Reproduce “PnP-CF: A Consensus Flow-Matching Solver for General Inverse Problems” 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
Dibyanshu Kumar, Magda Gregorova
PnP-CF extends flow-based plug-and-play reconstruction to nonlinear and learned forward operators. It alternates measurement-consistency and flow-denoising updates, carrying their disagreement forward with a dual variable and adapting a pullback between stochastic exploration and trajectory continuity. On 100 CelebA phase-retrieval images, it reports PSNR 30.33 dB, LPIPS 0.11, and SSIM 0.80; this comparison uses two restarts.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.