Retinex-RAWMamba: Bridging Demosaicing and Denoising for Low-Light RAW Image Enhancement

2024-09-11

arXiv.org

AI Summary

To address insufficient demosaicing-denoising coupling in low-light RAW image enhancement and color distortion/noise residue arising from cross-domain (RAW→sRGB) mapping, this paper proposes an end-to-end framework integrating degradation modeling with ISP priors. We introduce RAWMamba, the first Mamba-based scanning mechanism tailored for multi-CFA raw data, and design a Retinex-inspired illumination-reflectance decomposition module to jointly perform RAW-domain denoising and nonlinear exposure correction. Notably, this work pioneers the integration of Mamba’s sequence modeling into RAW processing, synergizing CFA-aware feature extraction with physics-driven decomposition. Evaluated on SID and MCR datasets, our method achieves state-of-the-art performance, yielding significant PSNR/SSIM improvements over prior arts. It effectively suppresses chromatic shifts and substantially enhances fine-detail recovery while preserving photometric fidelity.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Security and Privacy: Data transparency and provenance
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Low-light image enhancement, particularly in cross-domain tasks such as mapping from the raw domain to the sRGB domain, remains a significant challenge. Many deep learning-based methods have been developed to address this issue and have shown promising results in recent years. However, single-stage methods, which attempt to unify the complex mapping across both domains, leading to limited denoising performance. In contrast, two-stage approaches typically decompose a raw image with color filter arrays (CFA) into a four-channel RGGB format before feeding it into a neural network. However, this strategy overlooks the critical role of demosaicing within the Image Signal Processing (ISP) pipeline, leading to color distortions under varying lighting conditions, especially in low-light scenarios. To address these issues, we design a novel Mamba scanning mechanism, called RAWMamba, to effectively handle raw images with different CFAs. Furthermore, we present a Retinex Decomposition Module (RDM) grounded in Retinex prior, which decouples illumination from reflectance to facilitate more effective denoising and automatic non-linear exposure correction. By bridging demosaicing and denoising, better raw image enhancement is achieved. Experimental evaluations conducted on public datasets SID and MCR demonstrate that our proposed RAWMamba achieves state-of-the-art performance on cross-domain mapping.

Problem

• Low-light Imaging
• RAW Photo Processing
• Noise Reduction

Innovation

• Retinex-RAWMamba
• RDM module
• cross-domain mapping