Grid is Good: Adaptive Refinement Algorithms for Off-the-Grid Total Variation Minimization

TL;DR

This paper introduces an adaptive refinement algorithm for off-the-grid total variation minimization, improving solution accuracy by iteratively refining partitions based on dual problem resolution and constraint violations.

Contribution

The paper presents a novel adaptive refinement algorithm that avoids heuristics and guarantees convergence for total variation regularized measure optimization.

Findings

Convergence of the method is proven under mild conditions.

The algorithm achieves linear convergence rate with additional assumptions.

Numerical experiments confirm the theoretical results.

Abstract

We propose an adaptive refinement algorithm to solve total variation regularized measure optimization problems. The method iteratively constructs dyadic partitions of the unit cube based on i) the resolution of discretized dual problems and ii) on the detection of cells containing points that violate the dual constraints. The detection is based on upper-bounds on the dual certificate, in the spirit of branch-and-bound methods. The interest of this approach is that it avoids the use of heuristic approaches to find the maximizers of dual certificates. We prove the convergence of this approach under mild hypotheses and a linear convergence rate under additional non-degeneracy assumptions. These results are confirmed by simple numerical experiments.