Gaussian splines improve modeling of complex motion in 4D Gaussian splatting
GaussianSpline: Adaptive Bézier Splines for Highly Dynamic 4D Gaussian Splatting
GraphicsComputer Vision and Pattern Recognition
Summary
Capturing scenes where things move in complicated ways is hard for existing 4D modeling methods. The authors introduce GaussianSpline, which uses flexible curved paths called Bézier splines to better track individual moving elements over time. This lets the model focus more detail on fast-moving parts while keeping simpler parts efficient, resulting in clearer and more accurate videos of dynamic scenes. Their approach also ensures smooth transitions between motion segments, improving overall quality.
What this means in practice
- •For computer vision engineers: Improve dynamic scene reconstruction quality by capturing complex, fast-changing motion in video and 3D models with GaussianSpline trajectories.
- •For game developers: Create more realistic animations by accurately modeling non-uniform motion trajectories of objects and characters in 3D immersive environments.
Authors
Tom Yachil, Andrei Sharf
Abstract
We present GaussianSpline, a dynamic 4D Gaussian Splatting method that models per-Gaussian motion using adaptive, non-uniform piecewise cubic Bézier splines. Unlike approaches based on global deformations, fixed-order polynomials, or uniformly parameterized splines, GaussianSpline jointly learns the Bézier control points and the temporal partition of each Gaussian's trajectory. This adaptive representation allocates greater temporal capacity to highly dynamic regions while maintaining compact trajectories for near-static regions, enabling high-fidelity reconstruction of complex and rapidly changing motion. To jointly achieve globally smooth motion and local adaptivity, we enforce $G^1$ continuity between consecutive Bézier segments using a reflection-based constraint on their control points. This produces smooth, fully differentiable trajectories while preserving the flexibility of non-uniform temporal parameterization. Experiments on the Neural 3D Video benchmark, the Google Immersive dataset, as well as on synthetic scenes demonstrate that GaussianSpline substantially improves reconstruction quality in scenes with highly non-uniform and challenging dynamics, while maintaining competitive performance on general dynamic scenes compared with state-of-the-art 4D Gaussian Splatting methods.