Pith. sign in

REVIEW 3 major objections 5 minor 35 references

Particle Image Velocimetry Refinement via Consensus ADMM for Active Fluid Control

T0 review · 3 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read Running several PIV estimators in parallel and fusing them with consensus ADMM cuts flow-measurement error by up to 20 percent.

desk verdict A sensible consensus-ADMM fusion layer with a credible 20% EPE gain without outlier rejection, but the advertised outlier-rejection and real-world-control gains rest on an oracle threshold and an absent experiment. read the letter →

arxiv 2512.11695 v2 pith:2NQKVJ4B submitted 2025-12-12 physics.flu-dyn cs.CVeess.IVmath.OC

Alan Bonomi , Francesco Banelli , Antonio Terpin This is my paper
classification physics.flu-dyncs.CVeess.IVmath.OC PACS 47.80.Jk
keywords ParticleImageVelocimetryopticalflowconsensusADMMmulti-estimatorfusionend-pointerrorincompressibilityprioractivecontroloutlierrejection
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper argues that flow quantification is a fusion problem: different PIV/optical-flow algorithms, or different tunings of the same algorithm, fail on different regions of an image pair, so the best answer is to run several in parallel and reconcile them. The reconciliation is done by a consensus ADMM scheme that pulls the individual estimates toward agreement while enforcing smoothness, spatial-acceleration, and incompressibility priors. On the PIV benchmark, the fused field from three DIS tunings beats the best single tuning by up to 20% of average end-point error at a 60Hz inference rate, and the authors show larger gains for other estimators if outliers are first rejected. The deployable claim rests on the no-rejection case, because the outlier-rejection experiments use an oracle threshold that knows the true per-pixel error.

What carries the argument

The mechanism is consensus ADMM, which solves a multi-estimator fusion problem by assigning each base algorithm a local copy of the flow field, enforcing equality between local copies and a global consensus field via penalty and dual-update steps. The data term is a per-pixel convex loss (Huber, l1, or l2) around each estimator's output, whose update reduces to scalar proximal operators; the consensus step incorporates three priors—smoothness via finite differences, spatial acceleration via a Laplacian-of-Gaussian filter, and incompressibility via divergence—and is solved approximately by gradient descent. The per-pixel weights that encode confidence in each estimator can be constant, photom

What would settle it

Take the PIV test set, run each base estimator and the consensus fusion with no oracle information, and compare average EPE; if the fused field is not consistently at or below the best single estimator's EPE, the central claim fails. Alternatively, run the oracle-rejection experiment with the threshold computed from an independently tuned outlier detector rather than ground truth; if the gain drops to the no-rejection level, the headroom claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that combining the outputs of several heterogeneous flow estimators through a consensus ADMM formulation yields a flow field that is more accurate than the best individual estimator, for the same inference cost. Concretely, for a set of DIS estimators the consensus output reduces average end-point error on the PIV test set by up to 20% relative to the best single tuning, at a real-time rate of 60Hz; with oracle outlier rejection the improvement reaches up to 40% for the Farnebäck method. The authors also demonstrate the method's usefulness in an active flow control loop, where the refined estimates support an RL agent that reduces drag by 36% after two minutes of

Load-bearing premise

The 40% Farnebäck improvement (and the shape of the gain curves) assumes an oracle that knows the true per-pixel error at test time; the directly deployable claim is only the no-rejection 20% improvement.

Editorial extensions

If this is right

  • Any existing PIV or optical-flow estimator can be plugged into the consensus layer, so accuracy improvements do not require retraining or replacing the base method.
  • Because the fusion keeps the inference rate of the underlying estimators, closed-loop active flow control can use more accurate feedback without sacrificing real-time operation.
  • The oracle outlier-rejection results show a large headroom: if a reliable per-pixel confidence or outlier detector is available, consensus gains roughly double.
  • The incompressibility and smoothness priors encode physical constraints that are not present in most single estimators, so the consensus field is more physically plausible as well as more accurate.
  • Learning-based estimators could also be fused, though the paper leaves open how to fine-tune them to maximally benefit from the consensus layer.

Reading between the lines

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

  • A practical research target implied by the oracle results is a learned per-pixel confidence proxy that reproduces the oracle threshold without ground truth; success would raise the deployable gain from 20% toward the 40% level.
  • The complementarity observed across tunings suggests the consensus layer could also improve robustness to out-of-distribution flows, where any single estimator's error is large in different places.
  • One could test the method as a post-hoc regularizer for learning-based estimators that suffer catastrophic forgetting: fusing a fine-tuned and a generalist model may recover some of the lost generality.
  • For control applications, the demonstrated drag reduction indicates that flow feedback quality, not just estimator benchmark scores, is what improved; a direct experiment comparing single-estimator vs consensus feedback under identical control policies would isolate this effect.
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 multi-estimator fusion layer for Particle Image Velocimetry. Several flow estimators (DIS, DeepFlow, Farnebäck, or different tunings of DIS/Farnebäck) run in parallel on the same image pair, and their dense fields are reconciled by a consensus ADMM scheme. The data terms use per-pixel weighting with Huber, ℓ1, or ℓ2 losses; the consensus update enforces smoothness, spatial-acceleration, and incompressibility priors via a quadratic regularizer solved approximately with Adam. Experiments on the PIV dataset report relative average end-point-error (rAEPE) improvements, including up to 20% for DIS at τ=∞ (no outlier rejection) and up to 40% for Farnebäck when an oracle threshold is used. Ablations cover data-term choices, weighting strategies, and the effect of regularization. The abstract additionally claims a 60 Hz inference rate and successful deployment in a real-world active-flow-control setup with drag reduced by 36% and increased by 32%.

Significance. If the τ=∞ improvement is robust, the consensus layer is a useful, algorithm-agnostic postprocessor that can preserve real-time operation and is integrated into Flow Gym with a JAX implementation. The paper's strengths include a coherent convex formulation with closed-form proximal updates, controlled comparisons across base estimators, and reproducible code release. However, the most prominent advertised benefits—outlier-rejection gains and real-world active-flow-control results—either depend on oracle access to ground-truth EPE or are not documented in the main text. The significance is therefore conditional: the core τ=∞ result is plausible and valuable, but the current presentation overstates the deployable performance.

major comments (3)
  1. [Section 3.1, Figure 3, definition of w^τ] The outlier-rejection experiments use an oracle threshold: w^τ_{i,ℓ}=w_{i,ℓ} if the true per-pixel EPE is below τ, and 0 otherwise. Thus the finite-τ gains in Figure 3, including the 40% Farnebäck improvement, assume the algorithm knows the ground-truth error at test time. The τ=∞ curve (about 20% for DIS) is deployable, but the abstract and Section 3.1 highlight outlier rejection as an additional benefit without stating that it is oracle-based. The paper should either supply a practical confidence/outlier-detection scheme (e.g., a calibrated photometric-residual rule) or explicitly restrict the outlier-rejection claims to an upper-bound analysis.
  2. [Abstract vs. main text] The abstract states that the method was 'successfully deployed' in the real-world active-fluids-control setup of Terpin and D'Andrea, with drag reduced by 36% or increased by up to 32% after two minutes of interaction. No section of the main text describes this experiment, the hardware, the training protocol, or the results. This is a load-bearing advertised contribution and cannot be verified from the manuscript. The authors should add a full experimental section or remove the claim from the abstract.
  3. [Abstract and Section 3.1 (60 Hz claim)] The claim of a 20% EPE decrease 'at an inference rate of 60Hz' is not supported by any timing measurement or profiling setup in the body. The pipeline runs multiple base estimators plus K1=30 outer ADMM iterations, each with K2=30 Adam steps on a full-resolution field, so the runtime cost is nontrivial. The authors should report actual wall-clock timings (hardware, image size, batch size, breakdown per stage) or clearly state that 60 Hz refers only to the base DIS estimator, not to the full consensus pipeline.
minor comments (5)
  1. [Section 3.3 / Appendix B.1] The description of 'PE weighting' as 'inverse photometric error' is imprecise: the implemented weight in Eq. (18) also includes the gradient factor ||∇I_0||^2. Clarify the distinction in the text.
  2. [Appendix B.1, Proposition B.1] The proof of Proposition B.1 appears to have several inequalities with questionable directions (e.g., subtracting (E[C||Δ||^2])^2 and then dropping it). Since the weighting is motivated rather than used as a formal guarantee, this is not fatal, but the derivation should be cleaned up or explicitly labeled as a heuristic motivation.
  3. [Section 3.1] The phrasing 'Farnebäck tops the 40% improvement' is ambiguous: the 40% is relative to the best single Farnebäck tuning, not the absolute EPE. State the baseline explicitly in the caption or text.
  4. [Section 1, Example 1] The table header 'Raft32-cylinder' should be 'RAFT32-cylinder' for consistency. More importantly, the catastrophic-forgetting example is dramatic but the row labels 'Raft32-cylinder' etc. should specify which class was fine-tuned; the caption is currently terse.
  5. [General] The paper references 'Section C' for hyperparameter tuning, but the appendix is labeled 'C' only in the appendix heading; the main text should use 'Appendix C' consistently. Also, the regularization hyperparameters in Appendix C are tuned on a validation split; state the sensitivity of the results to these choices, since the method introduces several free parameters (λ_s, λ_acc, λ_div, ρ, η, K1, K2, Huber δ).

Circularity Check

1 steps flagged · score 4.0 of 10

Oracle threshold masks ground-truth errors to inflate 'outlier rejection' gains; core consensus derivation and the 20% DIS result are independent.

  1. fitted input called prediction [Section 3, definition of w^tau before Section 3.1; plotted in Fig. 3 ('Oracle threshold tau [px]'); advertised in Abstract and Section 4]
    "we define, for tau>=0, w^tau_{i,l} as w^tau_{i,l}=w_{i,l} if (EPE_i)_{p(l)}<=tau, where p(l) is the pixel entry corresponding to l, and w^tau_{i,l}=0 otherwise. By adopting this fully controlled outlier rejection, we decouple the results from a specific combination of algorithm and outlier rejection scheme."

    The rejection mask is built from the test-time ground-truth per-pixel EPE, which is exactly the quantity the evaluation then reports as rAEPE. Figure 3's 'Farneback tops the 40% improvement with outlier rejection' and the abstract's 'this performance boost can be increased further with outlier rejection' are measured under this oracle, so those extra gains are a test-time upper bound rather than a deployable capability. The paper supplies no practical detector, and its own Limitations concede 'sensitivity to severe outliers when proper outlier detection or confidence weighting is not in place.' The tau=infinity point (20% DIS) does not use the oracle and remains deployable, which is why this is partial rather than total circularity.

full rationale

The core derivation is not circular: the consensus-ADMM objective (4)-(6) optimizes a genuine data-fidelity-plus-priors problem over independently computed base estimates, with regularization weights lambda tuned on a validation split and accuracy measured against the external PIV test set [8]; the fused output does not restate its inputs. The headlined 20% EPE reduction for DIS is obtained at tau=infinity, i.e., with image-derived weights only (Appendix B.2), and is deployable. The circular component is confined to the outlier-rejection benefit: Section 3 defines w^tau from the test-time ground-truth per-pixel EPE, and Figure 3's rAEPE then reports improvement in that same EPE under the oracle mask (Farneback 'tops the 40% improvement with outlier rejection'; abstract: 'this performance boost can be increased further with outlier rejection'). Those extra gains are an oracle upper bound, not an implementable prediction; the paper's own Limitations flag this ('sensitivity to severe outliers when proper outlier detection or confidence weighting is not in place') without supplying a detector. Because the axis is labeled 'Oracle threshold tau' and the 20% headline survives tau=infinity, this is partial circularity, not total. Non-circularity concerns noted for the record: Appendix B converts the upper bound of Prop. B.1 into the weight formula w approx ||grad I0||^2/PE 'assuming small errors' - a rigor gap, not an input/output equivalence; the abstract's real-world drag numbers (36%/32%) are claimed with no corresponding body section, relying on the authors' own prior setup [1]; and the 60Hz inference claim has no timing measurement in the text. Self-citations (Flow Gym [18], [1]) are infrastructure/precedent, not load-bearing: the central evaluation is against external benchmarks and standard estimator implementations.

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

The method itself is a convex consensus optimization with hand-tuned regularization; the main extra assumptions are the oracle-based outlier selection in evaluation, the approximate inner ADMM solve, and the unbiasedness/small-error assumptions behind the confidence weighting.

free parameters (6)
  • lambda_s, lambda_acc, lambda_div = e.g., 3, 30, 1300 for DIS/Huber with Appendix B.2 weights
    Regularization weights tuned on a validation split of the PIV training set (Appendix C); they directly control the balance between data fidelity and the smoothness/acceleration/incompressibility priors.
  • ADMM penalty rho
    Appears in updates (6a)-(6c) and in Algorithm 2's gradient (N*rho/2 term), but its value is never reported.
  • Huber delta
    Used in the Huber proximal update (16); no value or tuning procedure is given.
  • Adam gain eta = 0.01
    Hand-selected descent gain for the approximate global update (Algorithm 2).
  • ADMM iteration counts K1, K2 = 30, 30
    Finite iterations; the convergence guarantee is asymptotic and not checked.
  • Outlier threshold tau = varied 0.1 to infinity as an oracle
    In Section 3, tau is applied to true per-pixel EPE, so it is a ground-truth-dependent selection parameter, not a practical tuning parameter.
assumptions (6)
  • domain assumption Brightness constancy (Eq. 2) is a good model for PIV image pairs.
    Used to define data terms and to motivate the weighting derivation in Appendix B.
  • standard math f_i and g are closed, proper, convex and a saddle point exists, giving ADMM convergence [26, Sec. 3.2].
    The chosen Huber/l1/l2 losses and quadratic regularizers satisfy this; the claim is cited to Boyd et al.
  • ad hoc to paper Assumption 1 (Appendix B): every base estimator is unbiased (E[û_i]=u*) with covariance sigma^2 I.
    Used to derive the gradient-adjusted photometric confidence weight; not tested and unlikely to hold exactly, especially for biased learning-based estimators.
  • domain assumption For Proposition B.1, u* is locally constant and the linearization remainder satisfies |r_i| <= C ||Delta û_i||^2.
    Needed to convert photometric residual into a variance bound; the constant-flow assumption limits applicability.
  • ad hoc to paper Approximate inner ADMM update by 30 Adam iterations with eta=0.01 still converges to the global optimum.
    The paper asserts this via [26] but does not verify the inexact-update error summability conditions.
  • ad hoc to paper Oracle outlier selection with true EPE is representative of deployable outlier rejection.
    Section 3 defines w_tau using ground-truth EPE; no practical confidence scheme is shown to match it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Particle Image Velocimetry Refinement via Consensus ADMM for Active Fluid Control." pith.science (2026). https://pith.science/paper/2NQKVJ4B

@misc{pith2026251211695,
  author       = {Pith},
  title        = {Pith review of: Particle Image Velocimetry Refinement via Consensus ADMM for Active Fluid Control},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2NQKVJ4B}},
  note         = {Machine review of arXiv:2512.11695}
}
read the original abstract

Particle Image Velocimetry (PIV) is among the central modalities for measuring flow fields across laboratory, industrial and environmental setting. Traditional PIV approaches typically depend on tuning parameters specific to the imaging setup, making the performance sensitive to variations in illumination, flow conditions, and seeding density. Similarly, state-of-the-art machine learning methods for flow quantification are fragile outside their training set. In our experiments, we observed that flow quantification would improve if different tunings (or algorithms) were applied to different regions of the same image pair. Motivated by this observation, we thus pose flow quantification as a multi-estimator fusion problem: several heterogeneous algorithms process the same image pair in parallel, and their dense flow fields are treated as complementary estimates. To fuse them, we adopt a consensus framework based on the alternating direction method of multipliers, incorporating priors such as smoothness and incompressibility. We perform several numerical experiments to demonstrate the benefits of this approach. For instance, we achieve a decrease in end-point-error of up to 20% of a dense-inverse-search estimator at an inference rate of 60Hz, and we show how performance can be increased with outlier rejection. Our method is implemented in JAX and integrated into Flow Gym, enabling reproducible comparisons with the state of the art and systematic evaluation across different base algorithms. Finally, we demonstrate successful deployment of our method in the same real-world active-fluids-control setup of Terpin and D'Andrea [1], where a reinforcement-learning agent uses our flow estimates to learn to minimize drag (down by 36%) or maximize it (up to 32%) with only two minutes of real-world interaction. Hardware and software are made available at ActiveFluidControl.com.

Figures

Figures reproduced from arXiv: 2512.11695 by the authors.

Figure 1
Figure 1. The proposed consensus ADMM pipeline for PIV refinement. PIV images (consecutive snapshots [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. On the left and middle, we report the estimates for the same image pair of two different tunings [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Benefits of consensus across different sets of algorithms [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Impact of different data terms fi(·); see Section 3.2. We report the rAEPE [%] (1) (lower is better) as a function of the outlier rejection threshold τ . • PE weighting: inverse photometric error as a confidence measure; cf. Appendix B.1. • Gradient-adjusted PE weighti…
Figure 5
Figure 5. Figure 5: Impact of different weighting schemes; see Section [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Impact of the regularization; see Section [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

35 extracted references · 3 linked inside Pith

  1. [1]

    Scarano, Iterative image deformation methods in PIV, Measurement science and technology (2001)

    F. Scarano, Iterative image deformation methods in PIV, Measurement science and technology (2001)

    2001

  2. [2]

    Westerweel, G

    J. Westerweel, G. E. Elsinga, R. J. Adrian, Particle image velocimetry for complex and turbulent flows, Annual Review of Fluid Mechanics (2013)

    2013

  3. [3]

    H. Wang, G. He, S. Wang, Globally optimized cross-correlation for particle image velocimetry, Experiments in Fluids (2020)

    2020

  4. [4]

    Astarita, Analysis of weighting windows for image deformation methods in PIV, Experiments in fluids (2007)

    T. Astarita, Analysis of weighting windows for image deformation methods in PIV, Experiments in fluids (2007)

    2007

  5. [5]

    F. F. J. Schrijer, F. Scarano, Effect of predictor–corrector filtering on the stability and spatial resolution of iterative PIV interrogation, Experiments in Fluids (2008)

    2008

  6. [6]

    S. Cai, J. Liang, Q. Gao, C. Xu, R. Wei, Particle image velocimetry based on a deep learning motion estimator, IEEE Transactions on Instrumentation and Measurement (2019)

    2019

  7. [7]

    S. Cai, J. Liang, S. Zhou, Q. Gao, C. Xu, R. Wei, S. Wereley, J.-S. Kwon, Deep-PIV: A new framework of PIV using deep learning techniques, in: Proceedings of the 13th International Symposium on Particle Image Velocimetry—ISPIV, 2019

    2019

  8. [8]

    S. Cai, S. Zhou, C. Xu, Q. Gao, Dense motion estimation of particle images via a convolutional neural network, Experiments in Fluids (2019)

    2019

Show all 35 references
  1. [9]

    Manickathan, C

    L. Manickathan, C. Mucignat, I. Lunati, Kinematic training of convolutional neural networks for particle image velocimetry, Measurement Science and Technology (2022)

    2022

  2. [10]

    Q. Gao, H. Lin, H. Tu, H. Zhu, R. Wei, G. Zhang, X. Shao, A robust single-pixel particle image velocimetry based on fully convolutional networks with cross-correlation embedded, Physics of Fluids (2021)

    2021

  3. [11]

    Lagemann, K

    C. Lagemann, K. Lagemann, S. Mukherjee, W. Schröder, Deep recurrent optical flow learning for particle image velocimetry data, Nature Machine Intelligence (2021)

    2021

  4. [12]

    Rabault, J

    J. Rabault, J. Kolaas, A. Jensen, Performing particle image velocimetry using artificial neural networks: a proof-of-concept, Measurement Science and Technology (2017)

    2017

  5. [13]

    Y. Lee, H. Yang, Z. Yin, PIV-DCNN: cascaded deep convolutional neural networks for particle image velocimetry, Experiments in Fluids (2017)

    2017

  6. [14]

    Q. Zhu, J. Wang, J. Hu, J. Ai, Y. Lee, PIV-FlowDiffuser: Transfer-learning-based denoising diffusion models for PIV, arXiv preprint arXiv:2504.14952 (2025)

    2025 arXiv

  7. [15]

    Y. A. Reddy, J. Wahl, M. Sjödahl, Twins-PIVNet: Spatial attention-based deep learning framework for particle image velocimetry using vision transformer, Ocean Engineering (2025)

    2025

  8. [16]

    Huang, X

    Z. Huang, X. Shi, C. Zhang, Q. Wang, K. C. Cheung, H. Qin, J. Dai, H. Li, Flowformer: A transformer architecture for optical flow, in: European conference on computer vision, 2022. 12

    2022

  9. [17]

    Terpin, R

    A. Terpin, R. D’Andrea, Using reinforcement learning to probe the role of feedback in skill acquisition, arXiv preprint arXiv:2512.08463 (2025)

    2025

  10. [18]

    Banelli, A

    F. Banelli, A. Terpin, A. Bonomi, R. D’Andrea, Flow Gym, Working paper (2025)

    2025

  11. [19]

    Dohare, J

    S. Dohare, J. F. Hernandez-Garcia, Q. Lan, P. Rahman, A. R. Mahmood, R. S. Sutton, Loss of plasticity in deep continual learning, Nature (2024)

    2024

  12. [20]

    C. J. Kähler, T. Astarita, P. P. Vlachos, J. Sakakibara, R. Hain, S. Discetti, R. La Foy, C. Cierpka, Main results of the 4th International PIV Challenge, Experiments in Fluids (2016)

    2016

  13. [21]

    G. R. Jassal, W. Thielicke, B. E. Schmidt, An optical flow algorithm with automatic parameter adjustment for fluid velocimetry, Journal of Open Research Software (2025)

    2025

  14. [22]

    Kroeger, R

    T. Kroeger, R. Timofte, D. Dai, L. V. Gool, Fast optical flow using dense inverse search, arXiv preprint arXiv:1603.03590 (2016)

    2016 arXiv

  15. [23]

    Weinzaepfel, J

    P. Weinzaepfel, J. Revaud, Z. Harchaoui, C. Schmid, DeepFlow: Large displacement optical flow with deep matching, in: IEEE International Conference on Computer Vision (ICCV), 2013

    2013

  16. [24]

    Farnebäck, Two-frame motion estimation based on polynomial expansion, in: Image Analysis, Proc

    G. Farnebäck, Two-frame motion estimation based on polynomial expansion, in: Image Analysis, Proc. 13th Scandinavian Conference on Image Analysis (SCIA), 2003

    2003

  17. [25]

    T. Liu, L. Shen, Fluid flow and optical flow, Journal of Fluid Mechanics (2008)

    2008

  18. [26]

    S. Boyd, N. Parikh, E. Chu, B. Peleato, J. Eckstein, Distributed optimization and statistical learning via the alternating direction method of multipliers, Foundations and Trends®in Machine Learning (2011)

    2011

  19. [27]

    P. J. Huber, Robust estimation of a location parameter, in: Breakthroughs in statistics: Methodology and distribution, Springer, 1992

    1992

  20. [28]

    D. Marr, E. Hildreth, Theory of edge detection, Proceedings of the Royal Society of London. Series B. Biological Sciences (1980)

    1980

  21. [29]

    D. P. Kingma, Adam: A method for stochastic optimization, arXiv preprint arXiv:1412.6980 (2014)

    2014 arXiv

  22. [30]

    Bradski, The OpenCV Library, Dr

    G. Bradski, The OpenCV Library, Dr. Dobb’s Journal of Software Tools (2000)

    2000

  23. [31]

    Stamhuis, W

    E. Stamhuis, W. Thielicke, PIVlab–towards user-friendly, affordable and accurate digital particle image velocimetry in MATLAB, Journal of open research software (2014)

    2014

  24. [32]

    Parikh, S

    N. Parikh, S. Boyd, et al., Proximal algorithms, Foundations and trends®in Optimization (2014)

    2014

  25. [33]

    Baker, I

    S. Baker, I. Matthews, Lucas-kanade 20 years on: A unifying framework, International journal of computer vision (2004)

    2004

  26. [34]

    Baker, I

    S. Baker, I. Matthews, Equivalence and efficiency of image alignment algorithms, in: Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001, 2001

    2001

  27. [35]

    A. C. Aitken, On least squares and linear combination of observations, Proceedings of the Royal Society of Edinburgh (1935). 13 A Choices of the data termfi(·) In this section, we derive the reformulations of (6a) when (7) is theℓ1,ℓ 2 norm or the Huber loss [27]. ℓ1 norm.The ...

    1935

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.