Pith. sign in

REVIEW 2 major objections 3 minor 30 references

$\Gamma$-Convergence of Weak-Type Nonlocal Functionals on Bounded Domains

T0 review · 2 major / 3 minor · reviewed 2026-08-27 · deepseek-v4-flash

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

Pith's one-line read The paper proves that, as the threshold tends to infinity, a weak-type nonlocal energy Gamma-converges to a universal constant times the p-Dirichlet energy or total variation, settling the bounded-domain question and extending it to all…

desk verdict The p=1 Gamma-liminf proof is unsound: a simple diagonal sequence of piecewise-constant interpolants of the affine profile contradicts the imported Lemma 2.4, so the paper's claimed lower bound and BV inference collapse. read the letter →

arxiv 2608.18414 v1 pith:ENA2IWZX submitted 2026-08-19 math.CA math.AP

Xiaosheng Lin , Dachun Yang , Sibei Yang , Wen Yuan , Yangyang Zhang This is my paper · ORCID
classification math.CAmath.AP MSC 49J4546E3526B3026D10
keywords Gamma-convergenceweak-typenonlocalfunctionalSobolevenergytotalvariationboundedcellformulaapproximationaffineconstant
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper establishes the variational limit, as the threshold $\lambda$ tends to infinity, of a weak-type nonlocal functional $G_{\lambda,p,\gamma}$ on bounded domains. The family $\Gamma$-converges in $L^p(\Omega)$ to a local functional $\Psi^{\mathrm{cell}}_{p,\gamma}$: a universal constant $C^{\mathrm{cell}}_{N,p,\gamma}$ times $\int_\Omega |\nabla u|^p\,dx$ when $p>1$ and $u\in W^{1,p}(\Omega)$, the same constant times the total variation $|Du|(\Omega)$ when $p=1$ and $u\in BV(\Omega)$, and infinity otherwise. The constant is characterized by an affine cell formula on the unit cube and is independent of the bounded domain $\Omega$; the theorem covers every $p\in[1,\infty)$ and every $\gamma\in(0,\infty)$. This resolves the bounded-domain $\Gamma$-convergence question for this family that was posed in a 2023 open-problem list, and it extends the earlier formulation, which had only considered $\gamma=N$, to all positive $\gamma$. The result matters because it turns a nonconvex, threshold-counting energy into a rigorous approximation of the classical Sobolev and total-variation energies, with the asymptotic energy density pinned down by a domain-independent cell constant.

What carries the argument

The load-bearing object is the affine cell constant $C^{\mathrm{cell}}_{N,p,\gamma}=m_{p,\gamma}(\ell)$, the minimal asymptotic energy needed to approximate the affine function $\ell(x)=x_N$ on the unit cube $Q_0$ as $\lambda\to\infty$. For the lower bound with $p>1$, the proof compares $G$ on small rotated cubes with this cell constant: at almost every Lebesgue point of $\nabla u$ with nonzero gradient, normalized blow-ups of $u$ converge to $\ell$, and a covering lemma converts the pointwise estimate into the global integral lower bound. For $p=1$, the blow-ups can converge to one-dimensional nondecreasing profiles $W_q(z)=q(z_N)$, so the proof must show the cell inequality $m_{1,\gamma}(W_q)\ge C^{\mathrm{cell}}_{N,1,\gamma}|Dq|(I_0)$ for every nondecreasing $q\in BV(I_0)$; this is the technical heart, using a truncation and partition argument, a boundary-layer energy estimate, and a cross-cube energy estimate. For the upper bound, near-minimizing sequences for the affine cell constant are modified so that they equal $\ell$ near the boundary of the cube, then scaled copies are glued into a continuous piecewise affine approximation of $u$; the boundary agreement suppresses interaction energy between neighbouring cubes.

What would settle it

Compute the cell energy of the single-jump profile $W_{q_j}(z)=\mathbf{1}_{\{z_N>0\}}-\tfrac12$ on $Q_0$, namely $m_{1,\gamma}(W_{q_j})$, whose total variation is $1$; if this number is strictly smaller than $C^{\mathrm{cell}}_{N,1,\gamma}$, then Proposition 4.11 is false and the claimed $\Gamma$-limit cannot hold. More generally, any sequence $v_j\to W_q$ in $L^1(Q_0)$ with $\liminf_j G_{\lambda_j,1,\gamma}(v_j;Q_0)<C^{\mathrm{cell}}_{N,1,\gamma}|Dq|(I_0)$ for some nondecreasing $q$ would break the lower bound.

Watch

Extended reading notes

Core claim

The central discovery is that the $\Gamma$-limit of the weak-type nonlocal energy is governed by affine profiles at small scales. For any sequence $\lambda_j\to\infty$ and any $u_j\to u$ in $L^p(\Omega)$, the liminf of $G_{\lambda_j,p,\gamma}(u_j;\Omega)$ is at least $\Psi^{\mathrm{cell}}_{p,\gamma}(u;\Omega)$, and for every $u\in W^{1,p}(\Omega)$ when $p>1$ (or $u\in BV(\Omega)$ when $p=1$) there is a recovery sequence $u_j\to u$ whose energy is asymptotically at most $\Psi^{\mathrm{cell}}_{p,\gamma}(u;\Omega)$. The constant $C^{\mathrm{cell}}_{N,p,\gamma}$ is defined as the infimum, over sequences $\lambda_j\to\infty$ and $u_j\to\ell$ in $L^p(Q_0)$, of $\liminf_j G_{\lambda_j,p,\gamma}(u_j;Q_0)$, where $Q_0=(-1/2,1/2)^N$ and $\ell(x)=x_N$ is the affine function on the unit cube; the proof shows this number is positive, finite, and independent of $\Omega$. The delicate case is $p=1$: normalized blow-ups of a $BV$ function at $|Du|$-almost every point converge to $W_q(z)=q(z_N)$ for some nondecreasing one-dimensional $BV$ function $q$ with $|Dq|(I_0)=1$, and the paper proves that the cell energy of every such profile is at least $C^{\mathrm{cell}}_{N,1,\gamma}|Dq|(I_0)$.

Load-bearing premise

The theorem stands on the claim that the affine cell constant is the true ground-state cost per unit of gradient or total variation, and in the $p=1$ case specifically on the cell inequality that every one-dimensional nondecreasing $BV$ profile $W_q$ costs at least $C^{\mathrm{cell}}_{N,1,\gamma}|Dq|(I_0)$; if any nondecreasing jump or Cantor profile were cheaper than the affine function, the $BV$ lower bound would fail.

Editorial extensions

If this is right

  • The weak-type nonlocal family $G_{\lambda,p,\gamma}$ is a valid variational approximation of the $p$-Dirichlet energy on bounded Lipschitz domains for every $\gamma\in(0,\infty)$, not only for $\gamma=N$.
  • For $p=1$, the $\Gamma$-limit exists and equals $C^{\mathrm{cell}}_{N,1,\gamma}|Du|(\Omega)$ even though the pointwise limit of $G_{\lambda,1,\gamma}$ on a fixed $BV$ function may fail to exist; variational convergence is more robust than pointwise convergence.
  • The limiting constant $C^{\mathrm{cell}}_{N,p,\gamma}$ is the same on every bounded interval ($N=1$) or bounded Lipschitz domain ($N\ge2$), so the asymptotic energy density of the threshold functional is a local, domain-independent quantity.
  • The recovery sequences are explicit: they are built from scaled copies of boundary-modified cell minimizers, giving a constructive route from nearly minimizing nonlocal configurations to Sobolev or $BV$ limit functions.
  • Since $\Gamma$-convergence passes minimizers and minimum values to the limit, adding lower-order perturbations to $G_{\lambda,p,\gamma}$ yields convergence of the corresponding nonlocal variational problems to the local Sobolev or total-variation problem.

Reading between the lines

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

  • The paper leaves the value of $C^{\mathrm{cell}}_{N,p,\gamma}$ implicit, giving only the variational cell characterization together with upper and lower bounds; a natural next step is to compute it in closed form, and by analogy with earlier nonlocal $\Gamma$-limits one testable conjecture is that the $\Gamma$-constant is strictly smaller than the pointwise constant for the affine profile.
  • Because the cell formula involves only the unit cube and the linear profile, the constant should be insensitive to the boundary geometry of $\Omega$; it is plausible that the bounded-Lipschitz hypothesis in the theorem can be relaxed to rougher bounded domains, since the lower bound is built from a covering argument and the upper bound from piecewise-affine density.
  • The $p=1$ cell inequality for nondecreasing one-dimensional profiles is an isolated one-dimensional statement; testing it numerically on a step profile, a Cantor profile, and a mixed profile would give quick evidence about strictness of the constant and about whether the same $C^{\mathrm{cell}}$ controls all blow-up profiles.
  • The constructive recovery sequences may be useful in imaging applications: they show that threshold-counting weak-type nonlocal energies, which are robust to outliers, can be tuned to approximate total-variation functionals while preserving their nonconvex weak-type structure.
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, and a circularity audit.

Referee Report

2 major / 3 minor

Summary. The paper proves that, for N≥1, p∈[1,∞), γ∈(0,∞), and Ω a bounded interval (N=1) or bounded Lipschitz domain (N≥2), the weak-type nonlocal functionals G_{λ,p,γ}(·;Ω) Γ-converge in L^p(Ω), as λ→∞, to a functional Ψ^{cell}_{p,γ} which is a positive multiple, given by a cell constant C^{cell}_{N,p,γ}, of the Sobolev energy for p>1 and of the total variation for p=1. The upper bound is obtained by modifying sequences realizing the affine cell constant and then using piecewise-affine approximation; the lower bound for p>1 uses blow-up, Poincaré's inequality, and a Morse covering argument, while the p=1 lower bound uses the fine structure of BV functions and a cell lower bound for one-dimensional nondecreasing profiles. The paper claims this answers Brezis Open Problem 9.3 and extends it to all γ>0.

Significance. If the proof were correct, this would be a substantial and valuable result: it settles a well-known open problem, identifies the Γ-limit constant through a parameter-free cell formula, and handles the delicate p=1 case where pointwise limits may fail. The paper is detailed, self-contained in its main arguments, and gives explicit constructions for the recovery sequences. However, the correctness of the p=1 lower bound and of the positivity of the cell constant is undermined by a false imported lemma, as detailed below.

major comments (2)
  1. [Section 2, Lemma 2.4] Lemma 2.4 is false as stated. For N=1, p=1, γ=1/2, let I0=(-1/2,1/2), ℓ(x)=x, and for each m∈N let u_m be the piecewise-constant left-endpoint interpolation of ℓ on a uniform mesh of size δ=1/m. Then u_m→ℓ in L^1(I0), and taking λ_m=δ^{-2γ}=m, only adjacent intervals contribute to G_{λ_m,1,γ}(u_m;I0). A direct computation gives G_{λ_m,1,γ}(u_m;I0)=2(m-1)δ/(γ+1)= (4/3)(1-1/m), so liminf_{m→∞} G_{λ_m,1,γ}(u_m;I0)=4/3. Thus C^{cell}_{1,1,1/2}≤4/3. On the other hand, Lemma 2.4 with K_{1,1}=2 yields the lower bound 2/log2(2^{3/2}-1)≈2.30>4/3 for any admissible sequence, a contradiction. The same construction is easily adapted to other γ>0. Since Lemma 2.4 is used in Lemma 2.5(ii) to prove the lower bound (2.6) for C^{cell} and in Theorem 4.2 to infer u∈BV(Ω) from finite liminf, the p=1 Γ-liminf proof is invalid as written. The authors need to correct Lemma 2.4 (or replace it by a valid lower bound) and re-examine the consequences for positivity and BV membership.
  2. [Section 4.4, Theorem 4.2] The proof of Theorem 4.2 relies essentially on Lemma 2.4 for the step 'By Lemma 2.4, we conclude that u∈BV(Ω)'. Since Lemma 2.4 is false, the p=1 liminf theorem lacks a valid proof. The covering argument and Proposition 4.11 do not by themselves supply the required coercivity. A repair is possible in principle, for instance by proving a correct positive lower bound with a smaller constant, but as it stands the central lower-bound claim for p=1 is unsupported.
minor comments (3)
  1. [Section 2, Lemma 2.2] The scaling identity in Lemma 2.2 appears to have the parameter inverted. With u(x0+rRz)=arw(z)+b, a direct change of variables gives G_{λ,p,γ}(u;x0+rRU)=|a|^p r^N G_{λ r^γ/|a|^p,p,γ}(w;U), not G_{r^γ|a|^p λ,p,γ}(w;U). The later applications, e.g., (3.26) and (4.59), use the former correct form, so this is likely a typographical error, but the lemma as printed is false.
  2. [Section 4.2, proof of Lemma 4.9] After equation (4.20), the sentence 'Equivalently, σ_r converges weakly to σ as r→∞' should read 'as r→0+'; the parameter is the shrinking radius, not the limit.
  3. [Throughout] The typesetting contains many spacing artifacts (e.g., 'LetN≥1', 'nota-tion', 'Brezis et al. [10] proved') that should be cleaned before publication.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the cell constant is variationally defined and both Gamma-liminf and Gamma-limsup are proved by independent cell arguments.

full rationale

The paper defines C_cell as the infimum of asymptotic cell energies over sequences converging to the affine function ℓ (Definition 2.1), and then proves the Gamma-liminf inequality (Theorems 4.1 and 4.2) and the Gamma-limsup inequality (Theorems 3.1 and 3.2) using that same variationally characterized constant. This is the standard cell-formula methodology for Gamma-convergence, not a reduction of the claimed result to its own input. The lower-bound proofs do not assume the target Gamma-limit: they obtain the constant by lower semicontinuity of m_{p,gamma} and, in the p=1 case, by Proposition 4.11, which transfers the lower bound from ℓ to arbitrary nondecreasing one-dimensional BV profiles through explicit gluing and comparison lemmas (Lemmas 4.12, 4.15, and 4.16). The external results used in the p=1 chain, Lemma 2.4 from Gobbino and Picenni [20] and Lemma 4.10 from Conti, Focardi, and Iurlano [11], are not author self-citations. Author self-citations in the introduction (for example, [12]-[14] and [30]) are contextual and are not load-bearing for the main derivation. The purported contradiction between an explicit p=1 sequence and Lemma 2.4 for γ<1 is a mathematical-correctness concern, not a circularity: it does not exhibit an equation that reduces to itself by construction, nor a fitted parameter renamed as a prediction. No circular step is present.

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

No free parameters are fitted to data. The only constant, C^cell, is defined as the infimum of a cell variational problem, not chosen by hand. No new physical or geometric entities are introduced. The central claim rests on standard results in BV theory and on two imported domain-specific results from [20] and [11].

assumptions (7)
  • standard math BV structure theorem: Du = grad u L^N + D^j u + D^c u with mutually singular parts, and rectifiability of J_u (Lemma 4.5, from [1]).
    Used in Lemma 4.4 to split blow-up analysis into absolutely continuous, jump, and Cantor parts.
  • domain assumption Gobbino-Picenni liminf bound [20, Theorem 1.1]: liminf_j G_{lambda_j,p,gamma}(u_j;U) is at least K_{N,p} log2/(2^{gamma+1}-1) times the Sobolev/BV energy.
    Imported as Lemma 2.4; it supplies the Gamma-liminf for limits outside W^{1,p}/BV and the lower bound for C^cell in Lemma 2.5(ii).
  • standard math Density of continuous piecewise affine functions in W^{1,p}(Omega) on bounded Lipschitz domains (Lemma 3.8, from [18, Prop 2.8] and [29, Thm 1]).
    Reduces the Gamma-limsup from general Sobolev functions to piecewise affine functions, where recovery sequences can be glued cube-by-cube.
  • standard math Strict BV approximation: every u in BV(Omega) is the L^1 limit of W^{1,1} cap C^infinity functions with total variation converging to |Du|(Omega) (Lemma 3.9, from [19, Thm 5.3]).
    Reduces the p=1 Gamma-limsup to the Sobolev case followed by a diagonal argument.
  • domain assumption Cantor-part blow-up lemma: for |D^c u|-almost every x, normalized blow-ups converge strictly in BV to W_{q_x^c} for a nondecreasing q_x^c (Lemma 4.10, from [11, Lemma 4.5]).
    This imported lemma is essential for the p=1 Gamma-liminf at Cantor points; its convergence mode and monotonicity feed directly into Proposition 4.11.
  • standard math Poincare inequality on cubes in the form ||g - average g||_{L^p(Q)} is bounded by a constant times diam(Q) ||grad g||_{L^p(Q)}.
    Used in (4.4) and (4.13) to show normalized blow-ups of Sobolev functions approach affine functions at Lebesgue points.
  • standard math Morse-type covering lemma for finite positive Radon measures (Lemma 4.3, from [1, Theorem 5.51]).
    Converts local cube lower bounds into the global Gamma-liminf lower bound.

how reviews work

0 comments
Cite this review

Pith. "Pith review of $\Gamma$-Convergence of Weak-Type Nonlocal Functionals on Bounded Domains." pith.science (2026). https://pith.science/paper/ENA2IWZX

@misc{pith2026260818414,
  author       = {Pith},
  title        = {Pith review of: $\Gamma$-Convergence of Weak-Type Nonlocal Functionals on Bounded Domains},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ENA2IWZX}},
  note         = {Machine review of arXiv:2608.18414}
}
abstract

Let $N\ge1$, $p\in[1,\infty)$, $\gamma\in(0,\infty)$, and $\Omega\subset\mathbb R^N$ be a bounded open interval when $N=1$ or a bounded Lipschitz domain when $N\ge2$. For any $\lambda\in(0,\infty)$ and any measurable function $u$, consider the weak-type nonlocal functional \begin{align*} G_{\lambda,p,\gamma}(u;\Omega) :=\lambda\iint_{\Omega\times\Omega} \mathbf 1_{\left\{(x,y)\in\Omega\times\Omega:\ x\neq y,\ \frac{|u(x)-u(y)|^p}{|x-y|^{p+\gamma}}\geq\lambda\right\}} |x-y|^{\gamma-N}\,dx\,dy. \end{align*} In this article, we prove that, as $\lambda\to\infty$, the family $G_{\lambda,p,\gamma}$ converges, in the sense of $\Gamma$-convergence in $L^p(\Omega)$, to the functional \begin{align*} \Psi_{p,\gamma}^{\mathrm{cell}}(u;\Omega):= \begin{cases} C_{N,p,\gamma}^{\mathrm{cell}}\displaystyle\int_\Omega|\nabla u|^p\,dx, &p\in(1,\infty)\ \hbox{and}\ u\in W^{1,p}(\Omega),\\[2mm] C_{N,1,\gamma}^{\mathrm{cell}}|Du|(\Omega), &p=1\ \hbox{and}\ u\in BV(\Omega),\\[1mm] \infty,&\hbox{otherwise}, \end{cases} \end{align*} where the positive constants $C_{N,p,\gamma}^{\mathrm{cell}}$ are independent of $\Omega$ and characterized by a cell formula. This gives an affirmative answer to the problem posed by Brezis [Open Problem~9.3, Rend. Lincei Mat. Appl. 2023].

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 30 canonical work pages

  1. [20]

    Gobbino and N

    M. Gobbino and N. Picenni, Gamma-liminf estimate for a class of non-local approximations of Sobolev and BV norms, J. Funct. Anal. 289 (2025), Paper No. 111106, 25 pp. 3, 4, 6, 7 42XiaoshengLin, DachunYang, SibeiYang, WenYuan andYangy angZhang

    work page 2025

  2. [1]

    Ambrosio, N

    L. Ambrosio, N. Fusco and D. Pallara, Functions of Bounded Variation and Free Discontinu- ity Problems, Oxford Mathematical Monographs, Oxford University Press, New York, 2000. 2, 18, 21, 22

    work page 2000

  3. [2]

    Antonucci, M

    C. Antonucci, M. Gobbino, M. Migliorini and N. Picenni, Optimal constants for a nonlocal approximation of Sobolev norms and total variation, Anal. PDE 13 (2020), 595–625. 3

    work page 2020

  4. [3]

    Bourgain, H

    J. Bourgain, H. Brezis and P. Mironescu, Another look at Sobolev spaces, In: Optimal Con- trol and Partial Differential Equations, pp. 439–455, IOS, Amsterdam, 2001. 2

    work page 2001

  5. [4]

    Brezis, How to recognize constant functions

    H. Brezis, How to recognize constant functions. A connection with Sobolev spaces, Russian Math. Surveys 57 (2002), 693–708. 2

    work page 2002

  6. [5]

    Brezis, Some of my favorite open problems, Rend

    H. Brezis, Some of my favorite open problems, Rend. Lincei Mat. Appl. 34 (2023), 307–335. 3

    work page 2023

  7. [6]

    Brezis and H.-M

    H. Brezis and H.-M. Nguyen, Non-local functionals related to the total variation and connec- tions with image processing, Ann. PDE 4 (2018), Paper No. 9, 77 pp. 2, 3

    work page 2018

  8. [7]

    Brezis and H.-M

    H. Brezis and H.-M. Nguyen,Γ-convergence of non-local, non-convex functionals in one dimension, Commun. Contemp. Math. 22 (2020), 1950077, 27 pp. 3

    work page 2020

Show all 30 references
  1. [8]

    Brezis and H.-M

    H. Brezis and H.-M. Nguyen, Non-local, non-convex functionals converging to Sobolev norms, Nonlinear Anal. 191 (2020), 111626, 9 pp. 3

    2020

  2. [9]

    Brezis, A

    H. Brezis, A. Seeger, J. Van Schaftingen and P.-L. Yung, Families of functionals representing Sobolev norms, Anal. PDE 17 (2024), 943–979. 2, 31

    2024

  3. [10]

    Brezis, J

    H. Brezis, J. Van Schaftingen and P.-L. Yung, A surprising formula for Sobolev norms, Proc. Natl. Acad. Sci. USA 118 (2021), e2025254118, 6 pp. 2

    2021

  4. [11]

    Conti, M

    S. Conti, M. Focardi and F. Iurlano, Phase-field approximation of a vectorial, geometrically nonlinear cohesive fracture energy, Arch. Ration. Mech. Anal. 248 (2024), Paper No. 21, 60 pp. 27

    2024

  5. [12]

    F. Dai, L. Grafakos, Z. Pan, D. Yang, W. Yuan and Y . Zhang, The Bourgain–Brezis– Mironescu formula on ball Banach function spaces, Math. Ann. 388 (2024), 1691–1768. 2

    2024

  6. [13]

    F. Dai, X. Lin, D. Yang, W. Yuan and Y . Zhang, Brezis–Van Schaftingen–Yung formulae in ball Banach function spaces with applications to fractional Sobolev and Gagliardo–Nirenberg inequalities, Calc. Var. Partial Differential Equations 62 (2023), Paper No. 56, 73 pp. 2

    2023

  7. [14]

    F. Dai, X. Lin, D. Yang, W. Yuan and Y . Zhang, Poincar ´e inequality meets Brezis–Van Schaftingen–Yung formula on metric measure spaces, J. Funct. Anal. 283 (2022), Paper No. 109645, 52 pp. 2

    2022

  8. [15]

    Dom ´ınguez and M

    O. Dom ´ınguez and M. Milman, New Brezis–Van Schaftingen–Yung–Sobolev type inequal- ities connected with maximal inequalities and one parameter families of operators, Adv. Math. 411 (2022), part A, Paper No. 108774, 76 pp. 2

    2022

  9. [16]

    Dom ´ınguez and M

    O. Dom ´ınguez and M. Milman, Bourgain–Brezis–Mironescu–Maz’ya–Shaposhnikova limit formulae for fractional Sobolev spaces via interpolation and extrapolation, Calc. Var. Partial Differential Equations 62 (2023), Paper No. 43, 37 pp. 2

    2023

  10. [17]

    Dom ´ınguez, A

    O. Dom ´ınguez, A. Seeger, B. Street, J. Van Schaftingen and P.-L. Yung, Spaces of Besov– Sobolev type and a problem on nonlinear approximation, J. Funct. Anal. 284 (2023), Paper No. 109775, 50 pp. 2

    2023

  11. [18]

    Ekeland and R

    I. Ekeland and R. T ´emam, Convex Analysis and Variational Problems, Classics in Applied Mathematics 28, Society for Industrial and Applied Mathematics, Philadelphia, PA, 1999. 17

    1999

  12. [19]

    L. C. Evans and R. F. Gariepy, Measure Theory and Fine Properties of Functions, Revised edition, Textbooks in Mathematics, CRC Press, Boca Raton, FL, 2015. 17

    2015

  13. [21]

    G. E. Karadzhov, M. Milman and J. Xiao, Limits of higher-order Besov spaces and sharp reiteration theorems, J. Funct. Anal. 221 (2005), 323–339. 2

    2005

  14. [22]

    Lahti, A sharp lower bound for a class of non-local approximations of the total variation, Math

    P. Lahti, A sharp lower bound for a class of non-local approximations of the total variation, Math. Ann. 392 (2025), 469–486. 4

    2025

  15. [23]

    Lahti and Z

    P. Lahti and Z. Li, Nonlocal functionals, total variation, and Gamma-convergence with re- spect to area-strict convergence, SIAM J. Math. Anal. 58 (2026), 2675–2701. 3

    2026

  16. [24]

    Nguyen,Γ-convergence and Sobolev norms, C

    H.-M. Nguyen,Γ-convergence and Sobolev norms, C. R. Math. Acad. Sci. Paris 345 (2007), 679–684. 3

    2007

  17. [25]

    Nguyen,Γ-convergence, Sobolev norms, and BV functions, Duke Math

    H.-M. Nguyen,Γ-convergence, Sobolev norms, and BV functions, Duke Math. J. 157 (2011), 495–533. 3, 4

    2011

  18. [26]

    Nguyen, Characterizations of the Sobolev norms and the total variation via nonlocal functionals, and related problems, C

    H.-M. Nguyen, Characterizations of the Sobolev norms and the total variation via nonlocal functionals, and related problems, C. R. Math. Acad. Sci. Paris 363 (2025), 1429–1455. 3

    2025

  19. [27]

    Picenni, New estimates for a class of non-local approximations of the total variation, J

    N. Picenni, New estimates for a class of non-local approximations of the total variation, J. Funct. Anal. 287 (2024), Paper No. 110419, 21 pp. 4

    2024

  20. [28]

    A. C. Ponce, A new approach to Sobolev spaces and connections toΓ-convergence, Calc. Var. Partial Differential Equations 19 (2004), 229–255. 3

    2004

  21. [29]

    Van Schaftingen, Approximation in Sobolev spaces by piecewise affine interpolation, J

    J. Van Schaftingen, Approximation in Sobolev spaces by piecewise affine interpolation, J. Math. Anal. Appl. 420 (2014), 40–47. 17

    2014

  22. [30]

    C. Zhu, D. Yang and W. Yuan, Generalized Brezis–Seeger–Van Schaftingen–Yung formulae and their applications in ball Banach Sobolev spaces, Calc. Var. Partial Differential Equations 62 (2023), Paper No. 234, 76 pp. 2 Xiaosheng Lin School of Mathematical Sciences, Jimei Universi...

    2023

Pith tools

Reviewed August 27, 2026 · model on record in the stance chip above.