Search papers, labs, and topics across Lattice.
This paper introduces HALO, a novel dual-prior-driven framework designed to enhance low-light remote sensing images by addressing the issue of attention drift that leads to structural blurring and color distortion. By leveraging an illumination-invariant semantic prior for regional homogeneity and a pseudo-3D topological prior for boundary heterogeneity, HALO effectively guides feature aggregation to maintain content consistency while preventing cross-boundary confusion. Extensive evaluations show that HALO outperforms existing methods across eight challenging benchmarks, significantly improving image sharpness and color fidelity for downstream Earth observation applications.
HALO's innovative dual-prior approach eliminates attention drift, achieving unprecedented clarity and color accuracy in low-light remote sensing imagery.
Restoring high-fidelity remote sensing imagery from extreme low-light degradation is indispensable for reliable Earth observation and downstream machine vision. However, under severe noise and illumination corruption, existing methods suffer from attention drift, erroneously aggregating features across distinct physical boundaries and causing severe structural blurring and color distortion. To address this, we propose HALO, a dual-prior-driven enhancement framework that formulates enhancement as a guided feature aggregation problem driven by foundation model priors. Specifically, an illumination-invariant semantic prior provides regional homogeneity as a positive bias for content-consistent aggregation, while a pseudo-3D topological prior provides boundary heterogeneity as a negative penalty to strictly prevent cross-boundary confusion. To cooperatively incorporate these two priors, we propose a Homogeneity-Heterogeneity Cooperative Attention Module (H2CAM) to resolve feature conflicts during cross-modal prior fusion. Extensive experiments demonstrate that HALO achieves state-of-the-art performance across 8 challenging synthetic and real-world remote sensing benchmarks, significantly improving physical boundary sharpness and color fidelity while maximizing the preservation of discriminative features for downstream Earth observation tasks.
