residuals cleanly separate clean measurements from present-but-wrong outliers of every family.

What would settle it

On a domain where the foundation’s relative ordering is systematically biased, measure whether MRAC’s MAD gate rejects most clean anchors or accepts multipath ones and underperforms the unprotected residual head on multipath cells; that outcome would refute the foundation-consistency premise.

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

If this is right

  • Robustness to missing anchors and robustness to wrong-valued anchors are distinct problems; training that only drops anchors does not protect against multipath.
  • A single shared residual head plus a ~50 µs CPU gate can serve anchor budgets from 5 to 200 without per-budget checkpoints.
  • The foundation’s relative-depth geometry is a cheap, sufficient witness for rejecting multipath, dropout, mixed-pixel, and uniform sensor outliers.
  • Existing residual-on-CFA pipelines can be made multipath-resistant as a drop-in wrapper with no architectural change and zero retraining.

Where Pith is reading between the lines

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

  • Any anchor-conditioned metric head that never consults the foundation’s relative ordering will inherit the same present-but-wrong blind spot.
  • The same consistency gate can wrap other frozen geometry foundations without retraining those foundations or the gate.
  • Sensor-fusion stacks that treat ToF or LiDAR returns as trustworthy sparse cues will systematically degrade under multipath unless a similar consistency check is added.

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

3 major / 6 minor

Summary. The paper studies sparse-anchor metric calibration of frozen monocular depth foundations under sensor outliers that are present with wrong values (multipath, mixed pixels, dropout, uniform), not merely missing. It shows that residual-on-CFA collapses under such corruption and that VI-Depth, while robust to dropout, has a structural multipath blind spot and falls behind an unprotected baseline on three of four datasets when anchors are present-but-wrong. The proposed MRAC is a parameter-free inference-time wrapper: a Theil–Sen fit plus MAD gate on foundation consistency selects inliers, then a single residual-on-CFA head call produces metric depth. On a 320-cell benchmark with same-backbone controls, MRAC wins 84% of same-backbone cells across four outlier families and, against VI-Depth, all twelve corrupted multipath cells and all sixteen KITTI cells, cutting KITTI multipath AbsRel from 0.489 to 0.151 at ~50 µs CPU overhead and no retraining, while serving K∈[5,200] from one checkpoint.

Significance. If the results hold, the paper makes a clear, practically useful contribution: it isolates a real failure mode of the strongest publicly deployed sparse-anchor calibrator, supplies a systematic outlier-robustness benchmark that the field lacked, and offers a drop-in, K-agnostic, zero-parameter fix with negligible latency. Strengths include same-backbone same-architecture controls (vanilla and B′), three-seed headline bars, explicit Theil–Sen breakdown analysis, gate precision/recall diagnostics, a K=150 re-evaluation of VI-Depth at its training budget, and honest disclosure of clean-cell cost and gate misses. The work is more diagnostic and engineering than algorithmically novel (Theil–Sen+MAD is classical), but the controlled evaluation and structural diagnosis of present-but-wrong vs missing anchors are valuable for metric depth deployment from frozen foundations.

major comments (3)
  1. [Abstract, §V-B, Table II] Abstract and §V-B headline the 3.2× KITTI multipath AbsRel reduction (0.489→0.151) against VI-Depth. On the same cell, unprotected vanilla already reaches 0.170 and B′ 0.164 (Table II), so most of the 3.2× gap is VI-Depth’s collapse relative to a competent residual-on-CFA baseline; MRAC’s same-backbone incremental gain is modest (~0.170→0.151). The structural blind-spot claim is well supported, but the abstract/intro should more carefully separate (i) VI-Depth worse than unprotected baseline from (ii) MRAC’s incremental same-backbone gain, so the wrapper’s contribution is not overstated relative to simply avoiding VI-Depth on multipath.
  2. [§III-D, §IV-E, §V] The entire 320-cell robustness grid (§IV-E, §V) injects synthetic outliers (uniform, near, dropout, mixed-pixel). The motivation and title center on real ToF multipath and related sensor faults, yet no experiment uses real multipath-corrupted range measurements. Synthetic families are sensor-grounded and the structural diagnosis of VI-Depth does not require real data, but transfer risk (outlier magnitude, spatial correlation, depth-range dependence) remains untested. A short real-sensor case study, or at least an explicit discussion of which multipath statistics the synthetic near model does and does not capture, is needed for the applied claim.
  3. [§III-E–F, §V-F, §IX-B, Fig. 5] On the headline KITTI near/25% cell, MAD-gate precision/recall is only 0.60/0.64 (§IX-B, Fig. 5, Table SII): ~40% of clean anchors are falsely rejected and ~36% of multipath outliers pass. The residual head R_θ is trained on clean anchors with K∈{5,10,25,50} and no outlier injection (§III-F). False rejection therefore changes effective K and spatial support relative to training, which may explain residual losses (e.g., KITTI near/40% where vanilla 0.275 edges MRAC 0.299; §V-F). The paper should analyze residual-head behavior under gated (sparser/biased) anchors and whether false rejects, not only missed outliers, drive the remaining same-backbone gaps.
minor comments (6)
  1. [Fig. 3, §V-C] Fig. 3 averages AbsRel over four datasets; shaded bands are ±1 std across datasets, which can mask per-dataset inversions (e.g., DIODE clean wins for VI-Depth). Consider per-dataset curves in the supplement or a note that the average is for trend only.
  2. [§VI-B, Table IV] Clean AbsRel for residual at K=50 is reported as 0.099 in the K-sweep (§VI-B) and 0.104 in the outlier harness (Table IV); the footnote attributes this to independent anchor draws. State the harness seed protocol once in §IV-E so readers do not treat the two as contradictory.
  3. [§IV-B, §IX-B] VI-Depth’s strongest published backbone (dpt-beit-large-512) was unreachable (§IV-B, §IX-B). A one-sentence note on whether the public swin2 config is the recommended deployment setting would help readers interpret the cross-backbone comparison.
  4. [§III-E, Supplement Table SI] κ sensitivity is swept and flat (§III-E, Table SI); consider stating in the main text that κ=1.5 is marginally better on most cells so practitioners know the default is slightly conservative for precision.
  5. [§II-C, §III-D] Related work on robust regression (§II-C) is appropriate; a brief pointer to robust depth-completion or multipath-compensation literature beyond the cited ToF surveys would better situate the sensor-outlier families.
  6. [Algorithm 1, §III-E] Algorithm 1 is clear; explicitly note that CFA is recomputed on M inside the single R_θ call (line 8) so readers do not assume the original a_cfa, b_cfa are reused.

Circularity Check

0 steps flagged

No significant circularity: empirical method paper whose claims are measured wins on external benchmarks, not reductions of fitted inputs or self-definitional derivations.

full rationale

The paper's load-bearing claims are empirical (residual-on-CFA collapses under present-but-wrong outliers; VI-Depth has a multipath blind spot; MRAC wins 84% of same-backbone cells and all twelve corrupted multipath cells, cutting KITTI multipath AbsRel 3.2 imes). These are established by a 320-cell grid on public datasets (NYUv2, KITTI, DIODE, SUN RGB-D) against same-backbone controls and the public VI-Depth checkpoint, not by deriving a quantity that equals its inputs by construction. The method itself (Theil–Sen slope + MAD gate on foundation consistency, then one residual-on-CFA call) is a standard robust-regression wrapper; κ=2 is the conventional two-MAD threshold whose flat sensitivity is tabulated, and training uses only clean anchors so robustness is not injected by construction. Citations (Theil–Sen, RANSAC, VI-Depth, Depth Anything V2, etc.) are external prior art, not self-citations of uniqueness theorems or ansätze by the present authors. No equation reduces a claimed prediction to a fitted parameter, and no uniqueness or first-principles result is imported from overlapping authors. The derivation chain is therefore self-contained and non-circular.

Axiom & Free-Parameter Ledger

1 free parameters · 3 axioms · 0 invented entities

The central claim rests on standard robust-statistics tools, the established residual-on-CFA substrate, and one domain premise (foundation consistency) plus a single conventional threshold. No new physical entities are postulated; free parameters are minimal and swept.

free parameters (1)
  • κ (MAD multiplier) = 2
    Single deployment knob set to the conventional value 2; sensitivity sweep over {1.5,2.0,2.5,3.0} shows AbsRel variation ≤0.011 on six of eight headline cells.
axioms (3)
  • domain assumption Legitimate anchors satisfy a single global affine relationship with the frozen foundation’s relative-depth field; outliers of every family violate it (foundation consistency).
    Stated as the operating principle of Sec. III-E; load-bearing for the claim that one gate handles multipath, dropout, mixed-pixel and uniform outliers without type-specific tuning.
  • standard math Theil–Sen estimator has asymptotic breakdown point ≈29% and resists correlated dropout better than RANSAC.
    Invoked in Sec. III-E and V-A; classical result of Theil (1950) and Sen (1968).
  • domain assumption Residual-on-CFA (global closed-form affine + lightweight residual head) is the correct calibration substrate to wrap.
    Adopted unchanged from VI-Depth and Marsal et al.; justified by zero-shot and K-sweep studies in Sec. VI.

pith-pipeline@v1.1.0-grok45 · 47407 in / 2718 out tokens · 25439 ms · 2026-07-11T21:40:40.480027+00:00 · methodology

0 comments
read the original abstract

Monocular depth foundations predict domain-general relative depth but lack absolute scale; a handful of sparse metric anchors from a range sensor can calibrate them to metric depth, an attractive alternative to metric-supervised training. Existing sparse-anchor calibration methods, however, assume the anchors are clean, whereas real sensors produce outliers that are present with the wrong value -- time-of-flight multipath, mixed pixels -- not merely missing. We show that the established residual-on-CFA calibration recipe collapses under such outliers, and that the strongest publicly deployed method, VI-Depth, has a structural multipath blind spot: robust to missing anchors, it falls behind an unprotected baseline on three of four datasets when anchors are present but wrong. We propose Multipath-Robust Anchor Calibration (MRAC), a parameter-free, inference-time wrapper that gates anchors by foundation consistency -- a Theil--Sen fit and a median-absolute-deviation test against the foundation's own relative-depth ordering -- before a single call to the calibration head. MRAC adds no learned parameters, runs its selection in $\approx 50\,\mu$s on CPU, and serves anchor budgets $K \in [5,200]$ from one checkpoint. On a $320$-cell benchmark with a same-backbone, same-architecture control, MRAC strictly wins $84\%$ of same-backbone cells across all four outlier families and, against VI-Depth, wins all twelve corrupted multipath cells and all sixteen KITTI cells, reducing KITTI multipath AbsRel by $3.2\times$ ($0.489$ to $0.151$) at zero retraining.

Figures

Figures reproduced from arXiv: 2607.04101 by Rajesh Misra, Sohag Roy, Swami Shastravidyananda, Tamal Maharaj.

Figure 1
Figure 1. Figure 1: Sparse-anchor calibration under multipath outliers. The strongest deployed method, VI-Depth, is robust to missing anchors but blind to present-but [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: MRAC pipeline. A frozen foundation F (DAv2, ∼330 M parameters) predicts a relative-depth field drel; the K sparse anchors are tested for consistency with that field by a Theil–Sen slope fit and a MAD inlier gate (|rk| ≤ κ σMAD); only the surviving cleaned anchors are passed to a single call of the residual-on-CFA head Rθ (0.47 M parameters), which produces the metric depth map. The robust selection adds no… view at source ↗
Figure 3
Figure 3. Figure 3: AbsRel versus outlier fraction p for the four outlier families, averaged over NYUv2, KITTI, DIODE, and SUN RGB-D. Four methods: vanilla (DAv2), B ′ (DAv2), MRAC on the B′ head (ours), and VI-Depth official. MRAC remains flat where the baselines diverge on dropout and where VI-Depth’s curve climbs on near (multipath). Shaded bands show ±1 standard deviation across the four datasets. TABLE III STRICT PER-CEL… view at source ↗
Figure 4
Figure 4. Figure 4: AbsRel versus anchor budget K on NYUv2 for three architectures. Residual-on-CFA dominates for K ≥ 5; K=1 is the architectural degenerate point (the closed-form fit needs K ≥ 2); the curve is U-shaped with a sweet spot at K∗=50. do not average out, and a single one can dominate a least￾squares fit (Sec. III-B). This is precisely why Gaussian-noise robustness is cheap and outlier robustness is not, and why t… view at source ↗
Figure 5
Figure 5. Figure 5: Mechanism of the multipath blind spot on three KITTI frames under [PITH_FULL_IMAGE:figures/full_fig_p011_5.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

45 extracted references · 25 linked inside Pith

  1. [1]

    Towards robust monocular depth estimation: Mixing datasets for zero-shot cross-dataset transfer,

    R. Ranftl, K. Lasinger, D. Hafner, K. Schindler, and V . Koltun, “Towards robust monocular depth estimation: Mixing datasets for zero-shot cross-dataset transfer,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 3, pp. 1623–1637, 2022, arXiv:1907.01341

    Pith/arXiv arXiv 2022

  2. [2]

    Vision transformers for dense pre- diction,

    R. Ranftl, A. Bochkovskiy, and V . Koltun, “Vision transformers for dense pre- diction,” inProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021, pp. 12 179–12 188, arXiv:2103.13413

    Pith/arXiv arXiv 2021

  3. [3]

    Depth anything v2,

    L. Yang, B. Kang, Z. Huang, Z. Zhao, X. Xu, J. Feng, and H. Zhao, “Depth anything v2,” inAdvances in Neural Information Processing Systems (NeurIPS), 2024, arXiv:2406.09414

    Pith/arXiv arXiv 2024

  4. [4]

    Monocular visual-inertial depth estimation,

    D. Wofk, R. Ranftl, M. M ¨uller, and V . Koltun, “Monocular visual-inertial depth estimation,” inProceedings of the IEEE International Conference on Robotics and Automation (ICRA), 2023, pp. 6095–6101, arXiv:2303.12134

    Pith/arXiv arXiv 2023

  5. [5]

    Recovering dense metric depth in indoor scenes from monocular depth foundation models and 2D LiDARs,

    R. Marsal, A. Chapoutot, P. Xu, and D. Filliat, “Recovering dense metric depth in indoor scenes from monocular depth foundation models and 2D LiDARs,” inEuropean Robotics Forum 2025 (ERF), Springer Proceedings in Advanced Robotics. Springer, 2025, vol. 36, pp. 236–241

    2025

  6. [6]

    Prompting depth anything for 4k resolution accurate metric depth estimation,

    H. Lin, S. Peng, J. Chen, S. Peng, J. Sun, M. Liu, H. Bao, J. Feng, X. Zhou, and B. Kang, “Prompting depth anything for 4k resolution accurate metric depth estimation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025, pp. 17 070–17 080, arXiv:2412.14015

    arXiv 2025

  7. [7]

    Depth anything with any prior,

    Z. Wang, S. Chen, L. Yang, J. Wang, Z. Zhang, H. Zhao, and Z. Zhao, “Depth anything with any prior,”arXiv preprint arXiv:2505.10565, 2025

    Pith/arXiv arXiv 2025

  8. [8]

    Sparse-lidar prompting of monocular geometry foundations: An empirical study toward long-range driving depth,

    K. Zheng, Q. Feng, X. Liu, W. Tan, and Y . Li, “Sparse-lidar prompting of monocular geometry foundations: An empirical study toward long-range driving depth,”arXiv preprint arXiv:2605.26456, 2026

    Pith/arXiv arXiv 2026

  9. [9]

    Depth anything: Unleashing the power of large-scale unlabeled data,

    L. Yang, B. Kang, Z. Huang, X. Xu, J. Feng, and H. Zhao, “Depth anything: Unleashing the power of large-scale unlabeled data,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 10 371–10 381, arXiv:2401.10891

    Pith/arXiv arXiv 2024

  10. [10]

    Repurposing diffusion-based image generators for monocular depth estimation,

    B. Ke, A. Obukhov, S. Huang, N. Metzger, R. C. Daudt, and K. Schindler, “Repurposing diffusion-based image generators for monocular depth estimation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 9492–9502, arXiv:2312.02145

    Pith/arXiv arXiv 2024

  11. [11]

    Moge-2: Accurate monocular geometry with metric scale and sharp details,

    R. Wang, S. Xu, Y . Dong, Y . Deng, J. Xiang, Z. Lv, G. Sun, X. Tong, and J. Yang, “Moge-2: Accurate monocular geometry with metric scale and sharp details,” inAdvances in Neural Information Processing Systems (NeurIPS), 2025, arXiv:2507.02546

    Pith/arXiv arXiv 2025

  12. [12]

    Depth pro: Sharp monocular metric depth in less than a second,

    A. Bochkovskii, A. Delaunoy, H. Germain, M. Santos, Y . Zhou, S. R. Richter, and V . Koltun, “Depth pro: Sharp monocular metric depth in less than a second,” arXiv preprint arXiv:2410.02073, 2024

    Pith/arXiv arXiv 2024

  13. [13]

    Metric3d: Towards zero-shot metric 3d prediction from a single image,

    W. Yin, C. Zhang, H. Chen, Z. Cai, G. Yu, K. Wang, X. Chen, and C. Shen, “Metric3d: Towards zero-shot metric 3d prediction from a single image,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023, pp. 9043–9053, arXiv:2307.10984

    Pith/arXiv arXiv 2023

  14. [14]

    Depth map prediction from a single image using a multi-scale deep network,

    D. Eigen, C. Puhrsch, and R. Fergus, “Depth map prediction from a single image using a multi-scale deep network,” inAdvances in Neural Information Processing Systems (NeurIPS), 2014, arXiv:1406.2283

    Pith/arXiv arXiv 2014

  15. [15]

    Deep ordinal regression network for monocular depth estimation,

    H. Fu, M. Gong, C. Wang, K. Batmanghelich, and D. Tao, “Deep ordinal regression network for monocular depth estimation,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 2002–2011, arXiv:1806.02446

    Pith/arXiv arXiv 2018

  16. [16]

    From big to small: Multi- scale local planar guidance for monocular depth estimation,

    J. H. Lee, M.-K. Han, D. W. Ko, and I. H. Suh, “From big to small: Multi- scale local planar guidance for monocular depth estimation,”arXiv preprint arXiv:1907.10326, 2019

    Pith/arXiv arXiv 1907

  17. [17]

    Adabins: Depth estimation using adaptive bins,

    S. F. Bhat, I. Alhashim, and P. Wonka, “Adabins: Depth estimation using adaptive bins,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 4009–4018, arXiv:2011.14141

    Pith/arXiv arXiv 2021

  18. [18]

    Neural window fully-connected crfs for monocular depth estimation,

    W. Yuan, X. Gu, Z. Dai, S. Zhu, and P. Tan, “Neural window fully-connected crfs for monocular depth estimation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 3916–3925, arXiv:2203.01502

    Pith/arXiv arXiv 2022

  19. [19]

    Zoedepth: Zero-shot transfer by combining relative and metric depth,

    S. F. Bhat, R. Birkl, D. Wofk, P. Wonka, and M. M ¨uller, “Zoedepth: Zero-shot transfer by combining relative and metric depth,”arXiv preprint arXiv:2302.12288, 2023

    Pith/arXiv arXiv 2023

  20. [20]

    A rank-invariant method of linear and polynomial regression analysis,

    H. Theil, “A rank-invariant method of linear and polynomial regression analysis,” Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen, Series A, vol. 53, pp. 386–392, 521–525, 1397–1412, 1950

    1950

  21. [21]

    Estimates of the regression coefficient based on kendall’s tau,

    P. K. Sen, “Estimates of the regression coefficient based on kendall’s tau,”Journal of the American Statistical Association, vol. 63, no. 324, pp. 1379–1389, 1968

    1968

  22. [22]

    Least median of squares regression,

    P. J. Rousseeuw, “Least median of squares regression,”Journal of the American Statistical Association, vol. 79, no. 388, pp. 871–880, 1984

    1984

  23. [23]

    Robust estimation of a location parameter,

    P. J. Huber, “Robust estimation of a location parameter,”The Annals of Mathe- matical Statistics, vol. 35, no. 1, pp. 73–101, 1964

    1964

  24. [24]

    Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography,

    M. A. Fischler and R. C. Bolles, “Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography,” Communications of the ACM, vol. 24, no. 6, pp. 381–395, 1981

    1981

  25. [25]

    Mlesac: A new robust estimator with application to estimating image geometry,

    P. H. S. Torr and A. Zisserman, “Mlesac: A new robust estimator with application to estimating image geometry,”Computer Vision and Image Understanding, vol. 78, no. 1, pp. 138–156, 2000

    2000

  26. [26]

    Hartley and A

    R. Hartley and A. Zisserman,Multiple View Geometry in Computer Vision, 2nd ed. Cambridge University Press, 2004

    2004

  27. [27]

    Sparse-to-dense: Depth prediction from sparse depth samples and a single image,

    F. Ma and S. Karaman, “Sparse-to-dense: Depth prediction from sparse depth samples and a single image,” inProceedings of the IEEE International Conference on Robotics and Automation (ICRA), 2018, pp. 4796–4803, arXiv:1709.07492

    Pith/arXiv arXiv 2018

  28. [28]

    Self-supervised sparse-to-dense: Self- supervised depth completion from lidar and monocular camera,

    F. Ma, G. V . Cavalheiro, and S. Karaman, “Self-supervised sparse-to-dense: Self- supervised depth completion from lidar and monocular camera,” inProceedings of the IEEE International Conference on Robotics and Automation (ICRA), 2019, pp. 3288–3295

    2019

  29. [29]

    Learning depth with convolutional spatial propagation network,

    X. Cheng, P. Wang, and R. Yang, “Learning depth with convolutional spatial propagation network,” inProceedings of the European Conference on Computer Vision (ECCV), 2018, pp. 103–119, arXiv:1808.00150

    Pith/arXiv arXiv 2018

  30. [30]

    Non-local spatial propagation network for depth completion,

    J. Park, K. Joo, Z. Hu, C.-K. Liu, and I. S. Kweon, “Non-local spatial propagation network for depth completion,” inProceedings of the European Conference on Computer Vision (ECCV), 2020, pp. 120–136, arXiv:2007.10042

    Pith/arXiv arXiv 2020

  31. [31]

    Learning guided convolutional network for depth completion,

    J. Tang, F.-P. Tian, W. Feng, J. Li, and P. Tan, “Learning guided convolutional network for depth completion,”IEEE Transactions on Image Processing, vol. 30, pp. 1116–1129, 2021

    2021

  32. [32]

    Penet: Towards precise and efficient image guided depth completion,

    M. Hu, S. Wang, B. Li, S. Ning, L. Fan, and X. Gong, “Penet: Towards precise and efficient image guided depth completion,” inProceedings of the IEEE International Conference on Robotics and Automation (ICRA), 2021, pp. 13 656–13 662

    2021

  33. [33]

    Com- pletionformer: Depth completion with convolutions and vision transformers,

    Y . Zhang, X. Guo, M. Poggi, Z. Zhu, G. Huang, and S. Mattoccia, “Com- pletionformer: Depth completion with convolutions and vision transformers,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, arXiv:2304.13030

    Pith/arXiv arXiv 2023

  34. [34]

    Bilateral propagation network for depth completion,

    J. Tang, F.-P. Tian, B. An, J. Li, and P. Tan, “Bilateral propagation network for depth completion,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 9763–9772, arXiv:2403.11270

    Pith/arXiv arXiv 2024

  35. [35]

    Unsupervised depth completion with calibrated back- projection layers,

    A. Wong and S. Soatto, “Unsupervised depth completion with calibrated back- projection layers,” inProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021, pp. 12 747–12 756, arXiv:2108.10531

    Pith/arXiv arXiv 2021

  36. [36]

    Analysis and removal of artifacts in 3-D LADAR data,

    J. Tuley, N. Vandapel, and M. Hebert, “Analysis and removal of artifacts in 3-D LADAR data,” inProceedings of the IEEE International Conference on Robotics and Automation (ICRA), 2005, pp. 2203–2210

    2005

  37. [37]

    3D measurements from imaging laser radars: How good are they?

    M. Hebert and E. Krotkov, “3D measurements from imaging laser radars: How good are they?”Image and Vision Computing, vol. 10, no. 3, pp. 170–178, 1992

    1992

  38. [38]

    Multipath interference compensation in time-of-flight camera images,

    S. Fuchs, “Multipath interference compensation in time-of-flight camera images,” inProceedings of the International Conference on Pattern Recognition (ICPR), 2010, pp. 3583–3586

    2010

  39. [39]

    Hansard, S

    M. Hansard, S. Lee, O. Choi, and R. Horaud,Time-of-Flight Cameras: Principles, Methods and Applications, ser. SpringerBriefs in Computer Science. Springer, 2013

    2013

  40. [40]

    Understanding and ameliorating mixed pixels and multipath interference in AMCW lidar,

    J. P. Godbaz, A. A. Dorrington, and M. J. Cree, “Understanding and ameliorating mixed pixels and multipath interference in AMCW lidar,” inTOF Range-Imaging Cameras. Springer, 2013, pp. 91–116

    2013

  41. [41]

    Indoor segmentation and support inference from rgbd images,

    N. Silberman, D. Hoiem, P. Kohli, and R. Fergus, “Indoor segmentation and support inference from rgbd images,” inProceedings of the European Conference on Computer Vision (ECCV), 2012, pp. 746–760

    2012

  42. [42]

    Are we ready for autonomous driving? the kitti vision benchmark suite,

    A. Geiger, P. Lenz, and R. Urtasun, “Are we ready for autonomous driving? the kitti vision benchmark suite,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2012, pp. 3354–3361

    2012

  43. [43]

    Unsupervised CNN for single view depth estimation: Geometry to the rescue,

    R. Garg, V . K. B G, G. Carneiro, and I. Reid, “Unsupervised CNN for single view depth estimation: Geometry to the rescue,” inProceedings of the European Conference on Computer Vision (ECCV), 2016, pp. 740–756, arXiv:1603.04992

    Pith/arXiv arXiv 2016

  44. [44]

    Diode: A dense indoor and outdoor depth dataset,

    I. Vasiljevic, N. Kolkin, S. Zhang, R. Luo, H. Wang, F. Z. Dai, A. F. Daniele, M. Mostajabi, S. Basart, M. R. Walter, and G. Shakhnarovich, “Diode: A dense indoor and outdoor depth dataset,”arXiv preprint arXiv:1908.00463, 2019

    Pith/arXiv arXiv 1908

  45. [45]

    Sun rgb-d: A rgb-d scene understanding benchmark suite,

    S. Song, S. P. Lichtenberg, and J. Xiao, “Sun rgb-d: A rgb-d scene understanding benchmark suite,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp. 567–576. UNDER REVIEW AT IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 13 Supplementary Material I. SENSITIVITY TO THEMADMULTIPLIERκ Section II...

    arXiv 2015