克服注意力漂移：面向低光照遥感图像增强的同质性-异质性引导特征聚合
Overcoming Attention Drift: Homogeneity-Heterogeneity Guided Feature Aggregation for Low-Light Remote Sensing Image Enhancement
浏览论文内容
中文总结 AI 辅助
针对低光照遥感图像增强中现有方法的注意力漂移问题，提出双先验驱动的HALO框架，通过H2CAM模块融合同质性与异质性先验，在8个基准上实现SOTA性能，提升边界清晰度与色彩保真度。
中文摘要 AI 辅助
从极端低光照退化中恢复高保真遥感图像，对于可靠的地球观测及下游机器视觉任务而言不可或缺。然而，在严重噪声与光照损坏下，现有方法存在注意力漂移问题，会错误地聚合不同物理边界间的特征，导致严重的结构模糊与色彩失真。为解决该问题，我们提出HALO，一种双先验驱动的增强框架，将增强任务建模为 foundation model（基础模型）先验驱动的引导特征聚合问题。具体而言，光照不变语义先验提供区域同质性，作为内容一致聚合的正向偏置；伪3D拓扑先验提供边界异质性，作为负向惩罚以严格防止跨边界混淆。为协同整合这两种先验，我们提出同质性-异质性协同注意力模块（H2CAM），以解决跨模态先验融合过程中的特征冲突。大量实验表明，HALO在8个具有挑战性的合成与真实世界遥感基准上实现了SOTA（state-of-the-art，当前最优）性能，显著提升了物理边界清晰度与色彩保真度，同时最大化保留了用于下游地球观测任务的判别性特征。
英文摘要
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.
发表机构
- School of Information and Artificial Intelligence, Yunnan University of Finance and Economics（云南财经大学信息与人工智能学院）
- Department of Computer Science, Hong Kong Baptist University（香港浸会大学计算机科学系）
- School of Electronics and Information, Northwestern Polytechnical University（西北工业大学电子信息学院）
机构由 AI 辅助整理，请以论文原文为准。
