Robust Low-light Scene Restoration via Illumination Transition

ICCV 2025
Ze Li1,2, Feng Zhang3,2, Xiatian Zhu4, Meng Zhang1,2, Yanghong Zhou2, P. Y. Mok1
1The Hong Kong University of Science and Technology, Hong Kong SAR
2The Hong Kong Polytechnic University, Hong Kong SAR
3Nanjing University of Posts and Telecommunications, Nanjing, China
4University of Surrey, Guildford, United Kingdom
Paper Code BibTeX

Abstract

Synthesizing normal-light novel views from low-light multiview images remains a challenging yet practical task due to the low visibility and high ISO noise challenges. Existing low-light enhancement methods often struggle to preprocess these images effectively due to their inability to structurally correlate multiple views. While state-of-the-art approaches have advanced by manipulating illumination-related components during rendering, they often introduce color distortions and artifacts. Moreover, they rely solely on NeRF's multi-view optimization, which offers limited denoising effectiveness. In this paper, we propose a novel Robust Low-light Scene Restoration framework termed (RoSe), which enables novel-view synthesis under normal lighting from low-light multiview images. Inspired by the 2D Retinex theory, we frame this task as an illuminance transition estimation problem in 3D space, further conceptualizing it as a specialized rendering task. This multiview-consistent illuminance transition field establishes a robust connection between low-light and normal-light conditions. By further exploiting the inherent low-rank property of illumination to constrain the transition representation, we achieve more effective denoising without complex 2D techniques or explicit noise modeling. To this end, we design a concise dual-branch architecture and propose a low-rank denoising module. Experiments demonstrate that RoSe significantly outperforms state-of-the-art models in both rendering quality and multiview consistency on standard benchmarks.

Method

Overview

Method Overview
Figure 2. An overview of the proposed RoSe framework.

Results

Comparisons with State-of-the-art Methods

Comparison with SOTA
Figure 1: Image/video enhancement (Zero-DCE+NeRF, RUAS+NeRF, LLVE+NeRF) vs. state-of-the-art models (LLNeRF, Aleth-Nerf) vs. our RoSe.

Qualitative Results

Novel View Synthesis
Figure 3: Normal-light novel view synthesis comparison in low-light conditions.

Analysis

Density Distribution
Figure 5: Density distribution of sampling points along the camera ray, with zoomed-in image pixels for better observation.

Citation

                @inproceedings{li2025robustlowlightscenerestoration,
                    title     = {Robust Low-light Scene Restoration via Illumination Transition},
                    author    = {Li, Ze and Zhang, Feng and Zhu, Xiatian and Zhang, Meng and Zhou, Yanghong and Mok, P. Y.},
                    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
                    year      = {2025}
                  }