2603.01860

Total: 1

#1 Multiresolution Adaptive Block-Coordinate Forward-Backward for Image Reconstruction [PDF] [Copy] [Kimi] [REL]

Authors: Edgar Desainte-Maréville, Marion Foare, Paulo Gonçalves, Nelly Pustelnik, Elisa Riccietti

Classical first-order optimization methods for imaging inverse problems scale poorly with image resolution. Wavelet based multilevel strategies can accelerate convergence under strong blur, but their fixed coarse-to-fine schedules lose effectiveness in moderate-blur or noise-dominated regimes. In this work, we propose an adaptive multiresolution block coordinate Forward-Backward algorithm for image restoration. Multiresolution block selection is driven by the local magnitude of the proximal update via a stochastic non-smooth Gauss-Southwell rule applied to the wavelet decomposition of the image. This adaptive selection strategy dynamically balances updates across scales, emphasizing coarse or fine blocks according to the degradation regime. As a result, the proposed method automatically adapts to varying blur and noise levels without relying on a predefined hierarchical update scheme.

Subjects: Signal Processing , Optimization and Control

Publish: 2026-03-02 13:41:20 UTC