Gamma-convergence of nonlocal energies for partitions

Thomas Gabard; Vincent Millot

Gamma-convergence of nonlocal energies for partitions

T0 review · 0 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The fractional perimeter functionals for multiphase partitions converge to a local energy with relaxed surface-tension coefficients.

desk verdict A careful, self-contained Gamma-convergence result for multi-phase fractional perimeters with arbitrary surface tensions; the relaxed limit is identified correctly and the proof holds up. read the letter →

arxiv 2506.20215 v1 pith:YLNFW4IR submitted 2025-06-25 math.AP math.FA

Thomas Gabard , Vincent Millot This is my paper · ORCID

classification math.APmath.FA MSC 49Q2049J4535R11

keywords Gamma-convergencefractionalperimeterpartitionsnonlocalrelaxationsurfacetensionhalf-spaceminimalitymax-flowmin-cut

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 a sharp-interface limit for a family of nonlocal energies defined on partitions of space into $m\ge 3$ measurable chambers. For any matrix $\sigma$ of positive surface-tension coefficients, the rescaled fractional $\sigma$-perimeter $(1-2s)P^\sigma_{2s}(\cdot,\Omega)$ is shown to $\Gamma$-converge as $s\to 1/2^-$ to $\omega_{n-1}$ times the classical perimeter functional $P^{\bar\sigma}_1(\cdot,\Omega)$, where $\bar\sigma$ is the largest componentwise-lower matrix satisfying the triangle inequality. Because $P^{\bar\sigma}_1$ is the lower semicontinuous envelope of $P^\sigma_1$, the limit is well posed even when $\sigma$ itself violates the triangle inequality. The paper also proves that limits of local minimizers of the nonlocal energies are local minimizers of the relaxed local energy. A sympathetic reader should care because this pins down how interfacial energies with incompatible surface tensions relax through phase nucleation as the interaction range shrinks.

What carries the argument

The load-bearing object is the asymptotic cell formula: for the unit cube $Q$ and upper half-space $H$, the coefficients $\Gamma_{ij}$ are defined by the lowest possible rescaled energies of partitions that converge to the two-chamber configuration $(H,H^c)$ inside $Q$. The paper reworks this cell formula until it is solved by the half-space partition itself, using a replacement lemma: given any multi-phase competitor, the max-flow min-cut theorem on the complete directed graph whose vertices are the chambers and whose edge capacities are the pairwise interaction energies produces a two-phase competitor of no greater energy. This reduces the multi-phase cell problem to the known two-phase half-space minimality for the fractional perimeter. The upper bound is carried out by polyhedral partitions and a direct computation of the pointwise limit, together with a relaxation step that replaces $\sigma$ by $\bar\sigma$; the convergence of minimizers uses a co-area gluing construction to compare competitors without breaking the partition constraint.

What would settle it

Compute, for a five-chamber matrix that violates the triangle inequality but is not $\ell^1$-embeddable, the minimum of $(1-2s)P^\sigma_{2s}(E,Q)$ among partitions of the unit cube that agree with the half-space partition outside $Q$, for a sequence $s\to 1/2^-$; if the value does not approach $\omega_{n-1}\bar\sigma_{ij}$, or if any competitor strictly beats the half-space partition for some $s$, the central claim collapses.

Watch

Extended reading notes

Core claim

The central discovery is that the relaxation already visible in local partition perimeters appears automatically in the sharp-interface limit of the nonlocal ones. The proof identifies the limiting matrix explicitly: $\bar\sigma_{ij}$ is the infimum over chains $i=i_0,i_1,\dots,i_H=j$ of $\sigma_{i_0i_1}+\cdots+\sigma_{i_{H-1}i_H}$, equivalently the metric closure of $\sigma$. In the cell problem for a pair of chambers, the half-space partition is the unique minimizer, and the cell energy equals $\omega_{n-1}\bar\sigma_{ij}$; the multi-phase case is reduced to this two-phase statement by a replacement argument based on the max-flow min-cut theorem. Consequently the $\Gamma$-limit is $\omega_{n-1}P^{\bar\sigma}_1$, not $\omega_{n-1}P^\sigma_1$, and any deficiency created by the failure of the triangle inequality is healed by nucleating intermediate phases in the limit.

Load-bearing premise

The proof imports from the two-phase theory the fact that a half-space minimizes the fractional perimeter among sets agreeing with it outside a small cube; if that statement failed for some $s\in(0,1/2)$, the cell constant would not equal $\omega_{n-1}\bar\sigma_{ij}$ and the lower bound would not match the upper bound.

Editorial extensions

If this is right

If the theorem is correct, the sharp-interface limit of fractional multiphase perimeters is a local functional even for coefficients that violate the triangle inequality, and the limiting coefficients are the shortest-path relaxed ones.
Any sequence of local minimizers of the nonlocal energies converges in $L^1_{\mathrm{loc}}$ to a local minimizer of $\omega_{n-1}P^{\bar\sigma}_1$, with convergence of the rescaled energies on domains whose boundary has zero relaxed perimeter.
The explicit shortest-path formula for $\bar\sigma$ makes the relaxation process computable: an interface between phases $i$ and $j$ may split into a chain of intermediate phases whose total surface tension is cheaper.
Compactness holds for equi-bounded sequences: limits of finite-energy partitions are Caccioppoli partitions, so existence and regularity theory for the relaxed problem become available.
For $m=3$ and $m=4$, the triangle inequality forces additive or nearly additive structure, so the half-space minimality is proved by explicit decompositions; for general $m$ it follows from the network-flow argument.

Reading between the lines

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

Inference: By analogy with known threshold-dynamics schemes, the $\Gamma$-convergence here suggests that curvature-driven network motions dissipating $P^\sigma_{2s}$ should converge, in the vanishing-interaction limit, to the relaxation flow of $\omega_{n-1}P^{\bar\sigma}_1$; the paper supplies the variational half of that statement but does not construct the flow itself.
Inference: The replacement lemma is likely robust: for any pairwise interaction kernel for which the two-phase half-space minimality holds, the same max-flow min-cut reduction should yield a relaxed $\Gamma$-limit with the metric-closure coefficients, so the phenomenon is not specific to the fractional kernel.
Inference: A testable quantitative prediction is that for a fixed small $s$ and a matrix violating the triangle inequality, minimizers should form thin layers of intermediate phases near interfaces, with layer width tending to zero as $s\to 1/2^-$ and the energy gap to $\omega_{n-1}P^{\bar\sigma}_1$ of order $(1/2-s)$; direct numerical simulation could verify the scaling.

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.

Desk Editor's Note

read the letter

The paper does what it says: it proves the Gamma-convergence of (1-2s) times the fractional sigma-perimeter, for partitions, to the local perimeter with the relaxed coefficient matrix sigma-bar. That relaxed matrix is the shortest-path distance induced by sigma, and the main theorems (1.2 and 1.3) pin down both the Gamma-limit and the behavior of local minimizers. This is genuinely new: previous work covered the single-phase case or additive coefficients, and the relaxation effect for non-additive matrices was open.

The proof is thorough. The compactness argument is clean, the blow-up lower bound is standard but executed carefully, and the upper bound via polyhedral partitions and the separate relaxation-by-hand is a nice choice. The most interesting piece is the max-flow min-cut replacement lemma (Lemma 5.3), which reduces the multi-phase half-space minimality problem to the two-phase result from Ambrosio–De Philippis–Martinazzi. I checked the path-decomposition step and the use of the triangle inequality for the relaxed matrix; it works. The gluing argument in Section 6 is also handled with the right amount of care, using the generalized co-area formula without pretending the partition constraints disappear.

Soft spots: the proof leans heavily on [3, Proposition 17] (half-space minimality for the two-phase fractional perimeter). That is a published, standard theorem, and the reduction to it is sound, but it is load-bearing. If that result ever turned out to have a gap, parts of this paper would need revisiting. That is not a criticism of this paper, just a statement about what is imported. There are also a few typos and minor notational inconsistencies (e.g., some subscripts appear as '1,h' instead of '1,k'), but nothing that obscures the mathematics.

Overall, the central claim is well supported. The proof is detailed enough that I found no hidden assumption or post-hoc fitting. The authors also give proper credit to prior work and do not overclaim; the introduction is honest about what is new and what is borrowed.

This paper deserves a serious referee. I would recommend accepting it after minor revisions, and I would bring it to our reading group. It is a solid contribution to the calculus of variations and will be useful for anyone working on nonlocal perimeter models or grain growth.

Referee Report

0 major / 6 minor

Summary. The paper proves that, as s→1/2^-, the family of fractional multi-phase perimeter energies (1-2s)P^σ_{2s}(·,Ω) Γ-converges, in the L^1(Ω)-sense for partitions, to ω_{n-1}P^{\barσ}_1(·,Ω), where \barσ is the shortest-path relaxation of the surface-tension matrix σ. This holds for an arbitrary symmetric matrix with positive off-diagonal coefficients, without any triangle inequality assumption. The authors also prove a compactness theorem and a convergence result for local minimizers, showing that limits of local minimizers are local minimizers of the relaxed local functional. The proof strategy combines a blow-up lower bound, a cell-formula reduction through gluing lemmas, a half-space minimality theorem obtained by a max-flow/min-cut replacement argument, and an upper bound via polyhedral approximation and relaxation. The main novelty is that the limit functional is the lower-semicontinuous envelope of the classical multi-phase perimeter, so coefficients violating the triangle inequality are relaxed in the limit, which the authors interpret as phase nucleation.

Significance. If correct, this result settles the sharp-interface asymptotics for multi-phase fractional perimeter functionals with arbitrary surface tensions, a question that is central to the variational analysis of nonlocal minimal clusters and to threshold-dynamics models for grain growth. The identification of the relaxed matrix \barσ as the shortest-path metric closure is clean and makes the relaxation mechanism explicit. The proof is unusually detailed: the key estimates in Lemmas 3.5, 4.2, 4.3 and 5.3 are written out, the gluing construction in Section 6 is carried through carefully, and the reliance on external results is limited to standard facts such as the two-phase half-space minimality theorem of [3] and the polyhedral approximation lemma of [5]. The max-flow/min-cut reduction of the multi-phase half-space problem to the two-phase case is an elegant and potentially reusable idea. The paper contains no circular reasoning: the relaxed coefficients are characterized independently in Lemma 3.4, and the Γ-limit is then derived from that characterization together with external minimality results.

minor comments (6)

[§5.2.2, four-phase case] In the displayed chain after the four-phase decomposition, the coefficient of P_{2s}(E_1∪E_4,Q) is written as (α*−α̃_6); it should be (α*−α̃_7). As printed, the sum of the coefficients does not equal α̃_1+α̃_2−α̃_6−α̃_7, and the displayed lower bound would not yield P^σ_{2s}(H_{12},Q).
[§5.4, Remark 5.4] The uniqueness claim for H_{ij} does not follow immediately from uniqueness of H in the two-phase problem: Lemma 5.3 gives a competitor F with P^σ(F)≤P^σ(E), and equality in that inequality only implies F=H_{ij}, not E=H_{ij}. The remark should either supply the additional argument or be weakened to an assertion of uniqueness up to the equality cases in the max-flow/min-cut reduction.
[§3–§5, notation] The original matrix σ and its relaxation \barσ are typographically very similar throughout Sections 3–5, and in several displayed formulas both appear in the same line (for example in Lemma 3.4, Proposition 4.1, and Lemma 5.3). Please use a clearly distinguishable notation, such as \barσ or \underlineσ, consistently in all statements and proofs.
[Lemma 4.2, before (4.2)] The displayed line starting with 'lim inf (1−2s_k)(1−2s_k)Eσ...' contains a duplicated factor (1−2s_k); it should read lim inf (1−2s_k)Eσ_{2s_k}(E_k,A_{1−δ,1}).
[Lemma 5.3, definition of F] In the definition of the competitor F after choosing the minimal cut, the clause '∅ for k≥3' is ambiguous because k is also used as the sequence index; it should read 'F_l=∅ for l=3,...,m' or use a different symbol.
[Lemma 3.7] Lemma 3.7 is imported from [5] with only a sketch of proof. Since it is load-bearing for the upper bound, please mark it explicitly as a quoted result with the precise reference to [5, Lemma 3.1] and either omit the sketch or label it as a sketch of the adaptation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the relaxed coefficients are independently characterized, and the Gamma-limit is derived from explicit constructions plus external half-space minimality.

full rationale

The derivation is self-contained and non-circular. The relaxed matrix \bar{\sigma} is characterized in Lemma 3.4 as the shortest-path infimum over chains of the original coefficients, and the paper proves directly that this is the largest triangle-inequality-satisfying matrix below \sigma; this characterization is independent of any Gamma-limit statement. The Gamma-limsup is built from the pointwise computation for polyhedral partitions (Lemma 3.5), using only the known half-space interaction asymptotics from [3, Lemma 9], followed by an explicit relaxation construction (Lemma 3.6) that inserts intermediate phases along shortest paths; this construction uses the same independent characterization of \bar{\sigma} and does not assume the desired limit. The Gamma-liminf is obtained by the blow-up method, then refined through the cell formula via the approximation Lemmas 4.2 and 4.3; the cell problem is evaluated using the external two-phase half-space minimality theorem [3, Proposition 17] together with a max-flow/min-cut replacement argument (Lemma 5.3). The triangle inequality used in that argument is exactly the property of \bar{\sigma} proved in Lemma 3.4, not an input borrowed from the conclusion. No parameter is fitted to any subset of the data, and no prediction is a renamed input. The cited results are external published theorems rather than a self-citation chain, and the authors' own prior work appears only as peripheral context. The central Gamma-convergence claim is therefore supported by computations performed in the paper and by independent external results.

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

No numbers are fitted. The input matrix σ is prescribed, not fitted, and the relaxed matrix \barσ is computed by an optimization formula, not estimated from data. The paper relies on standard measure theory, geometric measure theory, and graph theory results, none of which include the conclusion.

assumptions (7)

standard math Measurable partition and Caccioppoli partition framework
Defines the state space: partitions up to Lebesgue null sets, finite perimeter chambers, and L1 convergence of characteristic functions. Used throughout.
domain assumption Bounded open Ω with Lipschitz boundary
Required by Theorems 1.2 and 1.3 for the exterior interaction term and for the approximation arguments.
domain assumption Fixed m≥3 and positive symmetric off-diagonal matrix σ
The energy P^σ_{2s} is defined through σ; the m≥3 condition is where non-additive and non-triangular effects appear.
standard math Half-space minimality for two-phase fractional perimeter ([3, Prop. 17], [11])
Imported engine used in Sections 5.2 to 5.4 to identify the cell constant; the multi-phase minimality proof reduces to it.
standard math Polyhedral approximation of Caccioppoli partitions ([5, Lemma 3.1])
Used in Proposition 3.9 to extend the Gamma-lim sup from polyhedral partitions to all Caccioppoli partitions.
standard math Max-flow min-cut theorem and flow decomposition ([8], [24])
Used in Lemma 5.3 to build a two-phase competitor with no larger energy from an arbitrary multi-phase competitor.
standard math Vitali covering and De Giorgi-Letta machinery
Used in Proposition 3.2 and Theorem 1.3 to pass from pointwise cube estimates to measures.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Gamma-convergence of nonlocal energies for partitions." pith.science (2026). https://pith.science/paper/YLNFW4IR

@misc{pith2026250620215,
  author       = {Pith},
  title        = {Pith review of: Gamma-convergence of nonlocal energies for partitions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YLNFW4IR}},
  note         = {Machine review of arXiv:2506.20215}
}

read the original abstract

We prove that certain nonlocal functionals defined on partitions made of measurable sets Gamma-converge to a local functional modeled on the perimeter in the sense of De Giorgi. Those nonlocal functionals involve generalized surface tension coefficients, and are lower semicontinuous even if the coefficients do not satisfy the triangular inequality. It implies a relaxation process in the limit, and provides a novel effect compare to the known gamma-convergence of the fractional perimeter towards the standard perimeter.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 29 canonical work pages

[3]

Ambrosio, G

L. Ambrosio, G. De Philippis, L. Martinazzi : Gamma-convergence of nonlocal perimeter functionals. Manuscripta Math. 134 (2011), 377--403. 3pt

work page 2011
[5]

Baldo : Minimal interface criterion for phase transitions in mixtures of Cahn-Hilliard fluids

S. Baldo : Minimal interface criterion for phase transitions in mixtures of Cahn-Hilliard fluids. Ann. Inst. H. Poincaré Anal. Non Linéaire 7 (1990), no. 2, pp. 67–90 3pt

work page 1990
[1]

Ambrosio, A

L. Ambrosio, A. Braides : Functionals defined on partitions in sets of finite perimeter. II. Semicontinuity, relaxation and homogenization, J. Math. Pures Appl. 69 (1990), 307--333. 3pt

work page 1990
[2]

Almgren : Existence and regularity almost everywhere of solutions to elliptic variational problems with constraints, Mem

F.J. Almgren : Existence and regularity almost everywhere of solutions to elliptic variational problems with constraints, Mem. Amer. Math. Soc. 4 (165), 1976. 3pt

work page 1976
[4]

Ambrosio, L., N

L. Ambrosio, L., N. Fusco, D. Pallara : Functions of Bounded Variation and Free Discontinuity Problems. Oxford University Press, New York (2000) 3pt

work page 2000
[6]

Barles, C

G. Barles, C. Georgelin : A simple proof of convergence for an approximation scheme for computing motions by mean curvature, SIAM J. Numer. Anal. 32 (1995), 484--500. 3pt

work page 1995
[7]

Barrios, A

B. Barrios, A. Figalli, E.Valdinoci : Bootstrap regularity for integro-differential operators and its application to nonlocal minimal surfaces. Ann. Scuola Norm. Sup. Pisa Cl. Sci (2014) 3pt

work page 2014
[8]

J. A. Bondy, U. S. R. Murty : Graph theory with applications , American Elsevier Publishing Co., Inc., New York, 1976. 3pt

work page 1976

Show all 31 references

[9]

Bourgain, H

J. Bourgain, H. Brezis, P. Mironescu : Another look at Sobolev spaces. (2001) pp.439-455.hal-00747692 3pt

2001
[10]

Bretin, S

E. Bretin, S. Masnou : A new phase field model for inhomogeneous minimal partitions, and applications to droplets dynamics. Interfaces and free boundaries , 2017. 3pt

2017
[11]

Caffarelli, J.-M

L. Caffarelli, J.-M. Roquejoffre, O. Savin : Non-local minimal surfaces. Communications on Pure and Applied Mathematics, 63, 2009 3pt

2009
[12]

L. A. Caffarelli, P. E. Souganidis : Convergence of nonlocal threshold dynamics approximations to front propagation, Arch. Ration. Mech. Anal. 195 (2010), 1--23. 3pt

2010
[13]

Caffarelli, E

L. Caffarelli, E. Valdinoci : Uniform estimates and limiting arguments for nonlocal minimal surfaces. Calc. Var. 41, 203–240 (2011) 3pt

2011
[14]

Caselli, E

M. Caselli, E. Florit-Simon, J. Serra : Yau's conjecture for nonlocal minimal surfaces, preprint arXiv:2306.07100 . 3pt

arXiv
[15]

Cesaroni, M

A. Cesaroni, M. Novaga : Minimal clusters in the plane. Nonlinear Anal. , 199 (2020), 111945. 3pt

2020
[16]

H. Chan, S. Dipierro, J. Serra, E. Valdinoci : Nonlocal approximation of minimal surfaces: optimal estimates from stability, preprint arXiv:2308.06328 . 3pt
[17]

Colombo, F

M. Colombo, F. Maggi : Existence and almost everywhere regularity of isoperimetric clusters for fractional perimeters, Nonlinear Anal. 153 (2017), 243--274. 3pt

2017
[18]

Evans : Convergence of an algorithm for mean curvature motion, Indiana Univ

L.C. Evans : Convergence of an algorithm for mean curvature motion, Indiana Univ. Math. J. 42 (1993), 635--681. 3pt

1993
[19]

D\'avila : On an open question about functions of bounded variation, Calc

J. D\'avila : On an open question about functions of bounded variation, Calc. Var. Partial Differential Equations 15 (2002), 519--527. 3pt

2002
[20]

M. Deza, M. Laurent : Geometry of Cuts and Metrics. Algorithms and Combinatorics , Springer (1997). 3pt

1997
[21]

Esedoglu, F

S. Esedoglu, F. Otto : Threshold dynamics for networks with arbitrary surface tensions, Comm. Pure Appl. Math. 68 (2015), 808--864. 3pt

2015
[22]

Fonseca, S

I. Fonseca, S. M\"uller : Quasi-convex integrands and lower semi-continuity in L^1 , SIAM J. Math. Anal. 23 (1992), 1081--1098. 3pt

1992
[23]

Lahti, A

P. Lahti, A. Pinamonti, X. Zhou : BV Functions and nonlocal functionals in metric measure spaces, J Geom Anal 34 , 318 (2024). 3pt

2024
[24]

Leonardi : Infiltrations in immiscible fluids systems, Proc

G.P. Leonardi : Infiltrations in immiscible fluids systems, Proc. Roy. Soc. Edinburgh Sect. A 131 (2001), 425--436. 3pt

2001
[25]

Maggi : Sets of Finite Perimeter and Geometric Variational Problems: an Introduction to Geometric Measure Theory

F. Maggi : Sets of Finite Perimeter and Geometric Variational Problems: an Introduction to Geometric Measure Theory. Cambridge Studies in Advanced Mathematics , 2012. 3pt

2012
[26]

Merriman, J

B. Merriman, J. Bence, S. Osher : Diffusion generated motion by mean curvature, Computational Crystal Growers Workshop (1992). 3pt

1992
[27]

Millot, Y

V. Millot, Y. Sire, K. Wang : Asymptotics for the fractional Allen–Cahn equation and stationary nonlocal minimal surfaces. ARMA, Springer (2018) 3pt

2018
[28]

Morgan : Lowersemicontinuity of energy clusters

F. Morgan : Lowersemicontinuity of energy clusters. Proceedings of the Royal Society of Edinburgh: Section A Mathematics , 127:819–822, 1 1997. 3pt

1997
[29]

Morse : Perfect blankets

A.P. Morse : Perfect blankets. Trans. Amer. Math. Soc. , 61 (1947), 418-422. 3pt

1947
[30]

Savin, E

O. Savin, E. Valdinoci : Regularity of nonlocal minimal cones in dimension 2, Calc. Var. Partial Differential Equations 48 (2013), 33--39. 3pt

2013
[31]

White : Existence of Least-Energy Configurations of Immiscible Fluids, J

B. White : Existence of Least-Energy Configurations of Immiscible Fluids, J. Geom. Anal. 6 (1996), 151--161

1996

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [3]

Ambrosio, G

L. Ambrosio, G. De Philippis, L. Martinazzi : Gamma-convergence of nonlocal perimeter functionals. Manuscripta Math. 134 (2011), 377--403. 3pt

work page 2011

[2] [5]

Baldo : Minimal interface criterion for phase transitions in mixtures of Cahn-Hilliard fluids

S. Baldo : Minimal interface criterion for phase transitions in mixtures of Cahn-Hilliard fluids. Ann. Inst. H. Poincaré Anal. Non Linéaire 7 (1990), no. 2, pp. 67–90 3pt

work page 1990

[3] [1]

Ambrosio, A

L. Ambrosio, A. Braides : Functionals defined on partitions in sets of finite perimeter. II. Semicontinuity, relaxation and homogenization, J. Math. Pures Appl. 69 (1990), 307--333. 3pt

work page 1990

[4] [2]

Almgren : Existence and regularity almost everywhere of solutions to elliptic variational problems with constraints, Mem

F.J. Almgren : Existence and regularity almost everywhere of solutions to elliptic variational problems with constraints, Mem. Amer. Math. Soc. 4 (165), 1976. 3pt

work page 1976

[5] [4]

Ambrosio, L., N

L. Ambrosio, L., N. Fusco, D. Pallara : Functions of Bounded Variation and Free Discontinuity Problems. Oxford University Press, New York (2000) 3pt

work page 2000

[6] [6]

Barles, C

G. Barles, C. Georgelin : A simple proof of convergence for an approximation scheme for computing motions by mean curvature, SIAM J. Numer. Anal. 32 (1995), 484--500. 3pt

work page 1995

[7] [7]

Barrios, A

B. Barrios, A. Figalli, E.Valdinoci : Bootstrap regularity for integro-differential operators and its application to nonlocal minimal surfaces. Ann. Scuola Norm. Sup. Pisa Cl. Sci (2014) 3pt

work page 2014

[8] [8]

J. A. Bondy, U. S. R. Murty : Graph theory with applications , American Elsevier Publishing Co., Inc., New York, 1976. 3pt

work page 1976