Reproduce “GAIDA: Generative AI for Human-in-the-Loop Document Annotation” 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
Islam Mesabah, Ali Elhosary, Priyabanta Sandulu, Darko Obradovic, Sebastian J. Vollmer
GAIDA supports document annotation by using OCR to propose text regions and transcriptions and an LLM to suggest semantic labels, which annotators review and correct before export. On WildReceipt, assisted labeling took 0:26 per image versus 1:39 manually, a 73.7% reduction, with Micro F1 86.98%. This was a single-annotator, fixed-order comparison toward visible ground truth, not independent annotation quality.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.