Abstract
Low-light image enhancement (LLIE) aims to improve the illuminance of images due to insufficient light exposure. Recently, various lightweight learning-based LLIE methods have been proposed to handle the challenges of unfavorable prevailing low contrast, low brightness, etc. In this paper, we have streamlined the architecture of the network to the utmost degree. By utilizing the effective structural re-parameterization technique, a single convolutional layer model (SCLM) is proposed that provides global low-light enhancement as the coarsely enhanced results. In addition, we introduce a local adaptation module that learns a set of shared parameters to accomplish local illumination correction to address the issue of varied exposure levels in different image regions. Experimental results demonstrate that the proposed method performs favorably against the state-of-the-art LLIE methods in both objective metrics and subjective visual effects. Additionally, our method has fewer parameters and lower inference complexity compared to other learning-based schemes.
Abstract (translated)
弱光图像增强(LLIE)旨在通过不足的光线曝光来提高图像的亮度。最近，提出了各种轻量级基于学习的LLIE方法，以应对不良的低对比度、低亮度等挑战。在本文中，我们尽可能简化了网络架构。通过利用有效的结构参数重采样技术，我们提出了一种单卷积层模型(SCLM)，该模型提供全球弱光增强，作为粗劣增强结果。此外，我们引入了一个局部适应模块，学习一组共享参数，实现局部照明修正，以解决不同图像区域 vary 的曝光水平问题。实验结果显示， proposed 方法在客观指标和主观视觉效果方面表现良好，与先进的Llie方法相比，我们的方法和其他基于学习的方案相比，参数更少，推理复杂度更低。
URL
https://arxiv.org/abs/2305.14039
