EAGLES: Efficient Accelerated 3D Gaussians with Lightweight EncodingS

TL;DR

This paper introduces a memory-efficient and faster training method for 3D Gaussian splatting in scene synthesis, significantly reducing memory usage while maintaining high visual quality and enabling real-time rendering.

Contribution

It proposes quantized embeddings, a coarse-to-fine training strategy, and a pruning stage to reduce memory and accelerate training and rendering in 3D Gaussian splatting.

Findings

Reduces memory usage by over an order of magnitude

Achieves 10-20x faster training and inference speeds

Maintains high-quality scene reconstruction

Abstract

Recently, 3D Gaussian splatting (3D-GS) has gained popularity in novel-view scene synthesis. It addresses the challenges of lengthy training times and slow rendering speeds associated with Neural Radiance Fields (NeRFs). Through rapid, differentiable rasterization of 3D Gaussians, 3D-GS achieves real-time rendering and accelerated training. They, however, demand substantial memory resources for both training and storage, as they require millions of Gaussians in their point cloud representation for each scene. We present a technique utilizing quantized embeddings to significantly reduce per-point memory storage requirements and a coarse-to-fine training strategy for a faster and more stable optimization of the Gaussian point clouds. Our approach develops a pruning stage which results in scene representations with fewer Gaussians, leading to faster training times and rendering speeds for real-time rendering of high resolution scenes. We reduce storage memory by more than an order of magnitude all while preserving the reconstruction quality. We validate the effectiveness of our approach on a variety of datasets and scenes preserving the visual quality while consuming 10-20x lesser memory and faster training/inference speed. Project page and code is available https://efficientgaussian.github.io