NovaSplat: Model-Free 6D Object Pose Estimation via 3D Gaussian Splatting Onboarding and Refinement
NovaSplat: Model-Free 6D Object Pose Estimation via 3D Gaussian Splatting Onboarding and Refinement
This paper presents NovaSplat, a model-free pipeline for 6D object pose estimation that uses 3,3D Gaussian Splatting to build an object representation from a short smartphone video. The method outperforms all model-based methods on a household-object benchmark and model-free competitors on an industrial benchmark with textureless and specular surfaces. A faster inference variant, NovaSplat-Lite, is also introduced.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- code: not_reported
- paper: not_reported
- code: Reported pending Artifact link Open source S4
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
dustrial benchmark with textureless and specular surfaces, it outperforms all model-free competitors while substantially narrowing the gap to model-based ones. We further introduce NovaSplat-Lite, an inference variant that is $3.1\times$ faster at negligible accuracy cost. Code will be made available at \href{https://github.com/POSE-Lab/novasplat}{https://github.com/POSE-Lab/novasplat}.
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org