分层 4D 旋转高斯泼溅：面向长动态场景的压缩表示
Layered 4D-Rotor Gaussian Splatting: A Compressed Representation for Long Dynamic Scenes
针对长动态场景的高斯泼溅重建，提出分层 4D 表示与量化压缩流程，使存储降低 22.3 倍并支持 500+ FPS 实时渲染，可直接用于需要长时序、低显存的三维动态重建任务。
原文依据 we introduce Layered 4D-Rotor Gaussian Splatting (L4DRotorGS), a novel compressed representation designed for long dynamic scenes.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文致力于以存储高效的方式从多视角视频中重建长动态场景。高斯泼溅及其动态场景扩展的最新进展已展现出令人印象深刻的视觉质量，但仍受限于短时长（<10 秒）、大存储体积（>500 MB）以及高 GPU 显存占用。为克服这些局限，我们提出分层 4D 旋转高斯泼溅（L4DRotorGS），一种面向长动态场景的新型压缩表示。该方法将分层 4D 表示、高效训练与有效压缩集成于统一框架。具体而言，4D 高斯首先根据其时间跨度组织为层，然后划分为离散的时间桶。该结构允许仅选择性访问和渲染必要的 4D 高斯子集，从而大幅降低 GPU 显存需求。为进一步压缩表示，我们采用因子化协方差量化、分层压缩和残差码本量化等一系列技术，在保持高视觉保真度的同时实现高达 22.3 倍的压缩比。我们实现了高度优化的 C++/CUDA 框架，用于高效训练、压缩和实时渲染，在 RTX 3090 GPU 上达到超过 500 FPS。大量实验表明，L4DRotorGS 在存储效率、视觉质量和渲染速度方面具有优越性，在真实长动态场景的定量指标和感知质量上均持续优于先前方法。
英文原文摘要 1569 字符
We address the challenge of reconstructing long dynamic scenes from multi-view videos in a storage-efficient manner. Recent advances in Gaussian Splatting and its extensions to dynamic scenes have demonstrated impressive visual quality, but remain limited to short duration (<10 s), large storage size (>500 MB), and high GPU VRAM usage.To overcome these limitations, we introduce Layered 4D-Rotor Gaussian Splatting (L4DRotorGS), a novel compressed representation designed for long dynamic scenes. Our approach integrates a layered 4D representation, efficient training, and effective compression into a unified framework. Specifically, 4D Gaussians are first organized into layers based on their temporal extents and then partitioned into discrete temporal buckets. This structure allows for selective access and rendering of only the necessary subsets of 4D Gaussians, substantially reducing GPU memory requirements.To further compress the representation, we apply a series of techniques, Factorized Covariance Quantization, Layered Compression, and Residual Codebook Quantization, achieving a compression ratio of up to 22.3x while preserving high visual fidelity.We implement a highly optimized C++/CUDA framework for efficient training, compression, and real-time rendering, achieving over 500 FPS on an RTX 3090 GPU. Extensive experiments demonstrate the superior storage efficiency, visual quality, and rendering speed of L4DRotorGS, consistently outperforming prior methods in both quantitative metrics and perceptual quality on real-world long dynamic scenes.