Regularization by sup-inf convolutions on Riemannian manifolds: an extension of Lasry-Lions theorem to manifolds of bounded curvature

preprint2014arXivOpen access

Abstract

We show how Lasry-Lions's result on regularization of functions defined on Rn\mathbb{R}^n or on Hilbert spaces by sup-inf convolutions with squares of distances can be extended to (finite or infinite dimensional) Riemannian manifolds MM of bounded sectional curvature. More specifically, among other things we show that if the sectional curvature KK of MM satisfies −K0≤K≤K0-K_0\leq K\leq K_0 on MM for some K0>0K_0>0, and if the injectivity and convexity radii of MM are strictly positive, then every bounded, uniformly continuous function f:M→Rf:M\to\mathbb{R} can be uniformly approximated by globally C1,1C^{1,1} functions defined by (fλ)μ=sup⁡z∈Minf⁡y∈M{f(y)+12λd(z,y)2−12μd(x,z)2} (f_λ)^μ=\sup_{z\in M}\inf_{y\in M}\{f(y)+\frac{1}{2λ} d(z,y)^{2}-\frac{1}{2μ}d(x,z)^2\} as λ,μ→0+λ, μ\to 0^{+}, with 0<μ<λ/20<μ<λ/2. Our definition of (global) C1,1C^{1,1} smoothness is intrinsic and natural, and it reduces to the usual one in flat spaces, but we warn the reader that, in the noncompact case, this definition differs from other notions of (rather local) C1,1C^{1,1} smoothness that have been recently used, for instance, by A. Fathi and P. Bernard (based on charts). The importance of this regularization method lies (rather than on the degree of smoothness obtained) on the fact that the correspondence f↦(fλ)μf\mapsto (f_λ)^μ is explicit and preserves many significant geometrical properties that the given functions ff may have, such as invariance by a set of isometries, infima, sets of minimizers, ordering, local or global Lipschitzness, and (only when one additionally assumes that K≤0K\leq 0) local or global convexity. We also give two examples showing that this result completely fails, even for (nonflat) Cartan-Hadamard manifolds, whenever ff or KK are not bounded.

Explore this paper’s authors, topics and related work.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Reviews 0

Write a reviewWrite

No reviews yet.

Discussion 0

Add a commentWrite

No comments yet.