Optimized Minimal 3D Gaussian Splatting

Joo Chan Lee1, Jong Hwan Ko1, and Eunbyung Park2
1Sungkyunkwan University, 2Yonsei University
Preprint
arXiv Code

Optimized Minimal 3D Gaussian Splatting (OMG) focuses on minimizing storage requirements while using only a minimal number of Gaussian primitives. By proposing an efficient attribute representation, including sub-vector quantization, we achieve scene representation (>27 PSNR on Mip-NeRF 360 Dataset) with only 4.1 MB of storage while achieving 600+ FPS rendering!

Overview

We visualize qualitative examples (left) and the rate-distortion curve evaluated on the Mip-NeRF 360 dataset (right). All rendering speeds were measured on an NVIDIA RTX 3090 GPU, with values in parentheses in the left visualizations measured using an NVIDIA RTX 4090 GPU.


Recent studies have shown that high-quality rendering can be achieved with a substantially reduced number of Gaussians when represented with high-precision attributes. Nevertheless, existing 3DGS compression methods still rely on a relatively large number of Gaussians, focusing primarily on attribute compression. This is because a smaller set of Gaussians becomes increasingly sensitive to lossy attribute compression, leading to severe quality degradation. Since the number of Gaussians is directly tied to computational costs, it is essential to reduce the number of Gaussians effectively rather than only optimizing storage.
We propose Optimized Minimal Gaussians representation (OMG), which significantly reduces storage while using a minimal number of primitives. First, we determine the distinct Gaussian from the near ones, minimizing redundancy without sacrificing quality. Second, we propose a compact and precise attribute representation that efficiently captures both continuity and irregularity among primitives. Additionally, we propose a sub-vector quantization technique for improved irregularity representation, maintaining fast training with a negligible codebook size. Extensive experiments demonstrate that OMG reduces storage requirements by nearly 50% compared to the previous state-of-the-art and enables 600+ FPS rendering while maintaining high rendering quality.

Architecture





OMG learns per-Gaussian geometric and appearance features, applying Sub-Vector Quantization (SVQ) to all of them. The SVQ-applied geometric attributes are used for rendering, while the space feature based on the Gaussian center position is integrated into the appearance features to define the final appearance.

Results

OMG achieves remarkable efficiency alongside high performance, in terms of storage, rendering FPS, and training time.

BibTeX

@article{lee2025omg,
  author    = {Lee, Joo Chan and Ko, Jong Hwan and Park, Eunbyung},
  title     = {Optimized Minimal 3D Gaussian Splatting},
  journal   = {arXiv preprint arXiv:2503.16924},
  year      = {2025},
}