Skip to content

Latest commit

 

History

History
13 lines (7 loc) · 393 Bytes

README.md

File metadata and controls

13 lines (7 loc) · 393 Bytes

label_it

A tool for annotating image datasets, allowing the creation of key-value pairs(json) for each image, with OCR assistanced (PaddleOCR based).

It suitable for creating datasets of Donut/LayoutLMv3(or other e2e models).

USAGE

Left Mouse to zoom/pan, Right Mouse to draw a box and grab text in it, Esc to clean draw box

image

video