Pith. sign in

REVIEW 4 major objections 7 minor 46 references

Discrete Laplace Operator Estimation for Dynamic 3D Reconstruction

T0 review · 4 major / 7 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 the temporal order of unsynchronized photos can be recovered alongside 3D geometry by jointly estimating a discrete Laplace operator on the observations.

desk verdict A genuinely new formulation for synchronization-free dynamic 3D reconstruction, with an honest but simplified theory; the unordered-image variant leans hard on an assumption that can break, yet the paper deserves serious review. read the letter →

arxiv 1908.11044 v1 pith:GBHZHHNI submitted 2019-08-29 cs.CV

Xiangyu Xu , Enrique Dunn This is my paper · ORCID
classification cs.CV
keywords dynamic3DreconstructiondiscreteLaplaceoperatorunsynchronizedmulti-viewimagestemporalsequencingtri-convexoptimizationgraph-basedmissingdatainterpolationeventsegmentation
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

What the paper is trying to establish is that dynamic 3D reconstruction can be solved without timestamps, frame rates, or any precomputed temporal ordering: given a set of multi-view images of a moving scene, the 3D trajectories of the tracked points and the temporal adjacency of the images can be estimated together. The formulation turns each image into a graph node with an estimated 3D shape attribute, and writes the unknown spatio-temporal affinities as a discrete Laplace operator $L = D - A = D(I - W)$ on that graph. The joint estimate comes from minimizing a tri-convex cost that combines anisotropic smoothness of the 3D trajectories, a collapsing-neighborhood prior on the graph, multi-view ray convergence, and reprojection error, alternating over the structure matrix $X$, the degree matrix $D$, and the normalized weight matrix $W$. If this works, casually captured crowd-sourced images or unsynchronized cameras could be used for dynamic reconstruction without a synchronization pre-processing stage, and the temporal graph itself becomes an output of the estimation rather than an input.

What carries the argument

The load-bearing object is the discrete Laplace operator $L = D - A = D(I - W)$ on a fully connected graph whose vertices are the input images. The decomposition of the affinity matrix into a diagonal degree matrix $D$ and a row-stochastic weight matrix $W$ separates each node's local density from the relative weights of its neighbors, letting the graph's topology be learned rather than prescribed. The operator encodes three priors: the linear form $\|LX\|_F^2$ enforces anisotropic smoothness of the 3D motion, the quadratic form $\operatorname{tr}(X^\top L[A + A^\top]X)$ favors compact neighborhoods, and the spectral quadratic form $f^\top L f$ folds partial sequencing information into the graph. This single object converts dynamic reconstruction into a tri-convex program over $X$, $D$, and $W$ solved by alternating convex search.

What would settle it

Run the method on a synthetic motion-capture sequence whose trajectory is a figure-eight or back-and-forth path, render unsynchronized multi-view images with no timestamps, and compare the recovered ordering to ground truth by Kendall rank correlation; if the ordering collapses whenever the path revisits the same 3D location at widely separated times, the claim that temporal adjacency is recoverable from spatial adjacency is falsified.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is that the unknown temporal order of observations is carried by the same discrete Laplace operator that enforces smoothness of the 3D motion, and can therefore be recovered jointly with geometry. The proposed optimization minimizes over $X$, $D$, and $W$ the cost $\frac{1}{P}\|D(I-W)X\|_F^2 + \frac{\lambda_1}{P}\sum_{i,j} D_{ii}W_{ij}\|X_{i,:}-X_{j,:}\|_2^2$, plus terms that penalize distance to viewing rays and favor camera ray convergence, subject to $W$ being row-stochastic and non-negative and $D$ positive with unit trace. Minimizing the first term drives each estimated shape toward a barycentric combination of its temporal neighbors, while the second forces those neighbors to be spatially compact, so the graph connectivity approximates the true temporal chain of a smooth motion. The reported experiments show the approach staying competitive with methods that use full ground-truth sequencing and outperforming them at lower frame rates, irregular sampling, and missing-data levels of 10-50%.

Load-bearing premise

The load-bearing premise is that moments close in time are close in 3D space, and not merely that moments at the same place are close in time; for repetitive, oscillating, or self-intersecting motion, spatial proximity creates false temporal links and corrupts the recovered ordering and geometry.

Editorial extensions

If this is right

  • Dynamic reconstruction becomes possible from uncoordinated multi-view photo sets with no timestamps, at accuracy comparable to pipelines that require full ground-truth image order.
  • Under decreasing temporal sampling density, non-uniform sampling, and missing 2D observations, the learned affinity graph degrades more gracefully than fixed trajectory-basis or self-expressive dictionary baselines.
  • The spectral signature of the estimated Laplacian separates temporally disjoint events that are spatially co-located, so event segmentation falls out of the reconstruction without a separate clustering stage.
  • The same graph machinery associates feature tracks across multiple subjects, providing a data-association mechanism for multi-target dynamic scenes.

Reading between the lines

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

  • The paper leaves open whether the recovered Laplacian can be reused as a learned temporal prior to bootstrap reconstruction of new observations of the same scene; this is a natural next use of the estimated graph.
  • A testable extension is to replace Euclidean 3D proximity with acceleration-aware or trajectory-arc affinities, which could preserve correct temporal links on repetitive or self-intersecting motions while staying inside the same tri-convex framework.
  • The structure solve for $X$ is quoted as $O((NP)^3)$; scaling to video-length inputs would most likely require a faster linear solver for that block, a regime the paper does not explore.
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

4 major / 7 minor

Summary. The paper proposes a graph-theoretic framework for dynamic 3D reconstruction from unsynchronized multi-view images with unknown temporal sampling. Observations are represented as graph vertices, their 3D geometry as node attributes, and spatio-temporal affinities as edge weights. The method jointly estimates the 3D structure matrix X and a directed discrete Laplace operator parameterized as L = D(I - W) through the tri-convex objective in Eq. (7), with terms for anisotropic smoothness, neighborhood compactness, ray consistency, and multi-view reconstructability. Two variants are presented: one for unordered image sets (Sec. 4) and one for unsynchronized image streams with partial sequencing (Sec. 5). Experiments on motion capture data and multi-view image datasets show competitive or improved performance under decreasing frame rates, non-uniform sampling, missing data, and 2D noise, and the framework is extended to event segmentation and multi-target data association. A simplified reconstructability analysis in Sec. 7 derives error bounds for a fixed, ground-truth Laplacian.

Significance. If the results hold, the framework is a useful data-adaptive alternative to trajectory-basis and self-expressive dictionary methods, with the notable strength of learning temporal adjacency from geometry rather than assuming known sequencing. The problem formulation is clear, the optimization is specified in detail, and the synthetic experiments cover several practically relevant degradation conditions. The DTW arc-distance sequencing for image streams is a sensible way to mitigate the failure of Euclidean proximity for repetitive motion. However, the central claim of generality for asynchronous photography is not yet fully supported: the unordered-image variant inherits the failure modes of Assumption A3 without mitigation, and the theoretical analysis in Sec. 7 does not analyze joint topology estimation. With targeted additions (tests on repetitive motion in the unordered setting, reporting of all hyperparameters and error bars, and a clearer statement of the scope of Sec. 7), the contribution would be solid and likely of interest to the dynamic reconstruction community.

major comments (4)
  1. [Sec. 4, Eq. (15)] The unordered-photography variant estimates W and D from 3D Euclidean distances and the ray term (Eqs. (8), (10), (12)) without any temporal prior or monotonicity constraint. For repetitive or self-intersecting motion, spatial proximity can therefore create spurious temporal adjacencies, and the manuscript's own Assumption A3 (Sec. 1) concedes that spatial proximity does not imply temporal proximity. The mitigation in Sec. 5 and Table 1 uses DTW arc distance Z from intra-stream ordering, which is unavailable in the Sec. 4 setting; the repeating-motion row in Table 1 is computed with Z, not with the Euclidean distance used in asynchronous photography. Since the asynchronous-photography claim is central, the authors should either add a constraint that enforces a 1D path/temporal ordering in the unordered variant, demonstrate empirically on a repetitive/self-intersecting motion in the Sec. 4 setting, or explicitly restrict the claim. Section 8.3's co-located temporally disjoint events also need an explanation of why cross-event spatial proximity does not produce spurious edges.
  2. [Sec. 7, Eqs. (17)-(21)] The reconstructability analysis assumes L is fixed, encodes ground-truth temporal adjacency, and uses noise-free 2D observations. This does not analyze the joint estimation of X and L, which is the paper's central contribution, and the bounds in Eq. (21) therefore do not address the failure mode induced by estimating graph topology from spatial proximity. The analysis is useful for studying the effect of camera geometry and motion-plane incidence, but it should be framed as a simplified first-order analysis. The authors should add experiments or analysis that quantify the effect of topology-estimation error on the reconstruction bounds, or clearly state that the joint estimation aspect is not covered by the theoretical results.
  3. [Sec. 8.1, Eq. (15)] The hyperparameter lambda_1 appears in the optimization (Eqs. (13), (15)) but is never reported; Sec. 8.1 gives only lambda_2 = 0.0015 and lambda_3 = 0.02. Without lambda_1 values and a sensitivity study, the empirical results are not fully reproducible and it is unclear how robust the method is to hyperparameter choice. The authors should report the setting used for all three lambda parameters and provide a sensitivity analysis over a reasonable range.
  4. [Sec. 8.1 and Fig. 5] The synthetic experiments report averages over 20 executions but no standard deviations or error bars, and the multi-view image experiments are single runs. Several comparisons in Fig. 5 are close (e.g., with TB and HPF at high frame rates), so variance information is needed to support the claimed advantages. The authors should include error bars or per-trial statistics and specify the number of trials for the multi-view datasets.
minor comments (7)
  1. [Sec. 3.3, Eq. (7)] The sentence 'Based on the geometric properties encoded by the discrete Laplace operator the formulate the optimization problem' contains a typo and should read 'we formulate the optimization problem'.
  2. [Sec. 4, Eq. (11)] The definition of the ray error uses dnp, but Eq. (11) writes the same symbol; please clarify the notation and ensure the ray-error term is defined consistently for each point p and observation n.
  3. [Sec. 6] The initialization of W is not specified. The paper states initial values for D and X, but not for the variable block W; please describe the initial W used in the alternating scheme.
  4. [Sec. 8.2] The statement 'We unsynchronized images by removing concurrent observations' indicates that timestamps are used only to eliminate concurrency; please clarify whether any other timestamp information leaks into the method, since the goal is to operate without temporal metadata.
  5. [Abstract] The phrase 'arbitrary temporal sampling density and distribution' is stronger than Assumption A2, which requires sampling dense enough for approximate local linear interpolation; consider softening the wording to 'unknown and possibly irregular temporal sampling under a local-linearity assumption'.
  6. [Fig. 5] The subplot labels are small and the captions do not state whether error bars are intentionally omitted; please enlarge the labels and add a note that the plotted curves are means without variance.
  7. [Sec. 7, Eq. (20)] Please check the dimensions in the definition of b_n: the term L^T_{:,n} L X* is a column vector scaled by L^T_{:,n}, while the term \overleftrightarrow{L}_{n,:} X* is a row-scaled object; the expression may require a transpose to define b_n consistently as a scalar.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the joint DLOE formulation is self-contained and validated against external benchmarks; only minor non-load-bearing self-citation exists.

full rationale

The paper's central claim is that jointly estimating 3D structure X and the discrete Laplace operator via the decomposition A=DW in the tri-convex objective of Eq. (7) enables dynamic 3D reconstruction without timestamps. This is a genuine joint-estimation framework, not a derivation of a prediction from a fitted input. The key modeling choice, using spatial proximity as a proxy for temporal proximity, is explicitly introduced as Assumption A3 ('temporal proximity implies spatial proximity, but not vice-versa') and is acknowledged in Sec. 1 as not universally true. This is an assumption about the scene and an admitted limitation, not a circular reduction: the recovered graph topology is evaluated against external ground-truth orderings (Table 1) and external mocap/image benchmarks, rather than being defined by them. The reconstructability analysis in Sec. 7 assumes L is fixed and encodes ground-truth temporal adjacency, but this is an analytic bound for a subproblem, not a circular reuse of the method's output as its own evidence. The only notable self-citation is the comparison with SEDL [43], a prior method from the same group, but the comparison is empirical on motion-capture data and multi-view datasets, and the new formulation adds the D-W decomposition, directed graph, ray-convergence term, and spectral sequencing priors not present in SEDL. No equation in the paper reduces to its own input by construction, and no load-bearing claim rests on an unverified self-citation. Therefore the paper receives a low score reflecting only the minor, non-load-bearing self-citation, not any substantive circularity.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The method relies on three explicit motion assumptions from the introduction and on known camera parameters. The only free numbers are the three regularizer weights, one of which is not disclosed. No new physical entities are introduced; the discrete Laplace operator and arc distance are known concepts or derived measures.

free parameters (3)
  • lambda_1 = unreported
    Weight of the collapsing-neighborhood term T in Eq. (10). Only lambda_2 and lambda_3 are reported in Sec. 8.1; lambda_1 is never stated, so its value is an undisclosed tuning choice.
  • lambda_2 = 0.0015
    Weight of the reprojection term O in Eq. (11). Fixed in the motion capture experiments; the text says it depends on 2D noise level and mean camera distance, so it is chosen based on the data.
  • lambda_3 = 0.02
    Weight of the multi-view reconstructability term R in Eq. (12). Fixed in the motion capture experiments; chosen based on data.
assumptions (4)
  • domain assumption 2D observations are samples of the continuous motion of a 3D point set.
    Introduction, assumption A1. This grounds the whole formulation in continuous trajectories.
  • domain assumption Unknown and arbitrary temporal sampling density allows approximate local linear interpolation of 3D geometry.
    Introduction, assumption A2. Needed for the anisotropic smoothness prior to be valid.
  • domain assumption Temporal proximity implies spatial proximity, but not vice-versa.
    Introduction, assumption A3. Load-bearing for using spatial adjacency to infer temporal order; acknowledged as not universally true.
  • domain assumption Camera intrinsic and extrinsic parameters are known for every input image.
    Sec. 3.1 defines the input Theta as all 2D observations and camera parameters Kn, Mn. The reprojection and ray terms depend on these being accurate.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Discrete Laplace Operator Estimation for Dynamic 3D Reconstruction." pith.science (2026). https://pith.science/paper/GBHZHHNI

@misc{pith2026190811044,
  author       = {Pith},
  title        = {Pith review of: Discrete Laplace Operator Estimation for Dynamic 3D Reconstruction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GBHZHHNI}},
  note         = {Machine review of arXiv:1908.11044}
}
read the original abstract

We present a general paradigm for dynamic 3D reconstruction from multiple independent and uncontrolled image sources having arbitrary temporal sampling density and distribution. Our graph-theoretic formulation models the Spatio-temporal relationships among our observations in terms of the joint estimation of their 3D geometry and its discrete Laplace operator. Towards this end, we define a tri-convex optimization framework that leverages the geometric properties and dependencies found among a Euclideanshape-space and the discrete Laplace operator describing its local and global topology. We present a reconstructability analysis, experiments on motion capture data and multi-view image datasets, as well as explore applications to geometry-based event segmentation and data association.

Figures

Figures reproduced from arXiv: 1908.11044 by the authors.

Figure 1
Figure 1. Multi-view capture produces a set of unorganized [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Geometry of Remarks 1 & 2. At top: Arrows denote selected neighboring samples and dashed lines their convex hull. At bottom: Corresponding graph edge structure. linear topological structure, R(·) fosters strong convergence among viewing rays, while O(·) reduces reprojection errors. For simplicity, we define the problem variables in terms of L and X. However, given the explicit dependence of L on A, we’ll redefine th… view at source ↗
Figure 3
Figure 3. Arc distance between two observations of the same [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: In (a,b) error bounds specified in Eq. (21) get ”tighter” and reconstruction error is reduced when neighboring viewing [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Reconstruction error for motion capture data under different conditions. Reported averages over 20 executions. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Experiments on multi-view image capture. All datasets were devoid of concurrent observations. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Results on Dancing and Toddler [20]. Disjoint Dancing segments form an input datum. Spectral visualization of estimated affinity matrix reveal a triplet of clusters. For Toddler, we use DLOE for instance identification, see text for details. (a) Single iteration run ti…
Figure 8
Figure 8. Figure 8: Optimization run time and cost function ablation [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 44 canonical work pages

  1. [1]

    Metric multidimensional scaling (mds): ana- lyzing distance matrices

    Herv´e Abdi. Metric multidimensional scaling (mds): ana- lyzing distance matrices. Encyclopedia of measurement and statistics. Sage, Thousand Oaks, CA, pages 1–13, 2007

    work page 2007

  2. [2]

    Deformable motion 3d reconstruction by union of regularized subspaces

    Antonio Agudo and Francese Moreno-Noguer. Deformable motion 3d reconstruction by union of regularized subspaces. In 2018 25th IEEE International Conference on Image Pro- cessing (ICIP), pages 2930–2934. IEEE, 2018

    work page 2018

  3. [3]

    A scalable, efficient, and accurate solution to non-rigid structure from motion

    Antonio Agudo and Francesc Moreno-Noguer. A scalable, efficient, and accurate solution to non-rigid structure from motion. Computer Vision and Image Understanding, 167:121– 133, 2018

    work page 2018

  4. [4]

    Robust spatio- temporal clustering and reconstruction of multiple deformable bodies

    Antonio Agudo and Francesc Moreno-Noguer. Robust spatio- temporal clustering and reconstruction of multiple deformable bodies. IEEE Transactions on Pattern Analysis and Machine Intelligence, 41(4):971–984, 2019

    work page 2019

  5. [5]

    Trajectory triangulation of lines: Reconstruction of a 3d point moving along a line from a monocular image sequence

    Shai Avidan and Amnon Shashua. Trajectory triangulation of lines: Reconstruction of a 3d point moving along a line from a monocular image sequence. In Proceedings of the 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, volume 2, pages 62–66. IEEE, 1999

    work page 1999

  6. [6]

    Trajectory triangulation: 3d reconstruction of moving points from a monocular image sequence

    Shai Avidan and Amnon Shashua. Trajectory triangulation: 3d reconstruction of moving points from a monocular image sequence. IEEE Transactions on Pattern Analysis & Machine Intelligence, (4):348–357, 2000

    work page 2000

  7. [7]

    Unstructured video-based rendering: Interactive explo- ration of casually captured videos

    Luca Ballan, Gabriel J Brostow, Jens Puwein, and Marc Polle- feys. Unstructured video-based rendering: Interactive explo- ration of casually captured videos. ACM Transactions on Graphics, 29(4):Article–No, 2010

    work page 2010

  8. [8]

    Photo sequencing

    Tali Basha, Yael Moses, and Shai Avidan. Photo sequencing. In European Conference on Computer Vision, pages 654–667. Springer, 2012

    work page 2012

Show all 46 references
  1. [9]

    Feature- based sequence-to-sequence matching

    Yaron Caspi, Denis Simakov, and Michal Irani. Feature- based sequence-to-sequence matching. International Journal of Computer Vision, 68(1):53–64, 2006

    2006

  2. [10]

    Directed graph embedding

    Mo Chen, Qiong Yang, and Xiaoou Tang. Directed graph embedding. In IJCAI, pages 2707–2712, 2007

    2007

  3. [11]

    Fast and robust archetypal analysis for representation learning

    Yuansi Chen, Julien Mairal, and Zaid Harchaoui. Fast and robust archetypal analysis for representation learning. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1478–1485, 2014

    2014

  4. [12]

    Laplacians and the cheeger inequality for directed graphs

    Fan Chung. Laplacians and the cheeger inequality for directed graphs. Annals of Combinatorics, 9(1):1–19, 2005

    2005

  5. [13]

    Spectral graph theory

    Fan RK Chung. Spectral graph theory. Number 92. American Mathematical Soc., 1997

    1997

  6. [14]

    Feature-based multi-video synchro- nization with subframe accuracy

    Ahmed Elhayek, Carsten Stoll, Kwang In Kim, H-P Seidel, and Christian Theobalt. Feature-based multi-video synchro- nization with subframe accuracy. In Joint DAGM (German Association for Pattern Recognition) and OAGM Symposium, pages 266–275. Springer, 2012

    2012

  7. [15]

    Serialrank: Spectral ranking using seriation

    Fajwel Fogel, Alexandre d’Aspremont, and Milan V ojnovic. Serialrank: Spectral ranking using seriation. In Advances in Neural Information Processing Systems, pages 900–908, 2014

    2014

  8. [16]

    Synchroniza- tion of two independently moving cameras without feature correspondences

    Tiago Gaspar, Paulo Oliveira, and Paolo Favaro. Synchroniza- tion of two independently moving cameras without feature correspondences. In European Conference on Computer Vi- sion, pages 189–204. Springer, 2014

    2014

  9. [17]

    Bicon- vex sets and optimization with biconvex functions: a survey and extensions

    Jochen Gorski, Frank Pfeuffer, and Kathrin Klamroth. Bicon- vex sets and optimization with biconvex functions: a survey and extensions. Mathematical methods of operations research, 66(3):373–407, 2007

    2007

  10. [18]

    Reconstruction of a scene with multiple linearly moving objects

    Mei Han and Takeo Kanade. Reconstruction of a scene with multiple linearly moving objects. International Journal of Computer Vision, 59(3):285–300, 2004

    2004

  11. [19]

    Spatio- temporally consistent correspondence for dense dynamic scene modeling

    Dinghuang Ji, Enrique Dunn, and Jan-Michael Frahm. Spatio- temporally consistent correspondence for dense dynamic scene modeling. In European Conference on Computer Vision, pages 3–18. Springer, 2016

    2016

  12. [20]

    Panoptic studio: A massively multiview system for social interaction capture

    Hanbyul Joo, Tomas Simon, Xulong Li, Hao Liu, Lei Tan, Lin Gui, Sean Banerjee, Timothy Godisart, Bart Nabbe, Iain Matthews, et al. Panoptic studio: A massively multiview system for social interaction capture. IEEE Transactions on Pattern Analysis and Machine Intelligence, 41(1...

    2017

  13. [21]

    Space-time tradeoffs in photo sequencing

    Yael Moses, Shai Avidan, et al. Space-time tradeoffs in photo sequencing. In Proceedings of the IEEE International Conference on Computer Vision, pages 977–984, 2013

    2013

  14. [22]

    Documentation mocap database hdm05

    Meinard M ¨uller, Tido R ¨oder, Michael Clausen, Bernhard Eberhardt, Bj¨orn Kr¨uger, and Andreas Weber. Documentation mocap database hdm05. 2007

    2007

  15. [23]

    Linear sequence-to-sequence alignment

    Flavio Padua, Rodrigo Carceroni, Geraldo Santos, and Kiri- akos Kutulakos. Linear sequence-to-sequence alignment. IEEE Transactions on Pattern Analysis and Machine Intelli- gence, 32(2):304–320, 2010

    2010

  16. [24]

    3d reconstruction of a smooth articulated trajectory from a monocular image se- quence

    Hyun Soo Park and Yaser Sheikh. 3d reconstruction of a smooth articulated trajectory from a monocular image se- quence. In 2011 International Conference on Computer Vi- sion, pages 201–208. IEEE, 2011

    2011

  17. [25]

    3d reconstruction of a moving point from a series of 2d projections

    Hyun Soo Park, Takaaki Shiratori, Iain Matthews, and Yaser Sheikh. 3d reconstruction of a moving point from a series of 2d projections. In European Conference on Computer Vision, pages 158–171. Springer, 2010

    2010

  18. [26]

    3d trajectory reconstruction under perspective projec- tion

    Hyun Soo Park, Takaaki Shiratori, Iain Matthews, and Yaser Sheikh. 3d trajectory reconstruction under perspective projec- tion. International Journal of Computer Vision, 115(2):115– 135, 2015

    2015

  19. [27]

    Learning monocular 3d human pose estimation from multi-view images

    Helge Rhodin, J ¨org Sp¨orri, Isinsu Katircioglu, Victor Con- stantin, Fr´ed´eric Meyer, Erich M¨uller, Mathieu Salzmann, and Pascal Fua. Learning monocular 3d human pose estimation from multi-view images. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern ...

    2018

  20. [28]

    3d reconstruction from tangent-of-sight measurements of a moving object seen from a moving camera

    Dana Segal and Amnon Shashua. 3d reconstruction from tangent-of-sight measurements of a moving object seen from a moving camera. In European Conference on Computer Vision, pages 621–631. Springer, 2000

    2000

  21. [29]

    Trajec- tory triangulation over conic section

    Amnon Shashua, Shai Avidan, and Michael Werman. Trajec- tory triangulation over conic section. In Proceedings of the Seventh IEEE International Conference on Computer Vision, volume 1, pages 330–336. IEEE, 1999

    1999

  22. [30]

    Separable spatiotemporal priors for convex recon- struction of time-varying 3d point clouds

    Tomas Simon, Jack Valmadre, Iain Matthews, and Yaser Sheikh. Separable spatiotemporal priors for convex recon- struction of time-varying 3d point clouds. In European Con- ference on Computer Vision, pages 204–219. Springer, 2014

    2014

  23. [31]

    Kronecker-markov prior for dynamic 3d reconstruc- tion

    Tomas Simon, Jack Valmadre, Iain Matthews, and Yaser Sheikh. Kronecker-markov prior for dynamic 3d reconstruc- tion. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(11):2201–2214, 2017

    2017

  24. [32]

    Laplacian mesh processing

    Olga Sorkine. Laplacian mesh processing. In Eurographics (STARs), pages 53–70, 2005

    2005

  25. [33]

    Video synchronization from human motion using rank constraints

    Philip A Tresadern and Ian D Reid. Video synchronization from human motion using rank constraints. Computer Vision and Image Understanding, 113(8):891–906, 2009

    2009

  26. [34]

    Synchronizing video sequences

    Tinne Tuytelaars and Luc Van Gool. Synchronizing video sequences. In Proceedings of the 2004 IEEE Conference on Computer Vision and Pattern Recognition, VOL 1 , vol- ume 1, pages 762–768. Institute of Electrical and Electronics Engineers, 2004

    2004

  27. [35]

    General trajectory prior for non-rigid reconstruction

    Jack Valmadre and Simon Lucey. General trajectory prior for non-rigid reconstruction. In Proceedings of 2012 IEEE Conference on Computer Vision and Pattern Recognition , pages 1394–1401. IEEE, 2012

    2012

  28. [36]

    Spa- tiotemporal bundle adjustment for dynamic 3d reconstruction

    Minh V o, Srinivasa G Narasimhan, and Yaser Sheikh. Spa- tiotemporal bundle adjustment for dynamic 3d reconstruction. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1710–1718, 2016

    2016

  29. [37]

    Automatic adap- tation of person association for multiview tracking in group activities

    Minh V o, Ersin Yumer, Kalyan Sunkavalli, Sunil Hadap, Yaser Sheikh, and Srinivasa Narasimhan. Automatic adap- tation of person association for multiview tracking in group activities. arXiv preprint arXiv:1805.08717, 2018

    2018 arXiv

  30. [38]

    Grassmannian regularized structured multi-view embedding for image clas- sification

    Xinchao Wang, Wei Bian, and Dacheng Tao. Grassmannian regularized structured multi-view embedding for image clas- sification. IEEE Trans. Image Processing, 22(7):2646–2660, 2013

    2013

  31. [39]

    Motion guided video sequence synchronization

    Daniel Wedge, Du Huynh, and Peter Kovesi. Motion guided video sequence synchronization. In Asian Conference on Computer Vision, pages 832–841. Springer, 2006

    2006

  32. [40]

    Convolutional pose machines

    Shih-En Wei, Varun Ramakrishna, Takeo Kanade, and Yaser Sheikh. Convolutional pose machines. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 4724–4732, 2016

    2016

  33. [41]

    Deep graph laplacian regularization

    Jin Zeng, Jiahao Pang, Wenxiu Sun, Gene Cheung, and Ruichao Xiao. Deep graph laplacian regularization. arXiv preprint arXiv:1807.11637, 2018

    2018 arXiv

  34. [42]

    Sparse dynamic 3d reconstruction from unsynchronized videos

    Enliang Zheng, Dinghuang Ji, Enrique Dunn, and Jan- Michael Frahm. Sparse dynamic 3d reconstruction from unsynchronized videos. In Proceedings of the IEEE Inter- national Conference on Computer Vision, pages 4435–4443, 2015

    2015

  35. [43]

    Self-expressive dictionary learning for dy- namic 3d reconstruction

    Enliang Zheng, Dinghuang Ji, Enrique Dunn, and Jan- Michael Frahm. Self-expressive dictionary learning for dy- namic 3d reconstruction. IEEE Transactions on Pattern Anal- ysis and Machine Intelligence, 40(9):2223–2237, 2017

    2017

  36. [44]

    Spectral embedding of directed networks

    Quan Zheng and David B Skillicorn. Spectral embedding of directed networks. Social Network Analysis and Mining, 6(1):76, 2016

    2016

  37. [45]

    3d motion reconstruction for real-world camera motion

    Yingying Zhu, M Cox, and S Lucey. 3d motion reconstruction for real-world camera motion. In Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 1–8. IEEE Computer Society, 2011

    2011

  38. [46]

    Convolutional sparse coding for trajectory reconstruction

    Yingying Zhu and Simon Lucey. Convolutional sparse coding for trajectory reconstruction. IEEE Transactions on Pattern Analysis and Machine Intelligence, 37(3):529–540, 2015

    2015

Pith tools

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