The ability to extrapolate dynamic 3D scenes beyond the observed timeframe is fundamental to advancing physical world understanding and predictive modeling. Existing dynamic 3D reconstruction methods have achieved high-fidelity rendering of temporal interpolation, but typically lack physical consistency in predicting the future. To overcome this issue, we propose ParticleGS, a physics-based framework that reformulates dynamic 3D scenes as physically grounded systems. ParticleGS comprises three key components: 1) an encoder that decomposes the scene into static properties and initial dynamic physical fields; 2) an evolver based on Neural Ordinary Differential Equations (Neural ODEs) that learns continuous-time dynamics for motion extrapolation; and 3) a decoder that reconstructs 3D Gaussians from evolved particle states for rendering. Through this design, ParticleGS integrates physical reasoning into dynamic 3D representations, enabling accurate and consistent prediction of the future. Experiments show that ParticleGS achieves state-of-the-art performance in extrapolation while maintaining rendering quality comparable to leading dynamic 3D reconstruction methods.
@article{arxiv.2505.20270,
title = {ParticleGS: Learning Neural Gaussian Particle Dynamics from Videos for Prior-free Physical Motion Extrapolation},
author = {Jinsheng Quan and Qiaowei Miao and Yichao Xu and Zizhuo Lin and Ying Li and Wei Yang and Zhihui Li and Yawei Luo},
journal= {arXiv preprint arXiv:2505.20270},
year = {2025}
}