Pith. sign in

REVIEW 2 major objections 5 minor 53 references

Hybrid Camera Pose Estimation with Online Partitioning for SLAM

T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a hybrid monocular SLAM system, which partitions cameras into co-visibility blocks and aligns them with rotation averaging while skipping explicit loop closure, outperforms conventional BA-based systems in accuracy…

desk verdict A genuinely new online co-visibility partitioning scheme with strong benchmark results, but the written global alignment only averages rotations and never defines how translations and scales are aligned, which is load-bearing for the reported trajectory RMSE. read the letter →

arxiv 1908.01797 v2 pith:B273TC4B submitted 2019-08-05 cs.CV cs.RO

Xinyi Li , Haibin Ling This is my paper · ORCID
classification cs.CVcs.RO
keywords cameraposeestimationsimultaneouslocalizationandmappingbundleadjustmentmotionaveragingrotationonlinepartitioningco-visibilityLevenberg-Marquardt
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 proposes a hybrid camera pose estimation framework for monocular SLAM that replaces fixed-size temporal partitions with online blocks built from spatial co-visibility, then aligns the blocks with rotation averaging. The aim is to show that this combination lowers trajectory error and global alignment time relative to conventional bundle-adjustment (BA) systems while avoiding explicit loop closure. A self-adaptive Levenberg-Marquardt solver, initialized by propagating intermediate values from previous blocks, is claimed to speed up local optimization. On indoor and outdoor driving benchmarks, the reported trajectory RMSE and runtime are lower than several state-of-the-art monocular systems. A sympathetic reader would see the contribution as a practical route from off-line distributed motion averaging to real-time SLAM.

What carries the argument

The central object is the online co-visibility-based block partition together with single rotation averaging over shared cameras. Each block is a local BA subproblem whose cameras are spatially strongly connected, and blocks share cameras from previous blocks, yielding multiple estimates of inter-block relative rotations. The global alignment solves for the inter-block rotations by minimizing geodesic mean distances over the shared-camera measurements, using a globally convergent averaging algorithm on SO(3). A self-adaptive Levenberg-Marquardt solver with an error-bound-based damping rule carries the local optimization and is said to achieve quadratic convergence given the close-to-optimal initialization from propagated intermediate values.

What would settle it

On a long sequence with a clear loop, record the per-block estimates of a shared camera's position; if the translation components of those estimates disagree while the rotations agree, rotation-only averaging cannot be what holds the trajectory together, and the reported RMSE gains would have to come from somewhere else.

Watch

Extended reading notes

Core claim

The central claim is that camera poses can be estimated accurately and efficiently by solving small local BA problems on co-visibility-based blocks and then merging the blocks through single rotation averaging over shared cameras, with no explicit loop-closure stage. The partitioning scheme enforces a minimum average co-visibility within each block and re-adds previous cameras with sufficient overlap, producing dense inter-block data association. Shared cameras appear in multiple blocks, so relative rotations between block reference frames are measured many times and averaged via the geodesic mean on SO(3). The paper reports that this hybrid formulation outperforms conventional BA-based monocular SLAM systems in both trajectory RMSE and runtime on standard benchmarks, and that the global alignment runs 60-150x faster than keyframe-based global BA.

Load-bearing premise

The global alignment assumes that averaging only the rotations of shared cameras is enough to keep the whole trajectory consistent, even though the reported trajectory error includes translations and monocular scale.

Editorial extensions

If this is right

  • If the reported comparisons hold, monocular SLAM can match or beat conventional BA-based accuracy while skipping explicit loop closure, reducing global optimization to near-constant runtime per new block.
  • Because global alignment uses only rotations, the method's computational cost stays almost flat as the number of frames grows; the paper reports 60-150x faster global alignment than keyframe-based global BA.
  • The co-visibility partition with dynamic initialization lowers the number of local BA iterations needed, as shown by comparisons with fixed-size partitions.
  • The approach remains usable in fast, feature-poor scenes because it degrades to conventional fixed-size partitioning when co-visibility is low.

Reading between the lines

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

  • One implication the paper leaves open is that extending the rotation-only averaging to full similarity averaging (rotations plus translations and scale) might reduce translation drift further, especially on sequences with large scale changes; that extension is not tested.
  • The co-visibility-based blocks could be interpreted as a form of online submap fusion; a multi-session or multi-agent SLAM system could use the same partition-and-average scheme to merge maps without global BA.
  • Because the method skips explicit loop closure, its global consistency will likely depend on how many shared cameras a revisited scene produces; trajectories that revisit with little overlap may still need loop closure.
  • The self-adaptive damping rule is stated for the case where initialization is close to optimal; applying it to sequences with abrupt motion may require fallback to a more conservative solver.
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

2 major / 5 minor

Summary. The paper proposes a hybrid camera pose estimation framework for monocular SLAM. The method partitions incoming frames into blocks according to temporal and spatial co-visibility, performs local bundle adjustment per block with a modified Levenberg-Marquardt solver initialized from propagated intermediate values, and then aligns the blocks globally using single rotation averaging over camera poses shared between blocks. The authors claim that the approach improves both accuracy and efficiency compared to conventional BA-based systems and avoids explicit loop closing. Experiments are reported on TUM RGB-D and KITTI Odometry, with trajectory RMSE and runtime comparisons against several SLAM and SfM systems.

Significance. If the proposed pipeline were fully specified and validated, the combination of co-visibility-based partitioning, local BA, and rotation-only global alignment could be a meaningful contribution to real-time monocular SLAM. The authors present extensive experimental comparisons on standard benchmarks, which would support the claimed practical benefits. However, the algorithmic description is incomplete in a way that currently prevents verification of the central claims: the global alignment stage does not specify how inter-block translations and monocular scales are aligned, even though the reported trajectory RMSE is a translation-based metric. The paper also makes a mathematical claim about quadratic convergence of the modified LM solver without presenting a proof in the main text. These gaps are load-bearing for the claimed accuracy and efficiency advantages.

major comments (2)
  1. [Section III-C (Eqs. 10–12)] The global alignment optimizes only inter-block rotations. Local BA in Section III-B solves each block relative to its own reference frame, leaving an independent similarity ambiguity (scale and translation) per block. Equations (10)–(12) define only relative rotations R_ll' and the objective in Eq. (12) is invariant to independently re-scaling or translating any block about its reference camera; no equation, variable, or update step in Section III-C references translations or scales. The text explicitly states 'we never require the absolute rotation parameters' and 'to find the set of aligned rotations R_mb'. To compute the trajectory RMSE values reported in Tables I and II, one must obtain absolute translation vectors in a common coordinate system. The described algorithm cannot produce those translations. This is a missing core component, not a mere proof gap: the reported accuracy improvements cannot be attributed to the described global alignment, and the 'skip loop closure' claim is unsupported.
  2. [Section IV (Eqs. 16–19)] The paper claims the modified Levenberg-Marquardt solver achieves a quadratic convergence rate, but no proof or precise statement of the convergence result is provided in the main text. The text says 'The convergence properties are shown similar to [37]–[40] in the supplementary materials.' This is a load-bearing mathematical claim behind the efficiency advantages of the local optimization, and it should be either stated and proven in the main text or explicitly deferred with a clear pointer to a theorem that is actually presented. Currently, the claim is unverifiable.
minor comments (5)
  1. [Abstract and Section I-A] The abstract contains the literal LaTeX markup '\XL{we improve the Levenberg-Marquardt solver...}', which appears to be a leftover from revision; this should be removed.
  2. [Section IV, Eq. (19)] The update rule for ρ(η, α_k, ν_k; ν) is introduced with assertions about its inflection point and monotonicity properties, but the expression is not self-contained and the properties are not proven. Please clarify the domain and provide a reference or argument.
  3. [Section III-B, Eq. (15)] The notation ∇f(x·) = J^T(x·) is unconventional and may confuse readers; standard notation would define J as the Jacobian of f. Please make the notation consistent.
  4. [Section V, Tables I–II] The tables list RMSE and runtime for each method, but the total runtime of the proposed pipeline (including partitioning, local BA, and global alignment) is not reported; only the BA runtime is shown for 'OURS' columns. Please provide overall per-frame runtime to substantiate the real-time claim.
  5. [General] The paper would benefit from a pseudocode summary of the full algorithm, as the interactions between partitioning, local BA, propagation, and global alignment are described in prose and difficult to follow.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found: the derivations are self-contained and the reported comparisons are empirical, not constructed from the method's own outputs.

full rationale

Walking the derivation chain, I find no circular step. The local BA objective (Eq. 6) minimizes reprojection error over block-internal cameras and points; the global alignment (Eqs. 10-12) computes inter-block relative rotations from shared camera rotations via single rotation averaging; neither objective is defined in terms of the reported trajectory RMSE values in Tables I-II. Partitioning thresholds (gamma_thr=10, beta_thr=0.15, n_alpha=10, n_thr=50) are preset, not fitted to the evaluation outputs. The modified LM solver's convergence claim rests on external results (Yamashita-Fukushima, Fan-Pan, etc.) and does not import a conclusion from the authors' own prior work. There are no self-citations in the load-bearing argument and no 'uniqueness theorem' imported from the authors. The one substantive weakness - that Sec III-C's global alignment only solves for rotations, leaving inter-block translations and monocular scale unaddressed while trajectory RMSE includes translations - is a missing-component/correctness gap, not a circular reduction: the reported numbers are not equivalent to the method's inputs by construction. Therefore the circularity score is 0.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The system relies on four stated partitioning thresholds, an unspecified co-visibility edge threshold, and solver constants with unspecified values. It assumes the local error bound condition, close initialization, rotation averaging convergence, and, critically, that rotation-only global alignment suffices for trajectory consistency without explicit translation and scale alignment.

free parameters (7)
  • gamma_thr = 10
    Minimum local co-visibility score for block construction, set in Section III-A to balance accuracy and efficiency. No validation or ablation split is reported.
  • beta_thr = 0.15
    Minimum camera-block overlapping view ratio for adding previous cameras into the current block, set in Section III-A.
  • n_alpha = 10
    Upper bound on the number of added-in cameras per block, set in Section III-A to avoid excessive redundancy.
  • n_thr = 50
    Maximum block size used to keep local optimization real-time, set in Section III-A.
  • cov_thr = unspecified
    Co-visibility count threshold for edge weights in the maximum spanning forest initialization in Eq. 8; the numerical value is not stated.
  • nu = unspecified
    Given constant in (0,1) that controls the inflection point of the self-adaptive LM update rule in Eq. 19; no value is reported.
  • xi = unspecified
    Small lower bound that keeps the LM update factor strictly positive in Eq. 19; no value is reported.
assumptions (6)
  • domain assumption Local error bound condition (Eq. 16) holds for the local BA subproblems.
    Invoked in Section IV to claim quadratic convergence of the modified LM solver, but not verified for BA residuals in the paper.
  • domain assumption Propagated intermediate values initialize local BA sufficiently close to the solution set Omega.
    Used in Sections III-B and IV to justify fast convergence; no explicit bound or empirical validation is given.
  • standard math Rotation averaging on SO(3) with the Karcher mean converges globally for the inter-block graph.
    Follows Manton [35] and assumes the measured relative rotations admit a consistent global solution.
  • domain assumption Calibrated camera with known, fixed intrinsics K.
    Stated in Section II; standard for monocular SLAM but limits applicability to uncalibrated settings.
  • domain assumption Co-visibility-based blocks preserve global map consistency without explicit loop closure.
    Core design assumption in Sections III-A and III-C; not proven and not directly tested against full loop-closure systems.
  • ad hoc to paper Aligning only shared camera rotations across blocks is sufficient for global trajectory consistency.
    The global alignment in Section III-C solves only rotations (Eqs. 10 to 12), with no formulated translation or scale alignment, yet trajectory RMSE is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hybrid Camera Pose Estimation with Online Partitioning for SLAM." pith.science (2026). https://pith.science/paper/B273TC4B

@misc{pith2026190801797,
  author       = {Pith},
  title        = {Pith review of: Hybrid Camera Pose Estimation with Online Partitioning for SLAM},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/B273TC4B}},
  note         = {Machine review of arXiv:1908.01797}
}
read the original abstract

This paper presents a hybrid real-time camera pose estimation framework with a novel partitioning scheme and introduces motion averaging to monocular Simultaneous Localization and Mapping (SLAM) systems. Breaking through the limitations of fixed-size temporal partitioning in many conventional SLAM pipelines, our approach significantly improves the accuracy of local bundle adjustment by gathering spatially-strongly-connected cameras into each block. With the dynamic initialization using intermediate computation values, \XL{we improve the Levenberg-Marquardt solver to further enhance the efficiency of the local optimization.} Moreover, the dense data association between blocks by our co-visibility-based partitioning enables us to explore and implement motion averaging to efficiently align the blocks globally, updating camera motion estimations on-the-fly. Experiments on benchmarks convincingly demonstrate the practicality and robustness of our proposed approach by significantly outperforming conventional approaches.

Figures

Figures reproduced from arXiv: 1908.01797 by the authors.

Figure 1
Figure 1. The pipeline of our proposed algorithm: (a) the input image sequence and feature correspondences, (b) the online partitioning based on local and global co-visibility (§III-A), (c) the refinement of local pose-graph with semi-distributed initialization (§III-B), and (d) the progressive global alignment by motion averaging with common camera poses (§III-C). are estimated and maintained via rotation averaging. However,… view at source ↗
Figure 2
Figure 2. Our online partitioning scheme partitions the cameras into blocks based on both temporal and spatial co-visibility. The key is that one camera pose can be optimized within different blocks (e.g. C1 is estimated in B 1 and B 2 , marked as C 1 1 and C 2 1 , respectively) as long as it shares sufficient overlapping views with the blocks. constraints. New image frames are sequentially partitioned into current block B l … view at source ↗
Figure 3
Figure 3. Comparison with conventional local optimizations of fixed￾size partitions: (a) the average trajectory error for the local block with the peaks representing that a new block, and (b) iterations required for the convergence of the intra-block local BA. The experiments are conducted on Seq.02 of KITTI Odometry [32]. More details can be found in the supplementary materials. B. Local Optimization We first reformulate Eq.… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Comparison with conventional global BA on the accuracy and runtime. (a) shows the accumulated re-projection comparison with conventional hierarchical BA and incremental BA methods. (b) shows the run-time comparison with conventional keyframe-based global BA with fixed-…
Figure 5
Figure 5. Figure 5: Our proposed approach shows a trajectory estimation of comparably high quality in large scale outdoor scenes. Note on Seq.02 it is clearly shown that ORB-SLAM (in green) suffers from accumulated drifts although the trajectory contains many loops. Results on the rest of…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 52 canonical work pages

  1. [37]

    Convergence properties of a self-adaptive Levenberg- Marquardt algorithm under local error bound condition,

    J. Fan and J. Pan, “Convergence properties of a self-adaptive Levenberg- Marquardt algorithm under local error bound condition,” Computational Optimization and Applications , vol. 34, no. 1, 2006

    work page 2006

  2. [40]

    Levenberg-Marquardt methods with strong local convergence properties for solving nonlinear equations with convex constraints,

    C. Kanzow, N. Yamashita, and M. Fukushima, “Levenberg-Marquardt methods with strong local convergence properties for solving nonlinear equations with convex constraints,” Journal of Computational and Applied Mathematics, 2005

    work page 2005

  3. [1]

    Incremental Light Bundle Adjustment,

    V . Indelman, R. Roberts, C. Beall, and F. Dellaert, “Incremental Light Bundle Adjustment,” in BMVC, 2012

    work page 2012

  4. [2]

    Spatiotemporal Bundle Adjustment for Dynamic 3D Reconstruction,

    M. V o, S. G. Narasimhan, and Y . Sheikh, “Spatiotemporal Bundle Adjustment for Dynamic 3D Reconstruction,” in CVPR, 2016

    work page 2016

  5. [3]

    Adaptive relative bundle adjustment,

    D. Sibley and C. Mei, “Adaptive relative bundle adjustment,” in RSS, 2009

    work page 2009

  6. [4]

    Towards linear-time incremental structure from motion,

    C. Wu, “Towards linear-time incremental structure from motion,” in 3DV, 2013

    work page 2013

  7. [5]

    Structure-from-motion revisited,

    J. L. Schonberger and J.-M. Frahm, “Structure-from-motion revisited,” in CVPR, 2016

    2016

  8. [6]

    ICE-BA: Incremental, Consistent and Efficient Bundle Adjustment for Visual-Inertial SLAM,

    H. Liu, M. Chen, G. Zhang, H. Bao, and Y . Bao, “ICE-BA: Incremental, Consistent and Efficient Bundle Adjustment for Visual-Inertial SLAM,” CVPR, 2018

    work page 2018

Show all 53 references
  1. [7]

    FrameSLAM: From bundle adjustment to real-time visual mapping,

    K. Konolige and M. Agrawal, “FrameSLAM: From bundle adjustment to real-time visual mapping,” TRO, 2008

    2008

  2. [8]

    iSAM: Incremental smooth- ing and mapping,

    M. Kaess, A. Ranganathan, and F. Dellaert, “iSAM: Incremental smooth- ing and mapping,” in TRO, 2008

    2008

  3. [9]

    SLAM++-a highly efficient and temporally scalable incremental slam framework,

    V . Ila, L. Polok, M. Solony, and P. Svoboda, “SLAM++-a highly efficient and temporally scalable incremental slam framework,” IJRR, 2017

    2017

  4. [10]

    Efficient bundle adjustment with virtual key frames: a hierarchical approach to multi-frame structure from motion,

    H. Shum, Q. Ke, and Z. Zhang, “Efficient bundle adjustment with virtual key frames: a hierarchical approach to multi-frame structure from motion,” in CVPR, 1999

    1999

  5. [11]

    Submap-based bundle adjustment for 3D reconstruction from RGB-D data,

    R. Maier, J. Sturm, and D. Cremers, “Submap-based bundle adjustment for 3D reconstruction from RGB-D data,” in BMVC, 2014

    2014

  6. [12]

    Out-of-core bundle adjustment for large-scale 3D reconstruction,

    K. Ni, D. Steedly, and F. Dellaert, “Out-of-core bundle adjustment for large-scale 3D reconstruction,” in ICCV, 2007

    2007

  7. [13]

    Spectral partitioning for structure from motion

    D. Steedly, I. Essa, and F. Dellaert, “Spectral partitioning for structure from motion.” Georgia Institute of Technology, 2003

    2003

  8. [14]

    Efficient structure from motion by graph optimization,

    M. Havlena, A. Torii, and T. Pajdla, “Efficient structure from motion by graph optimization,” in ECCV, 2010

    2010

  9. [15]

    HSfM: Hybrid structure-from- motion,

    H. Cui, X. Gao, S. Shen, and Z. Hu, “HSfM: Hybrid structure-from- motion,” in CVPR, 2017

    2017

  10. [16]

    Robust optimization of factor graphs by using condensed measurements,

    G. Grisetti, R. K ¨ummerle, and K. Ni, “Robust optimization of factor graphs by using condensed measurements,” in ICIRS, 2012

    2012

  11. [17]

    Tectonic SAM: Exact, out-of-core, submap-based SLAM,

    K. Ni, D. Steedly, and F. Dellaert, “Tectonic SAM: Exact, out-of-core, submap-based SLAM,” in ICRA, 2007

    2007

  12. [18]

    Global structure-from-motion by similarity averag- ing,

    Z. Cui and P. Tan, “Global structure-from-motion by similarity averag- ing,” in ICCV, 2015

    2015

  13. [19]

    Very large-scale global SfM by distributed motion averaging,

    S. Zhu, R. Zhang, L. Zhou, T. Shen, T. Fang, P. Tan, and L. Quan, “Very large-scale global SfM by distributed motion averaging,” inCVPR, 2018

    2018

  14. [20]

    Parallel structure from motion from local increment to global averag- ing,

    S. Zhu, T. Shen, L. Zhou, R. Zhang, J. Wang, T. Fang, and L. Quan, “Parallel structure from motion from local increment to global averag- ing,” arXiv preprint arXiv:1702.08601 , 2017

    2017 arXiv

  15. [21]

    Efficient and robust large-scale rotation averaging,

    A. Chatterjee and V . Madhav Govindu, “Efficient and robust large-scale rotation averaging,” in ICCV, 2013

    2013

  16. [22]

    Robust relative rotation averaging,

    A. Chatterjee and V . M. Govindu, “Robust relative rotation averaging,” PAMI, 2018

    2018

  17. [23]

    Rotation averaging,

    R. Hartley, J. Trumpf, Y . Dai, and H. Li, “Rotation averaging,” IJCV, 2013

    2013

  18. [24]

    Simultaneous multiple rotation averaging using lagrangian duality,

    J. Fredriksson and C. Olsson, “Simultaneous multiple rotation averaging using lagrangian duality,” in ACCV, 2012

    2012

  19. [25]

    CSfM: community-based structure from motion,

    H. Cui, S. Shen, X. Gao, and Z. Hu, “CSfM: community-based structure from motion,” in ICIP, 2017

    2017

  20. [26]

    A consensus-based framework for distributed bundle adjustment,

    A. Eriksson, J. Bastian, T.-J. Chin, and M. Isaksson, “A consensus-based framework for distributed bundle adjustment,” in CVPR, 2016

    2016

  21. [27]

    Distributed very large scale bundle adjustment by global camera consensus,

    R. Zhang, S. Zhu, T. Fang, and L. Quan, “Distributed very large scale bundle adjustment by global camera consensus,” in ICCV, 2017

    2017

  22. [28]

    Progressive structure from motion,

    A. Locher, M. Havlena, and L. V . Gool, “Progressive structure from motion,” in ECCV, 2018

    2018

  23. [29]

    Visual SLAM: Why bundle adjustment,

    A. P. Bustos, T.-J. Chin, A. Eriksson, and I. Reid, “Visual SLAM: Why bundle adjustment,” in ICRA, 2019

    2019

  24. [30]

    How to conjugatec 1-close group actions,

    K. Grove and H. Karcher, “How to conjugatec 1-close group actions,” Mathematische Zeitschrift, vol. 132, no. 1, pp. 11–20, 1973

    1973

  25. [31]

    A method for the solution of certain non-linear problems in least squares,

    K. Levenberg, “A method for the solution of certain non-linear problems in least squares,” Quarterly of Applied Mathematics , 1944

    1944

  26. [32]

    Vision meets robotics: The KITTI dataset,

    A. Geiger, P. Lenz, C. Stiller, and R. Urtasun, “Vision meets robotics: The KITTI dataset,” IJRR, 2013

    2013

  27. [33]

    On the shortest spanning subtree of a graph and the traveling salesman problem,

    J. B. Kruskal, “On the shortest spanning subtree of a graph and the traveling salesman problem,” Proceedings of the AMS , vol. 7, no. 1, 1956

    1956

  28. [34]

    A randomized time-work optimal parallel algorithm for finding a minimum spanning forest,

    S. Pettie and V . Ramachandran, “A randomized time-work optimal parallel algorithm for finding a minimum spanning forest,”SIAM Journal on Computing, vol. 31, no. 6, 2002

    2002

  29. [35]

    A globally convergent numerical algorithm for computing the centre of mass on compact lie groups,

    J. H. Manton, “A globally convergent numerical algorithm for computing the centre of mass on compact lie groups,” in ICARCV, 2004

    2004

  30. [36]

    On the rate of convergence of the Levenberg-Marquardt method,

    N. Yamashita and M. Fukushima, “On the rate of convergence of the Levenberg-Marquardt method,” in Topics in Numerical Analysis . Springer, 2001

    2001

  31. [38]

    A self-adaptive trust region algorithm,

    L. Hei, “A self-adaptive trust region algorithm,” Journal of Computa- tional Mathematics, 2003

    2003

  32. [39]

    On the quadratic convergence of the Levenberg- Marquardt method without nonsingularity assumption,

    J. Fan and Y . Yuan, “On the quadratic convergence of the Levenberg- Marquardt method without nonsingularity assumption,” Computing, vol. 74, no. 1, 2005

    2005

  33. [41]

    Distinctive image features from scale-invariant keypoints,

    D. G. Lowe, “Distinctive image features from scale-invariant keypoints,” IJCV, 2004

    2004

  34. [42]

    Dense visual SLAM for RGB-D cameras,

    C. Kerl, J. Sturm, and D. Cremers, “Dense visual SLAM for RGB-D cameras,” in ICIRS, 2013

    2013

  35. [43]

    Kintinuous: Spatially Extended KinectFusion,

    T. Whelan, M. Kaess, M. Fallon, H. Johannsson, J. Leonard, and J. McDonald, “Kintinuous: Spatially Extended KinectFusion,”Computer Science and Artificial Intelligence Laboratory Technical Report , 2012

    2012

  36. [44]

    ElasticFusion: Dense SLAM Without A Pose Graph,

    T. Whelan, S. Leutenegger, R. Salas Moreno, B. Glocker, and A. Davi- son, “ElasticFusion: Dense SLAM Without A Pose Graph,” in RSS, 2015

    2015

  37. [45]

    LDSO: Direct sparse odometry with loop closure,

    X. Gao, R. Wang, N. Demmel, and D. Cremers, “LDSO: Direct sparse odometry with loop closure,” in IROS, 2018

    2018

  38. [46]

    ORB-SLAM: A Versatile and Accurate Monocular SLAM System,

    R. Mur-Artal, J. M. Montiel, and J. D. Tardos, “ORB-SLAM: A Versatile and Accurate Monocular SLAM System,” TRO, 2015

    2015

  39. [47]

    Efficient Non-Consecutive Feature Tracking for Robust Structure-From-Motion,

    G. Zhang, H. Liu, Z. Dong, J. Jia, T. T. Wong, and H. Bao, “Efficient Non-Consecutive Feature Tracking for Robust Structure-From-Motion,” TIP, 2016

    2016

  40. [48]

    VisualSFM: A visual structure from motion system,

    C. Wu, “VisualSFM: A visual structure from motion system,” 2011

    2011

  41. [49]

    A benchmark for RGB-D visual odometry, 3D reconstruction and SLAM,

    A. Handa, T. Whelan, J. McDonald, and A. J. Davison, “A benchmark for RGB-D visual odometry, 3D reconstruction and SLAM,” in ICRA, 2014

    2014

  42. [50]

    ORB: An efficient alternative to SIFT or SURF,

    E. Rublee, V . Rabaud, K. Konolige, and G. R. Bradski, “ORB: An efficient alternative to SIFT or SURF,” in ICCV, 2011

    2011

  43. [51]

    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, 1981

    1981

  44. [52]

    Ceres solver,

    S. Agarwal and K. M. et al., “Ceres solver,” http://ceres-solver.org

  45. [53]

    g 2 o: A general framework for graph optimization,

    R. K ¨ummerle, G. Grisetti, H. Strasdat, K. Konolige, and W. Burgard, “g 2 o: A general framework for graph optimization,” in ICRA, 2011

    2011

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.