Reproduce “NeuralSRNF: Neural Square Root Normal Fields for the Statistical Shape Analysis and Generation of Nonrigid 3D and 4D Objects” 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
Awais Nizamani, Hamid Laga, Guanjin Wang, Farid Boussaid, Mohammed Bennamoun, Anuj Srivastava
NeuralSRNF learns a conditional inverse from paired SRNFs and surfaces, recovering surfaces along paths computed in SRNF space. On 100 DFAUST test surfaces at 512 × 512 resolution, it reports inversion error 0.043 × 10⁻³ and time 0.60 seconds per surface, versus numerical inversion’s 2.655 × 10⁻³ and 69.10 seconds. It requires registered, genus-zero surfaces and class-specific training.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.