Leveraging Learned Image Prior for 3D Gaussian Compression

2025-10-16

AI Summary

To address the limited rate-distortion performance in 3D Gaussian Splatting (3DGS) compression—stemming from the absence of learnable priors—this paper proposes an image-prior-based post-processing restoration method. Our approach models compression-induced artifacts directly in the image space, taking as input a coarse reconstruction rendered from the compressed Gaussian parameters and its corresponding residual, then refining it via a deep neural network. The key contribution is the first integration of a learnable image prior into the 3DGS compression pipeline, enabling a plug-and-play, fully compatible optimization with existing compression methods. Experimental results demonstrate that our method significantly reduces storage overhead while consistently outperforming state-of-the-art approaches across quantitative metrics—including PSNR and LPIPS—thereby achieving superior trade-offs between visual quality and storage efficiency.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Learning on the Edge & Model Compression
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Compression techniques for 3D Gaussian Splatting (3DGS) have recently achieved considerable success in minimizing storage overhead for 3D Gaussians while preserving high rendering quality. Despite the impressive storage reduction, the lack of learned priors restricts further advances in the rate-distortion trade-off for 3DGS compression tasks. To address this, we introduce a novel 3DGS compression framework that leverages the powerful representational capacity of learned image priors to recover compression-induced quality degradation. Built upon initially compressed Gaussians, our restoration network effectively models the compression artifacts in the image space between degraded and original Gaussians. To enhance the rate-distortion performance, we provide coarse rendering residuals into the restoration network as side information. By leveraging the supervision of restored images, the compressed Gaussians are refined, resulting in a highly compact representation with enhanced rendering performance. Our framework is designed to be compatible with existing Gaussian compression methods, making it broadly applicable across different baselines. Extensive experiments validate the effectiveness of our framework, demonstrating superior rate-distortion performance and outperforming the rendering quality of state-of-the-art 3DGS compression methods while requiring substantially less storage.

Problem

• Compressing 3D Gaussian Splatting while preserving rendering quality
• Addressing compression artifacts using learned image priors
• Enhancing rate-distortion performance with compact Gaussian representation

Innovation

• Leveraging learned image priors for 3D Gaussian restoration
• Using coarse rendering residuals as side information
• Refining compressed Gaussians via restored image supervision