Reproduce “NAS-GS: Noise-Aware Sonar 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
Shida Xu, Jingqi Jiang, Jonatan Scharff Willners, Sen Wang
NAS-GS reconstructs sonar scenes and renders views from unseen positions. It represents structure with 3D Gaussians, computes sonar’s intensity and occlusion directions separately, and models noise with learnable Gaussian mixtures. In three simulated scenes, it outperformed Neusis and ZSplat* on reported image metrics. Offshore examples are qualitative, and the study does not compare directly with SonarSplat.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.