SGI: Structured 2D Gaussians for Efficient and Compact Large Image Representation

TL;DR

SGI introduces a structured, seed-based 2D Gaussian framework that efficiently compresses high-resolution images and accelerates optimization, outperforming prior methods in compression ratio and speed while maintaining or improving image quality.

Contribution

The paper proposes a novel seed-based structured Gaussian representation with a multi-scale fitting strategy for efficient high-resolution image encoding and faster convergence.

Findings

Achieves up to 7.5x compression over previous methods.

Speeds up optimization by 1.6x to 6.5x.

Maintains or improves image fidelity.

Abstract

2D Gaussian Splatting has emerged as a novel image representation technique that can support efficient rendering on low-end devices. However, scaling to high-resolution images requires optimizing and storing millions of unstructured Gaussian primitives independently, leading to slow convergence and redundant parameters. To address this, we propose Structured Gaussian Image (SGI), a compact and efficient framework for representing high-resolution images. SGI decomposes a complex image into multi-scale local spaces defined by a set of seeds. Each seed corresponds to a spatially coherent region and, together with lightweight multi-layer perceptrons (MLPs), generates structured implicit 2D neural Gaussians. This seed-based formulation imposes structural regularity on otherwise unstructured Gaussian primitives, which facilitates entropy-based compression at the seed level to reduce the total storage. However, optimizing seed parameters directly on high-resolution images is a challenging and non-trivial task. Therefore, we designed a multi-scale fitting strategy that refines the seed representation in a coarse-to-fine manner, substantially accelerating convergence. Quantitative and qualitative evaluations demonstrate that SGI achieves up to 7.5x compression over prior non-quantized 2D Gaussian methods and 1.6x over quantized ones, while also delivering 1.6x and 6.5x faster optimization, respectively, without degrading, and often improving, image fidelity. Code is available at https://github.com/zx-pan/SGI.