Reproduce “GaussianSpline: Adaptive Bézier Splines for Highly Dynamic 4D Gaussian Splatting” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Tom Yachil, Andrei Sharf
GaussianSpline reconstructs dynamic 3D scenes by learning the timing of a fixed number of motion segments for each Gaussian, placing shorter intervals around rapid change. At K=3, learned spacing improves all three metrics in a five-scene ablation. On the flame scene, it reaches 27.42 dB PSNR versus STG’s 23.96 dB; results vary by scene, and collisions can exceed the fixed segment budget.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.