Enhancing Low-light Light Field Images with A Deep Compensation Unfolding Network

TL;DR

This paper introduces DCUNet, a deep unfolding network designed to enhance low-light light field images by mimicking inverse problem optimization, incorporating deep compensation, and exploiting LF-specific features for superior restoration.

Contribution

The paper proposes a novel deep compensation unfolding network (DCUNet) with a multi-stage architecture and LF-specific modules for improved low-light light field image enhancement.

Findings

DCUNet outperforms state-of-the-art methods qualitatively and quantitatively.

It better preserves the geometric structure of light field images.

The framework effectively suppresses noise and estimation errors.

Abstract

This paper presents a novel and interpretable end-to-end learning framework, called the deep compensation unfolding network (DCUNet), for restoring light field (LF) images captured under low-light conditions. DCUNet is designed with a multi-stage architecture that mimics the optimization process of solving an inverse imaging problem in a data-driven fashion. The framework uses the intermediate enhanced result to estimate the illumination map, which is then employed in the unfolding process to produce a new enhanced result. Additionally, DCUNet includes a content-associated deep compensation module at each optimization stage to suppress noise and illumination map estimation errors. To properly mine and leverage the unique characteristics of LF images, this paper proposes a pseudo-explicit feature interaction module that comprehensively exploits redundant information in LF images. The experimental results on both simulated and real datasets demonstrate the superiority of our DCUNet over state-of-the-art methods, both qualitatively and quantitatively. Moreover, DCUNet preserves the essential geometric structure of enhanced LF images much better. The code will be publicly available at https://github.com/lyuxianqiang/LFLL-DCU.