Large Images Are Gaussians: High-Quality Large Image Representation with Levels of 2D Gaussian Splatting
Year: 2025
Type: conference-paper
Abstract: While Implicit Neural Representations (INRs) have demonstrated significant success in image representation, they are often hindered by large training memory and slow decoding speed. Recently, Gaussian Splatting (GS) has emerged as a promising solution in 3D reconstruction due to its highquality novel view synthesis and rapid rendering capabilities, positioning it as a valuable tool for a broad spectrum of applications. In particular, a GS-based representation, 2DGS, has shown potential for image fitting. In our work, we present Large Images are Gaussians (LIG), which delves deeper into the application of 2DGS for image representations, addressing the challenge of fitting large images with 2DGS in the situation of numerous Gaussian points, through two distinct modifications: 1) we adopt a variant of representation and optimization strategy, facilitating the fitting of a large number of Gaussian points; 2) we propose a Level-of-Gaussian approach for reconstructing both coarse low-frequency initialization and fine high-frequency details. Consequently, we successfully represent large images as Gaussian points and achieve high-quality large image representation, demonstrating its efficacy across various types of large images.
Language: English
Subfield: Computational Mechanics
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
Awards: G6740839966, Research and Development of an AI-enhanced System for Whole Slide Histopathology Image: Quantitative Assessment, Assisted Diagnosis and Prognostic Evaluation, Research and Development of an AI-enhanced System for Preoperative Meningioma Grading and Brain Invasion Prediction from Multi-modality Data
OpenAlex ID: W4409365737