Abstract:3D Gaussian splatting (3DGS) has recently emerged as an alternative representation that leverages a 3D Gaussian-based representation and introduces an approximated volumetric rendering, achieving very fast rendering speed and promising image quality. Furthermore, subsequent studies have successfully extended 3DGS to dynamic 3D scenes, demonstrating its wide range of applications. However, a significant drawback arises as 3DGS and its following methods entail a substantial number of Gaussians to maintain the high fidelity of the rendered images, which requires a large amount of memory and storage. To address this critical issue, we place a specific emphasis on two key objectives: reducing the number of Gaussian points without sacrificing performance and compressing the Gaussian attributes, such as view-dependent color and covariance. To this end, we propose a learnable mask strategy that significantly reduces the number of Gaussians while preserving high performance. In addition, we propose a compact but effective representation of view-dependent color by employing a grid-based neural field rather than relying on spherical harmonics. Finally, we learn codebooks to compactly represent the geometric and temporal attributes by residual vector quantization. With model compression techniques such as quantization and entropy coding, we consistently show over 25x reduced storage and enhanced rendering speed compared to 3DGS for static scenes, while maintaining the quality of the scene representation. For dynamic scenes, our approach achieves more than 12x storage efficiency and retains a high-quality reconstruction compared to the existing state-of-the-art methods. Our work provides a comprehensive framework for 3D scene representation, achieving high performance, fast training, compactness, and real-time rendering. Our project page is available at <a class="link-external link-https" href="https://maincold2.github.io/c3dgs/" rel="external noopener nofollow">this https URL</a>.

Mini-Splatting: Representing Scenes with a Constrained Number of Gaussians

Efficient Density Control for 3D Gaussian Splatting

Mini-Splatting2: Building 360 Scenes within Minutes via Aggressive Gaussian Densification

SG-Splatting: Accelerating 3D Gaussian Splatting with Spherical Gaussians

GeoGaussian: Geometry-aware Gaussian Splatting for Scene Rendering

Implicit Gaussian Splatting with Efficient Multi-Level Tri-Plane Representation

Feature Splatting for Better Novel View Synthesis with Low Overlap

CompGS: Efficient 3D Scene Representation via Compressed Gaussian Splatting

SWAG: Splatting in the Wild images with Appearance-conditioned Gaussians

3D Gaussian Splatting as Markov Chain Monte Carlo

Spacetime Gaussian Feature Splatting for Real-Time Dynamic View Synthesis

Compact 3D Scene Representation via Self-Organizing Gaussian Grids

SuperGaussians: Enhancing Gaussian Splatting Using Primitives with Spatially Varying Colors

Compressed 3D Gaussian Splatting for Accelerated Novel View Synthesis

FreeSplat: Generalizable 3D Gaussian Splatting Towards Free-View Synthesis of Indoor Scenes

4D Gaussian Splatting for Real-Time Dynamic Scene Rendering

Monocular Dynamic Gaussian Splatting is Fast and Brittle but Smooth Motion Helps

EfficientGS: Streamlining Gaussian Splatting for Large-Scale High-Resolution Scene Representation

Lightweight Predictive 3D Gaussian Splats

Compact 3D Gaussian Splatting for Static and Dynamic Radiance Fields

Speedy-Splat: Fast 3D Gaussian Splatting with Sparse Pixels and Sparse Primitives