Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

RePoseD: Efficient Relative Pose Estimation With Known Depth Information

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 ↗

Pith's one-line read This paper claims that monocular depth predictions, modeled as affine-unknown per image, reduce calibrated relative pose estimation to solving a quartic and yield faster, more accurate solvers.

desk verdict Solid minimal-solver paper with real new variants; the inverse-depth mismatch is a fixable soft spot, not a fatal flaw. read the letter →

arxiv 2501.07742 v3 pith:XJ2U2GXX submitted 2025-01-13 cs.CV

Yaqing Ding , Viktor Kocur , Václav Vávra , Zuzana Berger Haladová , Jian Yang , Torsten Sattler , Zuzana Kukelova This is my paper · ORCID
classification cs.CV
keywords relativeposeestimationmonoculardepthminimalsolversaffine-invariantscaleandshiftfocallengthRANSACcamerageometry
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper asks whether monocular depth estimates from modern networks actually improve two-view relative pose estimation over classical point-only methods. It answers yes, provided the depth is modeled as defined up to an unknown scale and shift per image, and it derives minimal solvers that jointly estimate those depth parameters with the pose. For calibrated cameras the problem reduces to a single quartic equation; for cameras with unknown shared or different focal lengths it reduces to small polynomial systems solved by Groebner basis elimination. On real datasets, the new solvers are faster and often more accurate than prior depth-aware solvers, and the scale-only variants can match or beat shift-modeling variants when the depth source is reliable.

What carries the argument

The load-bearing mechanism is distance-constraint elimination. Instead of solving directly for rotation and translation, the paper lifts each correspondence to a 3D point via its monocular depth and writes the unknown relative scale $s$ and shifts $u,v$ into the 3D coordinates. Pairwise differences cancel translation, and rotation is cancelled by observing that norms are rotation-invariant, leaving polynomial equations in $s,u,v$ (and focal lengths when they are unknown). For calibrated cameras, the substitution $c=s^2$ and Gauss-Jordan elimination express $\{cv^2, cv, c\}$ as quadratics in $u$; the consistency identity $g_2^2=g_1 g_3$ produces the quartic. This is the step that makes the solver minimal, closed-form, and fast.

What would settle it

Run the calibrated 3-point affine-invariant solver on real image pairs whose depth maps come from an estimator known to produce per-pixel scale variation or inverse-depth outputs, and compare median rotation and translation error against the 5-point point-only solver on the same matches; if the depth-aware solver is not consistently more accurate when depth is reliable, the central claim that affine-invariant monocular depth improves relative pose fails.

Watch

Extended reading notes

Core claim

The core discovery is an algebraic reduction: substituting the affine depth model $\eta_i=s_1(\alpha_i+u)$, $\lambda_i=s_2(\beta_i+v)$ into the projection equation, dividing by $s_1$, and subtracting equations eliminates translation, while taking norms of both sides eliminates rotation. The resulting pairwise-distance constraints contain only the relative scale $s$, the two shifts $u,v$ (or a subset), and the focal lengths. After Gauss-Jordan elimination, the calibrated case collapses to the identity $(cv)^2 = c(cv^2)$, which yields a quartic in $u$ that can be solved in closed form; the focal-length cases become small Groebner-basis solvers. The paper claims these solvers outperform the concurrent depth-aware solvers in speed and accuracy, and that in many realistic settings a scale-only solver is enough.

Load-bearing premise

The load-bearing premise is that, in each image, the true depth is a global affine function of the estimated depth with a single unknown scale and a single unknown shift; if a depth estimator violates this—for instance by returning inverse depth or varying its scale across the image—the solver's estimates can be biased.

Editorial extensions

If this is right

  • When monocular depths are accurate, the calibrated 3-point affine-invariant solver is competitive with or better than the 5-point point-only solver while being faster per RANSAC iteration.
  • Modeling an unknown shift is not always necessary: the scale-only 3-point solvers often match or beat shift-modeling solvers when depth estimates are good, contradicting the concurrent claim that shift modeling is always beneficial.
  • The proposed focal-length solvers recover both relative pose and unknown shared or different focal lengths from four depth-tagged correspondences, with smaller elimination templates than prior work.
  • An inverse-depth variant exists but is slower and less accurate inside RANSAC, so the paper does not recommend it for current affine-invariant inverse-depth networks.
  • Across several datasets and matchers, depth-aware solvers improve over point-only baselines mainly when depth is reliable; with poor depth or very strong matches, point-only solvers can still win.

Reading between the lines

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

  • Beyond the paper, if the affine-per-image depth model is violated—for example, by estimators that return inverse depth or that apply a different scale to different image regions—the quartic reduction no longer holds exactly; a testable extension would be a solver that estimates per-correspondence scale weights or fits the affine model robustly inside RANSAC.
  • Beyond the paper, the reported speed gain for the scale-only variant suggests a practical rule for pipeline designers: prefer the cheaper scale-only solver when the depth source is metric or reliably scale-calibrated, and reserve shift modeling for affine-invariant networks.
  • Beyond the paper, the same elimination-by-distance idea could be adapted to other two-view problems with depth priors, such as absolute pose with known scene depth or multi-camera rigs, though the paper does not test these settings.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes a family of minimal solvers for relative pose estimation from point correspondences augmented with monocular depth estimates. The key modeling assumption (Eq. 2) is that the true depth is a global affine function of the MDE output within each image, with one unknown scale and shift per image; the paper derives calibrated, shared-focal-length, and different-focal-length solvers that jointly estimate the relative pose with these depth parameters, using distance constraints to eliminate rotation and translation. The main calibrated solver (3PTsuv) reduces to a quartic, while the focal-length solvers (4PTsuvf, 4PTsuvf1,2) use Gröbner-basis or eigenvalue techniques. Experiments on ETH3D, Phototourism, and ScanNet with several MDEs, matchers, and RANSAC protocols compare the proposed solvers against prior depth-aware solvers (Madpose, Rel3PT, 3p3d, 4p4d) and classical point-based solvers, and the paper concludes that depth-aware solvers can outperform point-based methods when depth estimates are reliable, with scale-only solvers often matching or exceeding shift-modeling solvers.

Significance. If the results are sound, this is a useful contribution: it provides faster minimal solvers than the concurrent Madpose work, gives a systematic comparison across depth models and RANSAC variants, and offers evidence that modeling unknown shift is not always necessary. The algebraic derivations are plausible and the experimental scope is broad. The paper's main limitation is that the central affine-depth model is mismatched for inverse-depth MDEs used in the main experiments, and the experiments lack statistical validation of the small performance differences that support the headline claims.

major comments (3)
  1. [Section 3, Eq. (2); SM Sec. 6.2, Table 6]
  2. [Tables 3-5; SM Tables 8-15]
  3. [Section 4.3, Figure 3]
minor comments (5)
  1. [Section 1, Figure 1]
  2. [Table 1]
  3. [Section 3.1]
  4. [Section 3.2]
  5. [SM Sec. 6.2]

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the scale-shift-aware solvers are derived self-contained from an explicit depth model, and the empirical claims are evaluated against independent ground truth.

full rationale

RePoseD's central derivation is self-contained rather than circular. The calibrated solver starts from Eq. (4), the explicit affine-depth model eta_i = s1(alpha_i + u), lambda_i = s2(beta_i + v), substitutes it into the projection constraint, and eliminates rotation and translation via the Euclidean-distance identities ||s(beta_i+v) q~_i - s(beta_j+v) q~_j|| = ||(alpha_i+u) p~_i - (alpha_j+u) p~_j||. After Gauss-Jordan elimination, the monomials {c v^2, c v, c} are expressed as quadratics in u, and the algebraic identity (c v)^2 = c (c v^2) produces the quartic. This is standard elimination, not a reuse of outputs as inputs. The scale and shifts are unknowns of the same minimal system, solved anew for each correspondence sample; they are not constants fitted to benchmark ground truth, so no fitted input is later relabeled as a prediction. The experimental section compares against external datasets (Phototourism, ETH3D, ScanNet) with independent ground-truth poses and depths, so the claimed speed and accuracy advantages are empirically supported. The paper does cite prior work by overlapping authors, including [1], [15], [17], [35], and [38], but those citations are used as baseline solvers or as standard algebraic machinery; the quartic derivation does not import any unverified uniqueness theorem or ansatz from them. The affine model of Eq. (2) is an explicitly stated modeling assumption; whether it holds for inverse-depth MDEs such as Depth Anything v2 is a robustness and correctness concern, and the supplementary material even derives and tests an inverse-depth solver, which further shows that the modeling choice is not definitionally protected. The stated limitations about degenerate configurations and missing physical-feasibility checks for focal lengths are honest caveats and do not indicate circularity. Overall, no load-bearing step reduces by construction to its own inputs.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central model introduces two nuisance parameters per image (scale and shift) that are estimated from correspondences. No new physical entities are postulated. The key domain axiom is the global affine depth relation, which is the most fragile part of the modeling.

free parameters (2)
  • s (relative depth scale) = unknown per image pair
    Introduced in Eq. (4) to handle scale-invariant monocular depth. Estimated jointly with pose in every solver; central to all configurations.
  • u and v (depth shifts) = unknown per image pair
    Added in Eq. (2) for affine-invariant depth. Estimated jointly with pose; without modeling shifts the solver is biased on shifted depth maps.
assumptions (4)
  • domain assumption True depth equals a global affine function of estimated depth per image (Eq. 2)
    All affine-invariant solvers rest on this. MDE outputs may violate it, especially inverse-depth networks.
  • domain assumption Pinhole camera model with square pixels and principal point at image center for focal-length solvers
    Used in Section 3.2 to reduce unknown intrinsics to focal lengths f, f1, f2.
  • standard math Rotation preserves vector norms
    Used in Section 3.1 to eliminate R and obtain polynomial equations in s, u, v.
  • standard math Three or four point correspondences provide algebraically independent constraints
    Assumes generic positions, no degeneracies such as coplanar points or pure translation, which the paper notes as limitations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of RePoseD: Efficient Relative Pose Estimation With Known Depth Information." pith.science (2026). https://pith.science/paper/XJ2U2GXX

@misc{pith2026250107742,
  author       = {Pith},
  title        = {Pith review of: RePoseD: Efficient Relative Pose Estimation With Known Depth Information},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XJ2U2GXX}},
  note         = {Machine review of arXiv:2501.07742}
}
read the original abstract

Recent advances in monocular depth estimation methods (MDE) and their improved accuracy open new possibilities for their applications. In this paper, we investigate how monocular depth estimates can be used for relative pose estimation. In particular, we are interested in answering the question whether using MDEs improves results over traditional point-based methods. We propose a novel framework for estimating the relative pose of two cameras from point correspondences with associated monocular depths. Since depth predictions are typically defined up to an unknown scale or even both unknown scale and shift parameters, our solvers jointly estimate the scale or both the scale and shift parameters along with the relative pose. We derive efficient solvers considering different types of depths for three camera configurations: (1) two calibrated cameras, (2) two cameras with an unknown shared focal length, and (3) two cameras with unknown different focal lengths. Our new solvers outperform state-of-the-art depth-aware solvers in terms of speed and accuracy. In extensive real experiments on multiple datasets and with various MDEs, we discuss which depth-aware solvers are preferable in which situation. The code will be made publicly available.

Figures

Figures reproduced from arXiv: 2501.07742 by the authors.

Figure 1
Figure 1. Given point matches and their corresponding depths [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Speed accuracy evaluation for the case of two cameras [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Speed accuracy evaluation for different unknown focal [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Audio-Visual Camera Pose Estimation with Passive Scene Sounds and In-the-Wild Video

    cs.CV 2025-12 unverdicted novelty 8.0 of 10

    Passive scene audio, especially direction-of-arrival cues, improves relative camera pose estimation when combined with vision in real-world videos.

Reference graph

Works this paper leans on

73 extracted references · 65 canonical work pages · cited by 1 Pith paper

  1. [1]

    Fast relative pose estimation using relative depth

    Jonathan Astermark, Yaqing Ding, Viktor Larsson, and An- ders Heyden. Fast relative pose estimation using relative depth. In International Conference on 3D Vision (3DV) ,

    work page

  2. [2]

    Relative pose from sift features

    Daniel Barath and Zuzana Kukelova. Relative pose from sift features. In European Conference on Computer Vision, pages 454–469. Springer, 2022. 1

    work page 2022

  3. [3]

    Graph-cut RANSAC

    Daniel Barath and Ji ˇr´ı Matas. Graph-cut RANSAC. In Com- puter Vision and Pattern Recognition (CVPR), 2018. 2, 3

    work page 2018

  4. [4]

    Relative pose solvers us- ing monocular depth

    Daniel Barath and Chris Sweeney. Relative pose solvers us- ing monocular depth. In 2022 26th International Conference on Pattern Recognition (ICPR), 2022. 2

    work page 2022

  5. [5]

    A mini- mal solution for two-view focal-length estimation using two affine correspondences

    Daniel Barath, Tekla Toth, and Levente Hajder. A mini- mal solution for two-view focal-length estimation using two affine correspondences. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, 2017. 1

    work page 2017

  6. [6]

    Conic epipolar con- straints from affine correspondences

    Jacob Bentolila and Joseph M Francos. Conic epipolar con- straints from affine correspondences. Computer Vision and Image Understanding, 2014. 1

    work page 2014

  7. [7]

    Zoedepth: Zero-shot trans- fer by combining relative and metric depth

    Shariq Farooq Bhat, Reiner Birkl, Diana Wofk, Peter Wonka, and Matthias M ¨uller. Zoedepth: Zero-shot trans- fer by combining relative and metric depth. arXiv preprint arXiv:2302.12288, 2023. 2, 3

    arXiv 2023

  8. [8]

    Midas v3

    Reiner Birkl, Diana Wofk, and Matthias M ¨uller. Midas v3. 1–a model zoo for robust monocular relative depth estima- tion. arXiv preprint arXiv:2307.14460, 2023. 2, 3, 6, 7, 8, 4, 5

    arXiv 2023

Show all 73 references
  1. [9]

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

    Aleksei Bochkovskii, Ama ¨el Delaunoy, Hugo Germain, Marcel Santos, Yichao Zhou, Stephan R Richter, and Vladlen Koltun. Depth pro: Sharp monocular metric depth in less than a second. arXiv preprint arXiv:2410.02073, 2024. 1, 2, 3

    2024 arXiv

  2. [10]

    Hybrid Camera Pose Estimation

    Federico Camposeco, Andrea Cohen, Marc Pollefeys, and Torsten Sattler. Hybrid Camera Pose Estimation. In Confer- ence on Computer Vision and Pattern Recognition (CVPR) ,

  3. [11]

    Locally opti- mized ransac

    Ond ˇrej Chum, Ji ˇr´ı Matas, and Josef Kittler. Locally opti- mized ransac. In Pattern Recognition: 25th DAGM Sympo- sium, Magdeburg, Germany, September 10-12, 2003. Pro- ceedings 25, pages 236–243. Springer, 2003. 6

    2003

  4. [12]

    Using alge- braic geometry

    David A Cox, John Little, and Donal O’shea. Using alge- braic geometry. Springer Science & Business Media, 2006. 1, 2

    2006

  5. [13]

    Scannet: Richly-annotated 3d reconstructions of indoor scenes

    Angela Dai, Angel X Chang, Manolis Savva, Maciej Hal- ber, Thomas Funkhouser, and Matthias Nießner. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In Proceedings of the IEEE conference on computer vision and pattern recognition, 2017. 5, 4, 7

    2017

  6. [14]

    Superpoint: Self-supervised interest point detection and description

    Daniel DeTone, Tomasz Malisiewicz, and Andrew Rabi- novich. Superpoint: Self-supervised interest point detection and description. In Proceedings of the IEEE conference on computer vision and pattern recognition workshops , 2018. 6, 7, 8, 3, 4, 5

    2018

  7. [15]

    Revisiting the p3p problem

    Yaqing Ding, Jian Yang, Viktor Larsson, Carl Olsson, and Kalle ˚Astr¨om. Revisiting the p3p problem. In Computer Vision and Pattern Recognition (CVPR), 2023. 4, 5, 6, 7, 3

    2023

  8. [16]

    Noisy one-point homographies are surpris- ingly good

    Yaqing Ding, Jonathan Astermark, Magnus Oskarsson, and Viktor Larsson. Noisy one-point homographies are surpris- ingly good. In Computer Vision and Pattern Recognition (CVPR), 2024. 2

    2024

  9. [17]

    Fundamental matrix es- timation using relative depths

    Yaqing Ding, V ´aclav V´avra, Snehal Bhayani, Qianliang Wu, Jian Yang, and Zuzana Kukelova. Fundamental matrix es- timation using relative depths. In European Conference on Computer Vision (ECCV), 2024. 2, 3, 4, 5, 7, 8, 6

    2024

  10. [18]

    Roma: Robust dense feature matching

    Johan Edstedt, Qiyu Sun, Georg B ¨okman, M ˚arten Wadenb¨ack, and Michael Felsberg. Roma: Robust dense feature matching. In Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, 2024. 6, 7, 8, 3, 4, 5

    2024

  11. [19]

    Relative pose from deep learned depth and a single affine correspondence

    Ivan Eichhardt and Daniel Barath. Relative pose from deep learned depth and a single affine correspondence. In Euro- pean Conference on Computer Vision (ECCV), 2020. 1

    2020

  12. [20]

    Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography.Communications of the ACM, 1981

    Martin A Fischler and Robert C Bolles. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography.Communications of the ACM, 1981. 1

    1981

  13. [21]

    A minimal case solution to the calibrated relative pose prob- lem for the case of two known orientation angles

    Friedrich Fraundorfer, Petri Tanskanen, and Marc Pollefeys. A minimal case solution to the calibrated relative pose prob- lem for the case of two known orientation angles. In Euro- pean Conference on Computer Vision (ECCV), 2010. 1

    2010

  14. [22]

    Complete solution classification for the perspective-three-point problem

    Xiao-Shan Gao, Xiao-Rong Hou, Jianliang Tang, and Hang-Fei Cheng. Complete solution classification for the perspective-three-point problem. IEEE transactions on pat- tern analysis and machine intelligence, 2003. 4

    2003

  15. [23]

    Digging into self-supervised monocular depth estimation

    Cl ´ement Godard, Oisin Mac Aodha, Michael Firman, and Gabriel J Brostow. Digging into self-supervised monocular depth estimation. In International Conference on Computer Vision (ICCV), 2019. 1

    2019

  16. [24]

    Grayson and Michael E

    Daniel R. Grayson and Michael E. Stillman. Macaulay2, a software system for research in algebraic geometry. Avail- able at http://www2.macaulay2.com, 1992. 2

    1992

  17. [25]

    Relative pose estimation for multi-camera systems from point correspondences with scale ratio

    Banglei Guan and Ji Zhao. Relative pose estimation for multi-camera systems from point correspondences with scale ratio. In Proceedings of the 30th ACM International Confer- ence on Multimedia, 2022. 1, 2

    2022

  18. [26]

    An efficient hidden vari- able approach to minimal-case camera motion estimation

    Richard Hartley and Hongdong Li. An efficient hidden vari- able approach to minimal-case camera motion estimation. Trans. Pattern Analysis and Machine Intelligence (PAMI) ,

  19. [27]

    Multiple view ge- ometry in computer vision

    Richard Hartley and Andrew Zisserman. Multiple view ge- ometry in computer vision . Cambridge university press,

  20. [28]

    Metric3d v2: A versatile monocular geomet- ric foundation model for zero-shot metric depth and surface normal estimation

    Mu Hu, Wei Yin, Chi Zhang, Zhipeng Cai, Xiaoxiao Long, Hao Chen, Kaixuan Wang, Gang Yu, Chunhua Shen, and Shaojie Shen. Metric3d v2: A versatile monocular geomet- ric foundation model for zero-shot metric depth and surface normal estimation. arXiv preprint arXiv:2404.15506, 20...

    2024 arXiv

  21. [29]

    A construct- optimize approach to sparse view synthesis without camera pose

    Kaiwen Jiang, Yang Fu, Mukund Varma T, Yash Belhe, Xi- aolong Wang, Hao Su, and Ravi Ramamoorthi. A construct- optimize approach to sparse view synthesis without camera pose. In ACM SIGGRAPH 2024 Conference Papers, 2024. 2

    2024

  22. [30]

    Image Matching across Wide Baselines: From Paper to Practice

    Yuhe Jin, Dmytro Mishkin, Anastasiia Mishchuk, Jiri Matas, Pascal Fua, Kwang Moo Yi, and Eduard Trulls. Image Matching across Wide Baselines: From Paper to Practice. International Journal of Computer Vision, 2020. 5, 6, 8, 4

    2020

  23. [31]

    Critical motions in euclidean structure from motion

    Fredrik Kahl and Bill Triggs. Critical motions in euclidean structure from motion. In Computer Vision and Pattern Recognition (CVPR), 1999. 8

    1999

  24. [32]

    Repurpos- ing diffusion-based image generators for monocular depth estimation

    Bingxin Ke, Anton Obukhov, Shengyu Huang, Nando Met- zger, Rodrigo Caye Daudt, and Konrad Schindler. Repurpos- ing diffusion-based image generators for monocular depth estimation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024. 2, 3

    2024

  25. [33]

    Ro- bust self-calibration of focal lengths from the fundamen- tal matrix

    Viktor Kocur, Daniel Kyselica, and Zuzana Kukelova. Ro- bust self-calibration of focal lengths from the fundamen- tal matrix. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 5220– 5229, 2024. 6

    2024

  26. [34]

    Closed-form solutions to minimal absolute pose problems with known vertical direction

    Zuzana Kukelova, Martin Bujnak, and Tomas Pajdla. Closed-form solutions to minimal absolute pose problems with known vertical direction. In Asian Conference on Com- puter Vision (ACCV), 2010. 1

    2010

  27. [35]

    Poly- nomial eigenvalue solutions to minimal problems in com- puter vision

    Zuzana Kukelova, Martin Bujnak, and Tomas Pajdla. Poly- nomial eigenvalue solutions to minimal problems in com- puter vision. Trans. Pattern Analysis and Machine Intelli- gence (PAMI), 2012. 1, 2

    2012

  28. [36]

    A clever elimination strategy for efficient mini- mal solvers

    Zuzana Kukelova, Joe Kileel, Bernd Sturmfels, and Tomas Pajdla. A clever elimination strategy for efficient mini- mal solvers. In Computer Vision and Pattern Recognition (CVPR), 2017. 1

    2017

  29. [37]

    PoseLib - Minimal Solvers for Camera Pose Estimation, 2020

    Viktor Larsson and contributors. PoseLib - Minimal Solvers for Camera Pose Estimation, 2020. 6, 7, 8, 3, 4, 5

    2020

  30. [38]

    Effi- cient solvers for minimal problems by syzygy-based reduc- tion

    Viktor Larsson, Kalle ˚Astr¨om, and Magnus Oskarsson. Effi- cient solvers for minimal problems by syzygy-based reduc- tion. In Computer Vision and Pattern Recognition (CVPR),

  31. [39]

    Ground- ing image matching in 3d with mast3r

    Vincent Leroy, Yohann Cabon, and J´erˆome Revaud. Ground- ing image matching in 3d with mast3r. In European Confer- ence on Computer Vision . Springer, 2024. 2, 6, 7, 8, 3, 4, 5

    2024

  32. [40]

    Dngaussian: Optimizing sparse-view 3d gaussian radiance fields with global-local depth normaliza- tion

    Jiahe Li, Jiawei Zhang, Xiao Bai, Jin Zheng, Xin Ning, Jun Zhou, and Lin Gu. Dngaussian: Optimizing sparse-view 3d gaussian radiance fields with global-local depth normaliza- tion. In Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition, 2024. 2

    2024

  33. [41]

    Sinha, Michael F

    Hyon Lim, Sudipta N. Sinha, Michael F. Cohen, Matt Uyt- tendaele, and H. Jin Kim. Real-time Monocular Image-based 6-DoF Localization. IJRR, 34(4–5):476–492, 2015. 1

    2015

  34. [42]

    Lightglue: Local feature matching at light speed

    Philipp Lindenberger, Paul-Edouard Sarlin, and Marc Polle- feys. Lightglue: Local feature matching at light speed. In International Conference on Computer Vision (ICCV), 2023. 6, 7, 8, 3, 4, 5

    2023

  35. [43]

    Depth-guided sparse structure-from-motion for movies and tv shows

    Sheng Liu, Xiaohan Nie, and Raffay Hamid. Depth-guided sparse structure-from-motion for movies and tv shows. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, 2022. 2

    2022

  36. [44]

    Scale exploiting minimal solvers for relative pose with calibrated cameras

    Stephan Liwicki and Christopher Zach. Scale exploiting minimal solvers for relative pose with calibrated cameras. In BMVC, 2017. 1, 2

    2017

  37. [45]

    Distinctive image features from scale- invariant keypoints

    David G Lowe. Distinctive image features from scale- invariant keypoints. International journal of computer vi- sion, 60(2):91–110, 2004. 2

    2004

  38. [46]

    Fine-tuning image-conditional diffusion models is easier than you think, 2024

    Gonzalo Martin Garcia, Karim Abou Zeid, Christian Schmidt, Daan de Geus, Alexander Hermans, and Bastian Leibe. Fine-tuning image-conditional diffusion models is easier than you think, 2024. 1, 2, 3

    2024

  39. [47]

    Fast monocular visual- inertial initialization leveraging learned single-view depth

    Nathaniel Merrill, Patrick Geneva, Saimouli Katragadda, Chuchu Chen, and Guoquan Huang. Fast monocular visual- inertial initialization leveraging learned single-view depth. Robotics: Science and Systems (RSS) 2023, 2023. 2

    2023

  40. [48]

    Two efficient solutions for visual odometry using directional correspondence.Trans

    Oleg Naroditsky, Xun S Zhou, Jean Gallier, Stergios I Roumeliotis, and Kostas Daniilidis. Two efficient solutions for visual odometry using directional correspondence.Trans. Pattern Analysis and Machine Intelligence (PAMI), 2012. 1

    2012

  41. [49]

    An efficient solution to the five-point relative pose problem

    David Nist ´er. An efficient solution to the five-point relative pose problem. Trans. Pattern Analysis and Machine Intelli- gence (PAMI), 2004. 1, 2, 3, 4, 5, 7

    2004

  42. [50]

    Lambda twist: An ac- curate fast robust perspective three point (p3p) solver

    Mikael Persson and Klas Nordberg. Lambda twist: An ac- curate fast robust perspective three point (p3p) solver. In Proceedings of the European conference on computer vision (ECCV), pages 318–332, 2018. 4

    2018

  43. [51]

    Unidepth: Universal monocular metric depth estimation

    Luigi Piccinelli, Yung-Hsu Yang, Christos Sakaridis, Mattia Segu, Siyuan Li, Luc Van Gool, and Fisher Yu. Unidepth: Universal monocular metric depth estimation. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10106–10116, 2024. 1, ...

    2024

  44. [52]

    Vi- sion transformers for dense prediction

    Ren ´e Ranftl, Alexey Bochkovskiy, and Vladlen Koltun. Vi- sion transformers for dense prediction. InInternational Con- ference on Computer Vision (ICCV), 2021. 2, 3

    2021

  45. [53]

    From coarse to fine: Robust hierarchical localization at large scale

    Paul-Edouard Sarlin, Cesar Cadena, Roland Siegwart, and Marcin Dymczyk. From coarse to fine: Robust hierarchical localization at large scale. In Computer Vision and Pattern Recognition (CVPR), 2019. 1, 6

    2019

  46. [54]

    Superglue: Learning feature matching with graph neural networks

    Paul-Edouard Sarlin, Daniel DeTone, Tomasz Malisiewicz, and Andrew Rabinovich. Superglue: Learning feature matching with graph neural networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020. 5

    2020

  47. [55]

    Structure- from-motion revisited

    Johannes L Schonberger and Jan-Michael Frahm. Structure- from-motion revisited. In Computer Vision and Pattern Recognition (CVPR), pages 4104–4113, 2016. 1

    2016

  48. [56]

    A multi-view stereo benchmark with high- 10 resolution images and multi-camera videos

    Thomas Schops, Johannes L Schonberger, Silvano Galliani, Torsten Sattler, Konrad Schindler, Marc Pollefeys, and An- dreas Geiger. A multi-view stereo benchmark with high- 10 resolution images and multi-camera videos. In Computer Vision and Pattern Recognition (CVPR), 2017. 5, 6, 7, 3

    2017

  49. [57]

    A minimal solution for relative pose with un- known focal length

    Henrik Stew ´enius, David Nist´er, Fredrik Kahl, and Frederik Schaffalitzky. A minimal solution for relative pose with un- known focal length. In Computer Vision and Pattern Recog- nition (CVPR), 2005. 1

    2005

  50. [58]

    Recent developments on direct relative orientation

    Henrik Stewenius, Christopher Engels, and David Nist ´er. Recent developments on direct relative orientation. ISPRS Journal of Photogrammetry and Remote Sensing, 60(4):284– 294, 2006. 1, 2

    2006

  51. [59]

    City-scale localization for cameras with known ver- tical direction

    Linus Sv ¨arm, Olof Enqvist, Fredrik Kahl, and Magnus Os- karsson. City-scale localization for cameras with known ver- tical direction. Trans. Pattern Analysis and Machine Intelli- gence (PAMI), 2016. 1

    2016

  52. [60]

    Solving for relative pose with a partially known rotation is a quadratic eigenvalue problem

    Chris Sweeney, John Flynn, and Matthew Turk. Solving for relative pose with a partially known rotation is a quadratic eigenvalue problem. International Conference on 3D Vision (3DV), 2014. 1

    2014

  53. [61]

    Mlesac: A new ro- bust estimator with application to estimating image geome- try

    Philip HS Torr and Andrew Zisserman. Mlesac: A new ro- bust estimator with application to estimating image geome- try. Computer vision and image understanding, 2000. 3

    2000

  54. [62]

    Moge: Unlocking accurate monocular geometry estimation for open-domain images with optimal training supervision, 2024

    Ruicheng Wang, Sicheng Xu, Cassie Dai, Jianfeng Xiang, Yu Deng, Xin Tong, and Jiaolong Yang. Moge: Unlocking accurate monocular geometry estimation for open-domain images with optimal training supervision, 2024. 1, 2, 3, 6, 7, 8, 4, 5

    2024

  55. [63]

    Dust3r: Geometric 3d vi- sion made easy

    Shuzhe Wang, Vincent Leroy, Yohann Cabon, Boris Chidlovskii, and Jerome Revaud. Dust3r: Geometric 3d vi- sion made easy. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024. 2, 6

    2024

  56. [64]

    Gaussianob- ject: High-quality 3d object reconstruction from four views with gaussian splatting

    Chen Yang, Sikuang Li, Jiemin Fang, Ruofan Liang, Lingxi Xie, Xiaopeng Zhang, Wei Shen, and Qi Tian. Gaussianob- ject: High-quality 3d object reconstruction from four views with gaussian splatting. ACM Transactions on Graphics (TOG), 2024. 2

    2024

  57. [65]

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

    Lihe Yang, Bingyi Kang, Zilong Huang, Xiaogang Xu, Jiashi Feng, and Hengshuang Zhao. Depth anything: Unleashing the power of large-scale unlabeled data. In Computer Vision and Pattern Recognition (CVPR), 2024. 1, 2, 3

    2024

  58. [66]

    Depth any- thing v2

    Lihe Yang, Bingyi Kang, Zilong Huang, Zhen Zhao, Xiao- gang Xu, Jiashi Feng, and Hengshuang Zhao. Depth any- thing v2. arXiv preprint arXiv:2406.09414, 2024. 1, 2, 3, 6, 7, 8, 4, 5

    2024 arXiv

  59. [67]

    Relative pose estimation through affine corrections of monocular depth priors

    Yifan Yu, Shaohui Liu, R ´emi Pautrat, Marc Pollefeys, and Viktor Larsson. Relative pose estimation through affine corrections of monocular depth priors. arXiv preprint arXiv:2501.05446, 2025. 2, 3, 4, 5, 6, 7, 8

    2025 arXiv

  60. [68]

    Monosdf: Exploring monocu- lar geometric cues for neural implicit surface reconstruction

    Zehao Yu, Songyou Peng, Michael Niemeyer, Torsten Sat- tler, and Andreas Geiger. Monosdf: Exploring monocu- lar geometric cues for neural implicit surface reconstruction. Advances in neural information processing systems, 2022. 2

    2022

  61. [69]

    Cam- era pose voting for large-scale image-based localization

    Bernhard Zeisl, Torsten Sattler, and Marc Pollefeys. Cam- era pose voting for large-scale image-based localization. In International Conference on Computer Vision (ICCV), 2015. 1

    2015

  62. [70]

    Learned monocular depth priors in visual-inertial initializa- tion

    Yunwen Zhou, Abhishek Kar, Eric Turner, Adarsh Kow- dle, Chao X Guo, Ryan C DuToit, and Konstantine Tsotsos. Learned monocular depth priors in visual-inertial initializa- tion. In European conference on computer vision, 2022. 2 11 RePoseD: Efficient Relative Pose Estimation Wi...

    2022

  63. [72]

    constraints.txt

    More Details About the Solvers 6.1. Solvers Using Three Point Correspondences For calibrated camera pose estimation with monocular depth, all possible cases can be solved using three point correspondences and a varying number of monocular depth estimates. Similarly, for focal ...

  64. [73]

    Results for 3PTsuv(Inverse) This solver was derived to be used with affine-invariant inverse depths, e.g., obtained via Depth Anything [65]

    More Experiments 7.1. Results for 3PTsuv(Inverse) This solver was derived to be used with affine-invariant inverse depths, e.g., obtained via Depth Anything [65]. However, we observed that the 3PT suv(inverse) solver does not improve the accuracy even for affine-invariant inve...

  65. [2003]

    1, 2, 4, 5, 8, 6, 7 9

Pith tools

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