Multigrid methods for total variation (Q6908301)

In this interesting paper the authors study multigrid methods for total variation. To set the scene, the authors study composite optimization problems of the form: \(\min_{x\in X}(F(x)+G(x))\), where \(F\) is convex and also smooth and \(G\) is convex but also possibily nonsmooth on a Hilbert space \(X\). Based on a nonsmooth coherence condition the authors construct and prove the convergence of a forward-backward splitting method for several optimization problems of the type above. The method alternates between steps on a fine and a coarse grid, which is often used to solve numerous partial differential equations. The authors focus on total variation regularized inverse imaging problems, specifically their dual problems. The problems are chosen so that the authors are able to develop the relevant coarse-grid problems. The authors demonstrate numerically the performance of their methods on magnetic resonance imaging and variation denoising.\N\NThe paper is well written with a good set of references.\N\NFor the entire collection see [Zbl 1573.68017].