LagrangianSplats: Divergence-Free Transport of Gaussian Primitives for Fluid Reconstruction

TL;DR

This paper presents a novel fluid reconstruction method that enforces divergence-free flow constraints intrinsically, improving accuracy and physical validity in 3D fluid velocity estimation from sparse observations.

Contribution

The authors introduce a divergence-free kernel parameterization and a sliding window optimization scheme for more accurate and physically consistent fluid reconstruction.

Findings

Outperforms state-of-the-art methods in transport consistency.

Ensures flow incompressibility by construction.

Enables high-quality re-simulation and flow analysis.

Abstract

Reconstructing 3D fluid velocity fields from sparse 2D video observations is a highly ill-posed inverse problem, demanding both transport consistency with observed motion and physical validity under fluid laws. Existing methods typically impose these constraints through soft penalties, often leading to compromised accuracy and convergence issues. We introduce a reconstruction framework that structurally enforces both constraints. Specifically, we parameterize the reconstructed velocity using a continuous Divergence-Free Kernel representation, driving the advection of a Lagrangian 3D Gaussian Splatting representation. This formulation intrinsically guarantees both flow incompressibility and long-range transport coherence by construction. To enable the efficient optimization of such a constrained system, we introduce a novel Sliding Window scheme that propagates gradients over meaningful temporal horizons while maintaining tractable training costs. Experiments on synthetic and real-world datasets demonstrate that our method outperforms state-of-the-art baselines in both transport consistency and physical accuracy, enabling applications such as high-quality re-simulation and flow analysis.