FluxFlow: Conservative Flow-Matching for Astronomical Image Super-Resolution

TL;DR

FluxFlow introduces a conservative flow-matching approach with real-world benchmarking for astronomical image super-resolution, effectively handling atmospheric PSF variations and reducing hallucinations.

Contribution

It presents a novel flow-matching framework with observation-aware training and a test-time correction, plus a large real-world dataset for ground-to-space image super-resolution.

Findings

FluxFlow outperforms baseline methods in accuracy.

The method effectively suppresses hallucinations.

The DESI--HST Dataset provides a new benchmark for real-world evaluation.

Abstract

Ground-to-space astronomical super-resolution requires recovering space-quality images from ground-based observations that are simultaneously limited by pixel sampling resolution and atmospheric seeing, which imposes a stochastic, spatially varying PSF that cannot be resolved through upsampling alone. Existing methods rely on synthetic training pairs that fail to capture real atmospheric statistics and are prone to either over-smoothed reconstructions or hallucination sources with no physical counterpart in the observed sky. We propose FluxFlow, a conservative pixel-space flow-matching framework that incorporates observation uncertainty and source-region importance weights during training, and a training-free Wiener-regularized test-time correction to suppress hallucination sources while preserving recovered detail. We further construct the DESI--HST Dataset, the large-scale real-world benchmark comprising 19,500 real co-registered ground-to-space image pairs with real atmospheric PSF variation. Experiments demonstrate that FluxFlow consistently outperforms existing baseline methods in both photometric and scientific accuracy.