Table Transformer (Structure Recognition)

JSON →
microsoft vision
image

A DETR-based model fine-tuned to recognize table structure (rows, columns, cells) in document images.

vision
releasedNov 2021