Confident Splatting: Confidence-Based Compression of 3D Gaussian Splatting via Learnable Beta Distributions
Abstract
A novel lossy compression method using learnable confidence scores improves storage and computational efficiency in 3D Gaussian Splatting without sacrificing visual quality.
3D Gaussian Splatting enables high-quality real-time rendering but often produces millions of splats, resulting in excessive storage and computational overhead. We propose a novel lossy compression method based on learnable confidence scores modeled as Beta distributions. Each splat's confidence is optimized through reconstruction-aware losses, enabling pruning of low-confidence splats while preserving visual fidelity. The proposed approach is architecture-agnostic and can be applied to any Gaussian Splatting variant. In addition, the average confidence values serve as a new metric to assess the quality of the scene. Extensive experiments demonstrate favorable trade-offs between compression and fidelity compared to prior work. Our code and data are publicly available at https://github.com/amirhossein-razlighi/Confident-Splatting
Community
A new, lossy approach towards compressing 3DGS scenes with learnable confidence score and prunning in test-time (From 6M splats to 3M splats with less than 0.1 PSNR drop!)
Please visit the paper's website for code, the newly released dataset and interactive visualization!
https://amirhossein-razlighi.github.io/Confident-Splatting/
Also, our method is architecture-agnostic. meaning it works on original 3DGS, Gaussian Splatting MCMC, and all other architectures of Gaussian Splatting with no change!
Models citing this paper 1
Datasets citing this paper 1
Spaces citing this paper 0
No Space linking this paper
Collections including this paper 0
No Collection including this paper