LL-GaussianImage: Efficient Image Representation for Zero-shot Low-Light Enhancement with 2D Gaussian Splatting

2026-01-22

AI Summary

This work proposes the first training-free, zero-shot framework for low-light enhancement directly in the compressed representation domain of 2D Gaussian Splatting (2DGS), bypassing the conventional decompress–enhance–recompress pipeline that suffers from inefficiency and secondary distortion. By introducing a semantic-guided mixture-of-experts mechanism, a multi-objective collaborative loss, and a two-stage optimization strategy, the method establishes “compression-as-enhancement” as a novel paradigm. It achieves significant improvements in visual quality and effective artifact suppression while maintaining high compression ratios, thereby demonstrating the feasibility and superiority of direct processing in the 2DGS compressed domain.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Diffusion Models for Vision
• Data Mining & Knowledge Management: Data Compression

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

2D Gaussian Splatting (2DGS) is an emerging explicit scene representation method with significant potential for image compression due to high fidelity and high compression ratios. However, existing low-light enhancement algorithms operate predominantly within the pixel domain. Processing 2DGS-compressed images necessitates a cumbersome decompression-enhancement-recompression pipeline, which compromises efficiency and introduces secondary degradation. To address these limitations, we propose LL-GaussianImage, the first zero-shot unsupervised framework designed for low-light enhancement directly within the 2DGS compressed representation domain. Three primary advantages are offered by this framework. First, a semantic-guided Mixture-of-Experts enhancement framework is designed. Dynamic adaptive transformations are applied to the sparse attribute space of 2DGS using rendered images as guidance to enable compression-as-enhancement without full decompression to a pixel grid. Second, a multi-objective collaborative loss function system is established to strictly constrain smoothness and fidelity during enhancement, suppressing artifacts while improving visual quality. Third, a two-stage optimization process is utilized to achieve reconstruction-as-enhancement. The accuracy of the base representation is ensured through single-scale reconstruction and network robustness is enhanced. High-quality enhancement of low-light images is achieved while high compression ratios are maintained. The feasibility and superiority of the paradigm for direct processing within the compressed representation domain are validated through experimental results.

Problem

• low-light enhancement
• 2D Gaussian Splatting
• compressed domain processing
• image compression
• zero-shot learning

Innovation

• 2D Gaussian Splatting
• zero-shot enhancement
• compressed domain processing
• Mixture-of-Experts
• low-light image enhancement