DiffLight: Integrating Content and Detail for Low-light Image Enhancement

The Low Light Image Enhancement (LLIE) task has been a hotspot in low-level computer vision research. The camera sensor can only capture a small amount of ambient light signal in low-light condition, resulting in significant noise black pseudo artifacts in images, which not only degrade visual quality but also affect the performance of down-stream visual tasks. However, current methods often produce overly smoothed and distorted results, or introduce strong noise artifacts. Moreover, for recent UHD high-definition low-light images, due to GPU memory limitations, LLIE must be conducted in patches, leading to block artifacts. Faced with these challenges, we propose a dual-branch pipeline called DiffLight. Specifically, it consists of the Denoising Enhancement (DE) branch and the Detail Preservation (DP) branch. The DE-branch adopts a combination of DiffIR and LEDNet to reduce noise and enhance brightness, while the DP-branch utilizes a novel Light Full-Former (LFF) method, which comprises 20 Full-Attention (LFA) modules to preserve full-scale image details. To tackle block artifacts, we further introduce Progressive Patch Fusion (PPF) for patch fusion. Experimental results demonstrate that our approach is high-ranked in the CVPR2024 NTIRE Low Light Enhancement challenge and produced state-of-the (SOTA) results on other datasets.

1. Introduction

In low-light conditions, sensors typically capture only a tiny amount of light signal, leading to a significant amount of noise in the resulting images, consequently impairing the performance of downstream visual tasks such as image classification [25] and automatic driving [22]. The Low-light Image Enhancement (LLIE) task is to brighten images, enhance contrast, mitigate image degradation, and alleviate severe image noise caused by low-light conditions, aiming to restore the original image information and improve perceptual quality [50]. In addition, high-resolution imaging has experienced significant advancements due to the emergence of sophisticated imaging sensors and displays in recent years. The Ultra High Definition (UHD) LLIE task becoming a research focus, gradually.