Reproduce “From Prototype to Massing: Learning Graph Correspondence for Scalable Facade Modeling” 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
Yiliang Shao
The system routes designer-authored facade Patterns from a resolved prototype onto compatible segments of a larger massing. It represents segments as graphs and uses a shared encoder to retrieve correspondences, then transfers each segment’s authored Pattern. On a synthetic corpus, it reports 96.5% top-1 accuracy versus 76.5% for direct retrieval. This measures within-corpus retrieval, not generalization to unseen design families.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.