uakarsh/TiLT-Implementation

Implementation of the paper: Going Full-TILT Boogie on Document Understanding with Text-Image-Layout Transformer.

22
/ 100
Experimental

This helps automate extracting specific information and answering questions from scanned or PDF documents, even when the layout, images, and text are all important. It takes in document images (like forms, invoices, or research papers) and outputs structured data or answers to questions about the document's content. Anyone who regularly processes and needs to extract data from various types of documents, such as data entry specialists, researchers, or administrative staff, would find this useful.

No commits in the last 6 months.

Use this if you need to accurately pull structured data or answer questions from visually complex documents where text, images, and their arrangement on the page all contribute to understanding.

Not ideal if your primary need is for plain-text analysis without considering visual layout or if you require pre-trained models ready for immediate, high-accuracy deployment without further training.

document-processing information-extraction data-capture form-automation document-qa
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 0 / 25

How are scores calculated?

Stars

18

Forks

Language

Jupyter Notebook

License

MIT

Last pushed

Apr 23, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/uakarsh/TiLT-Implementation"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.