DEFormer: DCT-driven Enhancement Transformer for Low-light Image and Dark Vision

TL;DR

DEFormer introduces a DCT-driven transformer framework that leverages frequency domain information through a learnable frequency branch and cross domain fusion, significantly enhancing low-light image quality and dark vision tasks.

Contribution

The paper presents a novel DCT-driven enhancement transformer with a learnable frequency branch and cross domain fusion, addressing limitations of RGB-only methods in low-light image enhancement.

Findings

Achieved superior results on LOL and MIT-Adobe FiveK datasets.

Improved dark detection performance.

Enhanced detail restoration in dark areas.

Abstract

Low-light image enhancement restores the colors and details of a single image and improves high-level visual tasks. However, restoring the lost details in the dark area is still a challenge relying only on the RGB domain. In this paper, we delve into frequency as a new clue into the model and propose a DCT-driven enhancement transformer (DEFormer) framework. First, we propose a learnable frequency branch (LFB) for frequency enhancement contains DCT processing and curvature-based frequency enhancement (CFE) to represent frequency features. Additionally, we propose a cross domain fusion (CDF) to reduce the differences between the RGB domain and the frequency domain. Our DEFormer has achieved superior results on the LOL and MIT-Adobe FiveK datasets, improving the dark detection performance.