Reproduce “Bayesian Sequential Optimal Experimental Design for Single-Pixel Imaging using Reinforcement Learning” 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
Marcos Antonio Obando, Tianyuan Wang, Tanja Tarvainen, Tristan van Leeuwen, Felix Lucka
The authors train a reinforcement-learning policy to choose binary illumination masks for single-pixel imaging. The policy uses the evolving Bayesian belief and expected information gain as its reward, replacing repeated online mask optimization. In a uniform, noiseless one-active-pixel setting, learned policies approach the analytically optimal bisection strategy; the evidence is limited to controlled simulations, not general image reconstruction.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.