# Adaptive parameter selection for weighted-TV image reconstruction problems

## TL;DR

This paper introduces an efficient method for automatically selecting locally adaptive Total Variation regularization parameters in image reconstruction, combining local likelihood estimation with a global discrepancy principle, outperforming existing strategies.

## Contribution

It presents a novel hybrid approach that integrates local maximum-likelihood estimation with a global discrepancy principle for adaptive TV regularization parameter selection.

## Key findings

- Outperforms state-of-the-art parameter estimation methods
- Effective in various image reconstruction problems
- Demonstrates improved image quality and noise handling

## Abstract

We propose an efficient estimation technique for the automatic selection of locally-adaptive Total Variation regularisation parameters based on an hybrid strategy which combines a local maximum-likelihood approach estimating space-variant image scales with a global discrepancy principle related to noise statistics. We verify the effectiveness of the proposed approach solving some exemplar image reconstruction problems and show its outperformance in comparison to state-of-the-art parameter estimation strategies, the former weighting locally the fit with the data (Dong et al. '11), the latter relying on a bilevel learning paradigm (Hinterm\"uller et al., '17)