Reproduce “Fast, Accurate Manifold Denoising by Tunneling Riemannian Optimization” 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
Shiyu Wang, Mariam Avagyan, Yihan Shen, Arnaud Lamy, Tingran Wang, Szabolcs Márka, Zsuzsa Márka, John Wright
The paper presents denoising as finding clean signals on an unknown low-dimensional manifold. It learns landmarks, tangent spaces, and a graph, then traverses it with first-order steps and zero-order tunnels. On 20,000 synthetic gravitational-wave test points, this achieved a better measured multiplication-count versus MSE trade-off than nearest-neighbor search over shared landmarks. Its guarantee assumes an accurate graph, not the online learner.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.