Reproduce “Reading the Table: Sim-to-Real Game State Detection for the Chef's Hat Card Game” 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
Connor Gäde, Matthias Kerzel, Hauke Bünning, Jan-Gerrit Habekost, Mariana Barros, Stefan Wermter
The authors build a card-and-zone perception pipeline for Chef’s Hat that generates training labels from simulated camera geometry and varies scenes to reduce manual annotation. It detects card designs, segments game zones, and assigns cards to zones. On 100 rectified images from one robot camera, randomized training yields F1 0.906 without real-image fine-tuning; this tests perception, not gameplay.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.