Reproduce “Image Retrieval with a Bayesian Model of Relevance Feedback” 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
Dorota Glowacka, Yee Whye Teh, John Shawe-Taylor
Beta Experts models image-search preferences as sets of acceptable images rather than one exact target. After a user selects the closest image in a batch, it updates neighborhood beliefs and samples them to choose the next batch. In simulation, with five images per round and one acceptable target, it averaged 26 iterations, versus 117 for Auer–Leung and 422 for PicHunter; results apply to dataset targets.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.