Models / microsoft / Table Transformer (Structure Recognition)

Table Transformer (Structure Recognition)

JSON →
vision

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

imagevision
homepage ↗docs ↗
Specs
Context & limits
—
context window
—
max output
Lifecycle
Dates
released2021-11-01
Resources
homepagehttps://github.com/microsoft/table-transformer ↗
docshttps://huggingface.co/microsoft/table-transformer-structure-recognition ↗