Reproduce “DispFlow-GS: Displacement Flow Supervision with Motion Disentangling for Monocular Deformable 3D 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
Thai Duy Nguyen, Haitian Zhang, Addison Lin Wang
DispFlow-GS improves motion localization in monocular dynamic-scene reconstruction by rendering the scene at a later time from the first camera view and supervising Gaussian displacement splats against estimated scene motion. On NeRF-DS, mean MMAP rises from 0.6497 to 0.9015 and DRC from 0.6660 to 0.7081 versus Deformable-3DGS, while mean PSNR changes about 0.1%; this does not establish downstream robotic gains.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.