Towards Controllable Low-Light Image Enhancement: A Continuous Multi-illumination Dataset and Efficient State Space Framework

📄 arXiv: 2603.25296v1 📥 PDF

作者: Hongru Han, Tingrui Guo, Liming Zhang, Yan Su, Qiwen Xu, Zhuohua Ye

分类: cs.CV

发布日期: 2026-03-26

备注: 10 pages, 8 figures


💡 一句话要点

提出可控低光图像增强方法以解决现有方法的不足

🎯 匹配领域: 支柱二:RL算法与架构 (RL & Architecture) 支柱九:具身大模型 (Embodied Foundation Models)

关键词: 低光图像增强 可控增强 状态空间模型 去噪监督 多照明基准 图像处理 深度学习

📋 核心要点

  1. 现有低光图像增强方法难以处理环境条件和传感器参数的不确定性,导致亮度预测不准确。
  2. 本文提出可控低光增强(CLE)框架,通过将任务重新定义为条件问题,解决了亮度控制与色彩保真之间的矛盾。
  3. 在七个基准测试中,本文方法表现出色,显著减少了对gt-mean后处理的依赖,提升了图像增强的可控性。

📝 摘要(中文)

低光图像增强(LLIE)传统上被视为确定性映射,但这一范式难以应对任务的病态特性,导致预测与标签之间的亮度差异。为了解决这一根本性限制,本文提出了可控低光增强(CLE),将任务重新表述为一个良定义的条件问题。我们引入了CLE-RWKV框架,并支持Light100基准,该基准展示了连续的真实世界照明过渡。通过在HVI色彩空间中采用去噪监督策略,有效分离了照明调制与纹理恢复,实验结果表明该方法在七个基准上表现出竞争力和强大的可控性。

🔬 方法详解

问题定义:本文旨在解决低光图像增强任务中的病态特性,现有方法在处理未知环境条件和传感器参数时,常常导致亮度预测与真实标签之间的差异。

核心思路:提出可控低光增强(CLE)框架,将任务重新表述为条件问题,利用去噪监督策略在HVI色彩空间中分离照明调制与纹理恢复,从而实现更好的亮度控制和色彩保真。

技术框架:整体架构包括CLE-RWKV框架和Light100基准,前者通过空间到深度(S2D)策略适应高效状态空间模型(SSMs),后者提供连续的真实世界照明过渡数据。

关键创新:最重要的创新在于将低光图像增强任务转变为可控的条件问题,并通过去噪监督策略有效解决了亮度与色彩之间的矛盾,这与传统方法的确定性映射形成鲜明对比。

关键设计:在网络结构上,采用空间到深度(S2D)策略,将空间邻域折叠到通道维度,恢复局部归纳偏差,同时保持线性复杂度。损失函数设计上,强调亮度控制与色彩保真的平衡。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果显示,提出的方法在七个基准测试中均表现出色,相较于现有技术,显著降低了对gt-mean后处理的依赖,提升了图像的可控性和质量,证明了其在真实世界多照明条件下的有效性。

🎯 应用场景

该研究在低光图像处理领域具有广泛的应用潜力,尤其适用于夜间摄影、监控视频增强和医学成像等场景。通过提供可控的图像增强方法,能够在不同照明条件下提升图像质量,满足实际应用需求。

📄 摘要(原文)

Low-light image enhancement (LLIE) has traditionally been formulated as a deterministic mapping. However, this paradigm often struggles to account for the ill-posed nature of the task, where unknown ambient conditions and sensor parameters create a multimodal solution space. Consequently, state-of-the-art methods frequently encounter luminance discrepancies between predictions and labels, often necessitating "gt-mean" post-processing to align output luminance for evaluation. To address this fundamental limitation, we propose a transition toward Controllable Low-light Enhancement (CLE), explicitly reformulating the task as a well-posed conditional problem. To this end, we introduce CLE-RWKV, a holistic framework supported by Light100, a new benchmark featuring continuous real-world illumination transitions. To resolve the conflict between luminance control and chromatic fidelity, a noise-decoupled supervision strategy in the HVI color space is employed, effectively separating illumination modulation from texture restoration. Architecturally, to adapt efficient State Space Models (SSMs) for dense prediction, we leverage a Space-to-Depth (S2D) strategy. By folding spatial neighborhoods into channel dimensions, this design allows the model to recover local inductive biases and effectively bridge the "scanning gap" inherent in flattened visual sequences without sacrificing linear complexity. Experiments across seven benchmarks demonstrate that our approach achieves competitive performance and robust controllability, providing a real-world multi-illumination alternative that significantly reduces the reliance on gt-mean post-processing.