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

TL;DR

This paper introduces a controllable low-light image enhancement framework that reformulates the task as a conditional problem, utilizing a new dataset and efficient model design to improve luminance control and reduce post-processing needs.

Contribution

The paper proposes CLE-RWKV, a novel framework with a new benchmark Light100, employing a noise-decoupled supervision strategy and a Space-to-Depth approach for efficient dense prediction in low-light enhancement.

Findings

Achieves competitive performance across seven benchmarks.

Provides robust controllability and reduces reliance on gt-mean post-processing.

Introduces a new dataset with continuous real-world illumination transitions.

Abstract

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.