Reproduce “Smooth Flow Matching for Synthesizing Functional Data” 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

Jianbin Tan, Anru R. Zhang

Smooth Flow Matching (SFM) generates trajectories from sparse, irregular observations by learning a smooth pointwise transformation across flow time, observation time, and value. In MIMIC-IV, it modeled arterial blood pressure, respiratory rate, and heart rate separately; prediction using synthetic data performed best among synthetic methods. For arterial blood pressure, training took 30.28 seconds versus 410.90 for DSM; this measures training, not sampling.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.