Reproduce “An End-to-End Multi-Task Learning Model for Image-based Table Recognition” 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
Nam Tuan Ly, Atsuhiro Takasu
MTL-TabNet turns table images into HTML by predicting table structure, cell locations, and cell text in one model. A shared encoder and decoder feed branches for structure, bounding boxes, and character-level cell recognition, trained jointly on annotated tables. On PubTabNet validation, it reaches 96.67% overall TEDS, below VCGroup + ME’s 96.84%.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.