4D Gaussian Splatting as a Learned Dynamical System

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Asiimwe, Arnold Caleb, Vondrick, Carl
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918259575488512
author Asiimwe, Arnold Caleb
Vondrick, Carl
author_facet Asiimwe, Arnold Caleb
Vondrick, Carl
contents We reinterpret 4D Gaussian Splatting as a continuous-time dynamical system, where scene motion arises from integrating a learned neural dynamical field rather than applying per-frame deformations. This formulation, which we call EvoGS, treats the Gaussian representation as an evolving physical system whose state evolves continuously under a learned motion law. This unlocks capabilities absent in deformation-based approaches:(1) sample-efficient learning from sparse temporal supervision by modeling the underlying motion law; (2) temporal extrapolation enabling forward and backward prediction beyond observed time ranges; and (3) compositional dynamics that allow localized dynamics injection for controllable scene synthesis. Experiments on dynamic scene benchmarks show that EvoGS achieves better motion coherence and temporal consistency compared to deformation-field baselines while maintaining real-time rendering
format Preprint
id arxiv_https___arxiv_org_abs_2512_19648
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle 4D Gaussian Splatting as a Learned Dynamical System
Asiimwe, Arnold Caleb
Vondrick, Carl
Computer Vision and Pattern Recognition
We reinterpret 4D Gaussian Splatting as a continuous-time dynamical system, where scene motion arises from integrating a learned neural dynamical field rather than applying per-frame deformations. This formulation, which we call EvoGS, treats the Gaussian representation as an evolving physical system whose state evolves continuously under a learned motion law. This unlocks capabilities absent in deformation-based approaches:(1) sample-efficient learning from sparse temporal supervision by modeling the underlying motion law; (2) temporal extrapolation enabling forward and backward prediction beyond observed time ranges; and (3) compositional dynamics that allow localized dynamics injection for controllable scene synthesis. Experiments on dynamic scene benchmarks show that EvoGS achieves better motion coherence and temporal consistency compared to deformation-field baselines while maintaining real-time rendering
title 4D Gaussian Splatting as a Learned Dynamical System
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.19648