作者位次 #7/11ACL 2026Findings· Findings of ACL
Infinity-Parser: Layout-Aware Reinforcement Learning with High-quality Document Parsing Dataset
Baode Wang · Biao Wu · Weizhen Li · Meng Fang · Zuming Huang · Jun Huang · Yanjie Liang · Haozhe WANG · Ling Chen · Wei Chu · Yuan Qi
摘要
Document parsing from scanned images into structured formats remains a significant challenge due to its complexly intertwined elements such as text paragraphs, figures, formulas, and tables. Existing supervised fine-tuning methods often struggle to generalize across diverse document types, leadin...