Make It Flow: Animating Water in 3D Gaussian Splatting Scenes

Key points

• Explicit 3D scene representations such as 3D Gaussian Splatting (3DGS) have recently enabled high-quality, real-time novel view synthesis.1
• Our approach estimates 2D optical flow from reference images or model renderings and lifts it into a consistent 3D Eulerian motion field.1
• We propose a cinemagraph-inspired algorithm that applies this motion field to deform Gaussian primitives, producing temporally seamless looping animations.1
• We evaluate our method on a newly captured dataset of in-the-wild waterfall scenes and demonstrate clear improvements over prior work in dynamic Gaussian splatting and 3D cinemagraph generation in visual quality and computational efficiency.1

Sources (1)

• [1]Make It Flow: Animating Water in 3D Gaussian Splatting Scenes
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 9, 05:08 PM

  We present a method to animate water in static 3DGS scenes, enabling visually plausible motion without requiring dynamic capture or physical simulation.

  Explicit 3D scene representations such as 3D Gaussian Splatting (3DGS) have recently enabled high-quality, real-time novel view synthesis.

Extractive summary: sentences quoted from the sources.