Pith. sign in

REVIEW 2 major objections 3 minor 1 cited by

A $\operatorname{prox}$-Based Semi-Smooth Newton Method for TV-Minimization

T0 review · 2 major / 3 minor · reviewed 2026-05-22 · grok-4.3

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

Pith's one-line read A prox-based semi-smooth Newton method for total variation minimization is globally well-posed and locally superlinearly convergent after finite element discretization.

desk verdict The paper gives a prox reformulation of TV optimality conditions that yields a globally well-posed semi-smooth Newton scheme on conforming finite-element meshes, with local superlinear convergence. read the letter →

arxiv 2605.22728 v1 pith:H4DEWMEB submitted 2026-05-21 math.NA cs.NAmath.OC

Sören Bartels , Alex Kaltenbach This is my paper · ORCID
classification math.NAcs.NAmath.OC
keywords totalvariationminimizationsemi-smoothNewtonmethodproximaloperatorsfiniteelementdiscretizationprimal-dualoptimalityconvexoptimizationsuperlinearconvergence
checked against Cost.FunctionalEquationabsolute_floor_iff_bare_distinguishability
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 authors introduce a semi-smooth Newton method that uses proximity operators to handle the non-differentiable total variation term in minimization problems. They reformulate the primal-dual optimality conditions into a nonlinear operator equation that admits Newton-type derivatives. For conforming finite element approximations the resulting iteration is proven to be globally solvable in each step and to converge locally with superlinear rate. The same framework covers many other convex problems and maintains invariance between primal and dual variables.

What carries the argument

The prox-based reformulation of the primal-dual optimality conditions as a Newton-differentiable nonlinear operator equation.

What would settle it

Numerical computation of the discrete solutions on successively refined meshes for a problem with known exact solution, verifying that the number of Newton iterations remains bounded while the error decreases at a superlinear rate independent of the mesh size.

Watch

Extended reading notes

Core claim

The paper establishes that the primal-dual optimality conditions for the total variation minimization problem can be rewritten as a nonlinear operator equation possessing Newton-type differentiability. For a conforming finite element discretization this yields a semi-smooth Newton method that is globally well-posed at every iteration and converges locally superlinearly to the discrete solution.

Load-bearing premise

The reformulation of the primal-dual optimality conditions must produce a nonlinear operator equation that possesses Newton-type differentiability.

Editorial extensions

If this is right

  • The approach extends directly to a large class of convex minimization problems.
  • It coincides with established semi-smooth Newton methods for obstacle problems.
  • The method satisfies a primal-dual invariance.
  • Under suitable assumptions the scheme remains well-posed in the infinite-dimensional setting.
  • Numerical performance shows reliable reduction of the discrete primal-dual gap to machine precision and robustness to proximity parameters.

Reading between the lines

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

  • The superlinear convergence rate could make the method preferable to first-order proximal algorithms for high-accuracy solutions on fine meshes.
  • Primal-dual invariance might be used to construct efficient a posteriori error indicators for adaptive mesh refinement.
  • The method could be combined with multigrid or other acceleration techniques to further improve efficiency on large-scale problems.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Request a human review

A listed scientist reviews the paper for a fee and the review publishes here regardless of verdict. See the reviewers or get listed.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 3 minor

Summary. The paper proposes a prox-based semi-smooth Newton method for total variation (TV) minimization. Primal-dual optimality conditions are reformulated as a nonlinear operator equation with Newton-type differentiable structure. Well-posedness is analyzed in infinite dimensions with structural properties of the Newton-type derivatives identified. For conforming finite element discretizations, global well-posedness and local superlinear convergence of the semi-smooth Newton scheme are proved. The approach extends to a broad class of convex minimization problems, coincides with known methods for obstacle problems, satisfies primal-dual invariance, and numerical experiments demonstrate robust performance including reliable primal-dual gap reduction to machine precision, robustness to proximity parameters, and improved convergence basin relative to a primal semi-smooth Newton method.

Significance. If the global well-posedness and convergence results hold, the work supplies a theoretically supported solver for TV-minimization with guaranteed iteration existence from arbitrary initial data and local superlinear rates, extending prior semi-smooth Newton schemes while offering practical robustness on graded meshes. The identification of structural properties of the prox-based derivatives and the primal-dual invariance are notable strengths that could benefit variational imaging and related optimization tasks.

major comments (2)
  1. [§5] §5 (Discrete well-posedness and convergence): The proof that the semi-smooth Newton iteration is globally well-posed (i.e., the Newton step exists at every iterate for arbitrary initial data) requires uniform invertibility (or a merit-function guarantee) of the generalized derivative at all generated points. The structural properties derived for the Jacobian of the proximal mapping of the dual ball constraint do not explicitly exclude singularity when discrete gradients align with the boundary of the constraint set or for certain proximity-parameter values; this must be addressed to support the global existence claim independent of mesh size and initialization.
  2. [§4.2] §4.2 (Infinite-dimensional analysis): The well-posedness investigation identifies Newton-type differentiability but leaves open whether the generalized derivatives satisfy the conditions for global existence of the iteration in the continuous setting; the 'suitable additional assumptions' mentioned for infinite-dimensional well-posedness need to be stated explicitly and verified to be non-vacuous for the TV case.
minor comments (3)
  1. [Abstract] The abstract states that the method 'coincides with established semi-smooth Newton methods for obstacle problems'; a brief remark or reference to the precise coincidence (e.g., via choice of prox) would clarify the connection.
  2. [Numerical experiments] Numerical section: The reported robustness with respect to proximity parameters is supported by experiments, but quantitative bounds or sensitivity plots would strengthen the claim that performance remains reliable across a wide range of parameter choices.
  3. [§3] Notation: The definition of the Newton-type derivative (likely in §3) uses a generalized Jacobian; ensuring consistent notation between the continuous and discrete settings would aid readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We address the major comments point by point below, indicating the revisions we plan to incorporate.

read point-by-point responses

  1. Referee: [§5] §5 (Discrete well-posedness and convergence): The proof that the semi-smooth Newton iteration is globally well-posed (i.e., the Newton step exists at every iterate for arbitrary initial data) requires uniform invertibility (or a merit-function guarantee) of the generalized derivative at all generated points. The structural properties derived for the Jacobian of the proximal mapping of the dual ball constraint do not explicitly exclude singularity when discrete gradients align with the boundary of the constraint set or for certain proximity-parameter values; this must be addressed to support the global existence claim independent of mesh size and initialization.

    Authors: We appreciate the referee highlighting the need for greater explicitness in the invertibility argument. The structural properties of the proximal mapping Jacobian, including its monotonicity and the positive contribution from the regularization term on the active set, do ensure nonsingularity even when discrete gradients align with the boundary or for admissible proximity parameters. To make this fully rigorous and independent of mesh size and initialization, we will revise §5 by adding a dedicated lemma with case analysis for boundary alignment scenarios. This will strengthen the global well-posedness proof without altering the main claims. revision: yes

  2. Referee: [§4.2] §4.2 (Infinite-dimensional analysis): The well-posedness investigation identifies Newton-type differentiability but leaves open whether the generalized derivatives satisfy the conditions for global existence of the iteration in the continuous setting; the 'suitable additional assumptions' mentioned for infinite-dimensional well-posedness need to be stated explicitly and verified to be non-vacuous for the TV case.

    Authors: We agree that the suitable additional assumptions for infinite-dimensional well-posedness should be stated explicitly. In the revision of §4.2, we will list them clearly (e.g., a positive lower bound on the proximity parameter and suitable integrability of the data) and verify that they are non-vacuous for the TV-minimization problem by confirming they hold under the standard assumptions on the domain and forcing term used throughout the paper and in the numerical section. revision: yes

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: convergence claims follow from structural analysis of Newton derivatives under standard semismooth theory

full rationale

The derivation begins with a reformulation of primal-dual optimality conditions into a nonlinear operator equation possessing Newton-type differentiability. Structural properties of the associated generalized derivatives are identified, after which standard semismooth Newton convergence results are applied to establish global well-posedness and local superlinear convergence for the conforming finite-element discretization. These steps rely on the explicit form of the proximal mapping Jacobian for the dual ball and on mesh-independent initialization, without reducing the well-posedness statement to a self-definition, a fitted parameter renamed as prediction, or a load-bearing self-citation whose content is itself unverified. The reported coincidence with established obstacle-problem schemes and the extension to broader convex problems supply independent content. No equation or theorem in the chain equates the target result to its own inputs by construction.

Assumptions & free parameters 0 free parameters · 1 assumptions · 0 invented entities

The central claims rest on the existence of a Newton-type differentiable structure after prox reformulation and on standard properties of proximity operators for convex functionals; no explicit free parameters or invented entities are introduced in the abstract.

assumptions (1)
  • domain assumption Primal-dual optimality conditions admit a reformulation as a nonlinear operator equation with Newton-(type-)differentiable structure.
    Invoked to enable the semi-smooth Newton scheme in both infinite and discrete settings.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A $\operatorname{prox}$-Based Semi-Smooth Newton Method for TV-Minimization." pith.science (2026). https://pith.science/paper/H4DEWMEB

@misc{pith2026260522728,
  author       = {Pith},
  title        = {Pith review of: A $\operatornameprox$-Based Semi-Smooth Newton Method for TV-Minimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/H4DEWMEB}},
  note         = {Machine review of arXiv:2605.22728}
}
abstract

In this paper, we devise a $\operatorname{prox}$-based semi-smooth Newton method for the non-differentiable TV-minimization problem. To this end, the primal-dual optimality conditions are reformulated as a nonlinear operator equation with Newton-(type-)differentiable structure. We investigate the question of well-posedness of the resulting semi-smooth Newton scheme in the infinite-dimensional setting and identify structural properties of the associated Newton-type derivatives. For a conforming finite element discretization, we prove that the resulting semi-smooth Newton method is globally well-posed and locally super-linearly convergent. The approach extends to a large class of convex minimization problems, coincides with established semi-smooth Newton methods for obstacle problems, satisfies a primal-dual invariance, and, under suitable additional assumptions, is well-posed in the infinite-dimensional setting. Numerical experiments indicate a robust practical performance of the proposed method, including reliable reduction of the discrete primal-dual gap estimator to machine precision, robustness with respect to the choice of proximity parameters, an improved convergence basin compared to a canonical primal semi-smooth Newton method, and effective performance even for quadratically graded meshes using only a mesh-independent initialization criterion.

Figures

Figures reproduced from arXiv: 2605.22728 by the authors.

Figure 1
Figure 1. Residual decay of the prox-based semi-smooth Newton method (cf. Algorithm 3.1) for different choices of the regularization parameter ε = h β 7 with β ∈ {0.125, 0.25, 0.5, 1, 1.5, 2, 2.5, 3}. Left: Strategy 1 combining Algorithm 3.4 with Algorithm 3.1; only the iterations of Algorithm 3.1 are displayed, with k0 ∈ N indicating the number of preceding iterations of Algorithm 3.4. Right: Strategy 2 using Algorithm 3.1 w… view at source ↗
Figure 2
Figure 2. Residual decay of the prox-based semi-smooth Newton method (cf. Algorithm 3.1) for different choices of the proximity parameter γ = 2ℓ with ℓ ∈ {−2, −1, 0, 1, . . . , 9}. For Strategy 1, only the iterations of Algorithm 3.1 are displayed, with k0 ∈ N indicating the number of preceding iterations of Algorithm 3.4, which is identical for all ℓ ∈ {−2, −1, 0, 1, . . . , 9}. While γ = 1 does not always yield the smallest… view at source ↗
Figure 3
Figure 3. Residual decay (in terms of η 2 gap,h(u k h , zk h ), k ∈ N0, cf. (3.26c)) of the prox-based semi￾smooth Newton method (cf. Algorithm 3.1) and the primal semi-smooth Newton method (cf. Algo￾rithm 4.2) for the sequence of triangulations {Thi }i=0,...,7 and regularization parameters ε∈ {h, h2}. The results indicate that the convergence basin of the prox-based semi-smooth Newton method (cf. Algorithm 3.1) is more robus… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Locally refined triangulations Ti , i ∈ {0, 2, 4, 6}, obtained by red-green-blue refinement by marking all triangles intersecting the discontinuity set Ju = ∂B2 r (0) (marked in yellow; the inter￾face is marked in red). The plot on the right shows the grading exponents…
Figure 5
Figure 5. Figure 5: Residual decay of the prox-based semi-smooth Newton method (cf. Algorithm 3.1) un￾der mesh-refinement for the modified Strategies 1∗ and 2∗ , using the same initial iterate generated via Algorithm 3.4. Strategy 2∗ uses Armijo backtracking line search, whereas Strategy …

Discussion (0). Continue with ORCID to comment.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    the nonlinear mapping F_ε : A → B ... admits the root (z_ε, u_ε) ... Newton-type derivative J_Fε(y, v)(δy, δv) := [(1−J_prox( a ))∇δv − γ J_prox(a) δy; α δv − div δy]

  • IndisputableMonolith/Foundation/AbsoluteFloorClosure.lean absolute_floor_iff_bare_distinguishability unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    Theorem 3.2 ... Algorithm 3.1 is globally well-posed ... Newton derivative J_Fε_h is invertible with uniformly bounded inverse

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Forward citations

Cited by 1 Pith paper

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

  1. A $\operatorname{prox}$-Based Semi-Smooth Newton Method for Convex Variational Problems

    math.OC 2026-06 unverdicted novelty 5.0 of 10

    A prox-based semi-smooth Newton method is proposed for finite-element discretizations of convex variational problems, with global well-posedness and local superlinear convergence established under suitable assumptions...

Reference graph

Works this paper leans on

53 extracted references · 53 canonical work pages · cited by 1 Pith paper

  1. [1]

    H. Antil,S. Bartels,A. Kaltenbach, andR. Khandelwal, Variational problems with gradient constraints: A priori and a posteriori error identities,Math. Comput.(2025). doi:10.1090/mcom/4146

    work page doi:10.1090/mcom/4146 2025

  2. [2]

    Attouch,G

    H. Attouch,G. Buttazzo, andG. Michaille,Variational analysis in Sobolev and BV spaces. Applications to PDEs and optimization, 2nd revised ed.,MOS-SIAM Ser. Optim.17, SIAM; MOS,

    work page

  3. [3]

    doi:10.1137/1.9781611973488

    work page doi:10.1137/1.9781611973488

  4. [4]

    AuslenderandM

    A. AuslenderandM. Teboulle, Interior gradient and proximal methods for convex and conic optimization,SIAM J. Optim.16no. 3 (2006), 697–725. doi:10.1137/S1052623403427823

    work page doi:10.1137/s1052623403427823 2006

  5. [5]

    Balay et al.,PETSc Users Manual, Tech

    S. Balay et al.,PETSc Users Manual, Tech. Report ANL-95/11 - Revision 3.6, Argonne National Laboratory, 2015

    work page 2015

  6. [6]

    BartelsandA

    S. BartelsandA. Kaltenbach,A prox-Based Semi-Smooth Newton Method for Convex Varia- tional Problems, in preparation

    work page

  7. [7]

    Topological properties of the set of functions generated by neuralnetworksoffixedsize

    S. BartelsandA. Kaltenbach, Explicit A Posteriori Error Representation for Variational Problems and Application to TV-Minimization,Found. Comput. Math.(2024). doi:10.1007/s10208- 024-09676-5

    work page doi:10.1007/s10208- 2024

  8. [8]

    Bartels, The Allen–Cahn Equation, in Numerical Methods for Nonlinear Partial Differential Equations, S

    S. Bartels,Numerical methods for nonlinear partial differential equations,Springer Ser. Comput. Math.47, Cham: Springer, 2015. doi:10.1007/978-3-319-13797-1 1

    work page doi:10.1007/978-3-319-13797-1 2015

Show all 53 references
  1. [9]

    Bartels, Nonconforming discretizations of convex minimization problems and precise relations to mixed methods,Comput

    S. Bartels, Nonconforming discretizations of convex minimization problems and precise relations to mixed methods,Comput. Math. Appl.93(2021), 214–229. doi:10.1016/j.camwa.2021.04.014

    2021 doi

  2. [10]

    Bartels,L

    S. Bartels,L. Diening, andR. H. Nochetto, Unconditional stability of semi-implicit discretiza- tions of singular flows,SIAM J. Numer. Anal.56no. 3 (2018), 1896–1914. doi:10.1137/17M1159166

    2018 doi

  3. [11]

    Bartels,T

    S. Bartels,T. Gudi, andA. Kaltenbach, A priori and a posteriori error identities for the scalar Signorini problem,SIAM J. Numer. Anal.63no. 5 (2025), 2155–2186. doi:10.1137/24M1677691

    2025 doi

  4. [12]

    BartelsandA

    S. BartelsandA. Kaltenbach, Error estimates for total-variation regularized minimization problems with singular dual solutions,Numer. Math.152no. 4 (2022), 881–906. doi:10.1007/s00211- 022-01324-w

    2022 doi

  5. [13]

    BartelsandA

    S. BartelsandA. Kaltenbach, Explicit and efficient error estimation for convex minimization problems,Math. Comput.92no. 343 (2023), 2247–2279. doi:10.1090/mcom/3821

    2023 doi

  6. [14]

    Bartels,R

    S. Bartels,R. Tovey, andF. Wassmer, Singular solutions, graded meshes, and adaptivity for total-variation regularized minimization problems,ESAIM, Math. Model. Numer. Anal.56no. 6 (2022), 1871–1888. doi:10.1051/m2an/2022056

    2022 doi

  7. [15]

    H. H. BauschkeandP. L. Combettes,Convex analysis and monotone operator theory in Hilbert spaces, 2nd ed.,CMS Books Math./Ouvrages Math. SMC, Cham: Springer, 2017. doi:10.1007/978-3- 319-48311-5

    2017 doi

  8. [16]

    Beck,First-order methods in optimization,MOS-SIAM Ser

    A. Beck,First-order methods in optimization,MOS-SIAM Ser. Optim.25, SIAM; Philadelphia, MOS, 2017. doi:10.1137/1.9781611974997

    2017 doi

  9. [17]

    BeckandM

    A. BeckandM. Teboulle, Fast gradient-based algorithms for constrained total variation image denoising and deblurring problems,IEEE Trans. Image Process.18no. 11 (2009), 2419–2434. doi:10.1109/TIP.2009.2028250

    2009 doi

  10. [18]

    BeckandM

    A. BeckandM. Teboulle, A fast iterative shrinkage-thresholding algorithm for linear inverse problems,SIAM J. Imaging Sci.2no. 1 (2009), 183–202. doi:10.1137/080716542

    2009 doi

  11. [19]

    BoykovandV

    Y. BoykovandV. Kolmogorov, An experimental comparison of min-cut/max-flow algorithms for energy minimization in vision, inEnergy Minimization Methods in Computer Vision and Pattern Recognition,Lecture Notes in Computer Science2134, Springer, 2001, pp. 359–374. doi:10.1007/3- 5...

    2001 doi

  12. [20]

    Brezis,Functional analysis, Sobolev spaces and partial differential equations,Universitext, New York, NY: Springer, 2011

    H. Brezis,Functional analysis, Sobolev spaces and partial differential equations,Universitext, New York, NY: Springer, 2011. doi:10.1007/978-0-387-70914-7

    2011 doi

  13. [21]

    Chambolle, An algorithm for total variation minimization and applications,J

    A. Chambolle, An algorithm for total variation minimization and applications,J. Math. Imaging Vis.20no. 1-2 (2004), 89–97. doi:10.1023/B:JMIV.0000011325.36760.1e

    2004 doi

  14. [22]

    ChambolleandT

    A. ChambolleandT. Pock, A first-order primal-dual algorithm for convex problems with applications to imaging,J. Math. Imaging Vis.40no. 1 (2011), 120–145. doi:10.1007/s10851-010- 0251-1

    2011 doi

  15. [23]

    ChambolleandT

    A. ChambolleandT. Pock, Crouzeix-raviart approximation of the total variation on simplicial meshes,J. Math. Imaging Vis.62no. 6-7 (2020), 872–899. doi:10.1007/s10851-019-00939-3. prox-based semi-smooth Newton method for TV-minimization21

    2020 doi

  16. [24]

    T. F. Chan,G. H. Golub, andP. Mulet, A nonlinear primal-dual method for to- tal variation-based image restoration,SIAM J. Sci. Comput.20no. 6 (1999), 1964–1977. doi:10.1137/S1064827596299767

    1999 doi

  17. [25]

    T. F. Chan,H. M. Zhou, andR. H.-F. Chan, Continuation method for total variation denoising problems, inAdvanced Signal Processing Algorithms(F. T. Luk, ed.),2563, International Society for Optics and Photonics, SPIE, 1995, pp. 314–325. doi:10.1117/12.211408

    1995 doi

  18. [26]

    X. Chen,Z. Nashed, andL. Qi, Smoothing methods and semismooth methods for nondifferentiable operator equations,SIAM J. Numer. Anal.38no. 4 (2000), 1200–1216. doi:10.1137/S0036142999356719

    2000 doi

  19. [27]

    DarbonandM

    J. DarbonandM. Sigelle, A Fast and Exact Algorithm for Total Variation Minimization, in Pattern Recognition and Image Analysis(J. S. Marques,N. P ´erez de la Blanca, andP. Pina, eds.), Springer Berlin Heidelberg, Berlin, Heidelberg, 2005, pp. 351–359. doi:10.1007/11492429 43

    2005 doi

  20. [28]

    DieningandJ

    L. DieningandJ. Storn, Guaranteed upper bounds for iteration errors and modified Kaˇ canov schemes via discrete duality,Comput. Methods Appl. Math.25no. 3 (2025), 587–600. doi:10.1515/cmam-2025-0017

    2025 doi

  21. [29]

    D. C. DobsonandC. R. Vogel, Convergence of an iterative method for total variation denoising, SIAM J. Numer. Anal.34no. 5 (1997), 1779–1791. doi:10.1137/S003614299528701X

    1997 doi

  22. [30]

    EkelandandR

    I. EkelandandR. T ´emam,Convex analysis and variational problems,Class. Appl. Math.28, SIAM, 1999. doi:10.1137/1.9781611971088

    1999 doi

  23. [31]

    GilbargandN

    D. GilbargandN. S. Trudinger,Elliptic partial differential equations of second order, reprint of the 1998 ed.,Class. Math., Berlin: Springer, 2001

    1998

  24. [32]

    GoldfarbandW

    D. GoldfarbandW. Yin, Second-order cone programming methods for total variation-based image restoration,SIAM J. Sci. Comput.27no. 2 (2005), 622–645 (English). doi:10.1137/040608982

    2005 doi

  25. [33]

    GoldsteinandS

    T. GoldsteinandS. Osher, The split Bregman method for ℓ1-regularized problems,SIAM J. Imaging Sci.2no. 2 (2009), 323–343 (English). doi:10.1137/080725891

    2009 doi

  26. [34]

    GriesseandD

    R. GriesseandD. A. Lorenz, A semismooth Newton method for Tikhonov functionals with sparsity constraints,Inverse Probl.24no. 3 (2008), 19, Id/No 035007. doi:10.1088/0266-5611/24/3/035007

    2008 doi

  27. [35]

    Gr¨oger, A W 1,p-estimate for solutions to mixed boundary value problems for second order elliptic differential equations,Math

    K. Gr¨oger, A W 1,p-estimate for solutions to mixed boundary value problems for second order elliptic differential equations,Math. Ann.283no. 4 (1989), 679–687. doi:10.1007/BF01442860

    1989 doi

  28. [36]

    Hinterm¨uller,K

    M. Hinterm¨uller,K. Ito, andK. Kunisch, The primal-dual active set strategy as a semismooth Newton method,SIAM J. Optim.13no. 3 (2003), 865–888. doi:10.1137/S1052623401383558

    2003 doi

  29. [37]

    Hinterm ¨ullerandK

    M. Hinterm ¨ullerandK. Kunisch, Total bounded variation regularization as a bilater- ally constrained optimization problem,SIAM J. Appl. Math.64no. 4 (2004), 1311–1333. doi:10.1137/S0036139903422784

    2004 doi

  30. [38]

    Hinterm ¨ullerandG

    M. Hinterm ¨ullerandG. Stadler, An infeasible primal-dual algorithm for total bounded variation–based Inf-convolution-type image restoration,SIAM J. Sci. Comput.28no. 1 (2006), 1–23. doi:10.1137/040613263

    2006 doi

  31. [39]

    P. J. Huber, Robust estimation of a location parameter,Ann. Math. Stat.35(1964), 73–101. doi:10.1214/aoms/1177703732

    1964 doi

  32. [40]

    J. D. Hunter, Matplotlib: A 2D Graphics Environment,Computing in Science & Engineering9 no. 3 (2007), 90–95. doi:10.1109/MCSE.2007.55

    2007 doi

  33. [41]

    KaltenbachandT

    A. KaltenbachandT. Schreiber,Learning Proximity-Certified Warmstarts for TV-minimization, in preparation

  34. [42]

    Logg,K.-A

    A. Logg,K.-A. Mardal, andG. Wells(eds.),Automated solution of differential equations by the finite element method. The FEniCS book,Lect. Notes Comput. Sci. Eng.84, Berlin: Springer,

  35. [43]

    doi:10.1007/978-3-642-23099-8. [42]A. MarquinaandS. Osher, Explicit algorithms for a new time dependent model based on level set motion for nonlinear deblurring and noise removal,SIAM J. Sci. Comput.22no. 2 (2000), 387–405. doi:10.1137/S1064827599351751

    2000 doi

  36. [44]

    M. K. Ng,L. Qi,Y.-f. Yang, andY.-m. Huang, On semismooth Newton’s methods for total variation minimization,J. Math. Imaging Vis.27no. 3 (2007), 265–276. doi:10.1007/s10851-007- 0650-0

    2007 doi

  37. [45]

    Rademacher, ¨Uber partielle und totale Differenzierbarkeit von Funktionen mehrerer Vari- ablen und ¨ uber die Transformation der Doppelintegrale

    H. Rademacher, ¨Uber partielle und totale Differenzierbarkeit von Funktionen mehrerer Vari- ablen und ¨ uber die Transformation der Doppelintegrale. I, II.,Math. Ann.79(1920), 340–359. doi:10.1007/BF01498415

    1920 doi

  38. [46]

    L. I. Rudin,S. Osher, andE. Fatemi, Nonlinear total variation based noise removal algorithms, Physica D60no. 1-4 (1992), 259–268. doi:10.1016/0167-2789(92)90242-F. S. Bartels and A. Kaltenbach22

    1992 doi

  39. [47]

    Souiai,Newton Methods for Total Variation Minimization, Master’s thesis, Rheinische Friedrich- Wilhelms-Universit¨ at Bonn, 2010

    M. Souiai,Newton Methods for Total Variation Minimization, Master’s thesis, Rheinische Friedrich- Wilhelms-Universit¨ at Bonn, 2010

    2010

  40. [48]

    X.-C. Tai,R. Winther,X. Zhang, andW. Zheng, A uniform preconditioner for a Newton algorithm for total variation minimization and minimum-surface problems,SIAM J. Numer. Anal. 61no. 5 (2023), 2062–2083. doi:10.1137/22M1512776

    2023 doi

  41. [49]

    Ulbrich, Semismooth Newton methods for operator equations in function spaces,SIAM J

    M. Ulbrich, Semismooth Newton methods for operator equations in function spaces,SIAM J. Optim.13no. 3 (2003), 805–841. doi:10.1137/S1052623400371569

    2003 doi

  42. [50]

    C. R. VogelandM. E. Oman, Iterative methods for total variation denoising,SIAM J. Sci. Comput.17no. 1 (1996), 227–238. doi:10.1137/0917016

    1996 doi

  43. [51]

    WuandX.-C

    C. WuandX.-C. Tai, Augmented Lagrangian method, dual methods, and split Bregman iteration for ROF, vectorial TV, and high order models,SIAM J. Imaging Sci.3no. 3 (2010), 300–339. doi:10.1137/090767558

    2010 doi

  44. [52]

    Zeidler,Nonlinear functional analysis and its applications

    E. Zeidler,Nonlinear functional analysis and its applications. II/B: Nonlinear monotone operators. Transl. from the German by the author and by Leo F. Boron, New York etc.: Springer-Verlag, 1990

    1990

  45. [53]

    M. ZhuandT. F. Chan,An Efficient Primal–Dual Hybrid Gradient Algorithm for Total Variation Image Restoration, CAM Report 08-34, UCLA, Department of Mathematics, 2008. Appendix In this appendix, we provide the proof of Theorem 3.6(3.26). Proof (of Theorem 3.6(3.26)). ad (3.26a)...

    2008

Pith tools

Reviewed May 22, 2026 · model on record in the stance chip above.